microbiology and technology of - wordpress.com · 2009-03-23 · fermented foods,by a.h. rose, was...

Microbiology and Technology of Fermented Foods

FM_BW_Hutkins_277165 4/12/06 8:26 AM Page i

The IFT Press series reflects the mission of the Institute of Food Technologists—advancingthe science and technology of food through the exchange of knowledge. Developed in part-nership with Blackwell Publishing, IFT Press books serve as essential textbooks for academicprograms and as leading edge handbooks for industrial application and reference. Craftedthrough rigorous peer review and meticulous research, IFT Press publications represent thelatest, most significant resources available to food scientists and related agriculture profes-sionals worldwide.

IFT Book Communications Committee

Ruth M. PatrickDennis R. HeldmanTheron W. DownesJoseph H. HotchkissMarianne H. GilletteAlina S. SzczesniakMark BarrettNeil H. MermelsteinKaren Banasiak

IFT Press Editorial Advisory Board

Malcolm C. BourneFergus M. ClydesdaleDietrich KnorrTheodore P. LabuzaThomas J. MontvilleS. Suzanne NielsenMartin R. OkosMichael W. ParizaBarbara J. PetersenDavid S. ReidSam SaguyHerbert StoneKenneth R. Swartzel

FM_BW_Hutkins_277165 4/12/06 8:26 AM Page ii

Microbiology and Technology of Fermented FoodsRobert W. Hutkins

FM_BW_Hutkins_277165 4/12/06 8:26 AM Page iii

Titles in the IFT Press series

• Biofilms in the Food Environment (Hans P. Blaschek, Hua Wang, and Meredith E. Agle)• Food Carbohydrate Chemistry (Ronald E. Wrolstad)• Food Irradiation Research and Technology (Christopher H. Sommers and Xuetong Fan)• High Pressure Processing of Foods (Christopher J. Doona, C. Patrick Dunne, and Florence

E. Feeherry)• Hydrocolloids in Food Processing (Thomas R. Laaman)• Multivariate and Probabilistic Analyses of Sensory Science Problems (Jean-Francois

Meullenet, Hildegarde Heymann, and Rui Xiong)• Nondestructive Testing of Food Quality (Joseph Irudayaraj and Christoph Reh)• Preharvest and Postharvest Food Safety: Contemporary Issues and Future Directions

(Ross C. Beier, Suresh D. Pillai, and Timothy D. Phillips, Editors; Richard L. Ziprin, As-sociate Editor)

• Regulation of Functional Foods and Nutraceuticals: A Global Perspective (Clare M.Hasler)

• Sensory and Consumer Research in Food Product Development (Howard R. Moskowitz,Jacqueline H. Beckley, and Anna V.A. Resurreccion)

• Thermal Processing of Foods: Control and Automation (K.P. Sandeep)• Water Activity in Foods: Fundamentals and Applications (Gustavo V. Barbosa-Canovas,

Anthony J. Fontana Jr., Shelly J. Schmidt, and Theodore P. Labuza)

FM_BW_Hutkins_277165 4/12/06 8:26 AM Page iv

©2006 Blackwell PublishingAll rights reserved

Blackwell Publishing Professional2121 State Avenue, Ames, Iowa 50014, USA

Orders: 1-800-862-6657Office: 1-515-292-0140Fax: 1-515-292-3348Web site: www.blackwellprofessional.com

Blackwell Publishing Ltd9600 Garsington Road, Oxford OX4 2DQ, UKTel.: 144 (0)1865 776868

Blackwell Publishing Asia550 Swanston Street, Carlton, Victoria 3053, AustraliaTel.: 161 (0)3 8359 1011

Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is grantedby Blackwell Publishing, provided that the base fee of $.10 per copy is paid directly to the Copyright Clearance Center,222 Rosewood Drive, Danvers, MA 01923. For those organizations that have been granted a photocopy license by CCC,a separate system of payments has been arranged. The fee codes for users of the Transactional Reporting Service areISBN-13: 978-0-8138-0018-9; ISBN-10: 0-8138-0018-8/2006 $.10.

First edition, 2006

Library of Congress Cataloging-in-Publication Data

Hutkins, Robert W. (Robert Wayne)Microbiology and technology of fermented foods / Robert W. Hutkins.

—1st ed.p. cm.

Includes bibliographical references and index.ISBN-13: 978-0-8138-0018-9 (alk. paper)ISBN-10: 0-8138-0018-8 (alk. paper)1. Fermented foods—Textbooks. 2. Fermented foods—Microbiology—Textbooks. I. Title.

TP371.44.H88 2006664�.024—dc22 2006002149

The last digit is the print number: 9 8 7 6 5 4 3 2 1

FM_BW_Hutkins_277165 4/12/06 8:26 AM Page v

Contents

Preface ix

Acknowledgments xi

1 Introduction 3

2 Microorganisms and Metabolism 15

3 Starter Cultures 67

4 Cultured Dairy Products 107

5 Cheese 145

6 Meat Fermentation 207

7 Fermented Vegetables 233

8 Bread Fermentation 261

9 Beer Fermentation 301

10 Wine Fermentation 349

11 Vinegar Fermentation 397

12 Fermentation of Foods in the Orient 419

Index 457

FM_BW_Hutkins_277165 4/12/06 8:26 AM Page vii

Preface

In organizing this book, I have followedthe basic outline of the course I teach,Microbiology of Fermented Foods. Stu-dents in this course, and hopefully readersof this text, are expected to have had a ba-sic course in microbiology, at minimum, aswell as courses in food microbiology andfood science.An overview of microorgan-isms involved in food fermentations, theirphysiological and metabolic properties,and how they are used as starter cultureprovides a foundation for the succeedingchapters. Nine chapters are devoted to themajor fermented foods produced aroundthe world, for which I have presented bothmicrobiological and technological featuresfor the manufacture of these products. Iconfess that some subjects were consid-ered, but then not included, those beingthe indigenous fermented foods and thenatural fermentations that occur duringprocessing of various “non-fermented”foods, such as cocoa beans and coffeebeans. These topics are thoroughly cov-ered in the above mentioned texts.

One of my goals was to provide a histor-ical context for how the manufacture offermented foods evolved,while at the sametime emphasizing the most current sci-ence. To help accomplish this goal I haveincluded separate entries, called “Boxes,”that describe, in some detail, current topicsthat pertain to the chapter subjects. Someof these boxes are highly technical,whereas others simply provide sidebar in-formation on topics somewhat apart frommicrobiology or fermentation. Hopefully,the reader will find them interesting and apleasant distraction from the normal text.

This project started out innocentlyenough, with the simple goal of providinga resource to students interested in the mi-crobiology of fermented foods. Since1988, when I first developed a course infermentation microbiology at the Univer-sity of Nebraska, there has not been a suit-able student text on this subject that Icould recommend to my students. Peder-son’s Microbiology of Food Fermenta-tions had last been published in 1979 andFermented Foods, by A.H. Rose, was pub-lished in 1982. Brian Wood’s two volumeMicrobiology of Fermented Foods, pub-lished in 1998 (revised from an earlier1985 edition), is an excellent resource andis considered to be one of the most thor-ough texts on fermented foods, but it andother handbooks are generally beyond thescientific scope (and budget) of most stu-dents in a one-semester-long course. Fi-nally, there are many excellent resourcesdevoted to specific fermented foods. Therecently published (2004) Cheese Chem-istry, Physics and Microbiology (edited byFox, McSweeney, Cogan, and Guinee) is anoutstanding reference text, as are Jack-son’s Wine Science Principles, Practice,and Perceptions and Steinkraus’ Industri-alization of Indigenous Fermented Foods.However, their coverage is limited to onlythose particular foods.

I hope this effort achieves the dual pur-poses for which it is intended, namely tobe used as a text book for a college coursein fermentation microbiology and as a gen-eral reference on fermented food microbi-ology for researchers in academia, indus-try, and government.

ix

FM_BW_Hutkins_277165 4/12/06 8:26 AM Page ix

x Preface

Finally, in an effort to make the text eas-ier to read, I made a conscious decision towrite the narrative portion of the bookwith minimal point-by-point referencing.

Each chapter includes a bibliography fromwhich most source materials were ob-tained.The box entries, however, are fullyreferenced.

FM_BW_Hutkins_277165 4/12/06 8:26 AM Page x

Acknowledgments

Shoaf, Jennifer Huebner, Jun Goh, JohnRupnow, and the SMB group.The editorialstaff at Blackwell Press, especially MarkBarrett and Dede Pederson, have been in-credibly patient, for which I am very ap-preciative.

I thank my wife, Charla, and my kids,Anna and Jacob, for being such goodsports during the course of this project.Atleast now you know why I was busier thanusual these past two years.

Finally, I would not be in a position ofwriting an acknowledgment section,muchless this entire text, were it not for mygraduate mentors, Robert Marshall, LarryMcKay, Howard Morris, and Eva Kashket.Role models are hard to find, and I was for-tunate to have had four. My greatest inspi-ration for writing this book, however, hasbeen the many students, past and present,that have made teaching courses and con-ducting research on fermented foods mi-crobiology such a joy and privilege.

I am grateful to the many colleagues whoreviewed chapters and provided me withexcellent suggestions and comments. Anyquestionable or inaccurate statements,however, are due solely to the author (andplease let me know).To each of the follow-ing reviewers, I thank you again: Andy Benson, Larry Beuchat, Lloyd Bullerman,Rich Chapin, Mark Daeschel, Lisa Durso,Joe Frank, Nancy Irelan, Mark Johnson,Jake Knickerbocker, David Mills, DennisRomero, Mary Ellen Sanders, Uwe Sauer,Randy Wehling, and Bart Weimer.

For the generous use of electron micro-graphs, photos, and other written materi-als used in this text, I thank Kristin Ahrens,Andreia Bianchini, Jeff Broadbent, LloydBullerman, Rich Chapin (Empyrean Ales),Lisa Durso, Sylvain Moineau, Raffaele deNigris, John Rupnow,Albane de Vaux, BartWeimer, Jiujiang Yu, and Zhijie Yang.

For their encouragement and supportduring the course of this project, specialthanks are offered to Jim Hruska, Kari

xi

FM_BW_Hutkins_277165 4/12/06 8:26 AM Page xi

Microbiology and Technology of Fermented Foods

FM_BW_Hutkins_277165 4/12/06 8:26 AM Page 1


Copyright © 2006 by Blackwell Publishing

1

Introduction

“When our souls are happy, they talk about food.”Charles Simic, poet

than the raw materials from which they weremade. Despite the “discovery” that fermentedfoods tasted good and were well preserved, itmust have taken many years for humans to fig-ure out how to control or influence conditionsto consistently produce fermented food prod-ucts. It is remarkable that the means for produc-ing so many fermented foods evolved indepen-dently on every continent and on an entirelyempirical basis.Although there must have beencountless failures and disappointments, small“industries,” skilled in the art of making fer-mented foods, would eventually develop. Aslong ago as 3000 to 4000 B.C.E., for example,bread and beer were already being mass pro-duced by Egyptian bakeries and Babylonianbreweries. Likewise, it is clear from the histori-cal record that the rise of civilizations aroundthe Mediterranean and throughout the MiddleEast and Europe coincided with the productionand consumption of wine and other fermentedfood and beverage products (Box 1–1). It isnoteworthy that the fermented foods con-sumed in China, Japan, and the Far East werevastly different from those in the Middle East;yet, it is now apparent that the fermentationalso evolved and became established aroundthe same time.

Fermentation became an even more wide-spread practice during the Roman Empire, as

Fermented Foods and Human History

Fermented foods were very likely among thefirst foods consumed by human beings.This wasnot because early humans had actually plannedon or had intended to make a particular fer-mented food, but rather because fermentationwas simply the inevitable outcome that resultedwhen raw food materials were left in an other-wise unpreserved state.When, for example, sev-eral thousands of years ago, milk was collectedfrom a domesticated cow, goat, or camel, it waseither consumed within a few hours or else itwould sour and curdle, turning into somethingwe might today call buttermilk.A third possibil-ity, that the milk would become spoiled and pu-trid, must have also occurred on many occa-sions. Likewise, the juice of grapes and otherfruits would remain sweet for only a few daysbefore it too would be transformed into a pleas-ant, intoxicating wine-like drink. Undoubtedly,these products provided more than mere suste-nance; they were also probably well enjoyed foraesthetic or organoleptic reasons. Importantly, itmust have been recognized and appreciatedearly on that however imperfect the souredmilk, cheese, wine, and other fermented foodsmay have been (at least compared to modernversions),they all were less perishable and wereusually (but not always) safer to eat and drink

3

01CH_BW_Hutkins_277165 4/12/06 8:03 AM Page 3



4 Microbiology and Technology of Fermented Foods

new raw materials and technologies wereadopted from conquered lands and spreadthroughout the empire. Fermented foods alsowere important for distant armies and navies,due to their increased storage stability. Beerand wine, for example, were often preferredover water (no surprise there), because the lat-ter was often polluted with fecal material orother foreign material. During this era, themeans to conduct trade had developed, andcheese and wine, as well as wheat for bread-making, became available around the Mediter-ranean, Europe, and the British Isles.

Although manufacturing guilds for breadhad existed even during the Egyptian empire,by the Middle Ages, the manufacture of manyfermented foods, including bread, beer, andcheese, had become the province of craftsmenand organized guilds. The guild structure in-volved apprenticeships and training; oncelearned, these skills were often passed on tothe next generation. For some products, partic-ularly beer, these craftsmen were actuallymonks operating out of monasteries andchurches, a tradition that lasted for hundredsof years. Hence, many of the technologies and

Box 1–1. Where and When Did Fermentations Get Started? Answers from Biomolecular Archaeologists

Although the very first fermentations were certainly inadvertent, it is just as certain that humanbeings eventually learned how to intentionally produce fermented foods. When, where, andhow this discovery occurred have been elusive questions, since written records do not exist.However, other forms of archaeological evidence do indeed exist and have made it possible tonot only establish the historical and geographical origins of many of these fermentations, butalso to describe some of the techniques likely used to produce these products.

For the most part, investigations into the origins of food fermentations have focused on alco-holic fermentations, namely wine and beer, and have been led primarily by a research group atthe University of Pennsylvania Museum of Archaeology and Anthropology’s Museum AppliedScience Center for Archaeology (http://masca.museum.upenn.edu). These “biomolecular ar-chaeologists” depend not so much on written or other traditional types of physical evidence(which are mostly absent),but rather on the chemical and molecular “records”obtained from ar-tifacts discovered around the world (McGovern et al., 2004).

Specifically,they have extracted residues still present in the ancient clay pottery jars and vesselsfound in excavated archaeological sites (mainly from the Near East and China).Because these ves-sels are generally porous,any organic material was adsorbed and trapped within the vessel pores.In a dehydrated state, this material was protected against microbial or chemical decomposition.Carbon dating is used to establish the approximate age of these vessels, and then various analyti-cal procedures (including gas chromatography-mass spectroscopy, Fourier transform infraredspectrometry, and other techniques) are used to identify the chemical constituents.

The analyses have revealed the presence of several marker compounds, in particular, tartaricacid, which is present in high concentrations in grapes (but is generally absent elsewhere), andtherefore is ordinarily present in wine, as well (Guash-Jané et al., 2004; McGovern, 2003). Basedon these studies (and others on “grape archaeology”), it would appear that wine had been pro-duced in the Near East regions around present-day Turkey, Egypt, and Iran as long ago as the Neolithic Period (8500 to 4000 B.C.E.).

Recent molecular archaeological analyses have revealed additional findings. In 2004, it was re-ported that another organic marker chemical, syringic acid (which is derived from malvidin, apigment found in red wines), was present in Egyptian pottery vessels.This was not a real sur-prise, because the vessels were labeled as wine jars and even indicated the year, source, andvintner.What made this finding especially interesting, however, was that one of the vessels hadoriginally been discovered in the tomb of King Tutankhamun (King Tut, the “boy king”).Thus,not only does it now appear that King Tut preferred red wine, but that when he died (at aboutage 17), he was, by today’s standards, not even of drinking age.


Introduction 5

manufacturing practices employed even todaywere developed by monks. Eventually, produc-tion of these products became more priva-tized, although often under some form of statecontrol (which allowed for taxation).

From the Neolithic Period to the Middle Agesto the current era, fermented foods have beenamong the most important foods consumed byhumans (Figure 1–1). A good argument can bemade that the popularity of fermented foodsand the subsequent development of technolo-gies for their production directly contributed tothe cultural and social evolution of human his-tory. Consider, after all, how integral fermentedfoods are to the diets and cuisines of nearly allcivilizations or how many fermented foods andbeverages are consumed as part of religious cus-toms, rites, and rituals (Box 1–2).

Fermented Foods: From Art to Science

It is difficult for the twenty-first century readerto imagine that fermented foods, whose manu-facture relies on the intricate and often subtleparticipation of microorganisms, could havebeen produced without even the slightest no-tion that living organisms were actually in-volved. The early manufacturers of fermentedfoods and beverages obviously could not haveappreciated the actual science involved intheir production, since it was only in the last150 to 200 years that microorganisms and en-zymes were “discovered.” In fact, up until themiddle of the nineteenth century, much of thescientific community still believed in the con-cept of spontaneous generation. The very actof fermentation was a subject for philosophers

As noted above, the origins of wine making in the Near East can be reliably traced to about5400 B.C.E.The McGovern Molecular Archaeology Lab group has also ventured to China in aneffort to establish when fermented beverages were first produced and consumed (McGovern etal., 2004).As described in Chapter 12,Asian wines are made using cereal-derived starch ratherthan grapes. Rice is the main cereal used. Other components, particularly honey and herbs,were apparently added in ancient times.

As had been done previously, the investigators analyzed material extracted from Neolithic(ca. 7000 B.C.E.) pottery vessels. In this case, the specific biomarkers would not necessarily bethe same as for wine made from grapes, but rather would be expected to reflect the differentstarting materials. Indeed, the analyses revealed the presence of rice, honey, and herbal con-stituents, but also grapes (tartaric acid). Although domesticated grape vines were not intro-duced into China until about 200 B.C.E., wild grapes could have been added to the wine (as asource of yeast).Another explanation is that the tartaric acid had been derived from other na-tive fruits and flowers.Additional analyses of “proto-historic” (ca. 1900 to 700 B.C.E.) vessels in-dicate that these later wines were cereal-based (using rice and millet). Thus, it now appearsclear that fermented beverage technology in China began around the same time as in the NearEast, and that the very nature of the fermentation evolved over several millennia.

References

Guasch-Jané, M.R., M. Ibern-Gómez, C.Andrés-Lacueva, O. Jáuregui, and R.M. Lamuela-Raventós. 2004. Liq-uid chromatography with mass spectrometry in tandem mode applied for the identification of winemarkers in residues from ancient Egyptian vessels.Anal. Chem. 76:1672–1677.

McGovern, P.E. 2003.Ancient Wine:The Search for the Origins of Viniculture. Princeton University Press.Princeton, New Jersey.

McGovern, P.E., J. Zhang, J.Tang, Z. Zhang, G.R. Hall, R.A. Moreau,A. Nuñez, E.D. Butrym, M.P. Richards, C.-S.Wang, G. Cheng, Z. Zhao, and C.Wang. 2004. Fermented beverages of pre- and proto-historic China.Proc. Nat.Acad. Sci. 101:17593–17598.

Box 1–1. Where and When Did Fermentations Get Started? Answers from Biomolecular Archaeologists (Continued)



and alchemists, not biologists. Although theDutch scientist Antonie van Leeuwenhoek hadobserved microorganisms in his rather crudemicroscope in 1675, the connection betweenLeeuwenhoek’s “animalcules”and their biologi-cal or fermentative activities was only slowlyrealized. It was not until later in the next cen-

tury that scientists began to address the ques-tion of how fermentation occurs.

Initially it was chemists who began to studythe scientific basis for fermentation. In the late1700s and early 1800s, the chemists Lavoisierand Gay-Lussac independently described theoverall equations for the alcoholic fermenta-

About 14 million bushelsof grain per year must be imported into Rome tofeed the inhabitants

Wine grapes reintroduced intoAlsatia, France by Romanemperor Probus (after beingdisplaced for 200 years by wheat)

618

Kumyss, an ethanol-containing fermented mare's milk productis introduced in Asia.

Roquefort cheese "discovered"

1070

Laws regulating the price of breadand the amount of profit bakers can earn are enacted in England

120227716

Neanderthal man becomes Homo sapiens

As glaciers retreat inthe North, wild grainsbegin to grow in theNear East.

8,000

Agriculture begins and becomes thebasis of civilization. Domesticationof animals and cereal grains occurs

Sausage makingintroduced intoRome by Caesar

4811,00038,000

Soybeans introduced into China

1000 054

Julius Caesar reportsCheddar cheese beingmade in Britain

B.C.E.

A brewery, later namedas Löwenbräu in 1552,opens in Bavaria

Sherry, a type of fortified wine, is produced by thewine house, Valdespino, at Jerez de la Frontera, Spain.

1630

Soy sauce introduced in Japan by a company thatwill become Kikkoman

The Gekkeikan Sake Co.begins producing sake inKyoto, Japan. Today, itis the world's largestproducer.

1637

White Mission grapes are introduced intosouthern California

169714301383

The French Benedictine monk, Dom Pierre Pérignon, developstechniques for retaining carbondioxide in wine, leading to thedevelopment of Champagne.

1698

Figure 1–1. Developments in the history of fermented foods. From Trager, J., 1995. The Food Chronology.Henry Holt and Co., New York, and other sources.


Introduction 7

tion. Improvements in microscopy led Kützing,Schwann,and others to observe the presence ofyeast cells in fermenting liquids, including beerand wine. These observations led Schwann topropose in 1837 (as recounted by Barnett,2003) that “it is very probable that, by means ofthe development of the fungus, fermentation is

started.” The suggestion that yeasts were actu-ally responsible for fermentation was not widelyaccepted,however;and instead it was argued byhis contemporaries (namely Berzelius, Liebig,and Wöhler) that fermentation was caused byaerobic chemical reactions and that yeasts wereinert and had nothing to do with fermentative

Gruyère cheese isintroduced in France

1759

Guinness brewery established inDublin, Ireland by Arthur Guinness

1722

Moët and Chandon, now the world'slargest Champagne producer, beginsoperations in Épernay France

Thomas Jefferson, seeking todevelop viniculture in the Virginiacolony, plants grape vines atMonticello; later he will try to growolives. Both efforts are unsuccessful.

Captain Cook, by feeding sauerkraut tohis crewman, is awarded a medal by theRoyal Society for conquering scurvy.

1743 1773 1775

Frederick Accum, a professor at the SurreyInstitution flees to Berlin in order to escapeindustry wrath for publishing "A Treatise onAdulteration of Food and Culinary Poisons".This publication revealed the identities ofthose producers of wine, beer, bread,vinegar, cheese, pickles, and other productsthat were deliberately adulterated.

1816 1820 1822 1850

Port au Salut cheese isinvented by Trappist Monksin Touraine, France

Charles Heidsick produces hisfirst Champagne in Reims, France

Emmenthaler cheeseintroduced to America

2,000 bakeries operate in the U.S., but more than 90%of all bread consumed isbaked at home. Wheat consumption in the U.S. is over 200 pounds per person per year (in 2004, it is less than 150)

Sour dough bread becomesa staple among goldprospectors in California

The first modern chemistrytext book, Traitéélementaire de chime(Elements of Chemistry),by Lavoisier, is published.

90% of Americans are engagedin farming and food production.Today, only about 2% are.

Although reportedly around since the 12thcentury, Camembert cheese is "re-invented".The round little box that permits transport ofthe fragile cheese is invented 100 years later.

17901789 1791

Nicole-Barbe Cliquot invents the“remuage” technique, used to removeyeast sediment from champagne.

1796

Figure 1–1. (Continued)



processes.The debate over the role of microor-ganisms in fermentation was brought to an un-equivocal conclusion by another chemist, LouisPasteur, who wrote in 1857 that “fermentation,far from being a lifeless phenomenon, is a livingprocess” which “correlates with the develop-ment of . . . cells and plants which I have pre-pared and studied in an isolated and pure state”

(Schwartz, 2001). In other words, fermentationcould only occur when microorganisms werepresent.The corollary was also true—that whenfermentation was observed, growth of the mi-croorganisms occurred.

In a series of now famous publications, Pas-teur described details on lactic and ethanolicfermentations, including those relevant to milk

1857 18731864 1877

Brick cheese, amilder version ofLimburger inventedin Wisconsin

The H.J. Heinz Co.introduces WhiteVinegar and AppleCider Vinegar

Louis Pasteur, at age32, shows that bacteriaare responsible for thelactic acid fermentationin milk

Heineken beer,made using acustomized yeaststrain, is producedin Amsterdam

First varietal grape vinesare planted in California,symbolizing the beginningof the California wineindustry

The first pizzeria opens inthe U.S. in New York City

Lister isolates Lactococcuslactis from milk

1880 1895

1907 1916 1950 1964

Bacteriophage against lactic acidbacteria identified by New Zealandresearchers

Kraft introduces processcheese; Velveeta arrivesin 1928, Singles in 1947,and Cheese Whiz in 1953

Russian microbiologistEllie Metchnikoff isolatesLactobacillus from afermented milk product;his findings are publishedin "The Prolongation ofLife"

Yoplait yogurt introducedby SODIAAL, a Frenchdairy cooperative

Dannon Milk Products Inc.introduces yogurt in New York

1942

Miller Brewing Co.Introduces Miller Lite

Plasmids in lactic acidbacteria discovered byLarry Mckay

Saccharomyces cerevisiae genome sequenced

1972 1996 2001

Comparative genomeanalysis of elevenlactic acid and relatedbacteria is published

1974

Lactococcus lactis becomesthe first lactic acid bacteriumto have its genome sequenced

2006

Figure 1–1. (Continued)


Introduction 9

fermentations,beer, and wine.He also identifiedthe organism that causes the acetic acid (i.e.,vinegar) fermentation and that was responsiblefor wine spoilage.The behavior of yeasts duringaerobic and anaerobic growth also led to impor-tant discoveries in microbial physiology (e.g.,the aptly named Pasteur effect, which accountsfor the inhibitory effect of oxygen on glycolyticmetabolism).Ultimately,the recognition that fer-mentation (and spoilage) was caused by mi-croorganisms led Pasteur to begin working onother microbial problems, in particular, infec-

tious diseases. Future studies on fermentationswould be left to other scientists who had em-braced this new field of microbiology.

Once the scientific basis of fermentationwas established, efforts soon began to identifyand cultivate microorganisms capable of per-forming specific fermentations. Breweries suchas the Carlsberg Brewery in Copenhagen andthe Anheuser-Busch brewery in St. Louis wereamong the first to begin using pure yeaststrains, based on the techniques and recom-mendations of Pasteur, Lister, and others. By the

Box 1–2. Fermented foods and the Bible.

The importance of fermented foods and beverages to the cultural history of human societies isevident from many references in early written records. Of course, the Bible (Old and New Tes-taments) and other religious tracts are replete with such references (see below). Fermentedfoods, however, also serve a major role in ancient Eastern and Western mythologies.

The writers ,who had no scientific explanation for the unique sensory and often intoxicatingproperties of fermented foods,described them as “gifts of the gods.” In Greek mythology, for ex-ample, Dionysus was the god of wine (Bacchus, according to Roman mythology).The Iliad andthe Odyssey, classic poems written by the Greek poet Homer in about 1150 B.C.E., also containnumerous references to wine, cheese, and bread. Korean and Japanese mythology also refers tothe gods that provided miso and other Asian fermented foods (Chapter 12).

Fermented foods and the Bible

From the Genesis story of Eve and the apple, to the dietary laws described in the books of Exodusand Leviticus, food serves a major metaphoric and thematic role throughout the Old Testament.Fermented foods, in particular,are frequently mentioned in biblical passages, indicating that thesefoods must have already been well known to those cultures and civilizations that lived during thetime at which the bible was written.

In Genesis (9:20), for example, one of the first actions taken by Noah after the flood watershad receded was to plant a vineyard. In the very next line, it is revealed that Noah drank enoughwine to become drunk (and naked), leading to the first, but certainly not last, episode in whichdrunkenness and nakedness occur. Later in Genesis (18:8), Abraham receives three strangers(presumably angels), to whom he offers various refreshments, including “curds.”

Perhaps the most relevant reference to fermentation in the Bible is the Passover story.As de-scribed in Exodus (12:39), once Moses had secured the freedom of the Hebrew slaves, theywere “thrust out of Egypt, and could not tarry.”Thus, the dough could not rise or become leav-ened, and was baked instead in its “unleavened” state.This product, called matzoh, is still eatentoday by people of the Jewish faith to symbolically commemorate the Hebrew exodus.

Ritual consumption of other fermented foods is also prescribed in Judaism.Every Sabbath, forexample, the egg bread, Challah, is to be eaten, and grapes or wine is to be drunk, preceded byappropriate blessings of praise.

Fermented foods are also featured prominently in the New Testament.At the wedding in Cana( John 2:1–11), Jesus’ first miracle is to turn water into wine. Later ( John 6:1–14), another mira-cle is performed when five loaves of bread (and two fish) are able to feed 5,000 men.The Sacra-ment of Holy Communion (described by Jesus during the Last Supper) is represented by breadand wine.



early 1900s, cultures for butter and other dairyproducts had also become available.The dairyindustry was soon to become the largest userof commercial cultures, and many specializedculture supply “houses” began selling not onlycultures, but also enzymes, colors, and otherproducts necessary for the manufacture ofcheese and cultured milk products (Chapter 3).Although many cheese factories continued topropagate their own cultures throughout thefirst half of the century,as factory size and prod-uct throughput increased, the use of dairystarter cultures eventually became common-place. Likewise, cultures for bread, wine, beer,and fermented meats have also become thenorm for industries producing those products.

The Modern Fermented Foods Industry

The fermented foods industry, like all other seg-ments of the food processing industry, haschanged dramatically in the past fifty years.Cer-tainly, the average size of a typical productionfacility has increased several-fold,as has the rateat which raw materials are converted to fin-ished product (i.e., throughput). Althoughsmall, traditional-style facilities still exist, as isevident by the many microbreweries, smallwineries, and artisanal-style bakery and cheesemanufacturing operations, the fermented foodsindustry is dominated by producers with largeproduction capacity.

Not only has the size of the industrychanged, but so has the fundamental manner

in which fermented foods are produced (Table1–1). For example, up until the past forty or soyears, most cheese manufacturers used raw,manufacturing (or Grade B) milk, whereas pas-teurized Grade A milk, meeting higher qualitystandards, is now more commonly used. Manu-facturing tanks or vats are now usually en-closed and are constructed from stainless steelor other materials that facilitate cleaning andeven sterilization treatments. In fact, modernfacilities are designed from the outset with anemphasis on sanitation requirements, so thatexposure to air-borne microorganisms andcross-contamination is minimized.

Many of the unit operations are mechanizedand automated, and, other than requiring a fewkeystrokes from a control panel, the manufac-ture of fermented foods involves minimal hu-man contact. Fermented food production isnow,more than ever before, subject to time andscheduling demands. In the so-called “old days,”if the fermentation was slow or sluggish, it sim-ply meant that the workers (who were probablyfamily members) would be late for supper, andlittle else. In a modern production operation, aslow fermentation may mean that the workershave to stay beyond their shift (requiring thatthey be paid overtime), and in many cases, itcould also affect the entire production sched-ule, since the production vat could not beturned over and refilled as quickly as needed.Al-though traditional manufacturing practices maynot have always yielded consistent products, lotsizes were small and economic losses due to anoccasional misstep were not likely to be too se-

Table 1.1. Fermented foods industry: past and present

Traditional Modern

Small scale (craft industry) Large scale (in factories)Non-sterile medium Pasteurized or heat-treated mediumSeptic AsepticOpen ContainedManual AutomatedInsensitive to time Time-sensitiveSignificant exposure to contaminants Minimal exposure to contaminantsVarying quality Consistent qualitySafety a minor concern Safety a major concern


Introduction 11

rious. Besides, for every inferior cask of wine orwheel of cheese, there may have been anequally superior lot that compensated for theone that turned out badly. Even the absoluteworst case scenario—a food poisoning out-break as a result of an improperly manufacturedproduct—would have been limited in scopedue to the small production volume and narrowdistribution range.

Such an attitude, today, however, is simplybeyond consideration. A day’s worth of prod-uct may well be worth tens, if not hundreds ofthousands of dollars, and there is no way a pro-ducer could tolerate such losses,even on a spo-radic basis. Food safety, in particular, has be-come an international priority, and there isgenerally zero tolerance for pathogens or otherhazards in fermented foods. Quality assuranceprograms now exist throughout the industry,which strive to produce safe and consistentproducts. In essence, the fermented foods in-dustry has evolved from a mostly art- or craft-based practice to one that relies on modernscience and technology. Obviously, the issuesdiscussed above—safety, sanitation, quality, andconsistency—apply to all processed foods, andnot just fermented foods. However, the fer-mented foods industry is unique in one majorrespect—it is the only food processing indus-try in which product success depends on thegrowth and activity of microorganisms.The im-plications of this are highly significant.

Microorganisms used to initiate fermenta-tions are, unlike other “ingredients,” not easilystandardized, since their biochemical activityand even their concentration (number of cellsper unit volume) may fluctuate from lot to lot.Although custom-made starter cultures that areindeed standardized for cell number and activ-ity are readily available, many industrial fer-mentations still rely, by necessity, on the pres-ence of naturally-occurring microflora, whosecomposition and biological activities are oftensubject to considerable variation. In addition,microorganisms are often exposed to a varietyof inhibitory chemical and biological agents inthe food or environment that can compromisetheir viability and activity. Finally, the culture

organisms are often the main means by whichspoilage and pathogenic microorganisms arecontrolled in fermented foods. If they fail toperform in an effective and timely manner,the finished product will then be subject tospoilage or worse. Thus, the challenge con-fronting the fermented foods industry is tomanufacture products whose very productionis subject to inherent variability yet satisfy themodern era demands of consistency, quality,line-speed, and safety.

Properties of Fermented Foods

As noted in the previous discussion, fermentedfoods were among the first “processed” foodsproduced and consumed by humans. Theirpopularity more than 5,000 years ago was dueto many of the same reasons why they con-tinue to be popular today (Table 1–2).

Preservation

The preservation aspect of fermented foodswas obviously important thousands of yearsago, when few other preservation techniquesexisted. A raw food material such as milk ormeat had to be eaten immediately or it wouldsoon spoil.Although salting or smoking couldbe used for some products, fermentation musthave been an attractive alternative, due toother desirable features. Preservation was un-doubtedly one of the main reasons why fer-mented foods became such an integral part ofhuman diet.However,even today,preservation,or to use modern parlance, shelf-life (or ex-tended shelf-life), is still an important feature of fermented foods. For example, specializedcultures that contain organisms that produce

Table 1.2. Properties of fermented foods

Enhanced preservationEnhanced nutritional valueEnhanced functionalityEnhanced organoleptic propertiesUniquenessIncreased economic value



specific antimicrobial agents in the food arenow available, providing an extra margin ofsafety and longer shelf-life in those foods.

Nutrition

The nutritional value of fermented foods haslong been recognized, even though the scien-tific bases for many of the nutritional claimshave only recently been studied. Strong evi-dence that fermentation enhances nutritionalvalue now exists for several fermented prod-ucts, especially yogurt and wine. Fluid milk isnot regularly consumed in most of the worldbecause most people are unable to producethe enzyme �-galactosidase, which is neces-sary for digestion of lactose, the sugar foundnaturally in milk. Individuals deficient in �-galactosidase production are said to be lactoseintolerant, and when they consume milk, mild-to-severe intestinal distress may occur. Thiscondition is most common among Asian andAfrican populations, although many adult Cau-casians may also be lactose intolerant.

Many studies have revealed, however, thatlactose-intolerant subjects can consume yogurtwithout any untoward symptoms and cantherefore obtain the nutritional benefits (e.g.,calcium, high quality protein, and B vitamins)contained in milk. In addition, it has been sug-gested that there may be health benefits of yo-gurt consumption that extend beyond thesemacronutrients. Specifically, the microorgan-isms that perform the actual yogurt fermenta-tion, or that are added as dietary adjuncts, arenow thought to contribute to gastrointestinalhealth, and perhaps even broader overall well-being (Chapter 4).

Similarly, there is now compelling evidencethat wine also contains components that con-tribute to enhanced health (Chapter 10). Spe-cific chemicals, including several differenttypes of phenolic compounds, have been iden-tified and shown to have anti-oxidant activitiesthat may reduce the risk of heart disease andcancer.That wine (and other fermented foods)are widely consumed in Mediterranean coun-tries where mortality rates are low has led to

the suggestion that a “Mediterranean diet” maybe good for human health.

Functionality

Most fermented foods are quite different, interms of their functionality, from the raw, start-ing materials. Cheese, for example, is obviouslyfunctionally different from milk.However, func-tional enhancement is perhaps nowhere moreevident than in bread and beer.When humansfirst collected wheat flour some 10,000 yearsago, there was little they could do with it, otherthan to make simple flat breads. However, oncepeople learned how to achieve a leaveneddough via fermentation, the functionality ofwheat flour became limitless. Likewise, barleywas another grain that was widespread and haduse in breadmaking,but which also had limitedfunctionality prior to the advent of fermenta-tion. Given that barley is the main ingredient(other than water) in beer manufacture, couldthere be a better example of enhanced func-tionality due to fermentation?

Organoleptic

Simply stated, fermented foods taste dramati-cally different than the starting materials. Indi-viduals that do not particularly care for Lim-burger cheese or fermented fish sauce mightargue that those differences are for the worse,but there is little argument that fermentedfoods have aroma, flavor, and appearance at-tributes that are quite unlike the raw materialsfrom which they were made.And for those in-dividuals who partake of and appreciate Lim-burger cheese, the sensory characteristics be-tween the cheese and the milk make all thedifference in the world.

Uniqueness

With few exceptions (see below), there is noway to make fermented foods without fermen-tation. Beer, wine, aged cheese, salami, andsauerkraut simply cannot be produced anyother way. For many fermented products, the


Introduction 13

very procedures used for their manufacture areunique and require strict adherence. For exam-ple, Parmesan cheese must be made in a de-fined region of Italy, according to traditionaland established procedures, and then aged un-der specified conditions. Any deviation resultsin forfeiture of the name Parmesan. For those“fermented” foods made without fermentation(which includes certain fresh cheeses, sau-sages, and even soy sauce), their manufacturegenerally involves direct addition of acids and/or enzymes to simulate the activities normallyperformed by fermentative microorganisms.These products (which the purist might be in-clined to dismiss from further discussion) lackthe flavor and overall organoleptic propertiesof their traditional fermented counterparts.

Economic value

Fermented foods were the original members ofthe value-added category. Milk is milk, but addsome culture and manipulate the mixture justright, age it for a time, and the result may be afine cheese that fetches a price well above thecombined costs of the raw materials, labor, andother expenses.Grapes are grapes,but if grown,harvested, and crushed in a particular environ-ment and at under precise conditions, and thejuice is allowed to ferment and mature in an op-timized manner, some professor may well payup to $6 or $7 (or more!) for a bottle of the fin-ished product.Truly, the economic value of fer-mented foods, especially fermented grapes, canreach extraordinary heights (apart from the pro-fessor market). As noted in Chapter 10, somewines have been sold for more than $20,000per bottle. Even some specialty vinegars (Chap-ter 11) sell for more than $1,000 per liter. Itshould be noted that not all fermented foodscommand such a high dollar value. In truth, thefermented foods market is just as competitiveand manufacturers are under the same marketpressures as other segments of the food indus-try.Fermented foods are generally made from in-expensive commodities (e.g.,wheat,milk,meat,etc.) and most products have very modest profitmargins (some products,such as “current”or un-

aged cheese, are sold on commodity markets,with very tight margins).There is a well-knownjoke about the wine business that applies toother products as well, and that summarizes thechallenge in making fermented foods:“How doyou make a million dollars in the wine business?Easy, first you start with two million dollars.” Fi-nally, on a industry-wide basis, fermented foodsmay have a significant economic impact on a re-gion,state or country. In California, for example,the wine industry was reported to contributemore than $40 billion to the economy in 2004(according to a Wine Institute report; www.wineinstitute.org).A similar analysis of the U.S.beer industry (www.beerinstitute.org) reportedan overall annual impact of more than $140 bil-lion to the U.S.economy.

Fermented Foods in the Twenty-first Century

For 10,000 years, humans have consumed fer-mented foods. As noted above, originally andthroughout human history, fermentation pro-vided a means for producing safe and well-preserved foods. Even today, fermented foodsare still among the most popular type of foodconsumed. No wonder that about one-third ofall foods consumed are fermented.In the UnitedStates, beer is the most widely consumed fer-mented food product, followed by bread,cheese,wine,and yogurt (Table 1–3).Global sta-tistics are not available, but it can be estimatedthat alcoholic products head the list of mostpopular fermented foods in most of the world.In Asia, soy sauce production and consumptionranks at or near the top.Collectively, sales of fer-mented foods on a global basis exceeds a trilliondollars, with an even greater overall economicimpact.

Although fermented foods have been part ofthe human diet for thousands of years, as theworld becomes more multicultural and cuisinesand cultures continue to mix, it is likely that fer-mented foods will assume an even more impor-tant dietary and nutritional role. Foods such askimchi (from Korea), miso (from Japan), and ke-fir (from Eastern Europe) are fast becoming part



of the Western cuisine. Certainly, the desirableflavor and sensory attributes of traditional, aswell as new-generation fermented foods, willdrive much of the interest in these foods.

Consumption of these products also willlikely be increased as the potential beneficial ef-fects of fermented foods on human health be-come better established, scientifically and clini-cally.As noted above, compelling evidence nowexists to indicate that red wine may reduce therisk of heart disease and that live bacteria pre-sent in cultured milk products may positively in-fluence gastrointestinal health.Armed now withextensive genetic information on the micro-organisms involved in food fermentations thathas only become available in the last century,it is now possible for researchers to custom-produce fermented foods with not only specific

flavor and other functional characteristics, butthat also impart nutritional properties that ben-efit consumers.

ReferencesBarnett, J.A. 2003. Beginnings of microbiology and

biochemistry: the contribution of yeast research.Microbiol. 149:557–567.

Bulloch,W.1960.The History of Bacteriology.OxfordUniversity Press, London.

Cantrell, P.A. 2000. Beer and ale. In K.F. Kiple, K.C.Ornelas (ed). Cambridge World History of Food,p. 619–625. Cambridge University Press, Cam-bridge, United Kingdom

Steinkraus, K.H. 2002. Fermentations in world foodprocessing. Comp. Rev. Food Sci. Technol. 1:23–32.

Table 1.3. U.S. production and consumption of selected fermented foodsa

Food Production Consumptionb

Wine 2.3 � 109 L 9 LBeer 23 � 109 L 100 LCheese 4 � 109 Kg 14 KgYogurt 1.2 � 109 Kg 3.4 KgFermented meats nac 0.3 KgBread 7 � 109 Kg 25 KgFermented vegetables 0.8 � 109 Kg 2.8 Kg

aSources: 2001–2004 data from USDA,WHO, and industry organizationsbPer person, per yearcNot available


2

Microorganisms and Metabolism

“We can readily see that fermentations occupy a special place in the series of chem-ical and physical phenomena. What gives to fermentations certain exceptionalcharacters, of which we are only now beginning to suspect the causes, is the modeof life in the minute plants designated under the generic name of ferments, a modeof life which is essentially different from that of other vegetables, and from whichresult phenomena equally exceptional throughout the whole range of the chem-istry of living beings.”

From The Physiological Theory of Fermentation by Louis Pasteur, 1879

popular in Indonesia, is made by inoculatingsoybeans with the fungal organism Rhizopusoligosporus. The manufacturing process lendsitself, however, to chance contamination withother microorganisms, including bacteria thatsynthesize Vitamin B12, making tempeh a goodsource of a nutrient that might otherwise be ab-sent in the diet of individuals who consume thisproduct.

A Primer on Microbial Classification

For many readers, keeping track of the manygenus, species, and subspecies names assignedto the organisms used in fermented foods can bea challenging task. However, knowing which or-ganisms are used in specific fermented foods israther essential (to put it mildly) to understand-ing the metabolic basis for how microbial fer-mentations occur.Therefore, prior to describingthe different groups of microorganisms involvedin food fermentations, it is first necessary to re-view the very nature of microbial taxonomy(also referred to as systematics) and how micro-biologists go about classifying,naming,and iden-tifying microorganisms.

Although this might seem to be a thanklesstask, it is, after all, part of human nature to sortor order things; hence, the goal of taxonomy is

When one considers the wide variety of fer-mented food products consumed around theworld, it is not surprising that their manufac-ture requires a diverse array of microorganisms.Although lactic acid-producing bacteria andethanol-producing yeasts are certainly the mostfrequently used organisms in fermented foods,there are many other bacteria, yeast, and fungithat contribute essential flavor, texture, appear-ance, and other functional properties to thefinished foods. In most cases, more than oneorganism or group of organisms is involved inthe fermentation.

For example, in the manufacture of Swiss-type cheeses, thermophilic lactic acid bacteriafrom two different genera are required to fer-ment lactose, produce lactic acid, and acidifythe cheese to pH 5.2, a task that takes abouteighteen hours. Weeks later, another organism,Propionibacterium freudenreichii subsp. sher-mani, begins to grow in the cheese, producingother organic acids,along with the carbon diox-ide that eventually forms the holes or eyes thatare characteristic of Swiss cheese.

Even for those fermented foods in whichonly a single organism is responsible for per-forming the fermentation, other organisms maystill play inadvertent but important supportingroles.Thus, tempeh, a fermented food product

15





to achieve some sense of order among the mi-crobial world. Specifically, taxonomy providesa logical basis for: (1) classifying or arrangingorganisms into related groups or taxa; (2) es-tablishing rules of nomenclature so that thoseorganisms can be assigned names on a rationalbasis; (3) identifying organisms based on theaccepted classification scheme and nomencla-ture rules; and (4) understanding evolutionaryrelationships of species, one to another.

As will be noted later in this and successivechapters, rules for classification are not per-manently fixed, but rather can be amendedand re-defined in response to new, more dis-criminating methods. For the most part, thesenew classification methods are based on mo-lecular composition and genetic properties,which can also be used to determine phyloge-netic or evolutionary relationships betweenrelated organisms.

The three domains of life

According to modern taxonomy, life on thisplanet can be grouped into three branches ordomains—the Eukarya, the Bacteria, and theArchaea (Figure 2–1).This organization for clas-sifying all living organisms was proposed andestablished in the 1980s by Carl Woese and isbased on the relatedness of specific 16S rRNAsequences using a technique called oligonu-cleotide cataloging.This three-branch tree of lifedisplaced the classical taxonomy that had rec-ognized only two groups, eukaryotes and pro-karyotes, and that was based primarily on mor-phology and biochemical attributes. All of themicroorganisms relevant to fermented foods(and food microbiology, in general) belong to ei-ther the Eukarya or Bacteria. The Archaea,while interesting for a number of reasons, con-sists of organisms that generally live and grow in

Figure 2–1. Phylogenetic tree of life (based on 16S rRNA sequences). Courtesy of the Joint Genome Insti-tute (U.S. Department of Energy).


Microorganisms and Metabolism 17

extreme environments (e.g., very high tempera-ture, very low pH, very high salt), but rarely arethey associated with foods.

Classification of organisms as eukaryoticor prokaryotic is based on a variety of char-acteristics (Bergey’s Manual lists more than50 cytological, chemical, metabolic, molecu-lar, and reproductive properties), but the traditional distinguishing feature is the pres-ence of a nuclear membrane in eukaryotes.Included within the Eukarya domain are ani-mals, plants, protists, and fungi.The latter arerepresented by the Kingdom Fungi, which in-cludes both yeasts and molds.

Fungi

The Kingdom Fungi is very large, containingmany species, and although the exact numberis unknown, it is thought to possibly number asmany as 1.5 million. Only about 10% of thesehave been observed and described. In the cur-rent system of taxonomy (2004), four majorgroups or phyla are recognized among the truefungi or Eumycota.These are Chytridiomycota,Zygomycota,Ascomycota, and Basidiomycota.

The Chytridiomycota are primitive fungiand may be a link to ancestral fungi.They areconsidered true fungi based upon metabolicpatterns, the presence of chitin in their cellwalls, and small subunit rDNA sequence com-parisons with other fungi.The chytrids, as theyare sometimes called, live in wet soil, fresh wa-ter and marine environments, and are not com-mon in foods,nor are they involved in food fer-mentations. They may be unicellular or mayform simple branching chains of cells or primi-tive hyphae, which are coenocytic, that is,without cross walls.

The Zygomycota group, sometimes calledzygomycetes, also produces coenocytic hyphaeand reproduce sexually by fusion of two cellsknown as gametangia that form a zygospo-rangium containing zygospores.This group alsoproduces asexual sporangiospores, stolons, andrhizoids. Sporangiospores are produced in asac-like structure on an aerial stalk that extendsinto the air from the point of attachment to the

substrate by the rhizoid. The rhizoids are spe-cialized root-like hyphae that anchor the moldto its substrate,secrete enzymes,and absorb nu-trients.The stolons are rapidly growing hyphaethat run over the surface of the substrate fromone rhizoid to another. The Zygomycota maycontain as many as 900 species, but those im-portant in foods and food fermentations arefound primarily in the Rhizopus and Mucorgenera.

The Ascomycota is a very large phylum offungi, containing as many as half of all fungalspecies. The ascomycetes (members of the Ascomycota) have septate mycelia, with crosswalls or septa, and produce sexual spores,called ascospores within a structure known asan ascus (plural asci). The ascomycetes alsoproduce asexual reproductive states,where re-production occurs by production of asexualspores known as conidia, which are sporesborne on the end of special aerial fertile hy-phae (conidiophores). Conidia are producedfree, often in chains and not in any enclosingstructure.The asexual state or stage is referredto as the anamorphic state and the fungus iscalled an anamorph. This state is sometimescalled the “imperfect” stage or state of the fun-gus.The sexual state (ascogenous) is referredto as the teleomorphic state, and the as-cospore-producing fungus as a teleomorph.The sexual stage is also referred as the “per-fect” stage or state of the fungus.

Another group of fungi that are related tothe ascomycetes, but which are not given phy-lum status, are the deuteromycetes, sometimesreferred to as Deuteromycota. This is a groupof fungi that are similar to members of the as-comycetes,but which have no observed sexualstage in their life cycle. They are anamorphsand are sometimes (in older literature) referredas the “imperfects”or Fungi Imperfecti.

The zygomycetes, ascomycetes, and deu-teromycetes are fungi that are also calledmolds, or micro fungi, because of their smallsize and because they include the fungalspecies important in foods and food fermenta-tions. One group of ascomycetes, also knownas hemiascomycetes because they do not



produce a separate ascoma, are primarily sin-gle celled organisms known as yeasts. Yeastsmay produce ascospores within the yeast cell,but they also reproduce asexually by a processknown as budding, and are very important infoods and food fermentations.

The Basidiomycota includes certain plantpathogens, such as rusts and smuts, and macrofungi such as mushrooms,puffballs, and bracketfungi that grow on trees and fallen logs, as wellas ectomycorrhizal fungi that are associatedwith certain plant roots.Another group of fungithat form the most common type of mycor-rhizal associations with plant roots are the Glo-males, sometimes referred to as a fifth phylum,the Glomeromycota.

Bacteria

Within the Bacteria domain, there has beenless consensus among taxonomists on howto organize bacteria into higher taxa (e.g.,Kingdom, Sub-Kingdom, etc.).The Ninth Edi-tion of Bergey’s Manual of DeterminativeBacteriology (published in 1994) describedthree major categories of bacteria and one ofarchaeobacteria, and then further divided thebacteria categories into thirty different de-scriptive groups. According to the most cur-rent taxonomy (Garrity, et al., 2004), the Bacteria are now divided into twenty-fourdifferent phyla. Nearly all of the bacteria im-portant in food fermentations, including lac-tic acid bacteria, belong to a single phylum,the firmicutes. Beyond the phyla, bacteria canbe further divided into classes (and sub-classes), orders (and sub-orders), families, andgenera. Other details on their taxonomic po-sitions will be described below.

Nomenclature

Like all living organisms, microorganisms arenamed according to Latinized binomial no-menclature, meaning they are assigned twonames, a genus and a species. By convention,both the genus and species names are itali-

cized, but only the genus is capitalized.Thus,the name of the common food yeast is writ-ten as Saccharomyces cerevisae. For some or-ganisms, a trinomial system is applied to indi-cate a subspecies epithet, as is the case forthe dairy organism Lactococcus lactis subsp.lactis.

Microorganisms are named according to therules established by the appropriate governingbody. For bacteria and fungi, the InternationalCommittee on Systematic Bacteriology and theInternational Association for Plant Taxonomy,respectively, define the nomenclature rules.These rules are then published in the respec-tive “codes,” the International Code of Nomen-clature of Bacteria and the International Codefor Botanical Nomenclature. In addition, thereis a “running list,” called the List of BacterialNames with Standing in Nomenclature (www.bacterio.cict.fr), that provides updated bacter-ial nomenclature.To be included in these listsand considered as “valid,” a name must havebeen published in the scientific literature,alongwith a detailed description and relevant sup-porting data.

In some cases, a validly named organismmay be referred to by another name, indicatedas a synonym. In other instances, the name ofan organism may have been replaced by a newname, in which case the original name is indi-cated as a basonym.In situations where a namewas “unofficially” assigned to an organism anddoes not appear on the list, that name is con-sidered to be illegitimate and its use should bediscontinued. However, even if the name as-signed to a particular organism is supported bya valid publication, this does not mean that thename is or must be accepted by the scientificcommunity.

Although names are indeed based on offi-cial rules (where each taxon has a valid name),the utility of a given classification scheme, onwhich a given organism is named, is left up tothe scientists who use it.That is, a researchermay propose that a given organism be assigneda “new” name, and have the supporting evi-dence published in a valid journal, but othermicrobiologists are entitled to disagree with



the taxonomy and reject the proposed classifi-cation.

It is relevant to raise these issues, becausemany of the organisms used in food fermenta-tions have either undergone nomenclature re-visions or have been reclassified into new taxa.For example, the official name of the dairy or-ganism mentioned above, L. lactis subsp. lactiswas originally Bacterium lactis (the first or-ganism isolated in pure culture and named byJoseph Lister in 1873). It was renamed Strepto-coccus lactis in 1909, which is how this organ-ism was known for more than 70 years (andwhich still shows up occasionally in “current”texts), before the new genus, Lactococcus, wasadopted. In other cases, a name was proposed,then rescinded (see below).There are also in-stances of organisms that had been assigned“unofficial” names (see above), and throughfrequent use, had acquired some level of valid-ity, however undeserved. A good example ofthe latter situation was for Lactobacillus sporo-genes, an organism that is properly classified asBacillus subtilis.Yeast nomenclature, althoughunder the authority of the International Codefor Botanical Nomenclature, is also subject totaxonomical challenges and changes in classifi-cation (see below).

Microbial taxonomy and methods of analysis

If microbiology began with Pasteur in the mid-dle of the nineteenth century, then for the next120 years, microbial classification was basedprimarily on phenotypic characteristics. Al-though many of these traits remain useful as di-agnostic tools, by far, the most powerful meansof classifying microorganisms is now via mo-lecular techniques.

Originally, routine tests were based on nu-cleic acid composition (mol% G � C) andDNA-DNA and DNA-RNA hybridization.The lat-ter has long been considered the gold standardfor defining a species, in that organisms shar-ing high DNA homology (usually greater than70%) are regarded as members of the samespecies. However, more recently, the nucleic

acid sequence of the 16S rRNA region has be-come the most common way to distinguish between organisms, to show relatedness, andultimately to classify an organism to the genusor species level.

These sequences can be easily determinedfrom the corresponding DNA after PCR ampli-fication. By comparing the sequence from agiven organism to those obtained from refer-ence organisms in the 16S rRNA database, it ispossible to obtain highly significant matches orto infer relationships if the sequence is unique.In addition, DNA probes, based on genus- orspecies-specific 16S rRNA sequences from aparticular reference organism, are also usefulfor identifying strains in mixed populations.

Other techniques that essentially provide amolecular fingerprint and that can be used todistinguish between strains of the same speciesinclude restriction fragment length polymor-phism (RFLP), randomly amplified polymorphicDNA (RAPD), pulsed field gel electrophoresis(PFGE),and DNA microarray technology.Finally,it should be noted that nucleic acid-based meth-ods for classification, whether based on 16SrRNA sequences, DNA-DNA hybridization val-ues, or even whole genome analyses, are notnecessarily the be-all, end-all of microbial taxon-omy.Rather,a more holistic or integrative taxon-omy, referred to as polyphasic taxonomy, hasbeen widely adopted and used to classify manydifferent groups of bacteria. Polyphasic taxon-omy is based on genotypic as well as pheno-typic and phylogenetic information and is con-sidered to represent a “consensus approach” forbacterial taxonomy.

Bacteria Used in the Manufacture of Fermented Foods

Despite the diversity of bacteria involved di-rectly or indirectly in the manufacture of fer-mented foods, all are currently classified in oneof three phyla, the Proteobacteria, Firmicutes,and the Actinobacteria.Within the Firmicutesare the lactic acid bacteria, a cluster of Gram-positive bacteria that are the main organismsused in the manufacture of fermented foods.



This phylum also includes the genera Bacillusand Brevibacterium that contain species usedin the manufacture of just a few selected fer-mented foods.

The Proteobacteria contains Gram negativebacteria that are involved in the vinegar fermen-tation, as well as in spoilage of wine and otheralcoholic products. The Actinobacteria con-tains only a few genera relevant to fermentedfoods manufacture, and only in a rather indirectmanner. These include Bifidobacteriim, Kocu-ria, Staphylococcus, and Micrococcus. In fact,Bifidobacterium do not actually serve a func-tional role in fermented foods; rather they areadded for nutritional purposes (see below).

While species of Kocuria and the Staphylo-coccus/Micrococcus group are used in fer-mented foods, they are used for only one prod-uct, fermented meats,and for only one purpose,to impart the desired flavor and color.It is worthemphasizing that fermented foods may containmany other microorganisms, whose presenceoccurs as a result of inadvertent contamination.However, the section below describes onlythose bacteria whose contribution to fer-mented foods manufacture is known.

The Lactic Acid Bacteria

From the outset, it is important to recognizethat the very term “lactic acid bacteria” has noofficial status in taxonomy and that it is reallyjust a general term of convenience used to de-scribe a group of functionally and geneticallyrelated bacteria. Still, the term carries rathersignificant meaning among microbiologistsand others who study food fermentations,and,therefore, will be used freely in this text. Ac-cordingly, the lactic acid bacteria are generallydefined as a cluster of lactic acid-producing,low %G�C, non-spore-forming, Gram-positiverods and cocci that share many biochemical,physiological, and genetic properties (Table2–1).They are distinguished from other Grampositive bacteria that also produce lactic acid(e.g., Bacillus, Listeria, and Bifidobacterium)by virtue of numerous phenotypic and geno-typic differences.

In addition to the traits described above,other important properties also characterizethe lactic acid bacteria, but only in a generalsense, given that exceptions occasionally exist.Most lactic acid bacteria are catalase-negative,acid-tolerant, aerotolerant, facultative anaer-obes. In terms of their carbon and energyneeds, they are classified as heterotrophicchemoorganotrophs, meaning that they re-quire pre-formed organic carbon both as asource of carbon and energy.

Until recently, it was thought that all lacticacid bacteria lack cytochrome or electrontransport proteins and, therefore, could not de-rive energy via respiratory activity. This view,however true for the majority of lactic acidbacteria, has been revised, based on recentfindings that indicate some species may indeedrespire, provided the medium contains thenecessary nutrients (Box 2–1). Still, substratelevel phosphorylation reactions that occurduring fermentative pathways (see below) arethe primary means by which ATP is obtained.

The lactic acid bacteria as a group are oftendescribed as being fastidious with complex nu-tritional requirements, and indeed, there arespecies that will grow only in nutrient-rich,well-fortified media under optimized condi-tions. However, there are also species of lacticacid bacteria that are quite versatile with re-spect to the growth environment and thatgrow reasonably well even when the nutrientcontent is less than ideal. Furthermore, somelactic acid bacteria are actually known for theirability to grow in inhospitable environments,including those that often exist in fermentedfoods.This is reflected by the diverse habitats

Table 2.1. Common characteristics of lactic acid bacteria

Gram positiveFermentativeCatalase negative Facultative anaerobesNon-sporeforming Low mol% G � C Non-motileAcid-tolerant



Box 2–1. Lactic acid bacteria learn new tricks

Look up “lactic acid bacteria” in any older (or even relatively recent) text, and in the section onphysiological characteristics, it will likely be stated that these bacteria “lack cytochromes andheme-linked electron transport proteins” and are, therefore,“unable to grow via respiration.” Infact, this description was considered to be dogma for generations of microbiology students andresearchers who studied these bacteria, despite the occasional report that suggested otherwise(Ritchey and Seeley, 1976; Sijpesteijn, 1970). In the past few years, however, it has become ap-parent that this section on physiology of lactic acid bacteria will have to be re-written, becausebiochemical and genetic evidence, reported by researchers in France, now supports the exis-tence of an intact and functional respiratory pathway in Lactococcus lactis and perhaps otherlactic acid bacteria (Duwat et al., 2001; Gaudu et al., 2002).

Respiratory metabolism

Respiration is the major means by which aerobic microorganisms obtain energy. During respi-ration, electrons generated during carbon metabolism (e.g., citric acid cycle) are transported orcarried across the cytoplasmic membrane via a series of electron carrier proteins.These pro-teins are arranged such that the flow of electrons (usually in the form of NADH+) is toward in-creasing oxidation-reduction potentials. Along the way, protons are translocated across themembrane, effectively converting an oxidation-reduction potential into a proton electrochemi-cal potential.This proton potential or proton motive force (PMF) can drive transport, operateflagella motors, or perform other energy-requiring reactions within the membrane. It can alsobe used to make ATP directly via the ATP synthase.This reaction (called oxidative phosphoryla-tion) provides aerobes with the bulk of their ATP and is the coupling step between the respira-tory electron transport system and ATP formation.

Respiration in lactic acid bacteria

For respiration to occur in lactococci, several requirements must be met. First, the relevantgenes encoding for electron transport proteins must be present, and then the functional pro-teins must be made. Based on physiological and genetic data, these genes are present and therespiratory pathway is intact in a wild-type strain of Lactococcus lactis (Vido et al.,2004). In thisand presumably other strains of lactococci, this pathway is comprised of several proteins, in-cluding dehydrogenases, menaquinones, and cytochromes. In particular, cytochrome oxidase(encoded by cydAB) serves as the terminal oxidase and is essential (Gaudu et al., 2002). Miss-ing,however, is a system for making porphyrin groups.The latter is combined with iron to makeheme, which is required for activity of cytochrome proteins, as well as the enzyme catalase.Thus, heme (or a heme precursor) must be added to the medium for respiration to occur. Res-piratory growth follows a fermentative period and occurs primarily during the later stages ofgrowth. Interestingly,expression of the heme transport system is subject to negative regulation,mediated by the catabolite control protein CcpA (Gaudu et al., 2003). Recall that the activity ofthis regulator is high during rapid growth on glucose and that genes for other catabolic path-ways are repressed.Thus, it appears that CcpA might be the metabolic conduit between fer-mentation and respiration.

Importantly, it was observed that when lactococci were grown under respiring conditions(i.e., aerobic, with heme added), not only did cells perform respiratory metabolism, they alsogrew better (Duwat et al., 2001).That is, they grew for a longer time and reached higher celldensities compared to non-respiring cells.Moreover,during prolonged incubation at 4°C,respir-ing cells remained viable for a longer time, in part because the pH remained high (near 6.0),since less lactic acid was produced via fermentation.These findings are of considerable indus-trial importance, given that one of the main goals of the starter culture industry (Chapter 3)

(continued)



occupied by lactic acid bacteria,which includenot only plant material,milk, and meat,but alsosalt brines, low pH foods, and ethanolic envi-ronments.

Perhaps the most relevant properties of lac-tic acid bacteria are those related to nutrientmetabolism. Specifically, the main reason whylactic acid bacteria are used in fermented foodsis due to their ability to metabolize sugars andmake lactic and other acid end-products. Twofermentative pathways exist. In the homofer-mentative pathway,more than 90% of the sugarsubstrate is converted exclusively to lactic acid.In contrast, the heterofermentative pathway re-sults in about 50% lactic acid, with the balanceas acetic acid, ethanol, and carbon dioxide. Lac-tic acid bacteria possess one or the other of

these two pathways (i.e., they are obligate homofermentative or obligate heterofermenta-tive),although there are some species that havethe metabolic wherewithal to perform both(facultative homofermentative). These path-ways will be described in detail later in thischapter.

The genera of lactic acid bacteria

According to current taxonomy, the lactic acidbacteria group consists of twelve genera (Table2–2). All are in the phylum Firmicutes, Order,Lactobacillales. Based on 16S rRNA sequenc-ing and other molecular techniques, the lacticacid bacteria can be grouped into a broad phy-logenetic cluster, positioned not far from other

is to maximize cell biomass while maintaining cell viability. Thus, the respiration story has certainly captured the attention of the starter culture industry (Pedersen et al., 2005). Finally,since lactococci do not have an intact citric acid cycle, the electrons that feed the respiratorychain cannot be supplied from the citric acid cycle-derived NADH pool (Vido et al., 2004). In-stead, they must be generated though other pathways.

These recent findings on lactococci have not yet been extended to other lactic acid bacteria.However, based on several sequenced genomes, it appears that cyd genes may be present inother lactococci as well as some oenococci and leuconostocs (but are absent in Streptococcusthermophilus and Lactobacillus delbrueckii subsp. bulgaricus). If, and under what circum-stances these bacteria actually perform respiration, is not known.As had initially been reportedthirty years ago, and re-confirmed by these more recent studies, molar growth and ATP yieldsare higher during respiratory growth.Thus, it may well be that the respiratory pathway providesthese bacteria an alternative and efficient life-style choice, were they to find themselves backout in nature rather than in the confines of food environments (Duwat et al., 2001).

References

Duwat, P., S. Sourice, B. Cesselin, G. Lamberet, K.Vido, P. Gaudu,Y. Le Loir, F.Violet, P. Loubière, and A. Gruss.2001. Respiration capacity of the fermenting bacterium Lactococcus lactis and its positive effects ongrowth and survival. J. Bacteriol. 183:4509–4516.

Gaudu, P., G. Lamberet, S. Poncet, and A. Gruss. 2003. CcpA regulation of aerobic and respiration growth inLactococcus lactis. Mol. Microbiol. 50:183–192.

Gaudu, P., K.Vido, B. Cesselin, S. Kulakauskas, J.Tremblay, L. Rezaïki, G. Lamberet, S. Sourice, P. Duwat, and A.Gruss. 2002. Respiration capacity and consequences in Lactococcus lactis.Antonie van Leeuwenhoek82:263–269.

Pedersen, M.B., S.L. Iversen, K.I. Sørensen, and E. Johansen. 2005.The long and winding road from the re-search laboratory to industrial applications of lactic acid bacteria. FEMS Microbiol. Rev. 29:611–624.

Ritchey,T.W., and H.W. Seely. 1976. Distribution of cytochrome-like respiration in streptococci. J. Gen. Mi-crobiol. 93:195–203.

Sijpesteijn, A.K. 1970. Induction of cytochrome formation and stimulation of oxidative dissimilation byhemin in Streptococcus lactis and Leuconostoc mesenteroides.Antonie Van Leeuwenhoek 36:335–348.

Box 2–1. Lactic acid bacteria learn new tricks (Continued)



low G � C Gram positive bacteria (Figure 2–2).Five sub-clusters are evident from this tree,including: (1) a Streptococcus-Lactococcusbranch (Family Streptococcaceae), (2) a Lacto-bacillus branch (Family Lactobacillaceae), (3)a separate Lactobacillus-Pediococcus branch(Family Lactobacillaceae); (4) an Oenococcus-Leuconostoc-Weisella branch (Family Leu-conostocaceae), and (5) a Carnobacterium-Aerococcus-Enterococcus-Tetragenococcus-Vagococcus branch (Families Carnobacteri-aceae, Aerococcaceae, and Enterococcaceae).

It is worth noting that this phylogeny is notentirely consistent with regard to the morpho-logical and physiological characteristics ofthese bacteria. For example, Lactobacillus andPediococcus are in the same sub-cluster,yet theformer are rods and include heterofermenta-tive species, whereas the latter are homofer-menting cocci. Likewise, Carnobacterium are

obligate heterofermentative rods, and Entero-coccus and Vagococcus are homofermentativecocci.

Seven of the twelve genera of lactic acidbacteria, Lactobacillus, Lactococcus, Leucon-ostoc,Oenococcus,Pediococcus,Streptococcus,and Tetragenococcus, are used directly in foodfermentations. Although Enterococcus sp. areoften found in fermented foods (e.g., cheese,sausage, fermented vegetables), except for afew occasions, they are not added directly. Infact, their presence is often undesirable, inpart, because they are sometimes used as indi-cators of fecal contamination and also becausesome strains may harbor mobile antibiotic-resistance genes.

Importantly, some strains of Enterococcusare capable of causing infections in humans.Likewise, Carnobacterium are also undesir-able, mainly because they are considered as

Lactobacillus delbrueckii

Staphylococcus aureus

Bacillus subtilis

Listeria monocytogenes

Tetragenococcus halophilus

Lactobacillus casei

Lactobacillus brevis

Pediococcus pentosaceus

Lactobacillus plantarum

Lactobacillus gasseri

Lactobacillus johnsonii

Enterococcus faecalis

Vagococcus salmoninarum

Carnobacterium funditum

Aerococcus viridans

Oenococcus oeniLeuconostoc mesenteroides

Lactococcus lactis

Lactococcus cremorisWeissella cibaria

Streptococcus pneumoniae

Streptococcus thermophilus

Clostridium botulinum

0.1

Figure 2–2. Phylogeny of lactic acid and other Gram positive bacteria (based on 16s rRNA). The tree (un-rooted) was generated using the neighbor-joining method.



Tab

le 2

.2.

Gen

era

of la

ctic

aci

d ba

cter

ia a

nd t

heir

prop

ertie

s1,2

Cel

lFe

rmen

tati

on

Gro

wth

at:

Gro

wth

in N

aCla

t:G

row

th a

t p

H:

Lact

ic a

cid

Gen

us

Mo

rph

olo

gyro

ute

10°C

45°C

6.5%

18%

4.4

9.6

iso

mer

Lact

oba

cillu

sro

ds

ho

mo

/het

ero

3�

4�

��

��

D,L

,DL5

Lact

oco

ccu

sco

cci

ho

mo

��

��

��

LLe

uco

nost

oc

cocc

ih

eter

o�

��

��

�D

Oen

oco

ccu

sco

cci

het

ero

��

��

��

DPe

dio

cocc

us

cocc

i (te

trad

s)h

om

o�

��

��

�D

,L,D

LSt

repto

cocc

us

cocc

ih

om

o�

��

��

�L

Tetr

age

noco

ccu

sco

cci (

tetr

ads)

ho

mo

��

��

��

LA

eroco

ccu

sco

cci (

tetr

ads)

ho

mo

��

��

��

LC

arn

oba

cter

ium

rod

sh

eter

o�

��

��

�L

En

tero

cocc

us

cocc

ih

om

o�

��

��

�L

Va

goco

ccu

sco

cci

ho

mo

��

��

��

LW

eiss

ella

cocc

oid

het

ero

��

��

��

D,L

,DL

1A

dap

ted

fro

m A

xel

sso

n,2

004

2R

efer

s to

th

e ge

ner

al p

rop

erti

es o

f th

e ge

nu

s;so

me

exce

pti

on

s m

ay e

xis

t3Sp

ecie

s o

f La

ctoba

cillu

sm

ay b

e h

om

ofe

rmen

tati

ve,h

eter

ofe

rmen

tati

ve,o

r b

oth

4T

his

ph

eno

typ

e is

var

iab

le,d

epen

din

g o

n t

he

spec

ies

5So

me

spec

ies

pro

du

ce D

-,L-

,or

a m

ixtu

re o

f D

- an

d L

-lact

ic a

cid

.



spoilage organisms in fermented meat prod-ucts. Finally, species of Aerococcus, Vagococ-cus, and Weisella are not widely found infoods, and their overall significance in food isunclear. In the section to follow, the generalproperties, habitats, and practical considera-tions of the genera that are relevant to food fer-mentations will be described; descriptions ofthe important species will also be given. Infor-mation on the genetics of these bacteria, basedon genome sequencing and functional geno-mics, will be presented later in this chapter.

Lactococcus

The genus Lactococcus consists of five phylo-genetically-distinct species: Lactococcus lactis,Lactococcus garviae, Lactococcus piscium,Lactococcus plantarum, and Lactococcus raf-finolactis (Figure 2–3).They are all non-motile,obligately homofermentative, facultative anaer-obes, with an optimum growth temperaturenear 30°C. They have a distinctive microscopicmorphology, usually appearing as cocci inpairs or short chains.

One species in particular L. lactis, is amongthe most important of all lactic acid bacteria(and perhaps one of the most important organ-

isms involved in food fermentations, period).This is because L. lactis is the “work horse” ofthe dairy products industry—it is used as astarter culture for most of the hard cheesesand many of the cultured dairy products pro-duced around the world. There are actuallythree L. lactis subspecies: L. lactis subsp. lactis,Lactococcus lactis subsp. cremoris, and Lacto-coccus lactis subsp. hordinae. Only L. lactissubsp. lactis and L. lactis subsp. cremoris, how-ever,are used as starter cultures;L. lactis subsp.hordinae has no relevance in fermented foodmanufacture.Another variant of L. lactis subsp.lactis, formerly named L. lactis subsp. diaceti-lactis (or Lactococcus lactis subsp. lactis bio-var. diacetylactis), is distinguished based on itsability to metabolize citrate.This species is notincluded in the current List of Bacterial Names,and instead is encompassed within L. lactissubsp. lactis (see below). However, Lacto-coccus lactis subsp. diacetilactis is listed inBergey’s Taxonomical Outline of the Prokary-otes (2004 version).

Plant material has long been considered tobe the “original” habitat of both L. lactis subsp.lactis and L. lactis subsp. cremoris. The sugges-tion, however, that milk is now their “new”habitat is supported by several observations.

Lactococcus raffinolactis

Lactococcus plantarum

Lactococcus piscium

Lactococcus garvieae

Lactococcus lactis subsp. cremoris

Lactococcus lactis subsp. lactis

Lactococcus lactis subsp. hordniae0.01

Figure 2–3. Phylogeny of Lactococcus based on 16S rRNA sequence analysis.



First, they are readily isolated from raw milk;in fact, it is difficult to find L. lactis subsp.cremoris anywhere but milk. Second, bothspecies grow rapidly in milk, producing lacticacid and lowering the pH to below 4.5.Thus,they generally out-compete most potentialcompetitors. Finally, the genes required forgrowth in milk are located on plasmids (extra-chromosomal DNA), indicating they were ac-quired recently (relatively, speaking).

Specific plasmid-borne genes in lactococciencode for proteins involved in lactose trans-port and metabolism and casein hydrolysisand utilization.There is considerable selectivepressure, therefore, for milk-borne strains toretain these plasmids, since plasmid-cured de-rivatives grow poorly in milk. Indeed, somelactococcal plasmids are readily exchangedamong different strains (via conjugal transfer),and in other cases, plasmid DNA can integratewithin the chromosome, resulting in stable,highly adapted derivative strains. Plasmids,however, are common in lactococci, even instrains isolated from non-dairy sources.

Although L. lactis subsp. lactis and L. lactissubsp. cremoris differ in just a few seeminglyminor physiological respects, at least two ofthese differences are significant during milkfermentations. For example, the highest tem-perature at which most strains of L. lactissubsp. lactis are able to grow is 40°C, whereasmost L. lactis subsp. cremoris strains do notgrow above 38°C. In addition, L. lactis subsp.lactis has greater tolerance to salt (up to 4%)than does L. lactis subsp. cremoris.

What makes these differences relevant isthat both temperature and salt are among theprimary means by which the activity of thestarter culture is controlled during cheesemanufacture. Moreover, L. lactis subsp. cre-moris is generally considered to be, overall,less tolerant to the environmental stresses en-countered during both culture production andduring dairy fermentations. Thus, if L. lactissubsp. cremoris is used as a culture for cheesemaking (and it is generally accepted that thisorganism makes better quality products), thentemperature and salt concentrations must be

adjusted to account for this organism’s particu-lar growth requirements.

Complicating this discussion on the differ-ences between these two organisms, it has recently been noted that there are strains of L. lactis subsp. lactis that have a “lactis” geno-type, but a “cremoris-like” phenotype. Like-wise, there are strains of L. lactis subsp. cre-moris that have a “cremoris” genotype, but a“lactis-like” phenotype (Box 2–2). The use ofgenomics (discussed later in the chapter), willhelp to determine the genetic basis for theseindustrially relevant differences.

Streptococcus

The genus Streptococcus contains many di-verse species with a wide array of habitats. In-cluded in this genus are human and animalpathogens, oral commensals, intestinal com-mensals, and one (and only one) species, Strep-tococcus thermophilus, that is used in themanufacture of fermented foods. In general,streptococci are non-motile, facultative anaer-obes, with an obligate homofermentative me-tabolism.

Since the mid-1980s, there have been severalmajor taxonomical revisions within this genus.Some of these changes were especially relevantfor food microbiologists. For the previous fiftyyears, the streptococci were divided into fourmain groups: pyogenic, enterococcus, viridans,and lactic.These groupings, which were basedon the so-called Sherman scheme (published in1937, as described in Chapter 3), were revisedin subsequent years,but generally served as theprimary means for organizing streptococcalspecies. Starting in 1984, however, two majorrevisions were proposed and adopted.First, theenterococcus (or enteric streptococci), whichincluded Streptococcus faecalis, Streptococcusfaecium, and Streptococcus durans, weremoved to a new genus, Enterococcus.Then, in1985, two species that had been referred to as“lactic streptococci” (Streptococcus lactis andStreptococcus cremoris) were also assigned toa new genus, Lactococcus (see above).Thus, S.thermophilus is now the only member of this



Box 2–2. Lactococcus lactis—Differences Between the Subspecies

As might be expected, Lactococcus lactis subsp. lactis and Lactococcus lactis subsp. cremorisshare many phenotypic properties.In fact,these organisms are so similar that in the 1986 Bergey’sManual of Systematic Bacteriology, the lactis and cremoris strains were considered as belongingto a single species (Mundt, 1986). However, several phenotypic differences do exist, as describedearlier, and these differences have traditionally been used to distinguish between the two sub-species (Table 1).As noted previously, differences also exist at the molecular level. Currently, as-signment of a strain to the appropriate subspecies is based on the sequence of a 9 to 10 base pairregion of the VI segment in the 16S rRNA gene. Moreover, the genomes of both organisms havenow been sequenced, and other differences are likely to become apparent as comparative analy-ses of these genome are completed (Klaenhammer et al., 2002;Makarova et al., 2006).

Table 1. Distinguishing characteristics of dairy Lactococcus1.

Lactococcus lactis Lactococcus lactisPropertysubsp. lactis subsp. cremoris

Growth at 40°C � �

Growth in 4.0% NaCl � �

Growth at pH 9.2 � �

Growth in 0.1% methylene blue � �

Arginine hydrolysis � �

Acid from maltose � �

Acid from ribose � �

1Adapted from Teuber, 1992

Interestingly, it now appears that classification of L. lactis by both phenotypic and genotypiccriteria is not so clear-cut. In fact,one research group recently proposed that L. lactis strains canbe organized into one of five different groups (Kelly and Ward,2002).Group 1 consists of strainshaving a L. lactis subsp. lactis genotype and a “lactis” phenotype.These strains may be consid-ered as the common or typical L. lactis subsp. lactis. Group 2 strains also have a L. lactis subsp.lactis genotype, but have one particular distinguishing phenotypic property, namely the abilityto ferment citrate and to produce diacetyl. (As noted previously, these citrate-citrate-fermentingstrains had, for a long time, held subspecies status.) Strains in Group 3 have a cremoris pheno-type, but a L. lactis subsp. lactis genotype.They are not commonly found. In contrast, Group 4strains, having a L. lactis subsp. cremoris genotype but a lactis phenotype, are frequently iso-lated from milk and plant sources.Their lactic-like phenotype makes it difficult to distinguishthem from other lactis strains. Finally, Group 5 strains have a L. lactis subsp. cremoris genotypeand phenotype.They are found almost exclusively in milk and dairy environments, and are themost common and preferred strains for manufacture of cheese.

On a practical basis, it is probably easier to establish the genotype of a particular lactococcalstrain than it is to determine its phenotype (at least with regard to traits relevant to fermenta-tions).However, it is the phenotypic properties of a given strain that dictate whether that strainwill be useful in fermented foods manufacture. In some cases, properties beyond those nor-mally considered as part of a phenotype will be relevant, such as bacteriophage sensitivity orthe ability to produce good cheese flavor.A genetic method recently was devised to identify L.lactis strains to the subspecies level; it also provided a basis for determining phenotypes (No-mura et al., 2002).The method was based on the presence of glutamate decarboxylase activityin L. lactis subsp. lactis, which contains a functional gadB gene, and its absence in L. lactissubsp. cremoris, which contains insertions, deletions, and point mutations in gadB. Althoughthe sample set was small, the size of specific DNA fragments from the gadB gene (amplified by

(continued)



genus used in food fermentations (mainly yo-gurt and cheese).

That is not to say that this organism has beenexempt from taxonomical considerations.Origi-nally, this species was part of the Sherman viri-dans group (which included oral streptococci),and in the 1986 edition of Bergey’s manual, itwas listed as an “Other Streptococci.” Its physio-logical, structural, and genetic similarities toStreptococcus salivarius led to the recommen-dation in 1984 that it be reclassified as a sub-species of S. salivarius. Although this namechange was indeed adopted (and “Streptococcussalivarius subsp. thermophilus” is sometimesstill seen in the literature), subsequent DNA-DNA homology studies eventually led to therestoration of its original name.Thus, labels onyogurt products can justifiably claim the pres-ence of Streptococcus thermophilus rather thanthe less appetizing Streptococcus salivarius.

In several respects,S.thermophilus is not thatdifferent from the mesophilic dairy lactococci(L. lactis subsp. lactis and L. lactis subsp. cre-moris), as is evident by their close phylogenetic

position (Figure 2–2).Like the lactococci,S. ther-mophilus is highly adapted to a milk environ-ment in that it ferments lactose rapidly and pro-duces lactic acid in homolactic fashion. Theroute by which lactose is metabolized by S. ther-mophilus, however, is quite different from howL. lactis ferments this sugar (discussed later).Also, S. thermophilus has a higher temperatureoptima (40°C to 42°C), a higher maximumgrowth temperature (52°C), and a higher ther-mal tolerance (above 60°C). Its nutritional re-quirements are somewhat more demanding thanlactococci, in that S. thermophilus is weakly pro-teolytic and, therefore, needs pre-formed aminoacids. Salt tolerance,bile sensitivity, and a limitedmetabolic diversity are also characteristic of S. thermophilus. In fact, the statement made bySherman in 1937 that S.thermophilus“is markedmore by the things which it cannot do than byits positive actions”describes this organism quitewell. Finally, in contrast to lactococci, S. ther-mophilus strains contain few plasmids. Whenplasmids are found, they are generally small andcryptic (i.e.,having no known function).

PCR) from L. lactis subsp. lactis and L. lactis subsp. cremoris strains correlated with both theirgenotypes and phenotypes. Ultimately, this and other similar procedures could provide a rela-tively rapid and efficient means to screen lactococcal strains with the desired phenotype.

References

Bolotin,A.,P.Wincker,S.Mauger,O. Jaillon,K.Malarme, J.Weissenbach,S.D.Ehrlich,and A.Sorokin.2001.Thecomplete genome sequence of the lactic acid bacterium Lactococcus lactis ssp. lactis IL1403. GenomeRes. 11:731–753.

Goffeau,A., B. G. Barrell, H. Bussey, R.W. Davis, B. Dujon, H. Feldmann, F. Galibert, J. D. Hoheisel, C. Jacq, M.Johnston, E. J. Louis, H.W. Mewes,Y. Murakami, P. Philippsen, H.Tettelin, and S. G. Oliver. 1996. Life with6000 genes. Science 274:546–567.

Kelly, W., and L. Ward. 2002. Genotypic vs. phenotypic biodiversity in Lactococcus lactis. Microbiol.148:3332–3333.

Klaenhammer et al. (36 other authors). 2002. Discovering lactic acid bacteria by genomics.Antonie vanLeeuwenhoek 82:29–58.

Makarova, K.,Y.Wolf, et al., (and 43 other authors) 2006. Comparative genomics of the lactic acid bacteria.2006. Proc. Nat.Acad. Sci. In Press.

Mundt, J.O. 1986. Lactic acid streptococci. In Sneath, P.H.A., N.S. Mair, M.E. Sharpe, J.G. Holt. Bergey’s Man-ual of Systematic Bacteriology, Volume 2.Williams and Wilkins, Baltimore, Maryland. 1065–1071.

Nomura,M.,M.Kobayashi,and T.Okamoto.2002.Rapid PCR-based method which can determine both phe-notype and genotype of Lactococcus lactis subspecies.Appl. Environ. Microbiol. 68:2209–2213.

Teuber, M. 1992.The genus Lactococcus. In B.J.B.Wood and W.H. Holzapfel, ed. The Lactic Acid Bacteria,Volume 2, The genera of lactic acid bacteria. Blackie Academic and Professional pp 173—234.

Box 2–2. Lactococcus lactis—Differences Between the Subspecies (Continued)



Leuconostoc

The Leuconostoc belong to the Leuconosto-caceae Family, which also contains the closelyrelated genera Weissella and Oenococcus (seebelow). Leuconostocs are mesophilic, with optimum growth temperatures ranging from18°C to 25°C. Some species are capable ofgrowth at temperatures below 10°C. Micro-scopically, they appear coccoid or even some-what rod-like, depending on the compositionand form of the growth medium (liquid versussolid).The leuconostocs, in contrast to the ob-ligate homofermenting lactococci and strepto-cocci, are obligately heterofermentative. Ac-cordingly, they are missing an intact glycolyticpathway, and instead rely on the phosphoketo-lase pathway for metabolism of sugars (see be-low). Although leuconostocs can grow in anambient atmosphere, a reduced or anaerobicenvironment generally enhances growth. Plas-mids are common in Leuconostoc species and,when present, may encode for important func-tions including lactose and citrate metabolismand bacteriocin production.

The genus Leuconostoc has also been sub-ject to taxonomical revision, because severalspecies have been moved to other existing ornewly-formed genera. For example, Oenococ-cus oeni was formerly classified as Leuconos-toc oenos, and other Leuconostoc species havebeen reclassified as Weisella. The genus cur-rently consists of thirteen species (Figure 2–4).

Speciation is based on both genetic andphenotypic characteristics. The latter includecarbohydrate fermentation profiles, the abilityto produce the polysaccharide dextran, resis-tance to the antibiotic vancomycin, and vari-ous physiological properties (Table 2–3). Mostspecies are associated with particular habitats,including plant and vegetable material, milkand dairy environments, and meat products. Inaddition, some species are involved in foodspoilage (e.g., Leuconostoc gasicomitatum),whereas others are used in food fermentations.The latter include Leuconostoc mesenteroidessubsp. cremoris and Leuconostoc lactis, whichare used in dairy fermentations, and Leuconos-

toc mesenteroides subsp. mesenteroides, Leu-conostoc kimchii, and Leuconostoc fallax, thatare used in vegetable fermentations. In general,the habitats occupied by these strains are re-flected by the particular carbohydrates theyferment. Plant-associated strains, such as L.mesenteroides subsp. mesenteroides, fermentplant sugars (i.e., fructose, sucrose, arabinose,trehalose), whereas dairy strains (L. mesen-teroides subsp.cremoris and L. lactis) are morelikely to ferment lactose, galactose, and glu-cose. It is also possible for otherwise usefulspecies to cause spoilage, particularly in thecase of polysaccharide-producing strains thatform slime.

Heterofermentative metabolism of sugars byleuconostocs results in formation of a mixtureof end-products, including lactic acid, aceticacid,ethanol, and carbon dioxide.The presenceof CO2 is readily detected and can be used di-agnostically to separate these bacteria from ho-mofermentative streptococci, lactococci, andpediococci. Also, while leuconostocs generallydecrease the pH in the growth medium to be-tween 4.5 and 5.0, acid production by somespecies may be relatively modest, especiallywhen compared to homofermentative lacto-bacilli and other lactic acid bacteria.Thus,acidi-fication is not necessarily the major function ofthese bacteria during fermentations.

In sauerkraut and other vegetable fermenta-tions, for example, they lower the pH duringthe very early manufacturing stages and pro-duce enough CO2 to reduce the redox poten-tial in the food environment.These metabolicevents then create an atmosphere conducivefor growth of other microorganisms that areresponsible for more significant acid develop-ment. In addition, the heterofermentative end-products result in a more diverse flavor andaroma profile in fermented foods. The Leu-conostoc species used in dairy fermentationsare particularly important for flavor, in part, forthe same reasons as the vegetable-associatedstrains. However, dairy leucononstocs also pro-duce the four-carbon volatile, diacetyl, whichimparts a desirable buttery aroma to cultureddairy products (discussed below).



Oenococcus

This genus was established as recently as 1995,and contains only one species,O.oeni. Not sur-prisingly, O. oeni is located, phylogenetically,within the Leuconostocaceae branch; how-ever, it is somewhat distant from Leuconostocand Weissella (Figure 2–2). Although O. oenishares many phenotypic properties with Leu-conostoc sp. (e.g., heterofermentative metab-olism, mesophilic growth range), several important physiological differences exist. Inparticular, O. oeni is much more acid-tolerantthan Leuconostoc as it is able to commencegrowth in media with a pH below 5.0. In addi-tion,O.oeni is one of the most ethanol-tolerant

of all lactic acid bacteria and can grow in thepresence of 10% ethanol. In general, however,most strains of O. oeni are slow-growing andferment a limited number of sugars.

As implied by the etymology of its name,theuse of O. oeni in fermented foods is restrictedto only one application, namely wine making(oenos is the Greek word for wine). Despite itslimited use,however, the importance of O.oeniduring the wine fermentation cannot be over-stated.This is because O. oeni has the ability tode-acidify wine via the malolactic fermenta-tion, whereby malic acid is decarboxylated tolactic acid (described in detail in Chapter 10).Moreover, given the ability of O. oeni to fer-

Leuconostoc fallax

Leuconostoc pseudomesenteroides

Leuconostoc mesenteroides subsp. cremoris

Leuconostoc mesenteroides

Leuconostoc citreum

Leuconostoc lactis

Leuconostoc argentinum

Leuconostoc kimchii

Leuconostoc inhae

Leuconostoc gelidum

Leuconostoc gasicomitatum

Leuconostoc carnosum

Leuconostoc fructosum

Leuconostoc ficulneum

0.01

Figure 2–4. Phylogeny of Leuconostoc based on 16S rRNA sequence analysis.



ment glucose and fructose, but few other sug-ars, and its high tolerance to low pH and highethanol, wine or juice would appear to be thenatural habitat for this organism.

Pediococcus

The pediococci are similar, in many respects, toother coccoid-shaped, obligate homofermenta-tive lactic acid bacteria, with one main excep-tion.When these bacteria divide, they do so intwo “planes” (and in right angles).Thus, tetradsare formed, which can be observed visually.Cells may appear as pairs (and always sphericalin shape), but chains are not formed, as theyare for lactococci, streptococci, and leuconos-tocs. Pediococci, like other lactic acid bacteria,are facultative anaerobes, with complex nutri-tional requirements. They have optimumgrowth temperatures ranging from 25°C to40°C, but some species can grow at tempera-tures as high as 50°C.Several of the pediococciare also distinguished from other lactic acidbacteria by their ability to tolerate high acid(growth at pH 4.2) and high salt (growth at6.5% NaCl) environments (Table 2–4).

The pediococci can be found in diverse habi-tats, including plant material, milk, brines, ani-

mal urine, and beer. There are six recognizedspecies of Pediococcus (Figure 2–5); several areimportant in food fermentations. Two species,Pediococcus acidilactici and Pediococcus pen-tosaceus, are naturally present in raw vegeta-bles,where,under suitable conditions, they playa key role in the manufacture of sauerkraut andother fermented vegetables.These same speciesmay also be added to meat to produce fer-mented sausages. Despite their inability to fer-ment lactose, P. acidilactici and P. pentosaceusare frequently found in cheese, where they mayparticipate in the ripening process. Pediococciare also important as spoilage organisms in fer-mented foods, in particular, beer, wine, andcider. One species, Pediococcus damnosus, isespecially a problem in beer, where it producesdiacetyl,which in beer is a serious defect.

Plasmids are frequently present in pedio-cocci.The genes located on these plasmids mayencode for functions involved in sugar metabo-lism (e.g., raffinose and sucrose) and produc-tion of bacteriocins. The latter (described inChapter 6) are defined as anti-microbial pro-teins that inhibit closely related bacteria,including the meat-associated pathogens Lis-teria monocytogenes, Staphylococcus aureus,and Clostridium botulinum. Bacteriocin

Table 2.3. Characteristics of Leuconostoc and Oenococcus1

Leuconostoc mesenteroides subsp. Leuconostoc Oenococcus Propertycremoris mesenteroides dextranicum lactis oeni

% G+C 38–44 28–32 28–32 37–42 37–42Growth at 37°C � �2 � � �

Growth at pH 4.8 � � � � �

Dextran from sucrose � � � � �

Growth in 10% ethanol � � � � �

Acid from:arabinose � � � � �

fructose � � � � �

maltose � � � � �

melibiose � � � � �

salicin � � � � �

sucrose � � � � �

trehalose � � � � �

1Adapted from Bergey’s Manual of Determinative Bacteriology2Variable reaction, depending on strain



production is a particularly important trait inpediococci, since these bacteria are used asstarter cultures for the sausage fermentation.Thus, bacteriocin-producing pediococci couldprovide an additional level of preservationwhen used in sausage manufacture.

Tetragenococcus

Like the pediococci, Tetragenococcus are ho-mofermentative, tetrad-forming, facultativeanaerobes. They are mesophilic and neu-traphilic, with temperature optima generally

Table 2.4. Characteristics of Pediococcus1

Pediococcus Pediococcus Pediococcus Propertyacidilactici pentosaceus damnosus

% G�C 38–44 37 37–42

Growth at:35°C � � �

40°C � � �

50°C � � �

Optimum growth temperature (°C) 40 28–32 28–32

Growth at:pH 4.2 � � �

pH 7.0 � � �

Growth in:4.0% NaCl � � �

6.5% NaCl � � �

Arginine hydrolysis � � �

1Adapted from Bergey’s Manual of Determinative Bacteriology

Tetragenococcus halophilus

Pediococcus dextrinicus

Pediococcus pentosaceus

Pediococcus acidilactici

Pediococcus parvulus

Pediococcus inopinatus

Pediococcus damnosus0.01

Figure 2–5. Phylogeny of Pediococcus based on 16S rRNA sequence analysis.



between 25°C and 30°C and pH optima be-tween 6.5 and 8.0. The genus contains onlythree species, Tetragenococcus halophilus,Tetragenococcus muriaticus, and Tetrageno-coccus solitarius. Based on 16S rRNA se-quences, these bacteria are phylogeneticallymore closely related to Lactobacillus and Ente-rococcus, than to Pediococcus.

The genus was first recognized in 1990,when Pediococcus halophilus was re-classifiedas T. halophilus. A second species, T. muriati-cus, isolated from salty fish sauce, was pro-posed in 1997, and in 2005, Enterococcus soli-tarius was reclassified as T. solitarius. All ofthese species are characterized by their re-markable tolerance to salt. Not only is growthpossible in media containing up to 25% salt,but maximum growth rates require the pres-ence of 3% to 10% NaCl, and no growth occurswhen salt is absent.They also grow well whenthe water activity is low, presumably due totheir ability to maintain osmotic balance by ac-cumulating betaine, carnitine, and other com-patible solutes.

Lactobacillus

The genus Lactobacillus consists of more thaneighty species (Figure 2–6). In the last decade,microbial taxonomists have been very busy,proposing and validating new taxa. In one sin-gle month (January 2005), seven new speciesand subspecies were described in the pub-lished literature. About all that is common tothe species within this genus is that they are allnon-sporing rods, but even this description isnot wholly satisfactory.There are species thatappear rather short (�1.5 �m),whereas othersare more than 5 �m in length (some are re-portedly up to 10 �m).They may also have aslender, curved, or bent appearance. Whenviewed microscopically, it is difficult to be surethat a given strain is even a rod. Colony mor-phology is also variable on agar plates, withsome strains producing large round coloniesand others producing small or irregularcolonies. Not only does this genus contain farmore species, with extensive morphologicalheterogeneity, than any other genera of lactic

acid bacteria, but this group is also the mostecologically, physiologically, biochemically, andgenetically diverse.

Ecologically, lactobacilli occupy a wide rangeof habitats. Except for very extreme environ-ments, there are few locales where lactobacilliare not found, and they are often described asbeing ubiquitous in nature. Some species arenormal inhabitants of plant and vegetable mater-ial, and they are frequently found in dairy andmeat environments, in juice and fermented bev-erages, and in grains and cereal products.Theirpresence in the animal and human gastrointesti-nal tract (as well as in the stomach, mouth andvagina) has led to the suggestion (which hasgained substantial scientific support) that thesebacteria have broad “probiotic activity,”meaningthey promote intestinal as well as extra-intestinalhealth. In foods, they are involved not only inmany important fermentations, but are also fre-quently implicated in spoilage of fermented andnon-fermented foods.

The ability of lactobacilli to grow and persistin so many diverse environments and conditionsreflects their diverse physiological properties.Al-though most species are mesophilic, the genusalso contains species that are psychrotrophic,thermoduric, or thermophilic. Temperature op-tima varies widely, from 30°C to 45°C. Somespecies show high tolerance to salt, osmoticpressure,and low water activity.Acid-tolerance isa common trait of lactobacilli (many strains actu-ally prefer an acidic environment),and some alsoare ethanol-tolerant or bile-tolerant.Most speciesare aerotolerant, whereas others require morestrict anaerobic conditions.

Like all lactic acid bacteria, lactobacilli arefermentative,but,again, they are more metaboli-cally diverse than other lactics. In fact, one ofthe major ways in which the genus is subdi-vided into groups is based on the pathways theyuse to ferment sugars.As noted previously, lacticacid bacteria,in general,ferment sugars in eitherhomofermentative or heterofermentative fash-ion. However, some species of Lactobacillushave the genetic and physiological wherewithalto ferment sugars by either pathway. Thesespecies, therefore, are referred to as facultativeheterofermentative.



Lactobacillus aviariusLactobacillus salivarius subsp. salivariusLactobacillus salivarius subsp. saliciniusLactobacillus ruminisLactobacillus agilisLactobacillus maliLactobacillus animalisLactobacillus murinusLactobacillus perolensLactobacillus sakeiLactobacillus bifermentansLactobacillus coryniformisLactobacillus durianisLactobacillus manihotivoransLactobacillus rhamnosusLactobacillus zeaeLactobacillus paracasei subsp. paracaseiLactobacillus casei subsp. caseiLactobacillus sharpeaeLactobacillus pantherisLactobacillus kunkeeiLactobacillus brevisLactobacillus ferintoshensisLactobacillus kefiriLactobacillus buchneriLactobacillus hilgardiiLactobacillus vermiformeLactobacillus fructivoransLactobacillus lindneriLactobacillus sanfranciscensisLactobacillus versmoldensisLactobacillus farciminisLactobacillus mindensisLactobacillus alimentariusLactobacillus kimchiiLactobacillus paralimentariusLactobacillus arizonensisLactobacillus plantarumLactobacillus reuteriLactobacillus pontisLactobacillus frumentiLactobacillus mucosaeLactobacillus fermentumLactobacillus thermotoleransLactobacillus amylophilusLactobacillus acetotoleransLactobacillus amylolyticusLactobacillus gallinarumLactobacillus helveticusLactobacillus acidophilusLactobacillus amylovorusLactobacillus crispatusLactobacillus delbrueckii subsp. lactisLactobacillus delbrueckii subsp. bulgaricusLactobacillus psittaciLactobacillus jenseniiLactobacillus gasseriLactobacillus johnsonii0.01

Figure 2–6. Phylogeny of Lactobacillus based on 16S rRNA sequence analysis. Not all of the speciesare listed.



Accordingly, Group I lactobacilli consist ofobligate homofermenting species, Group IIcontains facultative heterofermenting species,and Group III contains obligate heteroferment-ing species (Table 2–5). These groups can befurther divided on the basis of other biochemi-cal properties.

Despite their metabolic diversity, lactobacilligenerally are quite fastidious, and many speciesrequire nutrient-rich environments. They arenot especially proteolytic or lipolytic, andamino acids, peptides, and fatty acids are usu-ally required for rapid growth. Some strains areparticularly demanding, requiring an array of vi-tamins, nucelotides, and other nutrients. Lacto-bacilli, of course also need fermentable carbo-hydrates, and, depending on their source orhabitat, they ferment a wide array of sugars.Thus, not only are most common sugars (e.g.,glucose, fructose, lactose) fermented, but manyplant-derived carbohydrates,such as cellobiose,amygdalin, and trehalose, are also used. Severalspecies even ferment starch (Table 2–5).

Numerous species of Lactobacillus are rele-vant in fermented foods. Some are added di-rectly in the form of starter cultures (Chapter3), but there are even more endogenous lacto-bacilli present in the raw material or equip-ment surfaces that indirectly contribute to themanufacture or to the finished properties offermented foods.

Starter culture lactobacilli are used primarilyin dairy and sausage applications.There are twomain species used as dairy starter cultures(mainly for cheese and yogurt), Lactobacillushelveticus and Lactobacillus delbrueckii subsp.bulgaricus.Other common dairy-related speciesinclude Lactobacillus casei and Lactobacillusacidophilus (both used frequently as probi-otics). Starter cultures for sausage fermentationscontain Lactobacillus plantarum or Lactobacil-lus sakei subsp.sakei.

Sourdough breads are also made using het-erofermentative Lactobacillus sanfranciscen-sis, Lactobacillus brevis, and other lactobacilli.Pure starter cultures for sourdough breads areavailable, but wild cultures, propagated “inhouse,”are still commonly used (Chapter 8).

For the production of sauerkraut,kimchi,andother fermented vegetables, the natural micro-flora is all that is necessary to initiate and per-form the fermentation, although pure startercultures containing L. plantarum and L. brevisand related species have become more com-mon for so-called “controlled pickle” fermenta-tions.

Finally, it is very important to emphasize,de-spite the general association of a particularspecies with a particular fermented food, that agiven species will often be used in or appear inquite dissimilar situations. In many cases, thespecies name may even add to the confusion.Thus,L. sakei was originally isolated from sake,Japanese rice wine, yet this same organism canbe isolated from fermented sausage and is evenused in sausage starter cultures. Likewise, L.plantarum is not just found in plant materialor in pickle brines, as its name might indicate,but also appears to be a normal inhabitant ofthe human intestinal tract.

As one would expect, the lactobacilli are ge-netically very diverse. Although the lactic acidbacteria,as a group,generally have a low %G�C(35 to 40), some lactobacilli have %G�C as lowas 32 and others are as high as 55. The lacto-bacilli also are positioned phylogenetically inone of two distinct branches. For example, L.plantarum and L. brevis belong to the “Lacto-bacillus-Pediococcus” branch, whereas L. del-brueckii,L.helveticus and L.acidophilus are lo-cated in a separate “Lactobacillus delbrueckii”branch (Figures 2–2 and 2–6). Plasmids arecommon in the lactobacilli, although for manyof these plasmids, the functions of the encodedgenes are not established. Lactose metabolismis encoded on at least one plasmid, and severalbacteriocin production and antibiotic-resistantsystems are also plasmid-encoded.

Other Bacteria Important in Food Fermentations

In addition to the lactic acid bacteria, severalother genera are involved in fermented foods.In most cases, these bacteria are used for a sin-gular purpose, that is, they are involved in just


36 Microbiology and Technology of Fermented FoodsTa

ble

2.5

.Fe

rmen

tatio

n ch

arac

teris

tics

of L

acto

baci

llus1

%G

row

thFe

rmen

tati

on

of:

Arg

inin

eR

epre

sen

tati

ve s

trai

ns

G�

Cat

15°

CC

el2

Fru

Gal

Lac

Mal

Star

chSu

cH

ydro

lysi

s

Gro

up

IO

bli

gate

ho

mo

ferm

enta

tive

Lact

oba

cillu

s a

cidoph

ilu

s34

–37

��

��

��

nd

3�

�

Lact

oba

cillu

s del

bru

eckii

49–5

1�

�4

��

��

nd

��

Lact

oba

cillu

s h

elve

ticu

s38

–40

��

��

��

nd

��

Lact

oba

cillu

s a

myl

oph

ilu

s44

–46

��

��

��

��

nd

Lact

oba

cillu

s a

myl

ovoru

s40

–41

��

��

��

��

nd

Lact

oba

cillu

s cr

ispa

tus

35–3

8�

��

��

�n

d�

�

Lact

oba

cillu

s ga

sser

i33

–35

��

��

��

nd

��

Lact

oba

cillu

s je

nse

nii

35–3

7�

��

��

�n

d�

�

Gro

up

II

Fac

ult

ativ

e h

eter

ofe

men

tati

veLa

ctoba

cillu

s pa

raca

sei

46n

d�

��

��

nd

��

Lact

oba

cillu

s cu

rva

tus

43n

d�

��

��

nd

��

Lact

oba

cillu

s pla

nta

rum

45�

��

��

�n

d�

�

Lact

oba

cillu

s sa

kei

43n

d�

��

��

nd

��

Lact

oba

cillu

s ba

vari

cus

43n

d�

��

��

nd

��

Lact

oba

cillu

s h

om

oh

ioch

ii36

��

��

��

nd

��

Lact

oba

cillu

s co

ryn

iform

is36

��

��

��

nd

��

Lact

oba

cillu

s a

lim

enta

riu

s36

nd

��

��

�n

d�

�

Gro

up

III

Ob

liga

te h

eter

ofe

rmen

tati

veLa

ctoba

cillu

s fe

rmen

tum

53�

��

��

�n

d�

�

Lact

oba

cillu

s sa

nfr

an

cisc

ensi

s37

nd

��

��

�n

d�

�

Lact

oba

cillu

s re

ute

ri41

nd

��

��

�n

d�

�

Lact

oba

cillu

s bu

chn

eri

45n

d�

��

��

nd

��

Lact

oba

cillu

s bre

vis

45n

d�

��

��

nd

��

Lact

oba

cillu

s kim

chii

35�

��

w5

��

��

�

Lact

oba

cillu

s ke

firi

41n

d�

��

��

nd

��

Lact

oba

cillu

s div

erge

ns

34�

��

��

�n

d�

�

1A

dap

ted

fro

m B

erge

y’s

Ma

nu

al of

Syst

ema

tic

Ba

cter

iolo

gy,V

olu

me

2(1

986)

2C

el �

cellu

lose

;Fru

�fr

uct

ose

;Gal

�ga

lact

ose

;Lac

�la

cto

se;M

al �

mal

tose

;Su

c �

sucr

ose

3N

ot

det

erm

ined

4V

aria

ble

rea

ctio

n,d

epen

din

g o

n s

trai

n o

r su

bsp

ecie

s5W

eak

po

siti

ve r

eact

ion



one application and perform only one majorfunction.These non-lactic acid bacteria repre-sent several different genera and include bothGram positive and Gram negative bacteria.Their taxonomical position and general prop-erties are given below. How they are actuallyused and the particular functions they performin fermented foods are described in detail inthe relevant chapters.

Acetobacter, Gluconobacter and Gluconoacetobacter

The only Gram negative bacteria used in themanufacture of fermented foods are the aceticacid-producing rods belonging to the generaAcetobacter, Gluconobacter, and Gluconoace-tobacter (in the Proteobacteria phylum,FamilyAcetobacteraceae).These bacteria are obligateaerobes, with a respiratory-only metabolism.They make acetic acid via oxidation of ethanol;some species may also have the capacity to fur-ther oxidize acetic acid completely to CO2 andwater.The acetic acid bacteria are mesophilic,with optimum growth temperatures about25°C to 30°C.Although acetobacteria are acid-tolerant, their preferred growth pH is generallybetween 5.3 and 6.3. Some species of Aceto-bacteraceae produce surface film and pig-ments. Recent changes in the taxonomy ofthese bacteria has resulted in the transfer ofseveral important species from Acetobacter toGluconoacetobacter.

It is important to note that other bacteria,including species of Acetobacterium andClostridium, also make acetic acid as a primarymetabolic end-product, but these bacteria areobligate anaerobes and rely on reductive path-ways (i.e., reduction of CO2). In wine, beer,cider, and other ethanol-containing products,Acetobacter, Gluconobacter, and Gluconoace-tobacter can act as spoilage organisms. How-ever, they are also the bacteria used in the man-ufacture of vinegar.

Industrially, vinegar can be produced usingeither pure culture or natural fermentations. Ingeneral, Acetobacter aceti is considered to bethe “vinegar bacterium,” because it is the most

commonly used species in pure cultureprocesses and is also found in natural vinegarfermentations. Other species, however, are alsofrequently isolated from, or are used in vinegarfermentations, including Acetobacter orleanen-sis, Acetobacter pasteurianus subsp. pasteur-ianus, Gluconoacetobacter europaeus, andGluconoacetobacter xylinus.

Bacillus

Species of the genus Bacillus are ubiquitous infoods, serving either as benign contaminants, asspoilage organisms,or occasionally as the causeof food poisoning syndromes. Strains of Bacil-lus subtilis, however, are also used for the pro-duction of an Asian fermented soybean productcalled natto (Chapter 12). These natto strainsare closely related to the widely used wild-typelaboratory strain, B. subtilis Marburg 168, ex-cept that the former contain several biochemi-cal and physiological properties that are re-quired for the natto fermentation. Specifically,the natto strains produce a capsular polysac-charide material that is responsible for the vis-cous texture and sweet flavor of this product.

Bifidobacterium

It is arguable whether species of Bifidobac-terium should be considered “involved” infood fermentations. Although they have a fer-mentative metabolism, these bacteria are notused in the manufacture of any fermentedfood,nor are they even found in most raw foodmaterials. Rather, bifidobacteria are added tocertain foods, mostly milk and fermented dairyproducts, strictly for their probiotic functions.The intestinal tract is their primary habitat, andtheir elevated presence in the human gastroin-testinal tract is correlated with a reduced inci-dence of enteric infections and overall intesti-nal health.Bifidobacteria are now so frequentlyused as probiotic adjuncts in foods that theyhave become a commercially important prod-uct line for starter culture companies as ingre-dients in yogurt and other dairy culture formu-lations (Chapter 3).



There are more than twenty-five recognizedspecies of Bifidobacterium (Figure 2–7), al-though only about ten are ordinarily used commercially as probiotics.These include Bifi-dobacterium bifidum, Bifidobacterium ado-lescentis, Bifidobacterium breve, Bifidobac-

terium infantis, Bifidobacterium lactis, andBifidobacterium longum. For many years (un-til the 1970s), bifidobacteria were classified inthe genus Lactobacillus. It is now clear thatthey are phylogenetically distinct from the lac-tic acid bacteria (Figure 2.2), and are in an en-

Bifidobacterium adolescentis

Bifidobacterium dentium

Bifidobacterium angulatum

Bifidobacterium pseudocatenulatum

Bifidobacterium catenulatum

Bifidobacterium subtile

Bifidobacterium breve

Bifidobacterium pseudolongum

Bifidobacterium longum

Bifidobacterium pullorum

Bifidobacterium saeculare

Bifidobacterium gallinarum

Bifidobacterium bifidum

Bifidobacterium scardovii

Bifidobacterium minimum

Bifidobacterium asteroides

Bifidobacterium coryneforme

Bifidobacterium indicum

Bifidobacterium thermophilum

Bifidobacterium boum

Bifidobacterium magnum

Bifidobacterium cuniculi

Bifidobacterium gallicum

Bifidobacterium animalis

Bifidobacterium choerinum

0.01

Figure 2–7. Phylogeny of Bifidobacterium based on 16S rRNA sequence analysis.



tirely different phylum (Actinobacteria). Bifi-dobacteria are Gram positive, non-motile, non-sporing rods with a high G�C content (55% to67 mol%). Cells often occur in pairs with a V-or Y-like appearance. They are strictly anaero-bic and catalase negative, with a temperatureoptima between 37°C and 41°C and a pH op-tima (for growth initiation) between 6.5 and7.0. Bifidobacteria are nutritionally fastidiousand require vitamins and other nutrients forgrowth.Their ability to use a wide array of car-bohydrates, including non-digestible oligosac-charides that reach the colon, may provide se-lective advantages in the colonic environment(discussed in Chapter 4). Sugar metabolism oc-curs primarily via the well-studied “bifidum”fermentation pathway that yields acetic acidand lactic acid. Bifidobacteria (except for B.longum) rarely contain plasmids.

Brevibacterium

The genus Brevibacterium (phylum,Actinobac-teria) are described as non-motile, non-sporing,non-acid-fast, irregular-shaped organisms thatbelong to the “coryneform” group. Like othercoryneform bacteria, Brevibacterium sp. areGram positive rods, but both their staining pat-tern and shape can vary, depending on the ageand condition of the cells. They also are highG�C organisms (60 to 67 mol%). Brevibac-terium are strictly aerobic, catalase-positivemesophiles, with an optimum growth tempera-tures between 20°C and 35°C. Most species aresalt-tolerant (�10%) and able to grow over awide pH range.There are currently eighteen rec-ognized species; the phylogeny of twelve repre-sentative species is shown in Figure 2–8.

Several are of medical importance and havebeen considered as opportunistic pathogens.One species, Brevibacterium linens, is impor-tant in fermented foods, mainly because it is in-volved in the manufacture of bacterial, surface-ripened cheeses, such as Limburger andMuenster. In these products,B. linens producesa yellow-orange-red pigment on the cheese sur-face that gives these cheeses their characteris-tic appearance. Their ability to hydrolyze pro-

teins and metabolize amino acids,especially thesulfur-containing amino acids, also contributesto cheese ripening and flavor development in avariety of cheeses.

Kocuria, Micrococcus, and Staphylococcus

These closely related genera (Phylum Firmi-cutes) consist of Gram positive, catalase posi-tive, non-motile, non-sporeforming, aerobiccocci. Only a few species are relevant to fer-mented foods, and their involvement is limitedto the manufacture of fermented sausages (al-though some may show up as inadvertent con-taminants in cheese and other products). Themain species include Kocuria varians (for-merly Micrococcus varians), Micrococcus lu-teus, Staphylococcus xylosus, and Staphylo-coccus carnosus. In fermented meats, thesebacteria are not used for acid formation (be-cause they are non-fermentative), but rather toenhance flavor and color development. In par-ticular, they produce aroma compounds, hy-drolyze lipids and proteins, and reduce nitrateto nitrite.The latter property is important (oressential) in European- or traditional-style dry-cured sausages in which the curing agent con-tains nitrate salts, rather than the active, nitriteform used in conventional sausage manufac-ture.This function is discussed in more detailin Chapter 6.

Propionibacterium

The propionibacteria are in the phylum Acti-nobacteria. They are high G�C (53 to 68mol%), non-sporing, Gram positive, non-motilerods, often with a characteristic club-shapedappearance.They may also appear coccoidal orbranched. Propionibacteria are catalase posi-tive (or variable), anaerobic to aerotolerantmesophiles. They are neutraphilic and growslowly at pH 5.0 to 5.2 (the pH that exists inSwiss-type cheese).

In general, the propionibacteria are associ-ated with two quite different habitats, the hu-man skin and dairy products. The former



group contains the cutaneous species Propi-onibacterium acnes and Propionibacteriumavidum, the organisms that cause acne. Thedairy group consists of several species thatare important in food fermentations, due totheir use in the manufacture of Swiss-typecheeses. The most frequently used dairyspecies include Propionibacterium freuden-reichii subsp. shermanii, Propionibacteriumfreudenreichii subsp. freudenreichii, Propi-onibacterium acidopropionici, and Propi-onibacterium jensenii. They have a mol%G�C content of 66% to 68% and are distin-guished from other propionibacteria basedmainly on morphological and biochemical

characteristics, in that they are usually shorterrods and ferment fewer sugars.

The ability to ferment lactose varies amongstrains, but in the absence of lactose, dairystrains are quite capable of using lactate as a car-bon and energy source.Metabolic end-productsinclude propionic acid, acetic acid, and carbondioxide. Small amounts of Vitamin B12 may alsobe produced. During growth in a peptide-richmedium, such as cheese, propionibacteria re-lease the amino acid proline via peptide hydrol-ysis. They also are lipolytic, releasing free fattyacids from triglycerides. Despite their role as“eye-formers” in Swiss-type cheese, they are notparticularly prolific CO2-producers.

0.01

Brevibacterium otitidis

Brevibacterium mcbrellneri

Brevibacterium paucivorans

Brevibacterium avium

Brevibacterium iodinum

Brevibacterium permense

Brevibacterium aureum

Brevibacterium epidermidis

Brevibacterium antiquum

Brevibacterium linens

Brevibacterium casei

Brevibacterium celere

Figure 2–8. Phylogeny of Brevibacterium based on 16S rRNA sequence analysis.



Yeasts and Molds Used in theManufacture of Fermented Foods

From 80,000 to 1.5 million species of fungi arebelieved to exist, but, fortunately for this dis-cussion, relatively few are involved in food fer-mentations.As described previously, yeasts andmolds are in the Eukarya domain and belong tothe Kingdom Fungi. Of the phyla within thiskingdom, most of the fungi relevant to fer-mented foods are classified as Zygomycota, As-comycota, or as deuteromycetes.These groupscontain several important genera of yeasts, in-cluding Saccharomyces, Kluyveromyces, andZygosaccharomyces, as well as the mold gen-era Aspergillus, Penicillium, and Rhizopus.

Fungi, as eukaryotes, differ from prokaryotesin several major respects. In particular, fungihave a much more complex physical structureand they reproduce in an entirely different man-ner than bacteria.Although molds and yeasts areboth fungi, molds are multicellular and filamen-tous, whereas yeast are unicellular and non-filamentous. In addition, molds can produceasexual and sexual spores, while yeast produceonly sexual spores or are non-sporulating.

The most common asexual spores pro-duced by molds are the sporaniga or coni-diospores type and can be further divided asconidia, arthrospores, and sporangiospores.

The most common molds used in fermentedfoods, Penicillium and Aspergillus, produceconidia, and Rhizopus and Mucor produce spo-rangia. Yeasts, in addition to producing asco-spores, also reproduce asexually by budding, aprocess in which part of the cytoplasm bulgesfrom the cell and eventually separates as a newcell. Budding can be either multilateral or polar,and is characteristic of the species.

Saccharomyces

It can reasonably be argued that the yeasts belonging to the genus Saccharomyces areamong the most important of all organismsused in fermented foods, perhaps more sothan even the lactic acid bacteria.These yeastsare required, after all, for the production of

beer, wine, and spirits (not to mention bread),products that have a combined, world-wideeconomic impact in the trillions of dollars. Inaddition, the main species, Saccharomycescerevisiae, is widely used as a model organismin biology and genetics, and its physiologicaland biochemical properties, as well as itsgenome sequence, have been well studied. Atthe gene level,Saccharomyces and other yeastsare much more complicated than are the bacte-ria.They have sixteen chromosomes, they canbe diploidal or polyploidal (i.e., more than oneset of chromosomes), and the size of theirgenomes is several times larger (more than6,000 protein-encoding genes compared toabout 2,000 in most lactic acid bacteria).

The taxonomy and nomenclature of Saccha-romyces has been subject to rather regular andfrequent revisions for more than a century.Al-though S. cerevisiae has long been one of themajor species used in wine, brewing, andbread-making applications, the specific strainsinvolved in the manufacture of these productsare clearly different from each other and fromlaboratory strains of S. cerevisiae.The frequentchanges in nomenclature have made it evenmore difficult to keep track of the particularspecies associated with a given fermentation.

Several examples illustrate this point. Priorto the 1970s, Saccharomyces ellipsoideus wasrecognized as one of the primary wine yeastsand was accorded species status. It was thenreclassified as S. cerevisiae (and demoted to“synonym” status), although the earlier namecontinues to be used. Similarly, another groupof wine strains classified as Saccharomycesuvarum were reclassified as Saccharomycesbayanus (or sometimes listed as Saccha-romyces bayanus subsp. uvarum). Finally, formany years, the species used in the lager fer-mentation (where lager is one of the two stylesof beer) was classified as Saccharomyces carls-bergensis,which was also referred to as its syn-onym, Saccharomyces uvarum. Both speciesthen merged as S. uvarum, only to be reclassi-fied yet again, first as Saccharomyces bayanusand then as Saccharomyces pastorianus,which is now the valid species name for the



lager yeast. In the most recent volume (1998)of The Yeast, A Taxonomic Guide, there werefourteen accepted species of Saccharomyces.However, in a 2003 report (FEMS Yeast Re-search 4:223-245), only seven species wererecognized: S. bayanus, Saccharomyces cario-canus, S. cerevisiae, Saccharomyces kudri-avzevii, Saccharomyces mikatae, Saccha-romyces paradoxus, and S. pastorianus.

Distinguishing between species of Saccha-romyces is based primarily on morphological,physiological, and biochemical properties.These yeasts usually have a round spherical orovoid appearance, but they may be elongatedwith a pseudohyphae.The sugar fermentationpatterns and the assimilation of carbonsources are key factors for speciation (Table 2–6;also see Chapter 9 for beer yeast speciation).Other specific diagnostic tests include hyphaeformation, ascospore formation, resistance tocycloheximide, and growth temperatures. Sev-eral physiological traits vary among the Sac-charomyces and are useful not only for classifi-cation, but are important for strain selection.Some strains, for example, are very osmophilicand halotolerant, and can grow in foods con-taining high concentrations of carbohydrates(e.g., high sugar grapes) or salt (soy sauce).

Nowhere, however, does the species ofyeasts influence the fermentation as much asin brewing.As noted above, there are two gen-eral styles of beer,ales and lagers.Each requiresa specific yeast, S. cerevisiae for ales and S. pas-torianus for lagers.These yeasts vary in severalrespects (reviewed in Chapter 9), but mainlyon the basis of where they settle or flocculatein the beer. Hence, ale yeasts are referred to astop-fermenting yeasts because they tend toflocculate at the top of the fermentation ves-sel, and lager yeasts are referred to as bottom-fermenting because they settle at the bottomof the fermentor. It is important to note thathowever useful these phenotypic characteris-tics are for classification, they are not alwaysconsistent with species assignments based onthe sequences of 18S rRNA and other regions.

Ecologically,Saccharomyces and other yeastsare ubiquitous in foods. In some fermented Ta

ble

2.6

.D

istin

guis

hing

cha

ract

eris

tics

of S

acch

arom

yces

spec

ies1

Ferm

enta

tio

n o

f:A

ssim

ilati

on

of:

Gro

wth

on

Gal

2M

alM

elM

e–gl

uG

alM

alTr

eM

elIn

uR

ibG

tlM

e–gl

u10

% N

aCl/

5% G

lu

Spec

ies

S.baya

nu

s�

��

��

��

��

��

��

S.ca

rioca

nu

s�

��

��

��

�L3

��

�L

S.ce

revi

sia

e�

��

��

��

��

��

��

S.ku

dri

avz

evii

�s4

��

��

��

�L

��

�

S.m

ika

tae

��

��

��

LL

��

��

�

S.pa

radox

us

��

��

��

��

��

��

�

S.pa

stori

an

us

��

��

��

��

��

��

�

1A

dap

ted

fro

m N

aum

ov

et a

l.,20

00.

Int.

J.Sy

st.E

vol.

Mic

rob

io.5

0:19

31�

1942

2A

bbre

viat

ion

s:G

al.g

alac

tose

;Mal

,mal

tose

;Mel

,mel

ibio

se;M

e–gl

u,m

eth

yl g

luco

sid

e;Tr

e,tr

ehal

ose

;In

u,i

nu

lin;R

ib,r

ibit

ol;

Gtl

,gal

acti

tol;

Glu

,glu

cose

3La

ten

t (d

elay

ed p

osi

tive

)4Sl

ow



foods, such as wine, their natural presence ongrapes and equipment is sufficient to initiate afermentation.Several different yeasts may be in-volved in “natural” or “spontaneous” wine fer-mentations, usually in a successive manner,where one species is dominant for a time, thengives way to others (Chapter 10). However, inmost modern wine fermentations, as well asbeer and bread fermentations, starter cultureyeasts are used, selected on the basis of theirphysiological and biochemical properties.Yeasts are also important in spoilage of fer-mented (and non-fermented) foods. For exam-ple, one of the serious defects in sauerkraut iscaused by the pink pigment-producing yeastRhodotorula.

Penicillium and Aspergillus

As previously mentioned, these molds areamong the most common and widespread infoods. In the older literature, they were oftenreferred to as “Fungi Imperfecti,”due to the ab-sence of a sexual stage in their life cycle. Peni-cillium and Aspergillus are mainly of concerndue to their role in food spoilage and as poten-tial producers of mycotoxins; however, somespecies of Penicillium and Aspergillus are alsoused to produce fermented foods. In fact, oneof the most famous of all organisms used in fer-mented foods is Penicillium roqueforti. Thismold gives Roquefort and other blue-veinedcheeses their characteristic color. It is alsolargely responsible for the flavor and aromaproperties of these cheeses. A related whitemold species, Penicillium camemberti, isequally important (and no less famous), due toits involvement in the manufacture of Camem-bert and Brie cheeses.

Although the role of Aspergillus in fermentedfoods manufacture is somewhat less prominent(at least in foods popular in Western cultures),species of Aspergillus are involved in some ofthe world’s most widely-consumed fermentedfoods. Specifically, two species, Aspergillusoryzae and Aspergillus sojae, are used in themanufacture of soy sauces and soy pastes andsake and other rice wines, so-called Oriental or

Asian fermented foods that are consumed liter-ally by billions of people (Chapter 12).

There are many different species of Penicil-lium and Aspergillus. As with the yeasts, speci-ation is based on morphology, structure, andspore type. Biochemical and physiologicalproperties of the fungi, however, are generallyless diagnostic. These fungi are heterotrophicand saprophytic, and are usually very good atbreaking down complex macromolecules (e.g.,protein,polysaccharides, lipids) via secretion ofproteinases, amylases, and lipases.The hydroly-sis products can then be used as substrates forgrowth.

Despite their readily observed structural dif-ferences, Penicillium and Aspergillus do sharesome common features. Both are filamentousand produce conidiospores, and are probablyrelated. Typical spore-bearing structures ofPenicillium and Aspergillus are shown in Fig-ure 2–9. Finally, Rhizopus oligosporus speciesare used to produce a fermented soybeanproduct known as tempeh that is popular inIndonesia and other Southeast Asian countries(Chapter 12).Typical morphological structuresof Rhizopus are shown in Figure 2–9.

Fermentation and Metabolism Basics

If one looks up fermentation in a biochemistrytextbook, the definition that appears is usuallysomething like this:“energy-yielding reactionsin which an organic molecule is the electronacceptor . . . .” Thus, in the lactic acid fermen-tation, pyruvic acid that is generated by the glycolytic pathway serves as the electron ac-ceptor, forming lactic acid. Likewise, in theethanolic pathway, acetaldehyde, formed bydecarboxylation of pyruvate, is the electron re-cipient (forming ethanol). So although this def-inition certainly is true for many of the fermen-tations that occur in foods, it is rather narrow.

For several fermented foods, important end-products are produced via non-fermentativepathways (as classically defined). For example,the malolactic fermentation that is very impor-tant in wine making is really just a decarboxyla-tion reaction that does not conform, strictly



speaking,to the definition stated above.Neitherdoes this definition apply in the manufacture oftempeh and other fungal-fermented foods, inwhich soy proteins and polysaccharides are de-graded and metabolized by fungi, but withoutproduction of glycolytic end-products. There-fore, in this text, the term fermentation will beused in a broader sense, accounting for themany metabolic processes that occur duringthe course of a food fermentation.

From the point of view of the microorgan-isms (which care even less about definitions),fermentation is merely the means by whichthey obtain energy. Microorganisms, after all,need energy to perform work (e.g., nutrienttransport and biosynthesis), to maintain chemi-cal and physical homeostasis (e.g., ionic and os-motic), and to grow and replicate. For the mostpart, the energy generated by fermentation is inthe form of ATP, and usually it is produced viametabolism of sugars (although, exceptions forboth of these claims exist, as suggested above).Whether sugar metabolism by microorganismsalso results in the conversion of milk into yo-

gurt or juice into wine is of no concern to theorganism. Of course, the same metabolic path-way may also lead to formation of end-productsthat cause milk or juice to become spoiled,which also matters little to the offending organ-ism. However, these outcomes do matter to yo-gurt and wine manufacturers.

Knowing how to control and manipulate mi-croorganisms and their metabolic activities canmean the difference between a pleasant-tastingcontainer of yogurt or soured milk that is tosseddown the drain, between success and failure,and between profit and loss. Therefore, under-standing the biochemical basis for metabolismof sugars, as well as other substrates, is essentialfor consistent production of fermented foodswith the expected biological, physical, chemi-cal,nutritional, and sensory characteristics.

Sugar Metabolism

The biochemical means by which microorgan-isms metabolize sugars was first studied by Pas-teur, Buchner, Schwann, and other early micro-

Figure 2–9. Principle fungi used in food fermentations. Shown schematically are Penicillium (A), Aspergillus(B), and Rhizopus (C). The spore structures are indicated by arrows (conidia in Penicillium and Aspergillus andsporangiospores in Rhizopus). Adapted from Bullerman, 1993.



biologists and biochemists. That food-relatedsugar fermentations, in particular, had attractedthe attention of these scientists was actuallyquite reasonable. After all, cultured milk prod-ucts,beer,and wine were industrially-importantproducts whose manufacture depended on fermentation of sugars.Thus, identifying the mi-croorganisms and pathways was essential tounderstanding how the manufacture of theseproducts could be improved and controlledand spoilage prevented. Metabolism and trans-port of sugars by lactic acid bacteria andethanol-producing yeasts will be reviewed inthe sections to follow.

Homofermentation

Lactic acid bacteria are obligate fermentors,and cannot obtain energy by oxidative or res-piratory processes (with the exception notedpreviously; Box 2–1). Technically, the precur-sor-product exchange systems, described be-low, provide an alternate way for these organ-isms to earn ATP “credits” by conserving theenergy that would ordinarily be used to per-form metabolic work. However, the substratelevel phosphorylation reactions that occurduring fermentation are by far the majormeans by which these cells make ATP. For ho-mofermentative lactic acid bacteria, hexosesare metabolized via the enzymes of the gly-colytic Embden-Meyerhoff pathway.

One of the key enzymes of this pathway isaldolase, which commits the sugar to the path-way by splitting fructose-1,6-diphosphate intothe two triose phosphates that eventuallyserve as substrates for ATP-generating reac-tions. The Embden-Meyerhoff pathway yieldstwo moles of pyruvate and two moles of ATPper mole of hexose (Figure 2–10). The pyru-vate is then reduced to L- or D-lactate by theenzyme, lactate dehydrogenase.More than 90%of the substrate is converted to lactic acid dur-ing homofermentative metabolism.

Importantly, the NADH formed during theglyceraldehyde-3-phosphate dehydrogenase re-action must be re-oxidized by lactate dehydro-genase, so that the [NADH]/[NAD�] balance is

maintained. Homofermentative lactic acid bac-teria include Lactococcus lactis, Streptococcusthermophilus, Lactobacillus helveticus, and L. delbrueckii subsp. bulgaricus (used as dairystarter organisms); Pediococcus sp. (used insausage cultures); and Tetragenococcus (usedin soy sauces).

Heterofermentation

Heterofermentative lactic acid bacteria me-tabolize hexoses via the phosphoketolasepathway (Figure 2–11). In obligate heterofer-mentative bacteria, aldolase is absent, and in-stead the enzyme phosphoketolase is present.Approximately equimolar amounts of lactate,acetate, ethanol and CO2 are produced, alongwith only one mole of ATP per hexose. Oxi-dation of NADH and maintenance of the[NADH]/[NAD�] balance occurs via the tworeductive reactions catalyzed by acetaldehydedehydrogenase and alcohol dehydrogenase.Many of the lactic acid bacteria used in foodfermentations are heterofermentative. In-cluded are L. mesenteroides subsp. cremorisand Leuconostoc lactis (used in dairy fermen-tations), L. mesenteroides subsp. mesen-teroides and Leuconostoc kimchii (used infermented vegetables), O. oeni (used in winefermentations), and Lactobacillus sanfrancis-censis (used in sourdough bread).

In general, the product yields for both path-ways may vary during actual fermentationprocesses,and depend on the type and concen-tration of substrate, the growth temperatureand atmospheric conditions, and the growthphase of the cells. Certainly some of the carbo-hydrate carbon is incorporated into cell mass.However, it has also been shown that undercertain conditions, such as sugar-limitation oraerobiosis,homofermentative lactococci can di-vert some of the pyruvate away from lactateand toward other alternative end-products,such as acetate and CO2.This so-called hetero-lactic fermentation may provide cells with addi-tional ATP or serve as a way to deal with excesspyruvate (Box 2–3).



ATP

ADP

ATP

ADP

Pi

H2O

NAD

NADH +H

glucose

glucose-6-phosphate

fructose-6-phosphate

fructose-1,6-diphosphate

hexokinase

phosphoglucose isomerase

phosphofructokinase

aldolase

dihydroxyacetone phosphateglyceraldehyde-3-phosphate

lactic acid

pyruvic acid

phosphoenolpyruvate

2-phosphoglycerate

3-phosphoglycerate

1,3-diphosphoglycerate

ADP

ATP

ADP

ATP

3-phosphoglycerate kinase

glyceraldehyde-3-phosphate dehydrogenase

phosphoglyceromutase

enolase

pyruvate kinase

lactate dehydrogenaseNADH +H

NAD

Figure 2–10. The Embden-Meyerhoff pathway used by homofermentative lactic acid bacteria. Thedashed line indicated the NAD/NADH oxidation-reduction part of the pathway.



hexokinaseATP

ADP

6-phosphogluconate

P

NAD

NADH + H

NAD

NADH + H

glucose-6-phosphatedehydrogenase

6-phosphogluconatedehydrogenase

glucose

glucose-6-phosphate

CO2

ribulose-5-phosphate

xyulose-5-phosphate

glyceraldehyde-3-phosphate

lactic acid

phosphoketolase

CoA

Pi

NADH + H

NAD

NADH + H

NAD

acetaldehydedehydrogenase

alcoholdehydrogenase

acetyl phosphate

acetyl CoA

acetaldehyde

ethanol

phospho-transacetylase

ribulose-5-phosphate-epimerase

Figure 2–11. The phosphoketolase pathway used by heterofermentative lacticacid bacteria.



Sugar Transport by Lactic Acid Bacteria

Although the catabolic pathways used by lacticacid bacteria to ferment sugars constitute a ma-jor part of the overall metabolic process, thefirst step (and perhaps an even more importantone) involves the transport of the substrateacross the cytoplasmic membrane.Transport isimportant for several reasons, not the least ofwhich is that the cell membrane is imperme-able to polar solutes (e.g., sugars, amino acids,

peptides). In the absence of transport systems,these solutes would be unable to transverse themembrane and gain entry into the cell. Second,the cell devotes a considerable amount of its total energy resources to support active trans-port. Third, some sugars are phosphorylatedduring the transport event,which then dictatesthe catabolic pathway used by that organism.Finally, transport systems may serve a regula-tory role, influencing expression and activity ofother transport systems.

Box 2–3. The Heterolactic Fermentation: Dealing with Pyruvate

Lactic acid bacteria, as previously noted, are either homofermentative, heterofermentative, orfacultative heterofermentative (where both pathways are present). However, even obligate ho-mofermentative strains have the potential to produce acetic acid,ethanol,acetoin,CO2 and end-products other than lactic acid.These alternative fermentation end-products are formed, how-ever, only under conditions in which pyruvate concentrations are elevated.

Such a scenario occurs when the rate of intracellular pyruvate formation exceeds the rate atwhich pyruvate is reduced to lactate (via lactate dehydrogenase).The pyruvate can arise fromsugars (see below), but also from amino acids. In either case, the excess pyruvate must be re-moved because it could otherwise become toxic.Moreover,when excess pyruvate is generatedas a result of sugar metabolism, the cells must also have a means for re-oxidizing the NADHformed from glycolysis (upstream in the pathway). Several different pathways in lactic acid bac-teria appear to serve this function (Figure 1; Cocaign-Bousquet et al., 1996; Garrigues et al.,1997). In addition, at least one of these alternative pathways include a substrate level phospho-rylation reaction and, therefore, provides the cells with additional ATP.

lactate

a-acetolactate

acetoin

lactate dehydrogenase

pyruvate dehydrogenase pyruvate-formate lyase

a-acetolactatesynthase

pyruvate

2,3-butanediol

acetaldehyde

formateacetate

ethanolacetaldehyde

carbon dioxideacetate

ethanol

Figure 1. Heterolactic end products from pyruvate metabolism.



The phosphoenolpyruvate-dependentphosphotransferase system

In general, several different systems are usedby lactic acid bacteria to transport carbohy-drates, depending on the species and the spe-cific sugar. The phosphoenolpyruvate (PEP)-dependent phosphotransferase system (PTS) isused by most mesophilic, homofermentativelactic acid bacteria, including lactococci andpediococci for transport of lactose and glu-cose. In contrast, other lactic acid bacteriatransport sugars via symport-type or ATP-dependent systems (described below). An-other group of transporters are the precursor-

product exchange systems that provide an al-together different (and more efficient) meansfor transporting sugars (Box 2–4).

The PTS consists of both cytoplasmic andmembrane-associated proteins that sequentiallytransfer a high-energy phosphate group fromPEP,the initial donor, to the incoming sugar (Fig-ure 2–12).The phosphorylation step coincideswith the vectorial movement of the sugar sub-strate across the membrane, resulting in the in-tracellular accumulation of a sugar-phosphate.Two of the cytoplasmic proteins of the PTS, En-zyme I and HPr (for histidine-containing pro-tein), are non-specific and serve all of the sugar

What are the conditions or environments that result in pyruvate accumulation and inductionof alternate pathways? There are several possible situations where these reactions occur. First,glycolysis is subject to several levels of regulation, such that when fermentation substrates arelimiting, the glycolytic flux tends to be diminished (Axelsson, 2004). Specifically, when the con-centration of fructose-1,6-diphosphate (a glycolytic regulator that forms early during glycoly-sis), is low, the activity of lactate dehydrogenase (an allosteric enzyme that occurs at the end ofglycolysis) is reduced.Thus, pyruvate accumulates.The glycolytic flux also may be decreasedduring growth on less preferred carbon sources, such as galactose, again resulting in excesspyruvate. At the same time that the lactate dehydrogenase activity is decreased, the enzyme,pyruvate-formate lyase, is activated. This enzyme splits pyruvate to form formate and acetylCoA. The latter is subsequently reduced to ethanol or phosphorylated to acetyl phosphate(both reactions releasing CoA). Importantly, the acetyl phosphate can be used as part of a sub-strate level phosphorylation reaction (via acetate kinase), resulting in the formation of an ATP.

Under aerobic environments,however,pyruvate-formate lyase is inactive,and other pathwaysbecome active. In the pyruvate dehydrogenase pathway, for example, pyruvate is decarboxy-lated by pyruvate dehydrogenase,and acetate and CO2 are formed. NADH that would normallyreduce pyruvate also is oxidized directly by molecular oxygen when the environment is aero-bic, rendering it unavailable for the lactate dehydrogenase reaction. Finally, excess pyruvate canbe diverted to end-products that may have a more functional role in certain fermented dairyproducts. Specifically, pyruvate can serve as a substrate for �-acetolactate synthase to form �-acetolactate.The �-acetolactate may then be further oxidized to form diacetyl, which has desir-able aroma properties.

References

Axelsson,L.2004.Lactic acid bacteria:classification and physiology,pp.1—66. In Salminen,S.,A.von Write,and A.Ouwehand.Lactic Acid Bacteria Microbiological and Functional Aspects Third Edition. MarcelDekker, Inc. New York, New York.

Cocaign-Bousquet,M.,C.Garrigues,P.Loubière, and N.Lindley.1996.Physiology of pyruvate metabolism inLactococcus lactis.Antonie Van Leeuwenhoek 70:253–267.

Garrigues, C., P. Loubière, N. Lindley, and M. Cocaign-Bousquet. 1997. Control of shift from homolactic acidto mixed-acid fermentation in Lactooccus lactis: predominant role of the NADH/NAD� ratio. J. Bacte-riol. 179:5282–5287.

Box 2–3. The Heterolactic Fermentation: Dealing with Pyruvate (Continued)



Box 2–4. Something For Nothing, or How Lactic Acid Bacteria Conserve Energy Via “Precursor-product” Exchange Systems

Transport of nutrients from the environmental medium across the cytoplasmic membrane andinto the cytoplasm is one of the most important of all microbial processes. For many microor-ganisms, including lactic acid bacteria, nearly 20% of the genome is devoted to transport func-tions, and cell membranes are literally crammed with a hundred or more transport permeases,translocators, and accessory proteins (Lorca et al., 2005). The function of transporters, ofcourse, is to provide a means for the cell to selectively permit solutes to move,back and forth asthe case may be, across a generally impermeable membrane.

For most nutrients (and other transport substrates), however, transport is not cheap, in thatenergy (e.g.,ATP) is usually required to perform “vectorial” work (i.e., to move molecules fromone side of a membrane to another).This is because nutrient transport generally occurs againsta concentration gradient,meaning that the concentration of the substrate is usually much lowerin the outside medium than it is on the inside (i.e., within the cytoplasm).Although passive orfacilitated diffusion, where the concentration gradient (higher outside than inside) drives trans-port occurs occasionally, this is not the normal situation, and instead some sort of active,energy-requiring process is required.

The metabolic cost of transport is especially high for lactic acid bacteria, given the limitedand generally inefficient means by which these bacteria make energy.After all, glycolysis gener-ates only two molecules of ATP per molecule of glucose or hexose fermented. If ATP or itsequivalent (e.g.,phosphoenolpyruvate or the proton motive force) is required to “move”mono-and disaccharides across the cytoplasmic membrane, the net energy gain by the cell may be re-duced by as much as 50%. On the other hand, if the driving force for solute transport does notdepend on an energy source, then the cell can conserve that un-spent energy or use it for otherreactions.

By what means can cells transport solutes without having to spend energy? In many lacticacid bacteria, metabolism of substrate (or precursor) molecules results in a large amount ofmetabolic products that must be excreted from the cell. In some cases, these products are notfurther metabolized; in other cases, they may even be toxic if allowed to accumulate intracellu-larly. In any event, a rather large concentration gradient is formed, where the inside concentra-tion of “product” is much higher than the outside concentration. Efflux of “product”molecules,therefore, is driven by the concentration gradient in a downhill manner.The carrier for efflux inthese situations,however,actually has affinity not just for the product,but also for the substrate.Moreover, it operates in opposing directions, such that product excretion (via the downhill con-centration gradient) can drive uphill uptake of the precursor substrate (against the concentra-tion gradient).These so-called precursor-product exchange systems, therefore, represent a novelmeans by which cells can accumulate nutrients and metabolic substrates without having toconsume much-needed sources of energy.

Examples of Precursor-product Exchange Systems in Lactic Acid Bacteria

Precursor-product exchange systems are widely distributed in lactic acid bacteria, and are usedto transport fermentation substrates, amino acids, and organic acids (Konings, 2002).These sys-tems can be electroneutral, without a net change in the electric charge across the cell mem-brane,or electrogenic,where a charge is generated across the membrane.For example,a neutralsugar precursor (lactose) exchanged for a neutral sugar product (galactose) is electroneutral, asis the exchange of the amino acid arginine for ornithine (Figure 1, panels A and C). However, adi-anionic precursor (citrate) exchanged for a mono-anionic product (lactate) is electrogenic,since it results in an increase of the transmembrane electric charge (Figure 1, panel B).



References

Konings, W.N. 2002. The cell membrane and the struggle for life of lactic acid bacteria. Antonie vanLeeuwenhoek 82:3–27.

Lorca, G.L., R.D. Barabote,V. Zlotopolski, C.Tran, B.Winnen, R.N. Hvorup,A.J. Stonestrom, E. Nguyen, L.-W.Huang, D. Kimpton, and M.H. Saier. 2006.Transport capabilities of eleven Gram-positive bacteria: com-parative genomic analyses. In press.

Box 2–4. Something For Nothing, or How Lactic Acid Bacteria Conserve Energy Via “Precursor-product” Exchange Systems (Continued)

esotcalag

esotcalag

esotcal

esotcal

ScaL

esadisotcalagb-

esoculg

dica citcal

yawhtap ME

tuo

ni

H+

H+

H+

H+

A

etartic -2

OC 2

H+

etavuryp -

etartic -2

lytecaid

PtiC

tuo

ni

etatcal -

etatecaolaxo -2

etatcal - esoculg

etateca

B

eninigra

eninigra

DcrA

tuo

ni

enihtinro

enillurtic

HN 3 OC + 2

HN 3

enihtinro

lyomabrac �P

PDA

PTA

C

iP

Figure 1. Precursor-product exchange systems in lactic acid bacteria.Three examples are shown. In Streptococcus thermophilus (and somestrains of Lactobacillus), lactose is transported by the LacS system viaproton-symport (panel A). Intracellular hydrolysis releases free galac-tose, which is not metabolized, but is instead secreted. The efflux reac-tion, which appears to be favored in most strains, then drives uptake oflactose. Citrate-fermenting lactic acid bacteria transport citrate via theCitP citrate permease (panel B). Metabolism of citrate, or glucose whenit is also available, generates lactic acid, whose efflux through CitP candrive citrate uptake. This reaction is electrogenic, in that a proton isconsumed within the cytoplasm, increasing the proton motive forceacross the membrane. Lactic acid bacteria capable of metabolizing theamino acid arginine do so via the arginine deiminase pathway (panelC). The products of this pathway are NH3, CO2, ornithine, and ATP. Or-nithine can be effluxed via the arginine transporter (ArcD), in exchangefor arginine. The cell, therefore, gains energy from arginine (one moleof ATP per arginine metabolized), without having to spend transportenergy.



phosphotransferase systems present in the cell.The other PTS components, however, are sub-strate-specific. The latter proteins form an En-zyme II complex, which most often consists ofthree protein domains: EIIA, EIIB, and EIIC.These EII domains may be present as individual

proteins or are combined or fused as one or twoproteins. For example, the lactose PTS EnzymeII complex in Lactococcus lactis contains threelactose-specific protein domains that exist inthe form of two proteins, i.e.,EIIAlac and EIIBClac

(where “lac” denotes lactose-specificity). As for

Figure 2–12. The phosphotransferase system (PTS) in Gram positive bacteria. As shown in panel A, thePTS cascade is initiated by the cytoplasmic proteins Enzyme I (EI) and HPr. Phosphorylated HPr (HPr~P) thentransfers the phosphoryl group (obtained originally from PEP) to the substrate-specific Enzyme II complex.The latter consists of several proteins or domains, shown here as EIIA, EIIB, and EIIC. However, dependingon the organism and the substrate, EII complexes may be organized differently, for example, as EIIA and EIIBC or as EIIABC. Regulation of the PTS is mediated, in part, via the phosphorylation state of HPr (panel B).Phosphorylation by HPr kinase results in formation of HPr[Ser~P], which, along with CcpA and fructosediphosphate (FDP), form a dimeric complex that recognizes and binds to CRE sites and prevents transcrip-tion of catabolic genes.



all of the EII proteins, the EIIB and EIIC domainsare membrane-associated.

In the case of hexoses, the product of thePTS is hexose-6-phosphate, which can thenfeed directly into the glycolytic pathway.ThePTS has the advantage, therefore, of sparingthe cell of the ATP that ordinarily would berequired to phosphorylate the free sugar.When lactose is the substrate, the product islactose-phosphate (or more specifically, glu-cose-�-1,4-galactosyl-6-phosphate). Lactose-phosphate is not hydrolyzed by the enzyme�-galactosidase, which is widespread in themicrobial world, but rather by phospho-�-galactosidase. The products of this reactionare glucose and galactose-6-phosphate. Theglucose is subsequently phosphorylated byhexokinase, and the glucose-6-phosphate thatforms feeds into the glycolytic pathway, as

described earlier. The galactose-6-phosphate,in contrast, is metabolized by the tagatosepathway (Figure 2–13), eventually leading tothe formation of the same triose phosphates,glyceraldehyde-3-phosphate and dihydroxy-acetone phosphate, that form during glycoly-sis. It is perhaps not unexpected that thestructural (and regulatory) genes coding forlactose transport and hydrolysis are locatedon the same operon as the galactose/tagatosegenes (Figure 2–14).

Symport and ABC Transport Systemsin Lactic Acid Bacteria

Although the PTS is widely distributed amonglactic acid bacteria, several species rely onother active transport systems to transport sug-ars.The latter include symport systems, driven

lactose-6-P

glucosegalactose-6-P

lactic acid

lactic acid

tagatose-6-P

tagatose-1,6-di-P

glyceraldehyde-3- phosphate

dihydroxyacetone phosphate

EM

ATP

ADP

phospho-b-galactosidase

galactose-6-Pisomerase

tagatose-6-Pkinase

aldolase

EM

Figure 2–13. Tagatose pathway in lactococci. Galactose-6-phos-phate is formed from hydrolysis of lactose-phosphate, the product ofthe lactose PTS. Isomerization and phosphorylation form tagatose-1,6-diphosphate, which is split by an aldolase, yielding the triosephosphates that feed into the EM pathway.



by ion gradients, and ATP-binding cassette(ABC) systems, fueled by ATP (Figure 2–15).Moreover, it is often the case that an organismuses a PTS for one sugar and a symport or ABCsystem for another.According to the TransportClassification system (www.tcdb.org), symportsystem transporters belong to a class of sec-ondary carriers within the Major Facilitator Superfamily.

These symport systems consist of a mem-brane permease that has binding sites for boththe substrate and a coupling ion, usually pro-tons.Transport of the substrate is driven by theion gradient across the membrane. The mostcommon ion gradient in bacteria is the protongradient (called the proton motive force or

PMF), although there are also symport systemsdriven by sodium ion gradients (such as themelibiose system that exists in Lactobacilluscasei).

The PMF is comprised of the sum of twocomponents: (1) the charge difference (��)across the membrane,where the outside is posi-tive and the inside is negative;and (2) the chem-ical difference (�pH) across the membrane,where the proton concentration is high on theoutside and low inside. The positively chargedprotons then flow “down” this gradient (i.e., to-ward the inside) in symport with the “uphill” in-tracellular accumulation of the solute. In lacticacid bacteria, symport systems exist for severalsugars (based on biochemical and genetic evi-

lacA lacB lacC lacD lacF lacElacR lacG

568105 586310141

galactose-6-P isomerase

171

tagatose-6-P kinase

tagatose-1,6-diP aldolase

326

phospho-b -galactosidase

lacX

468

EIIBC laclacEIIA

Figure 2–14. The lac operon in lactococci. The operon consists of four structural genes (lacABCD) codingfor enzymes of the tagatose pathway, two genes, lacFE, coding for lactose-specific PTS proteins, and the lacGgene coding for phospho-�-galactosidase. A divergently transcribed lacR gene codes for a repressor protein.The function of lacX is not known. Promoter sites and directions are shown by the arrows, and potential tran-scriptional terminators are shown as hairpin loops. The number of amino acid residues for each protein isgiven. Adapted from de Vos et al., 1990.

permease

out

in

H+

A

solute

H+soluteATP ADP

H+

H+

out

in

B

solute

solute

ATP ADP

NBD1 NBD2

BP

IMPI MP

++++

-- --

Figure 2–15. Solute transport in lactic acid bacteria. In Panel A, a proton symport system, fueled by theproton gradient, drives solute uptake. The proton gradient is formed via the proton translocating ATPase. Inpanel B, ATP hydrolysis drives transport directly via nucleotide binding domains (NBD). The solute is first cap-tured by a binding protein (BP) which delivers it to the integral membrane proteins (IMP).



dence), including lactose (in Lactobacillus bre-vis, L. delbrueckii, L. acidophilus), galactose (L. lactis), raffinose (P. pentosaceus), melibiose(L. lactis), and xylose (Lactobacillus pentosus).Although generation of the PMF is mediated viathe proton-translocating ATPase (also called theF0F1-ATPase),which requires ATP, lactic acid bac-teria may also make a PMF by alternative routesthat spare the cell of ATP (Box 2–4).

For those organisms that transport lactoseby a symport system, the intracellular productis lactose (chemically unaltered, in contrast tothe lactose PTS, where the product is lactose-phosphate). Intracellular hydrolysis of the freelactose then occurs via �-galactosidase, yield-ing glucose and galactose.The former is phos-phorylated by hexokinase to form glucose-6-phosphate, which feeds directly into theEmbden-Meyerhoff pathway.

To convert galactose into glucose-6-phos-phate requires the presence of the Leloir path-way (Figure 2–16), a three enzyme pathwaywhose expression is subject to strong negativeregulation in some lactic acid bacteria. In par-ticular, S. thermophilus and Lb. delbrueckisubsp. bulgaricus transport and hydrolyze lac-tose by a PMF and a �-galactosidase, respec-tively,but ferment only the glucose and not the

galactose moiety.Although these bacteria havethe genes encoding for the Leloir pathway en-zymes, the genes are poorly transcribed or ex-pressed, due to mutations within the promoterregion of the operon.As a result, the intracellu-lar galactose is largely unfermented and is ex-creted via the lactose transport system (LacS in S. thermophilus), a seemingly wastefulprocess.What makes this process especially in-teresting, however, is that LacS can catalyze alactose:galactose exchange reaction, such thatgalactose efflux drives lactose uptake (Box 2–4).Thus, what might initially appear to be aninefficient and wasteful act (secreting a per-fectly good energy source into the medium), is,instead,an efficient,energy-saving means of ex-changing a readily fermentable sugar for onethat is slowly metabolized, if at all.

The other major group of transport systemsused by lactic acid bacteria for accumulatingsugars are the ABC transport systems. Thesesystems consist of several proteins or proteindomains (Figure 2–15). Two of these proteinsor domains are membrane-spanning (i.e., inte-grated within the membrane) and serve asporters. Two intracellular proteins/domainsbind and hydrolyze ATP. An additional proteinis extracellular, but is tethered to the cell

glucose-1-P

UDP glu

UDP gal

glucose-6-P

galactose-1-Puridyl transferase

phosphoglucomutase

UDP galactoseepimerase

galactosegalactokinase

galactose-1-P

ATP ADP

homofermentative pathway

heterofermentative pathway

Figure 2–16. The Leloir pathway in lactic acid bacteria.



surface and has substrate-binding activity. Theenergy released from ATP hydrolysis drivestransport. In general,ABC systems are used totransport amino acids, peptides, and osmopro-tectants. However, several ABC systems are in-volved in sugar transport, including the mal-tose ABC systems in L. plantarum and theoligosaccharide-multiple sugar transport sys-tem in L.acidophilus.

Regulation of Transport Systems

As noted above, regulation of sugar metabolismin lactic acid bacteria is intimately connected tothe transport machinery. In L. lactis, for exam-ple, lactose fermentation is regulated biochemi-cally and genetically at the level of transcriptionthrough a signal transduction cascade mediatedvia the lactose transport system. Biochemicalregulation occurs in two major ways. First, sev-eral glycolytic enzymes are allosteric, and theiractivities are subject to the intracellular concen-tration of specific glycolytic metabolites.Whenlactose is plentiful, the high intracellular con-centration of fructose-1,6-diphosphate (and lowlevel of inorganic phosphate) activates pyruvatekinase, the enzyme that catalyzes substrate levelphosphorylation and ATP synthesis, and theNADH-dependent lactate dehydrogenase,whichoxidizes NADH�. In contrast, when lactose islimiting, pyruvate kinase activity decreases,causing a metabolic bottleneck in glycolysis.This cessation of glycolysis results in the accu-mulation of PEP, which primes the cell for PTS-mediated transport when sugars ultimately become available.

A second biochemical mechanism for regu-lating lactose metabolism involves the PTStransport proteins,and the cytoplasmic compo-nent HPr, in particular (Figure 2–12). As out-lined above, the PTS cascade starts when HPr isphosphorylated by Enzyme I. Ordinarily, phos-phorylation occurs at the histidine-15 [His-15]residue of HPr.However,an ATP-dependent HPrkinase can also phosphorylate HPr, but at a ser-ine residue [Ser-46]. When HPr is in the latterphosphorylation state (i.e., HPr [Ser-46-P]),

phosphorylation at the His-15 site does not oc-cur and the PTS cannot function. At the sametime, HPr [Ser-46-P] can bind directly to the in-terior side of other sugar permeases and pre-vent transport of those sugars,a process knownas inducer exclusion.

It is interesting to note that HPr kinase is ac-tivated by fructose-1,6-diphosphate, whichreaches high concentrations only when thesubstrate is plentiful. That is, HPr is evidentlyconverted to HPr [Ser-46-P] at just the timewhen one would expect Hpr to be in a non-phosphorylated state (i.e., in the active trans-port mode). Thus, it seems counterintuitivethat the cells would slow down transportwhen substrates are so readily available.

Why, one might ask, would the cell down-regulate the activity of the transport system intimes of plenty? One possible answer is thatthe cell is simply modulating or controlling itsappetite by not transporting more substratethan it can reasonably consume.This hypothe-sis is supported by the observation that HPrnot only exerts biochemical control on trans-port,but that HPr [Ser-46-P] also regulates tran-scription of sugar transport genes (Figure 2–12, Panel B). Specifically, HPr [Ser-46-P] inter-acts with a DNA-binding, trans-acting proteincalled CcpA (Catabolite control protein A).TheHPr [Ser-46-P]-CcpA complex (along with fruc-tose-1,6-diphosphate), binds to 14 base pairDNA regions called Catabolite Responsive Ele-ments (CRE). These CRE sites are located justupstream of the transcription start sites of cer-tain catabolic genes. When these CRE regionsare occupied by the HPr [Ser-46-P]-CcpA com-plex, transcription by RNA polymerase is effec-tively blocked or reduced and mRNA is notmade. This process, whereby transcription ofcatabolic genes, including other PTS genes, isblocked in the presence of a preferred sugar, iscalled catabolite repression.

Finally, sugar metabolism in lactic acid bacte-ria can also be genetically regulated directly viarepressor proteins. For example, in L. lactis, thelactose PTS operon is negatively regulated byLacR, a repressor protein encoded by the lacR



gene (Figure 2–14). When lactose is present,lacR expression is itself repressed, and tran-scription of the lac operon is induced.However,when lactose is unavailable or when cells aregrown on glucose, LacR is expressed and tran-scription of the lac genes is repressed.Althougha CRE site is also located near the transcrip-tional start site of the lac operon, suggestingthat the lac genes are subject to CcpA-mediatedrepression, it seems that LacR may have primaryresponsibility for regulating sugar metabolism.

Sugar Metabolism by Saccharomycescerevisiae

Metabolism of carbohydrates by S. cerevisiaeand related yeasts is, in general,not dramaticallydifferent from the lactic acid bacteria. Theseyeasts are facultative anaerobes that are able touse a wide range of carbohydrates. In S. cere-visiae, as in the homofermentative lactic acidbacteria, fermentable sugars are metabolizedvia the glycolytic pathway to pyruvate, yieldingtwo moles of ATP per hexose. However, S. cere-visiae, rather than making lactic acid, insteaddecarboxylates the pyruvate to form acetalde-hyde, which is then reduced to ethanol.Thus,the end-products of sugar fermentation by S.cerevisiae are ethanol and CO2. Furthermore,whereas most lactic acid bacteria are obligatefermentors and have limited means for respira-tory metabolism, S. cerevisiae has an intact cit-ric acid cycle and a functional electron trans-port system and can readily grow and respireunder aerobic conditions.

Glucose transport is the rate-limiting step inglycolysis by S. cerevisiae. Remarkably, S. cere-visiae has as many as seventeen functional glu-cose transporters (referred to as hexose trans-porters or Hxt proteins). Equally surprising,perhaps, is the fact that glucose transport bythese hexose transport systems occurs by facil-itated diffusion.These are energy-independentsystems, driven entirely by the concentrationgradient. Some of these transporters are consti-tutive or inducible, some have low affinity orhigh affinity for their substrates, and others

function not as transporters, per se, but ratheras molecular sensors (see below). Thus, theseemingly excessive redundancy inferred bysuch a large number of transporters insteadprovides these organisms with the versatilitynecessary to grow in environments with awide range of glucose concentrations.

This metabolic flexibility has practical sig-nificance. During beer or wine fermentations,for example, the initial fermentable carbohy-drate concentration is high, but then decreasessubstantially as the fermentation proceeds.Having transport systems that can function athigh, low, and in-between substrate concentra-tions is essential for Saccharomyces to fermentall of the substrate. Complete fermentation ofsugars ensures that these products can reachfull attenuation or dryness, qualities importantin beer and wine manufacture.

Regulation of sugar metabolism in S. cere-visiae is similar, in principle, to that in lacticacid bacteria, but dramatically different with re-gard to the actual mechanisms and the extent towhich other catabolic pathways are affected. Ingeneral, catabolite repression occurs when twoor more fermentable sugars are present in themedium and the cell must decide the order inwhich these sugars are to be fermented. Sinceglucose is usually the preferred sugar,genes thatencode for metabolism of other sugars are sub-ject to catabolite repression. In lactic acid bacte-ria, catabolite repression is mediated via thephosphorylation state of HPr, in concert withCcpA (see above).Repression of catabolic genesfor other sugars occurs at cre sites located nearpromoter regions. In S. cerevisiae, catabolite re-pression (more frequently referred to as glucoserepression) assumes a much broader function.Glucose not only represses transcription ofgenes for sugar use,but also turns off genes cod-ing for pathways involved in the citric acid cy-cle, the glyoxylate cycle, mitochondrial oxida-tion, and glycerol utilization.

How does glucose mediate this global re-sponse in S. cerevisiae? Although the answer isnot fully established, it appears that glucosegenerates a specific signal that is transmitted



via protein cascades and that ultimately resultsin protein-promoter interactions. Several suchcascades actually exist, accounting for situa-tions in which glucose is present at high orlow concentrations. The main pathway in-volves the repressor protein, Mig1 (Figure 2–17). This is a DNA-binding protein, producedwhen glucose is present. In the presence ofglucose, Mig1 is localized in the nucleus,where it (in conjunction with accessory pro-teins) binds to promoter regions, blockingtranscription of the downstream genes. Whenglucose is absent, however, Mig1 becomesphosphorylated by Snf1, a protein kinase, andis translocated out of the nucleus, resulting inde-repression of the regulated genes (i.e., thosegenes are then expressed). A second proteincascade pathway is also involved in glucose in-duction. In this system, the membrane proteinsSnf3 and Rgt2 sense the presence and level ofglucose and then generate a signal that eitheractivates or inactivates specific glucose trans-porters (whose activities depend on the glu-cose concentration).

Protein Metabolism

Like many bacteria, lactic acid bacteria are un-able to assimilate inorganic nitrogen and in-stead must rely on preformed amino acids tosatisfy their amino acid requirements. Sincemost foods contain only a small pool of freeamino acids, this means that lactic acid bacte-ria must first be able to degrade proteins andlarge peptides and then also be able to trans-port the free amino acids and small peptidesreleased during proteolysis. The proteolyticsystem in lactic acid bacteria, especially thosespecies involved in dairy fermentations, hasbeen well studied. For these bacteria, the milkprotein casein serves as the primary substrate.Casein metabolism by lactic acid bacteria notonly is important nutritionally for the organ-isms, but its degradation during cheese manu-facture has major implications for flavor andtexture development (Chapter 5). However, invegetable, bread, and other fermentations, pro-tein metabolism is somewhat less important.

The casein utilization system in lactic acidbacteria involves three main steps. First, caseinis hydrolyzed by proteinases to form peptides.Next, the peptides are transported into cellsvia peptide transport systems. Finally, the pep-tides are hydrolyzed by intracellular peptidasesto form free amino acids (Figure 2–18).Each ofthese steps are described below.

1. The proteinase system

Hydrolysis of casein by lactic acid bacteria oc-curs via a cell envelope-associated serine pro-teinase called PrtP.This enzyme is actually syn-thesized as a large inactive pre-pro-proteinase(�200 kDa). The “pre” portion contains aleader sequence, whose function is to directthe protein across the cytoplasmic membrane.The “pro” sequence presumably stabilizes theprotein during its synthesis.After both of theseregions are removed, the now mature pro-teinase remains anchored to the cell envelope.With casein as a substrate, more than 100products are formed by PrtP.The majority arelarge oligopeptides (up to thirty amino acidresidues), but most are between four and tenresidues.

2. Peptide transport systems

Despite their rather large size and bulkiness,lactococci and other lactic acid bacteria trans-port peptides directly into the cell without fur-ther extracellular hydrolysis. In addition, al-though there are a myriad of peptides formedby PrtP, relatively few transporters can deliverthese peptides into the cell, and even fewer areactually essential for growth in milk. For exam-ple, both DtpT and DtpP transport di- andtripeptides, respectively, but given that thesepeptides are generally not released during ca-sein proteolysis, the role of these transportersin nitrogen nutrition is probably not very criti-cal. Indeed, lactococcal strains with mutationsin the genes encoding for DtpT and DtpP (i.e.,dtpT and dtpP) grow fine in milk. Lactococciand other lactic acid bacteria instead rely on anoligopeptide transport system (Opp) to satisfy



their amino acid requirements. The Opp sys-tem transports about ten to fourteen differentpeptides of varying size (between four andeleven amino acid residues). In contrast todtpT and dtpP mutants, strains unable to ex-press genes coding for the Opp system do notgrow in milk.

3. Peptidases

In the final step of protein metabolism, thepeptides accumulated in the cytoplasm by theOpp system are hydrolyzed by intracellularpeptidases.There are more than twenty differ-

ent peptidases produced by lactococci and lac-tobacilli that ultimately generate the pool ofamino acids necessary for biosynthesis and cellgrowth. Included are endopeptidases (thatcleave internal peptide bonds) and exopepti-dases (that cleave at terminal peptide bonds).The latter group consist entirely of aminopep-tidases, as carboxypeptidases in lactic acid bac-teria have not been reported (although car-boxypeptidases may be present). For lacticacid bacteria to fully use peptides accumulatedby the Opp system requires the combined ac-tion of endopeptidases, aminopeptidases, di-peptidases, and tripeptidases.

Mig1

(- )

Snf1~P Snf1

Mig1~P Mig1

[glucose] low [glucose] high

[glucose] high

[glucose] low

nucleus

out

in

Figure 2–17. Catabolite repression in Saccharomyces. The key regulatory protein is Mig1, whose pres-ence in the nucleus or the cytoplasm depends on the glucose concentration. When glucose is low, Snf1~Pphosphorylates Mig1, trapping it in the cytoplasm, where it is inactive as a repressor, allowing expressionof catabolic genes. In the presence of glucose, Mig1 is not phosphorylated and remains in the nucleus,where it is able to repress catabolic gene expression.



Other Metabolic Systems

Although sugar metabolism, and the lactic andethanolic fermentations in particular, form thebasis for the manufacture of most fermentedfoods and beverages, other metabolic activitiesare also important.As noted above, protein me-tabolism serves the needs of the cells and pro-vides flavor attributes in cheese. In other dairyproducts, such as buttermilk and sour cream,lactic acid bacteria perform fermentations thatyield diacetyl, a compound that imparts thecharacteristic flavor of these products. InSwiss-type cheeses, the propionic acid fermen-tation results in formation of several metabolicend-products that likewise give these cheesetheir unique features. Finally, while fungi donot ferment substrates in the classical sense(see above), they nonetheless convert proteinsand fats into an array of products with strongaromatic and flavor properties.

The citrate fermentation

Several species of lactic acid bacteria can fer-ment citrate and produce the flavor compounddiacetyl. Initially, citrate is transported by a pH-dependent, PMF-mediated citrate permease(CitP).The intracellular citrate is split by citratelyase to form acetate and oxaloacetate (Figure2–19).The acetate is released directly into themedium, whereas the oxaloacetate is decar-boxylated to form pyruvate and CO2.The pyru-vate concentration, however, may increase beyond the reducing capacity of lactate dehy-drogenase,causing a glycolytic bottleneck due ashortage of NADH.Therefore, the excess pyru-vate is removed by a decarboxylation reactioncatalyzed by thiamine pyrophosphate (TPP)-de-pendent pyruvate decarboxylase.The product,acetaldehyde-TPP, then condenses with anothermolecule of pyruvate, via �-acetolactate syn-thase, forming �-acetolactate.The latter is unsta-

PrtP

cell wall membrane

PepAPepC PepX

PepN

casein

oligopeptides

amino acidsfree amino acids

di- andtripeptides

AA

Opp

DtpP

DtpT

Figure 2–18. Proteolytic system in lactococci. Milk casein is hydrolyzed by a cell envelope-associated pro-teinase (PrtP) to form oligopeptides. These oligopeptides are then transported across the membrane by theoligopeptide transport system (Opp). The intracellular oligopeptides are then hydrolyzed by cytoplasmic pepti-dases (e.g., PepA, PepC, PepN, and PepX) to form amino acids. Extracellular di- and tripeptides and free aminoacids in milk are transported by di- and tripeptide transporters (DtpT, DtpP) and amino acid (AA) transporters.The intracellular di- and tripeptides are then hydrolyzed to amino acids. Adapted from Hutkins, 2001.



ble in the presence of oxygen and is non-enzymatically decarboxylated to form diacetyl.

Because the citrate-to-diacetyl pathwaydoes not generate ATP, it would appear to holdno metabolic advantage for the organism. Infact, it costs the cell energy (supplied by thePMF) to transport citrate. Despite these obser-vations, it is now clear that the cell does deriveenergy during the citrate fermentation (Box 2–4). First, during the intracellular oxaloacetatedecarboxylation reaction, a cytoplasmic pro-ton is consumed, causing the cytoplasmic pHand the �pH component of the PMF to in-crease. Second, when both citrate and a fer-mentable sugar are present, the efflux of lac-tate can drive uptake of citrate.Moreover,since

lactate is monovalent and citrate (at physiolog-ical pH) is divalent, the citrate permease acts asan electrogenic precursor-product exchanger(i.e., making the inside more negative and theoutside more positive), resulting in a net in-crease in the �� or electrical component ofthe PMF. Collectively, therefore, citrate fermen-tation results in an increase in the metabolicenergy available to the cell.

The propionic acid fermentation

The propionic acid fermentation is performedby P. freudenreichii subsp. shermanii and re-lated species. In fermented foods, the relevanceof this pathway is limited to the manufacture of

acetate

citrate

citrate lyase

oxaloacetate

CO 2oxaloacetate decarboxylase

pyruvate

TPP

pyruvate decarboxylase

CO 2

acetaldehyde-TPP

-acetolactatesynthase

-acetolactate

O 2

CO 2

diacetyl

CO 2

-acetolactate decarboxylase

acetoin

acetoinreductase

diacetylreductase

2,3-butanediol

NADH+H NADNADH+H NAD

α

α

α

Figure 2–19. Citrate fermentation pathway in lactic acid bacteria. The dashed line indicates thenon-enzymatic, oxidative decarboxylation reaction.



Swiss-type cheeses. Although some strains ofpropionibacteria can ferment lactose, none isavailable during the time at which these bacte-ria are given the opportunity to grow (i.e., sev-eral weeks after the primary lactose fermenta-tion is complete). Instead, lactate is the onlyenergy source available. Lactate fermentationoccurs via the propionate pathway, whichyields two moles of propionate, and one eachof acetate and CO2 per three moles of lactatefermented. The cell nets one mole of ATP perlactate. In cheese, the actual amount of end-products varies as a result of condensation re-actions, co-metabolism with amino acids, orstrain variation. The propionic acid pathway is quite complex and requires several metal-containing enzymes and vitamin co-factors(Chapter 5).

Metabolism of molds

As noted above, the metabolic activities ofPenicillium, Aspergillus, and other fungi arequite unlike those of bacteria and yeasts. Thelatter have a mostly fermentative metabolism,growing on simple sugars and producing just afew different end-products. In contrast, fungalmetabolism is characterized by secretion ofnumerous proteolytic, amylolytic, and lipolyticenzymes. These enzymatic end-products thenserve as substrates for further metabolism.

Often, the metabolic pathways used byfungi result in an array of unique products. Forexample, when P. roqueforti, the blue mold organism, grows in cheese, substantial proteo-lysis occurs through elaboration of several ex-tracellular proteinases, endopeptidases, and exopeptidases. The resulting amino acids aresubsequently metabolized via deaminases anddecarboxylases releasing amines, ammonia,and other possible flavor compounds.

More end-products that are characteristic ofblue cheese flavors are generated, however,from lipid metabolism. About 20% of triglyc-erides in milk are initially hydrolyzed by P.roqueforti-produced lipases, releasing freefatty acids, including short chain, volatile fattyacids, such as butyric and caproic acids. Metab-

olism of free fatty acids via �-oxidation path-ways then yields a variety of methylketones,compounds that are responsible for the char-acteristic flavor and aroma of blue cheese. Sim-ilarly, growth of the Brie cheese mold, P.camemberti, results in a similar sequence ofmetabolic events. Proteinases and lipases dif-fuse through the cheese (since the mold growsonly at the surface), generating amino and freefatty acids. Subsequent metabolism of theamino acids results in formation of ammonia,methanethiol, and other sulfur compounds,presumably derived from sulfur-containingamino acids. Lipolysis of triglycerides and fattyacid metabolism by P. camemberti are just asimportant in Brie-type cheeses as in blue-veined cheese, and methylketones are abun-dant. In the cheese environment,both P.roque-forti and P. camemberti can use lactic acid as acarbon source, which causes the pH to rise, of-ten to near-neutral levels.

Metabolic Engineering

As described earlier in this chapter as well asthroughout this text, the key to a successfulfermentation is control. The difference be-tween fermentation and spoilage really boilsdown to controlled growth and metabolismversus uncontrolled growth and metabolism.Therefore, one way to improve a given fermen-tation process would be to impose eithergreater control on that process or,alternatively,simply engineer the preferred metabolic routedirectly into the organism of interest.The latterstrategy, referred to as metabolic engineering,provides for a more precise and consistentmetabolic result. For example, if increased di-acetyl production by a lactic acid bacterium isthe goal, rather than manipulate substrate oroxygen levels, the pathway can be geneticallymanipulated, such that carbon flow is directedto diacetyl rather than lactate (e.g., by inacti-vating lactate dehydrogenase). The availabilityof sequenced genomes, along with computa-tional tools, now makes it possible to screenthose genomes, not just for a particular en-zyme, but for entire pathways or clusters of



Box 2–5. Fermentation Microbiology: From Pasteur’s Ferment to Functional Genomics

In 1999, the American Society of Microbiology published a chronology of “Microbiology’s fiftymost significant events during the past 125 years” (ASM News, 1999).The list started with Pas-teur’s discovery in 1861 that yeast produced more ethanol during anaerobic fermentativegrowth compared to growth under aerobic, respiring conditions (the aptly called “Pasteur ef-fect”).The list ended with the report in 1995 that described, for the first time, the completegenome sequence of a bacterium (Fleischmann et al., 1995).

The latter event marked, perhaps, the end of one era and the start of another. From 1995 tothe present (June, 2005), more than 230 microbial genomes have been sequenced (and pub-lished), and another 370 are nearly finished (according to the National Center for Biotechnol-ogy Information,www.ncbi.nlm.nih.gov). Sequencing genomes has evolved from an expensive,multi-year, labor-intensive endeavor to one that is affordable, fast, and highly automated.At thesame time, the development of bioinformatics and various computational tools used to analyzeand compare genome information has also advanced at an equally rapid rate.

Genomes of Food Fermentation Organisms

Among organisms involved in food fermentations, the genome of the common yeast, Saccha-romyces cerevisiae, was published in 1996 (Goffeau et al., 1996).Although the sequence strainwas a lab strain, and was quite different from the typical bakers’ or brewers’ yeasts used in in-dustry, the sequence nonetheless provided valuable information on yeast biology and genetics.Several years later, the genome sequence of the lactic acid bacterium,Lactococcus lactis subsp.lactis, was reported (Bolotin et al., 2001). In the past several years, groups in the United States,Europe, and New Zealand have completed sequencing projects for more than 30 other lacticacid and related bacteria (Table 1).A major collaborative effort was also begun in 2002, whenthe Lactic Acid Bacteria Genome Consortium (in collaboration with the U.S. Department of En-ergy’s Joint Genome Institute) was organized (Klaenhammer et al., 2002). The genomes ofeleven commercially important bacteria were sequenced as part of this project (Figure 1;Makarova et al., 2006).

Genome Sequences are Only the Beginning

The actual output of a genome sequencing project is really just the linear order of nucleotidesarranged as a continuous circle. It is not until the nucleotide sequences (i.e., strings of Gs,Cs,As,and Ts) are analyzed computationally, that open reading frames (orf) and putative genes can beinferred or predicted. Comparison of those orfs (or the translated proteins) to genome databases provides a basis for assigning probable functions to those genes.The latter process,calledgenome annotation,usually results in about 70% of the orfs having assigned functions, althoughas databases grow, so does the likelihood a given orf will have a homolog (i.e., find a match to apreviously assigned protein).

The genome may also contain regions encoding for insertion sequence elements, prophages,and tRNA and rRNA. Of course, predicted function is just that—it is based on the statisticalprobability that a gene codes for a particular protein.To confirm an actual biological functionfor a gene requires some form of biochemical demonstration (e.g., creating loss-of-function mu-tations having the expected phenotype).Therefore, genome sequencing and annotation are re-ally only the first steps toward understanding how pathways are constructed, how regulatorynetworks function, and how these bacteria ultimately behave during food fermentations. Func-tional and comparative genomics not only addresses these questions, but also provides insightinto their phylogeny and evolution.

(Continued)



Table 1. Sequenced genomes of lactic acid bacteria1.

Organism Strain(s)

Lactobacillus acidophilus NCFMLactobacillus brevis ATCC 367Lactobacillus delbrueckiisubsp. bulgaricus ATCC 11842, DN-100107, ATCC BAA-365Lactobacillus caseiATCC 334, BL23Lactobacillus gasseri ATCC 33323Lactobacillus helveticus CNRZ 32, DPC 4571Lactobacillus johnsonii NCC 533Lactobacillus plantarum WCRS1Lactobacillus reuteri 100-23, DSM 20016TLactobacillus rhamnosus HN001Lactobacillus sakei 23KLactococcus lactis subsp. cremoris MG 1363, SK11Lactococcus lactis subsp. lactis IL 1403Leuconostoc mesenteroides ATCC 8293Leuconostoc citreum KM20Oenococcus oeni PSU-1Pediococcus pentosaceus ATCC 25745Streptococcus thermophilus LMG 18311, CNRZ 1066, LMD-9

1Adapted from Klaenhammer et al., 2002

What have the genomes revealed?

The genomes of the lactic acid bacteria are generally small, with several less than 2 Mb.Accord-ingly, the genomes reflect the rather specialized metabolic capabilities of these bacteria and thespecific environmental niches in which they live (Makarova et al., 2006). For example, mostspecies contain genes encoding for only a few of the carbon-utilization pathways (i.e., homo-and heterofermentation; pyruvate dissimulation; and pentose, citrate, and malate fermentation).These bacteria are auxotrophic for several amino acids, and, therefore, contain many genes en-coding for protein and peptide catabolism. In fact, given their overall dependence on obtainingnutrients from the environment, it is not surprising that the genomes are replete (between 13%and 18%) with transport system genes (Lorca et al., 2006).

Genome evolution analyses have revealed that the loss of biosynthetic genes over time, andthe corresponding acquisition of catabolic genes, are consistent with the view that the lacticacid bacteria have recently adapted from nutritionally-poor, plant-type habitats to nutritionally-complex food environments.This suggestion is supported by several observations.First, some ofthe lactic acid bacteria contain a large number of plasmids and transposons that accelerate hor-izontal gene transfer (genetic exchange between related organisms) and that promote adapta-tion to new environments. Second, there are a large number of non-functional pseudogenes(containing mutations or truncation) within the genomes that likely represent “gene decay” orloss of genes that are no longer needed (Makarova et al.,2006).For example,10% of the genomeof the nutritionally-fastidious Streptococcus thermophilus are classified as pseudogenes, withmany encoding specifically for energy metabolism and transport functions (Bolotin et al.,2004).As lactic acid bacteria have evolved, therefore, significant genome reduction, with mod-est genome expansion, has apparently become the norm (Makarova et al., 2006).

References

ASM News. Microbiology’s fifty most significant events during the past 125 years. Poster supplement. 65.

Box 2–5. Fermentation Microbiology: From Pasteur’s Ferment to Functional Genomics (Continued)



Bolotin,A.,P.Wincker,S.Mauger,O. Jaillon,K.Malarme, J.Weissenbach,S.D.Ehrlich,and A.Sorokin.2001.Thecomplete genome sequence of the lactic acid bacterium Lactococcus lactis ssp. lactis IL1403. GenomeRes. 11:731-753.

Fleischmann, R.D., M.D.Adams, et al., (38 other authors). 1995.Whole-genome random sequencing and as-sembly of Haemophilus influenza Rd. Science 269:496–512.

Goffeau,A., B. G. Barrell, H. Bussey, R.W. Davis, B. Dujon, H. Feldmann, F. Galibert, J. D. Hoheisel, C. Jacq, M.Johnston, E. J. Louis, H.W. Mewes,Y. Murakami, P. Philippsen, H.Tettelin, and S. G. Oliver. 1996. Life with6000 genes. Science 274:546–567.

Klaenhammer, T., E. Altermann, et al., (35 other authors). 2002. Discovering lactic acid bacteria by genomics.Antonie van Leeuwenhoek 82:29–58.

Lorca, G.L., R.D. Barabote,V. Zlotopolski, C.Tran, B.Winnen, R.H. Hvorup,A.J. Stonestrom, E. Nguyen, L.-W.Huang, D. Kempton, and M.H. Saier. 2006.Transport capabilities of eleven Gram-positive bacteria: com-parative genomic analyses. In press.

Makarova, K.,Y.Wolf, et al., (and 43 other authors) 2006. Comparative genomics of the lactic acid bacteria.2006. Proc. Nat.Acad. Sci. In Press.

Box 2–5. Fermentation Microbiology: From Pasteur’s Ferment to Functional Genomics (Continued)

5.0 µm

A

E

G

B

1.0 µm

H

C

F D

2.0 µm

I

Figure 1. Electron micrographs of lactic acid and related bacteria: A, Lactobacillus delbrueckkii subsp. bul-garicus; B, Lactobacillus brevis; C, Pediococcus pentosaceus; D, Lactobacillus casei; E, Lactococcus lactis; F,Brevibacterium linens; G, Lactobacillus helveticus; H, Streptococcus thermophilus; and I, Bifidobacteriumlongum. Scale bars are 3.0 �m, unless otherwise indicated. Not shown: Lactobacillus gasseri and Oenococcusoeni. Photos courtesy of J. Broadbent and B. McManus, Utah State University, and D. O’Sullivan, University of Minnesota, and with permission from the American Society for Microbiology (ASM News, March, 2005, p. 121–129)



pathways (Box 2–5). Moreover, a variety ofmolecular tools now exist that can be used toinactivate some pathways and activate others.

BibliographyAxelsson, L. 2004. Lactic acid bacteria: classification

and physiology, pp. 1–66. In Salminen, S., A. vonWrite, and A. Ouwehand. Lactic Acid Bacteria Mi-crobiological and Functional Aspects Third Edi-tion Marcel Dekker, Inc. New York, New York.

Bullerman, L.B. 1993. Biology and health aspects ofmolds in foods and the environment. J. KoreanSoc. Food Nutr. 22:359–366.

Dellaglio, F., L.M.T. Dicks, and S. Torriani. 1992. Thegenus Leuconostoc. In B.J.B. Wood and W.H.Holzapfel, ed.The Lactic Acid Bacteria,Volume 2,The genera of lactic acid bacteria. Blackie Acade-mic and Professional pp 235–268.

de Vos,W.M., I. Boerrigter, R.J. van Rooyen, B. Reiche,and W. Hengstenberg. 1990. Characterization ofthe lactose-specific enzymes of the phospho-transferase system in Lactococcus lactis. J. Biol.Chem. 265:22554–22560.

Gancedo, J.M. 1998.Yeast carbon catabolite repres-sion. Microbiol. Mol. Biol. Rev. 62:334–361.

Garrity, G.M., J. Bell, and T.G. Lilburn. 2004. Taxo-nomic Outline of the Prokaryotes. Bergey’s Man-

ual of Systematic Bacteriology, Second Edition.Release 5.0, Springer-Verlag. DOI: 10.1007/bergeysonline.

Hemme, D., C. Foucaud-Scheunemann. 2004. Leu-conostoc, characteristics, use in dairy technologyand prospects in functional foods. Int. Dairy J.14:467–494.

Hutkins,R.W.2001.Metabolism of starter cultures,p.207. In E.H. Marth and J.L. Steele (ed.), AppliedDairy Microbiology. Marcel Dekker, Inc., NewYork, NY.

Makarova, K., Y. Wolf, et al., (and 43 other authors)2006. Comparative genomics of the lactic acidbacteria. 2006. Proc. Nat.Acad. Sci. In Press.

Moore, D. and L. Novak Frazer. 2002. Essential Fun-gal Genetics. Springer-Verlag. New York, NY.

Naumov,G.I., S.A. James,E.S.Naumova,E.J. Louis, andI.N. Roberts. 2000.Three new species in the Sac-charomyces sensu stricto complex: Saccharo-myces cariocanus, Saccharomyces kudriavze-vii and Saccharomyces mikatae. Int. J. Syst. Evol.Microbiol. 50:1931–1942.

Pitt, J.I. and A.D. Hocking. 1999. Fungi and FoodSpoilage.Aspen Publishers, Inc.Gaithersburg,MD.

Samson, R.A., E.S. Hoekstra, J.C. Frisvad and O. Filtenborg. 2000. Introduction to Food andAirborne Fungi. Sixth Ed. Centraalbureau voorSchimmelcultures. Utrecht,The Netherlands.


3

Starter Cultures

“The connection between wine fermentation and the development of the sugarfungus is not to be underestimated; it is very probable that, by means of the devel-opment of the sugar fungus, fermentation is started.”

From A Preliminary Communication Concerning Experiments on Fermentation of Wine and Putrefaction by Theodor Schwann,

1837 (as recounted by Barnett, 2003.)

in the starting material? In other words, howare fermentations started? There are essen-tially three ways to induce or initiate a foodfermentation.The oldest method simply relieson the indigenous microorganisms present inthe raw material. Raw milk and meat, for ex-ample, usually harbor the very bacteria neces-sary to convert these materials into cheeseand sausage. Grapes and grape crushingequipment, likewise, contain the yeasts re-sponsible for fermenting sugars into ethanoland for transformation of juice into wine.

For these natural fermentations to be suc-cessful, however, requires not only that the“correct” microorganisms be present, but alsothat suitable conditions for their growth are es-tablished. Even if these requirements are satis-fied, however, there is no guarantee that theproduct will meet the quality expectations, besafe to consume, or even be successfully pro-duced. Still, many foods are produced by nat-ural fermentation, including some sausages(Chapter 6), wines (Chapter 10), and picklesand other fermented vegetables (Chapter 7).

Once a successful fermentation has beenachieved, a portion of that product could betransferred to fresh raw material to initiate afermentation. This method, called backslop-ping, is probably nearly as ancient as the nat-ural fermentation practice. It works for almost

Introduction

The successful manufacture of all fermentedproducts relies on the presence, growth, andmetabolism of specific microorganisms. In real-ity, however, it is possible to produce non-fermented counterparts of some fermentedproducts. For example, sour cream, cottagecheese, and summer sausage can be producedin the absence of microorganisms simply byadding food-grade acidulants to the raw mater-ial.Similarly,carbon dioxide can be produced indough by adding chemical leavening agents.Even products as complicated biochemicallyand microbiologically as soy sauce can be madeby chemical means. However, despite the tech-nical feasability of producing these productswithout fermentation,these non-fermented ver-sions generally lack the desired organolepticqualities that are present, and that consumersexpect, in the fermented products. This is be-cause microorganisms are responsible for pro-ducing an array of metabolic end-products andtextural modifications, and replicating those ef-fects by other means is simply not possible.

If microorganisms are indeed necessary for converting raw materials into fermentedfoods with the desired characteristics andproperties, then what is the best way to en-sure that the relevant organisms are present

67





any fermented food, and is still commonly usedfor beer, some cheeses and cultured dairy prod-ucts, sour dough bread, and vinegar. In addition,these methods are still practiced today for small-scale production facilities, as well as in less de-veloped countries and in home-made type prod-ucts. The principle, regardless of product, is the same. Any successfully fermented productshould contain the relevant number and type ofmicroorganisms, and, given a fresh opportunity,they will perform much the same as they hadthe previous time. Despite the detractors of thebackslopping technique (see below), it can beargued that this practice actually selects forhardy and well-acclimated organisms with manyof the desired traits necessary for successfulproduction.

The demonstration by Pasteur that fermen-tation (as well as spoilage) was caused by mi-croorganisms led Lister, Orla-Jensen, and otherearly microbiologists,more than a century ago,to isolate and identify the responsible organ-isms. Koch’s postulates regarding the germ-disease connection could then be applied tofermentation science. Thus, an organism iso-lated from soured or fermented milk could bepurified, re-introduced into fresh milk causingthe expected fermentation, and then re-iso-lated from the newly fermented product. Theimplication of this discovery—that pure cul-tures could be obtained and used to start fer-mentations, did not go unnoticed. Indeed,these observations resulted in a third way toproduce fermented foods, namely via the useof a starter culture containing the relevant microorganisms for that particular product.

Role of Starter Cultures

It is often argued by advocates of traditionalmanufacturing methods that natural fermenta-tions, whether initiated by the endogenousflora or by backslopping, yield products thathave unique or singular quality attributes. Nat-urally-fermented wines, for example, are oftenclaimed to be superior to wines made using astarter culture. Even if a slow or “stuck” fer-mentation occurs occasionally, it would cer-

tainly be worth it (the argument goes) to endup with a truly exceptional product. This atti-tude may be perfectly fine on a small scale ba-sis, given the inherent flexibility in terms oftime and quality expectations. In contrast,how-ever, modern large-scale industrial productionof fermented foods and beverages demandsconsistent product quality and predictable pro-duction schedules, as well as stringent qualitycontrol to ensure food safety.These differencesbetween traditional and modern fermentations,in terms of both how fermented products aremanufactured and their quality expectations,are summarized in Chapter 1,Table 1–1.

Simply defined, starter cultures consist ofmicroorganisms that are inoculated directlyinto food materials to overwhelm the existingflora and bring about desired changes in thefinished product. These changes may includenovel functionality, enhanced preservation, re-duced food safety risks, improved nutritionalor health value, enhanced sensory qualities,and increased economic value.Although somefermented foods can be made without a starterculture, as noted above, the addition of con-centrated microorganisms, in the form of astarter culture, ensures (usually) that productsare manufactured on a timely and repeatableschedule, with consistent and predictableproduct qualities. In the case of large volumefermentations, specifically, cheese fermenta-tions, there is also a volume factor that must beconsidered. In other words, there is no easyway to produce the amount of culture neces-sary for large scale cheese production withoutthe use of concentrated starter cultures (seebelow). For all practical purposes (but with afew notable exceptions), starter cultures arenow considered an essential component ofnearly all commercially-produced fermentedfoods.

History

As noted above, microbiology was not estab-lished as a scientific discipline until the 1860sand ’70s. And although Pasteur, Lister, Koch,Ehrlich, and other early microbiologists were


Starter Cultures 69

concerned about the role of microorganisms asa cause of infectious disease,many of the issuesaddressed by these scientists dealt with food-stuffs, including fermented foods and bever-ages. In fact, many of the members of the microbiology community at the turn of thetwentieth century were essentially food micro-biologists, working on very applied sorts ofproblems.The discovery that bacteria and yeastwere responsible for initiating (as well as spoil-ing) food fermentations led to the realizationthat it was possible to control and improve fer-mentation processes. Although the brewing,baking, and fermented dairy industries werequick to apply this new knowledge and toadopt starter culture technologies, other fer-mented food industries did not adopt startercultures until relatively recently.And some stillrely on natural fermentations.

One of the first such efforts to purify a starterculture was initiated in 1883 at the CarlsbergBrewery in Copenhagen, Denmark.There, EmilChristian Hansen used a dilution method to iso-late pure cultures of brewing yeast, derivedfrom a mixed culture that occasionally pro-duced poor quality beer. Subsequently, he wasable to identify which strain produced the best(or worst) beers. Hansen also was the first toisolate the two types of brewing yeast, the top (or ale) yeast and the bottom (or lageryeast). Eventually, all of the beer produced bythe Carlsberg Brewery was made using theHansen strain, Saccharomyces carlsbergensis(later reclassified as Saccharomycer pasto-nanus),which also became the lager type strain.

Although Pasteur had also proven that winefermentations, like those of beer, were per-formed by yeast, there was little interest at thetime in using pure yeast cultures for wine-making.That satisfactory wine could be madeusing selected strains was demonstrated by theGerman scientist Hermann Müller-Thurgau in1890; however, it wasn’t until the 1960s thatwine yeasts became available (and acceptable)as starter cultures for wine fermentations.

At around the same time and place, anotherHansen, Christian Ditlev Ammentorp Hansen,was working on the extraction of enzymes

from bovine stomach tissue. This work led tothe isolation of the enzyme chymosin, an es-sential ingredient in cheese manufacture. Priorto this time, chymosin had been prepared andused as a crude paste, essentially ground-upcalf stomachs. Using the Hansen process, chy-mosin could be partially purified in a stable, liq-uid form, and the activity standardized. Facto-ries dedicated to chymosin production werebuilt in Copenhagen in 1874 and in New Yorkin 1878. Later, Hansen developed proceduresto produce natural coloring agents for cheeseand other dairy products. By the end of thecentury, the Chr.Hansen’s company began pro-ducing dairy starter cultures, thus establishinga full-service business that continues even to-day to be a world-wide supplier of starters cul-tures and other products for the dairy, meat,brewing, baking, and wine industries.

In the United States, a culture industry de-voted to bakers’ yeast arose in the 1860s and’70s. Two immigrant brothers from Europe,Charles and Max Fleischmann, began a yeastfactory in Cincinnati, Ohio, producing a com-pressed yeast cake for use by commercial aswell as home baking markets. Bread pro-duced using this yeast culture was far supe-rior to breads made using brewing yeasts,which was the common form at the time.More than forty years later, in 1923, anotherbakers’ yeast production facility was built inMontreal, Canada, by Fred Lallemand, anotherEuropean immigrant.

With the exception of brewers’ and bakers’yeasts, all of the initial cultures developed andmarketed prior to the 1900s consisted of bacte-ria that were used by the dairy industry.Therewas particular interest in flavor-producing cul-tures that could be used for butter manufacture.Most of the butter produced in the UnitedStates and Europe was cultured butter, madefrom soured cream (in contrast to sweet creambutter which now dominates the U.S. market).The cream was soured either via a natural orspontaneous fermentation or by addition of pre-viously soured cream or buttermilk. Both meth-ods, however, often resulted in inconsistentproduct quality. By the 1880s, researchers in



Europe (Storch in Denmark and Weigman inGermany) and the United States (Conn) showedthat strains of lactic acid bacteria could begrown in pure culture, and then used to ripencream.

The first dairy starter cultures were liquidcultures, prepared by growing pure strains inheat-sterilized milk. The main problem withthese cultures was that they would becomeover-acidified and lose viability (discussed inmore detail later in this chapter).To maintain amore neutral pH, calcium carbonate was oftenadded as a buffer. Still, liquid cultures had a rela-tively short shelf life,which eventually led to thedevelopment of air-dried cultures. The latterwere rather crude preparations and were pro-duced simply by passing liquid cultures throughcheese cloth, followed by dehydration at 15°Cto 18°C.These cultures were more stable,yet re-quired several transfers in milk to revive the cul-ture and return it to an active state.Freeze drieddairy cultures had also become available by the1920s,but cell viability remained a problem,andeven these products required growth in inter-mediate or mother cultures to activate the cells.Frozen cultures, which are now among themost popular form for dairy cultures, were in-troduced in the 1960s. In the last twenty years,significant improvements in both freezing andfreeze-drying technologies have made thesetypes of cultures the dominant forms in thestarter culture market.

The technologies described above werelargely developed by starter culture companiesand by researchers at universities and researchinstitutes. Most of the culture houses startedout as family-owned enterprises and then grewto rather large companies with significant re-search and production capabilities. By the1980s, large pharmaceutical-based corpora-tions acquired many of these companies, al-though a few smaller ones remain, sellingmostly specialized products.

As noted above, cultures are not the onlyproduct the industry sells.There is a rather largeamount of culture media that is used to propa-gate starter cultures and this represents a sub-stantial market, as do the enzymes and coloring

agents used for cheese production. Finally, dueto the demanding nature of the cheese and cul-tured dairy products industries—successfulproduct manufacture requires consistent cul-ture performance—starter culture companiesoften provide nearly round-the-clock technicalservice and support to their customers. If a fer-mentation is delayed, slow, or is otherwise de-fective, the culture (and the culture supplier)will often be blamed. Cultures, after all, areabout the only food ingredient supplied in a bi-ologically active form.Enzymes also fall into thecategory, but are generally more easily standard-ized and stabilized.

A discussion of the history of starter culturescience and technology would be incompletewithout noting what has been perhaps themain driving force for much of the basic andapplied research on lactic starter cultures.Theobservation that bacteriophages,viruses whosehosts are bacteria, could infect and then killstarter culture bacteria was first made in the1930s.Since that time,but especially in the pastthirty years (coinciding with the advent of mo-lecular biology), several research groups in theUnited States and from around the world havedevoted significant attention to understandingand controlling bacteriophages (Box 3–1).Thisresearch has led to countless other importantand fundamental discoveries about lactic acidbacteria.

Starter Culture Microorganisms

Chapter 2 described many of the microorgan-isms that are important in fermented foods.Not all of these organisms, however, are pro-duced or marketed as starter cultures. Some ofthe products do not lend themselves to starterculture applications, and for others, the marketis simply too small or too specialized. For ex-ample, sauerkraut and pickle fermentations aremediated by the natural microflora; therefore,even commercial manufacturers can producethese products without a starter culture. Al-though cultures and manufacturing proce-dures now exist and can be used for controlledstarter culture-mediated pickle fermentations,


Starter Cultures 71

Box 3-1. Bacteriophages of Lactic Acid Bacteria

The physical structure, biology, taxonomy, and life cycle of bacteriophages are quite unlikethose of the bacteria,yeasts, and molds. In fact,bacteriophages,defined simply as viruses that at-tack bacteria, look and behave like nothing else in the biological world. In essence, bacterio-phages are just packages of DNA (or RNA) contained within a protective “head” which is at-tached to structures that enable them to adhere to and ultimately parasitize host cells.

Despite their simplicity, their importance in food microbiology cannot be overstated. Bacte-riophages that infect Escherichia coli O157:H7, Vibrio cholera, and other pathogens, for exam-ple,may carry genes encoding for toxins and other virulence factors,which are then transferredto and expressed by the host cell. In fermented foods made using lactic starter cultures, bacte-riophages are often responsible for causing manufacturing delays, quality defects, and otherdetrimental effects.The economic consequences of bacteriophage infection can be enormous,especially in large scale operations.Thus, understanding the ecology, mode of replication, andtransmission of bacteriophages are essential for minimizing the incidence of bacteriophageproblems and for developing strategies for their control.

Bacteriophage classification

Because bacteriophages are parasites and cannot grow outside of a host, they have no real phys-iological activities that can be used as a basis for systematic classification. That is, bacterio-phages are neither aerobic or anaerobic, they have no temperature or moisture optima, per se(aside from attachment kinetics or host-dependent functions), and biochemical or metabolicpathways do not exist. Thus, classification is based on other criteria, including morphology,structural composition, serology, host range, and DNA and genome analyses (McGrath et al.,2004). Morphological characteristics of bacteriophages that infect lactic acid bacteria are sum-marized in Table 1.

Table 1. Morphotypes of lactic acid bacteriophage.

Morphotypes Group Description Representative types

Myoviridae A Contractile tails 3-B1, 7E1Siphoviridae B1 Long non-contractile tails; small isometric head r1t, P335, sk1, blL170

B2 Long non-contractile tails;prolate head c2, blL67B3 Long non-contractile tails;elongated head JCL 1032

Podoviridae C Short non-contractile tails PO34, KSY1

From Brusow, 2001; Desiere et al., 2002; McGrath et al., 2004; and Stanley et al., 2003

Based on morphological distinctions, there are three main types of bacteriophages: morpho-types A, B, and C, with the B or Siphoviridae group being the most common type that infect lac-tic acid bacteria.The group B phages have long tails with heads that have either a small isomet-ric (B1 sub-group), prolate (B2), or elongated (B3) shape (Figure 1). The lactococcal phagesgenerally belong to the B1 or B2 group, whereas the phages that infect Streptococcus ther-mophilus are all in the B1 group. Furthermore, most of the lactococcal phages that have beencharacterized share structural similarity to three specific lactococcal phages: c2 (a B2 morpho-type),936,and P335 (morphotype B1).Therefore, it is common to refer to other phages as beingrepresented by one of these three types. For example, phages of the P335 type are consideredto be the most common type infecting industrial fermentations that rely on lactococci (Durmazand Klaenhammer, 2000).

Bacteriophage genomics has now become one of the most powerful means for characteriz-ing and classifying lactococcal and related phages. Sequence analyses of bacteriophage

(Continued)



genomes (more than thirty have been sequenced) have revealed that gene clusters are arrangedin modular fashion (Figure 2).That is, the genome consists of modules that code for particularfunctions, such as replication, structure,or assembly. In addition, these modules are transcribedeither during the early or late stages of the phage replication cycle. In some phages, a middlegene cluster also exists. Phage classification, then, can be based on similarities in the genetic or-ganization of different phages. Moreover, it is now recognized that the ability of lytic bacterio-phages to undergo frequent recombination is likely due to the modular nature of the phagegenome. Importantly, recombination by phage may result in the acquisition of new genetic in-formation, including genes that enable the phage to counter host defense systems (Rakonjac etal., 2005).

Box 3-1. Bacteriophages of Lactic Acid Bacteria (Continued)

Figure 1. Electron photomicrographs of representative lactococcal bacteriophages. All were all isolated fromcheese factories. Photos provided courtesy of Sylvain Moineau.

sk1

c2

Tuc2009

Middle

Early Lateori

Late Earlyori

Middle Late Earlyori

Replication Lysis

Morphogenesis

Packaging Lysis

Morphogenesis

ReplicationMorphogenesisPackaging Lysis

ReplicationMorphogenesisPackaging Lysis

Figure 2. Alignment of the three main classes of lactococcal bacteriophages: the prolate headed c2, thesmall isometric headed sk1 (936 class), and the small isometric headed Tuc2009 (P335 class). Early, middle,and late expressed regions and putative functions are indicated. Adapted from Desiere et al., 2002 andStanley et al., 2003.

Bacteriophage biology

Not all bacteriophages infect and lyse their host cells. Some bacteriophages have two routes ofinfection or “lifestyle” choices: lytic or lysogenic. Lytic phages infect a host, replicate inside thathost cell, and eventually lyse the cell, releasing new infectious phage particles. In contrast, a


Starter Cultures 73

culture systems for sauerkraut have not beenas readily adopted (Chapter 7). In othercases, the manufacturer may prefer to main-tain and propagate their own proprietary cul-tures. Many breweries, for example, have thenecessary laboratory facilities and personnelto support internal culture programs. Theproduction of fungal-fermented foods (de-scribed in Chapter 12) also frequently relieson house strains.

When starter culture technology first devel-oped (and for many years thereafter), the actualorganisms contained within most starter cul-tures were generally not well established interms of strain or even species identity. Rather,a culture was used simply because it worked,meaning that it produced a good product withconsistent properties. Such undefined cultures,however, are now less frequently used. Instead,the organisms present in modern commercial

lysogenic infection occurs when so-called temperate bacteriophages infect the host cell, butrather than initiating a lytic cycle, they can instead integrate their genome within the host chro-mosome.The phage then exists as a dormant prophage, and is replicated along with the hostgenome during cell growth.

Importantly, lysogenic cells may be immune to subsequent infection. Although the mecha-nisms dictating whether a temperate phage enters a lytic or lysogenic phase are not completelyknown (for lactic phages, that is), it appears that lytic and lysogenic genes are subject to tran-scriptional regulation. Specifically, repressors of these two pathways (i.e., lysis or lysogeny) ap-parently exist that prevent transcription of the relevant genes. It has also been demonstrated experimentally that prophages can be induced, such that the phage is excised from the chro-mosome and is converted from a lysogenic state to a lytic phase.The observation that prophageinduction may occur during cheese manufacture suggests that prophages may be a source ofnew phages within fermentation environments (e.g., cheese plants). In addition, the ability tomanipulate prophage induction by temperature, osmolarity, and other environmental cues,makes it possible to accelerate lysis of starter culture cells and release of enzymes necessary forcheese ripening (Feirtag and McKay, 1987; Lunde et al., 2005).

References

Brüssow, H. 2001. Phages of dairy bacteria.Annu. Rev. Microbiol. 55:283–303.Desiere,F., S.Lucchini,C.Canchaya,M.Ventura,and H.Brüssow.2002.Comparative genomics of phages and

prophages in lactic acid bacteria, p. 73—91. In Siezen, R.J., J. Kok,T.Abee, and G. Schaafsma (ed.). LacticAcid Bacteria: Genetics, Metabolism and Applications.Antonie van Leeuwenhoek 82:73–91. KluwerAcademic Publishers.The Netherlands.

Feirtag, J.M., and L.L. McKay. 1987. Thermoinducible lysis of thermosensitive Streptococcus cremorisstrains. J. Dairy Sci. 70:1779–1784.

Lunde, M.,A.H.Aastveit, J.M. Blatny, and I.F. Nes. 2005. Effects of diverse environmental conditions on �LC3prophage stability in Lactococcus lactis.Appl. Environ. Microbiol. 71:721–727.

McGrath, S., G.F. Fitzgerald, and D. van Sinderen. 2004. Starter cultures: bacteriophage, p. 163—189. In P.F.Fox,P.L.H.McSweeney,T.M.Cogan, and T.P.Guinee (ed.).Cheese: Chemistry, Physics, and Microbiology,Third ed., Volume 1: General Aspects. Elsevier Ltd., London.

Rakonjac, J.,P.W.O’Toole,and M.Lubbers.2005. Isolation of lactococcal prolate phage-phage recombinantsby an enrichment strategy reveals two novel host range determinants. J. Bacteriol. 187:3110–3121.

Stanley, E., S. McGrath, G.F. Fitzgerald, and D. van Sinderen. 2003. Comparative genomics of bacteriophageinfecting lactic acid bacteria p.45—94. In B.J.B.Wood,and P.J.Warner (ed.).Genetics of Lactic Acid Bac-teria. Kluwer Academic/Plenum Publishers. New York, New York.

Box 3-1. Bacteriophages of Lactic Acid Bacteria (Continued)



starter culture preparations are usually verywell defined, often to a strain level. They arealso carefully selected based on the precisephenotypic criteria relevant for the particularproduct (Table 3–1).

Bacterial starter cultures

The most important group of bacteria used asstarter organisms are the lactic acid bacteria(LAB); (Table 3–2). In fact, only a few non-lacticacid bacteria are commercially available asstarter cultures (see below). As described inChapter 2, the LAB consist of a cluster of Grampositive cocci and rods that share several phys-iological and biochemical traits. In general, they

are Gram positive, catalase negative het-erotrophs that metabolize sugars via either ho-mofermentative or heterofermentative metabo-lism. The LAB grow over a wide temperaturerange,although most LAB are either mesophilesor moderate thermophiles (with temperatureoptima of about 30°C and 42°C, respectively).Likewise, they vary with respect to salt toler-ance,osmotolerance,aerobiosis,and other envi-ronmental conditions, accounting, in part, forthe diversity of habitats with which they are associated.

Although twelve genera of LAB are now rec-ognized (Chapter 2), starter culture LAB be-long to one of four genera, with the dairy LABrepresenting the largest group. The latter in-clude species of Lactococcus, Streptococcus,Leuconostoc, and Lactobacillus. The dairystarters are generally grouped as mesophiles orthermophiles,depending on the product appli-cation, but even this designation has becomesomewhat blurry. In fact, thermophilic strainsof Streptococcus thermophilus are now some-times included in mesophilic starter cultures(e.g., for Cheddar cheese), and mesophilic Lac-tococcus lactis susbsp. lactis are occasionallyincorporated in thermophilic cultures (e.g., forMozzarella cheese).

In general, dairy cultures perform threemain functions: (1) to ferment the milk sugar,lactose, and acidify the milk or cheese; (2) togenerate flavor or flavor precursors; and (3) tomodify the texture properties of the product.Lactic acid bacteria used as starter cultures forother fermented food products have similarfunctions, although acidification is by far themost important. In sourdough bread, for exam-ple, starter culture strains of Lactobacillus san-franciscansis ferment maltose and lower thedough pH via production of lactic acid,but theculture also produces acetic acid and other fla-vor and aroma compounds.

In contrast,one of the lactic acid bacteria in-cluded in sausage starter cultures, Pediococcusacidilactici, essentially performs one job—tomake lactic acid and reduce the meat pH to alevel inhibitory to undesirable competitors (in-cluding pathogens). For some applications, the

Table 3.1. Desirable properties of starter cultures.

Culture Property

Dairy cultures Controlled lactic acid production rates

Short lag phasePhage resistanceEase of manufactureStability and consistencyProduce desired flavor and texturePreservation toleranceLack of off-flavors

Meat cultures Fast acidificationProduce desired flavorAntimicrobial activity

Beer cultures Rapid fermentationProduce desired flavorPreservation tolerance and stabilityFlocculationLack of off-flavorsProper attenuationGrowth at wide temperature rangeTolerant to osmotic, temperature,

and handling stresses

Wine cultures OsmotolerantEthanol tolerantFlocculation, sedimentationGrowth at low temperatureProduce consistent flavorMalolactic fermentation

Bread cultures Freeze tolerantProduce desired flavorProduce adequate leavening


Starter Cultures 75

lactic starter culture may serve a purpose quiteremoved from those described above. Thefunction of Oenococcus oeni, an organismused during the wine fermentation, is to con-vert malic acid to lactic acid.This reaction re-sults in a slight but critical increase in pH, andis necessary to soften or de-acidify overlyacidic wines (Chapter 10).

Although LAB are the most important andmost widely-used group of starter culture bac-teria, other non-lactic bacteria also are in-cluded in various starter cultures (Table 3–3).In the cheese industry, Propionibacteriumfreudenreichii subsp. shermanii is used tomanufacture Emmenthaler and other Swiss-type cheeses and Brevibacterium linens isused in the manufacture of so-called surface-ripened cheeses, such as Limburger, Muenster,and brick cheeses (Chapter 5). In Europe (andoccasionally in the United States), manufactur-ers of dry fermented sausages add Micrococcusspp. to the meat mixtures.This organism pro-duces the enzyme nitrate reductase which re-duces the curing salt, sodium (or potassium)nitrate, to the nitrite form, and in so doing, en-hances development of flavor and color attrib-utes (Chapter 6). Finally, the manufacture of

vinegar involves oxidation of ethanol by Aceto-bacter aceti and related species (Chapter 11).

Yeast starter cultures

Bread manufacturers are by far the largest usersof yeast starter cultures. Several different formsof yeast starter cultures are available, rangingfrom the moist yeast cakes, used exclusively bythe baking industry, to the active dry yeastpackages sold at retail to consumers. Yeastsused for the bread fermentation (i.e., bakers’yeast) are classified as Saccharomyces cere-visae, and are selected, in large part, on theirability to produce large amounts of carbondioxide in rapid fashion. Wine, beer, distilledspirits, and nearly all other alcoholic beveragesare also made using yeast cultures (Table 3–3).

Many of these ethanol-producing yeasts arealso classified as S.cerevisae, although they dif-fer markedly from those strains used for breadmanufacture. As noted above, many of thewines produced in Europe are made via nat-ural fermentations, relying on the naturally-occurring yeasts present on the grape surfaceand winery equipment. The trend for mostwine manufacturers, however, has been to use

Table 3.2. Lactic acid bacteria used as starter cultures.

Organism Application

Lactococcus lactis subsp. lactis Cheese, cultured dairy productsLactococcus lactis subsp. cremoris Cheese, cultured dairy productsLactococcus lactis subsp. lactis biovar. diacetylous Cheese, cultured dairy productsLactobacillus helveticus Cheese, cultured dairy productsLactobacillus delbrueckii subsp. bulgaricus Cheese, yogurtLactobacillus sanfranciscensis Sourdough breadLactobacillus casei Cheese, cultured dairy productsLactobacillus sakei SausageLactobacillus plantarum Sausage, fermented vegetablesLactobacillus curvatus SausageStreptococcus thermophilus Cheese, yogurtPediococcus acidilactici Sausage, fermented vegetablesPediococcus pentosaceus SausageTetragenococcus halophila Soy sauceOenococcus oeni WineLeuconostoc lactis Cheese, cultured dairy productsLeuconostoc mesenteroides subsp. cremoris Cheese, cultured dairy products,

fermented vegetables



starter culture yeasts.As with other starter cul-tures, strain selection is based primarily on thedesired flavor and sensory attributes of the fi-nal product. However, other production traits,including the ability to flocculate, to grow athigh sugar concentrations, and to produce ad-equate ethanol levels, are also relevant and aretaken into consideration. The brewing indus-try is another potential user of yeast startercultures, although large breweries generallymaintain their own proprietary cultures.As forwine, strain selection is based on characteris-tics relevant to the needs of the particularbrewer, and include factors such as floccula-tion, flavor development, and ethanol produc-tion rates.

Mold starter cultures

Despite the relatively few products made usingfungi, manufacturers of these products oftenprefer to use mold starter cultures rather thanto propagate their own cultures. Thus, fungalstarter cultures are available for several typesof cheeses, including both the blue-veinedtypes (e.g., Roquefort and Gorgonzola), as wellas the white surface mold-ripened cheeses(e.g., Brie and Camembert).The blue mold cul-tures consist of spore suspensions of Penicil-

lium roqueforti and white mold cheese cul-tures contain Penicillium camemberti. Fungalcultures are also used in the production ofmany of the Asian or soy-derived fermentedfoods (Chapter 12).The main examples includetempeh, made using Rhizopus microsporussubsp. oligosporus, and miso and soy sauce,made using Aspergillus sojae and Aspergillusoryzae. There is also a small but importantmarket for fungal starter cultures used to pro-duce European-style sausages and hams.

Strain identification

Although it may seem obvious that the preciseidentities of the organisms present in a starterculture should be known, strain identificationis not a trivial matter. In fact, microbial system-atics is an evolving discipline that has been sig-nificantly influenced by advances in genetics,genomics, and various molecular techniques(Box 3–2). Thus, the movement or reassign-ment of a given organism from one species toanother is not uncommon. In recent years, forexample, strains of lactic acid bacteria, brew-ers’ yeasts, and acetic acid-producing bacteriaused in vinegar production have all been re-classified into different species or assignednew names altogether.

Table 3.3. Other organisms used as starter cultures.

Organism Application

BacteriaBrevibacterium linens Cheese: pigment, surfacePropionibacterium freudenreichii subsp. shermanii Cheese: eyes in SwissStaphylococcus carnosus subsp. carnosus Meat: acid, flavor, color

MoldPenicillium camemberti Cheese: surface ripens whitePenicillium roqueforti Cheese: blue veins, protease, lipasePenicillium chrysogenum SausageAspergillus oryzae Soy sauce, misoRhizopus microsporus subsp. oligosporus Tempeh

YeastSaccharomyces cerevisiae Bread: carbon dioxide productionSaccharomyces cerevisiae Ale beersSaccharomyces pastorianus Lager beersSaccharomyces cerevisiae Wine


Starter Cultures 77

Box 3–2. Who’s Who in Microbiology: Identifying the Organisms in Starter Cultures

In the early days of microbiology, Pasteur, Lister, and their contemporaries could view an organ-ism microscopically,note its appearance,perhaps even describe a few physiological properties,and then assign a name to it (Krieg, 1988).Thus, classification was initially based on a limitednumber of phenotypic properties or traits.However, this situation changed in the early 1900s asmore biochemical, physiological, and genetic characteristics became known and as objectivemethods for assigning particular organisms to specific groups were adopted.Thus was bornethe science of microbial taxonomy and the development of classification schemes for bacteria.

Although classification and nomenclature are generally important for all biologists, how anorganism is named and the taxonomical group into which it belongs are especially relevant forstarter culture microbiologists. For many years, up until the 1970s, identification of lactic acidbacteria relied mainly on ecological, biochemical, and physiological traits (see below).The firstcomprehensive description of these bacteria was provided in 1919 by Orla-Jensen,and 20 yearslater the Sherman scheme was published (Orla-Jensen, 1919 and Sherman, 1937).As an aside,both of these historical reviews make for excellent reading; the Sherman paper was the first re-view article in the very first issue of the prestigious Bacteriological Reviews journal series.TheSherman scheme, which is still often used today, systematically placed streptococci into fourdistinct groups: pyogenic, viridan, lactic, and enterococci (Table 1).

Table 1. Sherman scheme for classification of streptococci1.

Growth in presence of: Survival

Representative Growth at: 6.5% methylene pH at 60°C, NH3 fromDivision species2 10°C 45°C NaCl blue (0.1%) 9.6 30 minutes peptone

Pyogenic Streptococcus � � � � � � �

pyogenesViridans Streptococcus � � � � � � �

thermophilusLactic Lactococcus

lactis subsp. � � � � � � �

lactisEnterococcus Enterococcus � � � � � � �

faecium

1Adapted from Sherman, 19372Names reflect current nomenclature

These early papers also reflect the changing nature of bacterial taxonomy and nomenclature.In the Orla-Jensen treatise, for example, a description is given of the dairy lactic acid bacteriumthat was originally called Bacterium lactis (by Lister in 1878), then changed, successively toStreptococcus acidi lactici, Bacterium lactis acidi, Bacterium Güntheri, Streptococcus lacti-cus, and then Streptococcus lactis. In 1985, Streptococcus lactis became Lactococcus lactissubsp. lactis (Schleifer et al., 1985). Interestingly, Orla-Jensen noted that “it would be temptingto employ the name Lactococcus” for these bacteria (which was the genus name first used bythe Dutch turn-of-the-century microbiologist, Beijerinck).

Although the initial classification schemes for lactic acid bacteria were based on phenotypicproperties, serological reactions have also been used to differentiate streptococci.The Lance-field reactions (named after Rebecca Lancefield) were based on the antigenic properties of cellwall-associated components, resulting in more than thirteen distinct groups.The dairy strepto-cocci (now L. lactis subsp. lactis and L. lactis subsp. cremoris) were found to possess the group

(Continued)



Despite the numerous changes that haveoccurred in microbial classification andnomenclature (especially since the early1980s), it is still essential that the microbialcontents of a starter culture be accurately de-scribed and that species identification bebased on the best available taxonomical infor-mation.First, culture suppliers need to includethe correct species names on their products,since Generally Regarded As Safe (GRAS) sta-tus is affirmed only for specific organisms. Acheese culture claiming to contain species ofStreptococcus thermophilus, but actually con-taining closely related strains of Enterococcus,would be mis-labeled. Second, accurate identi-fication is necessary for the simple reason that

culture propagation and production processesrequire knowledge of the organism’s nutri-tional and maintenance needs, which arespecies-dependent. The growth requirementsof Lactococcus lactis subsp. lactis are differ-ent, albeit only slightly, from the closely related Lactococcus lactis subsp. cremoris. Fi-nally, should the product manufacturer preferto include species information on the label ofa fermented product, the species name shouldbe correct. (This would be voluntary, since it isnot required.) For example, yogurt and othercultured dairy products that contain probioticbacteria often include species information,since some consumers may actually look forand recognize the names of particular species.

N antigen and became referred to as the Group N streptococci.Although the other major dairystreptococci, Streptococcus thermophilus, was not antigenic and could not be serologicallygrouped, the Lancefield groupings were and still are useful for classifying pathogenic strepto-cocci, including Streptococcus pyogenes and Streptococcus pneumonia (Group A) and Strepto-coccus agalactiae (Group B), as well as enteric streptococci (Group D; now referred to as En-terococcus). However, one taxonomist suggested that “serology is best forgotten when workingwith dairy streptococci” (Garvie, 1984).

As noted above,phenotypic traits are still used successfully to classify lactic acid bacteria.Themost informative distinguishing characteristics include: (1) temperature and pH ranges ofgrowth; (2) tolerance to sodium chloride, methylene blue, and bile salts; (3) production of am-monia from arginine; and (4) carbohydrate fermentation patterns. Some of these are physiolog-ically relevant in fermented foods (e.g., salt tolerance and growth at low or high temperature),whereas others are simply diagnostic (e.g., inhibition by methylene blue).

Based on these criteria (as well as microscopic morphology), it is, experimentally, rather easyto perform the relevant tests and to obtain a presumptive identification of a given lactic acidbacterium. Kits based on sugar fermentation profiles are also available that can be used to iden-tify species of lactic acid bacteria. Specific classification schemes for lactic acid bacteria, basedprimarily on phenotypes, are described in Bergey’s Manual (Holt et al., 1994). Other identifica-tion schemes that rely on membrane fatty acid composition, enzyme structure similarities, andother properties also exist, although they are now less often used.

Despite the value of traditional identification schemes, there is no doubt that the methods de-scribed above lack the power and precision of genome-based techniques. In fact, advances innucleic acid-based bacterial fingerprinting methods have led, not only to new identificationtools, but also to renewed interest in bacterial taxonomy. The ability to distinguish betweenstrains of the same species is important not only for identification purposes, but also because itprovides a basis for understanding the phylogenetic and evolutionary relationships betweenstarter culture organisms.Although morphological, biochemical, and other phenotypic charac-teristics remain useful for genus and species identification, molecular approaches that rely onnucleotide sequences have proven to be more reliable, more reproducible, and more robust.Several techniques, in particular, are widely used, including pulsed field gel electrophoresis

Box 3–2. Who’s Who in Microbiology: Identifying the Organisms in Starter Cultures (Continued)


Starter Cultures 79

For such products, the name of the organism(and perhaps even the strain) is especially important because probiotic activity dependson the actual species or strain. Unfortunately,species declarations on consumer productsare often incorrect or are outdated.

Starter Culture Math

Using starter cultures not only ensures a con-sistent and predictable fermentation, it alsoaddresses a more fundamental problem,namely how to produce enough cells to ac-commodate the inocula demands of large-scale fermentations. For most fermentedfoods, the first requirement of a starter cul-

ture is that it initiate a fermentation promptlyand rapidly. Although exceptions exists (e.g.,in the case of the secondary, carbon dioxide-evolving fermentation that occurs late in theSwiss cheese process), it is usually necessarythat the fermentation commence shortly afterthe culture is added. And while a short lagphase may certainly be acceptable or even ex-pected, a long lag phase is generally a signthat the culture has suffered a loss of cell via-bility.Thus, for a starter culture to function ef-fectively, it must contain a large number of vi-able microorganisms.

As the mass or volume of the food (or liq-uid) starting material increases, either largerstarter culture volumes or greater starter

(PFGE), restriction fragment length polymorphism (RFLP), ribotyping, 16s ribosomal RNA se-quence analysis, and various PCR-based protocols.These precise methods of strain identifica-tion have also become important for commercial reasons because proprietary organisms, withunique characteristics and production capabilities, are developed and used in fermentationprocesses.

Finally, the reader should be aware, as authors of Bergey’s Manual have warned, that bacteriado not really care into what group they are classified or what they are called—rather, classifica-tion is done for the sole benefit of microbiologists (Staley and Krieg,1986).Therefore, there is acertain degree of arbitrary decision-making involved in the classification process,despite effortsby taxonomists to be totally objective. In fact, as these authors noted, in bold type,“There is no‘official’ classification of bacteria.” Names of bacteria can certainly be ‘valid’ (as per the List ofBacterial Names with Standing in Nomenclature, available online at www.bacterio.citc.fr), butthe value of a given classification system depends entirely on its acceptance by the microbiol-ogy community.

References

Garvie, E.I. 1984.Taxonomy and identification of bacteria important in cheese and fermented dairy prod-ucts, p. 35–66. In F.L. Davies and B.D. Law (ed.). Advances in the Microbiology and Biochemistry ofCheese and Fermented Milk. Elsevier Science Publishers. London.

Holt, J.G., N.R. Krieg, P.H.A. Sneath, J.T. Staley, and S.T.Williams. 1994. Bergey’s Manual of DeterminativeBacteriology. Ninth Edition.Williams and Wilkins. Baltimore, Maryland.

Krieg, N.R., 1988. Bacterial classification: an overview. Can. J. Microbiol. 34:536–540.Orla-Jensen, S. 1919.The lactic acid bacteria. Copenhagen.Schleifer, K.H., J. Kraus, C. Dvorak, R. Kilpper-Bälz, M.D. Collins, and W. Fischer. 1985.Transfer of Streptococ-

cus lactis and related streptococci to the genus Lactococcus gen. nov. System.Appl. Microbiol. 6:183–195.

Sherman, J.M. 1937.The streptococci. Bacteriol. Rev. 1:3–97.Staley, J.T., and N.R. Krieg. 1986. Classification of procaryotic organisms: an overview, p. 965–968. In P.H.A.

Sneath, N.S. Mair, M.E. Sharpe, and J.G. Holt (ed.). Bergey’s Manual of Systematic Bacteriology, Volume2.Williams and Wilkins. Baltimore, Maryland.

Box 3–2. Who’s Who in Microbiology: Identifying the Organisms in Starter Cultures (Continued)



culture cell concentrations are required. Forexample, if a 1% starter culture inoculum isordinarily used for a given product, then 1kg (or 1 L) of starter culture would be nec-essary to inoculate 100 kg (or 100 L) of thefood substrate material.This modest-sized in-oculum could easily be produced from acolony or stock culture, simply by one ortwo successive transfers through an inter-mediate culture (e.g., 0.1 ml or one colonyinto 10 ml, followed by 10 ml into 1 L).

When pure cultures, rather than backslop-ping techniques, were first introduced a cen-tury ago in the dairy industry, this approach ofmaking intermediate and mother cultures wasgenerally how cultures were prepared andused. As the size of the industry increased,however, such that larger and larger starter cul-ture volumes were required, it was no longerfeasible for cheese manufacturers to preparecultures in this manner. In other words, acheese plant receiving 1 million L of milk perday would need 10,000 L of culture, plus all ofthe intermediate cultures (and incubations)necessary to reach this volume.

To address this situation, two general types ofcultures are manufactured and sold to the fer-mented food and beverage industries. The firsttype, often referred to in the dairy industry asbulk cultures, are used to inoculate a bulk tank.The bulk starter culture essentially is the equiva-lent of several intermediate cultures that tradi-tionally have been required to build up the cul-ture. After a suitable incubation period in theappropriate culture medium, the fully-grownbulk culture (which is akin to a mother culture)is then used to inoculate the raw material.Thestarter culture organisms comprising the bulkculture will remain viable for many hours, pro-vided they are protected against acid damage,oxygen, hydrogen peroxide, or other inhibitoryend-products. In the cheese and fermented dairyproducts industry, bulk cultures are routinelyused to inoculate production vats throughout amanufacturing day. Maintaining culture viabilityis still an important issue,however,as will be dis-cussed below.

A bulk culture is not warranted for manyfermented foods simply because raw materialvolumes are more modest.That is, the amountof culture necessary to inoculate the fermen-tation substrate can easily be met using theculture as supplied directly from the manufac-turer. For example, bakers’ yeasts may be sup-plied as yeast cakes, which are added directlyto the dough ingredients just prior to mixing.Similarly, meat starter cultures, whether infrozen can form or lyophilized packets, areadded directly to the meat mixture. Even dairycultures that are designed to be inoculated directly into the food material are now avail-able, thus eliminating the need to preparebulk cultures.

In the dairy industry, these cultures are re-ferred to as direct-to-vat set cultures.They haveseveral advantages that have made them verypopular. They eliminate the labor, hardware,and capital costs associated with the construc-tion, preparation, and maintenance of bulkstarter culture systems.They also eliminate left-over or wasted bulk culture. Although theywere initially produced as frozen concentrates,packaged in cans, they are now available aspourable pellets or lyophilized powders, mak-ing it easy to dispense the exact amount neces-sary for inoculation.

As described below, eliminating the bulk cul-ture fermentation also means that bacterio-phages have one less opportunity to infect theculture and cause trouble.The use of these cul-tures can also reduce mixed strain composi-tional variability. However, when direct-to-vatset cultures are used to inoculate large volumes,the culture must be highly concentrated to de-liver a sufficient inoculum into the raw material.Freezing, centrifugation, filtration, lyophiliza-tion, and other concentration steps may indi-rectly reduce culture viability,ultimately leadingto slow-starting fermentations.Improvements inconcentration technologies have minimizedsome of these problems. Still, direct-to-vat cul-tures are generally more expensive than bulkcultures, and, despite their convenience, maynot be economical for all operations.


Starter Cultures 81

Culture Composition

Mixed or undefined cultures

Mixed or undefined cultures were once themain type of culture used throughout the fer-mented foods industry. They typically containblends of different organisms, representingseveral genera, species, or even strains.The ac-tual identities of the organisms in a mixed cul-ture are often not known, and the individualspecies may or may not have been character-ized microbiologically or biochemically. Eventhe proportion of different organisms in amixed culture is not necessarily constant fromone product lot to another.

Nevertheless, undefined cultures are stillused as starter cultures for many applicationsbecause they have a proven history of suc-cessful use. In Italy, the famous ParmigianoReggiano cheese is made using a mixed cul-ture obtained by overnight incubation of thewhey collected from that day’s cheese pro-duction. Moreover, depending on the prod-uct, mixed cultures may have a particular ad-vantage. For example, mixed starter culturesare commonly used in the Netherlands for themanufacture of Gouda, Edam, and relatedcheese varieties.The fermentation part of thecheese-making process can be quite long, andit is not unusual for a given strain to be in-fected and subsequently killed by bacterio-phages that inhabit the cheese plant environ-ment (or that are present within the cultureitself ). However, given the diversity of lacticacid bacteria present in these mixed cultures,there are likely strains that are resistant tothat particular bacteriophage and that canthen complete the fermentation. In fact, it iswell established that frequent exposure to dif-ferent bacteriophages provides a natural andeffective mechanism for ensuring that phage-resistant strains will be present in repeatedlypropagated mixed cultures.

Despite their proven track record, undefinedmixed cultures are not without problems. Inparticular, product quality may be inconsistentand fermentation rates may vary widely, affect-

ing production schedules. For small-scale opera-tions, where time is somewhat flexible andwhere quality variations may be more tolerable,these issues are not so serious.For large produc-tion facilities,however,where precise schedulesare critical and consistent product quality is expected, mixed cultures have become lesscommon. Starter cultures comprised of definedstrains, with more precise biological and bio-chemical properties, are now prevalent. Thesecultures, referred to as defined cultures, simplyrefer to cultures that contain microbiologicallycharacterized strains.

Defined cultures

Defined starter cultures can be comprised of asingle individual strain or as a blend of two ormore strains. The origin of the defined strainsthat are used commercially varies. Some weresimply present in traditional mixed cultures andothers have been isolated from natural sources.For example, in the case of dairy starters, milkproduction habitats and cheese serve as goodsources; in the case of wine starters, grapes and wine-making equipment are good locationsto find suitable yeasts. Defined strains must be identified and characterized for relevantmetabolic and physiological properties, phage-resistance (in the case of dairy strains), andother desirable traits.For fungal starters,safety isan important issue and the inability of a putativeculture strain to produce mycotoxins is an es-sential criteria. In addition, when multiplestrains are assembled into a culture blend, allstrains must be compatible. That is, one strainmust not produce inhibitory agents (e.g., bacte-riocins, hydrogen peroxide, or killer factors)that would affect growth of other organisms orthat would cause one strain to dominate overothers.

Ever since the defined strain concept fordairy starters was introduced in the 1970s,there has been debate about whether single,paired, or multiple strain blends are preferred(Box 3–3). Multiple strains are used for someapplications, such as yogurt cultures, which



Box 3-3. Defined Strain Cultures: How Many Strains Are Enough?

Most large cheese factories in the United States rely on defined multiple strain starter cul-tures.These cultures, which typically contain up to five different strains, perform in a timelyand predictable manner and generally yield products with consistent quality attributes. Per-haps most importantly, they provide the manufacturer with a reasonable level of assurancethat phages will not be a problem, because if one of the strains was suddenly attacked andlysed, other strains in the culture would act as back-ups and would see the fermentationthrough to completion.

However popular the multiple strain approach has become, it is limited for several reasons.First, in the case of Cheddar cheese manufacture, there are not that many phage-unrelatedstrains of Lactococcus lactis subsp. lactis that have the appropriate cheese-making properties.There are even fewer distinct strains of Lactococcus lactis subsp. cremoris, which is generallyconsidered to be the “better”cheese-making strain.Although it is relatively easy to isolate spon-taneous phage-insensitive mutants (derived from good cheese-making strains), these derivativestrains often revert to a phage-sensitive phenotype or else lose plasmids or acquire other muta-tions that make them unsuitable for cheese-making.

Second, it is not so easy to produce and then maintain multiple strain blends that contain thecorrect strains and in the correct ratio. In theory, strains present in the multiple strain cultureare selected based on good cheese-making properties.However, in reality, there may still be vari-ation in product quality using multiple culture blends, at least more so than would occur if onlyone or two strains were used.Also, whenever multiple strain starters are used, phage diversitywithin the production environment would likely increase, whereas when a single or pairedstrain starter is used, the types of phages that one would expect to see in the cheese plantwould be limited.

Given the advantages of single or paired strain starters, why didn’t such cultures quickly be-come the norm in the cheese industry? Indeed, the very idea of defined strain cultures (firstconceived in the 1930s by researchers in New South Wales,and later employed by New Zealandresearchers) was based on the use of phage-resistant, single-strain cultures (recounted by O’Toole,2004).A simple test was also devised to provide a reliable basis for predicting whethera strain will remain resistant to bacteriophages during repeated cheese making trials (Figure 1;Heap and Lawrence, 1976).

Obviously, reliance on a single strain carries a perceived risk (despite the advantages notedabove),because the appearance of an infective phage could quickly decimate the culture and ruinthe fermentation. Even if the single strain was known to be highly phage resistant, it would behard to imagine that the cheese plant manager would get much sleep.Indeed, in the United Statesand Europe, cheese manufacturers were initially quite reluctant to adopt single strain starters(Sandine in 1977 described this as a possibly “frightening” practice). However, research in the1980s from several groups supported the earlier results from New Zealand and demonstrated thatdefined single or paired strains could be used successfully without phage infections (Hynd,1976;Richardson et al., 1980;Thunell et al., 1984;Timmons et al., 1988).

Another way to assuage the worries of the sleepless cheese manager would be to use a straingenetically configured to resist bacteriophage infection (reviewed in Walker and Klaenhammer,2003).Specifically,genes encoding for phage resistance could be introduced into the genome ofthe selected organism, conferring a bacteriophage-resistant phenotype. As described later inthis chapter (and in Chapter 5), there are multiple mechanisms by which lactic acid bacteriacan become phage-resistant.These strains can be constructed such that a single strain is stackedwith several different phage-resistant systems and is therefore insensitive to a variety of phage.Alternatively, isogenic derivatives,each carrying a different phage resistance system,can be con-structed and used in a rotation program.


Starter Cultures 83

References

Heap, H.A., and R.C. Lawrence. 1976.The selection of starter strains for cheesemaking. N.Z.J. Dairy Sci.Tech. 11:16–20.

Hynd, J.1976.The use of concentrated single strain cheese starters in Scotland. J.Soc.Dairy Technol.29:39–45.

O’Toole,D.K.2004.The origin of single strain starter culture usage for commercial Cheddar cheesemaking.Int. J. Dairy Technol. 57:53–55.

Richardson, G.H., G.L. Hong, and C.A. Ernstrom. 1980. Defined single strains of lactic streptococci in bulkculture for Cheddar and Monterey cheese manufacture. J. Dairy Sci. 63:1981–1986.

Sandine,W.E. 1977. New techniques in handling lactic cultures to enhance their performance. J. Dairy Sci.60:822–828.

Thunell, R.K., F.W. Bodyfelt, and W.E. Sandine. 1984. Economic comparisons of Cheddar cheese manufac-tured with defined-strain and commercial mixed-strain cultures. J. Dairy Sci. 67:1061–1068.

Timmons, P., M. Hurley, F. Drinan, C. Daly, and T.M. Cogan. 1988. Development and use of a defined strainstarter system for Cheddar cheese. J. Soc. Dairy Technol. 41:49–53.

Walker, S.A., and T.R. Klaenhammer. 2003.The genetics of phage resistance in Lactococcus lactis. In B.J.B.Wood and P.J.Warner (ed.), p. 291—315. The Lactic Acid Bacteria, Volume 1, Genetics of Lactic AcidBacteria. Kluwer Academic/Plenum Publishers. New York, New York.

Box 3-3. Defined Strain Cultures: How Many Strains Are Enough? (Continued)

Culture (24 hours) + Phage mixture

Milk

Incubate as follows:

Hold at 32°C for 70 minutesIncrease to 39°C over 30 minutes

Hold at 38°C for 160 minutesDecrease to 32°C over 40 minutes

Measure pH

WheyCoagulated Milk

At least 4 cycles

Figure 1. Heap-Lawrence Test. The principleof this simple but highly useful test is basedon the premise that strains capable of grow-ing and producing acid after repeated chal-lenges against bacteriophages have innatephage resistance. Furthermore, by performingthe phage challenge in a laboratory situationthat mimics cheese making, it is predictedthat these strains will behave similarly in anactual cheese production environment. There-fore, strains that “pass” the test could beused in industrial cheese manufacture.



require the presence of two different organ-isms, Streptococcus thermophilus and Lacto-bacillus delbrueckii subsp. bulgaricus. Sourcream cultures,likewise,contain acid-producinglactococci and flavor-producing Leuconostocsp. Many sausage cultures similarly containspecies of Lactobacillus and Pediococcus. How-ever, even for products requiring only a singleorganism (e.g.,Lactococcus lactis used for ched-dar cheese manufacture), paired or multiplestrains are often desired.

The rationale for this preference is basedmainly on bacteriophage concerns. Multiplestrain starters are blended such that they con-tain a broad spectrum of strains with dissimi-lar phage sensitivity patterns. That is, eachstrain is resistant to different phage types, sothat if one strain is infected, the other strainscan complete the fermentation.This approachis reminiscent of the natural mixed strain cul-tures described above, except that the definedstrain blends behave in a more predictableand consistent manner. Identifying whichstrains have become sensitive to the indige-nous phages in that particular plant requiresregular monitoring of phage levels in thecheese whey, and simple in-plant tests havebeen developed for this purpose (along withmore sophisticated titer assays conducted bythe culture supplier laboratories).When phagetiters reach some critical level or a particularstrain begins to show sensitivity, that strain isremoved from the blend and replaced with anew phage-resistant strain or derivative. Multi-ple strain cultures can also be rotated in a par-ticular order to achieve an even greater levelof security (see below).

Manufacture of Starter Cultures

Most starter cultures, including those pro-duced for the dairy,meat,baking, and other fer-mented foods industries, are mass-produced inmodern, large-volume fermenters. The manu-facturing process actually begins with a singlecolony isolate or stock culture, which is thengrown in a small volume prior to inoculationinto the production fermentor (Figure 3–1).

These fermentors operate under aseptic condi-tions, not unlike those used in the pharmaceu-tical industry for production of biomedicalproducts. They almost always operate in abatch mode, because continuous cell produc-tion systems have not yet been adopted by theindustry.The size of the fermentors may vary,from as little as 10 liters to several thousandliters. However, the basic operational parame-ters and control features are essentially thesame for both small and large batches.

The choice of media also varies, dependingon the organisms being grown and the nutri-ent requirements necessary for optimum cellproduction. For example, dairy starter culturesare frequently produced using milk- or whey-based media, whereas molasses or corn syrupcan be used as the basal medium for other lac-tic cultures.Specialized nutrients are also oftenadded to the medium. Water-soluble vitamins,such as the B vitamins, are required for opti-mum growth of lactococci and lactobacilli, andsome species of Streptococcus, Leuconostoc,and Lactobacillus also require specific aminoacids.Other compounds, such as the surfactantTween 80, are added to the growth medium topromote membrane stability of lactobacilli andother LAB during subsequent frozen storageand lyophilization.

To achieve high cell density and maximizebiomass production and cell viability, it is im-portant that inhibitory metabolic end-productsbe removed or neutralized. In particular, lacticacid can decrease the medium pH to a levellow enough to significantly inhibit cell growth.Although the optimum pH for cell growth de-pends on the specific organism (e.g., lacto-bacilli prefer a slightly more acidic medium pHthan lactococci), lactic acid bacteria, in gen-eral, grow best within a pH range of 5 to 6.Therefore, neutralization of the accumulatedacid via addition of alkali, usually gaseous NH3,NH4OH, Na2CO3, or KOH, is essential. Similarly,hydrogen peroxide can also accumulate to in-hibitory levels during culture production as aresult of peroxide-forming oxidation-reductionreactions. Hydrogen peroxide formation canbe reduced or removed by minimizing incor-


Starter Cultures 85

poration of oxygen and by adding catalase di-rectly to the fermentor during cell production.

Provided that optimum growth conditionsare provided, as described above, maximumcell densities of 109 to 1010 cells per ml are typ-

ically obtained. In general, lactic acid bacteriaare harvested either at late log phase or earlystationary phase. However, the optimum har-vest time depends on the specific organism.Atthis point, the cells can be directly packaged as

Isolate or stock culture(from culture collection)

Propagation(small volume)

Confirm purityand identity

Propagation(large volume)

Freeze

Concentrate(centrifugation or filtration)

Freeze

Lyophilization

Storage

Package

Package

Suspend incarrier medium

Cryoprotectants

Freeze

Package

Storage

Storage

Lyophilizedculture

Frozenconcentrated

culture

Frozen bulkculture

pH controlGrowthsupplements

Figure 3–1. Production of industrial starter cultures. Adapted from Buckenhüskes, 1993.



frozen liquid cultures, concentrated and thenfrozen, or lyophilized (freeze-dried).

Frozen liquid cultures are dispensed intometal cans or plastic containers with volumesranging from 100 ml to 500 ml, then rapidlyfrozen in liquid nitrogen (��196°C). This isthe most common form for lactic acid bacterialcultures, and it is how bulk dairy starters andmeat cultures are usually packaged. Concen-trated frozen cultures, in contrast, are used fordirect-to-vat set cheese applications. For bothtypes of frozen cultures, it is critically impor-tant that the cultures remain frozen (��45°C)and that temperature fluctuations be avoidedduring transport and during storage at themanufacturing site.This is because freeze-thawcycles result in the formation of ice crystals,which can then puncture and kill cells, reduc-ing cell number and viability when the cellsare thawed and used.

To produce concentrated frozen cultures,cells are first grown to high cell densities (asdescribed above) and then concentrated eitherby continuous centrifugation or cross-flowmembrane filtration. The cells may then bewashed to remove the spent medium and re-suspended in a suspension solution containingstabilizing agents (see below). The concen-trated cells are then packaged into cans (usu-ally 300 g to 500 g) at cell densities as high as1011 to 1012 cells per gram and are rapidlyfrozen.Thus, the concentrated culture containsabout the equivalent amount of cells thatwould normally be present in a fully grownbulk culture.Typically, a single 500 g can is suf-ficient to inoculate 5,000 kg of milk. Anotherway to freeze concentrate cells is via a pellettechnology, whereby concentrated cells aredropped directly into liquid nitrogen.The cellsfreeze instantaneously in the form of pellets,which can be collected and dispensed into pa-perboard containers or foil or plastic pouches.Provided they are maintained in a frozen state,the pellets are pourable and are easy to weighout and dispense.

Finally, concentrated cells can also be ob-tained by one of several different dehydrationprocesses. Spray drying, for example, can be

very effective for some biological materials;however, it is inapplicable for lactic acid bacte-ria, due to substantial loss in cell number andviability. In contrast, lyophilization, or freezedrying,has long been known to be a more gen-tle process for dehydrating and preservingstarter culture lactic acid bacteria. Althoughsome cell inactivation or injury may still occur,lyophilized cells are generally more stable thancells concentrated by other means.

The lyophilization process involves removalof water from a frozen material by applyingsufficient vacuum (�1.0 Torr) such that subli-mation of ice occurs. For lyophilization of cul-tures, cells are first grown to high cell density,harvested, and concentrated, as describedabove, and then subjected to the freeze-dryingprocess (usually about 5 Pascal to 20 Pascal foreighteen to thirty hours). Freeze-dried culturescan contain from 109 to 1012 cells/gram. Thecells are usually packaged in foil pouches orother oxygen-impermeable material. Althoughlyophilized cells are best maintained at �20°C,they show good stability even at refrigerationtemperature. Lyophilized cultures are now aspopular as frozen direct-to-vat set starter cul-tures, especially for yogurt and other culturedmilk products.

On occasion, growth of both frozen andlyophilized cells may be sluggish, with a pro-longed lag phase, after inoculation into thefood substrate. To maintain cell viability, vari-ous cryoprotectant agents are usually added toprotect the cells against freeze and stress dam-age. Depending on the cell species, cryopro-tectant agents include glycerol, lactose, su-crose, trehalose, ascorbate, and glutamate.Moreover, some microorganisms simply donot lyophilize well, and suffer a more seriousloss in cell number and viability. For example,there are several reports indicating that dairylactobacilli, including L. delbrueckii susbp.bulgaricus, L. helveticus, and L. acidophilus,may be particularly sensitive to lyophilization,and that special efforts must be made to en-sure that these organisms remain viable (Box3–4).This is important because one or more ofthese organisms are included in commercial


Starter Cultures 87

Box 3-4. Improving Cell Viability During Freezing and Lyophilization Processes

Although lyophilization is generally considered as one of the more gentle means for long-termpreservation of starter cultures, the initial freezing and subsequent sublimation steps do take atoll on microorganisms (Carvalho et al., 2004). Freezing, in particular, is detrimental to cells forseveral reasons. First, freezing causes ice crystals that can damage membranes and puncturecells.As water turns to ice, water activity decreases and the intracellular solutes become con-centrated,both of which cause osmotic stress.Within the lactic acid bacteria, lactobacilli are es-pecially sensitive to the freezing and lyophilization process (Monnet et al., 2003). In part, thismay be due to geometry—large rods have more surface area than small spherical cells, and aremore prone to crystal-induced damage.The drying step also influences survival and viability,be-cause cells are prone to collapse as vacuum is applied and as the temperature increases in thefreeze drying chamber (Fonseca et al., 2004).

Growth conditions

Several factors must be considered to achieve high survival rates and to enhance cell viabilityduring freezing and lyophilization. The first is how the cells are grown (Table 1; Carvalho et al.,2004). It is known that many bacteria, including lactic acid bacteria, when exposed to a de-crease in temperature, alter their membrane lipid composition by synthesizing and incorporat-ing specific fatty acids into the membrane (Béal et al., 2001; Murga et al., 2003; Wang et al.,2005).This change in membrane composition provides fluidity at low temperature and helps tominimize membrane damage. Therefore, lower, sub-optimal growth temperatures (e.g., 25°C)may enhance resistance to subsequent freezing conditions (Murga et al., 2000).The addition ofTween 80 (an oleate ester of sorbitan polyethoxylate) to the growth medium also may promotethis shift in fatty acid synthesis and similarly improve stability during freezing (Goldberg and Eschar, 1977).

Table 1. Factors affecting cell survival and viability during lyophilization.

Growth conditionsOsmoprotectantsCryoprotectantsFreezingDrying mediumStorageRehydration

Another means by which cells protect themselves against freezing and dehydration is by ac-cumulating cryoprotectant and osmoprotectant molecules.These compounds, also referred toas compatible solutes, can reach high intracellular concentrations (above 1 M) and function, inpart, by preventing denaturation of macromolecules within the cytoplasm and cell membrane.They also lower intracellular water activity and prevent water loss and eventual plasmolysis.

The most common protectant molecules used by lactic acid bacteria are betaine, carnitine,and proline.They are usually directly transported into the cell, but in some organisms, they canbe synthesized from precursor molecules. Including these substances in the growth medium al-lows the bacteria to stock up so that when freezing occurs, they will be loaded with protectantmolecules. Importantly, compatible solutes accumulate in response to shock conditions otherthan freezing and dehydration, so a quick heat shock may also activate their uptake.Althoughglycerol is also known to enhance cryotolerance, it is not clear that this compound is activelyaccumulated in lactic acid bacteria (Fonseca et al., 2001).

(Continued)



Sub-lethal stress

In addition to activating transport systems for compatible solutes, cold shock, heat shock, andosmotic shock may induce expression of genes involved in the general stress response systemin lactic acid bacteria.The genetic response to these (and other) shock conditions is mediatedby specific sigma factors that regulate transcription of genes whose protein products are nec-essary for defending the cell against stress damage or death.

The function of these so-called stress proteins varies, but, in general, involves repair or degra-dation of misfolded proteins and stabilization of macromolecules and membranes.Thus, greatercell survival and stability rates might be achieved during lyophilization by prior induction of thestress-response system.

This approach might be counter-intuitive in at least one respect. It is well established, for ex-ample, that rapid freezing is better than slow freezing, since the latter generates large ice crystalthat are very damaging to cells. However, cells of L. acidophilus, adapted to low temperature ei-ther by slow cooling or by exposure to a pre-freezing stress, were reported to have greater sur-vival rates compared to rapidly frozen cells, presumably because the stress-response system hadbeen induced in the adapted cells (Bâati et al.,2000).Induction of a cold shock response and iden-tification of cold shock genes were also reported for other lactic acid bacteria, although cryotol-erance was strain-dependent (Broadbent and Lin,1999;Kim and Dunn,1997,Wouters et al.,1999).

Freeze-drying medium

After cells are grown and harvested, they are usually resuspended in a suitable medium prior tofreezing and lyophilization. For dairy applications, the basal medium is typically milk-based.However, because the suspension medium will remain a component of the final cell mixture, itshould not only be optimized to maintain viability during freeze-drying, the medium should alsocontain components that enhance survival and preservation while the cells are in the dehy-drated state. Among the compounds that appear to be most effective are various sugars andsugar alcohols, including lactose, mannose, glucose, trehalose, and sorbitol, as well as aminoacids and phenolic antioxidants (Carvalho et al., 2003).The latter may be especially important,because exposure to oxygen or air is known to reduce viability of dehydrated cells. Membranecomponents, in particular, are very sensitive to oxidation reactions (Castro et al., 1996).

Storage conditions

Although lyophilized cells are generally considered to be quite stable, they are sensitive to envi-ronmental abuses that may occur during storage. High storage temperatures (above 20°C) andtemperature fluctuations significantly reduce survival rates and cell viability. Aerobiosis, asnoted above, may cause detrimental oxidation reactions. Likewise, high relative humidity con-ditions increase the water activity of the cell material and promote undesirable lipid oxidationreactions (Castro et al., 1995).

Rehydration

The conditions by which lyophilized cells are rehydrated can have a significant influence on theviability of the culture. In general, the best suspension medium approximates the medium inwhich the cells were originally dehydrated.Suspension media that are hypotonic or that are toowarm or cold should be avoided (Carvalho et al., 2004).

Cryotolerant mutants

As described above, there is both a genetic as well as a physiological basis for cryotolerance(the ability to remain viable during freezing). Genes that encode for compatible solute uptakesystems and for shock-response systems have been identified in several lactic acid bacteria

Box 3-4. Improving Cell Viability During Freezing and Lyophilization Processes (Continued)


Starter Cultures 89

(Kim and Dunn, 1997;Wouters et al., 1999, 2000b). It has also been reported that differences incryotolerance exists even between different strains of the same species. These observationssuggest that it may be possible to identify mutants that have enhanced crytolerance (Monnet etal., 2003). Indeed,when genes encoding for known cold-shock proteins were overexpressed ina strain of Lactococcus lactis, five- to ten-fold increases in survival rates were observed aftercells were exposed to multiple freeze-thaw cycles (Wouters et al., 2000a).

In another study, spontaneous mutants were obtained by passing Lactobacillus delbrueckiisubsp. bulgaricus cells through thirty growth and freeze-thaw cycles (Monnet et al., 2003). Intheory, these treatments would have selected for cryotolerant mutants, although there weresubpopulations with varying cryotolerant phenotypes.Although it is not clear what the natureof the mutation(s) may have been that conferred enhanced crytolerance, this approach appearsto be a useful means to generate natural cryotolerant strains.

References

Bâati,L.,C.Fabre-Gea,D.Auriol,and P.J.Blanc.2000.Study of the crytolerance of Lactobacillus acidophilus:effect of culture and freezing conditions on the viability and cellular protein levels. Int J. Food Micro-biol. 59:241–247.

Béal, C., F. Fonseca, and G. Corrieu. 2001. Resistance to freezing and frozen storage of Streptococcus ther-mophilus is related to membrane fatty acid composition. J. Dairy Sci. 84:2347–2356.

Broadbent, J.R., and C. Lin. 1999. Effect of heat shock or cold shock treatment on the resistance of Lacto-coccus lactis to freezing and lyophilization. Cryobiology 39:88–102.

Carvalho, A.S., J. Silva, P. Ho, P. Teixeira, F.X. Malcata, and P. Gibbs. 2003. Protective effect of sorbitol andmonosodium glutamate during storage of freeze-dried lactic acid bacteria. Le lait. 83:203–210.

Carvalho,A.S., J. Silva, P. Ho, P.Teixeira, F.X. Malcata, and P. Gibbs. 2004. Relevant factors for the preparationof freeze-dried lactic acid bacteria. Int. Dairy J. 14:835–846.

Castro,H.P.,P.M.Teixeira,and R.Kirby.1995.Storage of lyophilized cultures of Lactobacillus bulgaricus un-der different relative humidities and atmospheres.Appl. Microbiol. Biotechnol. 44:172–176.

Castro, H.P., P.M.Teixeira, and R. Kirby. 1996. Changes in the cell membrane of Lactobacillus bulgaricusduring storage following freeze-drying. Biotechnol. Lett. 18:99–104.

Fonseca, F., C. Béal, and G. Corrieu. 2001. Operating conditions that affect the resistance of lactic acid bac-teria to freezing and frozen storage. Cryobiology 43:189–198.

Fonseca, F., S. Passot, O. Cunin, and M. Marin. 2004. Collapse temperature of freeze-dried Lactobacillus bul-garicus suspensions and protective media. Biotechnol. Prog. 20:229–238.

Kim,W.S., and N.W. Dunn. 1997. Identification of a cold shock gene in lactic acid bacteria and the effect ofcold shock on cryotolerance. Curr. Microbiol. 35:59–63.

Monnet, C., C. Béal, and G. Corrieu. 2003. Improvement of the resistance of Lactobacillus delbrueckii ssp.bulgaricus to freezing by natural selection. J. Dairy Sci. 86:3048–3053.

Murga, M.L.F., G.M. Cabrera, G.F. de Valdez,A. Disalvo, and A.M. Seldes. 2000. Influence of growth tempera-ture on cryotolerance and lipid composition of Lactobacillus acidophilus. J.Appl. Microbiol. 88:342–348.

Wang,Y., G. Corrieu, and C. Béal. 2005. Fermentation pH and temperature influence the cryotolerance ofLactobacillus acidophilus RD758. J. Dairy Sci. 88:21–29.

Wouters, J.A., B. Jeynov, F.M. Rombouts,W.M. de Vos, O.P. Kuipers, and T.Abee. 1999.Analysis of the role of 7kDa cold-shock proteins of Lactococcus lactis MG1363 in cryoprotection. Microbiol. 145:3185–3194.

Wouters, J.A., F.M. Rombouts, O.P. Kuipers,W.M. de Vos, and T.Abee. 2000a.The role of cold-shock proteinsin low-temperature adaptation of food-related bacteria. System.Appl. Microbiol. 23:165–173.

Wouters, J.A., M. Mailhes, O.P. Kuipers,W.M. de Vos, and T.Abee. 2000b. Physiological and regulatory effectsof controlled overproduction of five cold shock proteins of Lactococcus lactis MG1363.Appl. Environ.Microbiol. 66:3756–3763.

Box 3-4. Improving Cell Viability During Freezing and Lyophilization Processes (Continued)



yogurt and other dairy cultures, which are in-creasingly produced in lyophilized forms.

Evaluating Culture Performance

Simply stated, the job of the starter culture is tocarry out the desired fermentation, promptlyand consistently, and to generate the expectedflavor and texture properties relevant to the spe-cific food product.The particular requirementsfor a given strain, however, depend entirely onthe application for that culture (Table 3–1). Ascustomers have come to expect particular andoftentimes exacting product specifications, thespecific traits and properties expressed by cul-ture organisms have also become quite demand-ing.Thus, strains of lactic acid bacteria used asdairy starter cultures are now selected based notjust on lactic acid production rates, but also onflavor- and texture-producing properties, saltsensitivity, compatibility with other strains,phage resistance, and durability during produc-tion and storage.

For example, strains of S. thermophilus andL.delbrueckii subsp.bulgaricus used as yogurtstarter cultures should have, at minimum,rapidacid development rates and good phage resis-tance, but other factors are just as important.That is, how much acetaldehyde the cultureproduces, whether or not it synthesizes ex-opolysaccharides, and to what extent it con-tributes to post-fermentation acidity are all fac-tors that influence strain selection. Moreover,strains of these bacteria that make them suit-able as thermophilic starter cultures for yogurtmanufacture does not necessarily mean theywill be useful as starters for Mozzarella orSwiss cheese manufacture, since the require-ments for the latter products are different fromthose of yogurt. Likewise, traits desirable forbrewing strains of Saccharomyces cerevisiae(e.g., flocculation and ethanol tolerance) arequite different from those important for bak-ers’ yeasts strains of S. cerevisiae.

Several tests are routinely performed to as-sess functional properties of starter cultures.For lactic acid bacteria, acid-production ratesare relevant,whereas for yeast cultures, rates of

CO2 evolution (or gassing rates) may be appro-priate.Fermentation performance for dairy cul-tures, for example, can be easily determined byinoculating heat-treated milk with a standard-ized inoculum (e.g., 0.1% to 1.0%), incubatingthe material at an appropriate temperature,and then either measuring the decrease in pHas a function of time or simply by determiningthe terminal pH or titratable acidity after a spe-cific incubation period (e.g., 6 or 18 hours).

For some cultures, especially frozen orlyophilized products, lag times may be impor-tant and can also be estimated from growthcurve data.Similar tests are performed for meatstarter cultures, whose main function is also toproduce lactic acid.

Another critical test used specifically fordairy cultures involves the determination ofbacteriophage resistance. In one specific pro-cedure, the Heap-Lawrence challenge test, con-ditions that simulate cheese making are used(Box 3–3). Cells are inoculated into culturetubes of milk, a representative bank or factoryphage mixture is added, and the tubes are in-cubated according to a typical cheese produc-tion time-temperature profile. If the pH at theend of the incubation cycle has reached the ex-pected target value (as easily judged by milkcoagulation), then it is assumed that the cul-ture was not affected by the added phage.Theentire process is then repeated up to six times(adding a portion of the whey from the previ-ous cycle).The absence of a sufficient pH de-crease for two successive cycles indicates thatthe culture has become sensitive to bacterio-phages and should be removed from use.

Similar to lactic cultures, yeast cultures usedfor wine, beer, and bread should also fermentrapidly, have good sensory properties, and bestable during storage. Depending on the prod-uct, these yeasts must also possess other spe-cific properties. For example, a critical physio-logical characteristic of wine and beer yeasts istheir ability to flocculate or sediment at the endof the fermentation. Cells that flocculate willsediment much faster relative to free cells. Sedi-mentation of cells enhances their separationfrom the wine or beer and reduces autolysis


Starter Cultures 91

and subsequent release of intracellular con-stituents. Importantly, flocculation is an inheri-table trait,meaning that the flocculation pheno-type is subject to genetic control and can belost or acquired. Identification of several floccu-lation genes that encode for proteins involvedin cell aggregation has made it possible to im-prove brewing yeast strains (Chapter 9).

Other performance characteristics of yeaststarters also depend on the specific product.Lager beer yeasts, for example, should be ableto grow at low temperature and produce fla-vorful end-products. The ability of brewingyeasts to ferment all available sugars (at leastthose that are fermentable), a property knownas attenuation, is a critical property. In winemanufacture, wine yeasts are also often se-lected based on growth at low temperatures(10°C to 14°C),as well as resistance to sulfitingagents. For some applications, wine yeastshould also have high ethanol tolerance andosmotolerance. Finally, the main requirementsfor bakers’ yeasts are to have a short lag phaseand to produce carbon dioxide rapidly in thedough (gassing rate). For some applications,such as frozen doughs, the yeasts should alsohave good cryotolerance, the ability to with-stand freezing conditions, so that the frozendough will rise after thawing.

Compatibility issues

One other absolute requirement for all multiple-strain starter cultures is that each organism mustbe compatible with all the other organismswithin a particular culture. For example, lacticacid bacteria are well-known producers of bac-teriocins,peroxides,and other antimicrobial sub-stances that can inhibit other organisms (see below). Even individual strains of a singlespecies may inhibit other strains of that samespecies. If a multiple strain L. lactis culture wasassembled without regard to compatibility is-sues, it would be entirely possible that one strainmay produce a bacteriocin that was inhibitoryto all other L. lactis strains in that culture. Inmuch the same manner, some wine and beeryeasts produce inhibitory substances called

“killer” factors that inhibit other yeast, thereforeinfluencing how mixed strain cultures are pre-pared. Compatibility also extends to the growthcharacteristics of the organisms in a given cul-ture. Organisms that produce acids and lowerthe pH quickly may prevent growth of their cul-ture partners, a situation encountered in yogurtproducts containing probiotic adjuncts.

How Starter Cultures Are Used

There are two general ways that starter cul-tures are used. As described above, frozen orlyophilized starter cultures can simply be inoc-ulated directly into the food substrate (i.e., di-rect-to-vat).This is the normal means by whichactive bakers’ yeast cultures, fermented meatstarter cultures, and many dairy cultures areadded to the starting food material. In the caseof frozen cultures, the cans are first thawed incold (and usually chlorinated) water immedi-ately prior to use. Frozen pellets can be addeddirectly. Lyophilized cultures can also be addeddirectly, but may require additional stirring ormixing in the vat to facilitate hydration.

Bulk cultures

In contrast to the direct-to-vat type cultures,some fermented foods are inoculated withbulk cultures (Figure 3–2). Bulk cultures areproduced to increase the size of the starter cul-ture (in terms of total cell number and volume)by several orders of magnitude (2,000 L to4,000 L is about an average size tank).

The dairy industry is by far the main user ofbulk cultures (although the liquid creamyeasts used by large bread manufacturers maybe considered bulk cultures).When bulk dairycultures are used, several additional steps arenecessary to ensure that the organisms havereached high cell densities and that they areactive and viable. First, specialized culture me-dia, formulated to provide optimum growthconditions, are almost always used for bulkstarter preparation. Bulk culture media mayalso contain phosphate buffers and otheragents that protect the cells from acid damage



and from infection by lytic bacteriophages(see below).

In general, bulk culture media are not thatdifferent from the fermentation media usedby culture manufacturers for mass produc-tion of starter culture cells. They contain abasal medium consisting of a fermentablecarbohydrate (usually lactose, but glucose orsucrose can also be used) and a nitrogensource (usually proteins derived from milk orwhey). Culture media are also supplementedwith additional sources of vitamins, minerals,and other nutrients. Yeast extract and cornsteep liquor are excellent sources of many ofthese materials.

Buffering activity is usually supplied in theform of carbonate, citrate, or phosphate salts.The latter also provides a phage-inhibitoryfunction via calcium chelation (see below). Insome cases, lactose concentrations may belowered to limit acid production and reducethe risk of acid injury to the cells (see below).All ingredients used in the bulk culture me-dium must be food grade.

Growth and activity of the starter cultureare also influenced by the incubation condi-tions. In general, the incubation temperaturefor a given culture is established based on opti-mum culture performance, and not necessarilyon optimum growth rates.This is because mostcultures usually contain more than one strain,and often more than one species.Thus, the pri-ority is maintaining the desired balance be-tween multiple strains.

For example, Mozzarella cheese culturescontain a mixture of two thermophilic lacticacid bacteria, S. thermophilus and Lactobacil-lus helveticus. The final ratio of these organ-isms in the bulk culture tank depends primar-ily on the incubation temperature. If theculture is incubated at 35°C to 41°C,growth ofS. thermophilus is favored, whereas incubationat 42°C to 46°C favors L.helveticus.

These organisms also differ physiologically,especially with respect to sugar and proteinmetabolism, such that the strain ratio in theculture will have a major impact on the func-tional properties of the finished cheese.Higher

Direct-to-vat

Bulk

BulkTank

ProductionVat

Lyophilized orfrozen cultures

Lyophilized orfrozen cultures

Figure 3–2. Bulk vs. direct-to-vat culture preparation.


Starter Cultures 93

S. thermophilus levels generally yield a Moz-zarella cheese having high residual lactose andgalactose concentrations and lower levels ofprotein hydrolysis. By adjusting the incubationtemperature the manufacturer can controlthese compositional characteristics,which ulti-mately determine how the cheese will performwhen used in pizza manufacture (Chapter 5).

Multiple strain mesophilic cultures are alsoused for the manufacture of sour cream andcultured buttermilk.These cultures contain twogeneral types of organisms, flavor-producersand acid-producers. The flavor-producing or-ganisms consist of one or more strains of citrate-fermenting, diacetyl-forming, heterofer-mentative Leuconostoc mesenteroides subsp.cremoris. Their temperature optimum is about18°C to 25°C. The acid-producers include ho-mofermentative L. lactis subsp. lactis, L. lactissubsp. cremoris, and L. lactis subsp. diacetilac-tis (that has both acid-forming and flavor-forming ability).The latter group have an opti-mum growth temperature of 28°C to 32°C. If a greater ratio of flavor-to-acid producers is desired in the finished culture, low incubation temperatures (�25°C) should beused, whereas a high incubation temperature(�25°C) favor the acid producers.

Controlling pH

Growth of lactic acid bacteria in milk or whey isaccompanied by a marked decrease in pH. Al-though lactic acid bacteria are copious produc-ers of lactic acid, most species are, in fact, trueneutraphiles and grow best at pH levels above6.0. Cells held at pH 5.0 or below for even amodest period of time will lose viability and be-have sluggishly when inoculated into the start-ing material.Therefore,preventing acid injury ordamage during the production and propagationof starter cultures is essential.

In general, there are two ways to minimizeor prevent acid damage to lactic starter cul-tures. Both involve controlling the pH as theculture grows in the bulk culture tank. One ap-proach relies on the addition of alkaline solu-

tions into the bulk culture tank to neutralizethe acid produced during fermentativegrowth. Neutralizing agents may includesodium or ammonium hydroxide or ammoniagas. In actual practice, these so-called externalpH control systems consist of a tank fitted witha pH electrode and pH monitoring device(some systems may control multiple bulktanks). When the pH has decreased below acritical pre-set threshold, a signal is sent to apump that then adds neutralizing agent to thetank until an upper pH limit is reached.As cellsgrow and produce acid, the pH never falls be-low the set threshold.The optimum pH rangedepends on the specific organisms;most exter-nal control systems for mesophilic dairy starterbacteria maintain a pH between 5.8 and 6.2.These systems yield bulk starter cultures withcell densities as much as ten times greater thanuncontrolled cultures, and with enhanced via-bility.Although lactate salts may accumulate asa by-product, resulting in higher media osmo-lality and ionic strength, they do not have sig-nificant effects on cell viability.

It is also possible to neutralize the bulk cul-ture tank via a manual one- or two-time addi-tion of neutralizing agent (sodium or potas-sium hydroxide) to the bulk culture tank.Thisapproach reduces some of the hardware ex-penses (pH electrodes, pumps, pH monitoringdevices) associated with automated externalpH control systems. Bulk cultures, producedunder pH control conditions, may subse-quently be cooled (or not) and used for up toseveral days.

An alternative approach to the external pHcontrol systems is to incorporate buffer saltsdirectly into the bulk culture medium. Theseso-called internal pH control systems have onemajor advantage—they do not require the pur-chase of the expensive external pH controlhardware. The development of internal pHcontrol media (in the late 1970s),however,wasnot quite so simple.

Although phosphate or citrate salts are com-mon ingredients in culture medium, it is notpossible to add enough of these salts to provide



the buffering power necessary to maintain anear-neutral pH during growth by LAB.This isbecause high phosphate concentrations can inhibit LAB used as starter cultures. Thus,the buffer agents must be incorporated into themedium in either an insoluble or encapsulatedform. Internal pH control buffers now includesodium carbonate and trimagnesium phos-phate in forms that are solubilized and releasedinto the medium as a function of low pH, suchthat the pH of the culture medium is main-tained above 5.0. It is necessary to providesome agitation to prevent settling of the bufferagents during bulk culture growth. Comparedto the external pH control systems, internal pHcontrol media provides similar advantages.Since higher cell densities are achieved, andcell viability is enhanced, less culture is needed.In addition, because the cells are maintained in a viable state, they can be used for a longer period of time.

Bacteriophages and Their Control

As noted earlier, problems caused by bacterio-phages have been the driving force for much ofthe research on starter cultures.Bacteriophagescapable of infecting the culture bacteria willlikely be present in open environments, wherelactic fermentations generally occur.Their abil-ity to multiply and spread rapidly (a single repli-cation cycle can be as short as 30 minutes) in acheese or yogurt factory may lead to decima-tion of the starter culture.

During the production of bulk cultures,bac-teriophage problems have become relativelyrare for several reasons.First, the culture mediais ordinarily prepared by heating to pasteuriza-tion temperatures high enough to inactivatebacteriophages. Also, most modern manufac-turing facilities employ aseptic conditionswithin the culture preparation areas, includingair handling systems to prevent entry of air-borne phages. Finally, closed, sterilizable tanksare usually used for culture preparation.

In contrast, cheese manufacturing condi-tions are not always so restrictive and opportu-nities for phage infection commonly exist. For

example, milk used for cheese manufacture isoften unpasteurized or may receive only a rela-tively modest heat treatment. Moreover, thecheesemaking process, in general, is often con-ducted in an open manner.Thus, the milk andcurds are readily accessible to air- or whey-borne phages.Furthermore, in high-throughputcheese production factories, where vats arefilled and re-filled several times per day, there isample time for phages that may initially be pre-sent at low levels to propagate and reach popu-lations high enough to cause slow or arrestedfermentations and poor quality product.

Ultimately, the lactic starter culture and thephages that infect them are in a sort of race—can the culture do its job (i.e., bring about asuccessful fermentation) before the phagecauses culture inhibition? In general, forCheddar-type cheeses, it only takes about 3 to3.5 hours for the fermentation to be com-pleted (the so-called set-to-salt time). How-ever, bearing in mind that a single phage caninfect a host cell and replicate within 30 min-utes, releasing fifty or more new phage parti-cles into the environment, it may not takelong for the phage population to exceed thatof the culture (Table 3–4).Although this exactscenario in which the fermentation is actuallyhalted prematurely by infectious bacterio-phages may be uncommon, there are morefrequent occasions when phage infections ei-ther cause production delays or downgradingof cheese quality, both of which are costly tothe manufacturer.

For many years, lactic starter cultures com-prised of mesophilic lactococci were the“workhorse” cultures. They were used in themanufacture of the most popular cheeses andcultured dairy products, including Cheddar andCheddar-like cheeses, mold-ripened cheeses,Dutch-type cheeses, cottage cheese, culturedbuttermilk,and sour cream (all described, in de-tail, in Chapters 4 and 5).Thus, bacteriophagesthat attacked lactococci, especially L. lactis sus-bsp. lactis, were really the only concern. Re-ports of phage problems occurring for prod-ucts made using other lactic acid bacterialcultures were relatively rare.


Starter Cultures 95

In the past twenty years, however, the inci-dence of phage infections against thermophiliccultures (consisting of S. thermophilus and ei-ther L. bulgaricus or Lactobacillus helveticus)has increased significantly. It is no coincidencethat there has been a huge increase in the pro-duction of Mozzarella cheese, yogurt, and otherproducts that rely on these thermophilic startercultures.The emergence of thermophilic bacte-riophage underscores the problem faced by thestarter culture industry and its customers:wher-ever and whenever lactic acid bacteria are usedon a large scale, phages that infect those organ-isms will undoubtedly appear as potential ad-versaries.

In response to the bacteriophage problem,the dairy starter culture industry, as well as thecheese and cultured dairy products industries,have adopted several strategies (Table 3–5; alsoreviewed in Chapter 5). First, and perhaps mostimportantly, high standards of hygiene must beapplied. In fact,sanitation and asepsis,appropri-ate plant design,and phage exclusion programs

are the first lines of defense against phages.San-itizing agents that are ordinarily used in dairyplants (e.g., chlorine and hypochlorite solu-tions) are also effective against bacteriophages.

All areas where starter cultures are handledand grown should be isolated from the rest ofthe processing facility. Because phages are fre-quently transmitted via air or airborne droplets,the contained areas should be maintained un-der positive pressure, and filtered air should beused within the starter culture room, as well asin the starter tanks. It is also important to locatethe actual production facilities down streamfrom the starter preparation area to ensure thatwaste flow (i.e., whey) does not contaminatethe culture area.The latter point is critical, be-cause whey is considered to be the major reser-voir of phages within a dairy plant, and the pri-mary vehicle by which phage disseminationoccurs. In fact, in other lactic acid fermenta-tions where opportunities for phage transmis-sion are limited, infection of the culture byphages does not occur. For example, in the

Table 3.4. Dynamics of a phage infection1.

Time (h:m) Phage/ml Cells/ml

0 1 1,000,0000:40 50 2,000,000 � 50 � 1,999,9501:20 2,500 3,999,900 � 2,500 � 3,997,4002:00 1.25 � 105 7,994,800 � 1.25 � 105 � 7,869,8002:40 6.25 � 106 15,739,600 � 6.25 � 106 � 9,489,6003:20 3.13 � 108 18,979,200 � 3.13 � 108 � less than 1

1Given the following assumptions:

Initial phage level � 1 phage/ml of milk

Initial cell level � 1,000,000 cells/ml

Phage latent period � 40 minutes

Cell generation time � 40 minutes

Average burst size � 50

Table 3.5. Phage control strategies.

Method Purpose or Function

Sanitation Kill and remove phages in plant environmentPlant design Keep phages out of production area, prevent cross contaminationPhage inhibitory media Prevent phages from attaching to and infectingculture cellsPhage resistant cultures Design starter cultures cells that will grow and perform well even in presence of phagesCulture rotation Prevent proliferation of phages by limiting access to suitable host



manufacture of fermented meats (Chapter 6),fermentation takes place within the individu-ally encased sausages, and spread of phagesfrom one link to another is not possible.

Another way to control phage is to usephage inhibitory media during the culture prop-agation step. As part of the infection process,calcium ions are required for phages to attach toand subsequently invade their bacterial hosts.The development of phage inhibitory medianearly fifty years ago was based on the principlethat by reducing ionic calcium or by incorporat-ing calcium chelating agents into the culturemedia, calcium ions would then be unavailablefor phage attachment.The main chelating agentsare phosphate and citrate salts, which have theadded advantage of providing buffering capac-ity (see above). On the flip side, however, somephages exist that can effectively attach to andinfect host cells even in the absence of calcium.

A third approach to reduce infection bybacteriophages takes advantage of the innateability of some starter culture bacteria to de-fend themselves against phage attack (dis-cussed below and in Chapter 5). Several typesof natural phage-resistance mechanisms exist,including inhibition of adsorption, restrictionof phage DNA, and abortive infection.Thus, itis possible to isolate strains that possess oneor more of these systems and to use them inindustrial fermentations.This approach, as de-scribed previously, forms the basis of mostdairy starter culture systems. In particular, themultiple defined strain starter cultures usedin cheese manufacture often contain as manyas five different phage unrelated strains.Thesecultures can be used either on a continuousbasis (the same culture every day) or rotatedsuch that strains with the same phage sensi-tivity pattern are not used for consecutive fer-mentations. Rotations programs can also beperformed even with single or paired strainstarter. In either case, it is important to moni-tor the phage levels in the whey or milk on aregular basis.

When phage titers reach a particular thresh-old, signifying a strain has become phage-sensitive, that strain is removed from the mix-

ture and replaced by a resistant strain. Sincephage proliferation requires susceptible hoststrains, when those strains are removed fromthe production environment, the backgroundlevel and accumulation of phages will be re-duced such that normal fermentation rates canbe achieved. This practice, however, is con-strained by the limited availability of phage-unrelated strains (discussed previously). Also,frequent exchange of one strain with anotherwithin a given culture may result in undesir-able variations in product quality. In fact, manyof the defined, phage-resistant cultures nowcontain only two or three strains to maintaingreater product consistency. In theory, itshould be possible to isolate new strains fromnature that are both phage-resistant and thathave good cheese making properties, a strat-egy that is now standard industry practice(Box 3–5).

Of course, if some strains are naturally resis-tant to phage infections, then there must be agenetic basis for the phage-resistant pheno-types described above. Indeed, genes responsi-ble for these phage-resistant phenotypes havebeen identified and characterized. Importantly,these genes can be introduced into phage-sensitive strains, making them phage-resistant.In addition, it is also possible to obtain sponta-neous phage-resistant mutants by simply ex-posing sensitive, wild-type strains to lytic bac-teriophages. The phage-resistant derivativesthat are then selected must be evaluated forcheese-making properties before they are reintroduced into a starter culture, becausepleiotropic mutations frequently occur thatrender them defective as cheese cultures (dis-cussed previously).

Engineered Phage Resistance

Genes in Lactococcus lactis subsp. lactis thatconferred a phage-resistance phenotype wereidentified in the early 1980s by researchers atNorth Carolina State University. These geneswere located on a 46 kb, self-transmissible plas-mid, named pTR2030. Initially, it was thoughtthat pTR2030 contained a single phage-


Starter Cultures 97

Box 3-5. Looking Far and Wide for Lactococci

Over the thousands of years that lactic acid bacteria have been used in the manufacture of fer-mented foods, and the 100 years that lactococci, in particular, have been used as dairy startercultures, these organisms have undoubtedly been exposed to considerable selective pressure.To grow and compete well in a milk environment, for example, an organism must transport andmetabolize lactose rapidly and also be able to hydrolyze milk proteins and use the resultingpeptides efficiently.Other traits, specific for cheese fermentations and for which strains are rou-tinely screened, include the ability to produce good cheese flavor and texture, to grow withinthe relevant temperature range, to resist bacteriophage infection, and to be amenable to large-scale propagation, handling, and storage.

Because of these strict phenotypic requirements, it has been suggested that relatively few dis-tinct strains of Lactococcus lactis subsp. lactis and Lactococcus lactis subsp. cremoris actuallyexist and that the overall genetic diversity of dairy lactococci is limited (Salama et al., 1995).Moreover, many of the strains that are marketed commercially, even those from disparate geo-graphical locations, are derived from common stock cultures or dairy products (Ward et al.,2004). It is reasonable to assume that in many cases, these cultures, even if marketed by differ-ent culture companies, may not be very different from one another.Therefore, the discovery ofnew strains that satisfy culture requirements would be quite valuable, especially since thesestrains might provide a rich source of new genetic information, including genes encoding fornovel phage resistance mechanisms.

In an effort to obtain new lactococcal strains that could potentially be used as starter cul-tures, researchers at Oregon State University obtained environmental and dairy samples from di-verse geographical locations (Salama et al., 1993, 1995). Samples (more than 200) were ob-tained from local plant sources, including vegetables, wildflowers, weeds, and grasses, as well asmilk and dairy products from China, Morocco, Ukraine, and the former Yugoslavia.Their screen-ing method was modeled after procedures used for large, rare clone library screening, and in-volved an enrichment step followed by colony hybridization using lactococci- or L. lactis subsp.cremoris-specific 16rRNA probes.The selected isolates were then characterized based on vari-ous phenotypic characteristics (growth at 10°C, 40°C, and 45°C or in 10% salt or at pH 9.2, andarginine hydrolysis). In general, L. lactis subsp. lactis were isolated from both plant and dairysources, whereas L. lactis subsp. cremoris could only be found in dairy samples.

A second level of screening was then performed, based on rapid acid production and flavorprofiles (i.e., tasting milk fermented with each strain).These results indicated that 61 out of 120strains grew well in milk and produced acceptable flavor characteristics.Additional investiga-tions indicated that ten new strains were phage resistant and suitable for cheesemaking (Ur-bach et al., 1997).

Collectively, these studies suggest that new strains, with the necessary phenotypes for dairyfermentations, certainly do exist in nature. However, having a rational isolation and screeningplan is also required if such “bio-prospecting”efforts are to be successful.

References

Salama, M.S.,W.E. Sandine, and S.G. Giovannoni. 1993. Isolation of Lactococcus lactis subsp. lactis from na-ture by colony hybridization with rRNA probes.Appl. Environ. Microbiol. 59:3941–3945.

Salama, M.S.,T. Musafija-Jeknic,W.E. Sandine, and S.G. Giovannoni. 1995.An ecological study of lactic acidbacteria: isolation of Lactococcus lactis subsp. cremoris. J. Dairy Sci. 78:1004–1017.

Urbach, E., B. Daniels, M.S. Salama,W.E. Sandine, and S.G. Giovannoni. 1997.The ldh phylogeny for environ-mental isolates of Lactococcus lactis is consistent with rRNA genotypes but not phenotypes.Appl. Env-iron. Microbiol. 63:694–702.

Ward, L.J.H., H.A. Heap, and W.J. Kelly. 2004. Characterization of closely related lactococcal starter strainswhich show differing patterns of bacteriophage sensitivity. J.Appl. Microbiol. 96:144–148.



resistance determinant (responsible for a heat-sensitive abortive infection phenotype), but itwas later discovered that other genes, encod-ing for a restriction and modification system,were also present. When the plasmid wastransferred via a simple conjugal mating proce-dure into a phage-sensitive, industrial cheese-making strain, transconjugants with resistanceto a broad range of lytic industrial phages wereobtained. This was a noteworthy accomplish-ment, in part, because it represented the firstapplication of biotechnology to improve dairystarter cultures, but also because the actualtechnique of gene transfer did not involve re-combinant DNA technology.Thus, these modi-fied strains could be used commercially.

Indeed, this approach was successful in actual cheese manufacturing environments.However, with prolonged use, bacteriophageseventually appeared in cheese plants that wereable to circumvent the resistance of thesestrains. Nonetheless, these early efforts markedthe beginning of a new era in starter culturetechnology and led to the development ofother molecular strategies aimed at controllingbacteriophages.One approach, for example, in-volved introducing different phage-resistancegenes into a single strain, on an individual ba-sis, thereby generating several isogenic phage-resistant derivatives. When used in a rotationscheme, the properties of the parental strainremain the same, while the resistance patternagainst different phage types is expanded.Other examples of engineered phage resis-tance are described in Chapter 5.

Starter Culture Technology in theTwenty-first Century

Phage is not the only concern of the dairystarter culture industry. How fast a given straingrows, what sugars it ferments, and what end-products are formed are also important issuesthat influence culture performance and thathave been addressed by various researchgroups. In particular, much of the research onlactic acid bacteria in the past thirty-five yearswas based on the pioneering work of McKay

and co-workers at the University of Minnesota(Box 3–6). It was the McKay lab that showedthat most of the phenotypic traits necessaryfor growth and activity of lactic starter cul-tures, including lactose fermentation, caseinhydrolysis, and diacetyl formation, were en-coded by plasmid DNA. In addition, genes en-coding for nisin production and immunity, forphage resistance, and for conjugal transfer fac-tors were also identified by the McKay group.

These discoveries, and the development ofgene transfer and gene exchange techniques,have made it possible to manipulate the physi-ological, biochemical, and genetic propertiesof lactic starter cultures in ways hardly imagin-able a generation ago. Cultures are now cus-tomized to satisfy individual customer de-mands. Not only are improved strains of lacticacid bacteria now used in dairy, meat, veg-etable, soy, and wine fermentations to satisfymodern manufacturing requirements and en-hance product quality, they also are being usedfor novel applications. Lactic acid bacteria,by virtue of their ability to survive digestiveprocesses and reach the intestinal tract, havebeen found to serve as excellent deliveryagents for vaccines and antigens.The use of lac-tic acid bacteria as probiotics has advanced tothe point where the medical community isnow actively involved in evaluating these bac-teria, in well-designed clinical trials, for theirability to treat important chronic diseases(Chapter 4). Certainly, the information minedfrom the “omic” fields of genomics, prote-omics, and metabolomics is bound to lead tomany other new applications (Chapter 2).

Of course, research on starter culture mi-croorganisms is not limited to lactic acid bac-teria. The genome of Saccharomyces cere-visiae was sequenced in 1996, and effortsaimed at manipulating the physiological andbiochemical properties of bread, beer, andwine yeasts have been ongoing for many years(Chapters 8, 9, and 10). Despite the apparentsimplicity of these fermentations, there aremany opportunities for improvement. Bakers’yeasts, for example, can be made more cryotol-erant so that leavening will still occur when


Starter Cultures 99

Box 3-6. Plasmid Biology, Gene Transfer Technology, Microbial Starter Cultures, and Softball: Life in the Larry McKayLaboratory

It can be fairly argued that research interest in lactic acid bacteria is at an all-time high. Nearlytwenty genomes have been sequenced, sophisticated culture improvement programs are un-derway throughout the world, and pharmaceutical companies have developed strains of lacticacid bacteria that are now being used to deliver human and animal vaccines.

Of course, lactic acid bacteria were among the very first groups of bacteria studied by thevery first microbiologists more than a century ago. Pasteur, Lister, Metchnikoff, and Koch all fo-cused (literally) their microscopes and attention on these bacteria.As the applied significanceof these bacteria became evident, researchers devoted their entire careers to understandingtheir physiological and biochemical properties and their particular performance characteristicsin fermented dairy foods. Indeed, great discoveries were made by Orla-Jensen, Sherman, andWhitehead, and research groups led by Paul Elliker, Bob Sellars, Marvin Speck, Bill Sandine,Robert Lawrence, and many others.

Research on lactic acid starter cultures moved into an entirely new,uncharted, and ultimatelyrevolutionary direction in the early 1970s. It was at that time when Larry McKay,a new assistantprofessor at the University of Minnesota, embarked on a research career that essentially createdthe field of starter culture genetics. The McKay lab, managed by his extraordinary assistant,Kathy Baldwin, also became the main teaching laboratory for the next thirty years, responsiblefor training a cadre of students and future scientists that has continued to make advances instarter culture research.

McKay had been a graduate student in the Sandine lab at Oregon State University when hefirst became interested in lactic acid bacteria and when he developed a keen sense of observa-tion. It was his ability to understand what those observations meant that led him to discover thebiochemical pathway by which lactic acid bacteria metabolized lactose, a key finding that wasof fundamental as well as applied significance.

When McKay arrived at Minnesota in 1971, he began to address another question that hadlong puzzled dairy microbiologists—specifically, why is it that some lactococci appear to losetheir ability to ferment lactose (referred to as a lac� phenotype, in contrast to a lactose-fermenting or lac+ phenotype).Similarly, there were also strains that had spontaneously lost theability to degrade milk proteins (a prt� phenotype).This loss of function phenotype had beenobserved by Orla-Jensen as long ago as 1919, by Sherman in 1937, and Hirsch in 1951, but asMcKay noted,“The question remains, however, as to how the loss of lactose metabolism is in-duced in cultures of lactic streptococci” (McKay et al., 1972).

The answer, while obvious today, was not so obvious in 1971, when McKay rightly hypothe-sized that these bacteria “could be carrying a genetic element which is responsible for the cells’ability to ferment lactose” and that “the loss of this element would cause the cell to becomelac�” (McKay et al., 1972). Shortly thereafter, the McKay lab provided the first experimentalproof that these genetic elements were indeed plasmids (Cords et al., 1974; the reader is re-minded that plasmids had only been “discovered”in 1969).The McKay group then published anentire series of seminal papers demonstrating that other essential traits in lactic acid bacteria,including protein metabolism, citrate fermentation, nisin production and resistance, conjuga-tion factors, and phage resistance, were also encoded by plasmid DNA.These (and subsequent)reports all had the McKay signature—they were written in a simple and concise style; the ex-perimental approach was straightforward,precise,and well controlled;and the results were pre-sented clearly and definitively.

As it was becoming clear that important plasmid-encoded functions were widespread in lac-tic acid bacteria and lactococci in particular,McKay quickly realized the implications.Already, in1974, McKay and Baldwin had shown that strains which had lost the ability to ferment lactoseor produce proteinase could be transduced by tranducing bacteriophages, generating “Lac�”

(Continued)



and “Lac�, Prt�” transductants (McKay, Cords, and Baldwin, 1973; McKay and Baldwin, 1974).The development of a conjugation or cell-to-cell mating system (Kempler and McKay, 1979;McKay et al.,1980;and Walsh and McKay,1981) provided a basis for transferring genes from oneorganism to another.

Although transformation, the direct uptake of DNA into the cells, proved to be more elusive,Kondo and McKay reported the first successful use of a transformation technique for lactococciin 1982 and later described optimized procedures in 1984. Over the next several years, studieson gene cloning, construction of food-grade cloning vectors, and mechanisms of gene integra-tion and conjugation were reported (Harlander et al.,1984;Froseth et al.,1988;Mills et al.,1996;Petzel and McKay, 1992). Strain improvement strategies were also devised, including effortsaimed at isolating strains that could accelerate cheese ripening, over-produce bacteriocins, andresist bacteriophage infection (Feirtag and McKay, 1987; McKay et al., 1989; Scherwitz-Harmonand McKay, 1987).

During his career, McKay enjoyed nothing more than perusing the lab, examining plates,looking at gels, or asking questions of the lab personnel. He was just as passionate about teach-ing and mentoring.His course on microbial starter cultures,which was taught every other year,was so valuable that students would often take the class a second time, just to stay caught up inthe field. Upon leaving the lab, McKay’s students and postdoctorates were always well pre-pared, scientifically and professionally, and assumed prestigious positions in academia and in-dustry. Of course, as productive as the McKay lab was for more than three decades, the lab didhave other important interests.There are stories, perhaps apocryphal, that upon interviewing apotential new graduate student, McKay’s final question would have something to do with theapplicant’s skills on the softball field and whether they could contribute to the department’schances for the next season.While there was no substitute for having a solid background in bio-chemistry, genetics, and microbiology, having a good arm at shortstop would certainly havehelped your chances at joining one of the best research laboratories in the world.

References

Cords, B.R., L.L. McKay, and P. Guerry. 1974. Extrachromosomal elements in group N streptococci. J. Bacte-riol. 117:1149–1152.

Harlander, S.K., L.L. McKay, and C.F. Schachtele. 1984. Molecular cloning of the lactose-metabolizing genesfrom Streptococcus lactis.Appl. Environ. Microbiol. 48:347–351.

Feirtag, J.M., and L.L. McKay. 1987. Isolation of Streptococcus lactis C2 mutants selected for temperaturesensitivity and potential in cheese manufacture. J. Dairy Sci. 70:1773–1778.

Froseth, B.R., R.E. Herman, and L.L. McKay. 1988. Cloning of nisin resistance determinant and replicationorigin on 7.6-kilobase EcoRI fragment of pNP40 from Streptococcus lactis subsp. diacetylactis DRC3.Appl. Environ. Microbiol. 54:2136–2139.

Kempler, G.M., and L.L. McKay. 1979. Genetic evidence for plasmid-linked lactose metabolism in Strepto-coccus lactis subsp. diacetylactis.Appl. Environ. Microbiol. 37:1041–1043.

Kondo, J.K., and L.L. McKay. 1982. Transformation of Streptococcus lactis protoplasts by plasmid DNA.Appl. Environ. Microbiol. 43:1213–1215.

Kondo, J.K., and L.L. McKay. 1984. Plasmid transformation of Streptococcus lactis protoplasts: optimizationand use in molecular cloning.Appl. Environ. Microbiol. 48:252–259.

McKay,L.L.,and K.A.Baldwin.1974.Simultaneous loss of proteinase- and lactose-utilizing enzyme activitiesin and reversal of loss by transduction.Appl. Microbiol. 28:342–346.

McKay, L.L., K.A. Baldwin, and P.M.Walsh. 1980. Conjugal transfer of genetic information in Group N strep-tococci.Appl. Environ. Microbiol. 40:84–91.

Box 3-6. Plasmid Biology, Gene Transfer Technology, Microbial Starter Cultures, and Softball: Life in the Larry McKayLaboratory (Continued)


Starter Cultures 101

frozen dough is thawed. Likewise, increasingthe metabolic capacity of these yeasts to fer-ment maltose and other sugars, whose use isordinarily repressed by glucose,would acceler-ate the fermentation rate and shorten the leav-ening step. Modern brewers’ yeast strains canbe modified such that they produce beers withspecific compositional characteristics (e.g.,low carbohydrate, low ethanol), that have spe-cific performance traits (e.g., attenuation, floc-culating properties), or that simply producebetter-tasting beer (e.g., no diacetyl or hydro-gen sulfite). Several desirable traits in wineyeasts have also been targeted for improve-ment. Development of osmophilic wine yeaststrains would be desirable due to their abilityto grow in high sugar-containing musts. Themalolactic fermentation, ordinarily performedby malolactic acid bacteria, could be carriedout by S. cerevisiae harboring genes encodingfor the malic acid transporter and malolacticenzyme.

Despite the opportunities that now exist tomodify and improve starter culture organisms,strain improvement programs have been lim-ited by regulatory and public perception con-siderations. While tools for introducing DNAinto food grade starter microorganisms or for

altering the existing genetic makeup of theseorganisms are now widely available, commer-cialization of genetically modified organisms(GMOs)on a world-wide basis has not oc-curred. Although GMOs are common in theUnited States, they are generally not marketedin Europe or the Far East. The debate overGMOs is not likely to abate anytime soon, evenas researchers continue to demonstrate the po-tential benefits GMO technologies may offer(Box 3–7).

Encapsulated and Immobilized Cells

In the fermentation industry, there has beenconsiderable interest in the development andapplication of encapsulated and immobilizedcell technologies. In general, encapsulationrefers to a process whereby cells are embed-ded or enrobed within a gel-containing shell.Encapsulated cells are metabolically active andfully capable of performing fermentations. Inaddition, encapsulated cells may be eitherfreely suspended in the medium or immobi-lized to an inert support material.Encapsulatedand immobilized cell systems offer several ad-vantages. First, they provide a means for con-ducting fermentations on a continuous rather

McKay, L.L., M.J. Bohanon, K.M. Polzin, P.L. Rule, and K.A. Baldwin. 1989. Localization of separate geneticloci for reduced sensitivity towards small isometric-headed bacteriophage sk1 and prolate-headed bac-teriophage c2 on pGBK17 from Lactococcus lactis subsp. lactis KR2. Appl. Environ. Microbiol. 55:2702–2709.

McKay,L.L.,K.A.Baldwin,and E.A.Zottola.1972.Loss of lactose metabolism in lactic streptococci.Appl.Mi-crobiol. 23:1090–1096.

McKay, L.L., B.R. Cords, and K.A. Baldwin. 1973.Transduction of lactose metabolism in Streptococcus lactisC2. J. Bacteriol. 115:810–815.

Mills,D.A.,L.L.McKay,and G.M.Dunny.1996.Splicing of a group II intron involved in the conjugative trans-fer of pRSO1 in lactococci. J. Bacteriol. 178:3531–3538.

Petzel, J.P., and L.L. McKay. 1992. Molecular characterization of the integration of the lactose plasmid fromLactococcus lactis subsp.cremoris SK11 into the chromosome of Lactococcus lactis subsp. lactis.Appl.Environ. Microbiol. 58:125–131.

Scherwitz-Harmon,K.M.,and L.L.McKay.1987.Restriction enzyme analysis of lactose and bacteriocin plas-mids from Streptococcus lactis subsp. diacetylactis WM4 and cloning of BclI fragments coding for bac-teriocin production.Appl. Environ. Microbiol. 53:1171–1174.

Walsh, P.M., and L.L. McKay. 1981. Recombinant plasmid associated with cell aggregation and high-frequency conjugation in Streptococcus lactis ML3. J. Bacteriol. 146:937–944.

Box 3-6. Plasmid Biology, Gene Transfer Technology, Microbial Starter Cultures, and Softball: Life in the Larry McKayLaboratory (Continued)



than batch mode. This increases throughputand generally reduces overall production costs.In addition, the cells could be recovered after afermentation and then used repeatedly (withinlimits) in subsequent fermentations. Becauseencapsulated cells are surrounded by a protec-tive, inert material, they may be more stableduring the fermentation, as well as during stor-age. In addition, cells that are encapsulatedwithin alginate beads or other matrices may beless sensitive to phage infections due to thelimited access phage would have to the cells’surface.

Despite these advantages and extensive re-search, the number of actual applications ofthese technologies are relatively few in number.For products such as cheese or yogurt, the non-fluid nature of the food medium places obviousrestrictions on the ability to pass substratethrough immobilized cell bioreactors or to re-cover encapsulated cells from fermented prod-ucts. Still, food fermentation processes areamong the applications receiving much of theattention.

One potential application of immobilizedcell technology is production of culture bio-

Box 3-7. Genetically Modified Organisms: Current Status and Future Prospects

As noted elsewhere in this chapter, as well as throughout this text, one of the most importantdevelopments in the starter culture industry has been the application of molecular biology toimprove starter culture microorganisms. It is now rather easy to modify or manipulate the phe-notype of starter bacteria or yeasts, either by the introduction of heterologous (i.e., foreign)DNA into selected organisms or by increasing expression or inactivating particular genes.Thesegenetically modified (GM) or genetically engineered organisms may then have traits, that whenused to make fermented foods, result in products having better flavor, texture,nutritional value,or shelf-life. In addition, GM organisms may possess other processing advantages, such as im-munity to bacteriophages, resistance to biological or chemical agents,and tolerance to low tem-perature and osmotic and other inhibitory conditions.

Despite the relative ease with which genetically modified organisms (GMOs) or microorgan-isms can be constructed in the laboratory, there are few actual examples of genetically modifiedstarter culture organisms that are currently used in the fermented foods industry.This is in con-trast to the GMO crops that are now fully integrated within the U.S. food supply. For example,most of the soybeans, corn, and canola produced in the United States come from geneticallymodified seeds. Moreover, even those commercially-available starter cultures that have been“modified” are not really genetically modified(i.e., made via recombinant DNA technology), butrather are derived via classical genetic techniques.They do not contain foreign DNA, and onlynatural gene exchange systems (i.e., conjugation or hybridization) are used.The notion of “nat-ural”gene transfer is worth emphasizing,since the European Union (EU) defines GMOs as beingaltered genetically via processes that do not occur naturally (Kondo and Johansen, 2002).

If GM starter culture organisms, possessing highly desirable properties, can be developed inthe laboratory, then why aren’t these cultures available in the marketplace? This simple ques-tion has many answers. First, GMO products must satisfy the regulatory requirements of thecountry or countries in which the products will be marketed. Because GMO regulations differthroughout the world (and even what constitutes a GMO varies from country to country), gain-ing approval from one jurisdiction in no way guarantees acceptance elsewhere. Moreover, theease with which approval for GMOs is obtained varies considerably. For example, in the UnitedStates, GM lactic acid bacteria can qualify as GRAS (generally recognized as safe), provided the“bioengineered”organism is not different, in any significant manner, from the original organism(i.e., it satisfies the substantial equivalence concept).

In the EU, however, the GMO application process is much more elaborate, requiring risk as-sessment and monitoring and detection plans. Labeling also is required in the European Union,



mass. Starter culture bacteria are mass pro-duced in a batch mode. Even when pH, atmos-phere and oxygen, nutrient feed, and other en-vironmental factors are properly controlled,high cell densities may be difficult to achieve.Continuous processes also do not fare well,mainly because plasmid-borne functions maybe lost and because such configurations can-not be used for co-culture or mixed strain situ-ations due to the difficulty in maintainingstrain balance. It is possible, however, to pro-duce continuous cell biomass from encapsu-lated cells. This approach takes advantage ofthe fact that cell-containing gel beads tend torelease free cells at a regular rate. Those freecells can then be collected on a continuous ba-

sis while the encapsulated cells are maintainedand contained either in an immobilized formor within in a fixed compartment.

Another application for encapsulated or im-mobilized cells is in liquid fermentations. Forexample, immobilized lactic acid bacteria canbe used in yogurt fermentations to “pre-ferment” the milk, on a continuous basis, priorto the batch incubation. In this approach,the milk pH is reduced to 5.7 during the pre-fermentation step, allowing an overall 50%reduction in the total fermentation time. Like-wise, immobilized lactic acid bacteria can beused to produce industrial lactic acid andother organic end-products from whey, wheypermeates, or corn-based feedstocks. Other

whereas no such requirement exists in the United States for most GMO foods.The labeling issueis not a minor point—in fact, it may be a deal-breaker, because many consumers are either op-posed to GMOs or confused enough by the GMO declaration on the label that it influencestheir purchasing decisions.

Although it would appear that the U.S.market (and perhaps others) would be more receptiveto GM cultures than the EU market, it must be recognized that most foods or food ingredients,including starter cultures, are marketed globally. Thus, it seems likely that a culture supplierwould commit resources to development of GM microorganisms only if approval on a nearworldwide basis is expected. Nonetheless, there are some cultures, derived using moleculartechniques, that are sold in the United States but are not marketed in Europe.

For example, a high diacetyl-producing derivative strain of Lactococcus lactis (MC010), ob-tained via a recombinant plasmid-mediated mutagenesis procedure, is now widely used in theUnited States while its approval in the EU is still being considered (Curic et al., 1999; Pedersenet al.,2005). Importantly, this strain contains no foreign DNA and is identical to the parent strainexcept for a mutation in a single gene. Moreover, there are now a variety of molecular ap-proaches that can be adopted to construct food-grade GM starter culture organisms. Plasmidvectors, for example, that are comprised of only lactococcal DNA with natural selection mark-ers (and devoid of antibiotic-resistance genes), are now available. It remains to be seen, how-ever, whether these so-called “self-cloning” strategies will result in greater acceptance by bothregulatory agencies and consumers.

References

Curic, M., B. Stuer-Lauridsen, P. Renault, and D. Nilsson. 1999.A general method for selection of �-acetolac-tate decarboxylase-deficient Lactococcus lactis mutants to improve diacetyl formation.Appl. Environ.Microbiol. 65:1202–1206.

Kondo, J.K., and E. Johansen. 2002. Product development strategies for foods in the era of molecularbiotechnology.Antonie van Leeuwenhoek 82:291–302.

Pedersen, M.B., S.L. Iversen, K.I. Sørensen, and E. Johansen. 2005.The long and winding road from the re-search laboratory to industrial applications of lactic acid bacteria. FEMS Microbiol. Rev. 29:611–624.

Box 3-7. Genetically Modified Organisms: Current Status and Future Prospects (Continued)



value-added fermentation products that couldbe produced on a continuous basis by immobi-lized cells include exopolysaccharides (Chap-ter 4), bacteriocins, and diacetyl.

Perhaps the encapsulated cell applicationswith the greatest potential are those involvingalcoholic fermentations, especially for wineproduction. Indeed, several such products, in-cluding both encapsulated yeasts and bacteria,are available commercially for use in specificproduct applications. Encapsulated strains of S.cerevisiae, for example,can be used in the man-ufacture of sweet wines (whose manufacture isdescribed in Chapter 10).These wines, in con-trast to dry wines, contain residual or unfer-mented sugar, due to cessation of the fermenta-tion via a decrease in temperature to nearfreezing,addition of sulfites,or removal of yeastby filtration. In contrast, bead-encapsulatedyeasts can be placed in nylon bags which arethen added to the must, such that when the de-sired sugar concentration is reached, the bags(and yeasts) can simply be removed. Encapsu-lated yeast cells can also be used to re-startstuck or sluggish wine fermentations and tometabolize malic acid to ethanol. Similarly, en-capsulation and immobilization of the malolac-tic bacterium Oenococcus oeni has been pro-posed to be an effective way to reduce malicacid and deacidify wine on a continuous basis.

Finally, it is relevant to mention that encap-sulated starter cultures, in an entirely naturalform, have long been used in the manufactureof kefir, a fermented dairy product widely con-sumed in Russia and Eastern Europe (Chapter4).The key feature of this product is that it istraditionally made by inoculating milk with aculture contained within inert particles calledkefir grains.These irregularly-shaped particles,which can be as large as 15 mm in diameterand consist of polysaccharide and milk pro-tein,harbor a complex microflora. Included arehomo- and heterofermentative lactic acid bac-teria,as well as several yeasts.The grains can beseparated from the fermented kefir by filtra-tion or sieving, and then washed and re-usedfor subsequent fermentations. Although kefircultures, which contain pure strains of Lacto-

bacillus spp. and Lactococcus spp., are nowavailable commercially and are often used inthe United States for kefir manufacture, tradi-tional kefir grains still remain popular in manyparts of the world.

Probiotics and Cultures Adjuncts

In addition to producing starter cultures forspecific food fermentations, the starter cultureindustry also manufactures microbial productsfor many other applications. For example,there is now a large market for probiotic mi-croorganisms in both fermented and non-fermented foods. As described in detail inChapter 4, probiotics are defined as “live mi-croorganisms which when administered in ad-equate amount confer a health benefit on thehost.” The main probiotic organisms includespecies of Bifidobacterium and Lactobacillus,although Saccharomyces and Bacillus sp., andeven E. coli, are also commercially available asprobiotics. In general, these organisms are pro-duced industrially in the same manner asstarter cultures (i.e., under conditions thatmaximize cell density). However, because theyare not used to carry out a subsequent fermen-tation, these culture products are of the direct-to-vat type, rather than in bulk culture form.

Another group of culture products used bythe cheese industry consists of lactic acid andrelated bacteria whose function is to acceler-ate and enhance cheese ripening and matura-tion. Specifically, these organisms, usuallystrains of Lactobacillus helveticus and Lacto-bacillus casei, produce peptidases and otherprotein-hydrolyzing enzymes that are neces-sary for proper flavor and texture develop-ment. Citrate-fermenting LAB that produce theflavor compound diacetyl as well as heterofer-mentative LAB that produce carbon dioxidemay also be added as adjunct cultures for spe-cific cheeses.

Finally,adjunct cultures and culture prepara-tions are now available that serve a food safetyand preservation function.That is, the microor-ganisms contained in these products do notnecessarily make fermentation acids or modify



texture or flavor, but rather they are includedin the culture because they inhibit pathogenicor spoilage organisms. The inhibitory activityof these organisms may be due to one of sev-eral substances, including hydrogen peroxide,organic acids, diacetyl, and a class of inhibitorycompounds called bacteriocins. The latter areproteinaceous, heat-stable materials producedby a given organism that inhibits other closelyrelated organisms (Chapter 6).

Lactic acid bacteria are prolific producersof bacteriocins, with all of the food-relatedgenera capable of producing bacteriocins.Therefore, they can be used either as part of alactic acid-producing starter culture or as anadjunct in dairy, meat, and other foods to in-hibit pathogens and spoilage organisms and toenhance shelf-life. Furthermore, some of theseorganisms used as adjuncts are capable of pro-ducing bacteriocin in the food, but with onlyminor production of acids or other fermenta-tion end-products.Thus, the sensory character-istics of the product are not affected by theproducer organism, a property that would beespecially important in non-fermented foodssuch as ready-to-eat meats.

Another means by which bacteriocins can be introduced into a food without adding live organisms has also been developed. In thisprocess, the producer organism is grown in adairy- or non-dairy-based medium, and then thespent fermentation medium is harvested, pas-teurized, and concentrated.This material wouldcontain the bacteriocin (as well as organicacids), which could then be added to foods as anatural preservative. Although these so-calledbioprotective products are mainly effectiveagainst Gram positive bacteria, some commer-cially-available products also inhibit yeasts,mold,and Gram negative spoilage bacteria, includingpsychrotrophs that spoil refrigerated foods.

The Starter Culture Industry

Although starter cultures are certainly themain product line, the starter culture industry,as noted previously,produces and markets a va-riety of other products. Dairy culture suppli-

ers, in particular, offer culture media, coagu-lants, colors, and other ancillary products usedin cheese manufacture. Starter culture compa-nies also provide technical services and sup-port, perhaps more so than any other ingredi-ent suppliers.This is because the manufactureof cheese, sausage, wine, and all fermentedproducts depends on the activity of an inher-ently unstable biological material, i.e., thestarter culture.Although there may certainly bequality issues that arise during the manufac-ture of corn flakes, canned peas, granola bars,or other non-fermented foods, the failure of abiologically-dependent process is not one ofthem.A cheese plant operations manager whois responsible for converting 2 million Kg ofmilk into 200,000 kg of cheese (worth at least$250,000) each day will not wait long beforecalling the culture supplier the minute he orshe suspects a culture problem. Thus, a well-trained support staff is a necessary componentof the modern culture industry.

Culture suppliers are also often asked toprovide customized cultures for specific appli-cations.This means that the phenotype (or insome cases, even the genotype) of each strainin an industrial culture collection must be cata-loged. Fermentation rates, product formationand enzyme profiles, sugar fermentation pat-terns, osmotic- or halo-tolerance behavior, andflocculation properties (for yeasts) should beincluded in the strain bank data. In the case oflactic acid bacteria and beer and wine yeasts,sensitivities to phages and killer yeasts, respec-tively, must be determined for each strain.When multiple strains are used in cultureblends, production of or sensitivity to antago-nism factors must be known.

Finally, the starter culture industry has al-ways been a major contributor to the researchon lactic acid bacteria, yeast, and fungi. Severalof the large culture companies have main-tained active research and development pro-grams that conduct basic and applied researchthat has led directly to major findings on thegenetics and physiology of starter culture or-ganisms. In other cases, their contribution hasbeen indirect, namely by providing research



funding, strains, and other materials to investi-gators at universities and research institutes.Industry sponsorship of conferences and sci-entific societies have also served an importantfunction in the advancement of starter cultureresearch.

BibliographyBuckenhüskes, H.J. 1993. Selection criteria for lactic

acid bacteria to be used as starter cultures forvarious food commodities. FEMS Microbiol. Rev.12:253–272.

Caplice, E., and G.F. Fitzgerald. 1999. Food fermenta-tions: role of microorganisms in food productionand preservation. Int. J. Food Microbiol. 50:131–149.

Cogan, T.M. 1996. History and taxonomy of startercultures, p. 1—23. In T.M. Cogan and J.-P.Accolas(ed.). Dairy Starter Cultures. VCH Publishers,Inc., New York, New York.

De Vuyst, L. 2000.Technology aspects related to theapplication of functional starter cultures. FoodTechnol. Biotechnol. 38:105–112.

Hammes W.P., and C. Hertel. 1998. New develop-ments in meat starter cultures. Meat Sci. 49 (sup-plement):S125–S138.

Mäyrä-Mäkinen A, and M. Bigret. 1999. Industrial useand production of lactic acid bacteria, p. 175—198. In S. Salminen,A. von Wright, and A. Ouwe-hand (ed.) Lactic Acid Bacteria. Microbiologicaland functional aspects. 3rd Ed., Marcel DekkerInc., New York, New York.

Pedersen, M.B., S.L. Iversen, K.I. Sørensen, and E.Johansen. 2005.The long and winding road fromthe research laboratory to industrial applicationsof lactic acid bacteria. FEMS Microbiol. Rev.29:611–624.

Ross, R.P., C. Stanton, C. Hill, G.F. Fitzgerald, and A.Coffey. 2000. Novel cultures for cheese improve-ment.Trends Food Sci.Technol. 11:96–104.

Sandine,W.E. 1996. Commercial production of dairystarter cultures, p. 191—206. In T.M. Cogan and J.-P. Accolas (ed.) Dairy Starter Cultures. VCHPublishers, Inc., New York, New York.


4

Cultured Dairy Products

“. . . during recent years, attention has been directed to soured milk to such an ex-tent that it has become necessary for all who are interested in the handling of milkand milk products to have a knowledge of the subject, as it seems clearly demon-strated that, under proper direction, there is every possibility of its forming an im-portant element in the prolongation of life.”

From The Bacillus of Long Life by Loudon Douglas, 1911

although growth of acid-tolerant yeast andmolds is possible, growth of pathogens rarelyoccurs.

Given the early recognition of the impor-tance of milk in human nutrition and its wide-spread consumption around the world, it is notsurprising that cultured dairy products haveevolved on every continent.Their manufacturewas already well established thousands of yearsago, based on their mention in the Old Testa-ment as well as other ancient religious texts andwritings.Yogurt is also mentioned in Hindu sa-cred texts and mythology.And although the man-ufacturing procedures, the sources of milk, andthe names of these products may vary consider-ably, they share many common characteristics.Thus, dahi (India), laban (Egypt, Lebanon), andjugart (Turkey) are all yogurt-like productswhose manufacture involves similar milk han-dling procedures and depends on the same ther-mophilic culture bacteria. Other products, inparticular, kefir and koumiss, evolved from Asia,and are made using various lactose-fermentingyeasts in addition to lactic acid bacteria.

Consumption of Cultured Dairy Products

Yogurt, sour cream (and sour cream-baseddips), and cultured buttermilk are the mostpopular cultured dairy products in the United

Introduction

Milk fermentations must undoubtedly beamong the oldest of all fermented foods. Milkobtained from a domesticated cow or camel orgoat, some thousands of years ago, would havebeen fermented within hours by endogenouslactic acid bacteria, creating a yogurt-like prod-uct. In fact, the ability to maintain milk in afresh state before souring and curdling had oc-curred would have been quite some trick, es-pecially in warm environments. Of course, fer-mentation and acid formation would havebeen a good thing since, in the absence of vi-able lactic acid bacteria, other bacteria, includ-ing pathogens, could have grown and causedunpleasant side effects.

Milk is particularly suitable as a fermenta-tion substrate owing to its carbohydrate-rich,nutrient-dense composition. Fresh bovinemilk contains 5% lactose and 3.3% protein andhas a water activity near 1.0 and a pH of 6.6 to6.7, perfect conditions for most microorgan-isms. Lactic acid bacteria are saccharolytic andfermentative, and, therefore, are ideally suitedfor growth in milk. In general, they will out-compete other microorganisms for lactose,and by virtue of acidification, will produce aninhospitable environment for would-be com-petitors. Therefore, when properly made, cul-tured dairy products have long shelf-lives and,

107





States.Yogurt accounts for more than half of allcultured dairy products consumed in theUnited States (Figure 4–1), with nonfat andlow-fat versions being the most popular (about90% of total yogurt sales). Per capita consump-tion of buttermilk in the United States has de-creased nearly 50% in the past twenty years,but consumption of sour cream (includingsour cream-based dips) has doubled and yo-gurt has tripled during that same time.

Nonetheless, total per capita consumptionfor all cultured dairy products in 2003 in theUnited States was less than 6.5 Kg (about 14pounds),whereas in some European countries,yogurt consumption alone is more than 20 Kgper person per year—the equivalent of ninety8-ounce cups (Table 4–1). Furthermore, in theNetherlands,France, and other European coun-tries, the dairy sections of food markets andgrocery stores contain numerous other tradi-tional as well as new forms of cultured milkand cream products. Many of these new prod-uct trends are beginning to catch on in theUnited States, and now kefir, fluid yogurts,

crème fraîche (a 50% fat sour cream), andother new cultured dairy products are readilyavailable in the market place.

Cultured dairy products and probiotic bacteria

Among the factors contributing to the in-creased consumption of yogurt and relatedproducts are the positive nutritional benefitsthese products are believed to provide.Throughout the world, yogurt, and specifically,the live cultures present in yogurt, have beenpromoted on the basis of the many health ben-efits these bacteria are believed to confer. Al-though the bacteria that comprise the normalyogurt starter culture, Streptococcus thermo-philus and Lactobacillus delbrueckii subsp.bulgaricus, may indeed enhance gastrointesti-nal health (to be discussed later), yogurt hasbecome widely used as a vehicle for delivery ofother microorganisms not ordinarily found inyogurt that may also improve human health.Such health-promoting bacteria are referred to

0

1

2

3

4

1980 1985 1990 1995 2000

raey rep nosrep rep gk

Yogurt

Buttermilk

Sour Cream

2003Figure 4–1. Consumption of cultured dairy products in the United States, 1980 to 2003. (Data for sourcream included sour cream dips.) From USDA, Economic Research Service data.


Cultured Dairy Products 109

as probiotics, organisms that confer a healthbenefit to the host (Box 4–1).

For example,Lactobacillus acidophilus andvarious species of Bifidobacterium are com-monly added to yogurt as probiotics, eventhough these organisms are not part of the yo-gurt starter culture, nor are they involved inany way in the actual yogurt fermentation.Their only contribution is as a culture adjunctintended to promote good health. Still, accord-ing to Dairy Management Inc., about 80% ofthe yogurt products produced in the UnitedStates contain L. acidophilus.The mechanismsby which these suggested health benefits actu-ally occur and the evidence to support theseclaims will be discussed later in this chapter.

In addition to probiotic bacteria, a groupof compounds called prebiotics are also nowbeing added to cultured milks—mainly yo-gurt and kefir in the United States, but manymore products in Europe, Japan, and Korea.Prebiotics are oligosaccharide-containing ma-terials that are neither degraded nor ab-sorbed during transit through the stomach orsmall intestine (Box 4–1).They end up in thecolon, where they are preferentially fer-mented by intestinal strains of lactobacilli

and bifidobacteria.Thus, they enrich the pop-ulation of those bacteria thought to con-tribute to gastrointestinal health.

Probiotics can now be found not only in yo-gurt, but also in a variety of other cultured andnon-cultured dairy products (Table 4–2). Insome of these products, as noted above, theprobiotic bacteria are not involved in the fer-mentation. However, some new products havebeen developed to take advantage of their fer-mentative ability as well as their probiotic ac-tivity. Examples include Yakult, a Japaneseproduct made using a special strain of Lacto-bacillus casei (strain Shirota) that was origi-nally isolated from the human intestinal tract,and Cultura, a European “bioyogurt”made withL. casei F19 (also a human isolate).

At the same time, there are also dairy prod-ucts containing probiotic lactic acid bacteriathat are not fermented but serve strictly as acarrier. The product known as Sweet Aci-dophilus Milk, for example, is simply fluid milksupplemented with L.acidophilus (added afterpasteurization). Similar unfermented productsmay also contain other probiotic bacteria.Maintaining the product at low refrigerationtemperatures (the same as for normal fluidmilk products) prevents the bacteria from fer-menting and souring the milk.

Fermentation Principles

The principles involved in the manufacture ofspecific cultured dairy products, which lacticacid bacteria are used for these products, andthe attributes and properties desired by thesebacteria are the focus of this chapter. Fer-mented dairy products from around the world,including India,Africa, Iceland, and other coun-tries, will be described, but the emphasis willbe on those products most widely consumedin the United States. In theory (and practice),some of these products can be made in the ab-sence of a starter culture simply by addingfood-grade acidulants to the milk mixture. In-deed, there are manufacturing advantages forfollowing this practice; however, these acid-setproducts are not fermented—in fact, they must

Table 4.1. Per capita consumption of yogurt1.

Country Kg per person per year

Australia (2003) 5.9Austria (1999) 8.2Bulgaria (2002) 24.7Canada (2003) 5.8Denmark (1999) 14.8Finland (2001) 17.0France (2001) 14.5Germany (2004) 16.8Italy (1999) 6.5Netherlands (2003) 21.2Spain (1999) 7.1Sweden (2002) 28.5Switzerland (1999) 13.0Tunisia (2002) 8.0Turkey (2004) 36.0Russia (2001) 1.5United Kingdom (1999) 4.8United States (2003) 3.4

1Data obtained from multiple sources for the given year



Box 4–1. Probiotics and Prebiotics

History

Humans began consuming cultured dairy products thousands of years ago.These yogurt-likeproducts were easy to produce, had good shelf-lives (so to speak), were free of harmful sub-stances,and had a pleasant sensory appeal.At some time, it is theorized, they were also regardedas having therapeutic value, even though there was no scientific basis for this notion (Shortt,1999).After all, the existence of bacteria and their role in fermentation weren’t even recognizeduntil Pasteur’s research in the 1860s.Then, at the beginning of the twentieth century, the Russ-ian scientist Elie Metchnikoff (who was working at the Pasteur Institute in Paris) suggested thatthe health benefits of fermented milk were due to the bacteria involved in the fermentation.

Specifically, Metchnikoff argued that these bacteria (which may have been the yogurt bacte-ria, Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus or other lac-tobacilli) inhibited putrefactive bacteria in the intestinal tract, thereby influencing the intestinalmicroflora such that overall health and longevity could be enhanced.At around the same timeand shortly thereafter, other scientists (as recounted by Shortt, 1999) reported that consump-tion of other lactobacilli (including Lactobacillus acidophilus) and bifidobacteria also had pos-itive health effects, including reducing the rate of infant diarrhea.

Interestingly, the observation a full century ago that bifidobacteria were present in the fecalcontents of breast-fed infants suggested that these bacteria were associated with good intestinalhealth and possibly foreshadowed the concept of prebiotics (see below).These early reports,byhighly regarded scientists and research institutes, attracted the attention of the medical com-munity, and by the 1920s, studies using bacteria therapy (with milk as the carrier vehicle) hadbegun. Unfortunately, many of these subsequent studies suffered from the absence of estab-lished measurement criteria, the use of mis-identified strains, and other design flaws (Shortt,1999).As described later in this chapter, in more recent years, these experimental limitationshave been recognized and addressed and now rigorous and appropriate methodologies are be-ing used (Reid et al., 2003).

Definitions

The term “probiotics,”as originally used in 1965, referred to the “growth-promoting factors”pro-duced by one microorganism that stimulated growth of another (i.e., the opposite of antibi-otics).This definition went through several permutations and by 1989, probiotics were definedas a “live microbial feed supplement which beneficially affects the host animal by improving itsintestinal balance” (Fuller, 1989).The current definition, adopted by the World Health Organiza-tion (part of the United Nations’ Food and Agriculture Organization), defines probiotics as “livemicroorganisms which when administered in adequate amounts confer a health benefit on thehost.”This latest derivation is important because it recognizes both the relevance of a live andsufficient dose as well as the many reports that indicate probiotics may have health benefitsthat extend beyond the gastrointestinal tract.Whether all of the benefits ascribed to probiotics(Table 1) can be validated is another matter, however, as discussed later in this chapter.

Prebiotics are defined as “non-digestible food ingredient(s) that beneficially affects the hostby selectively stimulating the growth and/or activity of one or a limited number of bacteria inthe colon, and thus improves host health” (Gibson and Roberfroid, 1995).To re-phrase this defi-nition, but with more detail, prebiotics are carbohydrate substances that escape digestion andadsorption in the stomach and small intestine and instead reach the colon.There, they are se-lectively metabolized by specific members of the colonic microflora. Prebiotics, therefore, en-rich the population of those generally desirable commensal organisms such as lactobacilli andbifidobacteria at the expense (theoretically) of their less desirable competition.



Table 1. Suggested health benefits of probiotic bacteria.

Reduce blood cholesterolMaintain intestinal healthAlleviate intestinal bowel diseasesModulate immune systemReduce incidence of gastrointestinal infectionsReduce incidence of urinary and vaginal infectionsAlleviate lactose intoleranceAnti-carcinogenic and anti-tumorogenicReduce incidence and severity of diarrheal diseases

Most of the prebiotics that are used commercially and that have received the most researchattention are either polysaccharides or oligosaccharides.They are either derived from plant ma-terials or are synthesized from natural disaccharide precursors (Figure 1). For example, inulin is

Box 4–1. Probiotics and Prebiotics (Continued)

CH2OHO

CH2

O

O

CH2

OHOH2C

CH2OH

OHOH2C

G

F

F

CH2OHO

CH2

O

O

CH2

OHOH2C

CH2OH

OHOH2C

n

G

F

F

OHOH2C

CH2OH

CH2OHO

CH2

O

O

CH2

OHOH2C

CH2

OHOH2C

OHOH2C

CH2OH

O

O

G

F

F

F

F

OHOH2C

CH2OH

CH2OHO

CH2

O

O

CH2

OHOH2C

CH2

OHOH2C

O

G

F

F

F

CH2OH

O

CH2OH

O OCH2OH

OO H OHGGaGa

CH2OH

O OCH2 CH2OH

OOO H OHG

GaGa

Figure 1. Structure of prebiotics. Shown (upper panel, left-to-right) are short chain fructooligosaccharides(FOS), containing two, three, or four fructose units linked �-1,2 and with a terminal glucose. On the far right,longer chain FOS are shown, where n can equal up to twenty or more fructose units. The lower panel showstwo forms of galactooligosaccharides (GOS), with galactose units linked �-1,4 (left) or �-1,6 (right), bothlinked to terminal glucose units.

(Continued)



a naturally-occurring plant polysaccharide (consisting of fructose units, linked �-1,2 with a ter-minal glucose residue) that can be used in its intact form or as a mixture of partially hydrolyzedfructooligosaccharide (FOS) molecules.The latter can also be synthesized from sucrose via atransfructosylating enzyme that adds one, two, or three fructose units to the sucrose backbone.

Another type of prebiotic oligosaccharide that has attracted considerable attention are thegalactooligosaccharides (GOS). These oligosaccharides are built from lactose via addition ofgalactose residues by �-galactosidases with high galactosyltransferase activity. Galactooligosac-charides are arguably the most relevant prebiotics being used in foods,since the GOS moleculesclosely resemble the oligosaccharides found in human milk.These human milk oligosaccharides(which also exist in milk from other species, but usually at lower levels) are now widelythought to be responsible for the bifidogenic properties associated with human milk. In fact, ithad long been suggested that there was something in human milk that promoted growth of bi-fidobacteria (the so-called “bifidus” factor) and that this factor accounted for the dominance ofthese bacteria in the colon of nursed infants.That infants fed mother’s milk suffered fewer in-testinal infections and were generally healthier than formula-fed infants provided circumstantialevidence that having a greater proportion of bifidobacteria (and perhaps lactobacilli) in thecolon would be desirable, not just for infants, but for the general population as well.

References

Fuller, R. 1989. Probiotics in man and animals. J.Appl. Bacteriol. 66:365–378.Gibson,G.R., and M.B.Roberfroid.1995.Dietary modulation of the human colonic microbiota: introducing

the concept of prebiotics. J. Nutr. 125:1401–1412.Lilly, D.M., and R.H. Stillwell. 1965. Probiotics: growth-promoting factors produced by microorganisms. Sci-

ence 147:747–748.Reid, G., M.E. Sanders, H.R. Gaskins, G.R. Gibson,A. Mercenier, R. Rastall, M. Roberfroid, I. Rowland, C. Cher-

but, and T.R. Klaenhammer. 2003. New scientific paradigms for probiotics and prebiotics. J. Clin. Gas-troenterol. 37:105–118.

Shortt, C. 1999. The probiotic century: historical and current perspectives. Trends Food Sci. Technol.10:411–417.

Box 4–1. Probiotics and Prebiotics (Continued)

Table 4.2. Commercial probiotic organisms used in dairy products1.

Organism Supplier or source

Lactobacillus acidophilus NCFM Danisco, Madison,WI, USALactobacillus acidophilus SBT-2062 Snow Brand Milk Products,Tokyo, JapanLactobacillus casei strain Shirota Yakult,Tokyo, JapanLactobacillus casei F19 Arla Foods, Skanderborgvej, DenmarkLactobacillus fermentum RC-14 Urex Biotech, London, CanadaLactobacillus gasseri ADH Danisco, Madison,WI, USALactobacillus johnsonii KA1 (NCC 533) Nestle, Lausanne, SwitzerlandLactobacillus plantarum 299v Probi, Lund, SwedenLactobacillus reuteri SD2112 (ATCC 55730) Biogaia, Stockholm, SwedenLactobacillus rhamnosus GR-1 Urex Biotech, London, CanadaLactobacillus rhamnosus GG (ATCC 53103) Valio Ltd., Helsinki, FinlandLactobacillus salivarius UCC118 University College, Cork, IrelandBifidobacterium longum SBT-2928 Snow Brand Milk Products,Tokyo, JapanBifidobacterium longum BB536 Morinaga Milk Industry, Zama City, JapanBifidobacterium breve strain Yakult Yakult,Tokyo, Japan

1Adapted from Reid, 2001 and Sanders, 1999.



be labeled as “directly set” to denote this man-ner of acidification—and often lack the flavorof the fermented versions.They certainly lackthe nutritional benefits that may be con-tributed by live active cultures.

The function of lactic acid bacteria in themanufacture of cultured dairy products isquite simple—they should ferment lactose tolactic acid such that the pH decreases and theisoelectric point of casein, the major milk pro-tein, is reached. By definition, the isoelectricpoint of any protein is that pH at which thenet electrical charge is zero and the protein isat its minimum solubility. In other words, asthe pH is reduced, acidic amino acids (e.g., as-partic acid and glutamic acid), basic aminoacids (e.g., lysine and arginine), and partialcharges on other amino acids become proto-nated and more positive such that at somepoint (i.e., the isoelectric point), the totalnumber of positive and negative charges onthese amino acids (as well as on other aminoacids) are in equilibrium.

For casein, which ordinarily has a negativecharge, the isoelectric point is 4.6.Thus, whensufficient acid had been produced to over-come the natural buffering capacity of milkand to cause the pH to reach 4.6, casein pre-cipitates and a coagulum is formed.Along theway, the culture may also produce other smallorganic molecules, including acetaldehyde, di-acetyl, acetic acid, and ethanol.Although theselatter compounds are produced in relativelylow concentrations, they may still make impor-tant contributions to the overall flavor profileof the finished product. The culture may alsoproduce other compounds that contribute tothe viscosity, body, and mouth feel of the prod-uct (see below). The choice of culture, there-fore, is dictated by the product being pro-duced, since different cultures generate flavorand aroma compounds specific to that product(Table 4–3).

The actual manufacture of cultured dairyproducts requires only milk (skim, lowfat, orwhole, or cream, depending on the product)and a suitable lactic starter culture. However,the process is not quite so simple, because de-

fects associated with flavor, texture, and ap-pearance are not uncommon. As will be de-scribed later, the most frequent and seriousproblem in the manufacture of many of theseproducts, especially yogurt, is syneresis.

Syneresis is defined as the separation (orsqueezing out) of water from the coagulatedmilk. To many consumers, the appearance ofthese pools of slightly yellow-green water(which is actually just whey) from the top of theproduct is considered unnatural and objection-able.Thus, to minimize syneresis problems, andto improve the body and texture of the finishedproduct, manufacturers perform several stepswhose purpose is to enhance the water bindingcapacity of the milk mixture. First, the milksolids are increased, either by adding dry milkpowder or by concentrating milk. Second, themilk mixture is heated well above ordinary pas-teurization conditions to denature the wheyproteins, exposing more amino acid residues tothe aqueous environment. Finally, most manu-facturers have incorporated stabilizers, thicken-ing agents,and other ingredients into the formu-lation to further reduce syneresis. However, afew countries, France, in particular, prohibitmany of these additional ingredients and instead

Table 4.3. Organisms used as starter cultures in themanufacture of fermented dairy products.

Product Organisms

Yogurt Streptococcus thermophilusLactobacillus delbreckii subsp.

bulgaricus

Buttermilk Lactobacillus lactis subsp. lactisLactobacillus lactis subsp. cremorisLeuconostoc lactisLeuconostoc mesenteroides subsp.

dextranicum

Sour Cream Lactobacillus lactis subsp. lactisLactobacillus lactis subsp. cremorisLeuconostoc lactisLeuconostoc mesenteroides subsp.

dextranicum

Kefir Lactobacillus kefiriLactobacillus kefiranofaciensSaccharomyces kefiri



rely on other means to reduce syneresis (dis-cussed below).

Yogurt Manufacture

Milk treatment

Yogurt can be made from skim (non-fat), re-duced fat, or whole milk.As is true for all dairyproducts, but especially so for yogurt andother cultured milks, it is important to usegood quality milk, free of antibiotics and otherinhibitory substances.The first step (Figure 4–2) involves adding nonfat dry milk to the milkto increase the total solids to 12% to 13%,sometimes to as high as 15%.Alternatively, thetotal milk solids can be increased by concen-trating the milk via evaporation. Other permit-ted ingredients (see below) may be added, andthe mix (only if it contains fat) is then usuallyhomogenized, although there are some spe-cialty manufacturers that produce unhomoge-nized, cream-on-the-top whole milk versions.

Yogurt is considered a fluid milk product bythe U.S. Food and Drug Administration (FDA)and must be made using pasteurized milk.However, most yogurt mixes receive a heattreatment well above that required for pasteur-ization. Thus, instead of pasteurizing milk for71.7°C for 15 seconds (the minimum required),mixes are heated to between 85°C and 88°C forup to 30 minutes. Other time-temperature con-ditions can also be used,but kinetically they areusually equivalent. Heating can be done inbatch mode (i.e., in vats), but continuous heat-ing in plate or tube type heat exchangers is farmore common.

The high temperature treatment not onlysatisfies all of the normal reasons for pasteur-ization (i.e., killing pathogens and spoilage or-ganisms and inactivating enzymes), but thesesevere heating conditions also perform two ad-ditional functions.First,even heat-resistant bac-teria and their spores are killed, making themixture essentially free of competing microor-ganisms. Second, the major whey proteins, �-lactalbumin and �-lactoglobulin, are nearly100% denatured at the high pasteurization

temperatures.These proteins exist in globularform in their native state but once denatured,amino acid residues are exposed and their abil-ity to bind water, via hydrogen bonding, is sig-nificantly enhanced. Denatured whey proteinsalso reduce the Eh and stabilize the milk gel.

Yogurt styles

The pasteurized milk is then cooled via a heatexchanger to the desired incubation tempera-ture, usually between 40°C and 43°C. Alterna-tively, the milk can be cooled as for conven-tional processing to 2°C to 4°C, and thenwarmed to the higher temperature later. Theincubation temperature is critical, since it willinfluence the activity of the culture and ulti-mately the properties of the finished yogurt(see below).

The route the mix takes next depends onthe type or style of yogurt being made.Thereare two general types of yogurt.Yogurt that ismixed with flavors, fruit, or other bulky ingre-dients is called stirred or Swiss-style yogurt.Forthis type, the mix is pumped into vats and theculture is added.The mixture is then incubatedsuch that the entire fermentation occurs in thevat.At the end of the fermentation, the mixtureis gently agitated and cooled, and the flavor in-gredients are introduced. The mixture is thenpumped into containers.

In contrast, mix can be inoculated with cul-ture, pumped immediately into the container,and then fermented directly in the container. Ifthis so-called fermented-in-the-cup style yogurtis to contain fruit or other bulky flavoring (i.e.,fruit-on-the-bottom or sundae-style), the fruitor flavoring material is first dispensed into thecup and the yogurt mix added on top, followedby incubation and fermentation.The consumermust do the stirring and mixing to incorporatethe flavoring throughout the product.

Yogurt cultures

For either style, the lactic culture represents acritical ingredient and must be carefully se-lected. Yogurt cultures, in general, and the ac-



tual strains, in particular, have a profound influ-ence on the flavor, texture, appearance, andoverall quality attributes of the finished prod-uct (Box 4–2).As noted above, starter culturesfor yogurt consist of two organisms, L. del-brueckii subsp. bulgaricus and S. thermo-philus. Many commercial products containmore than one strain of each organism, but thetraditional ratio of L.delbrueckii subsp.bulgar-icus and S. thermophilus (or rod to coccus)

has usually been about 1:1.Although this ratiocan be adjusted, depending on the desiredproperties of the final product, it is importantthat both organisms be present.

Yogurt made with only one of these twoorganisms will not turn out well, for severalreasons. In fact, it has long been known thatthese bacteria grow faster and perform betterwhen grown as a pair compared to whenthey are grown separately. The basis for this

Milk

Yogurt Mix

Heat Treatment

Cool

Inoculate

Dispense Incubate

Cool

Mix and Dispense

Homogenize

85°C for 30 minutes

40-45°C

Standardize (nonfat, lowfat, whole)

Nonfat milk solids

Stabilizers, other ingredients

2.0 – 2.5% thermophilic culture

Incubate

40-45°C, 3 to 6 hours

0-4°CCool

40-45°C

Swiss-StyleCup-Style

Figure 4–2. Manufacture of yogurt.



Box 4–2. Selecting Yogurt Cultures

Criteria used for selecting yogurt starter culture strains are based on several general perfor-mance characteristics, as well as other traits that provide specific functions (Table 1). Generalrequirements include vigorous and rapid growth, such that the culture is able to lower the pHto the target level within four to six hours.The culture should produce good flavor without bit-terness or excess acidity.Resistance to bacteriophages is also expected.The starter culture mustalso be tolerant of storage conditions, whether frozen or lyophilized, so that viability is retainedand long lag phases are avoided.

Table 1. Desired properties of yogurt cultures.

Stable during frozen or lyophilized storage Able to produce the “right”consistency or body, i.e., ropiness Viable and active after thawing or rehydration Produces good yogurt flavor, without excess acid or acetaldehydePrompt growth and fermentation; short lag phase No syneresisResistant to bacteriophage No acidity produced during storage (over-acidification)

There are also several specific characteristics expected from the culture, depending on thewants and needs of the manufacturer. For example, cultures may be selected on the basis ofwhether they produce viscous yogurts with strong body or little viscosity and a relatively mildbody. High-body yogurts are generally made using exopolysaccharide-producing cultures, as de-scribed elsewhere.

In addition to the use of customized cultures for body characteristics, there is also a prefer-ence among yogurt manufacturers in the United States to use cultures that produce so-calledmild flavored yogurts (Ott et al., 2000).These cultures make enough acid to cause milk coagula-tion, but not so much that the pH drops below 4.4 to 4.5. In contrast, traditional acidic yogurtsoften reach a pH of 4.0 or even lower.The less acidic yogurts are not only preferred by manyconsumers, but if probiotic adjuncts are added to the yogurt, their rate of survival during stor-age is generally enhanced at the higher pH.

Of the two organisms used for yogurt fermentations, Streptococcus thermophilus and Lacto-bacillus delbrueckii subsp. bulgaricus, it is the latter that is usually (but not always) responsi-ble for excessive acid production.This organism grows well at 42°C to 45°C, so one effectiveway to limit acid production is simply to incubate the yogurt at lower incubation temperatures(around 37°C). Alternatively, L. delbrueckii subsp. bulgaricus can be omitted altogether fromthe culture blend and replaced with a different strain (e.g., Lactobacillus casei) that may pro-duce less acid. In addition, it is possible to select for natural acid-sensitive derivative strains (thatnecessarily stop making acid at some particular low pH). Presumably, some of the commercialyogurt cultures promoted as being low acid or mild-flavor cultures contain these strains. Simi-larly, strains that are sensitive to low temperature could also be used,since they would cease fer-mentation when the product was shifted to refrigeration temperature.One example of this par-ticular approach was the isolation, via chemical mutagenesis, of a L. delbrueckii subsp.bulgaricus variant that synthesized a “cold-sensitive”�-galactosidase (Adams et al., 1994).

References

Adams, R.M., S.Yoast, S.E. Mainzer, K. Moon,A.L. Polombella, D.A. Estell, S.D. Power, and R.F. Schmidt. 1994.Characterization of two cold-sensitive mutants of the �-galactosidase from Lactobacillus delbrueckiisubsp. bulgaricus. J. Biol. Chem. 269:5666–5672.

Ott,A.,A. Hugi, M. Baumgartner, and A. Chaintreau. 2000. Sensory investigation of yogurt flavor perception:mutual influences of volatiles and acidity. J.Agri. Food Chem. 48:441–450.



synergistic growth has been the subject ofconsiderable study since the 1960s. That S.thermophilus always appears to grow first inco-culture experiments suggests the initialmilk environment is somehow less con-ducive for L. delbrueckii subsp. bulgaricus. Itis now thought that S. thermophilus lowersthe pH and Eh to levels preferred by L. del-brueckii subsp. bulgaricus. In contrast, S.thermophilus is weakly proteolytic and lacksthe ability to hydrolyze casein, but can stillmake do, at least initially, by relying on thesmall pool of free amino acids present in themilk. Later, when those amino acids are de-pleted, S. thermophilus benefits by its associ-ation with L. delbrueckii subsp. bulgaricus,since the latter produces a proteinase thatmakes peptides and amino acids available forall organisms that happen to be present.

Eventually, however, during the co-cultureenvironment, L. delbrueckii subsp. bulgaricuswill produce more acid than can be toleratedby S. thermophilus. In yogurt, therefore, the S.thermophilus population may begin to de-crease. If these organisms were continuallypropagated as a single mixed culture, the S.thermophilus would likely be displaced afterseveral transfers.Therefore, in the manufactureof commercial yogurt cultures, the strepto-cocci and lactobacilli are ordinarily grown sep-arately in rich media under species-specific op-timum conditions, harvested, and then mixedin the desired ratio.

Depending on the form of the culture andthe manufacturer’s instructions, the culturesare added to the yogurt mix to give an initialcell concentration of about 107 cells per gram.When bulk cultures are used, this translatesinto an inoculum of about 1.5% to 2% (on aweight basis).Highly concentrated cultures arenow available, in either frozen or lyophilizedforms, that require a much lower amount ofculture to achieve the same starting cell den-sity. The inoculated mixes are then incubated(either in the cup or vat) at 40°C to 45°C forfour to six hours or until a titratable acidity (aslactic acid) of 0.8% to 0.9% is reached and thepH is about 4.4 to 4.6. Depending on the cul-

ture activity and desired final pH, the fermenta-tion may be even shorter (i.e., three hours orless).The recent trend in the United States to-ward low acid yogurts means that the fermen-tation may be judged, by some manufacturers,as complete when the pH reaches 4.8 to 4.9,orjust as the mixture begins to coagulate. In con-trast, many traditional-style yogurts are fer-mented until the pH is near 4.0.

As noted above, the incubation temperaturecan have a profound effect on the fermentationand the overall product characteristics. In gen-eral, L. delbrueckii subsp. bulgaricus has ahigher temperature optima than S. thermo-philus. Growth of S. thermophilus is favored attemperatures below 42°C, whereas L. del-brueckii subsp. bulgaricus is favored above42°C.Thus, by shifting the incubation tempera-ture by just a few degrees, it is possible to influ-ence the growth rates of the two organisms, aswell as the metabolic products they produce.Since, for example,L.delbrueckii subsp.bulgar-icus is capable of producing greater amounts oflactic acid and acetaldehyde compared to S.thermophilus, high incubation temperaturesmay result in a more acidic and flavorful yogurt.

Culture metabolism

S. thermophilus and L. delbrueckii subsp. bul-garicus both make lactic acid during the yo-gurt fermentation.They are homofermentative,meaning lactic acid is the primary end-productfrom sugar metabolism, and both ferment lac-tose in a similar manner. Moreover, the specificmeans by which lactose metabolism occurs inthese bacteria not only dictate product forma-tion, but also have an important impact on thehealth-promoting activity these bacteria pro-vide (discussed later).

The first step involves transport of lactoseacross the cell membrane. As reviewed inChapter 2, there are two general routes bywhich this step can occur in lactic acid bac-teria. Mesophilic lactococci (i.e., Lactococcuslactis subsp. lactis and Lactococcus lactissubsp. cremoris) use the phosphoenolpyru-vate (PEP)-dependent phosphotransferase



system (PTS), in which lactose is phosphory-lated during its transport across the cyto-plasmic membrane. The high energy phos-phate group of PEP, following its transfer viaa protein cascade, serves as the phosphorylgroup donor for the phosphorylation reac-tion. It also provides the driving force forlactose transport.The product that accumu-lates in the cytoplasm, therefore, is lactose-phosphate, with the phosphate attached tocarbon 6 on the galactose moiety. In con-trast, S. thermophilus and L. delbrueckiisubsp. bulgaricus, the thermophilic culturebacteria, use a secondary transport system(called LacS) for lactose uptake.Transport oflactose occurs via a symport mechanism,with the proton gradient serving as the dri-ving force (although this is not the entirestory, as discussed below). Lactose is notmodified during transport and instead accu-mulates inside the cell as free lactose.

For both the PTS and LacS systems, the nextstep is hydrolysis, but since the substrates are different, the hydrolyzing enzymes must be dif-ferent, too. In lactococci, hydrolysis of lactose-phosphate occurs by phospho-�-galactosidase,forming glucose and galactose-6-phosphate.Glu-cose is subsequently phosphorylated (using ATPas the phosphoryl group donor) by glucokinase(or hexokinase) to form glucose-6-phosphate.The latter then feeds directly into the EmbdenMeyerhof glycolytic pathway, ultimately leadingto lactic acid.The other product of the phospho-�-galactosidase reaction, galactose-6-phosphate,is metabolized simultaneously by a parallel path-way called the tagatose pathway, which also re-sults in formation of lactic acid.

In S. thermophilus and L.delbrueckii subsp.bulgaricus, the intracellular lactose that accu-mulates is hydrolyzed by a �-galactosidase, re-leasing glucose and galactose. The glucose ismetabolized to lactic acid, the same as in thelactococci. However, most strains of S. ther-mophilus and L. delbrueckii subsp. bulgaricuslack the ability to metabolize galactose and in-stead secrete galactose back into the milk.Thisis despite the observation that a pathway forgalactose metabolism indeed exists in thesebacteria.

The reason why this pathway does not func-tion, especially in S. thermophilus, and whythis apparent metabolic defect occurs, hasbeen the subject of considerable study. It nowappears that the genes coding for the enzymesnecessary to metabolize galactose, the so-called Leloir pathway, are not transcribed orexpressed. Rather, these genes are strongly re-pressed,and the cell responds by excreting theunfermented galactose that it has accumu-lated.The extracellular galactose is also not me-tabolized even later, although genes coding fora galactose transport system apparently exist.

Although it would appear that S. ther-mophilus is being wasteful by not making effi-cient use of both monosaccharide constituentsof lactose, this is not the case. When S. ther-mophilus grows in milk,where the lactose con-centration is 5% (more than 140 mM), or in yo-gurt mix, which has an even higher lactoseconcentration due to added milk solids, sugarlimitation is not an issue. Rather, growth cessa-tion in yogurt occurs due to low pH or low tem-perature, not sugar availability. Interestingly,when S.thermophilus excretes galactose it doesso via the LacS permease, the same system thatis responsible for lactose uptake.Thus, LacS actsas an exchange system, trading an out-boundgalactose for an in-coming lactose (Chapter 2).This precursor-product exchange reactionspares the cell of energy that it would normallyspend to transport lactose. In fact, S. thermo-philus may be so well-adapted to growth in milkthat the galactose efflux reaction is kineticallyfavored,even in variant or mutant strains wherethe galactose genes are de-repressed and galac-tose pathway enzymes are expressed. In yogurt,galactose accumulation, even at concentrationsas high as 0.7% to 1.0%, is ordinarily of no majorconsequence (the exception may be for individ-uals prone to cataracts, since dietary galactosemay exacerbate that condition). However incheese manufacture, galactose can be the causeof serious technological problems, and galac-tose-fermenting strains of S. thermophilus maybe of considerable value (Chapter 5).

In some yogurt cultures, Lactobacillus hel-veticus is used instead of L. delbrueckii subsp.bulgaricus. This organism deals with lactose



and galactose in a somewhat similar manner.Transport of lactose, as noted above,occurs viaa LacS that has nearly the same activity as in S.thermophilus. Galactose efflux occurs afterhydrolysis by �-galactosidase, but not nearly tothe same extent. Instead, some of the galactoseis metabolized via the Leloir pathway at thesame time as glucose (Chapter 2). In addition,extracellular galactose can be transported andmetabolized. However, little of the secretedgalactose is fermented because the fermenta-tion period is so short.Even most of the lactoseis left unfermented. Since most yogurt is madewith high solids milk, the lactose concentra-tion in yogurt is often higher than that of milk,with levels of 6% to 7%.

Post-fermentation

The fermentation is considered completewhen the target acidity is reached and the yo-gurt is cooled quickly to below 4°C. In fact, forall practical purposes, cooling is really the onlyway to arrest the fermentation and stop furtheracid production. Cup-set yogurt must be verycarefully moved to coolers (0°C to 4°C) toavoid agitation which may disturb the gel, re-sulting in syneresis. For Swiss style yogurt,where the fermentation occurs in a vat, the yo-gurt is typically stirred and cooled in the vat,then mixed with fruit or other flavoring, andfilled into cups or containers. It is important torecognize that during the cooling period thepH may continue to drop by an additional 0.2to 0.3 pH units, so initiating the cooling stepeven when the pH is 4.8 to 4.9 may be war-ranted. In addition, some culture strains maycontinue to produce acid during refrigeratedstorage, albeit slowly. Over-acidification andpost-fermentation acidification are major prob-lems,since U.S.consumers generally prefer lessacidic yogurts (see below).

Yogurt Flavor and Texture

The most dominant flavor of yogurt is sour-ness, due to lactic acid produced by the starterculture. Most yogurts contain between 0.8%and 1.0% lactic acid and have a pH below 4.6.

In the absence of sweetener or added flavors,most consumers can detect sourness when thepH is below 5.0. Other organic acids, includingformic and acetic,may also be produced by theculture, but at much lower concentrations, andthey generally make only modest contributionsto yogurt flavor.

There are, however, other metabolic prod-ucts produced by the culture that accumulatein yogurt and contribute significantly to flavordevelopment. The most important of these isacetaldehyde, a two carbon aldehyde.Althoughnormally present at less than 25 ppm concen-tration, this is still sufficient to give yogurt itscharacteristic tart or green apple flavor. Both S.thermophilus and L. delbrueckii subsp. bul-garicus can produce acetaldehyde; however,the rate and amount produced depends on thestrain and on the growth conditions.

There appear to be at least two metabolicroutes by which acetaldehyde is produced(Figure 4–3). In one pathway, an aldolase en-zyme hydrolyzes the amino acid, threonine,and acetaldehyde and glycine are formed di-rectly. It is also possible to form acetaldehydefrom pyruvate (generated from glucose metab-olism).The pyruvate can either be decarboxy-lated, forming acetaldehyde directly, or con-verted first to acetyl CoA, and then oxidized toacetaldehyde. By understanding how thesepathways function, it may soon be possible toapply molecular metabolic engineering ap-proaches to enhance or control acetaldehydelevels in yogurt. Finally, diacetyl and acetoinmay also be produced by the yogurt culture;however, they are usually present at concentra-tions below a typical taste threshold.

In the United States, most yogurt is flavoredwith fruit and other flavorings. Fruits are usu-ally added in the form of a thick puree, with orwithout real pieces of fruit.These ingredientsobviously dilute or mask the lactic acid and ac-etaldehyde flavors. In fact, the trend in theUnited States is to produce mild-flavored yo-gurts with less characteristic yogurt flavor.Thus, strains that produce little acetaldehydeare often used in yogurt cultures.

The texture and rheological properties of yogurt are, perhaps, just as important to



consumers as flavor. Coagulated milk is essen-tially a protein gel that imparts viscosity,mouthfeel, and body to the finished product. Forma-tion and maintenance of the gel structure,therefore, is important for yogurt quality. Thegel properties are affected by the ingredientsin the yogurt mix, how the yogurt mix isprocessed and produced, culture activity, andpost-fermentation factors. For example, properheat treatment of the milk has a profound ef-fect on gel strength and, specifically, the water

holding capacity. If the gel is poorly formed ordisrupted syneresis will occur, leading to thedefect known as “wheying off” (see below).Tocontrol syneresis and maintain a suitable gelstructure, yogurt manufacturers in the UnitedStates commonly add stabilizers to the mix.Typically, stabilizers are hydrophilic polysac-charides whose purpose is to bind water.Themost popular stabilizers are naturally derivedgums and starches, including carrageenan, lo-cust bean, and guar gums; corn starch; tapioca;

CO

CH

H

H

H

acetaldehyde

glucose

C C

O OC

OHH

H

H

pyruvate

glycolysis

pyruvatedecarboxylase

acetyl CoA

aldehydedecarboxylase

threonine C

H

C

HO

CH C

H

H

O

O-

+NH3H

serine hydroxymethyltransferase

CO2

glycine

NADH +H

NAD

pyruvatedehydrogenase

CoA

Figure 4–3. Formation of acetaldehyde by yogurt bacteria from pyruvate and threonine. Adapted fromChaves et al. 2002.



and pectin.Gelatin is also frequently added as astabilizer.

In some European countries (e.g., France),stabilizers are not allowed in yogurt.Therefore,the ability of some strains of S. thermophilusand L. delbrueckii subsp. bulgaricus to pro-duce and secrete natural polysaccharide mater-ial directly in the yogurt during the fermenta-tion is an especially important trait. This isbecause many of these extracellular polysac-charides (exopolysaccharides or EPS) have ex-cellent stabilizing and rheological properties.The producer strains are now widely used andincluded in yogurt starter cultures to providethe desired body characteristics. A variety ofEPS produced by these bacteria have beenidentified (Box 4–3). Some EPS contain only asingle sugar (homopolysaccharides). In con-trast, other EPS consist of a heterogenous mix-ture of different sugar monomers (heteropoly-saccharides) in varying ratios. The lattercontain isomers of galactose and glucose,along with rhamnose and other sugars.

Defects

The stand-up comedian, George Carlin, used toask how one could tell when yogurt has gonebad.Yogurt is, after all, sour milk, so how couldit spoil? In fact, yogurt can suffer from flavor,texture, and appearance defects, and can evenbecome spoiled. In some cases, these defectsare caused by microorganisms (including thestarter culture), but often they are caused bymanufacturing issues.This is especially true forthe wheying-off or syneresis defect that isprobably the single most serious issue for yo-gurt consumers.

Although relatively rare, some chemically-derived flavor defects in yogurt can be causedby using poor-quality milk. For example, ran-cid or oxidized flavors are almost always dueto mis-handling of the milk (or nonfat drymilk).The most common flavor defects, how-ever, are caused by microorganisms. In somecases, microbial spoilage is due to the pres-ence of thermophilic sporeforming bacteria,including various species of Bacillus, that sur-

vived pasteurization and grew in the milk,pro-ducing bitter flavors.Yeasts and molds that arepresent in the fruit flavoring material can alsoproduce off-flavors, although it is usually justthe appearance of yeast or mold colonies thatcause consumers to reject such products. Ofcourse, bacteria, such as psychrotrophs thatgrew in the raw milk prior to pasteurization,can produce bitter and rancid flavors.

Despite the acidity and acetaldehyde flavornormally expected in yogurt, these attributescan also be considered as defects if present athigh enough concentrations or if they exceedconsumer tastes.Yogurts within a range of pH4.2 to 4.5 (0.8% to 1% lactic acid) are generallyacceptable, but the pH can drop to as low as4.0 (and about 2% lactic acid) under certaincircumstances. Excessive acid production, forexample, can occur if the yogurt is incubatedat too high a temperature or if cooling rates aretoo slow (both promote growth of L. del-brueckii subsp. bulgaricus). Even properlyproduced yogurt can eventually become tooacidic during low temperature storage due tothe ability of the culture to produce, albeitslowly, lactic acid.This post-fermentation acidi-fication is one of the main reasons why con-sumers may consider an otherwise perfectlyfine yogurt to be spoiled (despite George Car-lin’s protestations).

One way to prevent over-acidification is tocontrol culture activity. Strains have been iden-tified (either naturally or via molecular inter-ventions) that are sensitive to low pH or lowtemperatures, such that growth and acid for-mation are halted when the pH reaches a criti-cal value or when the temperature is suffi-ciently low (Box 4–2). Of course, one otherway to control acidity and to increase shelf-lifewould be to heat pasteurize the yogurt afterthe fermentation. This approach has beenlargely rejected by the industry because itwould negate most of the health benefits con-ferred by the culture and would harm the im-age yogurt enjoys as a healthy food product.Moreover, the National Yogurt Association, atrade organization representing yogurt manu-facturers, has suggested that refrigerated



Box 4–3. Exopolysaccharide production by lactic acid bacteria

The ability of lactic acid bacteria (LAB) to synthesize exopolysaccharides (EPS) is now re-garded as particularly important because they enhance texture development in fermenteddairy foods. Specifically, they can serve as natural thickeners, stabilizers, bodying agents, emul-sifiers, gelling agents, and fat replacers (De Vuyst and Vaningelgem, 2003; Ruas-Madiedo and delos Reyes-Gavilán, 2005).

They are especially applicable in yogurt and other cultured dairy products since they can re-duce syneresis and gel fracture, and can increase viscosity and gel strength (Duboc and Mollet,2001).Thus, incorporating EPS-producing strains in dairy starter cultures could eliminate thenecessity for adding other stabilizing agents. This is important in some European countrieswhere the addition of stabilizers is restricted.

Lactic acid bacteria that produce EPS do so in one of two general ways.Some strains synthesizeEPS as released or dis-attached material (referred to as “ropy” strains), whereas others producecapsular polysaccharide that remains attached to the cell surface (Broadbent et al., 2003). How-ever, it appears that the EPS from both ropy and capsular strains have useful functional propertiesin cultured milk products, regardless of how synthesis and attachment occurs.EPS may serve as adefense mechanism against bacteriophages by preventing their adsorption (Forde and Fitzgerald,1999; Lamothe et al., 2002). In addition, some LAB may produce EPS to adhere or stick to varioussurfaces, thereby promoting colonization of desirable habitats (Lamothe et al., 2002).

Structure of Exopolysaccharides

The molecular masses of EPS from LAB range from 10 kDa to greater than 1,000 kDa (Welman and Maddox, 2003).The number of monomers present within these EPS range fromabout 50 to more than 5,000.The most frequently found monosaccharides are glucose andgalactose, but rhamnose, mannose, fructose, arabinose and xylose, and the sugar derivatives,N-acetylgalactosamine and N-acetylglucosamine, are also common (Table 1).

In general, the EPS can be subdivided into two major groups: homopolysaccharides and het-eropolysaccharides (Jolly et al., 2002; Ruas-Madiedo et al., 2002). Homopolysaccharides arecomposed of only one type of monosaccharide, whereas heteropolysaccharides are composedof a repeating unit that contains two or more different monosaccharides. Examples of the ho-mopolysaccharides include: (1) dextrans and mutans, �-1,6 and �-1,3-linked glucose polymersproduced by strains of Leuconostoc mesenteroides and Streptococcus mutans, respectively;and (2) glucans,�-1,3-glucose polymers, produced by strains of Pediococcus and Lactobacillus;(3) fructans, �-2,6-linked D-fructose polymers produced by strains of Streptococcus salivarius,Lactobacillus sanfranciscensis, and Lactobacillus reuteri; and (4) polygalactans, galactose pro-duced by strains of Lactococcus lactis.

The heteropolysaccharides produced by lactic acid bacteria differ considerably betweenstrains in their composition, structure, and physicochemical properties (De Vuyst and Degeest,1999). Heteropolysaccharides most often contain a combination of glucose, galactose, andrhamnose, as well as lesser amounts of fructose, acetylated amino sugars, ribose, glucuronicacid, and noncarbohydrate constituents such as glycerol, phosphate, pyruvyl, and acetyl groups(Jolly et al., 2002).These EPS are the ones most commonly associated with yogurt, due to theirproduction by Lactobacillus delbrueckii subsp. bulgaricus, Lactobacillus helveticus, andStreptococcus thermophilus.The majority of the EPS produced by S. thermophilus contain re-peating tetrasaccharide units; however, several strains produce EPS consisting of hexasaccha-rides or heptasaccharides. Strains of L. delbrueckii subsp. bulgaricus typically produce hep-tasaccharides composed of glucose, rhamnose, and a high ratio of galactose, whereas L.helveticus produces heteropolysaccharides made up almost exclusively of glucose and galac-tose. Some mesophilic LAB, including Lactococcus lactis subsp. lactis, L. lactis subsp. cremoris,L. sakei, and L. rhamnosus, are also heteropolysaccharide EPS producers.



Table 1. Exopolysaccharide composition of various lactic acid bacteria.

Lactic acid bacteria (Strain) Exoploysaccharide Composition Ratio of Sugars

Lactobacillus delbrueckii subsp.bulgaricus

LY03 Heptasaccharide of glucose, galactose, Gal:Glu:Rha 5:1:1rhamnose

rr Heptasaccharide of glucose, galactose, Gal:Glu:Rha 5:1:1rhamnose

Lfi5 Heptasaccharide of glucose, galactose, Gal:Glu:Rha 5:1:1rhamnose

291 Pentasaccharide of glucose, galactose Glu:Gal: 3:2

Lactobacillus lactis subsp. cremorisB39 Heptasaccharide of glucose, galactose, Gal:Glu:Rha 3:2:2

rhamnoseH415 Pentasaccharide of galactose —SBT 0495 Pentasaccharide glucose, galactose, Gal:Glu:Rha:Pho 2.1:1.8:1.3:1

rhamnose, phosphate

Lactobacillus helveticus TY1-2 Heptasaccharide of glucopyranosyl, GluP:GalP:AGluP:galactopyranosyl, 2-

acetamido-2-deoxy-D-glucopyranosyl 3.0:2.8:0.9

Lactobacillus rhamnosus RW-9595M Heptasaccharide of glucose, galactose, Glu:Gal:Rha:Pyr 2:1:4:1rhamnose, pyruvate

Lactobacillus sakei 0-1 Pentasaccharide of glucose, rhamnose, Glu:Rha:G3P 3:2:1sn-glycerol 3-phosphate

Lactobacillus sanfranciscensis Trisaccharide of glucose and fructose Glu:Fru 1:2LTH2590

Streptococcus thermophilusS3 Hexasaccharide of galactose and rhamnose Gal:Rha 2:1Sy89 Tetrasaccharide of glucose and galactose Gal:Glu 1:1Sy102 Tetrasaccharide of glucose and galactose Gal:Glu 1:1Sfi6 Tetrasaccharide of glucose, and galactose, Gal:Glu:GalNAc 2:1:1

N-acetylgalactosamineSfi12 Hexasaccharide of galactose, rhamnose, Gal:Rha:Glu 3:2:1

glucoseSfi39 Tetrasaccharide of glucose and galactose Gal:Glu 1:1IMDO1 Tetrasaccharide of glucose and galactose1 Gal:Glu 3:1IMDO2 Tetrasaccharide of glucose and galactose1 Gal:Glu 3:1IMDO3 Tetrasaccharide of glucose and galactose1 Gal:Glu 3:1NCFB 859 Tetrasaccharide of glucose and galactose1 Gal:Glu 3:1CNCMl Tetrasaccharide of glucose and galactose1 Gal:Glu 3:1MR-lC Octasaccharide of galactose, rhamnose, Gal:Rha:Fuc 5:2:1

fucoseEU20 Heptasaccharide of glucose, galactose, Glu:Gal:Rha 2:3:2

rhamnoseOR 901 Hepatsaccharide of galactopyranose, GalP:RhaP 5:2

rhamnopyranose

1One galactose residue occurs as N-acetygalactosamine

Box 4–3. Exopolysaccharide production by lactic acid bacteria (Continued)

(Continued)



Production of Exopolysaccharides

Growth-associated EPS biosynthesis has been observed for several strains of LAB. In Streptococ-cus thermophilus LY03 and related strains, for example, there is a direct relationship betweenEPS yield and optimal growth conditions (de Vuyst et al., 1998). In contrast, EPS production inmesophilic LAB appears to be enhanced by sub-optimal temperatures, with maximum EPS pro-duction occurring at 20°C (Degeest and de Vuyst, 1999).The increase in EPS production whencells are growing at lower temperatures may be due to an increase in the number of isoprenoidlipid carrier molecules available for EPS formation,as occurs for Gram negative bacteria (Suther-land, 1972). Another possibility is that EPS synthesis may occur as a response to stress whencells are exposed to harmful or sub-optimal environments (Ophir and Gutnick, 1994). In eithercase, EPS production is likely correlated to EPS gene expression.

Gene organization for Exopolysaccharides

Genes encoding for EPS production exist as clusters, with specific structural, regulatory, andtransport functions (Figure 1). In general, these gene clusters are often located on plasmids inmesophilic LAB and on the chromosome in thermophilic LAB (de Vuyst et al., 2001). Genes orgene clusters involved in LAB exopolysaccharide biosynthesis have been sequenced for severalyogurt strains, as well as other mesophilic LAB (Broadbent et al., 2003; van Kranenburg et al.,1999b, 1999c).These clusters reveal a common operon structure, with the genes positioned ina particular order (Figure 1).

At the beginning of the gene cluster is epsA, which is suggested to be the region potentiallyinvolved in EPS regulation.The central region of the gene cluster, consisting of epsE, epsF, epsG,epsH, and epsI, encodes for enzymes predicted to be involved in the biosynthesis of the EPS re-peating unit.The products of epsC, epsD, epsJ, and epsK are responsible for polymerization andexport of the EPS. Interestingly, epsA, epsB, and epsC are also present in non-EPS producing


Doed A B C D E F G H I J K L SI3911

fro9.41

SI 1S Nfro1

SI R X A B C D E F G H I J K L Yfro

Doed A B C D E F G H I J K L M3

Xfro A B C D E F G H I J K L M N4 Yfro

2

Polymerization and Chain-length Determination

Biosynthesis of Repeating Unit

Polymerization and Export

Unknown, Not Specified or Outside of Cluster

Regulation

frO9.41

Figure 1. Functional comparison of EPS genes from lactic acid bacteria. 1, Streptococcus thermophilus NCFB 2393; 2,Lactobacillus lactis NIZO B40; 3, Streptococcus thermophilus Sfi6; 4, Lactobacillus delbrueckii subsp. bulgaricus Lif5.Gene maps were adapted from Almirón-Roig et al. 2000; van Kranenburg et al. 1999a; Stingele et al. 1996; andLamothe et al. 2002.



strains of S. thermophilus, indicating the possibility that a spontaneous genomic deletion of anancestral EPS gene cluster may have occurred (Bourgoin et al., 1999).This theory is supportedby the observation of genetic instability, including deletions, in the S. thermophilus genome.

Regulation of Exopolysaccharide Genes

Production of EPS by LAB is often unstable and yields can be highly variable (Ricciardi andClementi, 2000). Compared to other Gram positive bacteria, LAB generally produce loweramounts of EPS. Even when grown under optimized conditions, most LAB produce � 3 g per Lof EPS, whereas other organisms make as much as 10 g to 15 g per L. Therefore, several ap-proaches have been considered to improve EPS yield and stability (Boels et al., 2003a, 2003b;Levander et al., 2002; Stingele et al., 1999; Kranenburg et al., 1999a).

One strategy involves pathway engineering, such that carbon flow is diverted away from ca-tabolism and toward EPS biosynthesis. For example, synthesis of EPS ordinarily starts when glucose-6-phosphate is isomerized to glucose-1-phosphate by phosphoglucomutase (the prod-uct of the pgm gene).The glucose-1-phosphate is then converted, via UDP-glucose pyrophos-phorylase to UDP-glucose,which serves as a precursor for EPS biosynthesis.Over-expression ofpgm or galU (encoding for UDP-glucose pyrophosphorylase) led to an increase in the EPS yieldby S. thermophilus LY03 (Hugenholtz and Kleerebezem, 1999).

A second approach involves over-expression of EPS genes. For example, over-expression ofthe epsD gene (encoding the priming glycosyl transferase) in L. lactis resulted in a small in-crease in EPS production (van Kranbenburg et al., 1999c). Cloning the entire EPS gene clusteron a single plasmid with a high copy number could also have an effect of increasing EPS pro-duction, as shown for L. lactis NIZO B40, which produced four times more EPS than the parentstrain (Boels et al., 2003).

References

Almirón-Roig, E., Mulholland, F., Gasson, M.J., and A.M. Griffin. 2000.The complete cps gene cluster fromStreptococcus thermophilus NCFB 2393 involved in the biosynthesis of a new exopolysaccharide. Mi-crobiol. 146:2793–2802.

Boels, I.C.,M.Kleerebezem,and W.M.de Vos. 2003a.Engineering of carbon distribution between glycolysisand sugar nucleotide biosynthesis in Lactococcus lactis.Appl. Environ. Microbiol. 69:1129–1135.

Boels, I.C., R. van Kranenburg, M.W. Kanning, B.F. Chong, W.M. de Vos, and M. Kleerebezem. 2003b. In-creased exopolysaccharide production in Lactococcus lactis due to increased levels of expression ofthe NIZO B40 eps Gene Cluster.Appl. Environ. Microbiol. 69:5029–5031.

Bourgoin, F,A. Pluvinet, B. Gintz, B. Decaris, and G. Guedon. 1999.Are horizontal transfers involved in theevolution of the Streptococcus thermophilus exopolysaccharide synthesis loci? Gene 233:151–161.

Broadbent, J.R., D.J. McMahon, D.L.Welker, C.J. Oberg, and S. Moineau. 2003. Biochemistry, genetics, and ap-plications of exopolysaccharide production in Streptococcus thermophilus: a review. J. Dairy Sci.86:407–423.

Degeest, B. and L. De Vuyst. 1999. Indication that the nitrogen source influences both amount and size ofexopolysaccharides produced by Streptococcus thermophilus LY03 and modelling of the bacterialgrowth and exopolysaccharide production in a complex medium.Appl. Environ. Microbiol. 65:2863–2870.

De Vuyst, L., and B. Degeest. 1999. Heteropolysaccharides from lactic acid bacteria. FEMS Microbiol. Rev.23:157–177.

De Vuyst, L., and F. Vaningelgem. 2003. Developing new polysaccharides, p. 257—320. In B.M. McKenna(ed.), Texture in food, vol. 2. Semi-solid foods.Woodhead Publishing Ltd., Cambridge, United Kingdom.

De Vuyst, L., F. De Vin, F.Vaningelgem, and B. Degeest. 2001. Recent developments in the biosynthesis andapplications of heteropolysaccharides from lactic acid bacteria. Int. Dairy J. 11:687–707.


(Continued)



yogurt contain at least 108 viable organismsper gram at the time of manufacture.

Texture, body, and appearance defects arecommon in yogurt.Two of the more serious de-fects, weak bodied yogurt and free whey for-mation (i.e., wheying off), have several causes.They may occur when the solids content is toolow or when the mix was pasteurized at too

low of a temperature, both of which ultimatelyinfluence how much denatured whey proteinsare present. The gel structure will shrink andsyneresis will occur if the culture produces toomuch acid too fast during the fermentation,causing the pH to become too low.The sameresult will also occur if the temperature waslowered too rapidly during cooling. When yo-

De Vuyst, L., F. Vanderveken, S. Van de Ven, and B. Degeest. 1998. Production by and isolation of ex-opolysaccharides from Streptococcus thermophilus grown in milk medium and evidence for theirgrowth-associated biosynthesis. J.Appl. Microbiol. 84:1059–1068.

Duboc, P., and B. Mollet. 2001. Applications of exopolysaccharides in the dairy industry. Int. Dairy J. 11:759–768.

Forde,A., and G.F. Fitzgerald. 1999.Analysis of exopolysaccharide (EPS) production mediated by the bacte-riophage adsorption blocking plasmid, pCI658, isolated from Lactococcus lactis subsp. cremoris HO2.Int. Dairy J. 9:465–472.

Hugenholtz, J. and Kleerebezem,M.1999.Metabolic engineering of lactic acid bacteria:overview of the ap-proaches and results of pathway rerouting involved in food fermentations. Curr. Opin. Biotechnol.10:492–497. Jolly, L., S.J.F.Vincent, P. Duboc, and J.-R. Nesser. 2002. Exploiting exopolysaccharides fromlactic acid bacteria.Antonie van Leeuwenhoek 82:367–374.

Lamothe, G.T., L. Jolly, B. Mollet, and F. Stingele. 2002. Genetic and biochemical characterization of ex-opolysaccharide biosynthesis by Lactobacillus delbrueckii subsp. bulgaricus. Arch. Microbiol.178:218–228.

Levander,F.,M.Svensson, and P.Rådström. 2002.Enhanced exopolysaccharide production of metabolic en-gineering of Streptococcus thermophilus. Appl. Environ. Microbiol. 68:784–790.

Ophir,T. and Gutnick, D.L. 1994.A role for exopolysaccharides in the protection of microorganisms fromdesiccation.Appl. Environ. Microbiol. 60:740–745.

Ricciardi,A. and F. Clementi. 2000. Exopolysaccharides from lactic acid bacteria: structure, production andtechnological applications. Ital. J. Food Sci. 12:23–45.

Ruas-Madiedo, P. and C.G. de los Reyes-Gavilán. 2005. Methods for the screening, isolation, and characteri-zation of exopolysaccharides produced by lactic acid bacteria. J. Dairy Sci. 88:843–856.

Ruas-Madiedo, P., J. Hugenholtz, and P. Zoon. 2002.An overview of the functionality of exopolysaccharidesproduced by lactic acid bacteria. Int. Dairy J. 12:163–171.

Stingele, F., J.-R. Neeser, and B. Mollet. 1996. Identification and characterization of the eps (exopolysaccha-ride) gene cluster from Streptococcus thermophilus Sfi6. J. Bacteriol. 178:1680–1690.

Stingele, F., S.Vincent, E.J. Faber, J.W. Newell, J.P. Kamerling, and J.R. Nesser. 1999. Introduction of the ex-opolysaccharide gene cluster from Streptococcus thermophilus Sfi6 into Lactococcus lactis MG1363:production and characterization of an altered polysaccharide. Mol. Microbiol. 32:1287–1295.

Sutherland, I.W. 1972. Bacterial exopolysaccharides.Adv. Microbial Physiol. 8:143–213.van Kranenburg, R., I.C. Boels, M. Kleerebezem, and W.M. de Vos. 1999a. Genetics and engineering of mi-

crobial exopolysaccharides for food: approaches for the production of existing and novel polysaccha-rides. Curr. Opin. Biotechnol. 10:498–504.

van Kranenburg,R., I.I. van Swam, J.D.Marugg,M.Kleerebezem,and W.M.de Vos.1999c.Exopolysaccharidebiosynthesis in Lactococcus lactis NIZO B40: functional analysis of the glycosyltransferase genes in-volved in synthesis of the polysaccharide backbone. J. Bacteriol. 181:338–340.

van Kranenburg, R., H.R.Vos, I.I. van Swam, M. Kleerebezem, and W.M. de Vos. 1999b. Functional analysis ofglycosyltransferase genes from Lactococcus lactis and other gram-positive cocci: complementation, ex-pression, and diversity. J. Bacteriol. 181:6347–6353.

Welman,A.D., and I.S. Maddox. 2003. Exopolysaccharides from lactic acid bacteria: perspectives and chal-lenges.Trends Biotechnol. 21:269–274.




gurt is mixed with fruit or other flavorings, inthe case of Swiss or stirred style yogurts, orwhen stirred by consumers prior to consump-tion, the gel is further disrupted and thinningand syneresis occurs.

Nutritional Benefits of Yogurt

One of great all-time television commercialsaired in the late 1970s and featured an inter-view with a Russian centagenerian whoclaimed that his longevity was due to his dailyconsumption of yogurt. When asked who gothim started on his yogurt regimen, he proudlystated it was his mother, who’s smiling facethen moves into the television frame. Indeed,the popularity of yogurt, as implied by this ad-vertisement and as mentioned previously inthis chapter (Box 4–1), has long been due, inlarge part, to the purported health benefits as-cribed to yogurt consumption. In fact, this no-tion of yogurt as an elixir that fends off agingand promotes human health originated at leasta century ago, when the Russian immunologistand Nobel laureate Elie (Elia or Ilya) Metch-nikoff published The Prolongation of Life in1906.As Metchnikoff and many other microbi-ologists have since reported, it is the specificbacteria that either conduct the fermentation(i.e., S. thermophilus and L. delbrueckii subsp.bulgaricus) or that are added to yogurt in theform of culture adjuncts (e.g., lactobacilli andbifidobacteria) that are responsible for the de-sirable health benefits of yogurt.

Of course, yogurt has nutritional propertiesother than those derived from the culture or-ganisms. A single 170 g (6 ounce) serving ofplain, nonfat yogurt contains about 170 calo-ries and supplies 18% of the Daily Value re-quirements for protein, 30% for calcium, and20% for vitamin B12.Still, these are the same nu-trient levels one would get from milk (pro-vided one had accounted for the milk solidsnormally added to the yogurt mix). Althoughthere are reports that yogurt contains more vi-tamins than the milk from which it was made(due to microbial biosynthesis), these in-creases do not appear to be significant. Thus,if indeed yogurt has an enhanced nutritional

quality compared to milk, those differencesmust be due to the microorganisms found in yogurt. As noted earlier, there are many healthclaims ascribed to yogurt, and especially theprobiotic bacteria added as nutritional adjuncts.Among these claims are that these bacteria are anti-cholesterolemic and anti-tumorigenic,enhance mineral absorption, promote gastroin-testinal health, and reduce the incidence of enteric infections. There are also some rathersurprising health benefits that yogurt may pro-vide (Box 4–4). Despite volumes of research onthese claims,however, few have been unequivo-cally established (Box 4–5.).

Perhaps the health benefit that is most gener-ally accepted is the claim that yogurt organismscan reduce the symptoms associated with lac-tose intolerance. Lactose intolerance is a condi-tion that is characterized by the inability of cer-tain individuals to digest lactose. Its specificcause is due to the absence of the enzyme �-galactosidase, which is ordinarily produced andsecreted by the cells that line the small intestine(Figure 4–4). In individuals expressing �-galac-tosidase, lactose is hydrolyzed and the glucoseand galactose are absorbed across the epithelialcells and eventually enter into the blood stream(in the case of galactose, only after conversionto glucose in the liver). If �-galactosidase is notproduced in sufficient levels, however, the lac-tose remains undigested and is not absorbed. In-stead, it passes to the large intestine,where it ei-ther causes an increase in water adsorption intothe colon (via osmotic forces) or is fermentedby colonic anaerobes. The resulting symptomscan include diarrhea, gas, and bloating, leadingmany lactose intolerant individuals to omit milkand dairy products from their diet. Lactose in-tolerance has a genetic basis, affecting African,Asian, American Indian, and other non-Caucasian populations far more frequently thanCaucasian groups.These individuals could typi-cally tolerate lactose (i.e., milk) when young,but lose this ability during adulthood.As manyas 50 million people in the United States may belactose intolerant.

It has often been noted that lactose-intolerantindividuals could consume yogurt without ill effect. It is now known that the bacteria in



Box 4–4. Stressed out? Have a cup of yogurt and relax

It’s exam time and you’re feeling a bit stressed,maybe even depressed.How can you possibly re-lax so that you can study effectively? The solution is not have yet another cup of coffee, but tohave a cup of yogurt (yes, yogurt) instead.At least, that’s the suggestion made recently by a re-search group at the Consejo Superior de Investigaciones Cientificas research institute in Spain(Marcos et al., 2004).

In this study, two groups of college students were fed, on a daily basis, either plain milk ormilk fermented with Streptococcus thermophilus and Lactobacillus delbrueckii subsp.bulgar-icus (i.e., yogurt bacteria),plus a probiotic organism,Lactobacillus casei.The study began threeweeks prior to the start of an annual college testing period and continued for the three weeksduring which the tests were administered. Consumption of other fermented milk products wasnot allowed, although no other dietary restrictions were imposed.

Various indicators of stress and anxiety were measured at the start (baseline) and at the endof the study. Anxiety self-report tests and analyses for lymphocytes, phagocytic activity, andserum cortisol levels were included.The latter analyses can be used to assess the effect of stresson the immune response of the subject.Although no significant differences were reported be-tween the anxiety scores for the control (milk) and treatment groups (fermented milk), differ-ences were observed for other measures. Specifically, the total lymphocyte numbers werehigher in the treatment group, as was the number of CD56 cells (a specific subset or type of“killer cell” lymphocytes).

Because lymphocytes, and CD56 cells in particular, are part of the host defense system forfighting infections, the lower levels in the control group (compared to treatment subjects) mayindicate that these subjects could be exposed to a greater risk of infection. Conversely, the yogurt-eating group may have a more resilient immune system, and, therefore, are better pre-pared to tolerate stress.

Recently, it was suggested that probiotic bacteria may also be effective at alleviating symp-toms related to major depressive disorder or MDD (Logan and Katzman, 2005).This hypothesisis based, according to the authors, on several observations. First, there appears to be a relativelyhigh correlation between the number of individuals who suffer from MDD and who also haveintestinal disorders, atopic disease, endometriosis, malabsorption syndromes, small intestinebacterial overgrowth syndrome, and other chronic diseases. Second, the intestinal flora in thesepatients often contain lower levels of lactobacilli and bifidobacteria, and stress can lead to evengreater long-term reductions in these organisms.Third, as the authors note, the intestinal tract is“a meeting place of nerves, microorganisms, and immune cells.”Thus, the microflora, by virtueof their ability to produce various neurochemicals and to influence host production of cy-tokines and other bioactive compounds, may have intimate interactions with the immune andcentral nervous system.

Therefore,while more studies are certainly needed to establish the relationship between pro-biotic bacteria and stress and depression, students may, nonetheless, wish to consider the “chill-ing”effects of yogurt before the next big test.

References

Logan,A.C., and M. Katzman. 2005. Major depressive disorder: probiotics may be an adjuvant therapy. Med.Hypotheses. 64:533–538.

Marcos,A., J.Wärnberg, E. Nova, S. Gómez,A.Alvarez, R.Alvarez, J.A. Mateos, and J. M. Cobo. 2004.The effectof milk fermented by yogurt cultures plus Lactobacillus casei DN-114001 on the immune response ofsubjects under academic stress. Eur. J. Nutr. 43:381–389.



Box 4–5. Sorting Through the Pros and Cons of Probiotics

Few research areas in the food and nutritional sciences have received as much worldwide at-tention and study as the subject of probiotics. Since the late 1980s, there has been a rapid risein the amount of research being conducted on probiotic microorganisms (Figure 1). Included inthe scientific literature are descriptions of the biochemical and physiological properties of pro-biotic bacteria, their taxonomical classifications, criteria for their selection, and data from invitro and in vivo studies in which physiological effects on hosts have been tested. In addition,there is a plethora of information in the popular press, trade magazines, and especially the In-ternet, on probiotics.

The challenge for probiotic researchers has been to establish mechanisms and degrees of ef-ficacy of probiotics to prevent or treat particular diseases or conditions.These efforts are com-plicated by many factors, including:(1) the complexity of microbial communities of the alimen-tary canal; (2) the variety of potential probiotic organisms to evaluate; (3) the diversity ofpossible diseases or conditions that might be influenced by probiotics; and (4) the need forplausible mechanisms.Furthermore, there is a need for validated “biomarkers”—biological char-acteristics that predict underlying health or disease and that can be objectively measured.

Finally, clinical studies, designed like those in the pharmaceutical industry, have not (until re-cently) been done. In an effort to reconcile these issues, several scientific organizations (e.g., In-ternational Scientific Association for Probiotics and Prebiotics, the Canadian Research and De-velopment Centre for Probiotics, and Proeuhealth) have been formed to provide structure tothe debate, assess the state of the science, coordinate research, and, ultimately, formulate rec-ommendations.

Which organisms are probiotic?

In truth, the probiotic field is replete with mis-named,mis-characterized,and mis-represented or-ganisms. In some cases, an organism was long ago assigned to a particular species and the namestuck, even though the particular species was no longer officially recognized.As a result, someprobiotic-containing foods and formulations contain microorganisms whose value, as a probi-otic, at least, is rather dubious. Moreover, the lack of appropriate regulations or standards (withregard to probiotics in foods) means that for cultured dairy products and other foods that pro-fess to contain probiotic bacteria, there may or may not be organisms present at sufficient levelsand viability (assuming they are alive), or that have the expected functional characteristics.

0

100

200

300

400

500

2000s1990s1985 1989

snoitacilbup fo rebmu

N

Years

(Continued)

Figure 1. Publications on probiotics in thePubMed database. Citations were obtained using“probiotic” or “probiotics” as key words for eachof the year(s) indicated. Data for 2005 arethrough February.



A good example of the nomenclature confusion concerns the organism “Lactobacillus sporo-genes.”This organism was described in early editions of Bergey’s Manual (prior to 1939), buthas since been shown to be a misclassification and now “Lactobacillus sporogenes” is consid-ered to be scientifically invalid (Sanders et al., 2003).

It is important to recognize, however, that just because a given organism is appropriatelynamed does not ensure that the organism has probiotic activity. In other words, while solid evi-dence may exist to support the probiotic functions of a specific strain of Bifidobacterium bi-fidus or Lactobacillus acidophilus, that does not mean that all strains of those organisms meetthe minimum relevant criteria for a probiotic or that they have even been evaluated in any wayfor probiotic activity (Table 1).

Table 1. Functional characteristics of probiotic bacteria1.

Property Survival Colonization Bio-functional Safety

Acid tolerance +Bile tolerance +Stress tolerance +Compatibility with starter culture +Adherence factors +Attachment factors +Prebiotic metabolism +High growth rate +Competition factors +Produce short-chain acids +Immunomodulating +Hypocholesteremic +Antitumorogenic +Enhance gut barrier function +Lactose hydrolysis +Antimicrobial activity +Stimulate mucus production +Avirulent +Noninflammatory +Absence of side-effects +Absence of gene transfer +

1Note, the desirability of these properties for a given probiotic is target-dependent

In addition, the number and viability of cells present in the food vehicle or formulation maybe inadequate to promote a probiotic effect. For example, 109 cells per gram are thought to bethe minimum level that should be present in yogurt, although no regulatory requirement cur-rently exists. Moreover, questions have been raised regarding whether bacteria in yogurt actu-ally survive the route through the alimentary system. It was recently reported, for example, thatvolunteer subjects consuming yogurt daily for two weeks failed to shed yogurt bacteria in theirfeces at any time during the study (del Campo et al., 2005).Another complication concerns theclaim that organisms other than bifidobacteria and lactobacilli may also have probiotic effects.These include organisms not ordinarily associated with growth in the intestinal tract, such asBacillus, Sporolactobacillus, and Brevibacillus (Sanders et al., 2003).Whether or not these or-ganisms can actually survive digestion and have an impact in the gastrointestinal tract remainsuncertain. In addition,enterococci,Escherichia coli, and even yeasts (especially Saccharomycesboulardii) have been similarly promoted as probiotics and are also widely available. Finally,

Box 4–5. Sorting Through the Pros and Cons of Probiotics (Continued)



questions have been raised in recent years regarding the safety of lactobacilli and other lacticacid bacteria (despite their long time consumption in foods).

These safety concerns are based mainly on just a few sporadic cases in which probiotic or-ganisms were associated with endocarditis, bacteremia, or other invasive infections or underly-ing conditions (Borriello et al., 2003). Most of these cases, however, involved immunocomprim-ised or at-risk individuals undergoing treatments that would expose them to any opportunisticmicrobial agent.Moreover, the genome sequences from several probiotic bacteria are now avail-able (Chapter 2), and no functional virulence factors have yet been identified.

It is important to recognize that some bacteria used as probiotics are intrinsically resistant toantibiotics, and, in some cases, may even harbor antibiotic resistant genes on mobile genetic el-ements (Temmerman et al., 2003; von Wright, 2005). As noted above, Bacillus cereus may beused as a probiotic, despite the fact that some strains of this species cause foodborne disease.Thus, safety evaluations must be considered for any strain to be used a probiotic.

Making the Case for Probiotics

As noted above, literally thousands of research articles have been published on probiotics,with many suggesting that beneficial health effects occur as a result of consuming probioticsor probiotic-containing foods. However, these studies are frequently criticized for either lack-ing relevant controls or having relied on poor experimental design. Even when a particularstudy reports a positive result, it is frequently followed by another study in which such effectsare absent.

Therefore, determining which claims are supported by solid scientific evidence and identify-ing the appropriate criteria to evaluate this evidence have become critical goals (Pathmakanthanet al., 2000; Sanders et al., 2004). Indeed, meta-analyses and other retrospective studies, in whichpublished reports are critically evaluated, indicate that while some claims are weakly supportedby experimental or clinical data, others appear to be more convincing (Tables 2 and 3).

Table 2. Probiotics: evidence for intestinal health benefits1.

Reasonable evidence Suggestive evidence Potential use

Inflammatory bowel disease Pouchitis5 Crohn’s disease3,5

Chronic pouchitis2 Traveler’s diarrhea3,4 Ulcerative colitis3

Ulcerative colitis2 Episodic diarrhea4 Pouchitis3

Intestinal infectionsAcute diarrhea3

Relapsing Clostridium difficile4

Antibiotic-associated diarrhea4

1Adapted from Sartor, 20052Prevention of relapse3Treatment4Prevention5Postoperative prevention

For example, based on a review of the literature on the effect of probiotic bacteria on diar-rheal disease,serum cholesterol, the immune system,and cancer prevention, the only consistentfinding was a positive effect of Lactobacillus GG as a treatment against rotavirus intestinal in-fections (de Roos and Katan, 2000).This result was supported by a meta-analysis that also con-cluded that Lactobacillus therapy could be used safely and effectively to treat diarrheal diseasein children (Van Niel et al., 2002).


(Continued)



Table 3. Probiotics: recommendations according to evidence-based medicine1,2.

Recommendation Evidence Grade3 Level3 Claim, treatment, or benefit

Grade A Level 1A Treatment of infectious diarrhea in childrenPrevention of antibiotic-associated diarrheaPrevention of nosocomial and community acquired diarrhea in childrenTreatment of lactose malabsorption

Grade A Level 1B Prevention and maintenance of remission of pouchitisPrevention of post-operative infectionsPrevention and management of atopic diseases in children

Grade B Level 2 Prevention of traveler’s diarrheaPrevention of pancreatitis-associated sepsisMaintenance of remission of ulcerative colitisLowering of blood cholesterol

1According to this approach, recommendations regarding the effectiveness of a treatment for particular diseases are assigned a “grade”

that is based on the strength of the clinical evidence that supports the specific claim.The evidence is also assigned a strength “level” that

is based on the experimental methodology and the manner in which the supporting studies were performed. Level 1A and 1B evidence

are from multiple or single, respectively, randomized,placebo-controlled trails with low false-positive rates, adequate sample size, and ap-

propriate methodology. Level 2 evidence is from randomized, placebo-controlled studies with high false-positive rates, inadequate sam-

ple size, or inappropriate methodology. Accordingly, Grade A recommendations are supported by evidence from two or more Level 1

studies without conflicting evidence from other Level 1 studies.Grade B recommendations are supported by evidence from two or more

Level 1 studies with conflicting evidence from other Level 1 studies or by evidence from two or more Level 2 studies.2Adapted from Gill and Guamer, 20043Grade recommendations and hierarchies of evidence as defined by Fennerty, 2003

The ability of probiotic organisms (lactobacilli and Saccharomyces boulardii) to prevent di-arrheal disease was also suggested by a meta-analysis of nine trials, but treatment efficacy wasnot supported by the data (D’Souza et al., 2002).Another meta-analysis on the effect of probi-otics on antibiotic-associated diarrhea similarly concluded that the evidence for beneficial ef-fects was not definitive (Cremonini et al., 2002).

Objective Measures of Probiotic Activity

Determining whether a probiotic has actually had a biological effect is not always easy. Clearly,it is not difficult to measure changes in cholesterol or blood lipid levels or other relevant mark-ers in clinical trials (whether the treatment is a probiotic or a pharmaceutical drug). However,it is an entirely different matter when the outcome marker takes longer to become evident orwhen the marker provides an indirect measure of efficacy.

Double-blind, randomized, and placebo-controlled studies are essential, considering that inmany cases any probiotic effect is often based on the subjective judgment of human subjects.For example, a positive effect of probiotic feeding on digestion and intestinal well-being in oth-erwise healthy subjects may be difficult to substantiate. In contrast, objective measures, or bio-markers, can provide much more credible evidence. Biomarker measures may, in fact, show nodifferences between treatments, even when subjective scores appear to suggest that differ-ences do exist.For many studies,biomarkers are mandatory, since they may be the only reliable,non-invasive way to measure the efficacy of probiotics. Biomarkers, which must first be vali-dated for efficacy,may include immunological factors,GI tract populations,blood lipids, and fer-mentation metabolites. More than twenty colon cancer biomarkers were described in one re-cent report (Rafter, 2002).




yogurt are largely responsible for this effect,since heat-treated yogurt is not as effective at al-leviating lactose intolerance symptoms. It is in-teresting, however, that while yogurt can be tol-erated, other cultured milk products such assour cream and buttermilk cannot.The explana-tion for this observation is related directly to the

different routes by which lactose is metabolizedby the cultures used in the manufacture of theseproducts.

The fermentation that occurs in sour creamand cultured buttermilk production is per-formed mainly by lactococci (described below),whereas S. thermophilus and L. delbrueckii

References

Borriello,S.P.,W.P.Hammes,W.Holzapfel,P.Marteau, J.Schrezenmeir,M.Vaara,and V.Valtonen.2003.Safety ofprobiotics that contain lactobacilli or bifidobacteria. Clin. Infect. Dis. 36:775–780.

Cremonini, F., S. Di Caro, E.C. Nista, F. Bartolozzi, G. Capelli, G. Gasbarrini, and A. Gasbarrini. 2002. Meta-analysis: the effect of probiotic administration on antibiotic-associated diarrhoea.Aliment. Pharmacol.Ther. 16:1461–1467.

del Campo,R.,D.Bravo,R.Cantón,P.Ruiz-Garbajosa,R.García-Albiach,A.Montesi-Libois,F.-J.Yuste,V.Abraira,and F. Baquero. 2005. Scarce evidence of yogurt lactic acid bacteria in human feces after daily yogurtconsumption by healthy volunteers.Appl. Environ. Microbiol. 71:547–549.

de Roos, N.M., and M.B. Katan. 2000. Effects of probiotic bacteria on diarrhea, lipid metabolism, and car-cinogenesis: a review of papers published between 1988 and 1998.Am. J. Clin. Nutr. 71:405–411.

D’Souza,A.L., C. Rajkumar, J. Cooke, and C.J. Bulpitt. 2002. Probiotics in prevention of antibiotic diarrhoea:meta-analysis. Br. Med. J. 324:1361–1366.

Fennerty, M.B. 2003.Traditional therapies for irritable bowel syndrome: an evidence based appraisal. Rev.Gastroenterol. Disord. 3 (Supple. 2):S18–S24.

Gill, H.S., and F. Guarner. 2004. Probiotics and human health: a clinical perspective. Postgrad. Med. J.80:516–526.

Pathmakanthan, S., S. Meance, and C.A. Edwards. 2000. Probiotics: a review of human studies to date andmethodological approaches. Microb. Ecol. Health Dis. 12 (Supple. 2):10–30.

Rafter, J.J.2002.Scientific basis of biomarkers and benefits of functional foods for reduction of disease risk:cancer. Br. J. Nutr. 88 (Supple. 2):S219–S224.

Sanders, M.E.,T.Tompkins, J.T. Heimbach, and S. Kolida. 2004.Weight of evidence needed to substantiate ahealth effect for probiotics and prebiotics—regulatory considerations in Canada, E.U., and the U.S. Eur.J. Nutr. 44:303–310.

Sanders, M.E., L. Morelli, and T.A.Tompkins. 2003. Sporeformers as human probiotics: Bacillus, Sporolacto-bacillus, and Brevibacillus. Comp. Rev. Food Sci. Food Safety. 2:101–110. (www.ift.org/publications/crfsfs).

Sartor, R.B. 2005. Probiotic therapy of intestinal inflammation and infections. Curr. Opin. Gastroenterol.21:44–50.

Temmerman, R., B. Pot, G. Huys, and J. Swings. 2003. Identification and antibiotic susceptibility of bacterialisolates from probiotic products. Int. J. Food Microbiol. 81:1–10.

Van Niel, C.W., C. Feudtner, M.M. Garrison, and D.A. Christakis. 2002. Lactobacillus therapy for acute infec-tious diarrhea in children: a meta-analysis. Pediatrics 109:678–684.

von Wright, A. 2005. Regulating the safety of probiotics—the European approach. Curr. Pharm. Design.11:17–23.

Probiotic websites

www.proeuhealth.vtt.fiwww.isapp.netwww.crdc-probiotics/cawww.usprobiotics.org




subsp.bulgaricus are responsible for the yogurtfermentation. Recall that the latter organismshydrolyze lactose via �-galactosidase.When yo-gurt is consumed and these bacteria reach thesmall intestine, cell lysis occurs (due to bilesalts), intracellular �-galactosidase is released,and lactose is hydrolyzed.Thus, the yogurt cul-ture serves as an exogenous source of this en-zyme, substituting for the �-galactosidase notproduced by the host. In contrast, lactococciproduce phospho-�-galactosidase.The substrateof this enzyme is lactose-6-phosphate (the prod-uct of the lactose PTS), which can only beformed by intact cells.This enzyme does not hy-drolyze free lactose.Thus, when lactococci areingested and lysed in the small intestine, there

will be little, if any, impact on lactose digestion,and lactose-intolerant individuals will, unfortu-nately,get no relief.

Frozen Yogurt and Other Yogurt Products

Frozen yogurt is a product for which no fed-eral standard of identity exists.This has led toconfusion and even mis-representations aboutwhat this product actually is. Several statesnow have their own definitions of frozen yo-gurt (mainly dealing with fat levels), and theFDA has been petitioned to implement stan-dards of identity. The main reason why suchmeasures are thought necessary is to protect

colonic bacteria

Lactose Intolerance

stomach

colon

small

intestineGalGlu

E

Lac

Normal

+

Yogurt

Lac Lac

Lac Lac Lac

E

GalGlu

Lac

E

St Lb

acid + gas +

osmotic unbalance

E

Figure 4–4. In normal individuals (left panel), ingested lactose (Lac) passes through the stomach. Hydroly-sis occurs in the small intestine via endogenous �-galactosidase (E) produced by cells that coat the surface ofthe intestinal wall. The hydrolysis products glucose (Glu) and galactose (Gal) are subsequently adsorbed. Inlactose-intolerant individuals (center panel), those intestinal cells do not produce �-galactosidase, and thelactose escapes hydrolysis and instead reaches the colon where it causes colonic distress, either by virtue ofits metabolism by anaerobes or by osmotic disturbances. When yogurt is consumed (right panel), Strepto-coccus thermophilus (St) and Lactobacillus bulgaricus (Lb) serve as a source of exogenous �-galactosidase(after permeabilization in the small intestine) and restore normal lactose hydrolyzing ability to the host.



the image of yogurt as a healthy, nutritiousfood—one low in fat and calories and contain-ing health-promoting bacteria. Obviously, thereare marketing reasons for including “yogurt”onthe label of a product, whether or not thatproduct actually contains yogurt, or what con-sumers think is yogurt. For example, yogurt-covered pretzels, yogurt-covered raisins, andyogurt-containing salad dressings are made us-ing material that may have once been yogurt,but after processing and dilution with other in-gredients, little of the original yogurt charactermay actually be present. Similarly, ice creammix containing only a trace of yogurt could, intheory, be called frozen yogurt and gain themarket advantages associated with yogurt. Cer-tainly, and perhaps most importantly, the pres-ence of live yogurt organisms should not be as-sumed for these products.

As noted, some states have written defini-tions for frozen yogurt. However, while theremay be requirements for the presence of par-ticular organisms, there are generally no suchrequirements for minimum numbers. In Min-nesota, for example, frozen yogurts refer to a“Frozen dairy food made from a mix contain-ing safe and suitable ingredients including, butnot limited to, milk products. All or a part ofthe milk products must be cultured with acharacterizing live bacterial culture that con-tains the lactic acid producing bacteria Lacto-bacillus bulgaricus and Streptococus ther-mophilus and may contain other lactic acidproducing bacteria.” Some states require thatthe frozen yogurt products have a minimumacidity, ranging from 0.3% to 0.5%.

Cultured Buttermilk

Buttermilk is the fluid remaining after cream ischurned into butter. It is a thin, watery liquidthat is rarely consumed as a fluid drink. Be-cause it is rich in phospholipids (derived fromthe rupture of milk fat globules during churn-ing), it has excellent functional properties andis an especially good source of natural emulsi-fiers. It is typically spray dried and used as aningredient in processed food products. Cul-

tured buttermilk, in contrast, is made fromskim or low-fat milk that is fermented by suit-able lactic acid bacteria.The only relation thisproduct has to buttermilk is that butter gran-ules or flakes are occasionally added to providea buttery flavor and mouth feel. How, then, didthis product come to be called buttermilk? Inthe traditional manufacture of butter, it wascommon practice to add a mixed, undefinedlactic culture to cream prior to churning.Thelactic acid would provide a pleasant tart flavor,and the cream-ripened butter would be betterpreserved.The resulting by-product, the butter-milk, would also be fermented.

Cultured Buttermilk Manufacture

Cultured buttermilk is usually made from low-fat milk, although non-fat and whole milk ver-sions also exist. Non-fat milk solids are fre-quently added to give about 10% to 12%non-fat solids (Figure 4–5). Next, the milk isheated to 85°C to 88°C for thirty minutes.This not only pasteurizes the milk, but alsosatisfies other functional requirements, as de-scribed above for yogurt. The mix is thencooled to 21°C to 22°C and inoculated with amesophilic starter culture, specific for cul-tured buttermilk.

The starter culture for buttermilk usuallycontains a combination of acid-producing bac-teria and flavor-producing bacteria, in a ratio ofabout 5:1. Many culture suppliers now also of-fer body-forming (i.e., EPS-producing) strains.The acid producers include strains of L. lactissubsp. lactis and/or L. lactis subsp. cremoris.These bacteria are homolactic, and their func-tion is simply to produce lactic acid and lowerthe pH. For manufacturers who prefer a decid-edly tart, acidic product, the acid-producingstrains are sufficient.

It is more common, however, to includeflavor-producers in the starter culture.Among the flavor-producing bacteria used inbuttermilk cultures are L. lactis subsp. lactis(diacetyl-producing strains), Leuconostocmesenteroides subsp.cremoris, or Leuc. lactis.The latter organisms are heterofermentative,



producing lactic acid, as well as smallamounts of acetic acid, ethanol, and carbondioxide. These metabolic end-products con-tribute to the flavor and, in the case of thecarbon dioxide, to the mouth feel of the

product. Importantly, these bacteria also havethe metabolic capacity to ferment citric acidand to produce diacetyl, a compound thathas major impact on the flavor of culturedbuttermilk. Diacetyl has a buttery aroma and

Milk

Buttermilk Mix

Heat Treatment

Cool

Inoculate

Incubate

Cool and Agitate

Homogenize

85°C for 30 minutes

20-22°C

Standardize (nonfat, lowfat, whole)

Nonfat milk solids


0.5% mesophilic culture (containing acidproducing and citrate-fermenting strains)

20-22°C, 12 to 16 hours

Citrate, salt

Cool

Dispense

8-10°C

4°CFigure 4–5. Manufacture of cultured buttermilk.



flavor, and, according to buttermilk afficiona-dos, imparts a delicate and characteristic fla-vor. The ability of the starter culture to per-form the citrate fermentation, therefore, is acritical trait. How lactic acid bacteria convertcitrate into diacetyl has been a subject ofconsiderable interest, not only because of thepractical importance of the pathway in fer-mented foods, but also because it posed aparticular challenge to biochemists. For manyyears, two pathways, one enzymatic and theother, non-enzymatic, were thought possible.

In both pathways, the early steps are thesame (Figure 4–6;also see Chapter 2).Citrate isfirst transported by a citrate permease (CitP)that is pH-dependent, with an optimum activ-ity between pH 5.0 and 6.0. The intracellularcitrate is then hydrolyzed by citrate lyase to

form acetate and oxaloacetate. The oxaloac-etate is decarboxylated by oxaloacetate decar-boxylase to give pyruvate and carbon dioxide,and the acetate is released into the medium.Normally, lactic acid bacteria reduce pyruvateto lactate, which is then excreted. However,pyruvate reduction requires NADH, which isordinarily made during glycolysis, but which isnot made during citrate fermentation. In theabsence of reduced NADH, pyruvate would ac-cumulate inside the cell, eventually reachingtoxic or inhibitory levels.The excess pyruvateis instead oxidatively decarboxylated by thi-amine pyrophosphate (TPP)-dependent pyru-vate decarboxylase and acetaldehyde-TPP isformed (see Chapter 2 for details). The lattercompound condenses with another pyruvatemolecule to form �-acetolactate, a reaction

citrate

CO2 pyruvate

citrate

CitP

out

in

oxaloacetate2-

acetatecitratelyase

TPPpyruvate

decarboxylase

CO2

acetaldehyde-TPP

a-acetolactatesynthase

a-acetolactate/

O2

CO2

diacetyl acetoin

diacetylreductase

acetoinreductase

2,3-butanediol

NADH+HNADH+H NAD

NAD

oxaloacetatedecarboxylase

Figure 4–6. Citrate fermentation pathway inlactic acid bacteria. Adapted from Hutkins, 2001.



catalyzed by �-acetolactate synthase and requiring high concentrations of pyruvate.The�-acetolactate is unstable in the presence ofoxygen and is finally decarboxylated, nonenzy-matically, to form diacetyl. It is important tonote that diacetyl is not necessarily the termi-nal end-product of the pathway. Further reduc-tion of diacetyl can also occur, forming acetoinand 2,3-butanediol, compounds that con-tribute no flavor or aroma to the product.

Following the addition of the culture, themix is incubated at 21°C to 22°C for twelve tosixteen hours.At the end of the fermentation,when the titratable acidity has reached 0.85%to 0.90% and the pH has decreased to about4.5, the product is cooled to 2°C and agitatedto break up the coagulum.Salt is usually added,and, if desired, butter granules or flakes areadded.The finished product should be viscousand pourable.The product is pumped into con-tainers and distributed.

Factors Affecting Diacetyl Formationin Cultured Buttermilk

Even if the culture contains citrate-formingstrains, diacetyl formation does not always oc-cur in amounts necessary to impart the desiredflavor.Several reasons may account for reducedcitrate fermentation and diacetyl synthesis.First, there may simply not be enough citrate inthe milk. Although milk contains, on average,about 0.15% citrate, this amount varies, de-pending largely on the diet of the cow.There-fore, sodium citrate is frequently added to themilk to provide a consistent source of sub-strate. If the temperature during incubation istoo high (�24°C), growth of the homolacticlactococci will be favored, too much acid willbe produced, and the citrate-fermentors maybe inhibited. Not only will the product have a“lacks flavor” defect, but it will instead have aharsh acid flavor.

In contrast, if the incubation temperature istoo low (� 20°C), not enough acid is producedand the flavor will be flat or “lacks acid.”Acid pro-duction is also necessary for diacetyl formation,since the citrate transport system is not acti-

vated unless the pH is below 5.5. In fact, maxi-mum diacetyl synthesis occurs between pH 5.0and 5.5.Finally,once diacetyl is made, low pH in-hibits the reduction reactions that convert di-acetyl to acetoin and 2,3-butanediol.One otherfactor that is critical for synthesis of diacetyl isoxygen, which can stimulate diacetyl formationby as much as thirty-fold. Several mechanismsare responsible for the oxygen effect. As notedabove, pyruvate, the metabolic precursor of di-acetyl,can serve as the substrate for several alter-native enzyme reactions, including lactate dehy-drogenase. In the presence of high atmosphericoxygen, lactate dehydrogenase activity is re-duced, and the oxidative decarboxylation reac-tion responsible for diacetyl synthesis is en-hanced. In addition, by oxidizing NADH, oxygenalso slows the rate at which diacetyl is reducedto acetoin or 2,3-butanediol.For these reasons, itmay be useful to stir air into the final productduring the agitation step. In fact, maximum di-acetyl production occurs about seventy-twohours after manufacture, so fresh product maynot have as good a flavor as one that is slightly“aged”.The flavor and texture defects that occurin cultured buttermilk are, in general, similar tothose in yogurt.The most common flavor defectis “lacks flavor,” caused by insufficient diacetylformation due to reasons outlined above.Excessacidity is also objectionable to consumers.Texture defects include wheying off, too thin,ortoo viscous.The latter may be caused by body-forming bacteria that over-produce EPS. Milk-borne defects, such as yeasty flavor,unclean,andrancidity, are caused by poor quality milk, espe-cially milk that had been contaminated with psy-chrotrophic microorganisms.

Sour Cream

Despite the apparent differences in the appear-ance and texture, sour cream is actually quitesimilar to cultured buttermilk in several re-spects (Figure 4–7).The sour cream culture, forexample, is the same as that used for butter-milk.The incubation conditions and the flavorcompounds produced by the culture are alsosimilar for both products. There are, however,



several notable differences. For sour creammanufacture, cream (containing varying levelsof milkfat) is used instead of lowfat or skimmilk.The cream is pasteurized, but not quite atthe severe conditions used for buttermilk or yo-gurt.This is because for sour cream, denatura-tion of protein is not as crucial, since the milkfat will impart the desired creaminess, thick-ness, and body. The cream is homogenized,which also promotes a desirable heavy body.

Sour Cream Manufacture

In the United States the starting material forsour cream manufacture is cream at 18% to20% milkfat. Lower-fat versions, such as sourhalf-and-half (10% milkfat), are produced insome regions, as are higher-fat products (con-taining as much as 50% milkfat).The latter aregenerally produced as an ingredient for dipsand other sour cream products. There is no

Sour CreamMix

Heat Treatment

Cool

Inoculate

Dispense Incubate

Cool

Cool anddispense

Homogenize

85°C for 25 seconds

20-22°C

Standardize to 18 - 20% (less for reduced fat versions)


1% mesophilic culture

Incubate

20-22°C, 12 to 16 hours

0-4 °CCool

20-22°C

Cream

8-10°C

Figure 4–7. Manufacture of sour cream.



need to add additional nonfat dry milk, as foryogurt and buttermilk, since achieving a firmgel and thick body is not an issue for thesehigh-solids products. For the same reason, thecream is pasteurized at more typical time andtemperature conditions (85°C for twenty-fiveseconds). However, the cream must be homog-enized (usually twice) to produce a smooth-textured product with good viscosity.After themix is cooled, the sour cream culture is added.The culture, containing mesophilic acid-producing, flavor-producing, and body-formingstrains, is often the same as that is used for cul-tured buttermilk. The mix is then either filledinto cups and incubated or incubated directlyin vats (analogous to the two styles of yogurt).Incubation is at 20°C to 25°C for ten to sixteenhours. When the pH reaches about 4.4 to 4.6(about 0.7% to 0.9% lactic acid), the sour creamis cooled, either by moving the cup-fermentedproduct into coolers or, in the case of vat-fer-mented product, by stirring the product injacket-cooled vats.The product is then pumpedinto containers.Sour cream should have a simi-lar flavor profile as cultured buttermilk, withlactic acid and diacetyl predominating. Bodycharacteristics are especially important, andvarious gums and other stabilizing agents arefrequently added to the mix. Some manufactur-ers even add a small amount of chymosin toprovide additional firmness. When defects dooccur, they are often due to poor quality ingre-dients and post-pasteurization contaminationby acid-tolerant yeasts and molds and psy-chrotrophic bacteria.

Kefir

Kefir is a fermented dairy product that is de-scribed in many reference texts and has longbeen of academic interest to microbiologists,but was one for which few U.S. students (orconsumers for that matter) had any first-handknowledge. Although several brands of kefirare now available throughout the UnitedStates, it is still a mostly unknown product.Thisis despite the popularity of this productthroughout a large part of the world. In the

Middle East, Eastern Europe, and Central Asia,especially Turkey, Russia, Ukraine, Poland, andthe Czech Republic, kefir is one of the mostwidely consumed cultured dairy products, ac-counting for as much as 70% of all culturedmilk products consumed. It is interesting thatyogurt manufacturers have begun to introducefluid or pourable yogurt-like products that areonly slightly different than traditional kefir.

Kefir originated in the Caucasus Mountainin Russia. The traditional kefir manufacturingprocess, which is still widely practiced, relieson a mixed assortment of bacteria and yeast toinitiate the fermentation (Figure 4–8).The kefirfermentation is unique among all other dairyfermentations in that the culture organisms areadded to the milk in the form of insoluble par-ticles called kefir grains (Box 4–6). Moreover,once the fermentation is complete, the kefirgrains can be retrieved from the fermentedmilk by filtration and reused again and again.

The flavor of plain kefir is primarily due tolactic and acetic acids, diacetyl, and acetalde-hyde, produced by homofermentative and het-erofermentative lactic acid bacteria. However,because kefir grains also contain yeast, in addi-tion to lactic acid bacteria, other end-productsare formed that make the finished product quitedifferent from other cultured dairy products.This is because ethanol is produced when theyeasts ferment lactose, such that kefir can con-tain as much as 2% ethanol (accounting, per-haps, for its appeal). In the United States, thismuch ethanol would trigger action from theregulatory agencies (i.e., Alcohol and TobaccoTax and Trade Bureau).Thus, if yeast are presentin the culture (most kefir products made in theUnited States claim yeasts on their labels), theymust be low- or non-ethanol producers. Kefirgrains can also contain acetic acid-producingbacteria, such as Acetobacter aceti.

Kefir Manufacture

In the United States, kefir is usually made withlow-fat or whole milk (Figure 4–8).The milk ispasteurized at 85°C to 90°C for up to thirtyminutes (like yogurt), then cooled to 20°C to



22°C. Kefir grains are generally not used in theUnited States; instead, the milk is inoculatedwith a kefir starter culture containing Lacto-bacillus kefiranofaciens and Lactobacillus ke-fir. The mixture is incubated for sixteen totwenty-four hours or until a pH of 4.2 to 4.6 isreached. The coagulum is then gently stirred,flavoring ingredients (usually fruit) are added,and the product is then dispensed. Kefirshould contain 0.8% to 1.0% lactic acid alongwith other heterofermentative end-products(acetic acid, ethanol, and CO2). Acetaldehydeand diacetyl are also usually formed.The kefirwill have a tart flavor and a smooth, viscousbody. Production of CO2 by the starter culturealso provides effervescence and mouth feel.

Other Cultured Dairy Products

The emphasis of this chapter has been on thoseproducts produced and consumed in the UnitedStates. However, there are hundreds of otherproducts produced around the world that, al-though manufactured via similar processes,haveunique and interesting features (Table 4–4).Villi,for example, is widely consumed in Finland isknown for its high viscosity and musty flavorand aroma.The ropy texture is due to capsularEPS production by L. lactis subsp. lactis and L.lactis subsp. cremoris, and the musty flavor iscaused by growth of the fungal organism Geot-richum candidum.Another popular product isKoumiss,which was traditionally produced from

Boil for 10 minutes

Inoculate with kefir grains

Cool to 20°C

Raw milk

Traditional

Incubate at 20 °C for up to 24 h

Separate kefir grains by sieving

Cool to 4°C and store

Dispense into containers

Cool and gently agitate at 4°C

Heat at 85-95°C for 5-30 minutes

Dispense into containers

Modern

Homogenize

Standardize (lowfat, whole)

Nonfat milk solids

Stabilizers, flavors, other ingredients

Raw milk

Cool to 20°C

Inoculate with kefir starter culture

Incubate at 20°C for up to 24 h

Add fruit or bulky flavors

Figure 4–8. Traditional and modern manufacturing process for kefir. Otles and Cagindi. 2003.



Box 4–6. The Microbial Diversity of Kefir Grains

To microbial ecologists,kefir grains represent an amazingly diverse ecosystem.A single grain canharbor more than a million cells, with perhaps dozens of different species present (Table 1).The grains are essentially a natural immobilized cell bioreactor, with the kefir grain serving as aninert support material for the cells.

Table 1. Bacteria, yeast, and mold isolated from kefir grains1.

Bacteria Yeasts and Mold

Lactobacillus brevis Candida kefir2

Lactobacillus fermentum Candida marisLactobacillus kefiranofaciens3 Candida inconspicuaLactobacillus kefiri Candida lambicaLactobacillus parakefir Candida kruseiLactobacillus plantarum Saccharomyces cerevisiaeLactococcus lactis subsp. cremoris Kluyveromyces marxianusLactococcus lactis subsp. lactis Geotrichum candidumLeuconostoc mesenteroides subsp. cremoris Zygosaccharomyces sp.Leuconostoc sp.Streptococcus thermophilusAcetobacter aceti

1Adapted from various sources (in references)2syn Candida kefyr3Recently reclassified as Lactobacillus kefiranofaciens subsp. kefirgranum or subsp. kefiranofaciens (Vancanneyt

et al., 2004)

Attachment to the grains is mediated via one of several different types of polysaccharide-likematerial (kefiran being the prototype) that are produced by the resident organisms.The kefir or-ganisms do not appear to be randomly distributed; rather, some are contained mainly within thecore of the grains, whereas others are located primarily in the exterior. Symbiotic and synergis-tic relationships likely exist between different organisms, but these relationships are difficult toestablish in such a complex system (Leroi and Pidoux, 1993a and 1993b).

Although the precise chemical composition of kefir grains varies depending on the source,they generally contain about 80% to 90% water, 2% to 5% protein, and 8% to 10% carbohydrate(Abraham and De Antoni,1999;Garrote et al.,2001).Most of the protein faction is in the form ofcasein, whereas most of the carbohydrate portion is polysaccharide.The irregular, cauliflower-shaped grains range in size from 0.2 cm to more than 2 cm in diameter. In their dry form, theyare stable for many months. Although kefir grains are available commercially, many modernmanufacturers now use pure lyophylized cultures containing many of the strains ordinarilyfound in the grains.

The predominant organisms in kefir grains are lactic acid bacteria,with Lactobacillus speciesaccounting for up to 80% of the total (Garrote et al., 2001; Simova et al., 2002;Vancanneyt et al.,2004). Other studies have shown that homofermentative species, including Lactobacillus kefir-granum and Lactobacillus kefiranofaciens, were the most frequently isolated species in kefirgrains, whereas heterofermentative Lactobacillus kefir and Lactobacillus parakefir were lesscommon (Takizawa et al., 1998; Simova et al., 2002). Other lactic acid bacteria that have beenidentified include species of Lactococcus, Streptococcus, and Leuconostoc.Yeasts are also well-represented, and include lactose-fermenting (Saccharomyces kefir) and non-fermenting strains.

It was reported recently that kefir grains likely contain organisms that cannot easily be cul-tured and that are instead only viable within the confines of the grain environment (Witthuhnet al., 2005). Collectively, the kefir microflora has considerable metabolic diversity, which ex-



plains, in part, the incredible stability and viability of the grains, as well as the wide spectrum ofproducts produced during the fermentation.

References

Abraham,A.G., and G.L. De Antoni. 1999. Characterization of kefir grains grown in cow’s milk and in soyamilk. J. Dairy Res. 66:327–333.

Leroi, F., and M. Pidoux. 1993a. Detection of interactions between yeasts and lactic acid bacteria isolatedfrom sugary kefir grains. J.Applied Bacteriol. 74:48–53.

Leroi, F., and M. Pidoux. 1993b. Characterization of interactions between Lactobacillus hilgardii and Sac-charomyces florentinus isolated from sugary kefir grains. J.Applied Bacteriol. 74:54–60.

Garrote,G.L.,A.G.Abraham,and G.L.De Antoni.2001.Chemical and microbiological characterization of ke-fir grains. J. Dairy Res. 68:639–652.

Simova, E., D. Beshkova,A.Angelov,T. Hristozova, G. Frengova, and Z. Spasov. 2002. Lactic acid bacteria andyeasts in kefir grains and kefir made from them. J. Ind. Microbiol. Biotechnol. 28:1–6.

Takizawa, S., S. Kojima, S.Tamura, S, Fujinaga,Y. Benno, and T. Nakase. 1998.The composition of the Lacto-bacillus flora in kefir grains. System.Appl. Microbiol. 21:121–127.

Vancanneyt, M., J. Mengaud, I. Cleenwerck, K.Vanhonacker, B. Hoste, P. Dawynndt, M.C. Degivry, D. Ringuet,D. Janssens, and J. Swings. 2004. Reclassification of Lactobacillus kefirgranum Takizawa et al. 1994 asLactobacillus kefiranofaciens subsp. kefirgranum subsp. nov. and emended description of L. kefira-nofaciens Fujisawa et al. 1988. Int. J. Syst. Evol. Microbiol. 54:551–556.

Witthuhn, R.C.,T. Schoeman, and T.J. Britz. 2005. Characterisation of the microbial population at differentstages of Kefir production and Kefir grain mass cultivation. Int. Dairy J. 15:383–389.

Box 4–6. The Microbial Diversity of Kefir Grains (Continued)

Table 4.4. Cultured dairy products from around the world.

Product Origin Culture Organisms Unique Features

Villi Finland Lactococcus spp. Ropy textureLeuconostoc spp. Musty flavorGeotrichum candidum

Skyr Iceland Lactobacillus delbreckii Concentrated, highsubsp. bulgaricus protein content


Dahi India Lactobacillus delbreckii Yogurt-likesubsp. bulgaricus

Streptococcus thermophilusLactobacillus spp.

Koumiss Russia Lactobacillus delbreckii Mare’s milksubsp. bulgaricus � 1% ethanol

Lactobacillus acidophilusKluyveromyces spp.

Bulgarian Milk Bulgaria Lactobacillus delbreckii High acidsubsp. bulgaricus (�2% lactic acid)



mare’s milk.Similar to kefir in that lactic acid andethanol are both present, this product owesmuch of its popularity to its putative therapeuticproperties.The yogurt-like products dahi and la-ban are among the most widely-consumed cul-tured dairy products in India and the MiddleEast, respectively.

BibliographyChaves, A.C.S.D., M. Fernandez, A.L.S. Lerayer, I.

Mierau,M.Kleerebezem,and J.Hugenholtz. 2002.Metabolic engineering of acetaldehyde produc-tion by Streptococcus thermophilus. Appl. Envi-ron. Microbiol. 68:5656–5662.

Hutkins,R.W.2001.Metabolism of starter cultures,p.207241. In E.H. Marth and J.L. Steele (ed.), Ap-plied Dairy Microbiology. Marcel Dekker, Inc.,New York, NY.

Lucey, J.A. 2002. Formation and physical propertiesof milk protein gels. J. Dairy Sci.5:281–294.

Mistry, V.V. 2001. Fermented milks and creams. p.301–325. In E.H. Marth and J.L. Steele (ed.), Ap-plied Dairy Microbiology. Marcel Dekker, Inc.,New York, NY.

Otles, S. and O.Cagindi.2003.Kefir: a probiotic dairy-consumption nutritional and therapeutic aspects.Pak. J. Nutr. 2:54–59.

Reid,G.2001.Regulatory and clinical aspects of dairyprobiotics. FAO/WHO Expert Consultation onEvaluation of Health and Nutritional Properties of Powder Milk with Live Lactic Acid Bacteria.(Online at ftp://ftp.fao.org/es/esn/food/Reid.pdf)

Tamine,A.Y., and H.C. Deeth. 1980.Yogurt: technol-ogy and biochemistry. J. Food Prot.43:939–977.

Sanders, M.E. 1999. Probiotics. Food Technol. 53:67–77.

Sodini, I., F. Remeuf, S. Haddad, and G. Corrieu. 2004.The relative effect of milk base, starter, andprocess on yogurt texture: a review. Crit. Rev.Food Sci. Nutr. 44:113–137.

Vedamuthu, E.R. 1991. The yogurt story—past, pre-sent and future. Part I. Dairy Food Environ. Sanit.11:202–203. [Note, this is the first of a ten-partseries of articles on yogurt by E.R. Vedamuthuthat appeared in this journal through Vol. 12, No.6, 1992]


5

Cheese

“We are still, however, far from having arrived at a complete elucidation of all thequestions involved. It is particularly difficult to understand how various sorts ofhard cheese, apparently containing the same microflora, should each have its owncharacteristic taste and smell. There can hardly be any doubt that these sorts ofcheese in reality contain different species of bacteria, only we are unable to distin-guish them by the methods hitherto available.”

From The Lactic Acid Bacteria by S.Orla-Jensen,1918

twenty-five years from 8 Kg in 1980 to nearly14 Kg (30.1 pounds) per person per year in2003 (of U.S. made cheese).The most popularcheeses have been the American style (e.g.,Cheddar, Colby) and Italian style (e.g., Moz-zarella and pizza cheese) cheeses, accountingfor 41.5% and 40.6%, respectively,of all cheesesconsumed in the United States (Figure 5–1). Inaddition,another 0.75 Kg of imported cheese isconsumed per person per year. Worldwide,Greece (26 Kg per person per year),France (24Kg per person per year), and Italy (21 Kg perperson per year) are the leading consumers ofcheese, with other European countries not toofar behind (Figure 5–2).

On the production side, American and Italian-types cheeses are, by far, the maincheeses produced in the United States. Al-though the American-type cheeses accountedfor nearly 70% of all cheese produced in theUnited States in the 1960s, Italian-type cheesesand Mozzarella, in particular, are about to ex-ceed that of American-type cheese, based onthe current trend (Figure 5–3).

The U.S. cheese industry began in the mid-1800s, with factories opening first in New York(1851) and Wisconsin (1868).Prior to that time,cheese was mainly produced directly on farmsand sold locally. By the late 1800s, about 4,000cheese factories accounted for the nearly 100million Kg (217 million pounds) of cheese.Most

Introduction

Perhaps no other fermented food starts withsuch a simple raw material and ends up withproducts having such an incredible diversity ofcolor, flavor, texture, and appearance as doescheese. It is even more remarkable that milk,pale in color and bland in flavor, can be trans-formed into literally hundreds of differenttypes of flavorful, colorful cheeses by manipu-lating just a few critical steps. How so manycheeses evolved from this simple process un-doubtedly involved part trial and error, partluck, and plenty of art and skill. It is fair to as-sume that, until very recently, most cheesemakers had only scant knowledge of science,and microbiology in particular. Now, however,it is likely that few fermented foods requiresuch a blend of science, technology, and crafts-manship as does the making of cheese.

On a volume basis, the cheese industry isthe largest of all those involved in fermentedfoods manufacture. Of the 75 billion Kg (165billion pounds) of milk produced in the UnitedStates in 2001,more than one-third was used inthe manufacture of 3.7 billion Kg (8.1 billionpounds) of cheese. About a fourth of thatcheese was used to make various types ofprocessed cheese (discussed later).

On a per capita basis, cheese consumptionin the United States has increased in the past

145





cheese factories were located in the dairy-producing states of Wisconsin,Minnesota,Penn-sylvania,and New York.In the past twenty years,California has emerged as the leading producerof milk and the second leading (to Wisconsin)manufacturer of cheese. However, like other

segments of the food industry, more and moreproduct is made by fewer and fewer plants. In2001, there were half as many production facili-ties making American and Italian style cheese asthere were in 1980,while,at the same time,pro-duction capacity increased by 100%.

Other American (1.5)

Mozzarella(4.4)

Other Italian(1.3)

Swiss(0.5)

Other(0.7)

Cream(1.0)

Cheddar(4.3)

9

1412

21

25

20

8

22

2 2

15

7

3

17

10

0

4

8

12

16

20

24

28

Arg

netin

aAu

stra

liaC

anad

aD

enm

ark

Fran

ceG

erm

any

Gre

ece

Irela

ndIta

lyJa

pan

Mex

ico

Net

ehrla

nds

New

New

Zeal

adn

Rus

sia

Spai

nSw

eden

Uni

ted

Kin

gdom

14U

nite

dS

tate

s

Kg

30

26

7 8

Zeal

and

Egyp

t

6

Figure 5–1. Per capita consumption (Kg perperson per year) of different varieties of cheesein the United States in 2003. Adapted fromUSDA Economic Research Service statistics.

Figure 5–2. Worldwide per capita cheese consumption (Kg per person per year). Adapted from 2002 USDAand United Nations/FAO Agricultural Database statistics.


Cheese 147

The basic cheese manufacturing principleswill be outlined in this chapter, emphasizingthe key variables at each step that account forthe myriad types of cheeses that exist through-out the world. Manufacturing details for spe-cific cheeses representing the main cheesegroups or families will then be described.Along the way, an understanding of dairychemistry and the milk milieu and the role ofthe starter culture will be examined. Finally,how flavor and texture development occursduring cheese ripening, current problems andchallenges faced by the cheese industry, andother issues will be addressed.

Manufacturing Principles

Like so many fermented foods, the first cheesemade by human beings was almost certainly aresult of an accident. Some wandering nomad,as the legend goes, filled up a pouch madefrom the stomach of a calf or cow with a literor two of fresh milk.After a few hours, the milkhad turned into a solid-like material, and whenour would-be cheese maker gave the containera bit of a shake, a watery-like fluid quickly sep-

arated from the creamy white curd.This mod-erately acidic, pleasant-tasting curd and wheymixture not only had a good flavor, but it alsoprobably had a longer shelf-life than the freshmilk from which it was made.And despite therather crude production scheme, the productmade several thousands of years ago was notmuch different than many of the cheeses cur-rently produced and consumed even today.

Just what happened to cause the milk to be-come transformed into a product with such adecidedly different appearance, texture, andflavor? To answer that question, it is first neces-sary to compare the composition of the start-ing material, milk, to that of the product, thefinished cheese (Figure 5–4). Cow’s milk con-sists of, in descending order (and in generalconcentrations), water (87%), lactose (5%), fat(3.5% to 4%), protein (3.2% to 3.4%), and min-erals (�1%), mainly calcium. In contrast, a typi-cal cheese, such as Cheddar cheese, contains36% to 39% water, 30% to 32% fat, 26% to 28%protein, 2% to 2.5% salt, 1% mineral (mostlycalcium), and �1% lactose.

The differences should be evident—cheesecontains less water and more milk solids, in the

Percent of total U

.S. cheese production

Kg

prod

uced

in th

e U

.S. (

x106

)

1960 1970 1980 1990 2000 2005

1800 100

80

60

40

20

0

1500

1200

900

600

300

0

Figure 5–3. Production of American (�) and Italian (�) type cheeses in Kg and as a percent of total cheeseproduction (open symbols) from 1960 through 2003. Adapted from USDA Economic Research Service statistics.



form of fat and protein, than the milk fromwhich it was made. Thus, cheese making cansimply be viewed as a concentration process, inwhich the water portion, or whey, is removedand the solids are concentrated. In fact, as weshall see later, most of the steps involved incheese making are performed for the singularpurpose of removing water. Doing so not onlyconcentrates and solidifies the solid matter, butalso decreases microbial and enzymatic activi-ties such that cheese is much better preservedthan milk.As noted in the previous chapter,but-termilk, yogurt, sour cream, and other cultureddairy products are different from cheese in sev-eral respects,but chief among these differences

is that these products do not involve a water re-moval step.

Converting a liquid into a solid

Water removal and the concentration of proteinand fat occur via a combination of biochemical,biological, and physical-chemical events. Manyof these events happen at nearly the same timeand often have complex effects on one another.For example, exposure of cheese curds to bothhigh temperature and low pH enhances re-moval of water from the curd (a phenomenonknown as syneresis). But if the temperature istoo high, the microorganisms that produce the

Milk (100 Kg)

Total Protein (3.3 Kg)

Casein (2.5 Kg)

Whey proteins (0.8 Kg)

Lactose (5 Kg)

Fat (4 Kg)

Minerals (0.7 Kg)

Water (87 Kg)



Casein (2.4 Kg)

Lactose (0.2 Kg)

Fat (3.6 Kg)

Minerals (0.6 Kg)

Water (3.9 Kg)

Cheese (10.8 Kg) Whey (89.2 Kg)


Water (83.1 Kg)

Fat (0.4 Kg)

Minerals (0.1 Kg)

Lactose (4.8 Kg)

Casein (0.1 Kg)


Figure 5–4. Partition of milk into cheese and whey.


Cheese 149

acid that lowers the pH will be inactivated, re-sulting in poor syneresis (and poor qualitycheese). If, on the other hand, too much acid isproduced, then significant mineral loss (specifi-cally calcium) will occur, making the cheesecrumbly (which may or may not be a goodthing,depending on the cheese).

To further complicate the situation, remov-ing water from the curd also removes thesolutes dissolved in the curd.Thus, the amountof lactose in the curd is reduced as the curdsbecome dryer, and less substrate will be avail-able later for the culture.The point is that eachprocess step has consequences (hopefully, in-tended) not only on other cheese making ac-tivities, but also on the overall properties andquality of the finished cheese.

The conversion of liquid milk into a solidmass of cheese is done via coagulation (or pre-cipitation) of milk protein. Milk, as notedabove,contains about 3.3% protein.Of the pro-tein fraction, about 80% is casein (2.5% of themilk), and the remaining 20% are known col-lectively as whey proteins. For most cheeses,the protein portion consists almost entirely ofcasein. When milk coagulates and the coagu-lated material is separated, the soluble wheyproteins are released into the water or wheyfraction. The casein matrix not only containssome water (and whatever solutes are dis-solved or suspended in the water phase), butalso a large portion of the lipid fraction thatwas originally present in the milk, dependingon how coagulation occurs.

There are three ways the initial coagulationstep is accomplished. First, milk can be coagu-lated by acids produced by lactic acid bacteria,based on the same principle used for yogurt andother cultured dairy products (Chapter 4).When the milk pH reaches 4.6, casein is at itsisoelectric point and its minimum solubility,andtherefore it precipitates. However, unlike theprocess used for cultured products, the coagu-lum or gel is not left intact,but rather is then cutinto die-sized curds. After separation from thewhey and cooking and washing steps are com-pleted, the acid-precipitated curds are com-prised almost entirely of casein and water. In

fact, pure casein and caseinate salts (made byneutralizing acid casein), both of which are ofconsiderable commercial importance, are madevia the acid precipitation method. In cheesemaking, of course, the curds are then furtherprocessed,resulting in products such as Cottagecheese and farmers’cheese.It is important to re-alize that casein coagulates at pH 4.6 whetheracidification occurs via fermentation-generatedacids or simply by addition of food-grade acidsdirect into the milk. In fact, the latter process ispreferred by some producers of Cottage,cream,and other acid-coagulated cheeses, due in partto the ease of manufacture and the eliminationof starter cultures as an ingredient.However,be-cause theses products are not fermented (and,therefore,they must be labeled as acid-set), theirmanufacture will not be considered further inthis text.

The second and most common way to ef-fect coagulation is by the addition of the en-zyme chymosin (or rennet). This enzyme hy-drolyzes a specific peptide bond locatedbetween residues 105 (a methionine) and 106(a phenylalanine) in � casein. The hydrolysis of this bond is sufficient to cause a part of � ca-sein (the glycomacropeptide fraction) to dis-sociate from the casein micelle, exposing theanionic phosphates of �-casein. Thus, the remaining casein micelle becomes sensitive tocalcium-mediated precipitation (Box 5–1). Incontrast to acid-precipitated casein, the coagu-lated casein network formed by chymosintreatment traps nearly all of the milkfat withinthe curd. Most of the cheeses manufacturedaround the world rely on chymosin coagula-tion. It is worth emphasizing that even thoughchymosin,alone, is sufficient to coagulate milk,lactic starter cultures are also absolutely essen-tial for successful manufacture of most hardcheeses.The lactic acid bacteria that comprisecheese cultures not only produce acid and re-duce the pH, they also contribute to the rele-vant flavor, texture, and rheological propertiesof cheese, as described later.

Until relatively recently, most chymosinwas obtained from its natural source, thestomachs of suckling calves after slaughter,



Box 5–1. Casein Chemistry and Chymosin Coagulation

Casein, the main protein in milk,consists of several subunits, including �s1,�s2,�,and �.They areassembled as complexes called micelles that contain thousands of casein subunits, along withcalcium phosphate, that are held together primarily via hydrophobic and electrostatic interac-tions. Although some researchers have suggested that casein is organized in the form of sub-micelles,how casein micelles actually exist in milk is uncertain (Lucey,2002;Lucey et al.,2003).Moreover, it also interesting to understand how these micelles form into a gel when exposed tothe enzyme chymosin. Over the past forty years, several models have been proposed that de-scribe the casein micelle structure and how gel formation occurs.

One simplified model suggests that the hydrophobic �and � subunits are located within theinterior core of the micelle, and are surrounded by � casein, as shown below.Accordingly, a partof the kappa casein molecule (called the glycomacropeptide) protrudes from the micelle, pro-viding steric hindrance between neighboring micelles (Figure 1). Since this section of the � ca-sein is negatively charged, electrostatic repulsion between casein micelles also prevents mi-celles from coming into contact with other micelles.However,once this barrier is removed (viathe action of the enzyme chymosin; see below),glycomacropeptide is released and the micellescan interact with one other.There will still be a small negative charge on the micelles;however,soluble cationic calcium serves to neutralize the charge.This allows the micelles to come intocontact with one another, resulting in formation of a gel. Importantly, this gel will entrap fat,forming the basis of cheese manufacture (Figure 2).

glycomacropeptide

submicelle

calcium phosphate

Figure 1. Structural model of the casein micelle, showing individual submicelles. Adapted from Walstra,1999.


Cheese 151

In effect, � casein can be viewed as a stabilizing protein or physical barrier that protects thecasein micelle from a spontaneous calcium-induced coagulation. Similarly, according to anotherrecently proposed model (the Horne Model), casein molecules are “polymerized” via hydropho-bic cross-linking and calcium phosphate bridges (Horne,1998). Since � casein does not bind cal-cium and cannot cross-link, it acts a chain-terminator (thereby limiting the size of the micelles).

Chymosin is an aspartic protease (meaning that it contains aspartic acid residues within theactive site of the enzyme) synthesized by mucosal cells within the fore-stomach of youngcalves. Its physiological function in the suckling calf is presumably to transform casein from asoluble or dispersed form in the stomach to a solid state,where it can then be attacked more ef-ficiently by other gastric enzymes.The peptides and amino acids can then be adsorbed acrossthe intestinal wall, satisfying the nutritional requirements of the growing calf.

Chymosin performs a similar biochemical function in cheese manufacturing, albeit underquite different circumstances.When chymosin is added to milk, hydrolysis of � casein occursprecisely at the peptide bond located between residues 105 (phenylalanine) and 106 (methio-nine). As noted above, the product of this hydrolysis reaction, glycomacropeptide, separatesfrom the casein micelles. Because milk contains more than 1,000 mg/L calcium, about 10% ofwhich is in the ionic form, calcium-mediated coagulation readily occurs, and the first step incheese making has been accomplished.

It is also important to note that the casein subunits (and �casein, in particular) can also be at-tacked by chymosin and other proteinases present in the chymosin mixture by milk-derived pro-teinases (particularly plasmin) and by proteinases produced by microorganisms (including thestarter culture).The hydrolysis of casein by these proteinases may have serious consequences,both good and bad, during the manufacture of cheese, as will be discussed later in this chapter.

References

Horne, D.S. 1998. Casein interactions: casting light on the black boxes, the structure in dairy products. Int.Dairy J. 8:171–177.

Lucey, J.A. 2002. Formation and physical properties of milk protein gels. J. Dairy Sci. 85:281–294.Lucey, J.A., M.E. Johnson, and D.S. Horne. 2003. Perspectives on the basis of the rheology and texture of

properties of cheese. J. Dairy Sci. 86:2725–2743.Walstra, P. 1999. Casein micelles: do they exist? Int. Dairy J. 9:189–192.

Box 5–1. Casein Chemistry and Chymosin Coagulation (Continued)

caseinmicelle

fatwheyprotein

Figure 2. Chymosin-coagulated cheese matrix model. Caseinis envisioned as surrounding fat globules; small amounts ofwhey protein may also be present. Adapted from Hinrichs,2004.



via a salt extraction process. The relativelyhigh cost of this enzyme along with increas-ing demand and sporadic supply problemshave long driven cheese ingredient suppliersto develop less expensive, alternative enzymeproducts that could perform the same func-tion as chymosin. A genetically engineeredform of chymosin was approved by the FDAin 1990, and several such products were later

approved that have since captured much ofthe American chymosin market (Box 5–2).

Finally, it is possible to form a precipitate by acombination of moderate acid addition (pH6.0),plus high heat (� 85°C).Whey proteins aredenatured under these conditions, thus the pre-cipitate that form consists not only of casein,but also whey proteins. Fat may also be re-tained. Even in solutions where casein is absent

Box 5–2. Making Calf Chymosin in Fermentors

Chymosin, the enzyme that causes milk to coagulate, is an essential ingredient in the cheese-making process.Until the 1990s,calf chymosin (the major source) had been the most expensiveingredient (other than milk) used in cheese manufacture, adding about $0.03 to each kg ofcheese.This was because the chymosin supply depended on veal production, which was sub-ject to considerable market variations, as well as on production and purification costs.As thecheese industry grew worldwide,but especially in the United States in the 1960s, ’70s,and ’80s,the substantial increase in demand for calf chymosin created supply problems and led to evenhigher prices.Thus, cheese makers sought other, less expensive sources of coagulant.

Among the early non-calf rennet substitutes were bovine and porcine pepsin.These enzymescoagulate milk;however, they have a number of undesirable features,most important being thatcheese quality is not as good. In the 1960s, fungal enzymes (sometimes referred to as microbialrennets) derived from Mucor miehei and Mucor pusilus were isolated and commercialized.Al-though considerably less expensive than calf chymosin, these enzymes were also far from per-fect.They had much more non-specific casein hydrolysis activity, resulting in yield loss and fla-vor and texture defects. Some were heat stable and residual activity could be detected in thewhey, limiting the application of whey as an ingredient in other products. Still, price and otherconsiderations (e.g., kosher and vegetarian status) led these products, and their improved, sec-ond-generation versions, to gain a substantial portion of the coagulant market.

The search for enzymes with properties more like calf chymosin shifted to an entirely new di-rection in the mid- to late 1980s when recombinant DNA technology was developed. Duringthis time, several new biotechnology companies began projects to identify the gene coding forchymosin in calf abomasum mucosal cells and expressing the gene in suitable host cells(Marston et al., 1984; van den Berg et al., 1990; Dunn-Coleman et al., 1991). It was immediatelyrecognized that the actual genetic material of interest was not going to be the chymosin gene inthe form of DNA, but rather the mRNA transcribed from the DNA.

This is because it is the mRNA, not the DNA, that contains the actual coding regions in eu-karyotic cells. Most eukaryotic genes contain coding regions (exons) along with introns, intra-genic regions of non-coding DNA.The latter are excised from the mRNA following transcrip-tion, leaving behind the edited mRNA that is ultimately transcribed. In the case of the chymosingene, there are eight introns and nine exons.

The problem of identifying the chymosin gene was even more complicated because the genealso contains a sequence that encodes for the pro-enzyme form of the chymosin protein.Prochymosin contains a 42 amino acid N terminal region whose function is to maintain the en-zyme in a stable but inactive form thereby preventing the enzyme from hydrolyzing proteinswithin the producer cell itself.The chymosin genes also encode for a leader sequence that di-rects secretion of the protein across the membrane and out of the producing cell.This “pre” re-gion is cleaved during the secretion step, and the “pro” portion is ultimately autohydrolyzed inthe acid environment of the stomach, leaving active chymosin.


Cheese 153

As molecular biologists began this research, therefore, they first isolated total mRNA from themucosal cells (Figure 1). Following a simple purification step, the enzyme, reverse transcriptase,was used to make complimentary DNA (cDNA) directly from the mRNA.This cDNA was then lig-ated into a plasmid vector and cloned into E.coli to make a library of E.coli clones, representing,in theory, all of the different mRNA species transcribed by the mucosal cells (including clonescontaining the preprochymosin mRNA).The library was screened,either using anti-chymosin an-tibodies or chymosin gene DNA probes, to identify clones capable of expressing the chymosinprotein. Eventually, chymosin-expressing, recombinant E. coli clones were identified.

Yet another expression problem, however, quickly became apparent.When E. coli expressescertain heterologous genes, especially genes from eurkaryotic organisms, the gene products—in

Box 5–2. Making Calf Chymosin in Fermentors (Continued)

cDNA

mRNA

Calf mucosal cell

Cloned cDNA

cDNA library

Chymosin-expressing clone

Isolate total mRNA

Reverse transcriptase

Ligate to vector

Transform into host cells

Antibody or DNA probe

Figure 1. Making recombinant chymosin (see text for details).

(Continued)



this case, chymosin—are not secreted but instead are packaged inside the cell in the form of in-clusion bodies.Thus, to retrieve the chymosin, the cells had to be collected and disrupted,eithermechanically or chemically, to release the chymosin-containing inclusion bodies.The latter, then,had to be solubilized to liberate active chymosin. Nonetheless, despite these awkward and ex-pensive steps, genetically engineered chymosin won FDA approval and quickly gained accep-tance in the marketplace (Flamm, 1991). Subsequently, expression systems for eukaryotic organ-isms, including Kluyveromyces lactis and Aspergillus nidulans, were developed that led tosecreted product with high yield.Engineered chymosin is currently marketed worldwide by sev-eral major companies, including DSM Food Specialties (formerly Gist-Brocades), Chrs. Hansens,and Danisco.Among the production organisms currently used for these so-called recombinantchymosin products are K. lactis and Aspergillus niger.

Aside from the obvious cost advantage of the genetically engineered chymosins, other bene-fits were also realized.These products are derived from a fermentor, rather than calf stomachs,and are not subject to supply problems or price fluctuations.They can be classified as vegetar-ian and can more easily obtain kosher status. Importantly, the chymosin is relatively easy to pu-rify and can be produced free of other proteases that are sometimes present in calf chymosinpreparations. In fact, manufacturers claim that the engineered chymosin is 100% pure and hashigh specific activity.Thus,according to some experts,cheese made using these chymosin prod-ucts has organoleptic properties as good as (if not better than) cheese made with calf chy-mosin. (There was also a contrary argument—that pure chymosin makes a less flavorful agedcheese, since non-specific proteolysis is absent.)

Currently, engineered chymosin sells for about half the cost of what calf chymosin sold for inthe 1980s.Thus, it is not surprising that engineered chymosin has, for all practical purposes, en-tirely displaced calf chymosin in the United States. Calf chymosin has essentially become aniche product, with less than 10% of the coagulant market. However, in Europe, where manycheese manufacturers prefer or are required to produce cheese by traditional techniques, calfchymosin is more widely used. It is worth noting that unlike other foods produced via biotech-nology, when engineered chymosin was introduced, there was little public concern or protest.Perhaps this was because the notion of a product produced by microorganisms in fermentorswas more appealing to consumers than the image of calf stomach extracts. In addition, the FDAdetermined that no special labels would be required for cheese made with engineered chy-mosin, which they considered to be “not significantly different” from calf chymosin.

References

van den Berg J.A., K.J. van der Laken, A.J. van Ooyen,T.C. Renniers, K. Rietveld, A. Schaap, A.J. Brake, R.J.Bishop, K. Schultz, D. Moyer, M. Richman, and J.R. Shuster. 1990. Kluyveromyces as a host for heterolo-gous gene expression: expression and secretion of prochymosin. Bio/Technology 8:135–139.

Dunn-Coleman, N.S., P. Bloebaum, R.M. Berka, E. Bodie, N. Robinson, G.Armstrong, M.Ward, M. Przetak, G.L.Carter, R. LaCost, L.J.Wilson, K.H. Kodama, E.F. Baliu, B. Bower, M. Lamsa, and H. Heinsohn. 1991. Com-mercial levels of chymosin production by Aspergillus. Bio/Technology 9:976–981.

Flamm, E.L. 1991. How FDA approved chymosin: a case history. Bio/Technology 9:349–351.Marston, F.A.O., P.A. Lowe, M.T. Doel, J.M. Schoemaker, S. White, and S. Angal. 1984. Purification of calf

prochymosin (prorennin) synthesized in Escherichia coli. Bio/Technology 2:800–804.

Box 5–2. Making Calf Chymosin in Fermentors (Continued)

(i.e., whey), this process can result in enoughprecipitated whey protein to form a cheese. Ex-amples of precipitated cheeses include Ricottacheese, the Hispanic-style cheeses queso frescoand queso blanco, and Gjetost, a whey-derivedcheese popular in Norway.

Squeezing out water

Once milk is transformed from a liquid into asolid (actually a gel), the next goal is to removewater.The first step involves increasing the sur-face area of the single large gel mass by cutting


Cheese 155

it into literally millions of smaller cubes of curd(e.g., 1 m3 of a cheese gel cut into 1 cm3 parti-cles yields 106 such curd particles). Becausethe distance the water molecules must travel(from the interior of the gel to the outside en-vironment) decreases as the curd size de-creases, this step has the effect of substantiallyincreasing the rate of syneresis. Then, whenthese curds are stirred, and then heated, thecurds shrink and syneresis is further enhanced.

Acidification is another factor involved inconcentrating milk solids. As discussed pre-viously in Chapter 4, as the pH approaches the isoelectric point of casein (i.e., 4.6), water-holding capacity decreases and syneresis, orwhey expulsion, increases. In cheese manufac-ture, as for cultured milk products, acidificationis due to fermentation of lactose by the starterculture. As we will see later, the culture per-forms other critical functions; in particular, it isalso responsible for desirable flavor and textureproperties.But it is the ability to convert lactoseto lactic acid and to reduce the cheese pH thatmakes the culture essential. Finally, it is possibleto use physical force (i.e.,pressure or gravity) tosqueeze out even more water.

General Steps in Cheese Making

On a worldwide basis, there are probably thou-sands of different types of cheeses producedand consumed. As Charles De Gaulle, the for-mer French president, famously lamented,there are hundreds of different cheeses madein France alone1. Anyone who has visited a fro-mageri in Paris or a formaggio in Milan (or per-haps the National Cheese Emporium in Eng-land; Box 5–3), can certainly appreciate theincredible variety of cheeses that are available.How could there be so many? Are the proce-dures for making cheese so complex as to al-low for all the cheeses produced?

In reality, the basic manufacturing steps forall cheeses are surprisingly similar. However, itis the almost unlimited number of variablesthat exist at each of these steps that ultimatelyaccount for the myriad number of differentcheeses. Described below are the various in-gredients and manufacturing steps that are

used in cheese making, with an emphasis onthose variables that distinguish one type ofcheese from another.

Milk

As shown in Figure 5–5, the first variable startswith the milk itself.Although most cheeses aremade from cow’s milk, many cheeses are madeusing milk from other sources. For example,Feta and Chevre are ordinarily made from goatmilk, Roquefort and Romano are made fromsheep milk, and Mozzarella is often made fromthe milk of water buffaloes. These milks havedramatically different gross compositions frombovine milk; in particular, they all contain morefat (water buffalo and goat milk contain nearlytwice as much). However, even the composi-tion of cow’s milk varies according to thebreed of cow, the nature of the feed consumedby the cow, and even when the milk was ob-tained (i.e., morning versus evening).

Moreover, in cheese making, not only doesthe gross composition affect cheese proper-ties, but so does the specific composition ofeach of the milk constituents.The lipid portionof goat milk, for example, contains a higherpercentage of volatile, short-chain fatty acids,such that rancid flavor notes are most evidentwhen the triglycerides are hydrolyzed by li-pases.The fat content also has a profound in-fluence of other properties of cheese. Fat notonly contributes to the body and texture ofcheese, but it also serves as substrate for im-portant flavor-generating reactions performedby microorganisms. Also, many of the flavorconstituents derived from non-lipid substratesthat form during cheese ripening are soluble inthe lipid phase.For example,hydrophobic pep-tides derived from casein hydrolysis (many ofwhich are bitter) are found in the fat portion ofthe cheese.

For some cheeses, the milk is standardizedto give a fat content that is particular to a givencheese. In general, the minimum fat content formost cheeses is usually around 50% (on a drybasis). For example, Cheddar type cheeses aremade with whole milk, containing 3.5 to 4.5%milkfat. However, other cheeses are made with



Box 5–3. The Monty Python Cheese Shop Sketch

More than forty different types of cheese are mentioned,making it, for cheese science students,at least, the most educational television sketch ever presented.

CUSTOMER: Good Morning.OWNER: Good morning, sir.Welcome to the National Cheese Emporium!CUSTOMER:Ah, thank you, my good man.OWNER:What can I do for you, sir?CUSTOMER: I want to buy some cheese.OWNER: Certainly, sir.What would you like?CUSTOMER:Well, eh, how about a little red Leicester, sir.OWNER: I’m afraid we’re fresh out of red Leicester, sir.CUSTOMER: Oh, never mind, how are you on Tilsit?OWNER: I’m afraid we never have that at the end of the week, sir, we get it fresh on Monday.CUSTOMER:Tish tish. No matter.Well, stout yeoman, four ounces of Caerphilly, if you please.OWNER:Ah! It’s been on order, sir, for two weeks.Was expecting it this morning.CUSTOMER:‘Tis not my lucky day, is it? Aah, Bel Paese?OWNER: Sorry, sir.CUSTOMER: Red Windsor?OWNER: Normally, sir, yes.Today the van broke down.CUSTOMER:Ah, Stilton?OWNER: Sorry.CUSTOMER: Ementhal? Gruyere?OWNER: No.CUSTOMER:Any Norwegian Jarlsburg, per chance?OWNER: No.CUSTOMER: Lipta?OWNER: No.CUSTOMER: Lancashire?OWNER: No.CUSTOMER:White Stilton?OWNER: No.CUSTOMER: Danish Brew?OWNER: No.CUSTOMER: Double Gloucester?OWNER: No.CUSTOMER: Cheshire?OWNER: No.CUSTOMER: Dorset Bluveny?OWNER: No.CUSTOMER: Brie, Roquefort, Pol le Veq, Port Salut, Savoy Aire, Saint Paulin, Carrier de lest, BresBleu, Bruson?OWNER: No.CUSTOMER: Camembert, perhaps?OWNER:Ah! We have Camembert, yes sir.CUSTOMER: (surprised) You do! Excellent.OWNER:Yessir. It’s . . .ah, . . .it’s a bit runny . . .CUSTOMER: Oh, I like it runny.OWNER:Well, . . . It’s very runny, actually, sir.CUSTOMER: No matter. Fetch hither the fromage de la Belle France! Mmmwah!OWNER: I . . . think it’s a bit runnier than you’ll like it, sir.


Cheese 157

CUSTOMER: I don’t care how blinking runny it is. Hand it over with all speed.OWNER: Oooooooooohhh...!CUSTOMER:What now?OWNER:The cat’s eaten it.CUSTOMER: Has he.OWNER: She, sir.CUSTOMER: Gouda?OWNER: No.CUSTOMER: Edam? OWNER: No.CUSTOMER: Case Ness?OWNER: No.CUSTOMER: Smoked Austrian?OWNER: No.CUSTOMER: Japanese Sage Darby?OWNER: No, sir.CUSTOMER:You ... do have some cheese, don’t you?OWNER: Of course, sir. It’s a cheese shop, sir.We’ve got—CUSTOMER: No no ... don’t tell me. I’m keen to guess.OWNER: Fair enough.CUSTOMER: Uuuuuh,Wensleydale.OWNER:Yes?CUSTOMER:Ah, well. I’ll have some of that!OWNER: Oh! I thought you were talking to me, sir. Mister Wensleydale, that’s my name.CUSTOMER: Greek Feta?OWNER: Uh, not as such.CUSTOMER: Uh, Gorgonzola?OWNER: No.CUSTOMER: Parmesan? OWNER: No.CUSTOMER: Mozzarella?OWNER: No.CUSTOMER: Paper Cramer?OWNER: No.CUSTOMER: Danish Bimbo?OWNER: No.CUSTOMER: Czech sheep’s milk?OWNER: No.CUSTOMER:Venezuelan Beaver Cheese?OWNER: Not today, sir, no.CUSTOMER:Aah, how about a Cheddar?OWNER:Well, we don’t get much call for it around here, sir.CUSTOMER: Not much ca—It’s the single most popular cheese in the world!OWNER: No ’round here, sir.CUSTOMER: And what IS the most popular cheese ’round hyah?OWNER: ’Illchester, sir.CUSTOMER: Is it.OWNER: Oh, yes, it’s staggeringly popular in this manusquire.CUSTOMER: Is it.

Box 5–3. The Monty Python Cheese Shop Sketch (Continued)

(Continued)



milk adjusted to 3% fat (e.g., Swiss cheese) orless (part-skim Mozzarella). In part, this is be-cause the body characteristics of certain hardcheeses, for example, Swiss and Parmesan, re-quire less fat and a higher casein-to-fat ratio. Incontrast, some cheeses are made from milkthat is enriched with cream, such that the fatcontent of the cheese (so-called double creamcheese) will be 60% (or even 72% for triplecream cheese).

Another key variable involves the handlingof the milk, and in particular, whether the milkhas been heated or not. In the United States, allunripened cheese (aged less than sixty days)must be made from pasteurized milk, whereasonly aged cheese (held for more than sixty days

at a temperature not less than 1.7°C) can bemade from raw milk. In many other parts of theworld, including France, Italy, and other majorcheese-producing countries, there are no suchpasteurization requirements, and even freshcheeses can be made from raw milk (unlessthey are to be exported to the United States—then they must conform to U.S. requirements).

In reality, however, even cheese milk that isnot pasteurized is often heat-treated to sub-pasteurization conditions (even for agedcheese). The reason for pasteurizing milk, re-gardless of how that milk is to be used, is to killpathogenic and spoilage microorganisms.Given the concern about food safety in theUnited States, there has been a trend among

OWNER: It’s our number one best seller, sir!CUSTOMER: I see. Uuh ... ’Illchester, eh?OWNER: Right, sir.CUSTOMER:All right. Okay.‘Have you got any?’ he asked, expecting the answer ‘no.’OWNER: I’ll have to look, sir ... nnnnnnnnnnnnnnnno.CUSTOMER: It’s not much of a cheese shop, is it?OWNER: Finest in the district!CUSTOMER: (annoyed) Explain the logic underlying that conclusion, please.OWNER:Well, it’s so clean, sir!CUSTOMER: It’s certainly uncontaminated by cheese ...OWNER:You haven’t asked me about Limburger, sir.CUSTOMER:Would it be worth it?OWNER: Could be ...CUSTOMER: Have you got any Limburger?OWNER: No.CUSTOMER: Figures. Predictable, really I suppose. It was an act of purest optimism to haveposed the question in the first place.Tell me:OWNER:Yes sir?CUSTOMER: Have you in fact any cheese here at all.OWNER:Yes, sir.CUSTOMER: Really?OWNER: No. Not really, sir.CUSTOMER:You haven’t.OWNER: No sir. Not a scrap. I was deliberately wasting your time, sir.CUSTOMER:Well, I’m sorry, but I’m going to have to shoot you.OWNER: Right-O, sir.

The Customer takes out a gun and shoots the owner.

CUSTOMER:What a senseless waste of human life.

Box 5–3. The Monty Python Cheese Shop Sketch (Continued)


Cheese 159

large U.S. manufacturers to use pasteurizedmilk for most cheeses,even those that are aged.

Pasteurization, however, not only killspathogens and undesirable spoilage bacteria,but it also inactivates much of the endogenousmicroflora and enzymes ordinarily present inraw milk. Since the microflora and enzymesboth contribute to the overall flavor and tex-ture properties of the finished cheese, espe-cially if the cheese is aged, quality differencesbetween cheese made from raw or heat-treated milk can be significant. Thus, there isnow a debate between those who believe that

pasteurization should be required and thosewho contend that pasteurization is detrimentalto cheese quality (Box 5–4). It should be em-phasized, however, that whether the milk israw or pasteurized, it should still be of high mi-crobiological quality, free of antibiotics, andwithin the standards specified by governmentregulations.

Finally, one of the most obvious ways totreat milk to distinguish one cheese from an-other is by the color. Milk, and the finishedcheese, can be made more yellow by addingthe natural coloring agent, annato, or made

chymosin

lactic acid bacteria

raw or heat-treatedcow, sheep, goat

whole, reduced fat color

mesophilic

thermophilic

heterofermentative

homofermentative

adjunct organismsPropionibacterium

Penicillium

surface organisms

calf

engineered

add water

cut and cook

dilutes lactose

small or large curds

high or low temperature

curd handling

salt

brine

dry

rub

aging

hoop (shape)

milk

cheese

lipasesenzymes

Brevibacteriumyeasts, micrococci

Figure 5–5. General steps for manufacture of hard cheeses.



Box 5–4. Safety of Cheese Made From Raw Milk

In the United States, only cheese that is aged for more than sixty days can be made from rawmilk; all other cheeses must be made from milk that has been pasteurized.The rationale for thispolicy, which was established in the 1930s, is based on early research that indicated any patho-genic bacteria present in the raw milk would die during a sixty-day ripening period. Factors believed to contribute to cell death included low pH, low water activity, low Eh, high salt-in-moisture levels, the presence of organic acids and competing microflora,and the absence of fer-mentable sugars. In other words, the collective effect of these inhospitable conditions werethought to create an environment that would result in the eventual lysis and death of any path-ogenic bacteria that may have been present in the original cheese.

The sixty-day requirement was really nothing more than an approximation based on a ratherlimited number of studies. Moreover, these studies were done long before Listeria monocyto-genes,Escherichia coli O157:H7, and other pathogenic organisms were recognized as potentialcontaminants of raw milk.That these pathogens could be inherently more resistant to environ-mental extremes or that they could induce stress resistance defense systems were possibilitiesthat could hardly have been imagined.

It is now well-known that some of these pathogens can indeed tolerate all sorts of inhibitoryconditions.For example,L.monocytogenes is particularly capable of surviving,even growing,atacid pH and in the presence of high salt concentrations.Furthermore,when exposed to variousshocks, such as high,sub-lethal temperatures, this organism synthesizes a set of proteins that en-able it to tolerate other subsequent stresses.These observations led researchers to challenge theconclusions made seventy years ago, i.e., that a sixty-day aging process would result in raw milkcheese that is free of pathogens. Recent research, in fact, now indicates that these “newly” rec-ognized pathogens could survive aging and, if present at sufficiently high numbers in the rawmilk,could be present in the aged cheese.For example, several publications from the Marth lab-oratory at the University of Wisconsin showed that L. monocytogenes could survive more thanone year in Cheddar cheese and that levels greater than 107 cfu/g could be reached in Camem-bert cheese (summarized by Ryser, 1999).

In response to these findings, the FDA announced in 2002 that this policy would be reviewedand left open the possibility that all cheese, regardless of whether it was aged or not, wouldhave to made from pasteurized milk.Although manufacturers of aged cheese were undoubtedlyconcerned about this announcement, the greatest furor of opposition came from gourmet-typeconsumer groups.Their position, as articulated on various Web sites, is that cheese made fromraw milk is superior in flavor, texture,and overall sensory appeal to that of pasteurized milk,andthat any perceived food safety risks are exaggerated or non-existent. Indeed, despite the re-search showing that pathogens might, at least in theory, survive sixty days of aging, the foodsafety record for aged cheese is extraordinarily good.There are few published accounts of agedraw milk cheese having been responsible for foodborne disease, and in those cases where thecheese was implicated, other extenuating circumstances were involved (e.g., post-aging conta-mination by workers or equipment).

Despite the excellent record for aged raw milk cheese, fresh raw or pasteurized milk cheesehas been the cause of several serious foodborne disease outbreaks (Table 1). In the majority ofthese outbreaks, poor handling of the milk or cheese or poor sanitation was responsible. Post-pasteurization contamination of milk, in particular, is the most frequent cause of these out-breaks.However, there are some fresh or non-aged cheeses that,even when properly made,maystill pose a risk to certain consumers.Those at risk include the very young and very old, preg-nant women, and other immuno-compromised populations. Brie cheese, for example, is on the“do not eat” list for pregnant women and HIV-positive individuals, due to the potential (but in-frequent) occurrence of L. monocytogenes in this cheese.The use of sub-pasteurization treat-ments, such as “thermization” (performed by heating milk to about 65°C for fifteen to twenty


Cheese 161

more white by adding bleaching-type agents.Although the color has no effect on flavor ortexture, manufacturers have learned that visualappeal can have a pronounced effect on ac-ceptability and preference. In the UnitedStates, for example, consumers from the Mid-west, in general, prefer orange-colored Ched-dar cheese, whereas Northeasterners favorwhite Cheddar.

Starter cultures

The composition of the starter culture de-pends on the intentions of the cheese maker. Ifthe cheese-making procedure includes a stepwhere the curds will be exposed to high tem-peratures, such as during manufacture ofSwiss, Parmesan, and Mozzarella cheese, then athermophilic lactic acid bacterial culture ableto withstand those temperature must be used.If a particular flavor compound is desired, suchas diacetyl in Gouda cheese, then again, theculture must contain specific organisms capa-ble of producing those flavor compounds. Infact, culture technology has advanced now tothe point where each strain present in the cul-

ture can be selected on the basis of the spe-cific performance attributes desired by thecustomer. Thus, the culture can contribute toconsiderable variation in the finished cheese,and even modest changes in the culture com-position or amount can result in dramatic dif-ferences in the cheese.

As described in Chapters 2 and 3, culturesused for cheese fermentations consists of sev-eral genera and species of lactic acid bacteria.Aside from selecting a culture that gives the de-sired rate and extent of acid development andthat produces the desired flavor and texture,themain distinction for culture selection is basedon temperature.Mesophilic cultures,containingstrains of Lactococcus lactis subsp. lactis andLactococcus lactis subsp.cremoris,grow withina wide temperature range of about 10°C to40°C. Growth rates at the extreme ends of thisrange are, however, quite low and temperaturesmuch above the upper limit for growth arelethal. Moreover, the optimum temperature forgrowth of mesophilic cultures is about 28°C to32°C (depending on the strain), which is gener-ally very near the temperature range that themilk or cheese will be held during manufacture

seconds), may be somewhat effective against potential pathogens and provides a margin ofsafety at which some cheese makers may be comfortable.

Table 1. Safety record of cheese.

Year Cheese Cases Organism Cause

1970s Brie Enteroinvasive Poor sanitation347 Escherichi coli

1980s Brie Enterotoxigenic Unknown169 Escherichi coli

1980s Mont d’Or 122 Listeria monocytogenes Environment1984 Cheddar 2,000 Salmonella typhimurium Unknown1985 Mexican-style 300 Listeria monocytogenes Raw milk contamination; poor sanitation1989 Mozzarella 164 Salmonella javiana Poor sanitation1995 Brie 20 Listeria monocytogenes Raw milk contamination

Reference

Ryser, 1999. Incidence and behavior of Listeria in fermented dairy products, In E.H. Marth and E.T. Ryser(ed.) Listeria, Listeriosis, and Food Safety, 2nd ed. Marcel Dekker, New York.

Box 5–4. Safety of Cheese Made From Raw Milk (Continued)



(i.e., when fermentation is expected to occur).The mesophilic cultures are the workhorses ofthe cheese industry, and are used for the major-ity of cheeses.

Thermophilic cultures, mainly Lactobacillushelveticus and Streptococcus thermophilus, arealso widely used.They have a temperature op-tima around 42°C to 45°C and are able to growat temperatures as high as 52°C. Like themesophiles, however, temperatures just a fewdegrees higher than their maximum growthtemperature may result in thermal inactivation.This means that when curds are heated, thetemperature must not exceed that which theculture can tolerate. Otherwise, the culture willbe inactivated or injured and the subsequentfermentation will either be slow or not occur.However, it should be noted that thermal effectson microorganisms obey first-order kinetics,such that inactivation or killing occurs at a loga-rithmic rate.Thus, even when the temperaturereaches a lethal level, some cells will survive(depending on the exposure time) and be ableto grow (albeit slowly), once a compatible tem-perature is established.

Most of the strains that comprise mesophilicand thermophilic cultures are homofermenta-tive. However, heterofermentative lactic acidbacteria may be included in mesophilic cul-tures that are used for particular cheeses. Spe-cific heterofermentative species include Leu-conostoc mesenteroides subsp. cremoris andLeuconostoc lactis. Not coincidently, these or-ganisms also ferment citrate and produce di-acetyl.Thus, cheeses such as Gouda and Edamhave a buttery aroma, as well as a few smalleyes, due to CO2 formation. Finally, culturesmay also contain non-lactic acid bacteria forSwiss-type and surface-ripened cheeses, as wellas fungal spores for mold-ripened cheeses, aswill be discussed later.

Until relatively recently, bulk cultures grownin whey or milk without pH control were thedominant form in which cultures were used (re-viewed in Chapter 3). Culture inoculum forCheddar type cheeses was usually about 1%(w/w), which gave an initial cell concentrationof about 5 x 106 cells per ml of milk. Other

cheeses are made using more culture (e.g.,Moz-zarella is normally made with a 2% starter) orless culture (e.g., Swiss is made with 0.5% cul-ture). Although these traditional bulk culturesare still used, in the past twenty years, bulk cul-ture tanks began to be equipped with externalpH control systems. This simple technologymade it possible for operators to partially neu-tralize the medium as the cells grew, and, there-fore,maintained the pH at a level that was moreto the liking of the starter culture organisms.

Specialized culture media was also intro-duced that provided “internal”pH control.Withthe widespread use of these pH controls, bulkculture systems, culture viability, and cell den-sity (i.e.,cells per gram) have been significantlyenhanced. Thus, less culture is necessary, per-haps half as much on a weight or volume basis.In addition, highly concentrated, direct-to-vatcultures are also now widely used.These prod-ucts may contain ten or more times as manycells as traditional bulk cultures. One 500 mlcan, for example, is sufficient to inoculate10,000 L of milk (i.e., 0.005%).

Despite the consistent nature of moderncultures, their activity and cell density may stillbe somewhat variable.Thus, it is important forcheese manufactures to adjust inoculum levelsto satisfy production schedules and perfor-mance expectations. In the case of Cheddarcheese, if too much culture is added, acid de-velopment might occur too rapidly, resulting inearly loss of calcium and demineralization ofthe cheese (discussed later). In addition,a largeculture inoculum may lead to excessive pro-duction of proteolytic enzymes that can even-tually affect cheese yield, flavor, and texture. Incontrast, if not enough culture was added, fer-mentation and acid development is delayed,causing production schedule headaches aswell as opportunities for spoilage or patho-genic organisms to grow.

It is important to note that, despite the avail-ability of defined and consistent starter cul-tures, there are some cheeses made using moretraditional starter cultures. Parmigiano Reg-giano, the authentic version manufactured onlyin the Parma region of Italy, is made using whey


Cheese 163

as starter culture, a form of backslopping. Manyof the Dutch-type cheeses (e.g., Gouda andEdam) are still made using undefined strainsmaintained simply as a mixed culture.

Whether a pure, mixed, or undefined starterculture is used, the milk for most cheeses madewith a mesophilic culture is ordinarily broughtto a temperature between 30°C and 40°C(higher for cheeses made using thermophiliccultures). If early culture growth is to be en-couraged, the milk may be held for a period oftime within this temperature range. However,despite a mostly favorable temperature, littleculture activity and little acid developmentnormally occurs during this early stage of theprocess. This is because the cells either arecoming out of a somewhat dormant phase (inthe case of frozen or lyophilized direct-vat-setcultures), or are adapting from a rich, almostideal bulk culture medium to a milk mediumthat requires induction of at least some newbiochemical pathways, such as those involvingprotein hydrolysis and amino acid use. De-pending on the culture media and culturepreparation conditions, it is possible for cul-tures to be rather active, with relative short lagphases. In general, most cultures still experi-ence a lag phase (that may be quite extended,especially if a cooking step is also included),before active log phase growth occurs.

Although growth of the culture and produc-tion of fermentation end-products (other thanlactic acid) usually does not occur until later inthe cheese making process, delays due to cul-ture inhibition can cause serious problems forthe manufacturer. Large cheese factories thatprocess a million or more Kg of milk per dayrequire high throughput (i.e., the rate that milkis converted to finished cheese products). Abottleneck at the fermentation step may placeupstream operations are on hold, disruptingproduction schedules and perhaps causing em-ployees to work extra hours. Of course, if thefermentation is slow, sluggish, or fails alto-gether, cheese quality will be poor.

There are several possible causes of starterculture inhibition. Although state and federallaws require that milk be free of antibiotics, if

residues were present they could inhibit the lac-tic culture. However, it is now very rare thatmilk would contain undetected antibiotics.Milkmay also contain natural immunoglobulins thatbind to culture bacteria, forming clumps thateventually settle in the vat.This is particularly aproblem in Cottage cheese due to the long fer-mentation time. Another potential inhibitor isformed via the lactoperoxidase reaction.This re-action occurs when the enzyme lactoperoxi-dase oxidizes thiocyanate in the presence of hy-drogen peroxide to form hypothiocyanate.Thisreactive compound can be inhibitory to certainstarter culture lactic acid bacteria. Lactoperoxi-dase and thiocyanate are naturally present inmilk, and hydrogen peroxide can be producedby the endogenous microflora. It is also possibleto activate this reaction by inoculating raw milkwith hydrogen peroxide-producing lactic acidbacteria in an effort to control psychrotrophicspoilage bacteria.

Chemical agents used to sanitize cheese vatscan occasionally inhibit cultures if residues arenot adequately rinsed.By far,however, the maincause of culture inhibition are bacteriophages,viruses that infect bacteria—in this case, lacticacid bacteria (Chapter 3).Their detrimental rolein the cheese making process, and the meansby which phage problems can be controlled,will be discussed later in this chapter.

Coagulation

In many cheese factories, chymosin is added tothe milk immediately after or nearly at the sametime as the culture is added. Some cheese mak-ers allow a pre-ripening period to give the cul-ture a brief opportunity to produce a smallamount of acid and a slight lowering of themilk pH.Since chymosin is an acid protease (itsoptimum activity on �-casein occurs at pH 5.5),it will be more active as milk pH decreases.Thesolubility of calcium also increases as the pHdecreases. Thus, with pre-ripening, less chy-mosin can be used to give the same clot firm-ness. For similar reasons, it is also common toadd calcium chloride to the milk to promotecoagulation (and yield).



In any event, the amount of chymosinadded, and the length of the setting periodprior to cutting, depends on the cheese beingmade and the curd firmness desired. Usually,about 200 ml of single-strength chymosin per1,000 kilograms of milk will give a suitable co-agulation within about 30 minutes. For manylarge, automated manufacturers, the point atwhich the curd is sufficiently firm and readyfor cutting is based strictly on the clock (butalso on a priori knowledge of what times givethe best cheese). Although specialized instru-ments are available for this purpose, curd firm-ness is more often than not determined on asubjective basis, i.e., when the operator deemsit ready based on a simple cutting method.

Cutting and cooking

The coagulated mass is next cut using harp-like,wire knives that cut the curds into die-sized par-ticles.The knives are constructed such that thecurds can vary in size. Since this step is per-formed to enhance syneresis, the size, or moreimportantly, the surface area of the curd parti-cles, has a major influence on the rate of waterremoval from the curd. Hard, low-moisturecheeses like Parmesan and Swiss are typicallycut into kernel or wheat berry-sized curds,whereas soft,high-moisture cheeses are cut intodie-sized pieces.The size of the curd also influ-ences fat loss—more fat is retained in largecurds than in small ones. Regardless of the sizeof the curd at cutting, the actual compositionvaries relatively little.The curd is constructed ofa calcium-casein complex that contains en-trapped fat and bacteria (including starter culture organisms), as well as water and water-soluble components including whey proteins,enzymes,vitamins,minerals, and lactose.

Syneresis begins as soon as the curds are cut,and increases during the ensuing minutes whenthe curds are gently stirred.The initial rate de-pends on the starting pH of the curd, becauseprior acid development greatly enhancessyneresis. However, the cooking step is the pri-mary means of enhancing syneresis. All otherfactors being equal, the higher the temperature

and the longer the curds are cooked and stirred,the dryer will be the finished cheese.Thus, thecooking step is one of the major variables thatcheese manufacturers can manipulate to pro-duce different types of cheese.

As noted previously, the more water re-moved from the curd, the less lactose will beavailable for fermentation. Cheese manufactur-ers can, therefore, influence acid productionand cheese pH by modulating the cookingtime and temperature conditions. Whateverthe cooking temperature, however, the culturemust be able to withstand that temperature, orelse the cells will be attenuated (or worse yet,inactivated). Furthermore, when heat is ap-plied it must be done gradually via a step-wiseprogression, since too rapid heating causes theexterior of the curds to harden and actually re-duce syneresis. The cooking and stirring stephas so much of an effect on the finishedcheese that some experienced cheese makerscan tell how dry the cheese will be simply bytouch and the feel of the curd.Of course, somesoft cheeses, like Brie, are not cooked at all, butrather are stirred at the setting temperature. Fi-nally, although the actual heating step usuallyoccurs via indirect heat transfer through jack-eted vats, it is also possible to inject steam di-rectly into the whey-curd mixture.

Curd handling

Perhaps the most influential step during thecheese making process involves the means bywhich the curd is handled during and after thecooking and stirring steps. For many cheeses,the whey is removed when the desired acidityis reached,when the curd has been cooked for asufficient length of time, or when it is suffi-ciently firm or dry.There are several means bywhich the curd is separated and the whey is re-moved. In traditional Cheddar cheese manufac-ture, the curds are simply pushed to the sides ofthe cheese vat and the whey is drained downfrom the center, with screens in place at thedrain end to prevent curd loss.Alternatively, thecurds can be collected in cheese cloth andhoisted above the whey, as in traditional Swiss


Cheese 165

cheese manufacture. In more modern, large pro-duction factories, where cheese vats must becleaned and re-filled, curds and whey are typi-cally pumped to draining tables or Cheddaringmachines, where whey separation occurs. Simi-larly, the curd-whey mixture can be added di-rectly into perforated cheese hoops, where thewhey drainage step is completed.

Although there are many ways to manipu-late the curd to alter the properties of thecheese, one simple twist in the separation stepis worth special mention. If, during the drain-ing step,water is added back to the curds, thenso-called “sweet” or low-acid cheeses will re-sult. This is due to lactose dilution. Since thelactose concentrations in the curd and wheyare ordinarily in equilibrium, when whey is re-moved and replaced by water, lactose will dif-fuse from curd to water, leaving the curd withmarkedly less lactose available for subsequentfermentation by the starter culture. Thus, theinitial pH of Colby, Gouda, Havarti, and Edamcheeses are generally in the range of 5.2 to 5.4.If the added water is warm (i.e., about 35°C ),cooking and syneresis will continue (as is thecase for Gouda,Edam,and Havarti).However, ifthe water is cold (about 15°C), the moisturecontent of the cheese may increase (e.g.,Colby).This washing step is also used for Moz-zarella, in part for pH control, but more so toremove lactose and galactose from the curd(discussed later).

Once the curds are separated from thewhey, several things begin to happen. First, thestarter culture finds itself at a temperature con-ducive for growth, and soon the fermentationof lactose to lactic acid occurs. A subsequentdecrease in curd pH and an increase in thetitratable acidity (expressed as percent lacticacid) of the expressed whey is evident. Thecurds, almost immediately after whey is re-moved, mat or stick together. The mattedcurds, helped by piling slabs on top of one an-other, begin to stretch out and become plastic-like, a process known as Cheddaring. Alterna-tively, the dry curds can be stirred to facilitatewhey removal and to lower the moisture in thefinished cheese.

Salting

Salt is an essential ingredient that provides fla-vor, enhances syneresis, and contributes tothe preservation of most cheeses. Even thesimple step of salting, however, represents animportant variable during cheese manufac-ture. Salt can be applied directly (i.e., in dryform) to the milled curds, as in the case ofCheddar, or salt can be rubbed onto the sur-face of hooped cheese, as in the case of someblue cheese varieties (e.g., Gorgonzola andRoquefort). Alternatively, some cheeses, suchas Swiss, Mozzarella, and Parmesan, can beplaced in brines. Obviously, when salting oc-curs via brining methods, the amount of saltthat ends up in the cheese is a function of thediffusion rate into the cheese, as well as thegeometry of the cheese block, the duration ofbrining, and brine strength. If the cheese isshaped or cut into small units and left in thebrine, as with Feta cheese, salt concentrationscan be very high (�3%). In contrast, largeblocks or wheels of brined Swiss cheese typi-cally contain less salt (�1%), especially in theinterior sections. Brined cheeses that are thenallowed to air dry develop natural rinds, dueto surface dehydration.

Aging

The last step in the cheese making processhas as much influence as any previous stepwith regard to the properties and qualities ofthe finished cheese. As noted in Chapter 1, itis a fine line that separates the production ofa perfectly flavored, three-year old Cheddarcheese and a bitter, rancid, sour Cheddarcheese that is quickly rejected by any discern-ing consumer. Although the distinctly differ-ent properties of both of these two cheeseare the result of microbial and enzymatic ac-tivities, there is one clear distinguishing fac-tor. The key difference is that the gourmetcheese is produced when aging occurs undercontrolled conditions, whereas the rejectedcheese occurs when control is absent or lost.As a general rule, any cheese that is intended



for aging must be manufactured, from thevery start, differently than an unaged cheese.The handling of the milk, the cheese pH, themoisture and salt content, and the water/saltratio, in particular, all are important determi-nants that influence aged cheese quality.

In addition to its impact on the finishedcheese,aging or ripening is also one of the mostcomplex and most variable of all cheese makingsteps.This is due largely to the enzymes and mi-croorganisms that are primarily responsible forflavor and texture changes that occur duringripening.The enzymes in cheese may occur nat-urally in the milk or be added directly in theform of rennet, chymosin, or lipase extracts. En-zymes are also derived from starter culture bac-teria, adjunct organisms, or endogenous milk-borne organisms. Furthermore, the availabilityof substrates and the pH and Eh conditions inthe cheese influence the activity of these en-zymes and the types and amounts of productsthat are formed. Similarly, microorganisms incheese originate from the milk, the environ-ment, and the starter culture.Although the tem-perature in aging rooms is generally low (usu-ally around 3°C to 7°C, but sometimes muchhigher),and the cheese milieu is not particularlyconducive for growth (as noted above),metabo-lism of the various substrates in cheese by intactorganisms still occurs.A ripening cheese repre-sents a rather vibrant ecosystem.

Types of Cheese

Given the hundreds of cheeses producedworldwide, it is obviously not possible to dis-cuss each particular one. There are, however,several ways to categorize the many differenttypes of cheese into manageable groups,based, for example, on their level of hardness(e.g., from soft to hard), moisture content,cooking temperature, or extent of aging. In thesections that follow, the manufacturing proce-dures for different cheeses and the role of mi-croorganisms involved in their production willbe reviewed, based on the primary propertiesof those cheeses and their distinguishing char-acteristics (Table 5–1). For the most part, only

the most well-known and widely consumedcheeses are discussed and only general proce-dures are described. The reader seeking de-tailed procedures is advised to consult otherexcellent sources for specific manufacturingdetails (see Bibliography).

Acid-coagulated cheeses

In the United States, the most popular of theacid-precipitated cheeses are Cottage cheeseand cream cheese. And although per capitaconsumption of Cottage cheese (all varieties)has declined in the past twenty years bynearly 30% (despite modest increases in low-fat versions), cream cheese per capita con-sumption has increased by more than 100%(from less than 0.5 Kg to more than 1 Kg perperson per year). This increase in creamcheese consumption is undoubtedly due toan equal increase in the popularity of bagelsand cheesecakes. The availability of flavored,whipped, and low- and reduced- fat creamcheese products has also contributed to thisincrease. Other cheeses in this category, in-cluding bakers’ cheese and farmers’ cheese,have only a small share of the market.

These cheeses rely on the fermentation oflactose to lactic acid by a suitable starter cul-ture, such that a pH of 4.6 or below is reached.In Cottage cheese manufacture, the starting ma-terial is simply skim milk.Often,nonfat dry milkis added to increase the throughput, or theamount of cheese produced per vat, and to im-prove body. The milk is always pasteurized, asrequired by law, since this is a fresh or non-agedproduct. A mesophilic lactic starter culture,con-taining strains of Lactococcus, is then added at arate or amount that depends on the productionschedule preferences of the manufacturer. For afast make, as much as a 5% culture inoculum isadded (that is, 5% of the total milk volume, byweight, is culture). The inoculated milk ismixed, then allowed to incubate quiescently inthe vat at 30°C to 32°C, the optimum tempera-ture for the culture.At this inoculum level and atthis temperature, an active culture can coagu-late the milk in five hours or less.


Cheese 167

In contrast, if a 1% inoculum is added and theincubation temperature is set at 20°C to 22°C,well below the optimum for growth (essentially,room temperature),coagulation may take twelveto sixteen hours. Thus, a busy cheese makercould,under the first scenario,produce multiplebatches of cheese each day out of the same vat.However, it is also possible, in the alternativeprocedure, to inoculate or set the milk at 5:00p.m., go home, sleep, and return to work early

the next morning and have the cheese ready forthe next step.Of course, inoculum and tempera-ture regimens in between the long-set and short-set make times are also possible.Finally,althoughnot required for coagulation, a small amount ofchymosin (1 to 2 ml per 1,000 Kg) is sometimesadded to promote a more firm coagulum (usu-ally for large curd cottage cheese).

Once a pH of about 4.7 is obtained, a softcurded gel mass is formed.The gel is then cut

Table 5.1. Properties of major cheese groups.

Cheese Starter Culture Other organisms Salt Moisture pH

Cheddar typeCheddar mesophilic1 1.5 37 5.5Cheshire mesophilic 1.7 38 4.8Colby mesophilic 1.5 39 5.5

Dutch typeGouda mesophilic Leuconostoc sp. 2.0 41 5.8Edam mesophilic Leuconostoc sp. 2.0 42 5.7

Cheese with eyesEmmenthal thermophilic2 Propionibacterium 0.7 35 5.6Gruyere thermophilic Propionibacterium 1.1 33 5.7

Grating typeParmesan thermophilic 2.6 31 5.4Romano thermophilic 5.5 23 5.4

Pasta filataMozzarella thermophilic 1.2 53 5.2Provolone thermophilic 3.0 42 5.4

Blue moldRoquefort3 mesophilic Penicillium roqueforti 3.5 40 6.4Gorgonzola mesophilic Penicillium roqueforti 2.5 45 6.2Stilton mesophilic Penicillium roqueforti 2.3 39 6.2

External moldBrie4 mesophilic Penicillium camemberti 1.6 52 6.9Camembert4 mesophilic Penicillium camemberti 2.5 49 6.9

Surface ripenedHavarti mesophilic 1.9 43 6.4Muenster mesophilic Brevibacterium linens 1.6 42 6.4Limburger5 mesophilic Brevibacterium linens 2.0 45 6.8

BrinedFeta mesophilic 3.0 53 4.5

1Mesophilic cultures � Lactococcus lactis subsp. lactis and Lactococcus lactis subsp. cremoris2Thermophilic cultures � Streptococcus thermophilus, Lactobacillus delbrueckii subsp. bulgaricus, and/or Lactobacillus helveticus3Citate-fermenting Leuconostoc or Lactococcus sp. may be added4Streptococcus thermophilus may be added5For Limburger and other surface-ripened cheeses, species of Arthrobacter, Micrococcus, and yeasts may also be present.

Data from Guinee and Fox, 2004; Marcos et al., 1981; and other sources



using a pair of cheese knives that resemblesquare or horizontal-shaped harps. One knifecontains vertical wires, and the other horizon-tal wires.The distance between the wires deter-mines the curd size, which in turn dependson the product being made (e.g., large,medium,or small curd Cottage cheese productsare available). Following the appropriate passesthrough the gel, die-shaped cubes are formed.

Almost immediately, whey becomes appar-ent and the curds begin to separate.The curdsafter cutting are very soft and fragile, and mustbe handled carefully to avoid shattering or frac-ture. Not only do fractured curds result in prod-uct defects, but, importantly, the small, brokencurds or “fines” are lost when the whey isdrained, resulting in loss of yield.Thus, the cutcurds are initially left undisturbed for about fif-teen minutes, and then are stirred gently.

Stirring continues as the curds are heatedby raising the temperature in the jacketed vats.The temperature must be raised slowly at first(whether from 20°C or from 32°C) to preventthe curd exterior from cooking too fast anddrying out. If this phenomenon,known as case-hardening, occurs, water molecules in the inte-rior of the curd cannot escape and are heldwithin the cheese. Ultimately, the curd-wheymixture is heated to about 52°C to 56°C, usu-ally within one and a half hours (but accordingto some procedures, as long as three hours),with constant stirring.

Heating not only accelerates the rate ofsyneresis, the driving out of water from thecurd, but temperatures above 45°C also arrestthe fermentation. In fact, inactivating the cul-ture at this step is the only way to keep the pHfrom decreasing too low. If the coagulatedcheese is not cut until the pH has alreadyreached 4.6 or below, then by the time thecook temperature is sufficiently high enough toinactivate the culture, the curds may already betoo acidic. Cooking also inactivates coliformsand other heat-sensitive microorganisms (espe-cially psychrotrophs, such as Pseudomonas).

After cooking has been completed, thewhey is drained and cold (4°C) water is ap-plied to the curds to quickly reduce the tem-

perature. The curd usually is washed two orthree times. The water used for this washingstep must be slightly acidic so that the precipi-tated casein is not re-solubilized.Also, the washwater is often chlorinated to ensure thatspoilage microorganisms are not inadvertentlyadded back to the curds.The water is drained,leaving behind what is called dry curd Cottagecheese.This product has a bland, acidic flavor(pH about 4.6 to 4.8), and contains mostly pro-tein (20%) and water (nearly 80%). It is usedmostly as an ingredient in lasagna, blintzes, andother prepared products. It is far more com-mon to add a cream-based dressing to the drycurds, producing the familiar creamed Cottagecheese products of varying fat levels (generallyranging from 0% to 4%). The cream dressingstypically contain gums and thickening agents,salt, emulsifiers, anti-mycotic preservatives(e.g., natamycin, sorbates, and other organicacids), and flavoring agents. Of the latter, di-acetyl distillates are frequently added to impartbuttery-like flavor notes. Because the dressingcontains lactose (which can be fermented), itis important that the starter culture bacteria beinactivated during the cooking step.

The packaged dry or creamed Cottagecheese is a perishable product and has a shelf-life of only two to three weeks under refrigera-tion conditions. However, the addition of newgeneration preservatives, such as Microgard,and the application of modified atmosphere(using CO2) and aseptic packaging, may in-crease shelf-life to as long as forty-five days.Cottage cheese is susceptible to spoilage forseveral reasons. The cheese vats are typicallyopen (although most newer vats are enclosed),and exposure to air and environmental mi-croorganisms can be significant.Because the fi-nal pH of creamed products can be as high as5.4, and the water activity is also high (0.98 to0.99),Pseudomonas and other psychrotrophicGram negative bacteria can grow and producefruity, rancid, and bitter off-flavors. Yeast andmolds are the other main spoilage organisms,causing appearance as well as flavor and aromadefects. Flavor defects, such as high acidity andbitterness, are the result of excess growth by


Cheese 169

the starter culture. Other quality defects, in-cluding shattered, gummy, or soft curd, are usu-ally caused by manufacturing flaws.

Manufacture of cream cheese (and its lowerfat version, Neufchatel) is similar in principle toCottage cheese, but the starting material, themanufacturing steps, and the finished productare quite different. First, cream cheese is madefrom pasteurized and homogenized milk con-taining as much as 12% fat. Most other cheesesare made using non-homogenized milk becausehomogenization results in a soft curd, which,in the case of cream cheese, is desirable. Amesophilic culture is added, followed by eithera long or short set incubation.The acid-inducedcoagulum that forms at pH 4.7 is stirred (notcut) while the temperature is raised to as highas 73°C.The curds are then separated from thewhey by special centrifuge-type devices or ul-trafiltration systems. Although the resultingcheese material can be packaged as is, mostcream cheese is mixed with other ingredients,including cream and gums, and then homoge-nized or mixed.The temperature is maintainedabove 72°C throughout the process. Packagingis done under aseptic conditions (i.e., in roomsunder positive pressure with high efficiency airfiltration systems in place),giving this product along shelf-life (� 45 days).

Cheddar family

Cheddar cheese is the most popular cheeseconsumed in the United States, with per capitaconsumption in 2001 of nearly 6 kg (12.7pounds) per person per year (including otherCheddar types). Approximately 1.2 billion Kg(2.7 billion pounds) are produced each year inthe United States. Several cheeses, includingColby, Monterey Jack, and other washed orstirred curd cheeses,are closely related to Ched-dar cheese and are collectively considered asAmerican type cheeses. However, this designa-tion can be confusing since American cheese isoften considered synonymous with processedAmerican cheese, a totally different product.

Cheddar cheese was first made in the villageof Cheddar, England. It is different from most

cheeses due to a specific curd handling tech-nique called Cheddaring. During Cheddaring,the curds are separated from the whey and al-lowed to mat, after which the matted curdslabs are flipped and stacked.The Cheddaringpractice, like many great discoveries, probablyoccurred as a result of an accident.Perhaps thecheese maker had drained the whey, and in-stead of stirring the curds prior to filling formsor hoops, he or she was delayed or distracted.The curds then matted, staying warmer thanusual, and then began to stretch and becomeplastic. The cheese that was then producedhad a unique texture and body that led to theadoption of the Cheddaring process.The man-ufacture of Cheddar cheese has evolved from atraditional, strictly batch operation to a moremodern process that is more automated andmechanized.

The early steps in Cheddar cheese manufac-ture are rather straightforward.Whole milk, ei-ther pasteurized or raw, is brought to a temper-ature of 30°C to 32°C, and is inoculated with amesophilic lactic culture containing strains ofL. lactis subsp. lactis and L. lactis subsp. cre-moris. The culture, or rather the specificstrains present in that culture, is selected basedon the desired properties expected in the fin-ished cheese. Sometimes S. thermophilusstrains are added to promote rapid acidifica-tion. If the cheese is intended for the processmarket, then speed is the main criteria, andfast-growing strains of L. lactis subsp. lactis areoften used. If, however, the cheese is to beaged, then appropriate flavor development willlikely drive selection of the culture (usuallycontaining L. lactis subsp.cremoris).Two prop-erties, however, are required: the strains mustferment lactose and they must be able to hy-drolyze and use proteins as a nitrogen source.Of course,other properties, as well as the formof the culture (whether bulk set or direct-to-vat set, frozen or lyophilized,mixed or defined)are also important, and are discussed later inthis chapter.After culture addition, chymosin isadded, coagulation occurs, and the curd is cutusing medium sized knives (giving curds about0.6 to 0.8 cm in diameter). Following a short



five- to ten-minute “healing”period to allow thecurds to form, the curds are gently stirred.Heatis then applied gradually, about 0.3 degrees per minute, to a final curd-whey temperatureof about 38°C or 39°C. As the curds arecooked, the stirring speed is increased to pro-mote heat transfer, and stirring is maintainedfor an additional forty-five minutes. Althoughthe starter culture bacteria are containedwithin the curd, as a general rule, little fermen-tation should occur during the cooking or stir-out steps. A decrease of only 0.1 to 0.2 pHunits is normal.Although L. lactis subsp. lactishas a slightly higher temperature range forgrowth compared to L. lactis subsp. cremoris,both grow slowly, if at all, at the upper end ofcommon cooking temperatures. If however,lower cooking temperatures are used, growthof the culture can occur, resulting in fermenta-tion and lactic acid formation. Although fer-mentation during this step may be desirablefor some cheeses, for Cheddar it generally isnot. This is because acidification is accompa-nied by demineralization of casein as the cal-cium becomes solubilized. When the whey isdrained, the remaining curds contain less cal-cium, resulting in cheese that holds fat poorlyand is less elastic, with a short, brittle texture.The importance of calcium during later stagesof the Cheddar process is discussed below.

Once most of the whey is drained, the curdsare assembled on both sides of the vat, wherethey quickly mat or stick together. By the timethe last of the whey is removed, the curds haveturned into a cohesive mass. A knife is thenused to cut the matted curds into approxi-mately 20 cm x 80 cm slabs. Depending on thecheese maker’s preferences, the slabs are theneither rotated or flipped or flipped and stacked,a process known as Cheddaring. Because thetemperature of the Cheddar slabs is maintainedat 28°C to 32°C, the culture can begin to growin earnest, increasing from about 106/g to 108/gof curd within the Cheddaring period. As thelactic culture grows, it performs two criticalfunctions. First, it ferments lactose in the curd,in homolactic fashion, to lactic acid. Second, ithydrolyzes casein via a cell wall-anchored pro-

teinase,forming a variety of variously sized pep-tides.The consequences of both of these activi-ties will be discussed below.

During the Cheddar process, several biologi-cal, chemical, and physical changes occur thatgive this cheese it’s characteristic properties.However,without an active starter culture,noneof these changes will occur.As the culture pro-duces lactic acid and as pH becomes moreacidic, calcium that was initially associated withthe negatively-charged amino acid residues ofcasein (e.g., serine phosphate) is displaced byprotons. In this state, the casein is more soluble,smooth,and elastic, as is evident as the Cheddarslabs stretch and become plastic-like.Eventually,however, if the pH of the cheese becomes tooacidic (less than 4.95), the cheese will becomeshort and crumbly. Fermentation during Ched-daring, what Cheddar masters refer to as dryacid, is much preferred over acid formation dur-ing the cooking step, or wet acid development.In addition to the changes in casein structurethat occur as a result of acid formation, gravita-tional forces from stacking the Cheddar slabs ontop of one another also affect cheese structure.This facilitates linearization and stretching ofthe casein and allows the curds to knit whenpressed (see below).

The longer the cheese is Cheddared, themore the culture will grow and with it, thegreater will be the acidity.The extent of Ched-daring, and with it the extent of acid forma-tion,depends, therefore,on the desired proper-ties of the finished cheese. Lactic acid levels of0.4% to 0.6% are normally achieved by the endof Cheddaring (about one and a half to twohours after whey separation), resulting in a fin-ished cheese pH of 5.0 to 5.2. In addition,longer Cheddaring times also result in morecell mass produced. Since the lactic acid bacte-ria in the culture produce enzymes that de-grade milk proteins, higher concentrations inthe curd at this stage likely results in greaterproteolysis at later stages, i.e., during ripening.

Although the traditional Cheddar process isstill practiced in the United States in smallcheese factories, most mid-sized and large op-erations use different procedures and produce


Cheese 171

a somewhat different type of product. This isbecause large cheese factories require muchgreater throughput than can be obtained bytraditional procedures. Thus, setting, coagula-tion, and cutting steps are performed in en-closed vats,and the cut curds are then pumpedto another location for further processing.Thisallows the manufacturer to clean, sanitize, andre-fill the original vat.

In addition, modern operations have elimi-nated the labor-intensive flipping and stackingCheddaring steps by employing alternativemethods. Cheese makers learned that a similartexture could be obtained if the dry curds, afterdraining,were simply stirred in the vat.An addi-tional variation of this process involves addingcold water to the curds. As noted above, thisstep results in less lactose in the curd, and italso increases the moisture. The resultingcheese, known as Colby, is a less acidic, highmoisture cheese,compared to traditional Ched-dar. In some modern cheese factories, curds arecontinuously conveyed to the top of Cheddar-ing towers. By the time the curds have de-scended from these vertical towers and emergeout at the bottom, they have achieved a similarlevel of Cheddaring as traditional curd.

After Cheddaring, or when the pH is about5.2 to 5.4, the plastic, elongated Cheddar slabsare chopped up in a special milling device(called,appropriately,a Cheddar mill) to reducethe slabs to uniform, thumb-sized pieces.TheseCheddar curds are bland, with a squeaky, rub-bery texture. Next, salt is applied, in an amountranging from 2% to 3%, and in a manner thatpermits even salt distribution (i.e., while thecurds are continuously being stirred).

Salting is a critical step, since it has a pro-found effect on the quality of the finishedcheese. This is because, in addition to provid-ing flavor, salt has a major influence on control-ling microbial and enzymatic activities in theripening cheese.Although 2% salt might not beexpected to have much of an effect on mi-croorganisms or enzymes in food, one mustconsider that cheese consists of two phases, afat phase and a water phase. It is in the waterphase that the salt is dissolved, and likewise,

that is where the microorganisms live. SinceCheddar cheese contains no more than 39%water, the relevant salt concentration, i.e., thatwith which the microorganisms must contend,will be more than 5% (2/39 � 100 � 5.1%).Atthis concentration, the growth of the starterculture bacteria and many other microorgan-isms and the activities of microbial and milk-derived enzymes are effectively controlled.This is not to imply that microbial and enzy-matic activities are actually halted, becausesalt-tolerant organisms and enzymes remain ac-tive in the presence of high salt concentra-tions. Rather, the salt provides a means tocheck or contain those activities and to createa selective environment that aids in establish-ing a desired microflora.

The active salt concentration is often re-ferred to by cheese manufacturers as the salt-in-moisture or S/M ratio.For example, a cheesewith 2.4% salt at a moisture of 38% will have aS/M of 6.3.The S/M value is arguably the maindeterminant affecting cheese ripening. Thehigher the S/M level, the more inhibitory will that environment be to microorganismsand enzymes. In contrast, for cheeses havinglow S/M values (i.e., below 4.5), the microflorawill not be effectively constrained or con-trolled and production of off-flavors and otherspoilage defects may occur. However, cheeseripening rates, which are also a function of mi-crobial and enzymatic activities, will also beinhibited if the S/M is too high. Thus, cheesemanufacturers must carefully adjust salt con-centrations and moisture levels (as well as pH)to achieve the desired ripened cheese proper-ties (Figure 5–6). In general, an S/M between4.5 and 5.0 is desirable.

The next step involves filling forms, hoops,or barrels with the salted curds. The size ofthese forms range from 9 Kg to as high as 290Kg (20 pounds to 640 pounds). Obviously, theform also gives shape to the cheese, varyingfrom rectangular blocks to barrels. Many of thecheese forms have collapsible ends, such thatpressure can be applied to enhance the trans-formation of the curds into a solid mass and tosqueeze out whey (through perforations in the



hoops). Usually, about 20 psi is applied fortwelve to sixteen hours.The cheese is then re-moved from the forms,and, for the 9 Kg and 18Kg blocks, placed in Cryovac bags, vacuum isapplied, and the bags are sealed. The cheesethen is moved into cold rooms, ranging in tem-perature from 4°C to 12°C for storage or aging.

Cheese with Eyes

Swiss cheese, the most famous of the eye-containing cheeses, is considered by experi-enced cheese makers to be the easiest cheeseto make but the hardest cheese to make well.This is because the manufacture of high qualitySwiss cheese, and proper eye development, inparticular, depends on two rather independentprocesses. First, Swiss cheese requires excel-lent curd handling technique, such that condi-tions are correct for eyes to form. Second,there must be precise control over the mi-croorganisms that are involved in the fermen-tation and produce the gas that ultimately re-sults in eye formation. It is possible, easy infact, to produce a cheese that tastes like Swisscheese, but that has either small eyes, largeeyes, irregularly-shaped eyes, too many eyes,too few eyes,or no eyes.The goal is to producea cheese with the correct flavor and body char-

acteristics, and that has the right amount ofuniformly distributed, evenly-shaped roundeyes with the desired size.The trick is to man-age the early cheese-making steps such thatthe curd has just the right texture or elasticitynecessary to accommodate the carbon dioxidethat is produced much later in the process.

The manufacture of Swiss cheese seemseasy because the steps involved in curd mak-ing appear rather simple and straightforward.However, care must be taken from the outset.First, the milk is usually standardized to 3% fat.If too much or not enough fat is present in themilk, the body will be too soft or too hard, lead-ing to poor eye development. Second, the milkis given a modest heat treatment, about 56°Cfor fifteen seconds, which is below that usedfor pasteurization but which still is effective atinactivating many of the milk-borne bacteriathat could otherwise cause problems duringthe ripening period. Finally, the quality of thecheese also depends heavily on the perfor-mance of the bacteria used in the mixedspecies starter culture, as well as the ratio ofthose organisms.

To achieve the texture and body character-istics necessary for proper eye development,Swiss cheese must have the correct pH andmoisture content and retain (or lose) the right

S/M4.0 – 6.0

MSNF50 – 56

pH5.0 – 5.4

FDM50 – 57

FDM52 – 56

S/M4.7 – 5.7

MSNF52 – 54

pH5.1 – 5.3

Premium cheese(First Grade)

Second Grade Figure 5–6. Relationships between S/M (salt tomoisture ratio), MSNF (percent of moisture in thenon-fat portion), FDB (percent of fat on a dry ba-sis), and pH in Cheddar cheese at fourteen daysafter manufacture. To obtain a Premium cheese(or First Grade, according to New Zealand stan-dards), the more narrow compositional specifica-tions are usually necessary (compared to Grade Bcheese which has broader compositional specifi-cations). Adapted from Lawrence et al., 2004.


Cheese 173

amount of calcium.The curds are cut to aboutrice-sized particles, and then are cooked to amuch higher temperature than for Cheddar-type cheeses.Typically, the temperature will beraised over a forty-minute period from 32°C to35°C at setting to as high as 55°C, and then isheld at that temperature during stir-out for an-other forty-five minutes or until the curd is suf-ficiently dry.Thermophilic cultures that are tol-erant of high temperature must be used (seebelow), although there are some manufactur-ers who cook to a lower temperature and addmesophilic L. lactis subsp. lactis.

It is important to note that reducing themoisture in the curd will also result in less lac-tose in the curd. As noted above, controllingthe amount of lactose in the curd is an effec-tive way to limit the amount of acid that willultimately be produced during fermentation, aconsideration that will be further discussedlater.High cooking temperatures also slow aciddevelopment, remove water, and prevent solu-bilization of calcium phosphate into the whey.The whey is drained at the end of cooking,while the pH is still high (about 6.3). Thiskeeps phosphate in the curd, which promotesbuffering and keeps the pH from becomingtoo low later during the fermentation.

In the traditional process, when the curdsare sufficiently firm, they are collected ordipped into cheese cloth, which is then placedinto a large round form. In a modified versionused by many U.S. manufacturers, the curds areallowed to settle under the whey, sometimeswith weights applied,then the whey is drained,the matted curd is cut into block-sized sections,and placed into forms. In the larger operations,curd-whey mixtures are pumped into drainingvats and then the curd is filled into large forms.For all of the processes, the cheese blocks arethen pressed and held for up sixteen hours atnear ambient temperature.

Unlike Cheddar cheese, in which the fer-mentation occurs in the vat or on draining ta-bles, for Swiss cheese there has been no op-portunity, up this point, for fermentation.Rather, the fermentation occurs after thecheese is out of the vat and filled into forms. In

addition, the fermentation takes a much longertime, as much as twenty-four hours. Thecheese, however, will still be warm followingthe cooking step, so the internal temperaturemay be as high as 35°C for several hours.Theactual fermentation is quite different from thatwhich occurs during Cheddar cheese manu-facture. In Cheddar cheese, the culture con-tains mesophilic lactococci, and althoughmore than one species may be present (e.g., L.lactis subsp. lactis and L. lactis subsp. cre-moris), they are obviously closely related andserve an almost identical function. In contrast,the Swiss cheese starter culture contains threedifferent organisms, from three different gen-era:Streptococcus thermophilus,Lactobacillushelveticus, and Propionibacterium freudenre-ichii subsp. shermanii.All three are essential,and all are responsible for the fermentationpattern that is unique to Swiss cheese.

The culture that is initially added to the milkcontains different proportions of each organ-ism.The S. thermophilus generally outnumbersthe L. helveticus by as much as ten to one.Therefore, at the outset, growth of S. ther-mophilus occurs first, in part because it is pre-sent at a higher concentration, but also be-cause its simple physiological requirementsare more easily met, especially compared tothe more fastidious L. helveticus. As noted inChapter 4, S. thermophilus and L. helveticus(or Lactobacillus delbrueckii subsp. bulgari-cus) have a synergistic relationship, such thatgrowth of one organism promotes growth ofthe other. In Swiss cheese, S. thermophilusgrowth is stimulated by amino acids and pep-tides released from casein via L.helveticus pro-teinases. Growth of the latter does not com-mence until the pH and Eh within the cheeseare sufficiently reduced by S. thermophilus, aperiod that may take as long as twelve hours.

Another factor that influences the outcomeof the Swiss cheese fermentation relates di-rectly to the metabolic properties of these twoorganisms. In fact, one of the most interestingpeculiarities of all lactic acid bacteria occurswhen S. thermophilus ferments lactose inSwiss cheese (actually, the same phenomena



also occurs in Mozzarella cheese, yogurt, oranytime S. thermophilus grows in milk). Thisorganism actively transports lactose from theextracellular environment across the cell mem-brane and into the intracellular cytoplasm.Theenzyme, �-galactosidase, then immediately hy-drolyzes the accumulated lactose to glucoseand galactose. Glucose is phosphorylated toglucose-6-phosphate, which then feeds intothe glycolytic pathway and is rapidly metabo-lized to lactic acid. However, most strains of S.thermophilus do not express the enzymesnecessary for galactose phosphorylation andmetabolism (enzymes that comprise the Leloirpathway), and instead secrete or efflux thegalactose back into the extracellular medium.In the case of Swiss cheese, galactose will ap-pear in the curd almost as fast as lactose is con-sumed. Researchers have learned that the ex-cretion of galactose not only coincides with

lactose consumption, but that the efflux reac-tion actually provides a driving force for lac-tose uptake (described in Chapter 2).

Thus, in the first several hours of the fermen-tations,S. thermophilus grows and ferments lac-tose and excretes galactose (Figure 5–7). L. hel-veticus begins active growth after about eightto twelve hours and competes with S. ther-mophilus for the remaining lactose, which issubsequently consumed. However, one of themain roles of L.helveticus is to then ferment thegalactose left behind by S. thermophilus, suchthat after about eighteen to twenty-four hoursall of the carbohydrate in the curd has been fer-mented. If the curd contained just the rightamount of lactose at the start of the fermenta-tion (i.e., after cooking), and all of the lactoseand its constituent monosaccharides were fer-mented to completion, then the final pH shouldbe very near 5.2 � 0.1. If there is too much lac-

0 4 8 12 16 20 240

60

50

40

30

20

10

Time in hours (after pumping)

noitartnecnoc etardyhobraC

)druc fo marg/selo

mm(

0 4 8 12 16 20 244

9

8

7

6

Time in hours (after pumping)

druc ni stnuoc lairetcaB

)gufc gol(

5

A

B

Lactose

Galactose

Glucose


Lactobacillus helveticus

Figure 5–7. Carbohydrate utilization byStreptococcus thermophilus and Lactobacillushelveticus during the Swiss cheese fermenta-tion. The upper pane (A) shows carbohydrateconcentrations during twenty-four hours afterthe curds have been pumped from the vat. Thelower panel (B) shows growth of the culture or-ganisms during the same period. Adapted fromGiles, et al., 1983 (N.Z. J. Dairy Sci. Technol.18:117-123) and Turner et al., 1983 (N.Z. J.Dairy Sci. Technol. 18:109-115).


Cheese 175

tose or too little lactose at the start,due to insuf-ficient or excessive cooking, then the pH canend up being less than 5.0 or above 5.4.Both ofthese situations could result in poor qualitycheese, as described below.

After the primary fermentation period, thecheese blocks or wheels are placed into brinescontaining 20% salt for as long as three days.Theblocks may be flipped and additional salt maybe applied.Then the blocks are removed and al-lowed to air dry in coolers at 10°C to 15°C forfive to ten days. As noted for Cheddar cheese,salt provides flavor and influences the activitiesof microorganisms and enzymes present in thecheese, although the average concentration inSwiss cheese is much lower (� 1%). In the caseof traditional Swiss cheese and other brinedcheeses, salt also helps form a natural rind, dueto dehydration at the surface of the cheese.Thehard rind provides an excellent natural protec-tive barrier or casing. Brining, in contrast to di-rect or dry salt methods,also creates a salt gradi-ent,with the concentration decreasing from thesurface toward the interior of the cheese.Thisalso affects the development of the microflora.

The next step is perhaps the most crucial.The dried blocks or wheels are moved intowarm rooms where the temperature is main-tained at 20°C to 25°C. It is during the ensuingthree to four weeks that growth of Propioni-bacterium freudenreichii subsp. shermaniioccurs.Although this bacterium is morphologi-cally, physiologically, and genetically distantfrom the lactic acid bacteria, it is similar in thatit prefers an anaerobic atmosphere and has afermentative metabolism. Propionibacteria areneutrophiles and are salt-sensitive, so low pHand high salt conditions are inhibitory. Theyare added to the milk as part of the starter cul-ture, but at a much lower rate—the inoculatedmilk contains only about 102 to 103 cells perml.And although these bacteria are moderatelyresistant to high temperature, growth does notoccur until the cheese is moved into the warmroom. When these bacteria are grown on lac-tose or other fermentable carbohydrates, largeamounts of propionic and acetic acids andlesser amounts of carbon dioxide are pro-

duced. However, as explained above, by thetime Swiss cheese is placed in the warm room,there should not be any carbohydrate still inthe curd and available for fermentation. Sowhat does P. freudenreichii subsp. shermaniiuse as a substrate? In fact, this organism is ca-pable of metabolizing the lactic acid producedby the lactic starter culture via the propionatepathway (Figure 5–8).This pathway yields pro-pionic acid, acetic acid, carbon dioxide andATP. If, however, any lactose is still available,then more of these products will be formed.Moreover, excess production of carbon diox-ide, in particular, has serious consequences forthe final product.

As noted above,P. freudenreichii subsp. sher-manii is initially present at relatively low levelsin the cheese. As it grows in the warm room,small micro-colonies within the curd matrix areformed. Likewise, the fermentation end prod-ucts are evolved in that same vicinity and dif-fuse out into the neighboring region within thecurd. Although the acids are readily dissolved,the carbon dioxide molecules will diffusethrough the curd only until they reach weakspots, where they will then collect along withother CO2 molecules made by other micro-colonies.Eventually,enough CO2 molecules willhave accumulated,and voila—an eye is formed.However, this entire sequence of events de-pends on several hard-to-control variables.

First, CO2 formation must be slow andsteady. If too much CO2 is produced all at onceor the curd is too firm, the gas pressure can ex-ceed the ability of the curd to sustain the gasand large, even exploded holes are formed. Ifthe body of the curd is too soft and the weakspots too numerous, then many small eyes willform.The hard rind produced as a result of thebrining and drying steps also serves an impor-tant role; without a rind, the CO2 could theo-retically escape clear out of the cheese. Ofcourse,CO2-impermeable bags provide an easyremedy for this problem (thus was born rind-less Swiss, as described below).Obviously, timeis a critical factor as well—too much or notenough incubation time will result in less thanperfect eye development. Experienced cheese



makers have adopted various professionaltricks, from tapping the cheese and listeningfor just the right echo, to visually examiningthe expansion of the wheel or block.

Finally, when the cheese leaves the warmroom, it is moved into a cooler at 2°C to 5°Cfor aging, which can vary from three months to two or more years. In addition to the acidproducts (i.e., propionic and acetic acids),P. freudenreichii subsp. shermanii is also re-sponsible for producing other important end-products. In particular, P. freudenreichii subsp.shermanii produces specialized peptidasesthat release proline, which has a sweet-like fla-vor.Various other amino acids and peptides arealso metabolized, generating nutty flavors thatare characteristic of Swiss cheese.

Variations of the traditional Swiss or Em-menthaler cheese (made in Switzerland) in-clude Gruyere (France), Jarlsberg (Norway),and Samsoe (Denmark).All of these have a nat-ural rind and varying levels of eye formation. Inthe United States, rindless versions are morecommon, with the cheese blocks wrapped inCO2-impermeable plastic wrapping.

Mozzarella and Pasta Filata Cheese

As a recently as a generation ago, Mozzarellawas still considered a mostly ethnic cheese,used primarily as an ingredient in Italian cui-sine.The popularity of this cheese—it is nowequal to that of Cheddar cheese among Ameri-can consumers (Figure 5–3), is due to oneproduct: pizza. Of the more than 1.2 billion Kg(2.7 billion pounds) of Mozzarella and relatedcheeses consumed each year, about 70% isused by the food service industry as an ingredi-ent on pizza. And while Cheddar productionhas grown by about 60% since 1980, over thatsame period, Mozzarella production has in-creased by 400%. In fact, in the last twentyyears, many cheese factories that once madeCheddar converted their operation to Moz-zarella manufacturing. And despite the appar-ent differences in flavor and appearance be-tween Cheddar and Mozzarella cheese, bothshare several common manufacturing steps.

In the United States, all Mozzarella cheese ismade from pasteurized milk. In large part, thisis because Mozzarella is considered a fresh

succinyl CoA

PEP malate

methylmalonyl CoA

oxaloacetate fumarate

succinatepropionyl CoA

ADP

ATP

lactate

CO2

acetatepropionate

pyruvate

acetyl CoA

acetyl-P

Figure 5–8. The propionic acid pathway of propionibacteria. Only the key intermediate compounds areshown. Adapted from Hutkins, 2001.


Cheese 177

cheese that is essentially never ripened.Thus,legal requirements dictate that the milk be pas-teurized. Although some Mozzarella cheese ismade from whole milk, most is made from re-duced fat or partially skimmed milk. Specialconsiderations for the manufacture of reducedfat Mozzarella and other low- or reduced fatcheeses will be described later. Although al-most all Mozzarella cheese made in the UnitedStates comes from cow’s milk, it is worth not-ing that some traditional Italian Mozzarella ismade from the milk of water buffaloes.

After the milk is pasteurized and standard-ized, a thermophilic culture is added.Althoughthe organisms used for Mozzarella, Streptococ-cus thermophilus and Lactobacillus helveti-cus, are the same as those used for Swisscheese, the specific strains and ratios are likelyquite different. Strain selection for Mozzarellacheese is based, like all cultures, on the desiredproperties of the particular cheese.This is es-pecially true for Mozzarella cheese, where thestrains may have a profound effect on theproperties of the finished cheese (Box 5–5).

Box 5–5. The Science of Making Pizza

Given that the U.S. pizza industry is a $32 billion industry, that 350 slices of pizza are eatenevery second of every day, and that Mozzarella and pizza-type cheeses are the main ingredientson a pizza, it is not surprising that Mozzarella producers are under considerable pressure to pro-duce a cheese with the functional properties desired by the pizza manufacturer.

Of course, pizza restaurants are the main users of Mozzarella, with several companies domi-nating the industry (Dominos, Pizza Hut, Godfathers, and Papa Johns). When considering thecheese, each pizzeria seems to have their own particular preference. Some pizza chains desirecheese that stretches a particular distance when a slice is moved from plate to mouth. Otherswant a cheese that remains white and resists browning even when the pizza is baked quickly atextremely high temperatures.Thus,Mozzarella production is now often customized to the exactneeds of the customer, such that even before the vat is filled with milk, the manufacturer isready to produce a cheese meeting not only compositional requirements, but functional speci-fications as well.

Because of these considerations, there is now much interest in understanding and predictinghow a given cheese will perform during pizza manufacture and how manufacturing conditionscan be manipulated to produce a cheese that has specific functional properties (Kindstedt,1993). Unlike most other cheeses, Mozzarella, at least when used on pizza, is prized less for itsflavor and more for its physical and functional attributes. Mozzarella, after all is a bland, slightlyacidic cheese, with no aged cheese flavor. However, it has a unusual ability to stretch, retain fat,melt evenly, and provide a chewy mouth feel.These properties, however, are not necessarily au-tomatic, and are influenced by the culture, the coagulant, the manufacturing conditions, andpost-manufacturing handling and storage. Ultimately, these factors affect pH, loss of calcium,proteolysis, and the overall composition of the finished cheese.

The metabolic activities of the thermophilic starter culture bacteria have a major impact onthe finished product.After the curds are cooked and the whey is drained, the curds are eitherCheddared or dry stirred. During this time, Streptococcus thermophilus and Lactobacillus hel-veticus (or Lactobacillus delbrueckii subsp. bulgaricus) ferment lactose, producing lactic acidand causing the curd pH to decrease. However, both organisms metabolize only the glucoseportion of lactose; the galactose is released back into the curd.

When the pH reaches 5.2, the curds head off to the cooker-stretcher (operating at 85°C),which heats the curd to a temperature near 60°C.This process inactivates enzymes (althoughsome residual chymosin might remain) and the starter culture (or at least reduces the cell con-centration several fold).Thus, there is essentially no opportunity left for this galactose (and anyremaining lactose) to ferment. Instead, it will remain intact and be present in the finished cheese.



When the cheese is exposed to high baking temperatures (which dry the cheese surface and reduce the water activity), these reducing sugars react with amino acids, via the non-enzymaticMaillard reaction, to from brown pigments.Although moderate browning may be desirable,exces-sive browning or blistering is undesirable. Because most pizza manufacturers prefer white, non-browning cheese, steps must be taken to reduce the galactose concentration in the cheese.Thiscan be done either mechanically, by simply washing the curds, or biologically, by using culturesthat can ferment galactose.The latter may consist of mesophilic Lactococcus lactis subsp. lactisthat have no problem fermenting galactose, but their use requires other changes in the manufac-turing process (i.e., lower cooking temperature). Selected galactose-fermenting strains of S. ther-mophilus and L.helveticus can also be used,and although such cultures are available from starterculture suppliers, their availability appears limited (Mukherjee and Hutkins, 1994). Plus, evengalactose-fermenting strains do not ferment galactose as long as lactose is still available. It is alsotechnically possible to engineer such strains,but this strategy has not yet been adopted.

The culture, and specifically, the proteolytic activity of L. helveticus, also affects other impor-tant properties.The ability of Mozzarella cheese to stretch is largely a function of protein struc-ture. If, however, the culture is proteolytic (or residual coagulant activity is present in the curdafter cooking), protein hydrolysis will occur, and the shortened casein strands will lose theirability to stretch.This may or may not be a good thing,depending on the preference of the pizzamanufacturer (Dave et al., 2003).

If a short, non-stretching cheese is desired, then greater proteolysis may be encouraged. How-ever, manipulating any one variable to affect a given property will likely affect some other prop-erty. As noted previously, extensive proteolysis makes the cheese “soft” and “gummy,” and lessamenable to shredding. In addition, it is the protein network that traps fat in Mozzarella. Duringbaking, the fat melts and liquefies, but is retained by the protein matrix. If proteolysis has oc-curred and the protein network is disrupted, the fat is no longer contained and leaks out. On apizza, the fat forms unsightly pools of melted oil, a defect known as oiling-off. Finally, since pro-teolysis increases the free amino acid concentration, the more proteolytic the culture (or the co-agulant), the greater the rate of the Maillard reaction and the potential for browning problems.

Ordinarily, the amount of time between manufacture of Mozzarella and its use on a pizza isonly a few weeks.Thus, proteolysis in Mozzarella is nowhere near as extensive as in an agedcheese. Still, even moderate proteolysis can affect cheese functionality. Many Mozzarella pro-ducers, therefore, freeze the cheese within a few days of manufacture, giving it just enough ag-ing time to develop appropriate properties. Some manufacturers have even developedprocesses in which the cheese is frozen within hours of manufacture.

Manufacturing variables have also been modified to affect other functional properties.For ex-ample, most Mozzarella is eventually used in a shredded or diced form, so producing a cheesethat is easy to shred, without gumming up the equipment, is important.These modificationshave led some to question whether the cheese produced is actually Mozzarella, as defined by the standards of identity. Nonetheless, it seems likely that the demand for Mozzarella (or Mozzarella-type) cheese with particular functional characteristics, combined with significanteconomic pressures, will continue to be a driving force for innovative research within the Moz-zarella and starter culture industries.

References

Mukherjee,K.K., and R.W.Hutkins.1994. Isolation of galactose fermenting thermophilic cultures and theiruse in the manufacture of low-browning Mozzarella cheese. J. Dairy Sci. 77:2839–2849.

Dave, R.I., D.J. McMahon, C.J. Oberg, and J.R. Broadbent. 2003. Influence of coagulant level on proteolysisand functionality of mozzarella cheeses made using direct acidification. J. Dairy Sci. 86:114–126.

Kindstedt, P.S. 1993. Effect of manufacturing factors, composition, and proteolysis on the functional char-acteristics of mozzarella cheese. Crit. Rev. Food Sci. Nutr. 33:167–187.

Box 5–5. The Science of Making Pizza (Continued)


Cheese 179

After rennet addition, coagulation, and cut-ting, the curd-whey mixture is pumped intodraining tables, where the curds are gentlystirred and then cooked to 56°C. The whey isdrained and the curds are either allowed to mat,as far Cheddar, or are dry-stirred (as for stirred-curd Cheddar). Frequently, the curds will bewashed after draining to reduce lactose levels.Mozzarella is a brine-salted cheese;however,saltmay also be added directly to the curds prior tothe cooking-stretching step to minimize thelengthy brine-salting time ordinarily required.

The fermentation of lactose in Mozzarella issimilar to that which occurs in Swiss cheese. Inboth fermentations, galactose appears in thecurd as a result of the inability of S. ther-mophilus to efficiently metabolize both of themonosaccharide moieties of lactose. In Swisscheese, there is plenty of time for the compan-ion strain (i.e., L. helveticus) to eventually fer-ment the galactose and remaining lactose.However, the situation is quite different forMozzarella. This is because when the pH dur-ing Cheddaring or dry-stirring reaches 5.2, thecurds are moved to a cooker-stretcher devicethat exposes the curds to temperatures as highas 85°C. Although the curd temperature maynot reach such a high temperature (usually thecurd is about 54°C to 57°C), this treatment isstill sufficient to inactivate many of the starterculture bacteria present in the cheese. Thus,any galactose or lactose still in the curd at thecooking-stretching step will likely remain inthe cheese for the duration of its manufactureand storage.The consequences of this are dis-cussed in Box 5–5.

In addition to its effect on the culture andresidual enzymes, the cooking-stretching stepalso has an important impact on the physicalproperties of the cheese. Prior to this step,Mozzarella curds are not that much differentthan other rennet set curds. However, thecheese exiting the cooking-stretching ma-chines has properties unlike any other cheese.The cooking-stretching machines use augers toknead, stretch, and convey the curds in an up-ward manner (about a 30° incline), all thewhile exposing the curds to a hot water-steam

mixture at 85°C to 90°C. In the United Statesand Italy, almost all Mozzarella is stretched inhigh through-put continuous cooker-stretcherdevices. Only a relatively few very small manu-facturers rely on traditional techniques—man-ually kneading the curd in hot water, an ardu-ous and time-consuming task.

Although what actually happens to the ca-sein matrix during this step is not clearly un-derstood, it appears that the casein networkbecomes linearized, giving the fibrous prop-erty characteristic of Mozzarella cheese anddesired by most consumers. It is critical thatthe cooking-stretching step occurs when thecurd pH reaches 5.2,otherwise the cheese willbe short-textured and will stretch poorly. Fi-nally, the molten, plastic, and fluid curds arethen dropped into a hopper and filled intomolds of varying loaf sizes and shapes. Thecheese sets up quickly in the molds and is thendropped into a cold salt brine. After brining(about sixteen hours, but much less if dry salthad been added to the curd), the cheese isdried and packaged. In Italy and in small U.S.factories that employ traditional manufactur-ing practices, Mozzarella can be formed intosmall round shapes, which are then packagedin water or dilute brines.

Mozzarella cheese, as noted above, is notaged; however, even within just a few days orweeks, important functional properties canchange. One of the first changes that occurs isthe equilibration of calcium,that which exists aspart of the casein complex and that in the formof calcium phosphate. Importantly, the loss ofcalcium from the casein complex improves themelting and stretching properties of the cheese.

The other main change in the functionalproperties of Mozzarella is due to protein hy-drolysis. Proteolysis can occur as a result ofresidual proteinases released by the lacticstarter culture, as well as by residual coagu-lant (which appears to be the main source ofproteinase activity). Although flavor changesdue to protein hydrolysis are usually minor,texture properties can be significant. As thelong casein strands are hydrolyzed, the cheesestretches less. Also, the hydrolyzed casein is



less able to hold or contain the fat.The lattermay result in a defect known as “oiling off”that is readily apparent when the cheese isused in pizza and other cooked products.Also, since most Mozzarella cheese is mar-keted in a shredded form, excessive proteoly-sis makes the cheese less shreddable. Forthese reasons, fresh Mozzarella is best usedwithin two to three weeks. Alternatively, thecheese can be frozen shortly after manufac-ture or after shredding (the preferred form formost of the large pizza manufacturers), and al-though this will prevent proteolysis, freezingis not without its own set of effects.

Mozzarella is but one of several so-calledpasta filata (Italian for stretched curd) types ofcheese. Another well-known variety is Pro-volone cheese.This cheese is made essentiallyas for Mozzarella except that a lipase prepara-tion is added to the milk to promote lipid hy-drolysis and release of free fatty acids. Lipasesare either fungal-derived or are from animal tis-sues. A slightly rancid but pleasant flavor, de-scribed as piquant, develops during a short ag-ing period. Provolone is often shaped intopear-shaped balls (prova is Italian for ball), andis frequently smoked.

Hard Italian Cheese

Given the popularity of Mozzarella cheese andItalian cuisine, in general, it is not surprisingthat production of other Italian cheeses, suchas Parmesan, Romano, and other hard gratingtypes, have increased by more than 120% inthe last twenty years.Although Parmesan is theprototype of the hard Italian grating typecheeses, there are many variations of thischeese, including Romano,Asiago, and Grana.

As recently as 2002, the Food and Drug Ad-ministration’s official Standard of Identity forParmesan was changed to accommodatenewer methods of manufacturing, such thatless aging was required. In contrast, Parmesanmade in Italy must conform to rigid manufac-turing procedures if it is to be called Parmi-giano Reggiano.The milk, for example, must beobtained from cows raised in a specific region

(the Po Valley) and fed a specified diet. Onlyraw milk can be used, and it must be standard-ized in a specific manner (which includes anatural creaming step). Calf rennet is used forcoagulation. The cheese is ultimately aged forat least twelve months, but usually longer (twoto three years). Despite these considerablechallenges, annual production of ParmigianoReggiano is still rather high (about 100 millionkg); however, even in Italy it is expensive (of-ten more than $18 per Kg). Recently, the Euro-pean Union ruled that only cheese made fol-lowing these prescribed procedures could becalled Parmesan. Other famous Europeancheeses, including Roquefort and Feta, have re-ceived similar EU protections (the so calledDesignation of Protected Origin or DPO).

In the United States, where manufacturersare not bound by EU rules, Parmesan is madefrom cow milk standardized to about 2.4% to2.8% fat. The milk is inoculated with a ther-mophilic culture and rennet is added. Thecurds are cut to give small curds, similar toSwiss, cooked to 45 to 48°C, and stirred for upto forty-five minutes. Some acid developmentduring the stir-out step occurs, and when thepH reaches about 5.9, the whey is drained.Thecurds are placed into forms and pressed fortwelve to sixteen hours,during which time thethermophilic culture ferments the lactose.Next, the cheese is brined for two weeks anddried for up to six weeks (to obtain a moisturecontent of 32%). Aging is essential for Parme-san cheese, and until recently, ten months ofaging was required to satisfy the U.S. Standardsof Identity (and to produce good flavor).

In 1999, Kraft successfully petitioned theFDA to permit cheese aged for a minimum ofsix months, instead of the nine months, to becalled Parmesan, provided the original charac-teristics and properties were still maintained.Apparently, the use of enzyme technologies,combined with a slight increase in ripeningtemperatures,can produce a cheese that,whilenot intended to be the quality equivalent of atraditional,well aged Parmesan, still has a cleanand acceptable flavor that is quite suitable as agrating cheese. One of the more popular varia-


Cheese 181

tions of Parmesan,Romano, is made in a similarmanner as Parmesan,but it contains lipase and,at least in Italy, is made from sheep’s milk.

Dutch-type Cheeses

Among the most pleasant-tasting, colorful, andmicrobiologically complex cheeses are the so-called Dutch-type cheeses, of which Edam andGouda are the most well-known. In fact, thereare comparable cheeses produced throughoutthe world, using similar manufacturing proce-dures. Although eyes are usually present inthese cheeses, there are fewer of them and theyare much smaller than in Swiss cheese.The tex-ture and flavor is also completely different fromSwiss—these cheeses are softer, with a sweet,mild, buttery flavor.They are also easily distin-guished from other cheeses by the characteris-tic red wax that covers the round wheels.

Although the basic manufacturing proce-dures for these cheeses are similar to those al-ready described for other cheeses, there areseveral unique steps. The starter culture con-tains, in addition to acid-forming lactococci,one or more flavor-forming lactic acid bacteria.The latter include various species of Leuco-nostoc, including Leuconostoc mesenteroidessubsp. cremoris and Leuconostoc lactis. Notonly are these organisms heterofermentative,producing lactic acid, acetic acid, ethanol, andcarbon dioxide, but they are also capable offermenting citrate. Fermentation of citrate bythese bacteria results in formation of diacetyl(see Chapter 2), which imparts the buttery,creamy flavor that characterizes these cheeses.In addition, the citrate fermentation pathwaygenerates carbon dioxide, accounting for theeyes that are often present.

After culture and rennet are added to themilk (held at about 30°C to 32°C), the curdsare then cut and stirred. Next, a unique manu-facturing step occurs.A portion of the whey isdrained, and then hot water is added to raisethe temperature to 37°C to 38°C and to verynear the original volume.Stirring continues un-til the desired curd firmness is achieved, andthe remaining whey-water mixture is drained.

This washing step reduces the lactose concen-tration in the curd,making less available for thelactic culture. However, by using hot water, themoisture content is not increased. Ultimately,the cheese will be sweeter, and less acidic,with a final pH between 5.4 and 5.6.

Surface-ripened by Bacteria

Among the cheeses that are the most aromaticand flavorful, Limburger, Muenster, Brick, andother surface-ripened cheeses are right at thetop of the list.A well-ripened Limburger can fillthe room with the strong sulfury volatile com-pounds that are produced by Brevibacteriumlinens.This organism (and others that are gen-erally present; see below) is neither added tothe milk nor curds, but rather is applied to thecheese after its manufacture.As profound as isthe aroma and flavor of this cheese, its orange-red appearance is nearly as dramatic. In fact,the natural pigment produced by B. linens thataccounts for the color of these cheeses has re-cently found use as a coloring agent in otherfood products.

The reader is probably wondering why theMuenster or Brick cheese common to U.S. con-sumers lacks the punch described above, de-spite having the expected appearance.This isbecause the early manufacturers of thesecheeses, who typically had emigrated fromGermany and other European countries wherethese cheese were popular, realized that U.S.consumers preferred more mild-flavoredcheeses.Thus, the manufacturing process hadto be modified to accommodate the particularpreferences of the new American customers.Thus was born the mild-flavored Muenster andBrick cheeses that are so widely consumed to-day. Only Limburger has retained the tradi-tional flavor and appearance properties of theoriginal. Still, the preference for the mild con-tinues even today—Limburger production inthe United States in 2001 was less than 50% ofthat in 1980,and Brick and Muenster continuesto outsell Limburger by more than 100-fold.

Limburger and related cheeses are initiallymade according to rather standard procedures.



Whole milk is inoculated with a mesophilic lac-tic culture (L. lactis subsp. lactis) and chymosinis added.The coagulated milk is cut and stirred,but the curds are cooked to more moderatetemperatures, usually between 30°C and 34°C.The curds are collected into open-ended,brick-shaped forms, which are turned every three tofive hours to promote whey drainage. Afterabout twelve hours, the cheese is brined or drysalted at the surface and held in warm (20°C)and humid (90% relative humidity) rooms.

Traditional manufacturers may rely on theendogenous flora that is present on the shelvesin the ripening room to initiate surface ripen-ing. In addition to B. linens, micrococci, yeast,and other organisms are likely present and willparticipate in the ripening process by raisingthe pH and producing flavor and color com-pounds. The presence of Corynebacterium,Arthrobacter, and other coryneform bacteriain these cheeses is common, although theirrole in flavor and pigment production is notyet established. Alternatively, a pure B. linensculture can be applied. Under these warm, hu-mid, and aerobic conditions, ripening does nottake long, and there is considerable surfacegrowth, color formation, and flavor develop-ment within just a couple of weeks.The cheeseis packaged and moved into a 10°C cooler foranother month or two.These cheeses can eas-ily become over-ripened, so they must be keptat low temperature. The packaging materialsusually include parchment, wax paper, and foilto minimize oxygen availability (and to containvolatile aroma).

The strong aroma and flavor of Limburgerand related cheeses is due to production ofseveral volatile compounds. In defense of thesecheeses, it is fair to say that the bark is worsethan the bite, in that the flavor is nowhere nearas strong as the aroma might portend. In fact,the use of surface smears in cheese making hasfast become popular in the United States, espe-cially among small, farmstead or artisan-typecheese makers. Thus, domestic versions ofTilsiter, Raclette, Beaufort, Gruyere, and othersimilarly-produced cheeses are now morewidely available. In July, 2005, a Wisconsin-

made Beaufort style cheese received the bestin show award at the American Cheese Societycompetition, besting nearly 750 other cheeses.

Mold-ripened Cheese

Aside from the fact that the two main mold-ripened cheeses, the blue-type and the Brie-type, are both fungal fermentations, they sharefew common properties. The blue moldcheeses contain visible mold growth through-out, at the surface, and within the curd. Themold responsible, Penicillium roqueforti, pro-duces blue-green mycelia, in addition to a myr-iad of enzymes that ultimately generate typicalblue cheese flavors. In contrast, the Brie-typecheeses are made using Penicillium camem-berti, which produces white mycelia, andgrows only at the surface.Blue cheese is acidic,salty, brittle and crumbly, while the Brie-typecheeses are satiny smooth, soft, and creamy,with a near neutral pH.

Blue-mold Ripened Cheese

Although blue mold-ripened cheeses are madethroughout the world, three specific typeshave achieved a significant measure of fame towarrant their own name (and have DPO sta-tus). Roquefort, perhaps the most well-knownof all blue cheeses, must be made according toa strict set of manufacturing requirements. Forexample, the milk must come from specially-bred sheep that have grazed in the Causses re-gion of France. It is neither pasteurized, stan-dardized, nor homogenized, and the cheesemust ultimately be aged in caves within thatsame region. In Italy, the manufacture of Gor-gonzola cheese is similarly restricted to the PoValley region of Northern Italy and its manu-facture is also subject to specified procedures.This cheese is made from cow’s milk, but isotherwise very similar to Roquefort. Gor-gonzola, however, is usually not quite asstrongly flavored as Roquefort. Finally, the Eng-lish representative to the blue-mold family iscalled Stilton. It also is made from cow’s milkand only in three counties of central England.


Cheese 183

Other blue mold-ripened cheeses are ordinar-ily referred to simply as blue cheese, althoughregional varieties, often times of excellent sin-gular quality, exist worldwide.

In general, the manufacture of blue mold-ripened cheese requires several specializedsteps. In large part, the goal is to ensure thatthe aerobic mold can grow well within the in-terior of the cheese, where the Eh is ordinarilylow.Although the blue mold organism,P.roque-forti, is capable of growing at high CO2 andlow O2 levels, air incorporation is still an im-portant feature of the process.The coagulatedmilk is cut, when very firm, into larger thannormal-sized curds (as much as 2.5 cm). Asnoted earlier, the larger the curds at cutting,the higher the moisture will be in the cheese.However, an additional effect is to create amore open or porous texture such that diffu-sion of air is increased.

Frequently, the mesophilic lactic starter cul-ture, in addition to containing L. lactis,will alsoinclude heterofermentative Leuconostoc or citrate-fermenting lactic acid bacteria. Thesebacteria produce CO2 that contributes to theopen texture that enhances oxygen diffusionand gas exchange. Spores of P. roquefortii canbe added to the curds or to the milk, alongwith the lactic starter culture, prior to setting.Later, once the cheese is hooped and the lac-tose fermentation is complete, the cheesewheels are brine- or dry-salted.At this point, aset of large bore needles (diameter about 0.24cm) may be used to pierce the cheese and pro-vide a means for air to penetrate the inner por-tions. Later, when the cheese is cut vertically,one can see that mold growth followed thespike lines where oxygen was available.

Next, the cheese is aged for several weeks ina warm (10°C to 12°C), humid (90% to 95% rel-ative humidity) environment that promotesgrowth of the P.roqueforti throughout the inte-rior of the cheese. Lower ripening tempera-tures (4°C to 8°C) and longer times may also beused. Since the only way to arrest further moldgrowth is to cut off its supply of oxygen, theripened cheeses are then wrapped in oxygen-impermeable foil. Although some consumers

like blue cheese flavor, they may not be fond ofthe moldy appearance; thus, some manufactur-ers may limit growth of P. roqueforti, giving asomewhat whiter cheese with only a fewstreaks of blue. Even consumers who enjoyblue veined cheese may object to extensivegrowth at the surface, so the wheels are oftencleaned prior to final packaging. Pimaricin mayalso be applied to control surface growth.

Although other organisms, including yeastand bacteria, are often present in these cheeses,especially at the surface, the flavor compoundscharacteristic of blue cheese clearly are gener-ated primarily via mold growth and metabolism.Fungi, in general, are prolific producers of pro-teases, peptidases, and lipases, and P. roquefortiis no exception. Thus, release and subsequentmetabolism of protein hydrolysis and lipolysisproducts are important in blue cheese flavor.Ofparticular importance are ammonia and amines,derived from amino acid metabolism, andmethyl ketones,derived from free fatty acids.

Free fatty acids, themselves, may also con-tribute to cheese flavor, but their metabolism,via �-oxidation pathways to 2-heptanone, 2-nonanone, and other ketones, are primarily re-sponsible for the flavor of blue cheese.As muchas 20% of the triglycerides in the milk may behydrolyzed.When blue cheese is made from rawmilk,natural milk lipases may also contribute toformation of free fatty acids. These lipases areloosely associated with the surface of the caseinmicelles and are dislodged by agitation. In addi-tion, lipases’ substrates (triglycerides) are ex-posed when the fat globule is disrupted.There-fore, some manufacturers add a portion ofhomogenized raw cream to the cheese milk toaccelerate flavor development.

Finally, it is important to note that moldgrowth during blue cheese ripening is accom-panied not just by flavor development,but alsoby a marked increase in pH. This occurs be-cause P. roqueforti, not having lactose onwhich to grow (all of the lactose is fermentedby the starter culture), instead uses lactic acidas an energy source.Therefore,consumption oflactic acid causes the pH to rise from about 4.6to above 6.0.Production of ammonia and other



amines also contributes to the increase in pH.This increase in pH, in turn, promotes flavorproduction because many of the fungal decar-boxylases and other enzymes involved inmethyl ketone formation have neutral pH op-tima. However, the rise of pH to near neutralitymay also have food safety and preservationconsequences in that acid-sensitive organismsmay be able to grow once the pH reaches non-inhibitory levels.

White-mold Ripened Cheese

The white mold-ripened cheeses, of whichCamembert and Brie are the most well-known,are primarily made in France, where they arealso among the most popular. These cheesesvary only slightly; Camembert is made princi-pally in the Normandy region of France,whereas most Brie is produced in Melun andMeaux, just outside of Paris. Brie wheels areusually a bit larger,with bacteria on the surfacecontributing to flavor development.

Although similar versions of both Brie andCamembert are made in Germany, Switzer-land, and other European countries, as well asSouth America and even the United States,manufacturing conditions are not that differ-ent. In general, whole cow’s milk is used, anddouble or even triple cream versions exist inwhich the milk is supplemented with cream(not for the faint of heart). Since these cheesesare aged for as little as a few weeks, in theUnited States, the milk must be pasteurized(Box 5–4). In France, however, raw milk ismost frequently used. The milk is inoculatedwith a mesophilic starter culture, which is al-lowed to grow and produce acid before thechymosin is added. Spores of P. camembertican also be added to the milk (or applied laterto the surface of the hooped cheese).A relatedspecies, Penicillium caseicolum is also used,because it produces a whiter mycelia.

As with blue cheese, the coagulum is firmwhen cut into large curds particles.The curdsare gently stirred, but not cooked. Rather, theyare almost immediately filled into forms.Thesesteps decrease syneresis and result in cheese

with high moisture levels.The hooped cheesesare flipped several times over an eighteen-hour period, during which time the lactose isfermented and the pH decreases to as low as4.7. As with blue cheese, the cheese is thenbrined or dry-salted, and, if not already addedto the milk, spores are applied, but only ontothe surface.

The cheese wheels are then moved intoripening rooms (similar to those used for bluecheese, e.g., 10°C, high humidity).The cheesesits on shelves designed to promote contactwith air, and are turned periodically. It takesonly a week or two for a white mycelium matto form across the entire surface. However, theextent of mold growth and subsequent flavordevelopment depends on the target market, sothe ripening time may be several weeks longer.In cheese made from raw milk, a diverse mi-croflora, consisting of yeasts, brevibacteria, en-terococci and staphylococci, and coliforms,may also emerge during extended ripening.Thus, not only will there be more flavor andtexture development (see below) as ripeningcontinues, but the surface mat may becomeless white and more pigmented. In the UnitedStates, a more mild flavored, white-mattedproduct is preferred, so ripening times are usu-ally short. To control growth of mold, thecheese is wrapped and either held for moreripening or else stored at �4°C.

Not only do the Brie-type cheeses have a dif-ferent appearance from the blue mold cheeses,but they also have a quite different flavor andtexture. Despite these differences, the progres-sion of flavor and texture development is simi-lar. The proteinases and peptidases producedby P.camemberti are similar to those producedby P. roqueforti. Subsequent production of am-monia, methanethiol, and other sulfur com-pounds are derived from amino acid metabo-lism and are characteristic of Brie-type cheese.The scent of ammonia can be striking in a well-aged Brie. Lipolysis of triglycerides and fattyacid metabolism by P. camemberti are also im-portant, and methyl ketones can be abundant.

It is interesting that whereas blue cheese iscrumbly and brittle, Brie-type cheeses are soft


Cheese 185

and creamy. The creamy, even fluid texture ofthese cheeses is now thought to occur as a re-sult of protein hydrolysis, as well as the in-crease in pH due to ammonia. In particular, notonly are fungal proteases important in texturedevelopment, but �s1 casein hydrolysis by chy-mosin and the natural milk protease, plasmin,is also involved.

For many of the cheeses discussed in thischapter, the geometry of the forms or hoopsthat give shape to the cheese have not beendescribed in much detail. In many cases,whether the cheese is collected and shapedinto rectangular blocks or round forms orwhether the hoops or forms contain 5 Kg or200 Kg has only modest bearing on the prop-erties of the cheese.This is most definitely notthe case for Brie-type cheese, where shape hasa major impact on flavor and texture develop-ment. Since ripening depends on the fungigrowing exclusively on the surface, the rate offlavor and texture development within the in-terior of the cheese is necessarily a function ofthe diffusion rate of enzymes and enzyme reac-tants and products from the exterior.

Although Camembert wheels usually have asmaller diameter (about 10 cm to 12 cm) thanBrie (about 30 cm to 32 cm), their heights areessentially the same (about 3.5 cm) and thegeometric center of these cheeses is nevermore than 2 cm from the surface.Thus, the en-zymes and reactants (e.g., ammonia) producedat the surface have only a short distance totravel before reaching the center. If Brie cheesewere to be made in thicker wheels, like thoseused for blue cheese (e.g., 15 cm to 20 cm di-ameter), it would obviously take a long time forthe reactants to diffuse into the center region.During that period,the exterior portions wouldhave been exposed to those enzymes far toolong.The cheese might be perfectly ripened inthe center,but well over-ripened at the surface.

Finally, as with P. roqueforti, growth of P.camemberti in the manufacture of Brie-typecheeses depends on lactic acid as an energysource.The subsequent rise in pH (from 4.6 toas high as pH 7.0 at the surface) is similarly dueto lactate consumption and ammonia produc-

tion. This return to a neutral pH also con-tributes to the diversity of microorganisms thatgrow on the surface of these cheeses (seeabove) and, more importantly, poses a specialfood safety risk due to the possible presence ofListeria monocytogenes (see Box 5–4).

Pickled Cheese

High-salt, high-acid cheeses were likely amongthe first cheeses intentionally produced by hu-mans. They would have had a much longershelf-life than other similarly produced softcheeses that had lower salt content. Thesecheeses have long been popular in Greece,Turkey, Egypt, Israel, and other Middle Easterncountries, as well as the Balkan region of East-ern Europe. In the past twenty years, produc-tion and consumption of these cheeses hasspread throughout Europe, the United King-dom, the United States, New Zealand, and Aus-tralia.The most popular cheese in this categoryis undoubtedly Feta cheese, which has its ori-gins in Greece. Other similar types exist andvary based mostly on salting method (whetherapplied before or after curd formation). Al-though they are often packaged dry in theUnited States, it is more common for retail Feta-type cheeses to be packaged in tubs contain-ing salt brine.

The general manufacturing procedures forFeta-type cheese start with a pasteurizationstep, since the cheese is usually consumedfresh (aged for as little as a few weeks). InGreece and other European countries, raw milkis often used. The milk itself can be obtainedfrom cows, sheep, or goats. In many modern fa-cilities, ultrafiltered milk is used.The milk is in-oculated with a mesophilic starter containingL. lactis subsp. lactis and L. lactis subsp. cre-moris. The culture is added at the rate of 2%and allowed to ripen in the milk for as long astwo hours prior to addition of chymosin. Thelarge inoculum and long ripening time resultsin considerable acid development and a de-crease in pH even before chymosin is added.

A rennet paste, a crude preparation contain-ing a mixture of enzymes, is used in traditional



Feta manufacture. Other proteases as well as li-pases are among the enzymes present in thispaste.The latter can also be added separately,in the form of a lipase extract obtained fromanimal or microbial sources. The coagulationtime can be an hour or longer.The coagulumwill be very firm at cutting.The large curds (2cm) are stirred for only a short time before be-ing dipped and filled into hoops, which areflipped and turned for several hours.Fermenta-tion occurs quickly and the pH will reach 4.7or lower within eight hours.The cheese is theneither placed into a 18% salt brine for twelve totwenty-four hours, or dry salted for one tothree days.The salted cheeses are then placedin a 10% brine and held until packaging anddistribution.

The Feta-type cheeses have a very short,crumbly texture, due, in part, to low pH (�4.7)and high salt.As the pH approaches the isoelec-tric point of casein (pH 4.6), less water is re-tained. Also, high brine concentrations causethe casein network to shrink, releasing evenmore water from the cheese. Both of these fac-

tors make Feta crumbly.The main flavors in Fetaare due to lactic acid and salt, but, dependingon the presence of lipases (exogenous or frommilk), there can also be considerable hydrolysisof triglycerides and formation of free fattyacids. It is the release of short, volatile fattyacids, including acetic, propionic, butyric, andvaleric acids, that account for the rancid-like fla-vor notes characteristic of these cheeses.

Processed and Cold Pack Cheese

Although cheese is the main ingredient, pro-cessed cheese (or what many consumers mis-takenly call American cheese) is not a fer-mented food. However, because processedcheese is so popular in the United States andthere is so much confusion regarding the differ-ences between natural cheese and processedcheese (Box 5–6), it is worthwhile to describethe manufacture of processed cheese.

Briefly, processed cheese is made by addingemulsifying salts to natural cheese, along withwater and other dairy and non-dairy ingredi-

Box 5–6. “American” Cheese

Ask any American kid (even college kids) what kind of cheese he or she prefers and the likelyanswer will be “American”or perhaps even a popular brand name.What they are really referringto, in most cases,at least, is one of several types of processed cheese products. In fact,more thanhalf of the cheese now purchased at retail (i.e., from supermarkets) is processed cheese. In con-trast, the term “American” has no legal meaning, but is simply an adjective that describes thosenatural cheeses belonging to the Cheddar-style family.

The processed cheese category consists of three general types of products, each with a spe-cific standard of identity that describes in detail the types of ingredients that are allowed, how itis to be manufactured, and the gross composition of the finished product.Pasteurized processedcheese is made from a single type of cheese and is labeled accordingly. Thus, pasteurizedprocessed American cheese is made from Cheddar or other American-type cheese. It contains amaximum of 43% moisture and a minimum of 27% milkfat. Pasteurized processed cheese foodmay contain additional optional ingredients, and more water and less fat are allowed (44% and23%,respectively).Pasteurized processed cheese spread contains even more water (44% to 60%),less fat (20%), and stabilizers are allowed. Many of the loaf-type products fit this description.

Not only do these products provide consumers with convenience and functionality, they canbe considered an excellent example of value-added food technology. Cheese is, after all, a com-modity whose dollar value is rather modest.Aging can certainly increase its value, but wholesaleblock and barrel prices for Cheddar cheese in the United States during 2004 averaged only about$1.50 per pound.Considering that it may cost the cheese manufacturer as much as $1.30 just forthe milk that goes into a pound of cheese, earning a reasonable profit making commodity-stylecheese is a real challenge.


Cheese 187

ents.The mixture is then agitated while beingheated to about 70°C or higher.The emulsifica-tion that occurs in processed cheese is differentthan other true food emulsions. In the case ofprocess cheese, sodium or potassium polyphos-phate and citrate emulsifying salts raise the pHand displace calcium ions from the casein com-plex, resulting in more soluble sodium casein.The latter contains both lipophilic and hy-drophilic regions that then forms an emulsion-like mixture with the lipid portion of thecheese.After heating,the cooled mixture is thenformed into slices or loaves or filled into jars orcans.The finished products have excellent func-tionality and convenience features. Due to theheat treatment and other inherent conditions(pH, water activity, antimycotic agents), theseproducts also have long shelf-lives,even at ambi-ent temperature.

The quality of the cheese used as the start-ing material is among the factors that influenceprocessed cheese manufacture.A typical blendconsists of 15% aged and 85% young (or cur-rent) cheese.The aged cheese provides flavor;the young cheese provides elasticity, body, andemulsifying properties. Too much of the for-mer may cause a soft, soupy body, too much ofthe latter may lead to hard body and brownpigment formation. Cold pack cheese, in con-trast, is made by mixing or grinding differenttypes of natural cheese in the absence of heat.Various optional ingredients, including color,spices, and other flavoring agents, as well as an-timycotic agents, can also be added. Theseproducts have excellent flavor, but typicallylack the functionality of process cheeses.

Cheese Ripening

Freshly made cheese has essentially none of theflavor,aroma,rheological,or appearance proper-ties of aged or ripened cheese.Rather, the meta-morphosis from a bland, pale, rubbery mass ofprotein and fat into a flavorful, textured fusionof complex substances takes time and requirespatience.Although efforts to reduce aging timeand accelerate the ripening process have beensomewhat successful (see below), for the most

part, cheese ripening is a sequential process,with each step relying on a preceding step. Inother words, a particular flavor compound maybe present in a cheese only as a result of severalpreceding metabolic steps.

For example, methyl ketones, which con-tribute to the characteristic flavor of bluecheese, are synthesized by P. roqueforti fromfree fatty acids.Thus, methyl ketone formationdepends on release of these acids from triglyc-erides via lipolytic enzymes produced by mi-croorganisms or naturally present in the milk.Similarly, hydrogen sulfide, which is an impor-tant flavor note in aged Cheddar cheese (ormay also be a defect; see below), is derivedfrom sulfur-containing amino acids that formvia protein and peptide hydrolysis.

Of course, not only is ripening a sequentialprocess, but it also is subject to potentialchaos and disarray, such that the cheese mayripen poorly or unexpectedly. In some cases,over-production of an otherwise desirablecompound occurs, resulting in serious flavordefects. For example, hydrogen sulfide, at con-centrations in the ppb range, imparts a pleas-ant aroma in Cheddar cheese, but when pre-sent at ppm levels, the cheese is nearlyinedible. In contrast, while blue cheese maycontain 90 ppm to 100 ppm of methyl ke-tones, a small amount (1 ppm to 2 ppm) isperfectly fine in Cheddar cheese. Controllingthe ripening process, then, is key to success-ful and consistent production of aged andwell-ripened cheese.

Many factors contribute to the ripeningprocess, including live microorganisms, deadmicroorganisms, enzymes, and chemical andphysical reactions.Manipulating and controllingthese activities and reactions depends on char-acteristics intrinsic to the cheese, such as mois-ture, pH, salt, and Eh, as well as those extrinsicfactors that are influenced by the cheese manu-facturer.The latter include the source and han-dling of the milk, the temperature and humidityof the ripening room or environment,and othermanipulations performed by the manufacturer.It is the collective result of these events that dic-tate the properties of the ripened cheese.



As noted above, cheese contains mostly wa-ter, protein, and fat as it leaves the vat andmoves into the ripening coolers. Althoughsome protein hydrolysis certainly occurs dur-ing the lactic fermentation by starter culturebacteria and by the action of milk proteasesand the coagulant, �s1, �s2, and �-caseins arestill intact. Similarly, the triglycerides in milkare also mostly unaffected by the early cheese-making steps. However, within just a few days,enzymes begin to attack these substrates andinitiate the ripening process.

Proteolysis in cheese

It was long argued that milkfat was the primaryconstituent responsible for cheese flavor.Whileit is certainly true that many cheese flavors areeither evolved from the lipid fraction or are sol-uble in the lipid phase, it is now generally ac-cepted that, for most cheeses (the main excep-tion being the blue mold-type cheeses), it is theprotein fraction that makes the more importantoverall contribution (Table 5–2). Cheeses madeunder controlled conditions in which proteoly-sis does not occur develop neither the flavornor texture of a normal cheese.The recognitionthat proteolysis has such a profound influenceon cheese flavor and ripening has led to a de-tailed understanding of many of the specificsteps involved in the degradation of milk pro-teins and the metabolism of the hydrolysis prod-ucts (Figure 5–9).

For the most part, the relevant organisms re-sponsible for these activities are lactic acidbacteria, either those added as part of thestarter culture or those present as part of theordinary microflora. Lactic acid bacteria can-not synthesize amino acids from ammonia andinstead require pre-formed amino acids forprotein biosynthesis and cell growth. How-ever, milk contains only a small amount of freeamino acids (�300 mg/L), which are quicklyassimilated.Thus, growth of lactic acid bacteriain milk depends on the ability of these cells tometabolize proteins in milk. The means bywhich large casein molecules are brokendown by these bacteria into their componentamino acids (and beyond) consists of fourmain steps: (1) hydrolysis of casein; (2) trans-port of casein-derived amino acids, peptides,and oligopeptides into the cytoplasm of starterand non-starter lactic acid bacteria; (3) intracel-lular hydrolysis of accumulated peptides; and(4) metabolism of amino acids.

In Cheddar-type cheeses, starter culture lac-tic acid bacteria, notably, L. lactis subsp. lactis,produce a cell-wall anchored proteinase thathydrolyzes specific casein subunits (Box 5–1),producing as many as 100 or more differentpeptides varying in length from four to thirtyamino acid residues.Although proteinases areproduced by several strains of L. lactis subsp.lactis and other lactic acid bacteria, the en-zymes are structurally quite similar (in termsof their DNA and amino acids sequences).

Table 5.2. Sources of flavor compounds in cheese1.

Protein (casein) Carbohydrate Lipid

peptides lactate fatty acidsamino acids acetate keto acidssulfur compounds pyruvate estersammonia and amines ethanol methyl ketonespyruvate diacetyl lactonesacetate acetoinaldehydes 2,3-butanediolalcohols acetaldehydeketo acids

1Adapted from McSweeney, 2004 and Singh et al., 2003


Cheese 189

However, the specific casein substrates andhydrolysis products of the different PrtP en-zymes can vary considerably.

For example, some proteinases hydrolyze�S1-, �-, and �-caseins, whereas others havepreference for �-casein and relatively little ac-tivity for �S1- and �-caseins. It should be notedthat residual chymosin and milk proteinases(mainly plasmin) may also contribute to thepeptide pool, especially during the early stagesof cheese ripening. However, it is the starterand non-starter lactic acid bacteria that are re-sponsible for most of the subsequent caseinhydrolysis. In fact, when cheese is made with

starter strains that do not produce cell wallproteinase, casein degradation occurs to a verylimited extent (and cheese flavor and texturedo not develop). Likewise, cheese made in theabsence of non-starter lactic acid bacteria issimilarly bland.

Hydrolysis of the casein-derived peptideswas previously thought to occur by extracellu-lar peptidases produced by starter or non-starter lactic acid bacteria.It now is generally ac-cepted that peptide hydrolysis does not occuroutside the cells, but rather the peptides aretransported via various peptide transport sys-tems and are subsequently hydrolyzed by intra-

casein

oligopeptides

largepeptides

amino acidsdi/tripeptides oligopeptides

chymosinplasmin

proteinase

Dtp AA Opp Opp

oligopeptidesamino acidsdi/tripeptides

amino acids

amines

keto acids

carbonyls

sulfur compounds

peptidases

CO2NH3

Figure 5–9. Proteolysis during cheese ripening. The proteolytic system in lactococci starts with the hydroly-sis of casein by a cell envelope-associated proteinase (PrtP). The main products are oligopeptides, which arethen transported across the cell membrane by the oligopeptide transport system (Opp). Any free amino acidsand di- and tripeptides in the milk are similarly transported by amino acid (AA) and di- and tripeptide trans-porters (DtpT, DtpP), respectively. Once inside the cell, the oligopeptides are hydrolyzed by various peptidases,including PepA, PepC, PepN, and PepX. The di- and tripeptides are also hydrolyzed. The free amino acids arethen metabolized or used for protein biosynthesis (not shown). Adapted from McSweeney, 2004, andHutkins, 2001.



cellular peptidases.The main transporter, calledthe oligopeptide transport system (OPP), is pre-sent in most starter lactic acid bacteria andtransports oligopeptides across the cytoplasmicmembrane and into the cell. Mutants defectivein OPP, like mutants defective in proteinase pro-duction, grow poorly in milk because they areunable to use casein as a nitrogen source.Starterlactococci also have amino acid as well as di-and tri-peptide transport systems;however, theyare somewhat less important since most of thepeptides present in cheese are larger than threeamino acid residues.

Once inside the cells, the peptides can behydrolyzed by a variety of peptidases, depend-ing on their size and amino acid content.Aminopeptidases hydrolyze peptides at theamino end (i.e., the peptide bond located be-tween the amino terminal amino acid residueand the penultimate residue), whereas car-boxypeptides act at the carboxy end.Endopep-tidases also exist that hydrolyze peptide bondswithin a peptide. Lactococcal peptidases alsohave specificity with regard to the amino acidsand their position within a given peptide bond.For example, aminopeptidase P (PepP) from L.lactis subsp. lactis hydrolyzes peptides contain-ing proline at the penultimate residue.

Protein hydrolysis also has a major impacton texture.The solubility of coagulated caseinis limited, and is even less at low pH, makingyoung cheese plastic and rubbery. However, ascheese ages and casein is degraded, the pep-tides are more soluble, and elasticity increases.Eventually, however, when the casein and casein-derived peptides are more extensivelydegraded, the cheese develops a crumbly, shorttexture.

As noted above, cheese flavor and texturedevelopment is largely a function of the prote-olytic capacity of the bacteria present in theripening cheese. However, for the most part,the starter culture bacteria are not maintainedvery long after the first few weeks of aging.That is not to say that their impact is not signif-icant. In fact, as the starter bacteria lyse (due tosalt, low water activity, lack of fermentable sub-strates, and other factors), their entire array of

intracellular enzymes are released into thecurd matrix. Thus, the various peptidases im-portant for peptide degradation are free to hy-drolyze available substrates. Likewise, enzymesthat act directly on amino acids, triglycerides,and other substrates are also released, as arethe substrates themselves.

Finally,non-starter lactic acid bacteria,as wellas other indigenous bacteria, play an important(perhaps the single most important) role incheese ripening. In particular, several Lacto-bacillus spp., including Lactobacillus casei,Lactobacillus paracasei, and Lactobacillusplantarum, increase in number (from 102 pergram to more than 107) during aging, and pro-duce peptidases and other enzymes that gener-ate aged cheese flavor. Of course, some of theseadventitious organisms can also produce off-flavors.Heterofermentative lactobacilli are espe-cially of concern due to their ability to produceacetic acid and CO2 from residual sugars.

Current Issues in Cheese Technology

Given the important role of microorganisms inthe manufacture of cheese, it should come asno surprise that among the most important is-sues faced by the cheese industry, most are mi-crobiological in nature (not withstandingyield, costs, and other economic issues). Thekey to making any high quality fermentedfood, including cheese, is to control the activi-ties of the microbial starter culture and otherorganisms.This means ensuring that they growwhen they are supposed to, that agents thatwould otherwise interfere with their growthare controlled, and that they produce the cor-rect amount and types of enzymes, flavors, andother products.

Although consumer trends often influenceand raise specific sorts of problems, severallong-term issues continue to challenge cheesetechnologists.These questions include: (1) howto improve and accelerate cheese flavor devel-opment and reduce bitterness,especially in low-fat cheese; (2) how to control bacteriophagesthat infect, inhibit, and sometimes inactivatestarter cultures; (3) how to enhance shelf-life


Cheese 191

and prevent growth of spoilage microorganismsand ensure that cheese is free of potential path-ogenic microorganisms;and (4) how to increasethe value of whey using microorganisms.

Bitterness and accelerated ripening

For most cheeses, production of good agedcheese flavor and texture requires that there besignificant hydrolysis of casein. At the sametime, the main flavor defect in cheese is bitter-ness, which is also caused by casein hydrolysisand subsequent formation of bitter peptides.Thus,aging cheese is like walking a tightrope—

it requires a fine sense of preparation, balance,and proportion.A single misstep may lead to anunacceptable end. Accelerated ripening—ob-taining good cheese flavor and texture in halfor even a third of the time—is, to continue theanalogy, like running across the tightrope.There is tremendous economic incentive, how-ever, to decrease aging time, because the agingprocess adds about one cent per Kg per monthto the cost of the cheese. Reduced fat cheeseposes another set of flavor problems (Box 5–7).Not surprisingly, considerable research hasbeen aimed at developing accelerated cheeseaging programs.

Box 5–7. Reduced Fat Cheese—Taking the Fat (But Not the Flavor) Out of Cheese

Cheese, like other dairy products, is heavily regulated by the federal government, and standardsof identity exist for most of the cheeses consumed in the United States (Figure 1).These stan-dards (contained within the Code of Federal Regulations,Title 21, Part 133) not only describethe ingredients allowed in the manufacture of various cheeses, they specifically spell out the fatand compositional requirements for them. Importantly, the minimum amount of fat (on a dry orsolids basis) and maximum amount of water are specified. Until relatively recently, Cheddarcheese that contained less than 50% fat on a dry basis (or in the dry matter or FDM) could notbe called cheese. Similar standards exist for other cheeses, such that most contain at least 45%fat and derive more than 75% of their calories from fat.

Because milkfat was viewed, for hundreds of years, as a precious and desirable commodity,these standards were established to protect consumers and make sure they received full valuefor their money.Products labeled as cheese,but containing less than the required amount,wereconsidered inferior and in violation of the law. This public perception of milkfat, however,

Reduced fat = 25% reduction in total fat per reference amountwhen compared to an appropriate reference food

Light = food that has 50% less fat per reference amount whencompared to an appropriate reference food, when 50% or moreof its calories are from fat, or in the case of food containing lessthan 50% of its calories from fat, less than one-third lesscalories from fat or less than 50% fat per reference amountwhen compared to an appropriate reference food.

Nonfat = Less than 0.5% fat per reference amount whencompared to an appropriate reference food

Lowfat = Three grams or less fat per reference amount.

(Continued)

Figure 1. Reduced Fat Cheese Standards (Implemented 1996).



The main approach of most aging strategiesis to promote proteolytic and other enzymaticactivities.As described above and in Figure 5–9, the first step in the use of casein by the lac-tic starter culture is hydrolysis of casein by thecell wall anchored proteinase (PrtP). Of thehundred or more peptides that are produced,most contribute little to overall cheese flavor.However, several of these peptides have a bit-ter flavor. If the starter culture cannot degradethese bitter peptides, the cheese will be bitter.Starter culture strains that produce PrtP (or re-lated proteinases), but that are otherwise un-

able to degrade the bitter peptides so pro-duced, are referred to as bitter strains.

In general, bitter peptides are hydrophobicand contain proline,and are hydrolyzed only byspecific intracellular aminopeptidases, includ-ing PepN and PepX. Bitter strains lack the abil-ity to produce these enzymes. It is interestingthat most commercial strains of L. lactis subsp.lactis, the most widely used organism in thecheese industry, are bitter. In contrast, most ofthe L. lactis subsp. cremoris strains are non-bitter. However, L. lactis subsp. lactis growsfaster and, at least in unripened cheese (e.g.,

changed in the 1960s and 70s, as consumers in the United States began to reduce their con-sumption of animal fats and to seek out lower-fat products. Sales of low-fat fluid milk productsbegan to increase, and a market for other low-fat dairy products became evident.Although thecheese industry was well aware of the change in consumer attitudes and had been working onlower-fat versions, these new products could not be called cheese, due to existing standards. Itwas not until the 1990s that the U.S.government modified the definitions of cheese so that low-fat versions could still be labeled as cheese.These new standards for cheese follow the reducedfat, low-fat, and nonfat definitions used for other foods (see below).

The resolution of the labeling problem, however, did not lead to immediate consumer accep-tance of reduced fat cheese. Making cheese with low-fat milk is technically no different thanconventional cheese manufacture,but the products that were initially introduced suffered fromserious flavor and body defects.This situation has slowly begun to change, and as of the year2000, about two-thirds of consumers had purchased reduced or non-fat cheese. However, manyof the problems that plagued the initial products have yet to be resolved and the market forthese products is still relatively modest.

Making reduced fat cheese with flavor and texture properties comparable to full fat products,it turns out, is not easy, for two main reasons. First, when milkfat is taken out of cheese, it is es-sentially replaced by casein and water.Thus, the body becomes more firm and less smooth andpliable. Second, a higher moisture content means that the salt-in-moisture ratio decreases, re-sulting in an undesirable increase in microbial and enzymatic activities.All other things beingequal, the cheese ages poorly, with potential for significant production of bitterness and otheroff-flavors. More moisture also results in more lactose in the curd, which may lead to excessiveacid production. Not only does this affect flavor, but as the pH decreases, calcium is lost, whichmay cause other body defects.

Manufacturers have adopted several strategies to produce more acceptable reduced fatcheeses.First, gums and other fat mimics can be added to give the body, texture, and mouth feelof full-fat cheese.To control acidity, the curds can be washed to remove excess lactose.Alterna-tively, researchers at the University of Wisconsin developed a process, based on enhancing thebuffering capacity in the curd (by retaining calcium phosphate), that results in higher acid lev-els but moderate pH (Johnson et al., 2001).This also resulted in a lower level of calcium boundto casein and a softer body.An experimental centrifugation method in which the fat is removedafter manufacture and aging was recently described (Nelson and Barbano, 2004).

Despite these improvements,flavor,either the lack thereof or the presence of undesirable fla-vors, remains a problem.According to some researchers,milkfat not only serves as an important

Box 5–7. Reduced Fat Cheese—Taking the Fat (But Not the Flavor) Out of Cheese (Continued)


Cheese 193

cheese destined for the process kettle), its in-ability to degrade bitter peptides is not an issue.

Even cheese made with non-bitter strains, ifaged long enough, can eventually accumulateenough bitter peptides to develop bitter fla-vor. Bitter peptide production is further in-creased at higher ripening temperatures (i.e.,�12°C), which is perhaps the most commonway to accelerate cheese ripening. Althoughsome bitterness is unavoidable, the addition ofadjunct cultures, or strains capable of produc-ing aminopeptidases, has been promoted as ameans of accelerating the aging process (Box5–8). Adjunct cultures,which typically containselected strains of Lactobacillus and Lactococ-cus, can be added directly as part of the starter

culture. However, because excess acid produc-tion is to be avoided, variants that are unableto ferment lactose are preferred.

The observation that some strains of lacto-cocci and streptococci were autolytic led toanother innovative means for accelerating theripening process. As described previously,ripening depends largely on the presence ofenzymes released by bacteria as they lyse dur-ing the aging process. Ordinarily, cell lysis oc-curs only when the cells are no longer able tomaintain cell wall integrity, due to the lack ofenergy sources and the generally inhospitableconditions (i.e., low pH, high salt, etc.) presentwithin the cheese. In contrast, if lysis and en-zyme release were to occur during the early

flavor-generating substrate (i.e., fatty acids), it also performs as a solvent in cheese, since manyof the important Cheddar cheese flavor compounds are found in the lipid phase (Midje et al.,2000).Although others contend that cheese flavor is associated with the water-soluble fraction(McGugan et al., 1979; Nelson and Barbano, 2004), it is clear that the quality of reduced fatcheese could be improved by increasing flavor intensity.

One popular approach to improve the flavor of reduced fat cheeses (and conventionalcheese,as well) has been to add selected flavor-producing lactic acid bacteria to the milk, in theform of adjunct cultures (Fenelon et al., 2002; Midje et al., 2000). As described below (Box 8),adjunct cultures are added to the milk in much the same way as the starter culture. However,their function is to produce enzymes and flavor compounds, rather than lactic acid.Adjunct cul-tures typically contain strains of Lactococcus and Lactobacillus that produce aminopeptidasesthat specifically hydrolyze bitter peptides.Although bitter peptides are often present in full-fatcheese, they can be even more of a problem in reduced fat cheese, in part because they parti-tion more in the non-lipid or serum phase,where they are more readily detected.Also, the lowersalt-in-moisture ratio typical of reduced fat cheese may result in higher cell numbers and greaterproteinase production and activity by the starter culture bacteria, which ultimately leads to bit-ter peptide accumulation.The use of non-bitter strains may,however, significantly reduce bitter-ness problems in these cheeses.

References

Fenelon, M.A.,T.P. Beresford, and T.P. Guinee. 2002. Comparison of different bacterial culture systems forthe production of reduced-fat Cheddar cheese. Int. J. Dairy Technol. 55:194–203.

Johnson, M.E., C.M. Chen, and J.J. Jaeggi. 2001. Effect of rennet coagulation time on composition, yield, andquality of reduced-fat Cheddar cheese. J. Dairy Sci. 84:1027–1033.

McGugan W.A., D.B. Emmons, and E. Larmond. 1979. Influence of volatile and nonvolatile fractions on intensity of Cheddar cheese flavor. J. Dairy Sci. 62:398–403.

Midje, D.L., E.D. Bastian, H.A. Morris, F.B. Martin,T. Bridgeman, and Z.M.Vickers. 2000. Flavor enhancementof reduced fat Cheddar cheese using an integrated culturing system. J. Agric. Food Chem. 48:1630–1636.

Nelson,B.K., and D.M.Barbano.2004.Reduced-fat Cheddar cheese manufactured using a novel fat removalprocess. J. Dairy Sci. 87:841–853.

Box 5–7. Reduced Fat Cheese—Taking the Fat (But Not the Flavor) Out of Cheese (Continued)



stages of aging, the ripening process could beshortened.

Bacteriophages

Bacteriophages (or phages, for short) are ar-guably the number one problem in cheese pro-duction.They are certainly the main reason whystarter culture activity is sometimes inhibited.Infection of starter cultures by phages may re-sult in slow or sluggish fermentations;occasion-ally the fermentation may fail completely. Phage

infections ultimately can lead to prolongedmake times and inferior or poor quality cheese,both of which cause substantial economic hard-ships on the manufacturer. Although phageswere once thought to be primarily a problemfor cheeses made using mesophilic lactococcalcultures, it is now recognized that thermophiliccultures (especially S. thermophilus) are alsosusceptible to attack by phages.

As discussed in Chapter 3, classification oflactic phages is based mainly on morphologi-cal criteria (i.e., head shape and size and tail

Box 5–8. Adding Flavor With Adjunct Cultures

The development of cheese flavor and texture is a complicated process, involving numerous enzyme-catalyzed reactions.The enzymes involved in these reactions originate from three mainsources: the coagulant, the milk, and bacteria.The latter are either normally present in milk dueto unavoidable contamination or are deliberately added.And although it is certainly possible tomake cheese without bacteria (i.e., aseptic cheese), the finished product would contain few, ifany of the flavor or texture characteristics one might expect.This would be especially true if thecheese were aged, since aged flavor and texture would simply not develop.

While the coagulant and natural milk enzymes certainly contribute to cheese flavor and tex-ture development, their roles are considered to be rather limited, whereas the presence of bac-teria is essential (El Soda et al., 2000). Moreover, it is not just the lactic starter culture bacteriathat are responsible for the desirable changes that occur during cheese ripening—bacteria pre-sent as part of the natural milk flora also serve as a vital reservoir of enzymes that are necessaryfor proper cheese maturation. If, however, the milk is pasteurized or if it is produced and han-dled under strict hygienic conditions, the natural microflora will be rather limited,both in num-ber and diversity.The finished cheese will ripen slowly and, in some cases,never achieve the de-sired quality expectations.

The realization that the natural milk flora is important for cheese ripening led investigatorsfirst to identify the relevant organisms, and then to consider the possibility of adding them di-rectly to the milk to boost or enhance flavor development (Reiter et al., 1967).These studies re-vealed that the bacteria that were most frequently associated with good cheese flavor were lac-tic acid bacteria, but not the same species or strains used as the starter culture.Therefore, as agroup, they are referred to as non-starter lactic acid bacteria (NSLAB).Cheese flavor is enhancedwhen the correct NSLAB, which generally consist of species of Lactococcus and Lactobacillus,are added to cheese milk. It also appears that the presence of these NSLAB may out-compete“wild”NSLAB that might otherwise cause flavor defects.They are now commercially available inthe form of adjunct cultures, and are widely used for accelerated ripening programs as well asfor reduced fat cheese, which often suffer from flavor problems (Box 5–7).

When adjunct cultures were first considered for modifying cheese flavor, cells attenuated byfreeze or heat shocking, physical-chemical treatment, or genetic means were used (El Soda etal., 2000).This was done to make the cells nonviable and to minimize subsequent acid produc-tion and metabolism.However,as noted above,adjunct cultures are now commercially availableand are added directly to the cheese milk, albeit at much lower inoculation rates compared tothe starter culture (Fenelon et al., 2002).

Among the species currently used commercially are strains of Lactobacillus helveticus, Lac-tobacillus casei, and Lactobacillus paracasei. In particular, the specific NSLAB used as adjunctcultures are selected based largely on their ability to produce aminopeptidases and to lyse dur-


Cheese 195

length). Importantly, they vary widely with re-spect to host range and virulence. Somephages have a narrow host range, infectingonly a few host strains,whereas others are ableto attack many different strains.Although tem-perate phages that lysogenize their hosts mayoccasionally become virulent, lytic phages are,by far, the more serious.

Following infection, lytic phages may repli-cate and release as many as 100 or more newinfective phage particles from a single host cellwithin thirty to forty minutes. Even if the initial

phage concentration in the milk is low, phagereplication may be so rapid that enough phagesare produced to significantly inhibit or deci-mate the starter culture before that culturedoes its job.In the example shown in Table 5–3,a single phage with a latent period of thirtyminutes and a burst size of fifty will outnumbereven a fast-growing culture in less than threehours. In a cheese operation, that means thecurd pH may never reach the desired level, leav-ing the manufacturer with a “dead” vat and ahuge headache. Phages that are more virulent

ing cheese aging.The aminopeptidases released by cell lysis are then able to degrade the bitterpeptides that accumulate in cheese as a result of partial casein hydrolysis by chymosin andstarter culture proteinases.Thus, the cheese becomes less bitter, and the released amino acidscan be metabolized to yield desirable end products.

The latter point deserves emphasis, because de-bittering is not the only function of cultureadjuncts.There appears to be an association, for example, between strains with high glutamatedehydrogenase activity (which forms �-ketoglutarate from glutamate) and formation of desir-able cheese flavor and aroma (Tanous et al., 2002). Similarly, lactic acid bacteria that metabolizeglutathione release cysteine into the cheese,which serves as precursor for sulfur-containing fla-vor compounds.

Recently, the use of non-lactic acid bacteria as cultures adjuncts has been proposed, basedlargely on the ability of these organisms to metabolize amino acids and generate various flavorcompounds (Weimer et al., 1999). Specifically, strains of Brevibacterium linens (the organismused for Limburger and other surface-ripened cheese) have been reported to enhance flavor de-velopment via production of sulfur-containing volatiles, such as methanethiol, and fatty acids,such as isovaleric (Ganesan et al., 2004).

References

El Soda, M., S.A. Madkor, and P.S.Tong. 2000.Adjunct cultures: recent developments and potential signifi-cance to the cheese industry. J. Dairy Sci. 83:609–619.

Fenelon, M.A.,T.P. Beresford, and T.P. Guinee. 2002. Comparison of different bacterial culture systems forthe production of reduced-fat Cheddar cheese. Int. J. Dairy Technol. 55:194–203.

Ganesan,B.,K.Seefeldt, and B.C.Weimer.2004.Fatty acid production from amino acids and �-keto acids byBrevibacterium linens BL2.Appl. Environ. Microbiol. 70:6385–6393.

McGugan W.A., D.B. Emmons, and E. Larmond. 1979. Influence of volatile and nonvolatile fractions on in-tensity of Cheddar cheese flavor. J. Dairy Sci. 62:398–403.

Midje,D.L.,E.D.Bastian,H.A.Morris,F.B.Martin,T.Bridgeman,and Z.M.Vickers.2000.Flavor enhancement ofreduced fat Cheddar cheese using an integrated culturing system. J.Agric. Food Chem. 48:1630–1636.

Reiter, B.,T.F. Fryer,A. Pickering, H.R. Chapman, R.C. Lawrence, and M.E. Sharpe. 1967.The effect of micro-bial flora on the flavor and free fatty acid composition of Cheddar cheese. J. Dairy Res. 34:257-272.

Tanous, C.,A. Kieronczyk, S. Helinck, E. Chambellon, and M.Yvon. 2002. Glutamate dehydrogenase activity:a major criterion for the selection of flavour-producing lactic acid bacteria.Antonie van Leeuwenhoek82:271–278.

Weimer,B.,K.Seefeldt,and B.Dias.1999.Sulfur metabolism in bacteria associated with cheese.Antonie vanLeeuwenhoek 76:247–261.

Box 5–8. Adding Flavor With Adjunct Cultures (Continued)



than the one cited in the example above, orthat are present initially at much higher levels,may inactivate the culture even faster. Con-versely, if the culture itself is more resistant tothe phages present in the plant, or if the phagelevels are low when its host is introduced intothe cheese environment, then the effects ofthat phage may be much less severe.

Ever since the phage problem was first rec-ognized (seventy years ago!), manufacturersand scientists have developed strategies to pre-vent or at least minimize the incidence and im-pact of phage infections. In fact,phages and theproblems they cause cheese manufacturershave provided the driving force not only formuch of the scientific research on dairy startercultures, but also for the development of ahighly sophisticated starter culture industry.Re-search on bacteriophages has advanced to a remarkable extent in the past two decades dueto advances in molecular biology and biotech-nology by groups in Australia,New Zealand, theNetherlands, United Kingdom, the UnitedStates and Switzerland.

Several control strategies have evolved fromthese phage research programs, and are nowpart of most cheese manufacturing operations.However,before describing these approaches, itis important to note that the first requirement ofany phage control plan is to ensure that the

starter culture propagation environment, aswell as the cheese production areas, be well sanitized and that phage entry points are as re-strictive as possible. Phages are sensitive tohypochlorite and heat, and can, therefore, be in-activated from equipment and culture media.Phages move about via whey aerosols, air, andpersonnel, so these represent important controlpoints.The bottom line is that, unless a cheesemanufacturer practices excellent sanitation, hasappropriate air handling systems in place (e.g.,HEPA filters, positive pressure in culture rooms,etc.), segregates whey processing to separate areas, and limits personnel traffic, more high-tech phage control measures will have littlechance of being effective.

Even when good sanitation programs are inplace, phages cannot be eliminated from thecheese manufacturing environment. After all,wherever there are growing lactic acid bacte-ria, there will be phages that attack them.Given that reality, the culture industry offers anumber of products designed to minimize orprevent phage problems. These include: (1)phage inhibitory media; (2) phage-resistant cul-tures; and (3) culture rotation programs.

The bulk culture tank is the first placewhere phage propagation can occur.Tradition-ally, milk or whey was used as the culturemedium, and although good growth of lactic

Table 5.3. The phage problem during industrial cheese fermentations1.

Phage absent Phage present

Time Cells/ml pH Cells/ml Phage/ml pH

0:00 1 � 106 6.6 1 � 106 1 6.60:30 2 � 106 6.5 2 � 106 � 1 50 6.51:00 4 � 106 6.3 4 � 106 � 50 2,500 6.31:30 8 � 106 6.1 8 � 106 � 2,500 125,000 6.22:00 1.6 � 107 5.8 1.6 � 107 � 125,000 6.25 � 106 6.02:30 3.2 � 107 5.5 2.5 � 107 � 6.25 � 106

� 1.9 � 107 3.1 � 108 5.73:00 6.4 � 107 5.2 1.9 � 107 � 3.1 � 108

� 0 3.1 � 108 5.6

1This simulated fermentation profile is based on the following assumptions:

Culture generation time � 0.5 hours

Phage replication time � 0.5 hours

Final desired pH after 3 hours � 5.2

Average phage burst size � 50


Cheese 197

starter bacteria could be obtained, phageswere also well-suited to this simple medium.Phage-inhibitory media, in contrast, are alsomilk- or whey-based media, but, in addition tosome added nutrients and growth factors, theyalso contain phosphate and citrate salts.Thesesalts bind calcium,which is required by phagesfor adsorption to host cells (the first step in theinfection process). There are some phages,however, that are unaffected by the lack of cal-cium and replicate fine in phage-inhibitory me-dia. Thus, despite its widespread use, this ap-proach has only a modest overall effect onphage control. Moreover, this medium only af-fects phage propagation within the bulk tank,and not later, during cheese manufacture.

Dairy microbiologists have long searchedfor lactococci and other lactic acid bacteriathat naturally resist bacteriophages. Althoughsuch strains exist, in many cases, these isolateswere often found to lack other relevant charac-teristics necessary for cheese production. Forexample, they either grew too slowly in milk,produced poor-quality cheese, or eventuallybecame phage sensitive. An alternative ap-proach that has been more successful has beento isolate spontaneous phage-resistant mu-tants. These mutants or variants can be ob-tained by repeated exposure of cheese produc-tion strains to the predominant phage presentin a given environment.To ensure that fermen-tative properties have not been altered, themutants must be re-evaluated for acid produc-tion rates while exposed to phages and underconditions mimicking those that occur duringcheese manufacture (the so-called Heap-Lawrence procedure, named after two NewZealand researchers).

The most powerful and effective means ofdeveloping phage-resistant cultures has beento exploit the innate ability of certain strains toresist lytic phage infections. Several mecha-nisms have been identified that are responsiblefor the phage resistance phenotype in lacticacid bacteria. In general, phage-resistant strainsblock infection at one of several points in thelytic cycle (Figure 5–10). In some strains,phageadsorption fails to occur, due either to modifi-

cation of cell wall attachment sites or interfer-ence by polysaccharides. Once adsorption oc-curs, phages ordinarily inject DNA into thehost cell to initiate phage DNA replication.Thisstep is inhibited in some cells, although themechanism is not clear. Following injection ofphage DNA, the host cell can attack that DNAvia expression of restriction enzymes that hy-drolyze DNA at specific restriction sites.

Finally, if the phage has survived all of thesehost defense systems, its replication can still be inhibited via a mechanism known asabortive infection.Although some phages maystill be released, and the infected cell is killed,the burst size is significantly reduced, and theimpact on the remaining cells is minimized.

The discovery in the 1980s and 1990s of themechanisms described above coincided withthe observation that genes coding for most ofthese traits in lactococci were located on plas-mids.Thus, it was possible to mobilize the rele-vant plasmid from one lactic acid bacterium toanother, and to confer a phage resistance phe-notype into the recipient strain. This strategywas first used, in 1986, to develop a phage-resistant cheese-making strain of L. lactissubsp. lactis.The plasmid, pTRK2030, actuallyencoded for several phage-resistant mecha-nisms (at the time, the researchers were awareof only a restriction system). Because conjuga-tion, a natural means of genetic exchange, wasused to effect gene transfer, the phage-resistanttransconjugant strain required no regulatoryapproval.Once introduced into a cheese plant,it performed extremely well.

The application of this and other naturalphage resistance mechanisms to improvestarter cultures quickly followed. If one plas-mid is good, two (or more) reasoned the Klaen-hammer group, would be better.Thus, a stack-ing strategy was developed in which a singlestrain contained different plasmids, each cod-ing for unique phage resistance properties.Al-ternatively, the different plasmids can be sepa-rately introduced into a single strain, yieldingmultiple isogenic strains harboring a uniquephage resistant plasmid.These strains can thenbe used in a rotation program (see below) to



prevent any one phage from reaching highenough levels to cause problems.

As noted above, it is inevitable that a phagecapable of infecting a given starter culture strainwill eventually emerge in a cheese plant.The in-fecting phage may be naturally occurring, or

more likely, is the result of DNA modification,mutation, or a gene acquisition event.There is,after all, no other phage-host relationship that ismore dynamic than that which occurs in a largecheese plant.Every day,there may be as many as1018 bacteria growing in such a plant,all serving

DNAReplication

Injection

Packaging

Lysis

AdsorptionAdsorptionInhibition

Restriction/Modification

AbortiveInfection

bacterial cell

phage

Phage

Replication

Figure 5–10. Lytic cycle of lactic bacterio-phages and steps at which phage resistancemechanisms function.


Cheese 199

as potential hosts for phages.Thus,staying a stepor two or three ahead of the phage is no easymatter, and engineering more durable andlonger-lasting phage resistance into lactic acidbacteria is a major challenge.The availability ofgenome information (for both the host cells andtheir phages) in the hands of clever microbiolo-gists has made it possible to deploy powerfulstrategies for development of new phage-resistant strains (Box 5–9).

Despite the opportunity for phages to findand infect host cells, the only scenario thatwould result in a particular phage actually af-fecting a fermentation would be if the phagewere initially present at reasonably high lev-els. Phages can only replicate, and thereforeincrease in number and reach danger levels,when they infect a suitable host. If the hostcell is not present, the phage will not last longin that environment, especially when good

Box 5–9. How to Outsmart a Virus

It is remarkable that the main biological problem that confronts the modern and sophisticatedfermented dairy foods industry is viruses, nature’s simplest life form. Given the discovery in the1980s that some lactic acid starter culture bacteria are armed with natural anti-viral defense sys-tems, one might think that controlling these viruses, or bacteriophages, would be a relativelyeasy task. However, this is clearly not the case, as “new” bacteriophages continue to emerge,much to the frustration of cheese manufacturers.

The inhibition of milk fermentations by bacteriophages was first described by New Zealandresearchers in 1935 (Whitehead and Cox).Although phage infections of starter culture bacteriawere initially only occasionally reported, such occurrences became more common as the sizeof the cheese industry grew.As previously noted, many innate phage-resistant mechanisms, andthe genes responsible for those mechanisms,have been identified in lactococci and other lacticacid bacteria and subsequently used to construct new phage-resistant strains.However, the con-ferred resistance has generally been effective only for the short-term, because new virulentphages invariably appear within a relatively short time. Given the parasitic nature of bacterio-phages (i.e., they can only replicate on suitable hosts) and the incredible opportunity for phage-host interactions, it is really no surprise that wherever lactic acid bacteria live and grow,phagescapable of infecting those bacteria will eventually arise. This realization has inspired re-searchers, especially the Todd Klaenhammer group at North Carolina State University, to de-velop novel and quite clever strategies for combating these phages and protecting starter cul-tures from phage infections.These “intelligent” approaches (also referred to as “artificial,” sincethey do not exist naturally) rely on molecular trickery to either prevent phage adsorption or toshort-circuit the phage lytic cycle and thereby block phage replication.

Pips and Pers

The first stage of the phage infection process occurs when a phage recognizes and then at-taches to a specific site on the surface of the host cell. One such receptor site, referred to as aphage infection protein,or pip,has been identified in Lactococcus lactis subsp. lactis C2 (Gelleret al., 1993; Garbutt et al., 1997). Infection by phage c2 does not occur in cells lacking an intactcopy of pip. Since mutations in the pip gene do not appear to have any other phenotypic effecton the cell, engineered pip-defective cells would be insensitive to phages that depend on Pipfor binding. Recently, similar results were obtained for another class of lactococcal phages thatdo not recognize pip, but instead bind to membrane proteins called receptor-binding proteins(Dupont et al., 2004).

Replication of the phage genome occurs soon after a phage has infected a host cell. Replica-tion begins at a unique location within the phage genome, the so-called origin of replication (or

(Continued)



ori site), which is characterized by several direct and inverted repeated sequences.The ori re-gion, therefore, serves as the recognition site for the DNA replication machinery. If, however,there were multiple “decoy” copies of the ori region within the cell, then the DNA replicationsystem would, in theory, be diluted out and phage genome replication would be reduced. In-deed, this approach, called phage-encoded resistance, or Per, was shown by Hill et al. (1990) towork in Lactococcus lactis. In this case, a phage genome fragment (containing the ori region)from phage �50 was introduced into a phage-sensitive strain via a multi-copy plasmid vector.The resulting transformant had a phage-resistant phenotype, and, as predicted, the greater theexpression of the plasmid, the more phage resistant was the cell.

Suicide Traps

For a given phage to inhibit a fermentation, that phage must first infect a host cell, replicate in-side that cell and make more phage particles, and then lyse the cell to release the fully assem-bled and infectious phage. If the lytic cycle is somehow blocked prematurely, then viable phageparticles will not be formed or released.Although these naturally occurring abortive infectionsystems (Abi) are widespread in lactic acid bacteria, it is also possible to contrive or engineersuch a system by genetic means.This strategy, envisioned by O’Sullivan et al. (1996) and Djord-jevic et al. (1997), involved construction of a plasmid that contained lactococcal-derived genesencoding for a restriction/modification system (LlaIR) positioned downstream from a promoterregion obtained from a lactococcal phage (�31).When strains harboring this plasmid were at-tacked by phage, the restriction enzyme encoded by LlaIR was induced and expressed, whichthen lysed host DNA, leading to cell death.Thus, the single infected cell dies, but phage propa-gation is blocked. Or, as the authors eloquently stated,“infected cells . . . undergo programmedcell death in an altruistic fashion,” sparing the remaining population from a more severe phageinfection (Djordjevic et al., 1997).

Antisense mRNA

When DNA is transcribed into mRNA, the relevant transcription signals are present on only oneof the two strands of DNA, the coding or sense strand.The non-coding or antisense strand is nottranscribed. If,however, the transcription elements (consisting of a promoter and terminator re-gions) are positioned around the antisense strand, the latter will also be transcribed.Assumingenough of the antisense RNA transcripts are produced, they will hybridize with the sensemRNA, preventing translation and protein synthesis. Phage replication can be reduced by tar-geting genes encoding for essential phage proteins, although in lactococci, this approach hasbeen met with mixed results (Kim et al., 1992;Walker and Klaenhammer, 1998). However, bycombining antisense technology with a Per approach, it was possible to obtain “explosive” am-plification of the number of antisense ori transcripts and achieve a more significant reductionin phage proliferation (McGrath et al., 2001;Walker and Klaenhammer, 2000).

Phages strike back

As noted above, the phage-host relationship is extremely dynamic.There may easily be morethan 1015 cells in a large cheese plant and just as many viable phages. Even if the frequency ofspontaneous mutations within the phage population is low (� 10-9), many “new” phages willcertainly appear on a regular basis.Some of these mutations will confer an ability to infect a pre-viously resistant cell.However,given the nature of the phage infection,DNA packaging, and cellburst processes, co-mingling of host and phage DNA is also likely to occur via recombination-type events.

Box 5–9. How to Outsmart a Virus (Continued)


Cheese 201

sanitation is practiced. This rationale formsthe basis of the culture rotation programs thatare widely used in the cheese industry. Agiven culture, which may contain only one ortwo, but as many as six different strains, willbe used for a day or two, then followed by adifferent culture containing strains with dif-ferent phage sensitivity patterns.Thus, phages

that infect one or more of the strains presentin the initial culture may have begun to be-come established in the plant, but absenttheir host cells (which were rotated out),these phages will quickly die out and de-crease in number. Rotation programs can besupplemented by various phage monitoringprocedures in which phage titers (i.e., phage

Thus, lytic phages that have acquired host or prophage DNA will emerge within phage pop-ulations. In some cases, it turns out, the acquired DNA provides the phage with the means nec-essary to counter or circumvent the resistance of the host cell. For example, in one well-knownstudy in which a once resistant L. lactis commercial starter culture strain (LMA12-4) harboringa phage resistance plasmid (pTR2030) had become phage-sensitive, it was shown that thephage (�50)had acquired a methylase gene that was carried by the host cell (Sanders et al.,1986;Alatossava and Klaenhammer, 1991). Since methylases modify host DNA so that it is notdegraded by host restriction enzymes, the presence of this methylase gene in phage �50 en-ables this phage to self-methylate its DNA and thereby evade restriction by the host cell.

References

Alatossava,T., and T.R.Klaenhammer.1991.Molecular characterization of three small isometric-headed bac-teriophages which vary in their sensitivity to the lactococcal phage resistance plasmid pTR2030.Appl.Environ. Microbiol. 57:1346–1353.

Djordjevic, G.M., D.J. O’Sullivan, S.A.Walker, M.A. Conkling, and T.R. Klaenhammer. 1997.Triggered-suicidesystem designed as a defense against bacteriophage. J. Bacteriol. 179:6741–6748.

Dupont,K.,T. Janzen,F.K.Vogensen, J. Josephsen,and B.Stuer-Lauridsen.2004. Identification of Lactococcuslactis genes required for bacteriophage adsorption.Appl. Environ. Microbiol. 70:5825–5832.

Geller,B.L.,R.G. Ivey, J.E.Trempy,and B.Hettinger-Smith.1993.Cloning of a chromosomal gene required forphage infection of Lactococcus lactis subsp. lactis C2. J. Bacteriol. 175:5510–5519.

Garbutt, K.C., J. Kraus, and B.L. Geller. 1997. Bacteriophage resistance in Lactococcus lactis engineered byreplacement of a gene for a bacteriophage receptor. J. Dairy Sci. 80:1512–1519.

Hill, C., L.A. Miller, and T.R Klaenhammer. 1990. Cloning, expression, and sequence determination of abacteriophage fragment encoding bacteriophage resistance in Lactococcus lactis. J. Bacteriol.172:6419–6426.

Kim, S.G., and C.A. Batt.Antisense mRNA-mediated bacteriophage resistance in Lactococcus lactis.Appl.Environ. Microbiol. 57:1109-1113.

O’Sullivan,D.J., S.A.Walker,and T.R Klaenhammer.1996.Development of an expression system using a lyticphage to trigger explosive plasmid amplification and gene expression. Biotechnology 14:82–87.

McGrath, S., G.F. Fitzgerald, and D. van Sinderen. 2001. Improvement and optimization of two engineeredphage resistance mechanisms in Lactococcus lactis.Appl. Environ. Microbiol. 67:608–616.

Sanders, M.E., P.J. Leonhard, W.D. Sing, and T.R. Klaenhammer. 1986. Conjugal strategy for construction of fast acid-producing, bacteriophage-resistant lactic streptococci for use in dairy fermentations.Appl.Environ. Microbiol. 52:1001–1007.

Walker,S.A.,and T.R Klaenhammer.1998.Molecular characterization of a phage-inducible middle promoterand its transcriptional activator from the lactococcal bacteriophage �31. J. Bacteriol. 180:921–931.

Walker, S.A., and T.R Klaenhammer. 2000.An explosive antisense RNA strategy for inhibition of a lactococ-cal bacteriophage.Appl. Environ. Microbiol. 66:310–319.

Whitehead, H.R., and G.A. Cox. 1935.The occurrence of bacteriophage in starter cultures of lactic strepto-cocci. N. Z. J. Sci.Technol. 16:319–320.

Box 5–9. How to Outsmart a Virus (Continued)



concentrations) are enumerated or estimated.This allows cheese manufacturers to continueusing the same culture (therefore, minimizingproduct variation) and only make a switchwhen the phage levels reach a particularthreshold.

Microbial Defects, Preservation, and Food Safety

Given that cheese is made in a mostly open,non-sterile environment, using non-sterile rawmaterials, and is exposed to or held at non-lethal, non-inhibitory temperatures, it is notsurprising that all sorts of microorganisms cangain entry in cheese. Although the combinedeffects of moderately low pH, moderately highsalt concentrations, and low Eh provide a bar-rier against some microorganisms, there areothers, including spoilage organisms, and inrare cases, pathogens, that can grow in thecheese environment. Thus, microbial defectsare not uncommon in cheese and can be re-sponsible for significant economic loss. Ofcourse, chemical and physical defects (e.g., ox-idation, crystal formation, discoloration, me-chanical openings) can also be responsible fordecreased shelf-life or consumer rejection,however, they generally are less serious.

Microorganisms can cause several generaltypes of spoilage in cheese.These include ap-pearance defects, texture defects, and flavorand aroma defects. Appearance defects arethose that may cause the consumer to reject aproduct based strictly on sight. Perhaps thebest example is cheese contaminated withmold growth. Although there are many differ-ent fungi capable of growing on cheese, Peni-cillium spp. are the most common. Fungi areobligate aerobes, so growth occurs almost ex-clusively on the surface, where oxygen is avail-able. However, mold growth on packagedcheese, even vacuum-packaged products, canalso occur, provided there is enough oxygenpresent. Mold growth is more common inblock cheese, and even though it can betrimmed prior to cutting and packaging, trim-ming is time-consuming and results in loss of

yield. Since some fungi that grow on cheesebelong to species known to produce mycotox-ins, there has been concern that mycotoxinscould have diffused in the cheese and thattrimming visible mold from the surface maynot be sufficient to render the product safe.However, numerous studies have revealed thatmycotoxins are not produced on cheese, evenwhen the fungi is theoretically capable of syn-thesizing toxins. Presumably, the cheese envi-ronment restricts or inhibits expression oftoxin biosynthesis pathways.

Appearance spoilage also occurs due to for-mation of undesirable pigments. The mostcommon and best-studied example is the pinkdefect (or pink ring) that occurs around theexterior of Parmesan-type cheeses.The defectis worse after prolonged aging, and if thecheese is subsequently grated, a pink-to-browncolor may occur throughout the grated cheese.Several factors contribute to the formation ofthe pink color, including oxygen and the pres-ence of reducing sugars and free amino acids(i.e., reactant of the Maillard browning reac-tion).The starter culture, and the L. helveticusstrains in particular, appears to be the mainculprit, by virtue of its ability to produce oxi-dized pigments from the amino acid, tyrosine.

Finally holes or slits may form as a result ofgas production. Several organisms are capableof producing gas (mainly carbon dioxide) incheese, including yeast, coliforms, propionibac-teria, and clostridia. The clostridia, especiallyClostridium tyrobutyricum, are more commonin milk obtained from silage-fed cows and canproduce enough CO2 to cause formation oflarge holes (�3 cm in diameter). However, theheterofermentative lactic acid bacteria are prob-ably the most common cause of the slit defect.Relevant species include Lactobacillus fermen-tum and Lactobacillus curvatus.These bacteriaare either naturally present in the milk or arenormal contaminants in a cheese plant; thus,they may be particularly difficult to control, es-pecially in raw milk cheese. Residual lactose orgalactose in the cheese provides substrate forthese anaerobic bacteria, whose growth is en-hanced within the interior of large cheese


Cheese 203

blocks or barrels where the cheese is slow tocool and where low Eh conditions prevail.

The most common texture defects includeslime, crystal or sandy mouth feel, and poorbody. Polysaccharides that impart a slimy tex-ture to the cheese can be produced by psy-chrotrophic Gram negative bacteria such asPseudomonas and Alcaligenes, as well as byyeast and lactic acid bacteria. Crystals thatform on cheese are usually comprised of calcium lactate and are mainly a physical-chemical problem, caused by low solubility ofcalcium lactate salts at the cheese surface.However, pediococci, lactobacilli, and othernon-starter lactic acid bacteria may contributeto this problem by converting L(�) lactate toits isomer, D(�) lactate, which is less soluble.Tyrosine-containing crystals can also form indry cheeses (e.g., Parmesan and Romano).

Flavor defects can be caused not only bycontaminating organisms, but also by thestarter culture itself. If the culture is given theopportunity, it may produce too much lacticacid, causing an acid defect. As described ear-lier, certain peptides generated by starter cul-ture proteinases are bitter-tasting, and unlessfurther metabolized, cause a bitter defect. Het-erofermentative, non-starter lactic acid bacteriaproduce ethanol and acetic acid in addition toCO2.Acetic acid imparts a highly objectionablevinegar-like sour flavor to the cheese. Thesebacteria may also hydrolyze triglycerides, re-leasing free fatty acids that contribute fruity fla-vors into the cheese. Production of hydrogensulfide and other sulfur-containing compounds,ammonia, and other volatiles by bacteria andfungi, are also common flavor defects.

Minimizing the entry of spoilage organismsinto the cheese milk and subsequently into thecheese is one of the most important steps inpreventing spoilage. In most cases, the cheesewill only be as good as the milk from which itwas made.The use of high quality milk, appli-cation of good sanitation practices, and propercheese manufacture are essential. Pasteuri-zation of milk, even for aged cheese, is nowcommon, not only to inactivate pathogens (seebelow), but also to kill potential spoilage or-

ganisms. Some specific spoilage problems canbe addressed by the use of chemical preserva-tives,either added to the milk or applied to thecheese. For example, mold growth can be in-hibited by sorbic acid (usually in the form ofpotassium sorbate salts) or the antibioticnatamycin (also known as pimaricin). In someEuropean countries, but not in the UnitedStates, lysozyme and sodium nitrate can beadded to milk to control clostridia.

Although spoilage of cheese by microorgan-isms is a constant concern, the presence ofpathogens in cheese occurs only infrequently,and rarely does foodborne disease result (Box5–4).However,due to the serious public healthconsequences, as well as potential economicloss due to recalls and liability, attention tofood safety is an absolute requirement. Just asingle positive test for L. monocytogenes or E.coli O157:H7 may initiate a product recall.Thus, the goal is not just to prevent the growthof pathogens, but also to reduce or eliminatetheir very presence in cheese. In the UnitedStates, this means more and more cheese is be-ing made from pasteurized milk, and exposureto environmental sources of contamination isminimized (e.g., via the use of enclosed vats).Most cheese manufactures have also adoptedHazard Analysis Critical Control Points(HACCP) plans that are designed to anticipateand prevent food safety problems.

Another type of foodborne disease that occa-sionally occurs in cheese is caused by the pres-ence of the biogenic amines histamine and tyra-mine, but is unrelated to foodborne pathogens.In sensitive individuals, these amines (mainlyhistamine) can cause headaches, nausea, andcramps and other gastrointestinal symptoms.These compounds are produced from theamino acids histidine and tyrosine, respectively,via specific amino acid decarboxylases.Bacteriathat produce these enzymes and form theseamines include Lactobacillus buchneri andother lactobacilli, lactococci, and enterococci.Aged cheese, particularly aged Cheddar, Gouda,and Swiss, are more likely to contain biogenicamines due to the higher concentration of freeamino acid substrates that accumulate during



more extensive proteolysis. Because the bac-teria responsible for amine production are ordi-narily present as raw milk contaminants, pas-teurization effectively minimizes this problem.

Whey Utilization

As shown earlier in this chapter (Figure 5–4),water, in the form of whey, is released whenmilk is converted into cheese. In fact, the wheyaccounts for 90% of the original milk volume.The dilute nature of the whey (about 92% to94% water) and low protein concentration(�1%) have historically contributed to the per-ception (at least in the United States) that wheyhas little economic value. However, in otherparts of the world, whey is more widely used,especially in the manufacture of whey-derivedcheese. For example, in Norway, the goat wheycheeses Gjetost and Mysost are among themost popular of all cheeses (accounting forabout 20% of cheese consumption), and Ri-cotta, another whey-based cheese, is popularnot only in Italy,but throughout Europe and theUnited States.Whey cheeses are also producedin Greece (Manouri,Anthotyros), Portugal (Re-quejño), and the Czech Republic (Urda).

Whey cheese manufacture is based on theprinciple that whey proteins will precipitatewhen slightly acidified (pH 6) and then heatedto high temperatures (�70°C).Although wheyalone is ordinarily used as the starting material,milk can also be added. Much of the Ricottaproduced in the United States, for example,contains a portion of skim or whole milk.Thefinished cheese, in this case actually consists ofa casein-whey co-precipitate.Whey cheeses arenot fermented,and since the lactose content ofwhey may be as high as 75% (on a solids basis),these cheeses will also contain 3% to 4% lac-tose and are generally sweet.Although Ricottais a very soft cheese with a white, creamycolor, other whey cheeses, such as Gjestot, canbe very hard and a light brown color.The tan-to-brown color results from lactose carameliza-tion and Maillard browning reactions that oc-cur during high temperature heating.

In the United States, where the market forwhey cheeses has been extremely small, mostcheese manufacturers instead used whey as an-imal feed or fertilizer or,more likely, simply dis-carded it altogether. In the past thirty years thispractice has changed for several reasons. First,whey is now considered to be environmentalproblem. It has a high biological oxygen de-mand (BOD), and, given that nearly 25 millionkg are generated each year, discharging thismuch high BOD material down the drain orinto the environment exceeds that whichsewage treatment plants can handle.

Second, the dairy and food processing in-dustries have developed products and applica-tions in which economic value can be derivedfrom whey and whey components. Many ofthese applications involve separation technolo-gies (e.g.,ultrafiltration and reverse osmosis) inwhich the more valuable whey protein frac-tions are obtained. What remains from theseprocesses is a lactose-rich material (whey per-meate) that can be used directly, or further pu-rified, as food- or chemical grade lactose.

Importantly, it is technologically possible touse whey,whey permeate,or whey-derived lac-tose as feedstocks for various industrial fermen-tations, including the production of organicacids, ethanol, and other small molecules. How-ever, given their current market value and thecapital and processing costs that would be nec-essary, it does not make economic sense at thepresent time to produce these commoditychemicals from whey fermentations.

BibliographyCoffey, A., and R.P. Ross. 2002. Bacteriophage-

resistant systems in dairy starter strains: molecu-lar analysis to application.Antonie van Leeuwen-hoek. 82:303–321.

Fox, P.F., and P.L.H. McSweeney. 2004. Cheese: anoverview, p. 1—18. In P.F. Fox, P.L.H. McSweeney,T.M. Cogan, and T.P. Guinee (ed.). Cheese: Chem-istry, Physics and Microbiology, Third Edition,Volume 1:General Aspects.Elsevier Ltd.,London.

Guinee,T.P., and P.F. Fox, 2004. Salt in cheese: physi-cal,chemical and biological aspects,p.205—259.In P.F. Fox, P.L.H. McSweeney,T.M. Cogan, and T.P.


Cheese 205

Guinee (ed.).Cheese: Chemistry, Physics and Mi-crobiology, Third Edition, Volume 1: General As-pects. Elsevier Ltd., London.

Hutkins, R.W. 2001. Metabolism of starter cultures,p. 207—241. In E.H. Marth and J.L. Steele (ed.),Applied Dairy Microbiology. Marcel Dekker,Inc., New York, NY.

Kosikowski, F.V., and V.V. Mistry. 1997. Cheese andFermented Milk Foods. Volume 1: origins andprinciples. 3rd ed. F.V. Kosikowski, Great Falls,Virginia.

Kosikowski, F.V., and V.V. Mistry. 1997. Cheese andFermented Milk Foods. Volume 2: Proceduresand analysis. 3rd ed. F.V. Kosikowski, Great Falls,Virginia.

Lawrence, R.C., J. Gilles, L.K. Creamer,V.L. Crow, H.A.Heap, C.G. Honoré, K.A. Lawrence, and P.K.Samal. 2004. Cheddar cheese and related dry-salted cheese varieties, p. 71–102. In P.F. Fox,P.L.H. McSweeney, T.M. Cogan, and T.P. Guinee(ed.). Cheese: Chemistry, Physics and Microbiol-ogy, Third Edition, Volume 2: General Aspects.Elsevier Ltd., London.

Lucey, J.A., M.E. Johnson, and D.S. Horne. 2003.Perspectives on the basis of the rheology andtexture properties of cheese. J. Dairy Sci. 86:2725–2743.

Marcos, A., M. Alcala, F. Leon, J. Fernandez-Salguero,and M.A.Estaban.1981.Water activity and chemi-cal composition of cheese. J. Dairy Sci. 64:622–626.

McSweeney, P.L.H. 2004. Biochemistry of cheeseripening. Int. J. Dairy Technol. 57:127–144.

McSweeney, P.L.H. 2004. Biochemistry of cheeseripening: Introduction and Overview, 347—360.In P.F. Fox, P.L.H. McSweeney,T.M. Cogan, and T.P.Guinee (ed.). Cheese: Chemistry, Physics and Microbiology, Third Edition, Volume 1: GeneralAspects. Elsevier Ltd., London.

Pintado, M.E., A.C. Macedo, and F.X. Malcata. 2001.Review: technology, chemistry and microbiologyof whey cheeses. Food Sci.Tech. Int. 7:105–116.

Singh,T.K., M.A. Drake, and K.R. Cadwallader. 2003.Flavor of cheese: a chemical and sensory per-spective. Comp. Rev. Food Sci. Food Safety. 2:139–162.

Footnotes1“How can you govern a country which has 246 va-

rieties of cheese?” Charles De Gaulle, as quotedin Les Mots du General, 1962.


6

Meat Fermentation

“Stomach of goat, crushedsheep balls, soft fullpearls of pig eyes,snout gristle, fresh earth,worn iron of trotter, slateof Zaragoza, dried cat heart,cock claws. She grindsthem with one hand andwith other fistsmountain thyme, basil,paprika, and knobs of garlic.And if a tooth of stink thistlepulls blood from the roundblue marbled handall the better forthis ruby of Pamplona,this bright jewel of Vich,this stained crownof Solsona, thissalami.”

From “Salami” by Philip Levine, poet

must have been far more troubling to eat meatthat lacked obvious signs of spoilage only tosuffer hours or days later from a serious illnessof one form or another. Comminuted, orchopped or ground, meat products must havebeen even more prone to spoilage.

It probably did not take long, therefore, forsome aspiring food scientists to develop waysto preserve meat. Drying was undoubtedlyamong the first of these technologies provento be successful. It was probably observed thatsmoking and salting were also effective. If,however, a small amount of sugar was added toa meat mixture,and if the salt level was not toohigh and other conditions were just right, thenan entirely different type of product could be

Introduction

If the origins of fermentation and the makingand eating of fermented foods was based onfood preservation, then meat serves as one ofthe best examples of the application of thistechnology. When, many thousands of yearsago,wild or domesticated animals were slaugh-tered for food, unless some sort of low temper-ature storage was available, either the entireanimal had to be cooked and consumed withina relatively short amount of time or the meatwould spoil. Even if these early consumerswere less discerning than the more sophisti-cated consumers of today, rotten meat stillmust have had limited appeal. Of course, it

207





produced, one that had a tart but appealing fla-vor and lasted a long time, especially if com-bined with drying. Moreover, if the salt alsocontained sodium or potassium nitrate, ahighly desirable cured meat color and flavorwas also achieved.

History and Evolution of theFermented Meats Industry

Like other fermented foods, recorded refer-ences to fermented meats date back thousandsof years. The manufacture of these productslikely originated in southern Europe and areassurrounding the Mediterranean Sea during theRoman era, although there were probablyAsian counterparts that appeared around thesame time. Even though it is not absolutelyclear from the historical records whetherthese early sausage products were actually fer-mented, it is difficult to imagine, given the cir-cumstances, that there wasn’t some sort of nat-ural fermentation occurring.

The manufacture of these early fermentedmeat products most certainly occurred as a re-sult of accidents,and repeating these successeson a consistent basis must have been very diffi-cult. We know that the key to produce thesemeat products via a natural fermentation relieson creating conditions that select for theproper organisms (discussed below).However,the technique of backslopping—taking a por-tion of a good finished product and adding it tothe starting raw material—was likely adoptedto increase the probability of success, despitethe fact that the rationale behind this step wasnot understood. This technology is still prac-ticed throughout the world, even though, aswe shall see later, there are now far more reli-able methods for ensuring a consistent fermen-tation. In any event, the technology of fer-mented meats production remained an art formany years. In fact, sausage manufacturingwas, like the manufacture of other fermentedfoods, performed by craftsmen, monks, andother trained specialists.

The actual number and types of productsthat evolved were many, and like other fer-

mented foods, depended largely on geography.Certainly, preservation was the likely drivingforce for many of the processing practices thatwere adopted. In warmer areas, such as thosein the Middle East and around the Mediter-ranean, spices were often added, and a dryingstep was common. Thus, dry, peppery, andspicy sausages, such as Genoa salami and pep-peroni, evolved in Italy. In colder, more North-ern areas, where sausage technology is morerecent, spices were rarely added, and insteadproducts were usually smoked or sometimescooked following fermentation. Most of themoist and semi-dry German-style sausages,such as Thuringer,Lebanon bologna,and cerve-lat, are of this variety.

The drying,cooking, and smoking steps, theaddition of salt and spices, as well as the incor-poration of nitrate salts, not only added flavorand appeal, but also would have given theseproducts a long shelf-life, even at ambient tem-perature. That these products are enclosedwithin casings also contributes to their preser-vation, since post-processing microbial conta-minants are excluded. In fact, the preservationof fermented meats serves as a perfect exam-ple of what food scientists now refer to as thehurdle or barrier concept of food preserva-tion. Preservation is achieved not by any onespecific process, but rather by a combinationof processes, such that fermentation is com-bined with drying, salting, smoking, cooking,and antimicrobial chemicals to ensure safetyand to enhance preservation (discussed laterin more detail).

The fermented meats industry, compared toother fermented foods industries, is relativelynew. In fact, it was not until well into the twen-tieth century that producers of fermentedmeats began to develop and apply modernproduction technologies. Previously, sausagemanufacture was mostly confined to craftsmenand at-home producers who made products ona batch-by-batch basis. Product quality, consis-tency, and especially safety were not alwaysachieved. Equipment for grinding, mixing, andstuffing procedures, as well as fermentationchambers, only became available in the past


Meat Fermentation 209

sixty to seventy years. Moreover, pure curingagents, synthetic casings, and starter cultureshave only been available since the 1960s. It isno coincidence that it was during this periodthat the industry increased in size and thatrapid, high throughput manufacturing meth-ods capable of producing products of consis-tent quality and safety developed.

Today the variety of fermented meat prod-ucts available around the world is nearly equalto that of cheese. In Spain, for example, thereare at least fifty different types of fermentedsausages, and in Germany there are more than350. Not only is this variation due to the meatsource (i.e., beef, pork, goat, sheep, etc.), butthe cut of meat, the amount and coarseness ofthe fat, and the casing material used to formthe shape all have a profound influence on the

finished product. The level of dryness andwhether or not smoking is applied or moldgrowth permitted to occur are especially im-portant factors and form the basis of fer-mented meats classification (Table 6–1).

It is worth emphasizing, especially for read-ers less familiar with sausage nomenclature,that many of the sausage products available inthe marketplace are simply cured, comminutedproducts, meaning they contain curing saltsand undergo many of the same processingsteps as do fermented sausages, but they arenot fermented.Thus, frankfurters, bologna, andbreakfast sausages are not fermented and willnot be discussed in this chapter. In contrast,there are fermented meat products that aremade from whole, intact meat materials,such asthe country hams popular in selected regions

Table 6.1. Examples of fermented meats and sausages.

Product Meat Areas Produced Features

Moist sausages(aw � 0.94)Lebanon bologna Beef North America Low pH (�4.9)

SmokedMortadella Pork Italy, France, USA Low pH

Un-smokedCooked

Semi-dry sausages(aw � 0.90 � 0.95)Summer sausage Beef/Pork USA Moderate pH

(4.9 � 5.0)Smoked

Thuringer cervelat Pork Italy, France, USA Low pHUn-smokedCooked

Dry sausages(aw � 0.90)Pepperoni Beef/Pork Italy, North America Low pH

Un-smokedSpiced

Salami Pork Europe, USA, Mexico Low aw (�0.85)Un-smokedCooked

Chorizo Pork/Beef Italy Highly spicedLow acid

Whole meatsCountry-cured hams Pork North America Heavily saltedParma hams Pork Italy Heavily salted



in the United States and Europe.These productswill be discussed separately.

Although fermented meats are producedand consumed worldwide, they are most pop-ular in Europe (Table 6–2). More than 600 mil-lion kg of fermented sausage are consumed inGermany, Spain, France, and Italy.About 3% to5% of all meat consumed in these countries isin the form of fermented sausages.

Meat Composition

Fresh meat is a nutrient-rich medium and is, infact, one of the best for supporting growth ofmicroorganisms. Skeletal bovine muscle con-tains nearly 20% of high quality protein, about2% to 3% lipid, and a small amount of carbohy-drate,non-protein nitrogen,and inorganic mate-rial.The balance, about 75%, is water, so the wa-ter activity (aw), not surprisingly, is nearly 0.99.The pH of the fresh tissue,before rigor, is 6.8 to7.0, but decreases to about 5.6 to 5.8 followingrigor, due to post-mortem glycolysis by endoge-nous enzymes present within muscle cells.

These values may vary somewhat,dependingon the animal and its status before slaughter,butregardless of the source of meat (e.g., pork usu-ally has a slightly higher pH), these conditionsare just about perfect for most bacteria, includ-ing spoilage organisms and pathogens. Conse-quently, raw meat is highly perishable. And al-though the interior of intact meat tissue issterile, the exterior surface harbors a wide arrayof aerobic and facultative bacteria. In its com-minuted form, i.e., chopped or ground, meat

spoilage can be even more rapid, because thesurface microflora become homogeneously dis-tributed throughout the meat.And even thoughoxygen may also be incorporated into the mix-ture, enhancing growth of aerobes, anaerobicpockets can also develop within the interior,providing suitable environments for anaerobicbacteria. Included among the latter are Clostrid-ium botulinum and other related species capa-ble of causing serious, sometimes fatal food poisoning disease. Since fermented sausages are typically made using raw,comminuted meat,controlling the resident flora, especiallypathogens, is of utmost importance. If control isnot maintained, the results can be disastrous.

Fermentation Principles

In contrast to the lactic acid fermentation thatoccurs in milk, the meat fermentation hasbeen,until recently,considerably less well stud-ied and understood. In fact, the use of pure, de-fined starter cultures in the fermented meatsindustry is a relatively recent development (be-gun only in the 1950s and ‘60s). Before the useof meat starter cultures, the most common wayto start the fermentation practice was,as notedabove, backslopping.

Backslopping works for several reasons.First, backslopping ordinarily selects for thosebacteria that are well suited for growth in thesausage environment. Strains that are slow toscavenge for carbohydrates, are inhibited byfermentation acids, or are sensitive to salt or ni-trite are not maintained. Instead, they are dis-placed by more competitive bacteria that haveparticular metabolic and physiological advan-tages in that environment.Thus, even a prolificacid-forming strain does have much of a chanceif it is not also tolerant of salt and nitrite andable to grow in a low oxygen environment.

Second, the bacterial population that be-comes established during repeated transfersand fermentations is heterogenous in nature,consisting of multiple species and strains. If, forexample, one such strain were to suddenly dieor otherwise be lost due to the presence ofbacteriophages or an inhibitory agent, then the

Table 6.2. Consumption of fermented meats.

Country Kg/person/year

Canada1 0.9Finland1 3.1France2 3.1Germany2 7.1Italy2 4.5Spain2 2.7United Kingdom2 0.1United States1 0.3

1Estimated from 2001 WHO statistics2Estimated from Fisher and Palmer, 1995



remaining strains, acting as back-up, would stillbe able to complete the fermentation.

Finally, backslopping is effective due simplyto the size of the inoculum, which is usuallyaround 5%, but which can be as high as 20% ofthe total mass. Such a large inoculum providesreasonable assurance that the desired organ-isms will overwhelm the background flora andthat undesirable interlopers will have littlechance of competing.

Despite backslopping’s long history andwide use, there are several drawbacks associ-ated with it. Such products often have inconsis-tent quality and fermentations can be unreliableand difficult to control. In large production facil-ities in Europe, and especially in the UnitedStates, short and consistent fermentation times,standardized production schedules, and consis-tent product quality have become essential. Amanufacturer producing a single, small batch ofproduct can tolerate delays,and may even cometo appreciate differences in product quality.However, inconsistent product quality and pro-duction delays due to sluggish fermentationsare unacceptable for large manufacturers whohave substantial employee payrolls and tightproduction schedules.Above all, the entire back-slopping process can be considered microbio-logically risky, since any deviation from thenorm (i.e., slow or delayed fermentation) maypermit growth of Staphylococcus aureus, Liste-ria monocytogenes, C. botulinum, or otherpathogens of public health significance.

Although obvious, the following point can-not be over-emphasized: sausage is made fromraw meat that may well contain pathogenic or-ganisms. If a cooking step is not included, fer-mentation represents the primary means ofpreservation and the main barrier againstpathogens. Since the actual fermentation cantake a long time (from twelve to thirty-sixhours and in some cases, even longer), a slowor failed fermentation may not be discoveredright away, permitting growth of pathogenicbacteria. If the bacteria are able to reach highlevels, even subsequent acid production (oreven a heating step) may not be sufficient toinactivate these pathogens.

It is important to mention, as noted above,that it is entirely possible to produce fer-mented sausages without any type of culture atall.This practice is not uncommon, and is stillpracticed in many parts of Europe. One cansimply prepare the sausage mixture and waitfor the natural lactic flora to take over. Thismethod is successful because the formulationand fermentation conditions (i.e., salt, nitrate,low temperature, and an anaerobic environ-ment) provide sufficient selection of desirablelactic acid bacteria. However, for obvious foodsafety reasons, none of the large, modern oper-ations in the United States rely on natural fer-mentations. Finally, there is one other way toproduce sausages that have a tangy or acid-likeflavor without a starter culture and without re-lying on the indigenous microflora. The meatmixture can be directly acidified using a food-grade acidulant (e.g., glucono-�-lactone),which results in a product with sensory prop-erties that mimic (somewhat) those of fer-mented sausage.

Meat Starter Cultures

Once microbiologists began to study andidentify the microorganisms present in fer-mented sausages in the 1940s, it became clearthat lactic acid bacteria were the primary or-ganisms responsible for the fermentation.Thisconclusion was based on the fact that the pre-dominant organisms isolated from naturallyfermented sausages were species of Lacto-bacillus. When the isolates were propagatedand re-inoculated into fresh meat, a well-fermented sausage could be produced withall the expected characteristics.

Patents, based on using these bacteria asmeat starter cultures, were assigned in Europeand the United States. However, application ofthis technology was initially unsuccessful. Thiswas because the Lactobacillus strains that hadbeen identified and used successfully in trial sit-uations were difficult to mass produce in a formconvenient for sausage manufacturers. For anyorganism to function well as a starter culture, itmust not only satisfy the performance criteria,



but it must also be present in high numbersand be viable at the time of use. In the case ofthe Lactobacillus cultures, cell viability fol-lowing lyophilization (or freeze drying, themain form of starter culture preservation) waspoor, leading to slow and unacceptable fer-mentation rates.

Demand for a culture that could be used forfast, consistent, large-scale production of fer-mented sausage eventually led to the discoveryof other lactic acid bacteria that not only hadthe relevant performance characteristics, butalso the durability required for commercial ap-plications. One organism, classified as Pedio-coccus cerevisae, was found to have theseproperties. Even though pediococci are notnormally found in fermented sausage (theyare, however, involved in vegetable fermenta-tions), this organism was introduced in theUnited States in the late 1950s as the first meatstarter culture. Strains of this species, whichwere later re-classified as either Pediococcusacidilactici or Pediococcus pentosaceus, arestill widely used today.

In addition, improvements in starter culturetechnology and the development of frozenconcentrated cultures (see Chapter 3) en-hanced culture viability such that Lactobacil-lus starter cultures are also now available. Ini-tially, the Lactobacillus strains that wereisolated from fermented sausage and used asstarter cultures were classified as Lactobacil-lus plantarum. Other strains of Lactobacillusthat had different physiological properties andthat performed well in sausage manufacturewere subsequently isolated. Thus, the closelyrelated species, Lactobacillus sake and Lacto-bacillus curvatus, are also used in starter cul-ture preparations.

Although the different Pediococcus and Lac-tobacillus strains used as starter cultures per-form the same basic role—to ferment sugarsand produce organic acid—they vary with re-spect to several important physiological andbiochemical properties (Table 6–3).These dif-ferences influence how they are used as startercultures. First, different species have differenttemperature optima and different thermal tol-

erances. For example, L. sake and L. curvatusare considered psychrotrophic, meaning theyare capable of growth, albeit slowly, at temper-atures as low as 4°C.Thus, they are suitable forfermentations conducted at cool ambient tem-peratures.

As discussed in Box 6–1, many of the Euro-pean sausages are fermented at low tempera-tures (20°C) to control the fermentation rate andto provide sufficient time for nitrate-reductionand subsequent color and flavor development.In contrast,P.acidilactici has a growth optimumnear 40°C and is preferred by manufacturers interested in fast, high temperature fermenta-tions. Metabolic differences also exist betweenthese organisms.Whereas all pediococci and lac-tobacilli ferment glucose, other sugars are fer-mented only by specific strains. Pediococci, forexample,do not ferment lactose. Sausage formu-lations must, therefore, account for the meta-bolic capacity of the culture.

There are occasions when starter culturemetabolism can lead to problems and defectsin the finished product.All pediococci are con-sidered to be homofermentative, producingonly lactic acid from glucose. However, undersome circumstances, such as during sugar limi-tation, heterofermentative products such asethanol, carbon dioxide, and acetic acid can beformed. Similarly, small amounts of these endproducts can also be produced by homofer-mentative L. plantarum during sugar limita-tion or aerobic growth. Of the heterofermenta-tive end products that are formed, acetic acidis especially undesirable in fermented sausagebecause it imparts a sour, vinegar-like flavorrather than the tart, tangy flavor contributed bylactic acid.

The ability of some strains to produce per-oxides also is a serious problem. Dependingon the starter culture strain and the level ofoxygen in the environment, hydrogen perox-ide can be formed directly in the fermentingsausage. Hydrogen peroxide can react withheme proteins in the muscle tissue to formundesirable, green pigments. In addition, hy-drogen peroxide and peroxide radicals pro-mote lipid oxidation, a serious flavor defect.



Table 6.3. Properties of bacteria used in meat start cultures1.

Minimum Temperature Acid from Nitrate PrimaryOrganism Temperature Optimum Glucose2 Reductase FunctionLactobacillus sakei 4°C 32°C — 35°C � � acidLactobacillus curvatus 4°C 32°C — 35°C � � acidLactobacillus plantarum 10°C 42°C � �3 acidPediococcus pentosaceus 15°C 28°C — 32°C � � acidPediococcus acidilactici 15°C 40°C � � acidKocuria varians4 10°C 25°C — 37°C � � flavor, aromaStaphylococcus carnosus 10°C 30°C — 40°C � � flavor, aromaStaphylococcus xylosus 10°C 25°C — 35°C � � flavor, aroma

1Adapted from Bergey’s Manual of Systematic Bacteriology, Volume 2, 8th Ed.2Under anaerobic conditions3Some strains reduce nitrate under low glucose, high pH conditions4Formerly Micrococcus varians

Box 6–1. Micrococcaceae, Nitrate, and “Old World” Sausage

In the United States, most fermented sausage starter cultures contain only lactic acid bacteria.The fermentation temperatures are high and fermentation times are short. In contrast, for manyEuropean products (and a few in the United States), the microbial flora contain species ofStaphylococcus, Micrococcus, and Kocuria in the Family Micrococcaceae (Hammes and Hertel,1998;Tang and Gillevet, 2003). In Europe, fermentation temperatures are low and fermentationtimes are long.

Not coincidently, nitrite salts are, by far, the most common curing agents used in fermentedsausage manufacture in the United States, whereas in Europe, nitrate salts are more common.These differences reflect not only the different manufacturing requirements,but also the desiredquality attributes relative to “new world”and “old world”sausage-making technologies and styles.

To achieve the expected color, flavor, and antimicrobial property peculiar to all cured meats(whether fermented or not), either nitrate or nitrite must be added as a curing agent. However,it is the nitrite form, and not the nitrate, that actually reacts with the meat pigments and pro-vides the curing effect. If nitrate is used as the curing agent, it must first be converted to nitrite(Figure 1). This conversion is simply a reduction reaction catalyzed by the enzyme nitrate reductase.

NO3- NO2

- H+NO NO-Mb Nitrosyl myochromogen

(Pink)Nitrate Nitrite Nitric oxide

Myoglobin (Mb)(Purple-red)

Nitrosylmyoglobin

heat

Nitratereductase

H2O2O2

H+

(Bright red)Gray, Green, Brown

Figure 1. Nitrate and nitrite reactions in meat. Nitrate is first reduced to nitrite by nitrate reductase, an en-zyme produced by strains of Micrococcus and Staphylococcus. Nitrite is further reduced to nitric acid, whichthen reacts with myoglobin at low pH to form nitrosyl myoglobin. When heated, nitrosyl myochromogen(also called nitrosyl hemochromogen) is formed, giving a pink color. The latter can be oxidized to form anundesirable gray, green, or brown appearance.

(Continued)



In general, bacteria that produce nitrate reductase are respiring aerobes that use nitrate as anelectron acceptor during growth under anaerobic or low oxygen conditions.Among the organ-isms having high nitrate reductase activity are species of Staphylococcus, Micrococcus, andKocuria. Thus, if the sausage formulation contains nitrate, then its conversion to nitrite will de-pend on the presence of nitrate reductase-producing strains of these bacteria, either those nat-urally present in the raw meat or added in the form of a starter culture. In the latter case, itmight seem rather odd to add staphylococci to food, since some coagulase-positive staphylo-cocci (e.g., Staphylococcus aureus) are pathogenic and produce exo-toxins. However, thestrains of Staphylococcus carnosus and Staphylococcus xylosus used for sausage fermentationsare coagulase-negative and non-toxigenic. Still, despite this caveat, the use of these strains insausage fermentations is not acceptable in some areas.This also points out why it is importantto accurately identify strains of Staphylococcus that might end up in a starter culture (Morot-Bizot et al., 2004).

If nitrate must be converted to nitrite to produce cured sausage, then the obvious question toask is why add nitrate in the first place? Why depend on the nitrate-reducing bacteria to per-form this important function? In other words, why wouldn’t all sausage manufacturers simplyadd nitrite as the curing agent, as is widely done in the United States? To answer these ques-tions, one must consider what other functions the micrococci and other nitrate-reducing bac-teria might perform in fermented sausage, other than to reduce nitrate. It turns out there areseveral.

The actual nitrate-to-nitrite reaction occurs rather slowly during the sausage fermentation be-cause the prevailing conditions (low temperature and low pH) are less than optimal for nitratereductase activity. It can take several weeks for the conversion of nitrate to nitrite.This affordsthe micrococci, kocuriae, and staphylococci the opportunity to secrete lipases and other en-zymes that ultimately generate flavor precursors in the meat. If, instead of ripening the sausage at20°C, it was fermented at 38°C to 40°C, the lactic acid bacteria would produce acid too quickly,and the non-lactic acid bacteria would be inhibited and unable to reduce nitrate or otherwise in-fluence the properties of the finished product. It is also argued that slow conversion of nitrate tonitrite enhances color development (although why this should be the case is not understood).

Among the flavor compounds produced by staphylococci,micrococci and kocuriae are meta-bolic end products resulting from protein and fatty acid metabolism. Most strains are lipolyticand many are also proteoyltic. In addition, amino acid metabolism may also generate flavor andaroma compounds.For example, it was recently reported that metabolism of leucine by Staphy-lococcus xylosus resulted in formation of various metabolites, such as 3-methylbutanol and 2-methylpropanol, that contribute to the flavor properties of fermented meats (Beck et al., 2004).

References

Beck, H.C., A.M. Hansen, and F.R. Lauritsen. 2004. Catabolism of leucine to branched-chain fatty acids inStaphylococcus xylosus. J.Appl. Microbiol. 96:1185–1193.

Hammes,W.P., and C. Hertel. 1998. New developments in meat starter cultures. Meat Sci. 49 (Supple. 1):S125–S138.

Morot-Bizot, S.C., R.Talon, and S. Leroy. 2004. Development of a multiplex PCR for the identification ofStaphylococcus genus and four staphylococcal species isolated from food. J. Appl. Microbiol.97:10871094.

Tang, J.S., and P.M. Gillevet. 2003. Reclassification of ATCC 9341 from Micrococcus luteus to Kocuria rhi-zophila. Int. J. Syst. Evol. Microbiol. 53:995–997.

Box 6–1. Micrococcaceae, Nitrate, and “Old World” Sausage (Continued)



Although some strains may also produce cata-lase or catalase-like enzymes that degrade hy-drogen peroxide, the production of these en-zymes cannot always be counted on toinactivate accumulated peroxides.

Protective Properties of Cultures

It has long been suggested that lactic acidstarter cultures were responsible for preserva-tion effects beyond their acidification and pH-lowering effects. During growth in meat, theyscavenge sugars and other nutrients faster thancompetitors, and they lower the oxidation-reduction potential (Eh) of the environmentsuch that growth of aerobic organisms is inhib-ited. In the 1980s it was discovered that someof the strains used as meat starter cultures mayprovide additional preservation effects via pro-duction of antagonistic agents that inhibitedother bacteria. Although some strains couldproduce inhibitory levels of hydrogen perox-

ide, experimental evidence indicated that asubstance other than hydrogen peroxide wasresponsible for the inhibitory effects. It wassubsequently discovered that these bacteriawere capable of producing a class of sub-stances known as bacteriocins.

Bacteriocins are proteinaceous substanceswith bactericidal activity, usually against bacte-ria that are closely related to the producer or-ganism.Among the bacteria used as starter cul-tures for fermented sausage, several specieshave been shown to produce bacteriocins, in-cluding strains of P. acidilactici, L. plantarum,and L. sakei. Moreover, several of these bacte-riocins—in particular the plantaricins andsakacins—have a somewhat broad spectrum ofactivity, inhibiting L.monocytogenes,S.aureus,Clostridium sp., and other Gram positivespoilage organisms. Greater application ofthese so-called bioactive or protective culturesis likely to increase (Box 6–2), perhaps leadingto their use in traditionally non-fermented

Box 6–2. Improving the Safety and Preservation of Fermented Meats Using Bioprotective Cultures

It has long been recognized that lactic acid bacteria have inhibitory activity against other bacteria,above and beyond that due to the lactic acid they produce.Some lactic acid bacteria, for example,also produce acetic acid, hydrogen peroxide, and diacetyl, all of which can inhibit potentialspoilage or pathogenic organisms.However, the antimicrobial compounds that have attracted themost attention are the bacteriocins.These agents are now being used, in a variety of delivery for-mats, in a wide range of foods, including fermented meats.Although fermented meats,by virtue oftheir intrinsic properties (organic acids, low pH,low aw,smoke,spices),are not prone to microbialspoilage or infrequently serve as vehicles for food borne pathogens, the addition of a bacteriocinto these products provides manufacturers with one more barrier and an extra margin of safety.

Bacteriocins are defined as proteinaceous compounds produced by bacteria that inhibit otherclosely related bacteria.They are distinguished from antibiotics in that bacteriocins are ribosomally-synthesized (rather than via de novo synthesis from precursor molecules), they have a more nar-row antimicrobial spectrum of activity, and their mode of action is mediated primarily via distur-bances of the cytoplasmic membrane. Although the bacteriocins relevant to this discussion areproduced by lactic acid bacteria, nearly every bacterial genus reportedly contains species capableof producing bacteriocins. In fact, so many bacteriocins have been described in the literature (aPubMed search on the term bacteriocin led to more than 2,762 published article hits since 1980,with more than 700 published since 2000), their classification has become a serious challenge.

In 1993, a four-class system, based on genetic, chemical, and physiological properties oflactic acid bacteria-produced bacteriocins, was described (Klaenhammer, 1993). Class I bac-teriocins are referred to as lantibiotics, because they contain the amino acid lanthionine.Nisin, the most well-studied bacteriocin, belongs to this class. Class II bacteriocins consist ofseveral small Listeria-active non-lanthionine peptides that are heat stable, making them good

(Continued)



antimicrobials for cooked sausage products.The large, heat-sensitive bacteriocins belong toClass III, and the complex, lipid- or carbohydrate-containing bacteriocins belong to Class IV(although whether this class actually exists is questionable). Most of the bacteriocins used infood applications belong to Class I or II.

Table 1. Application of bacteriocins in fermented sausages.

Product Producer(s) Target(s) Reference

Dry sausage Pediococcus acidilactici PAC 1.0 Listeria Foegeding et al., 1992

Salami Lactobacillus plantarum MCS1 Listeria Campanini et al., 1993

Dry sausage Lactobacillus sakei CTC 494 Listeria Hugas et al., 1995Fermented sausage Lactobacillus sakei CTC 494 Listeria Hugas et al., 1996

Lactobacillus curvatusLTH 1174

Fermented sausage Lactobacillus sakei LB 706 Listeria Hugas et al., 1996Lactobacillus curvatus

LTH 1174

Spanish-style dry sausage Enterococcus faecium CCM 4231 Listeria Callewaert et al., 2000Enterococcus faecium RZS C13

Dry sausage Staphylococcus xylosus DD-34 E. coli Lahti et al., 2001Pediococcus acidilactici PA-2 Listeria

Dry sausage Lactobacillus bavaricus MI-401 E. coli Lahti et al., 2001Lactobacillus curvatus MIII Listeria

Merguez sausage Lactococcus lactis M Listeria Benkerroum et al., 2003

Cacciatore sausage Enterococcus casseliflavus Listeria Sabia et al., 2003IM 416K1

North European- Lactobacillus rhamnosus Listeria Työppönen et al., 2003type dry sausage LC-705

Lactobacillus rhamnosusE-97800

Dry sausage Lactobacillus plantarum Listeria Työppönen et al., 2003

There are two main reasons why bacteriocins are used in meat products: (1) to preventgrowth of spoilage organisms and the production of spoilage end products, and (2) to inhibit orkill pathogens. Lactic acid bacteria, enterococci, Brochothrix, and clostridia are among thespoilage organisms inhibited by specific bacteriocins. However, control of pathogens and Liste-ria monocytogenes, in particular, is the main motivation for the use of bacteriocins in fer-mented meat products (E. coli O157:H7 is Gram negative and is largely unaffected by bacteri-ocins).The United States has a zero-tolerance policy for this organism in these products, so anymeasure that would reduce the frequency of a positive test result for Listeria would be veryuseful. It should also be noted that these same concerns about Listeria and other pathogenshave led manufacturers of non-fermented, ready-to-eat meat products (e.g., hot dogs, luncheonmeats) to consider the use of bacteriocins and bacteriocin-producing cultures, but under con-ditions in which fermentation does not occur or is minimized (Vermeiren et al., 2004).

Many of the bacteriocins produced by lactic acid bacteria are inhibitory or bacteriocidal to L.monocytogenes, including nisins A and Z (Class I) and pediocin PA-1 and sakacin A (Class II).Thelatter are produced by strains of Pediococcus acidilactici and Lacobacillus sakei, respectively.Because both P. acidilactici and L. sakei are ordinarily used as starter cultures for fermented

Box 6–2. Improving the Safety and Preservation of Fermented Meats Using Bioprotective Cultures (Continued)



meats, if the appropriate bacteriocin-producing strains were included in the culture blend, bac-teriocins could be produced in situ during the fermentation,as first demonstrated by Foegedinget al (1992). Many other examples have since been described (Table 1). In most of these experi-mental trials, however, the level of Listeria inhibition is modest, usually only one to two logs.

Although including bacteriocin-producing strains in a starter culture is a convenient way tointroduce bacteriocins into a product, there is no assurance that the bacteriocin will always beproduced on a consistent basis and at levels necessary to achieve the desired effect.Thus,otherapproaches have been considered to ensure that sufficient amounts are actually delivered. Forexample, a pure bacteriocin can either be added directly to the sausage batter or applied to thesurface in the form of a dip or spray. It is also possible to incorporate bacteriocins into packag-ing films. However, nisin is the only bacteriocin that has been purified and granted “generallyrecognized as safe”(GRAS) status (and only for specific applications).Purification processes andclearing regulatory hurdles are expensive activities, which explains why other bacteriocinshave not yet been commercialized.

An alternate strategy is to use the non-purified, pasteurized, and concentrated fermentationmaterial obtained following growth of a bacteriocin-producing strain in a food-grade medium.Products containing such a mixture, in either a dehydrated or paste form, contain active bacte-riocin along with the organic acids that were also produced. Importantly, these products(which are commercially available as shelf-life extenders), do not necessarily require regulatoryapproval, and are simply labeled as “cultured corn syrup”or “cultured whey.”

References

Benkerroum,N.,A.Daoudi, and M.Kamal.2003.Behavior of Listeria monocytogenes in raw sausages (mer-guez) in presence of a bacteriocin-producing lactococcal strain as a protective culture. Meat Sci.63:479–484.

Callewaert,R.,M.Hugas,and L.De Vuyst.2000.Competitiveness and bacteriocin production of enterococciin the production of Spanish-style dry fermented sausages. Int. J. Food Microbiol. 57:33–42.

Campanini,M., I.Pedrazzoni,S.Barbuti,and P.Baldini.1993.Behavior of Listeria monocytogenes during thematuration of naturally and artificially contaminated salami: effect of lactic acid bacteria starter cul-tures. Int. J. Food Microbiol. 20:169–175.

Chen,H.,and D.G.Hoover.2003.Bacteriocins and their food applications.Comp.Rev.Food Sci.Food Safety.2:82–100. (Available at www.ift.org/publications/crfsts.)

Foegeding, P.M.,A.B.Thomas, D.H. Pilkington, and T.R. Klaenhammer. 1992. Enhanced control of Listeriamonocytogenes by in situ-produced pediocin during dry fermented sausage production.Appl.Environ.Microbiol. 58:884–890.

Hugas, M., M. Garriga, M.T. Aymerich, and J.M. Monfort. 1995. Inhibition of Listeria in dry fermentedsausages by the bacteriocinogenic Lactobacillus sakei CTC494. J.Appl. Bacteriol. 79:322–330.

Hugas, M., B. Neumeyer, F. Pages, M. Garriga, and W.P. Hammes. 1996.Antimicrobial activity of bacteriocin-producing cultures in meat products. 2. Comparison of bacteriocin-producing lactobacillli on Listeriagrowth in fermented sausages. Fleichwirtschaft 76:649–652.

Klaenhammer,T.R. 1993. Genetics of bacteriocins produced by lactic acid bacteria. FEMS Microbiol. Rev.12:39–85.

Lahti, E., T. Johansson, T. Honkanen-Buzalski, P. Hill, and E. Nurmi. 2001. Survival and detection of Es-cherichia coli O157:H7 and Listeria monocytogenes during the production of dry sausage using twodifferent sausage cultures. Food Microbiol. 18:75–85.

Sabia, C., S. de Niederhausern, P. Messi, G. Manicardi, and M. Bondi. 2003. Bacteriocin-producing Enterococ-cus casseliflavus IM 416K1, a natural antagonist for control of Listeria monocytogenes in sausage(“cacciatore”). Int. J. Food Microbiol. 87:173–179.

Työppönen, S.,A. Markkula, E. Petäjä, M.L. Suihko, and T. Mattila-Sandholm. 2003. Survival of Listeria mono-cytogenes in North European type dry sausages fermented by bioprotective meat starter cultures.FoodControl 14:181–185.

Vermeiren, L., F. Devlieghere, and J. Debevere. 2004. Evaluation of meat born lactic acid bacteria as protec-tive cultures for the biopreservation of cooked meat products. Int. J. Food Microbiol. 96:149–164.

Box 6–2. Improving the Safety and Preservation of Fermented Meats Using Bioprotective Cultures (Continued)



sausages. For example, various research labora-tories (including the author’s laboratory) haveshown that P. acidilactici can be incorporatedinto frankfurter mixtures, and even in the ab-sence of fermentation,can reduce the L.mono-cytogenes population.

Micrococcaceae Cultures

Most meat starter cultures available in theUnited States contain species belonging to twogenera of lactic acid bacteria, Lactobacillus andPediococcus. In Europe,a quite different type ofstarter culture has been used. Most of the cul-tures used for European or European-style fer-mented sausages contain not only lactic acidbacteria,but also totally unrelated organisms be-longing to the family Micrococcaceae.These in-clude species of coagulase-negative Staphylo-coccus, Micrococcus, and Kocuria. In fact,whenlactic acid bacteria starter cultures were first in-troduced in the United States nearly fifty yearsago for sausage manufacture, the first Europeancultures contained only Micrococcus. These mi-crococci cultures are still available.

Whereas the main function of the lacticstarter culture—to produce lactic acid andlower the pH—is generally considered to beessential, the inclusion of Micrococcaceae inmeat starter cultures, while important, isstrictly optional.These bacteria are not fermen-tative and they produce no acid end products.Moreover, although they are metabolically ac-tive, they hardly even grow in the sausage.Rather, these bacteria are included in startercultures to convert nitrate to nitrite via expres-sion of the enzyme nitrate reductase.Along theway, they help form flavor and enhance color(Box 6–1). Is it important to note that becausethey are mesophilic, the fermentation tempera-ture must fall within the range of 18°C to 25°C.Thus, the lactic acid bacteria present in the cul-ture must also be capable of growth withinthis range. If fermentation were to occur athigher temperatures (e.g., 32°C to 40°C), therapid acid development would inhibit growthof the micrococci and the benefits they pro-vide would not be achieved.

In summary, then, there are at least five func-tions performed by meat starter cultures(Table 6–4).The culture must: (1) produce lac-tic acid and lower the pH; (2) produce desir-able flavors; (3) out-compete spoilage andpathogenic microorganisms for substrates andnutrients; (4) lower the Eh, since Salmonella,S. aureus, and other pathogens grow betteraerobically; and (5) in the case of the Micro-coccaceae cultures, enhance flavor and colordevelopment via reduction of nitrate.Althougha number of factors account for the excellentsafety record of fermented meat products (Box 6–3), the role of the culture in producingsafe, high-quality products cannot be over-emphasized (see below).

Principles of Fermented SausageManufacture

There are actually only a few general steps in-volved in fermented sausage manufacture.First, the ingredients are selected, weighed,mixed, and stuffed into casings. Second, thestuffed sausages are held under conditionsnecessary to promote a fermentation. Third,the sausage is subjected to one or more post-fermentation steps whose purpose is to affectflavor, texture, and preservation properties.These latter steps can range in duration fromas little as one week in the case of moist orsemi-dry sausages to more than two monthsfor very dry, strongly flavored sausages suchas Italian salamis.

Table 6.4. Desirable properties of meat starter cultures1.

Bacteria Fungi

Non-pathogenic Non-pathogenicNo toxins produced No toxins producedGrows well in meat Competitive at the surfaceStable Firm surface myceliumProduces good flavor Proteolytic and lipolyticNitrate/nitrite resistant Moldy aromaSalt-tolerantBioprotectiveEasy to identify

1Adapted from Hammes and Knauf, 1994



Box 6–3. Pathogens, Toxins, and the Safety of Fermented Sausage

Given that fermented sausages are made from raw meat and many are never heat processed orcooked, it is not unreasonable to question the microbiological safety of these products. Theemergence of several serious foodborne pathogens in the 1980s, such as enterohemorrhagic Escherichia coli (EHEC) O157:H7 and Listeria monocytogenes, has raised additional concerns,since these organisms are frequently present in raw meat, have high mortality rates, and aremore hardy than other food borne pathogens.

Of course, if not properly manufactured, even traditional foodborne pathogens, such asStaphylococcus aureus, Clostridium botulinum, and Salmonella, can present significant foodsafety risks in fermented sausages. Finally, there are several food safety hazards other than thosecaused by bacterial pathogens. Parasites, fungi, viruses, and other agents can potentially conta-minate meat and sausage products. In addition,an unusual but not uncommon form of food poi-soning is caused by the presence of biogenic amines that are produced by bacteria ordinarilypresent in fermented sausage.

Despite these concerns, manufacturers of fermented sausages have long considered theseproducts to be safe, and they have generally withstood the test of time. Since 1994, there havebeen few food poisoning outbreaks associated with fermented meat products (Table 1).Theirsafety is undoubtedly due to the presence of multiple antimicrobial barriers or hurdles that ex-ist in fermented sausages (Table 2). After all, for a given organism to survive and grow in theseproducts, it would have to overcome each of these individual barriers.That is, at minimum, itwould have to be nitrite-resistant, acid-resistant, salt-resistant, and osmotolerant. Moreover,when barriers are combined, there is a synergistic effect, such that the net antimicrobial effectof multiple barriers is greater than would be predicted based on their singular effects. For ex-ample, a strain of Salmonella may be fully capable of surviving a pH of 5.2, provided that allother conditions are optimal for growth.But if another hurdle is put in place (e.g., the water ac-tivity is reduced from 0.99 to 0.97), then pH 5.2 may be sufficient to inhibit this organism.

Table 1. Ten-year safety record for fermented meats.

Year Country Product Cases Organism

1994 USA Dry-cured salami 23 Escherichi coli 0157:H71995 Australia Mettwurst 23 Escherichi coli 0111:H�

1995 Germany Teewurst �300 Escherichi coli 0157:H�

1998 Canada Genoa salami 39 Escherichi coli 0157:H71999 Canada Sausage 155 Escherichi coli 0157:H72001 USA Fermented beaver 3 Clostridium botulinum

Table 2. Antimicrobial barriers in fermented meats.

Property Level, range, or function

pH 4.5–5.5aw 0.7–0.9salt 2–4%Eh ReducedAcids 1–2%Competition ExclusionNitrite 125 ppmCasing Exclusion

(Continued)



The problem, however, is now more complicated, because it has been recognized recentlythat some pathogens appear to be tolerant even to multiple barriers. Some strains of E. coliO157:H7, for example, are much more tolerant to low pH and organic acids than are normal E.coli strains. Furthermore, L. monocytogenes is resistant to low pH, low water activity (aw), highsalt, and nitrite, and can even grow at refrigeration temperatures.And although there have beenfew food poisoning outbreaks caused by these organisms in fermented sausage, challenge ex-periments indicate that E. coli O157:H7, L. monocytogenes, and other pathogens can theoreti-cally survive typical sausage manufacturing procedures.Thus, even the mere presence of theseorganisms in ready-to-eat meat products would likely initiate a recall. Indeed, the U.S. Depart-ment of Agriculture has established rules that now require a minimum five-log reduction ofpathogenic organisms during the manufacture of uncooked, ready-to-eat fermented sausage.Manufacturers are also required to develop and implement HACCP plans that describe the rele-vant intervention steps necessary to produce safe products.

Of course, the simplest solution to many of these potential problem organisms and the most ef-fective way to ensure a greater than five-log reduction would be to include a heating step some-where during the process. However, many manufacturers are unwilling to accept the changes inthe sensory properties caused by heating, and instead rely on ensuring that natural barriers aresufficient. It is also possible to include additional intervention measures into the product orprocess. For example, meat starter cultures capable of producing bacteriocins and other antago-nistic agents that are inhibitory to pathogens are now available (discussed previously).

As noted above, bacterial pathogens are not the only food safety issue of concern to the fer-mented meats industry. Parasites and viruses can also be present in raw meat, and toxigenicfungi can contaminate sausage during fermentation and ripening.The parasite most commonlyassociated with meat, and pork in particular, is the tapeworm Trichinella spiralis, the causativeagent of trichinosis. Although certified Trichinella-free pork is available in the United States(and in Europe), it is also possible to inactivate this nematode by a freezing treatment, as perUSDA guidelines. Otherwise, a cooking step is required to destroy the cysts.

In contrast to parasites, viruses have not been considered as a serious food safety problem infermented meats, and this is still largely true.Viruses found in meat are not usually pathogenic tohumans, they do not replicate, and they are mostly sensitive to the acidification and drying stepsused in sausage manufacture. Although not a human pathogen, the spread of the foot-and-mouthvirus in the United Kingdom and Europe during 2001 had devastating economic and social ef-fects on all segments of the meat processing industry,even though it posed no threat to humans.

This situation was different from that which occurred in the 1980s and 1990s when “madcow disease” (or bovine spongiform encephalopathy) appeared in the United Kingdom. In thelatter case, the causative agent was an unusual type of infectious protein called a prion that oc-curred in beef and that was fully capable of causing an extremely rare but fatal disease.Theprion is not destroyed by fermentation, heat, or other food processing techniques, so the onlymeans of control is to prevent the transmission of the prion during animal production.

Several different genera of fungi, including species of Penicillium and Aspergillus, are fre-quent contaminants of fermented meats. For many products, including both whole fermentedmeats (i.e.,hams), as well as fermented sausages, their growth is encouraged,due to their abilityto produce flavor-generating enzymes (as discussed previously). However, some strains isolatedfrom mold-fermented products are capable of producing mycotoxins. Moreover, inoculation ofham and sausage with toxigenic fungi and incubation under optimized conditions results intoxin formation. Despite these findings, the presence of these toxins in mold-fermented meatproducts appears to occur rarely, if at all, a situation not unlike that for fungal-ripened cheeseand other mold-fermented products.

Still, there is much concern about the potential for mycotoxin production in fermentedmeats, and there is now a trend to use defined, nontoxigenic strains rather than the wild or

Box 6–3. Pathogens, Toxins, and the Safety of Fermented Sausage (Continued)



house strains that have commonly been used. For example, Penicillium nalgiovense, a fungalmeat starter culture,has many desirable properties,but since some strains produce mycotoxins,only strains demonstrated to be non-toxin producers are in commercial use. Undesirable moldgrowth on sausage can also be controlled by antimycotic agents such as sorbic acid and, if per-mitted, the antibioic pimaricin. Smoke, which is usually applied to high moisture, but not dryproducts, also contributes antimycotic constituents.

Finally, another biologically-active group of microbial end products found in a wide variety offermented foods, including fermented meats, are referred to as biogenic amines (Suzzi and Gar-dini,2003).These compounds are formed via decarboxylation of amino acids by various bacteriathat are commonly found in fermented meats, including lactic acid bacteria, enterococci, Enter-obacteriaceae, and Micrococcaceae. In some individuals, ingestion of biogenic amines results ina particular food poisoning syndrome marked by headache,nausea,and dilation of blood vessels.

In fermented meats, the most common biogenic amine capable of causing food poisoningsymptoms is tyramine,derived from the amino acid tyrosine. In general,dry fermented sausagescontain about 50 to 300 mg tyramine per kg (Table 3). Histamine, derived from histidine, mayalso be present, but usually at ten-fold less concentration. However, only when concentrationsare very high (�1,000 mg per Kg of dry weight) do these products pose a health risk in mostnormal individuals.

Table 3. Biogenic amines in European fermented dry sausages1,2.

Product (Country) Tyramine Histamine Cadaverine Putrescine

Soppressata (Italy) 178 22 61 99Salsiccia (Italy) 77 0 7 20Sobrasada (Spain) 332 9 13 65Fuet (Spain) 191 2 19 72Salchichón (Spain) 281 7 12 103Chorizo (Spain) 282 18 20 60

1Adapted from Suzzi et al., 20032Concentrations given in mg/Kg

Other biogenic amines may also be formed in meat at appreciable levels, including pu-trescine and cadaverine, but these compounds do not generally elicit symptoms describedabove. Since formation of biogenic amines requires the presence of free amino acids, theamount produced depends on the extent of protein hydrolysis that had occurred in the food.Thus, the longer the meat is aged or fermented, the higher will be the concentration of aminoacid substrates and the more likely it is that the product will contain biogenic amines (assumingthe relevant decarboxylating enzymes are also present).

Because some lactic acid bacteria have the ability to produce amino acid decarboxylases,starter cultures strains should be screened to eliminate such strains from use. Moreover, it maybe possible to use starter cultures that have inhibitory activity against potential bioamine pro-ducing bacteria, either by virtue of their ability to produce acids rapidly, produce bacteriocins,or outcompete them for nutrients. Finally, some staphylococci and lactic acid bacteria produceamine oxidases, enzymes that cause oxidative deamination of amines, and therefore, act to de-toxify biogenic amines.

References

Suzzi, G., and F. Gardini. 2003. Biogenic amines in dry fermented sausages: a review. Int. J. Food Microbiol.88:41–54.

Box 6–3. Pathogens, Toxins, and the Safety of Fermented Sausage (Continued)



Ingredients

It is entirely possible to manufacture a fer-mented sausage with just a handful of ingredi-ents. In fact, only five ingredients are essential:meat, sugar, salt, culture, and a curing agent.Asdiscussed above, fermented meats can indeedbe made without adding a culture, but mostlarge-scale manufacturers would not dream ofmaking product without a culture.Likewise, fer-mented sausages also can be manufactured inthe absence of the curing agent.However, theseagents, either in the form of nitrite or nitrate,perform such important microbiological andorganoleptic functions that they are nearly uni-versally used. That being said, there is a small(but growing) market among organic foods pro-ponents for reduced or even nitrite-free meatproducts. Even if the organoleptic propertiesprovided by nitrite could be provided by otheragents (a big if), removing nitrite from curedmeat products would expose these products toa potentially serious food safety threat. Still, pro-vided that other barriers are in place, especiallylow pH and low temperature, theoretically safeproducts can be produced.

Meat

Of the ingredients listed above, the main ingre-dient is obviously the meat, which contributes

uct matrix, but also the fat, which providesmuch of the flavor.The fat-containing cuts usu-ally are chopped or ground separately from theleaner portions to impart a desired appearanceand flavor.The grind also affects texture and ac-cordingly determines the type of product. Forexample, some sausages (e.g., Plockworst)have large visible fat particles, whereas others(e.g., cervelat) are ground to a fineness suchthat the fat particles are so small as to be indis-tinguishable from the sausage matrix.

The fat and lean portions may even be de-rived from different animals. Beef fat containsmore unsaturated lipids than pork fat, and ismore susceptible to oxidation reactions thatmay result in undesirable rancid flavors.Thus,

many sausage products, such as the popularU.S. product summer sausage, are typicallymade with mixtures of beef and pork. Cuts arealso important—shanks, chucks, and bull meathave binding properties that are especially im-portant in sausage manufacture. Obviously,there is a trend to use less expensive cuts, buthigh quality meats are often still used.

Sugar

The next essential ingredient is the sugar orcarbohydrate.Although glucose, in the form ofthe polymer glycogen, is initially present inmuscle tissue of slaughtered animals, glycogenstores are quickly depleted during the post-mortem period.Thus, fresh meat contains littlefermentable sugar, and addition of sugar isnecessary. Glucose is most common in theUnited States, and is added to about 1% to 2%of the total batter weight. Since the amount ofacid produced by the lactic culture is directlyrelated to the amount of available glucose, thesugar concentration in the batter can be ad-justed, in general, to give a particular final pH.Also, higher sugar levels promote faster fer-mentations, which are preferred in the UnitedStates. In contrast, many European fermentedsausage manufacturers prefer less tanginessand more diverse flavor development.Achiev-ing these characteristics require slower fer-mentation rates, thus less rapidly fermentablesugar is added (as little as 0.1% to 0.2%).

Salt

Salt is an essential ingredient in all types ofsausage products (fermented or not). Salt,added in concentrations of 2.4% to 3%, per-forms several critical functions. First, it is re-sponsible for extracting and solubilizing themuscle proteins, which are ordinarily in an in-soluble form. Once extracted and solubilized,the proteins form a “sticky” film around themeat particles,creating an emulsion-type struc-ture. Second, salt provides flavor. Finally, salt isthe primary means, at least initially, for control-ling the microflora. Although salt at a concen-


not only the protein and the bulk of the prod-


tration of even 3% might not appear to be suf-ficiently high enough to inhibit many organ-isms, the actual concentration within the aque-ous phase is considerably higher.

Culture

Most commercial cultures for sausage are sup-plied in either a frozen or lyophilized form.Frozen cultures, which are more common inthe United States, are supplied as thick slurriesin peel-back or flip-top cans ranging in sizefrom 20 ml to 250 ml. Cell densities typicallyrange from 108 to 109 cells per ml.A typical 70-ml can is sufficient for about 150 kg of sausagebatter.These cultures are shipped frozen underdry ice and users are instructed to store thecans at �40°C or below. The cans should bethawed in cold water prior to use.

Proper handling of frozen cultures is ab-solutely necessary to maintain culture viabilityand to ensure that culture performance (i.e.,rapid fermentation) is not impaired.Lyophilizedcultures, which are less commonly used, havethe advantage of not requiring low-temperaturestorage.They are stable at refrigeration or evenambient temperatures.As free-flowing powders,they are easily measured and distributed intothe batter.These cultures,which can contain upto 1010 cells per gram, are usually more expen-sive,however, than frozen cultures.

Curing agents

Finally, with few exceptions, fermented meatproducts include nitrite or nitrate as curingagents.These are added as either the sodium orpotassium salt. Although nitrite salts are nowused far more frequently, until the 1970s, ni-trate salts were more common.For reasons dis-cussed previously, some sausage manufactur-ers, still prefer nitrate. In any case, nitrite isadded at a maximum of 156 ppm for dry andsemi-dry sausages. Since a single ppm trans-lates to 1 gram per 1,000 kg, only 156 grams(about one-third of a pound) are all that can beadded to a 1,000-kg batch of sausage (morethan 2,200 pounds). Despite this relatively

small amount, nitrite performs a number of im-portant microbiological and organoleptic func-tions. In fact, without nitrite (or nitrate), fer-mented meat products would be far lesspopular and considerably less safe to eat.

Nitrite is mainly added to sausages (and notjust those that are fermented) because of its ef-fectiveness as an antimicrobial agent. In partic-ular, nitrite inhibits the out-growth of C. botu-linum spores (Box 6–4), making it one of themost powerful anti-botulinum agents availableto the processed meats industry.Although fer-mented meats contain combinations of or-ganic acids and salt, both effective antimicro-bials, neither provide a sufficient degree ofinhibition against this organism. It should beemphasized, however, that nitrite alone, at thelevels currently used, does not entirely inhibitC.botulinum.Rather, it is the combined effectsof nitrite along with organic acids, low pH,andlow aw that effectively control the growth ofthis organism during the manufacture and stor-age of fermented sausage.

The other reasons for adding nitrite to fer-mented meats are related to the organolepticproperties this agent imparts.Nitrite fixes color,acts as an antioxidant and prevents a warmed-over flavor, and imparts a desirable cured meatflavor.

Spices, flavoring, and other ingredients

A wide variety of spices, seasonings, and otherflavoring agents are often added to fermentedsausages. These include pepper (black andred), paprika, garlic, mustard, mace, and car-damom.These flavorings can be added in theirnatural form or as extracts. Levels vary, de-pending on the nature of the product and con-sumer preferences. In general, moist, smokedsausages that are popular in Germany andnorthern Europe are only slightly spiced,whereas the dried, non-smoked products con-sumed in southern Europe (e.g., Italy andSpain) and other Mediterranean regions aremore heavily spiced (Table 6–1). Among theoptional ingredients commonly added to fer-mented meats, ascorbate and erythorbate are



perhaps the most important. Both providesimilar functional roles as nitrite in that theyinhibit autooxidation and increase color andflavor intensity.

Manufacture of Fermented Sausage

Cutting and Mixing

Once the ingredients have been collected andweighed, the process is remarkably simple (Fig-ure 6–1).The meat portions (lean and fat) areusually ground separately in a silent cutter—arotating, bowl-shaped device that chops andmixes the sausage batter—or a similar grindingdevice to produce varying degrees of coarse-

ness.The meat, along with all of the remainingingredients, are then combined in a silent cut-ter. Mixing should be done to minimize or ex-clude oxygen, which not only can interferewith color and flavor development, but also isless conducive to growth of the starter culture.Above all,mixing,as well as all the grinding andweighing, must be performed at low tempera-tures for both quality and safety.The USDA re-quires that the temperature in the processingroom be 40°C or below. In fact, in the UnitedStates, all meat processors must have a HazardAnalysis Critical Control Points (HAACP) plan,with the temperature during processing listedas a critical control point (Box 6–5).

Box 6–4. Nitrite as an Anti-botulinum Agent

Using nitrate and nitrite salts as ingredients in fermented and non-fermented sausage is a ratherrecent practice, adopted only within the last century. It is likely, however, that sausage productshave long contained nitrate salts, due to the contamination of salt with potassium nitrate, in theform of saltpeter (Cammack et al., 1999).

The development of the desirable cured meat color and flavor by nitrate, and more specifi-cally, the nitrate reduction product, nitrite, must have certainly led to widespread use of salt-peter in meat processing.However, the discovery that nitrite also had inhibitory activity againstClostridium botulinum eventually led to the routine addition of nitrite salts in these products.The physiological mechanisms by which this organism is inhibited by nitrite,however,have yetto be established, although several possible explanations have been proposed (reviewed inCammack et al., 1999 and Grever and Ruiter, 2001).

As noted previously (Box 6-1), when nitrite (NO3–) is added to foods, it is quickly converted

to nitric oxide (NO) via dissimilatory nitrite reductases produced by the indigenous flora. It isalso possible that chemical reduction of nitrite to nitric oxide can also occur.Nitric oxide is a re-active species, especially in the presence of metal ions.When nitric oxide-metal ion complexesare formed, the biological and chemical availability of those metals is decreased.Thus, micro-organisms would be unable to acquire iron and other essential minerals from the medium.

In muscle systems, iron is present in the form of the heme-containing proteins, myoglobin,oxymyoglobin (oxidized), and metmyoglobin (reduced).When these proteins react with nitricoxide, the resulting nitrosyl myoglobin products give nitrite-cured meats their characteristiccolor.The strong binding of heme iron (which is enhanced by heating) is considered to be oneof the factors responsible for nitrite-mediated inhibition of Clostridium botulinum and otherbacteria.

Not only does nitric oxide react with heme iron, but it also binds to other iron- and iron-sulfur-containing proteins and enzymes. In particular, clostridia depend on ferridoxin and theenzymes pyruvate-ferridoxin oxidoreductase and hydrogenase for pyruvate metabolism andATP generation (Figure 1).The nitrosyl-protein complexes that are formed in the presence of nitric oxide block pyruvate metabolism and subsequent ATP synthesis.The suggested net result,therefore, is the inhibition of outgrowth by germinated spores.



Stuffing

After the batter has been sufficiently mixed,it ismoved to a stuffer,a device that pumps the mixinto casings. The casings are essentially longtubes that give the product its characteristicshape. The diameter of the casings can varyfrom less than 1.5 cm to more than 9 cm.As thetubes are filled, they are tied off or cut to givedesired section lengths,again depending on theproduct being made. Lengths can vary from 5cm to 100 cm. Shape and diameter size are im-portant,not only because they are specific for agiven product, but more importantly, because

they influence the rate of drying, cooking,smoking, and ultimately the flavor and textureof the finished product. Casings must be per-meable to both moisture and smoke.

Two general types of casing materials areused, natural and synthetic. These terms aresomewhat nebulous, since the latter are oftenmade from natural sources. Traditionally, cas-ings were made of animal intestines.Althoughthere is still a significant market for productsmade using natural casings, especially in Eu-rope, synthetic casings have several advantagesare now widely used.They are usually made us-ing cellulose or collagen.

References

Cammack, R., C.L. Joannou, X.-Y. Cui, C.T. Martinez, S.R. Maraj, and M.N. Hughes. 1999. Nitrtie and nitrosylcompounds in food preservation. Biochim. Biophys.Acta. 1411:475–488.

Grever, A.B.G., and A. Ruiter. 2001. Prevention of Clostridium outgrowth in heated and hermeticallysealed meat by nitrite—a review. Eur. Food Res.Technol. 213:165–169.

White, D. 2000. The physiology and biochemistry of prokaryotes, second edition. Oxford UniversityPress, Inc., New York.

Box 6–4. Nitrite as an Anti-botulinum Agent (Continued)

pyruvate acetyl CoA

Fdox Fdred

CoA

2H+H2

pyruvate-ferridoxinoxidoreductase

hydrogenase

acetyl-P

Pi CoA

phospho-transacetylase

ADP

ATP

acetatekinase

acetate

CO2

Figure 1. Metabolism of pyruvate to acetate in clostridia. The “phosphoroclastic” conversion of pyruvate toacetate is mediated by several iron-containing proteins. Pyruvate-ferridoxin oxidoreductase oxidizes pyruvate,using ferridoxin (Fd) as the electron acceptor. Reduced ferridoxin is then re-oxidized by a third iron-containingprotein, hydrogenase. Protons serve as the electron acceptor, and hydrogen gas is formed. Phospho-transacetylase and acetate kinase catalyze the final two steps, with acetate and one mole of ATP formed asend products. Importantly, both pyruvate-ferridoxin oxidoreductase and ferridoxin are able to form complexeswith nitric oxide. This would effectively block this pathway, depriving cells of ATP, as well as increasing theconcentration of pyruvate to potentially toxic levels. Adapted from White, 2000.



Fermentation

Once the sausage batter is packed into casings,the material is moved into specially designedripening chambers where the fermentation oc-curs. These facilities, often referred to as thegreen room or smoke house or simply thehouse have controls for maintaining tempera-ture, humidity, and air movement. Moreover,control systems in modern facilities are fullyprogrammable so conditions can be ramped oradjusted depending on how the fermentationis proceeding or on the particular specifica-tions of the manufacturers. Thermocouplesand pH probes inserted directly into productsamples can feed the appropriate informationinto a computer to provide constant monitor-

ing, record-keeping, and feed-back control.Thus, the entire fermentation can proceed inthe absence of a full-time operator.

Fermentation parameters vary dependingon the culture and the desired product quali-ties. In general, lower incubation temperaturesrequire longer fermentation times. For exam-ple, at the low temperature range of 21°C to24°C (70°F to 75°F), which is very near ambi-ent, fermentation can take as long as two tothree days. At 29°C to 32°C (85°F to 90°F),twelve to sixteen hours of fermentation will berequired. In the United States, where fasteroverall production times are preferred, the in-cubation temperature can be as high as 37°Cto 40°C (98°F to 102°F) for as little as twelve toeighteen hours.

Grind

Mix

Stuff

Ferment

Dry / Smoke / Cook

To desired coarseness

22-40°C

MeatPork

Beef

SaltGlucoseNitrate/NitriteAscorbateSpicesCulture

Semi-Dry DryMoist

Figure 6–1. Manufacture of fermented sausage.



Box 6–5. Ensuring Meat Safety Through HACCP

In July 1996, the Pathogen Reduction/Hazard Analysis and Critical Control Point system becameregulatory law for all USDA-regulated facilities.This system,abbreviated as HACCP,was designedto apply scientific principles to process control in food production.

The processing steps for manufacture of a representative fermented sausage product are il-lustrated in a flow diagram (Figure 1).Through careful analysis all hazards that are consideredreasonably likely to occur are then identified (Table 1). Control measures designed to preventhazards are applied at specific points in the manufacturing process; these are called critical con-trol points.

Receiving packagingmaterials

Receiving non-meatmaterials

Storage of non-meatfood ingredients

Weighing non-meatfood ingredients

Receivingstarter

cultures

Storageof startercultures

Prepare starterculture

Rework

FermentingCCP #3

DryingCCP #4

Slicing/peelingCCP #5

Packaging andlabeling

Finished product storage

ShippingStorage of

packaging materials

Receiving raw meatCCP #1

Cold storage of frozen andrefrigerated meat

CCP #2

Tempering frozenraw meat

Weighing Raw Meat

Combine Ingredients

Processing: chopping,grinding, mixing, stuffing

Receivecasings

Storecasings

Preparecasings

Figure 1. Generic process flow diagram for fermented dry sausage. The main critical control points (CCP)are shown.

(Continued)



Since individual culture strains may have dif-ferent temperature optima and tolerances, se-lection of cultures that perform under the se-lected conditions is critical. It is also important,for some applications, that the fermentationrate not proceed too fast.This is particularly rel-evant when flavor- or color-producing organ-isms (e.g., Micrococcus) are used, because theymay be inhibited by fast lactic acid producers.However, if the fermentation is too slow, theremay be sufficient opportunity for pathogens orspoilage organisms to grow. Ultimately, the pH

at the end of fermentation should be less than5.1, which, by itself, will not make the productshelf-stable, but which does provide a reason-able protective barrier against most foodbornepathogens. In addition, fermented sausages atpH 5.0 to 5.1 will have only a slight tart flavorthat may actually be indistinguishable fromnon-fermented products.

Alternatively, depending on the culture, theincubation conditions, and the amount of sub-strate (i.e., fermentable sugar) provided, muchhigher acidities can be achieved.Typically, sum-

In the fermentation step, for example, the pH must attain 5.3 or below within six hours.Con-trol of this step may have the greatest impact on assuring the safety of the final product.Criticallimits are established and then continuously monitored and recorded to verify that the processis under control.

Corrective actions are developed for each critical control point and documented in theHACCP plan.These corrective actions are applied whenever deviation from a critical limit oc-curs; specific and exacting records are kept.A separate HACCP program is designed and docu-mented for each manufacturing process used in an establishment.

The HACCP method accompanied by proper prerequisite programs such as Good Manufac-turing Practices and Sanitation Standard Operating Procedures can be an effective means of as-suring the safety and quality of food products.

Table 1. Identification of critical control points during the manufacture of fermented semi-dry sausage.

Critical Control Point Hazard Reasoning Control Measure

Receiving Raw Salmonella Reasonably likely to be present in Supplier certification to guarantee Meat E. coli 0157:H7 raw meat ingredients laboratory analysis indicating

Listeria absence of Salmonella and monocytogenes E. coli 0157:H7

Cold Storage Salmonella Salmonella and E. coli 0157:H7 Refrigeration temperature not to E. coli 0157:H7 may grow if temperature is not exceed 40°F; freezer

maintained below minimum temperature not to exceed 30°Fgrowth requirement

Fermentation Staphylococcus Potential for growth and toxin Must attain pH of 5.3 or below aureus production due to fermentation within 6 hours

failure

Drying Salmonella Potential for growth and toxin Attain established moisture to Staphylococcus production due to drying protein ratio (1.6:1 for

aureus failure pepperoni and 1.9:1 for sausage)

Slicing/peeling Listeria Potential for contamination from Sanitizer effective against monocytogenes environment or handling L.monocytogenes applied to

food contact surfaces every 4 hours

Box 6–5. Ensuring Meat Safety Through HACCP (Continued)



mer sausage has a pH around 4.9 but can be aslow as 4.7.When the pH reaches 4.8 or less, adefinite “tangy”flavor becomes apparent.

Fermentation rooms are also equipped withair movement devices and humidistats to main-tain the relative humidity (RH) at desired lev-els. Since the RH in the atmosphere surround-ing a food directly influences the water activityof that food (where aw � 100 � RH), as RH isdecreased, the more moisture is lost to the at-mosphere and the lower will be the aw in thesausage.Typically, the RH should be about 10%less than the aw � 100 of the finished product.

Cooking, drying, and smoking

Several different treatments and combinationsof treatments can be applied at end of the fer-mentation.These include cooking, drying, andsmoking. In the United States, fermentedsausages are often cooked after fermentation,whereas, in Europe and elsewhere, rawsausages are the norm, and post-fermentationheating steps are rarely applied.

In general, properly made dry, fermented,uncooked sausages, like salami and pepperoni,are still considered to be shelf-stable and ready-to-eat. Cooking, however, does provide impor-tant advantages. Cooking not only inactivatesthe culture and stops the fermentation, but italso kills pathogenic microorganisms that mayhave been present in the raw meat. Thus, thecooking step, at least in the United States, maybe an important component of HACCP pro-grams (Box 6–5) used by sausage manufactur-ers, since it serves as a terminal process stepproviding manufacturers (and consumers)with reasonable assurance of product safety.

Furthermore, in the United States, any pork-containing sausage must, by federal regula-tions, be cooked to destroy Trichinae, a nema-tode that infects swine and that can cause thedisease Trichinella in humans. The USDA hasdeveloped various time-temperature regimensthat are effective (e.g., 58.3°C), and which caneasily be implemented as part of the cookingprocess. Exceptions are allowed if the pork iscertified as Trichinae-free or if the pork is

frozen according to the USDA guidelines forfreeze-inactivation of the Trichinae cysts.

If a cooking step is included, the productcan be moved to a separate chamber or, as ismore common, the fermentation chamber it-self is equipped with heating capability. Thetemperature is slowly raised until the desiredinternal temperature (usually 60°C to 62°C) isachieved. Following cooking—or fermenta-tion, if there is no cooking step—the productcan be smoked and/or dried. In general,Mediterranean sausages (e.g., Genoa and Mi-lano) are dried, but not smoked, and northernsausages (e.g., German) are smoked, but notdried. In contrast, some semi-dry products,such as summer sausage, are slightly smoked.

If the product is to be dried, the chamberenvironment is set at about 7°C to 13°C (45°Fto 55°F) and 70% to 72% RH. Good air move-ment is necessary to rapidly remove water va-por and any condensate that collects at the sur-face.The rate of drying is also critical. If the RHis too low and the temperature too high, dry-ing will initially be rapid. However, at thesehigh drying rates, the surface will become de-hydrated and form a hard, water-impermeableskin.This phenomenon, called case hardening,results in slower drying and poor product qual-ity, since water molecules are unable to diffusethrough the hardened surface and are trappedwithin the sausage interior.

Drying times depend on product specifica-tions and desired quality characteristics. Obvi-ously, large diameter products require longerdrying times than small diameter products.Thelonger the drying, the more water is lost, andthe lower will be the final water activity. Semi-dry products are typically dried to removeabout 20% to 30% of the original water and togive a final moisture of about 45% to 50%.The aw of semi-dry products ranges from 0.90to 0.94. Such products may be stable at ambi-ent temperature, with a shelf-life of aboutthirty days. Examples include Thuringer, cerve-lat, Lebanon bologna, summer sausage, andsemi-dry salami. Dry fermented sausages willlose about 35% water, giving a final moisture of35% and an aw between 0.85 and 0.91. Dry



products can be shelf-stable for seventy-fivedays or more. Pepperoni and salami, especiallythe Italian, Mediterranean, and eastern Euro-pean varieties,are the most common examples.

Most modern facilities are equipped withprogrammable systems that allow operators toset fermentation and cooking parameters, aswell as air movement, smoking, and dryingconditions. Recording devices provide arecord of these conditions during the manufac-turing process. Thus, once the sausages areloaded into the ripening chamber, they can beleft on their own, more or less, until theprocess is completely finished.

Mold-ripening

Many of the European-style sausages areripened by mold. These products are particu-larly popular in Hungary and Romania, as wellas throughout the Mediterranean region. Formany of these products, fungal growth can beextensive,with the mycelia covering the entiresurface. Mold-fermented sausages are notnearly as common in the United States. How-ever, there are some whole meat products, inparticular, country-cured hams, that are fer-mented by yeasts and fungi and that are verypopular in various regions of the United States.Although technically not fermented, aged meatowes its improved flavor and texture proper-ties, in part, to growth of surface fungi, in par-ticular, Thamnidium elegans.

The role of fungi in fermented whole meatsand sausages is similar to that of other mold-fermented products.That is, the fungi are notinvolved in the primary lactic acid fermen-tation, but rather their function is to enhanceflavor and texture properties. This is accom-plished by production and secretion of pro-teinases and lipases and their diffusion into themeat, generating flavor and aroma products ortheir precursors.Ammonia, released from pro-tein and amino acid metabolism, and ketonesand other rancid and oxidized flavor notes, de-rived from fatty acid metabolism, are amongthe end products that accumulate in mold-ripened sausages. In this respect, the fungi pro-

duce organoleptic properties not unlike thatfound in mold-ripened cheese.

The manufacture of mold-ripened sausages issimilar to that of normal fermented sausage, inthat ingredients are mixed and stuffed into cas-ings. Inoculation of sausages or whole meatswith fungi can occur either naturally, via thefungal spores present in the manufacturing en-vironment, or by dipping or spraying the prod-uct with a defined spore suspension. Salt ap-plied to the surface of country-cured and Parmahams reduces the water activity and provides aselective environment for resident yeasts andfungal organisms. Yeast and fungal starter cul-tures, which are becoming more widely avail-able for fermented sausage and ham manufac-ture, are generally comprised of strains ofDebaryomyces hansenii, Penicillium nalgio-vense, Penicillium camemberti, and Penicil-lium chrysogenum. The main problems withnaturally-occurring fungi are (1) they may yieldproducts with inconsistent quality; and (2) per-haps more importantly, they may also producemycotoxins.Thus, the use of pure fungal startercultures, selected on the basis of their func-tional properties as well as their inability to pro-duce mycotoxins, have obvious advantages,since greater quality control and product safetycan be achieved (Table 6–4).

Flavor of Fermented Meats

Like other fermented foods made with a lacticacid starter culture, the main flavor com-pounds are acids, principally lactic and acetic,derived via metabolism of sugars. In many ofthe U.S.-produced products that are curedwith nitrite, fermented at high temperaturesfor a short time, and cooked following fermen-tation, there is only a brief opportunity for de-velopment of other flavors. In contrast, a muchmore complex array of flavor compounds isproduced in sausages in which nitrate is usedas the curing agent, that are fermented slowly,and are uncooked.

Flavor development may be especially en-hanced if micrococci are included in thestarter culture.The ripening process for these



sausage products is not unlike that for agedcheese, in that microbial as well as endogenousenzymes act on proteins and fats in the rawmaterial, generating hydrolysis products thatcontribute to the flavor and texture of the fin-ished product. Importantly, the source of themicrobial enzymes may be the lactic or micro-coccal starter culture organisms or naturally-occurring bacteria, yeasts, and molds presentin the raw meat material.

Defects and Spoilage of Fermented Meats

Defects of fermented meats can occur before,during, or after manufacturing. Like all fer-mented foods, the production of high qualityproducts depends largely on the microbiologi-cal quality of the raw ingredients. Perhaps thisis even more so for fermented sausage, sincethe starting material,meat, is raw and cannot beheat-processed to inactivate spoilage or otherundesirable microorganisms. Thus, any organ-isms present in the raw meat will be present inthe sausage batter and may even survive fer-mentation.

Of course, fermentation acids kill or inhibitmany organisms,and, if the sausage is cooked af-ter fermentation, most of the remaining organ-isms will be killed. However, spoilage productsmay have already been produced. For example,psychrotrophic bacteria, such as Brochothrixthermosphacta and various Pseudomonasspp., may produce ammonia and other volatileoff-odors and flavors in the meat during storageeven before fermentation. Lipases and other en-zymes may also be produced by these organ-isms prior to fermentation, resulting in rancid,“cheesy,”or bitter end-products.

The other main group of spoilage organismsare lactic acid bacteria.These bacteria are partof the natural meat microflora, and are quitetolerant of the barriers that ordinarily controlmicrobial activity in fermented sausage (i.e.,low pH, low aw, low Eh,nitrite, salt,etc.).Underappropriate conditions, lactic acid bacteria cancause flavor, color, and tactile defects, and, asdiscussed in Box 6–3, some strains are also re-

sponsible for foodborne disease, due to theirproduction of biogenic amines. Of course, thecausative organisms are not necessarily en-dogenous to the product, because the starterculture itself is comprised of lactic acid bacte-ria, including L.sake,L.curvatus,L.plantarum,P. acidilactici, and other species capable, intheory,of producing these defects.Therefore, itis important that screening and selection ofstrains for starter cultures be based, in part, ontheir ability (or inability) to produce undesir-able end-products.

Among the spoilage products formed by lac-tic acid bacteria,hydrogen peroxide is probablythe most problematic. It is produced primarilyby lactobacilli, but only by specific strains, andonly under specific conditions. Oxygen is re-quired for production of hydrogen peroxide by lactic acid bacteria; it serves as a reactant inthe hydrogen peroxide-generating reactionsand also induces expression of the enzymes in-volved in these reactions.Thus, minimizing ex-posure to air during mixing and subsequentsteps is critical.

Once formed, hydrogen peroxide partici-pates in several undesirable reactions. First, itreacts with heme-containing pigments, espe-cially nitrosyl myochromogen, formed as a re-sult of curing.When the heme iron is oxidizedby peroxide, the desirable pink color is lostand an undesirable green pigment is formed(Box 6–1). Second, hydrogen peroxide canform hydroxyl radicals (e.g., O2

–) that serve asinitiators of lipid oxidation reactions.

One way to limit the formation of hydrogenperoxide in fermented meats is to include cata-lase-producing strains in the starter culture.Al-though some lactic acid bacteria produce asmall amount of catalase or a pseudo-catalase,micrococci (present in some cultures) pro-duce much greater levels of this enzyme. An-other less common defect caused by lacticacid bacteria is slime formation. In particular,strains of L. sakei have been shown to produceexo-polysaccharides (i.e., slime) in vacuum-packaged meat products, and have occasion-ally been implicated in this form of spoilage infermented sausages.



In general, microbial spoilage can best beprevented by keeping psychrotrophic bacteriaout of the raw meat and keeping other poten-tial spoilage organisms out of the finished fer-mented product via comprehensive sanitationprograms. Few organisms should be present incooked fermented sausages, provided post-processing contamination does not occur.Vac-uum or modified atmosphere packaging withcarbon dioxide inhibits aerobic psychrotrophs,but not facultative and anaerobic spoilage or-ganisms, such as B. thermosphacta and lacticacid bacteria. Antimicrobial agents, includingorganic acids and bacteriocins,may be effectiveagainst these bacteria, as described previously.

ReferencesBarbuti, S., and G. Parolari. 2002.Validation of manu-

facturing process to control pathogenic bacteriain typical dry fermented products. Meat Sci.62:323–329.

Cook, P.E. 1995. Fungal ripened meats and meatproducts, p. 110–129. In G. Campbell-Platt andP.E. Cook (ed.), Fermented Meats. Blackie Acade-mic and Professional (Chapman and Hall).

Fernández, M., J.A. Ordåñez, J.M. Bruna, B. Herranz,and L. de la Hoz. 2000. Accelerated ripening of

dry fermented sausages. Trends Food Sci. Tech-nol. 11:201–209.

Fisher, S., and M. Palmer. 1995. Fermented meat pro-duction and consumption in the EuropeanUnion, p. 217–233. In G. Campbell-Platt and P.E.Cook (ed.), Fermented Meats. Blackie Academicand Professional (Chapman and Hall).

Hammes, W.P., A. Bantleon, and S. Min. 1990. Lacticacid bacteria in meat fermentation. FEMS Micro-biol. Rev. 87:165–174.

Hammes, W.P., and C. Hertel. 1998. New develop-ments in meat starter cultures. Meat Sci. 49, No.Suppl. 1, S125–S138.

Hammes, W.P., and H.J. Knauff. 1994. Starters in theprocessing of meats. Meat Sci. 36:155–168.

Incze,K.1998.Dry fermented sausages.Meat Sci.49,No. Suppl. 1, S169–S177.

Jessen, B. 1995. Starter cultures for meat fermenta-tions, p. 130–159. In G. Campbell-Platt and P.E.Cook (ed.), Fermented Meats. Blackie Academicand Professional (Chapman and Hall).

Lücke, F.-K. 1994. Fermented meat products. FoodRes. Int. 27:299–307.

Lücke, F.-K. 1998. Fermented Sausages, p. 441–483.In B.J.B.Wood (ed.), Microbiology of FermentedFoods, Volume 2. Blackie Academic and Profes-sional (Chapman and Hall).

Suzzi,G., and F.Gardini.2003.Biogenic amines in dryfermented sausages: a review. Int. J. Food Micro-biol. 88:41–54.


7

Fermented Vegetables

“Oh, when I think of my LenaI think of a girl who could cookSome sweet sauerkraut that would swim in your mouthLike the fishes that swim in the brook”

From “Bring Back My Lena to Me,” by Irving Berlin, 1910

then the mixture was packed into suitable con-tainers and stored at an ambient temperature. Ifthe salt concentration was not too high, thispractice would have established ideal conditionsfor growth of naturally-occurring lactic acid bac-teria.The ensuing fermentation would have notonly enhanced preservation, but it would havealso created highly desirable flavor and aromacharacteristics.

It is believed that fermented vegetable tech-nology actually began more than 2,000 yearsago.In Asia and the Far East, fermented productswere made from cabbage, turnips, radishes, car-rots, and other vegetables endemic to the localareas.This technology was exported to Europesometime in the 1500s, with regional vegeta-bles, such as European round cabbages, servingas the starting materials. Eventually, Europeansettlers to the New World brought with themcabbages and procedures for the manufactureof sauerkraut. It is also likely that other fer-mented vegetables,pickles and olives in particu-lar,were produced and consumed in the MiddleEast, at least since biblical times.

Fermented vegetables have long been a stapleof Middle-East, Western and Far East diets, notonly because of their enhanced preservationand desirable flavor and texture properties, butalso because these products had important nu-tritional properties. For example, sauerkraut has

Introduction

Wherever vegetables are grown and con-sumed, it is almost certain that fermented ver-sions exist. Moreover, despite the wide diver-sity of vegetables produced around the world,the principles involved in the manufacture offermented vegetables are very near the same.Like other fermented foods, readily observedvariations certainly exist, and are based on aes-thetic preferences, as well as the types of rawmaterials available in particular regions.

For example,cabbage is widely used as a fer-mentation substrate, but the actual cultivarthat is used varies depending on culture andgeography. Thus, in Germany, mild-tasting,white European cabbage is transformed via fer-mentation into sauerkraut, whereas in SouthKorea, Chinese cabbage is mixed with pep-pers, radishes, other vegetables, garlic, andspices, and fermented to make a much spicierversion called kimchi.

The manufacture of fermented vegetablesmost likely evolved from simply dry-salting orbrining vegetables.Salting vegetables was a com-mon means of food preservation and was prac-ticed for thousands of years in Europe, the Mid-dle East,and Asia,and for several centuries in theAmericas. In general, salt or brine was added tothe fresh raw material as a preservation aid, and

233





long been known to have anti-scurvy properties,due to the high vitamin C content of cabbage.Sauerkraut was an essential food for navies andseafarers, who had little access to the fresh veg-etables and fruits that normally would haveserved as a source of vitamin C. Cabbage alsocontains high concentrations of thiocyanatesand other sulfur-containing compounds thatmay have antimicrobial activity. In the Far East,and Korea in particular, kimchi, has become themost popular of all fermented vegetables.Again,kimchi has unique and desirable flavor and sen-sory attributes,but its popularity is also due to itsperceived nutritional properties (see below).

Products and Consumption

In the United States, there are essentially onlythree fermented vegetable products that areproduced and consumed on a large scale basis.These include sauerkraut, pickles, and olives.The raw materials for these products—cabbage,cucumbers, and olives—are high moisturefoods, with little protein or fat (except forolives), and just enough fermentable carbohy-drate to support a fermentation (Table 7–1).

Other fermented vegetables, such as pep-pers, cauliflower, and green tomatoes, are alsoproduced, but these are not nearly as popular(at least in the West).There are also many acidi-fied or pickled vegetable products that aremade by adding mixtures of vinegar, salt, andflavoring materials to fresh vegetables. In fact,most of the pickle products consumed in theUnited States are not fermented, but rather aresimply “pickled” by packing fresh cucumbersin vinegar or salt brines (discussed later). Like-wise, most of the olives consumed in theUnited States are similarly produced.

Although more than 800 million Kg of fer-mented vegetables are produced annually inthe United States, on a per capita basis,Ameri-cans are not particularly heavy consumers offermented vegetables. In the United States, percapita consumption of sauerkraut for the pastten years has averaged about 0.6 Kg. Another0.5 Kg of olives are also consumed. Althoughper capita U.S.consumption of pickles is some-what higher, at about 4 Kg, less than half ofthose pickles are of the fermented variety. Incontrast, Germans eat about 1.8 Kg of sauer-kraut per person per year, and Syrians eatnearly 6 Kg of fermented olives. Remarkably,Koreans consume more than 43 Kg per year ofkimchi (or 120 g per day!).

Production Principles

In a general sense, fermented vegetable technol-ogy is based on the same principles as other lac-tic acid fermentations, in that sugars are con-verted to acids, and the finished product takeson new and different characteristics. In reality,however, the actual production of fermentedvegetables occurs quite differently.For example,whereas cheese, cultured dairy products, andfermented meats are usually produced usingstarter cultures, the fermented vegetable indus-try still relies on natural lactic microflora tocarry out the fermentation.

Compared to the relatively few strains usedfor dairy and meat fermentations, the lactic acidbacteria that are ultimately responsible for veg-etable fermentations are quite diverse. Severalgenera are usually involved, including both het-erofermentative and homofermentative species.In addition, although the plant-based substrates(i.e., cabbage,cucumbers,olives) ordinarily con-tain the relevant lactic acid bacteria necessary

Table 7.1. Composition of substrates used in vegetable fermentations.

Vegetable H20 Carbohydrate Protein Fat

Cabbage 92% – 94% 5% – 6% 1% —Cucumbers 95% 2% – 3 % 1% —Olives 78% – 80% 2% – 4% 1% – 2% 12% – 14%


Fermented Vegetables 235

to perform a lactic fermentation, they also har-bor a complex microflora consisting of otherless desirable organisms. In fact, the resident lac-tic acid bacteria population represents only asmall faction of the total microflora present inthe starting material.And unlike dairy fermenta-tions, where pasteurization can substantially re-duce the indigenous microflora present in rawmilk, no such heating step can be used to pro-duce fermented vegetables.

Although chemical pasteurization proce-dures have been developed for some productsand can effectively reduce the resident flora(see below), these applications, for the mostpart,are not widely employed.Therefore,the es-sential requirement for a successful fermenta-tion is to create environmental conditions thatare conducive for the lactic acid bacteria, butinhibit or otherwise restrict the non-lactic flora.

The microflora of fresh vegetables

Plant material, including edible vegetables,serves as the natural habitat for a wide varietyof microorganisms (Table 7–2). The endoge-nous or epiphytic flora consist of yeast, fungi,and both Gram positive and Gram negativebacteria.The plant environment is exposed tothe air and the surfaces of plant tissue have a high Eh. Thus, aerobic organisms, such asPseudomonas, Flavobacterium, Bacillus, andvarious mold species, would be expected todominate freshly harvested material, as is in-deed the case.

However, facultative anaerobes, including Enterobacter, Escherichia coli, Klebsiella, andother enteric bacteria, as well as sporeformingclostridia, are also part of the resident flora.Vari-ous yeasts, including Candida, Saccharomyces,Hansenula, Pichia, and Rhodotorula, may alsobe present. Lactic acid bacteria, mainly speciesbelonging to the genera Lactobacillus, Pedio-coccus, Streptococcus, and Enterococcus, are ordinarily present, but at surprisingly low numbers. In fact, whereas the total populationof Pseudomonas,Flavobacterium,Escherichia,and Bacillus may well reach levels as high as107 cells per gram, lactic acid bacteria are nor-

mally present at only about 103 cells per gram.Thus, the lactic acid bacteria are outnumberedby non-lactic competitors by a thousand timesor more,putting them at a serious disadvantage.

Given the diversity of microorganisms ini-tially present in the raw material and the nu-merical disparity between the lactic and non-lactic bacteria, it would seem that rather severemeasures must be adopted to establish the se-lective environment necessary for a successfullactic acid fermentation. Actually, selection isbased on only a few simple factors: salt, temper-ature, and anaerobiosis.Thus,under appropriateconditions, most non-lactic acid bacteria willgrow slowly, if at all. In contrast, lactic acid bac-teria will generally be unaffected (but not to-tally,), and will instead grow and produce acidicend products. The acids, along with CO2 thatmay also be produced, creates an even morestringent environment for would-be competi-tors.Within just a few hours, lactic acid bacteriawill begin to grow, a lactic acid fermentation

Table 7.2. Representative microflora of vegetables.

Organisms Log CFU/g

Aerobic bacteria 4 — 6PseudomonasFlavobacteriumMicrococcusStaphylococcusBacillusLactic acid bacteria 0.7 — 4LactobacillusPediococcusStreptococcusTetragenococcusLeuconostocEnteric bacteria 3 — 3.5EnterococcusEnterobacterKlebsiellaEscherichiaYeasts and Mold 0.3 — 4.6FusariumAscochytaAspergillusPenicilliumRhodotorula

Adapted from Nout and Rombouts, 1992



will commence, and the number of competingorganisms will decline.

The lactic acid fermentation that occursduring most vegetable fermentations dependsnot on any single organism, but rather on aconsortium of bacteria representing severaldifferent genera and species (Table 7–3).Thatis, a given organism (or group of organisms)initiates growth and becomes established for aparticular period of time.Then,due to accumu-lation of toxic end products or to other in-hibitory factors, growth of that organism willbegin to slow down or cease. Eventually, theinitial microbial population gives way to otherspecies that are less sensitive to those in-hibitory factors. Microbial ecologists refer tothese sorts of processes as a succession.This isone reason why vegetable fermentations areordinarily conducted without starter cultures,since duplicating a natural succession of or-ganisms likely would not be achieved on a con-sistent basis.

Manufacture of Sauerkraut

Few fermented foods are produced in such aseemingly simple process as is sauerkraut (Fig-ure 7–1). Only two ingredients, cabbage andsalt, are necessary, and once these ingredientsare properly mixed, there is little that the man-ufacturer needs to do until the fermentation iscompleted.The simplicity of the process is re-flected by the U.S. Standards, which states thatsauerkraut is the “product of characteristicacid flavor, obtained by the full fermentation,

chiefly lactic, of properly prepared and shred-ded cabbage in the presence of not less than 2percent nor more than 3 percent of salt.”Afterfermentation, sauerkraut should contain notless than 1.5% acid (expressed as lactic acid).

The manufacture of sauerkraut starts withthe selection of the raw substrate material.Al-though various cabbage cultivars exist, whitecabbage is typically used because it has a mild,slightly sweet flavor and contains 5% or morefermentable sugars (mostly equimolar amountsof glucose and fructose,with a small amount ofsucrose). Cabbage used to make sauerkrautshould be fully mature, and should contain fewouter leaves. Some manufacturers allow thecabbage heads to wilt for a day or two.

Shredding and salting

Once the outer leaves and any spoiled leavesare removed, the cabbage heads are washedand the core is drilled out.The cabbage (alongwith the core) is shredded (according to themanufacturers specifications) to make a slaw.The shredded leaves are then weighed andconveyed directly to tanks or are depositedfirst into tubs or carts and then transferred intotanks. Salt can be added as the slaw is con-veyed or it can be added to the slaw when it ar-rives in the tanks. In either case, both theamount of salt added and the means by whichmixing and distribution occur are critical.

Usually, between 2% and 2.5% salt is added(by weight), although 2.25% is generally con-sidered to be the optimum. Problems are al-

Table 7.3. Main lactic acid bacteria involved in vegetable fermentations.

Sauerkraut Kimchi Pickles Olives

Leuconostoc Leuconostoc Leuconostoc Leuconostoc mesenteroides mesenteroides mesenteroides mesenteroides

Leuconostoc fallax Leuconostoc kimchii Lactobacillus plantarum Lactobacillus plantarumLactobacillus plantarum Leuconostoc gelidum Lactobacillus brevis Lactobacillus brevisLactobacillus brevis Leuconostoc inhae Pediococcus pentosaceusPediococcus pentosaceus Leuconostoc citreum

Lactobacillus plantarumLactobacillus brevisLactococcus lactisWeissella kimchii



most certain to occur if too much or notenough salt is added or if the salt is not uni-formly distributed, because salt performs sev-eral essential functions during the sauerkrautfermentation.Very soon after the salt is mixedwith the shredded cabbage, water begins todiffuse out from the interior of the plant tissueto the exterior medium,due to simple osmosis.The brine that forms also contains sugars andother dissolved nutrients that diffuse out withthe water.Thus, it is this water phase that ulti-mately serves as the location for most of themicrobial activity.

Next, salt (dissolved in the brine) providesthe selective conditions that discourage growth

of most of the non-lactic microorganisms thatwould otherwise compete with the lactic mi-croflora.Although salt at a concentration of only2.25% is, by itself, not ordinarily sufficient to in-hibit all of the indigenous, non-lactic bacteria, itis enough to provide the lactic acid bacteriawith a substantial growth advantage. Further-more, combined with other environmental factors, the selective effects of this relativelymoderate salt concentration can be increasedappreciably (discussed below). Moreover, oncethe pH has been decreased by the productionof organic acids, the combination of salt plusacid contributes significantly to the long preser-vation properties of the finished product.Finally, salt imparts a desirable flavor to theproduct and helps to maintain a crisp texture bypreventing softening of the tissues.

Mixing

The shredded and salted cabbage is thenplaced into tanks and mixed well to distributethe salt.As noted above,mixing is an importantstep, because localized regions within therather heterogenous material may containmore or less than the 2.25% salt that was addedto the bulk mixture. Within those pockets,therefore, it is entirely possible that the saltconcentration may vary considerably, perhapsby as much as 0.1%. This may result in eithertoo little or too much inhibitory control overthe organisms that reside in that microenviron-ment. If spoilage organisms were able to grow,their products (e.g., slime, pigments, off-flavors) could accumulate and,when the sauer-kraut is mixed prior to packaging, contaminatethe entire batch of product. It is worth notingthat high salt levels can promote spoilage justas readily as low salt levels. For example, the“pink” defect (discussed below) is caused bygrowth of salt-tolerant yeasts that ordinarilywould be suppressed by lactic acid bacteriawhose growth is impaired at high salt levels.

The sauerkraut fermentation was tradition-ally performed in wooden barrels.Wood-stavetanks are still used; however, concrete vatsare now common. The latter are lined with

Pasteurize Refrigerate

Remove outer leaves and core

Cabbage

Shred and salt

Convey to tanks and mix

Fermentation

Package

Wash

Figure 7–1. Manufacture of sauerkraut. Adapted fromHarris, 1998.



fiberglass or plastic, and can hold as much as50,000 Kg.The cabbage is covered with a plas-tic, tarp-like material, large enough to drapeover the sides of the tank. Water (or brine) isthen placed on top to weigh down the cab-bage and to drive out and exclude air.This alsoreduces exposure to air-borne organisms, for-eign matter, and insects. The weight further enhances formation of a brine, which sooncompletely covers the shredded cabbage.

Fermentation

The sauerkraut fermentation has long been thesubject of interest among food microbiologistsas well as microbial ecologists. In fact, many ofthe biochemical and microbiological details ofthe sauerkraut fermentation were described aslong ago as the 1930s. This interest has un-doubtedly been due, in large part, to the verynature of the fermentation process, in that itinvolves several different naturally-occurringmicroorganisms acting as part of a complexecosystem. Recent reports suggest that bacte-riophages may also play an important role inthe microbial ecology of the sauerkraut fer-mentation (Box 7–1).

The manufacture of sauerkraut and manyother fermented vegetables depends on a suc-cession of organisms that are naturally pre-sent in the raw material. Some appear early onin the fermentation, perform a particularfunction, and then, for all practical purposes,disappear from the product. Other organisms,in contrast, emerge later in the fermentationand then remain at moderate to high levelsthroughout the duration of the fermentationand post-fermentation process. However,growth of those organisms that occur late inthe process depends on those organisms thathad grown earlier and that had establishedthe correct environmental conditions.

Microbial activity begins as soon as a brinehas formed. Initially, the atmosphere is aerobic,with redox potentials (or Eh values) of over200 mV. However, the combined effects ofphysical exclusion of air and residual respira-

tion and oxygen consumption by plant cellsquickly reduce the Eh and make the environ-ment anaerobic. Thus, pseudomonads, fungi,and other obligate aerobic microorganismsthat may initially be present at high levels,havelittle opportunity for growth. Some of these or-ganisms are also salt-sensitive, further reducingtheir ability to grow in this environment. Still,at the temperatures used during the sauerkrautfermentation (20°C to 25°C), many other in-digenous salt-tolerant, mesophilic, facultativeorganisms might be expected to grow, includ-ing Enterobacter, E. coli, Erwinia, and othercoliforms. Instead, these organisms persist foronly a short time, perhaps as little as a fewhours, due to competition by lactic acid bacte-ria and the inhibitory effects of the acids pro-duced by these bacteria.

The lactic fermentation in sauerkraut oc-curs in a series of overlapping stages or se-quences. These stages and the succession ofmicroorganisms associated with each stagehave been very well studied. Remarkably, thefermentation almost always follows the exactsame pattern (Figure 7–2).

The first stage, variously referred to as theinitiation or heterolactic or gaseous phase, ismarked by growth of Leuconostoc mesen-teroides.This organism is salt-tolerant and has arelatively short lag phase and high growth rateat low temperatures (15°C to 18°C). Impor-tantly, it metabolizes sugars via the heterofer-mentative pathway, yielding lactic and aceticacids, CO2, and ethanol. The acidic environ-ment (0.6% to 0.8%, as lactic acid) created bygrowth of L. mesenteroides not only inhibitsnon-lactic competitors, but it also favors otherlactic acid bacteria.The production of CO2 alsocontributes to making the environment evenmore anaerobic (as low as �200 mV), whichagain favors the more anaerobic lactic acidbacteria. Eventually, however, as the acid con-centration approaches 1.0%, L. mesenteroidesis, itself, inhibited, and within four to six days,this organism is barely detectable.

In the next stage or primary, homolactic,or non-gaseous phase, the decrease in the



Box 7–1. Bacteriophages Get Into the Mix

To microbial ecologists, there are few model systems better suited for the study of microbial in-teractions than that which occurs during the manufacture of sauerkraut and other fermentedvegetables. Consider what transpires in the course of a typical vegetable fermentation.Thereare, for example, major shifts in the environment, from aerobic to anaerobic, and from a neutralto acidic pH. The availability of nutrients, and fermentable carbohydrates, in particular, de-creases during the fermentation.At the same time, there is a succession of microorganisms,suchthat some species,present at the outset, are displaced by other microbial communities, and can-not even be detected at later stages.

In addition to these environmental and extrinsic factors, bacteriophages have recently beenrecognized as having a major influence on the microbial ecology and diversity of fermentedvegetables (Lu et al., 2003 and Yoon et al., 2002). Phages, it now appears from these studies, arewidespread in commercial sauerkraut fermentations and may directly influence microbiologicalsuccession during the fermentation. Over a two-year period, these investigators isolated morethan twenty-six distinct phage types that were capable of infecting lactic acid bacteria, includ-ing species of Leuconostoc, Lactobacillus, and Weissella sp.

Importantly, the appearance of some members of the natural microflora was correlated withthe appearance of their homologous phages.That is, phages capable of infecting Leuconostocwere only isolated during the first few days of the fermentation, when the flora are dominatedby Leuconostoc. Phages that infect Lactobacillus, in contrast, were not found until later in thefermentation,when Lactobacillus plantarum and other species had become established.Thus,these phage infection events coincided with the main population shift that occurs during thesauerkraut fermentation, from mainly heterofermentative species to mainly homofermentativespecies.

Further characterization of these phages revealed that some were stable in low pH environ-ments (pH 3.5) and were capable of persisting in the tanks for as long as sixty days. It appearsfrom this work, therefore, that the well-studied phenomenon of succession that occurs duringthe sauerkraut fermentation may be mediated not just by the changing environment,but also bythe emergence of bacteriophages.

There is another very practical reason why bacteriophages are important in vegetable fer-mentations.Although many manufacturers still rely on natural fermentations to produce theseproducts, starter cultures are now being used more frequently in pickle production, and appli-cations for their use in the manufacture of other fermented vegetables is expected to increase.However, like most industrial fermentation processes in which pure culture strains are used ona repeated basis, infective bacteriophages will invariably emerge. If the phage population in apickle production environment, for example,was able to reach some critical level, infection andlysis of starter culture strains may occur to the point at which the fermentation fails.

Because cucumbers (and other raw materials used in fermented vegetable manufacture) can-not be heated, nor are these fermentations conducted under aseptic conditions, it is not possi-ble to exclude phages from the production environment.Thus, the continuous use of startercultures in large-scale manufacture of pickles and other fermented vegetables may provide idealconditions for phage proliferation.Although there are no reports of phage problems in this in-dustry, it has been suggested that development of phage control systems may still be warranted.

References

Lu,Z.,F.Breidt,V.Plengvidhya, and H.P.Fleming.2003.Bacteriophage ecology in commercial sauerkraut fer-mentations.Appl. Environ. Microbiol. 69:3192–3202.

Yoon, S.S., R. Barrangou-Poueys, F. Breidt, Jr.,T.R. Klaenhammer, and H.P. Fleming. 2002. Isolation and char-acterization of bacteriophages from fermenting sauerkraut.Appl. Environ. Microbiol. 68:973–976.



Leuconostoc population coincides with thesuccession of several other lactic acid bacte-ria, most notably Lactobacillus plantarumand, to a lesser extent, Lactobacillus brevis.Although L. plantarum is a facultative hetero-fermentor (meaning it has the metabolic ca-pacity to ferment different sugars via homo-or heterofermentative pathways) and L.brevisis an obligate heterofermentor, both organ-isms are strong acid producers, nearly dou-bling the acid content to about 1.4% to 1.6%.They are also quite stable in this acidic envi-ronment and dominate the fermentation dur-ing this period (especially L. plantarum). It isnot unusual, however, for other lactic acidbacteria, including Pediococcus pentosaceus,Pediococcus acidilactici, Lactobacillus cuva-tus, and Enterococcus sp., to be present during the primary fermentation. Finally, asthe acidity approaches 1.6% and the pH de-

creases below 4.0, only the acid-tolerant L.plantarum is able to grow.The entire processcan take up to one to two months, and the fer-mentation is generally considered completewhen the acidity is at about 1.7%, with a pHof 3.4 to 3.6.

End products

Although lactic acid is the major compoundproduced during the fermentation, other meta-bolic end products are also formed. Impor-tantly, many of these products contribute tothe overall flavor of sauerkraut. In particular,end products produced by Leuconostoc sp.and other heterofermentative lactic acid bacte-ria are essential for good-tasting sauerkraut.Asmuch as 0.3% acetic acid and 0.5% ethanol canbe present in the finished sauerkraut. In addi-tion, these bacteria may also synthesize small

102

108

107

106

105

104

103

109

4

3

6

5

7

0 4 8 12 16 20 24 3228

pH

Days

marg /rebmun llec goL

1.0%

1.5%

1.8%

coliforms

Leuconostocmesenteroides

Lactobacillis brevis

Lactobacillis plantarum

Figure 7–2. Fermentation succession. Idealized model for successive growth of lactic acid bacteria duringthe sauerkraut fermentation. The approximate acidities (as lactic) at varying pHs are indicated. Adapted frommultiple sources.



amounts of diacetyl, acetaldehyde, and othervolatile flavor compounds. Finally, the CO2 thataccumulated during the initiation stage of thefermentation provides carbonation and en-hances mouth feel.

Mannitol is another end product that accu-mulates during the sauerkraut fermentation. Itis formed directly from fructose by heterofer-mentative leuconostocs and lactobacilli via theNADH-dependent enzyme, mannitol dehydro-genase.As shown in Figure 7–3, regeneration ofNAD in the heterofermentative pathway ordi-narily occurs twice, first when acetyl CoA is reduced to acetaldehyde and then when ac-etaldehyde is reduced to ethanol. However,

fructose, when available in excess, can serve asan alternative electron acceptor.This allows thecell to use acetyl phosphate as a phosphorylgroup donor in the acetate-generating reactioncatalyzed by acetate kinase.As a consequence,one additional molecule of ATP is synthesizedby substrate level phosphorylation.

More than 100 mM mannitol can be formedin sauerkraut by this pathway (with a com-mensurate amount of ATP made for the cell).Eventually, some of this mannitol may be con-sumed by lactobacilli that emerge later in thefermentation.Although most of the mannitol isproduced directly from fructose, the appear-ance of mannitol even when fructose has been

ATP

ADP

CO2

NAD

NADH + H

glucose

glucose-6-P

6-phosphogluconate

ribulose-5-P

xyulose-5-P

acetyl P

lactic acid

ATP

ADP

fructose

fructose-6-P

glyceraldehyde-3-P

acetaldehyde

NAD

NADH + H

acetate

ATPADP

mannitol

NADNADH + H

ethanol

acetyl CoA NADH + H

NAD

NADH + H

NAD

acetate kinase

mannitol dehydrogenase

Figure 7–3. Mannitol formation in Leuconostoc. Phosphorylated glucose and fructose are metabolized bythe heterofermentative (i.e., phosphoketolase) pathway. The reactions leading to lactic acid (from glyceralde-hyde-3-P) are not shown. Only the relevant enzymes are indicated; enzymes for all other reactions are given inChapter 2. P=Phosphate. Adapted from Salou et al., 1994 Appl. Environ. Microbiol. 60:1459-1466.



depleted suggests that some might be formedindirectly from glucose (following its conver-sion to fructose via glucose isomerase).

Packaging and processing

In the United States, commercial products areusually thermally processed, much like otherhigh-acid foods at about 75°C, prior to packag-ing in cans or jars. Such products are essen-tially commercially sterile and are stable atroom temperature. There is also a market fornon-pasteurized, refrigerated sauerkraut that ispackaged in glass jars or sealed plastic bags(polybags). These products also have a longshelf-life, provided antimycotic agents, such asbenzoate and sulfite salts, are added and theproduct is kept cold.

Spoilage and defects

Although chemical or physical reactions are oc-casionally responsible for sauerkraut spoilage,most defects are caused by microorganisms(Table 7–4). Defects are more likely to occur ifthe production conditions are not properlycontrolled, leading to deviations in the fermen-tation pattern.

The two most common factors that influencethe fermentation and that may lead to quality de-fects are temperature and salt.If,for example,thetemperature is too high (�30°C) or if too muchsalt is added (�3%), L. mesenteroides will be in-

hibited,heterofermentative end-products will beabsent, and the flavor will be harsh.Worse yet,when prompt acid formation is delayed by highsalt conditions (or pockets of high salt, due touneven mixing), growth of salt-tolerant yeastsmay occur. Growth of Rhodotorula sp. is a par-ticular problem,due to the undesirable pink pig-ment this yeast produces.

If, in contrast, the temperature is too low(�10°C) or too little salt is added (�2%),variousGram negative bacteria, including Enterobacter,Flavobacterium, and Pseudomonas, may grow.Some of these bacteria are capable of producingpectinolytic enzymes that cause a “soft kraut”de-fect, resulting from pectin hydrolysis.

Finally, some strains of L.mesenteroides pro-duce dextrans and other polysaccharides thatgive a slimy or ropy texture to the product.Other lactic acid bacteria, including L. plan-tarum, may produce capsular materials thatcause similar texture defects.

Kimchi

As noted previously, the Korean version ofsauerkraut is called kimchi.Although the man-ufacture of kimchi and sauerkraut are very sim-ilar, the characteristics of the finished productsare quite different. The main difference be-tween sauerkraut and kimchi is that the latterproduct contains ingredients other than sim-ply cabbage and salt. Kimchi, for example, isusually made from cabbage, but other vegeta-

Table 7.4. Microbial defects in fermented vegetables.

Product Defect Causative organisms

Sauerkraut Pinking RhodotorulaRopy/slimy Leuconostoc, LactobacillusSoftening Enterobacter, Flavobacterium, Pseudomonas

Pickles Bloating/floating Bacillus, Aeromonas, Achromobacter, Aerobacter, Fusarium, PenicilliumSoftening Fusarium, Penicillium, Ascochyta

Olives Gassy/floater/fish eyes Enterobacter, Citrobacter, Klebsiella,Escherichia, Aeromonas, Hansenula,Saccharomyces

Softening Bacillus, Aeromonas, Achromobacter,Aerobacter, Fusarium, Penicillium

Zapatera/malodorous Clostridium, Propionibacteria



bles, including radishes and cucumbers, canalso be used, alone or mixed with cabbage.Other vegetables, spices, and flavoring agentsare also commonly added to kimchi, depend-ing on the particular type of kimchi being pro-duced. Garlic, green onion, ginger, and red pep-pers are among the typical ingredients, but

fish, shrimp, fruits, and nuts, can also be added.Kimchi, therefore, has a much more complexflavor and texture profile. However, kimchi hasalso been suggested to have unique nutritionalproperties, conferred in part by the raw mate-rials, but also by the fermentative microorgan-isms and their end products (Box 7–2).

Box 7–2. Health Properties of Kimchi

In Korea, kimchi is arguably the most popular of all fermented foods.As many as 100 differenttypes are produced using various raw materials and processes. On an annual basis, about 360million kg are produced commercially, with a large amount also made directly in homes. Percapita consumption is more than 100 grams per day, accounting for about 12% of the total foodintake. In Korea, it is common to eat kimchi at every meal, all year round.

The manufacturing steps involved in kimchi production are nearly the same as those used forits Western counterpart, sauerkraut (Figure 1).The main difference is that kimchi contains otheringredients that impart considerably more flavor and texture properties.While the popularity of

Chinese cabbage

Trim and cut

Wash

Brine

Rinse and drain

Blend

Ferment

saltseafood

seasoningsWash Cut

red peppergarlic

gingergreen onions

Cut

(Continued)

Figure 1. Kimchi flow chart. Adapted from Cheigh, 1999.



kimchi can certainly be ascribed to the combination of these desirable sensory characteristics,there are also nutritional reasons that contribute to the its widespread consumption.For example,kimchi contains appreciable amounts of ascorbic acid (vitamin C),B vitamins,calcium,iron,potas-sium,dietary fiber, and naturally occurring antioxidants. It is also important to note that kimchi ismost often eaten in its uncooked or raw state, and that a live microflora is ordinarily present.

Kimchi has been recognized as having other important nutritional and health-promoting prop-erties and has even been regarded as a functional food (i.e., those with health benefits beyondtheir ordinary nutritional constituents). It is thought to reduce the risks of stomach, colon, andliver cancers,due to the putative antimutagenic and anticarcinogenic constituents present in kim-chi.Although convincing evidence is still needed to support most of the claims, epidemiologicalstudies indicate that consumption of kimchi may be associated with reduced incidence of variousdisease states among the Korean population. For example, a significant decrease in the risk of de-veloping gastric cancer was correlated with increased intake of kimchi made from low-nitratevegetables, such as Chinese cabbage (Kim et al.,2002).The same study,however,also showed thatconsumption of kimchi made from high-nitrate vegetables actually increased the risk.

If, indeed, kimchi has health-promoting properties beyond those present in the raw material,then those effects must be due either to the kimchi microflora directly, to products produced bythe microflora during the fermentation,or to transformation of kimchi constituents.Various lacticacid bacteria isolated from kimchi have been reported to have antimutagenic activity, based onAmes tests. For example, Lactobacillus plantarum KLAB21 was found to produce extracellularglycoproteins that had antimutagenic activity against aflatoxin B1 and several chemical mutagens(Rhee and Park, 2001).A methanol extract from kimchi similarly reduced induced cytotoxicity inC3H/10T1/2 cells (Cheigh, 1999). In another recent study, four Leuconostoc mesenteroidesstrains isolated from kimchi were able to deplete nitrite, a potential carcinogen, from broth cul-tures, suggesting that this property would contribute to the safety of kimchi (Oh et. al., 2004).Anactive chemopreventative component was also recently identified in kimchi extracts (Park et. al.,2003).This plant-derived compound,�-sitosterol, inhibited proliferation of human cancer cells bydecreasing DNA synthesis and cell signaling pathways.What role the fermentation plays in thisprocess is not clear,however.Finally, it was recently suggested by researchers in Korea that kimchisomehow provided protection against the viral agent that causes severe acute respiratory syn-drome (SARS).This claim was based largely on the observation that while SARS was widespread inChina,Hong King, and Taiwan in 2003, there were few suspected cases in South Korea.

Kimchi is made from a wide variety of raw materials, including several that may have biologi-cal activity (e.g., garlic, radishes, peppers, ginger, and onion).The lactic acid microflora is alsocomplex and variable.Thus, identifying the actual components that may be responsible for thesuggested health benefits of kimchi, and then justifying the nutritional claims, remains a signifi-cant challenge.

References

Cheigh, H. 1999. Production, characteristics and health functions of kimchi. In Proceedings of the Interna-tional Symposium on the Quality of Fresh and Fermented Vegetables, p. 405—419, J.M. Lee, K.C. Gross,A.E.Watada, and S.K. Lee (ed).Acta Hort. 483. ISHS.

Kim,H.J.,W.K.Chang,M.K.Kim,S.S.Lee,B.Y.Choi.2002.Dietary factors and gastric cancer in Korea: a case-control study. Int. J. Cancer 97:531–535.

Oh, C.K., M.C. Oh, and S.H. Kim. 2004.The depletion of sodium nitrite by lactic acid bacteria isolated fromkimchi. J. Med. Food 7:38–44.

Park, K.-Y., E.-J. Cho, S.-H. Rhee, K.-O. Jung, S.-Y.Yi, and B.Y. Jhun. 2003. Kimchi and an active component,�-sitosterol, reduce oncogenic H-Rasv12-induced DNA synthesis. J. Med. Food. 6:151–156.

Rhee, C.-H., and H.-D. Park. 2001.Three glycoproteins with antimutagenic activity identified in Lactobacil-lus plantarum KLAB21.Appl. Environ. Microbiol. 67:3445–3449.

Box 7–2. Health Properties of Kimchi (Continued)



Principles of Pickle Production

In a very general sense,pickles refer to any veg-etable (or fruit) that is preserved by salt oracid. Certainly, the vegetable most often associ-ated with pickles is the cucumber. Currently,about half of the total cucumber crop (1 bil-lion Kg) in the United States is used for pickles.The acid found most often in pickled productsis lactic acid, derived from a lactic fermenta-tion.As previously noted, however, not all pick-les are fermented. In fact, less than half of thepickles consumed in the United States undergoa lactic acid fermentation. Rather, acetic acidcan be added directly as the pickling acid,omitting the fermentation step.The pickle sliceon the top of a fast-food hamburger is probablynot the fermented type.

Pickles and pickling technology have a longand rich history. The cucumber was broughtfrom India to the Middle East about 4,000years ago, and cured versions were eaten atleast 3,000 years ago. Cleopatra endorsedpickle consumption, claiming they were re-sponsible (in part) for her beauty.The ancientRoman historian and natural scientist Pliny theElder and his contemporaries, Roman emper-ors Julius Caesar and Tiberius, were all fond ofpickles. Columbus introduced pickles to theAmericas, eventually leading to the origins of apickle industry on the Lower East Side of NewYork City. American founding fathers GeorgeWashington, John Adams,and Thomas Jeffersonreportedly derived inspiration from the pickle.

In the United States, pickles are generally di-vided into three different groups, based on theirmeans of manufacture. Fresh-packed pickles aresimply cucumbers that are packed in jars, cov-ered with vinegar and other flavorings, then pas-teurized by heat.They have a long shelf-life,evenat room temperature. Fresh packed pickles arecrisp, mildly acidic, and are the most popular.Refrigerated pickles are also made by packingcucumbers jars with vinegar and various flavor-ings,but they are not heated. Instead these pick-les are refrigerated, giving them a crisp, crunchytexture and bright green color.Although a slightfermentation may occur, refrigerated pickles

have a shorter shelf-life than fresh-packed pick-les.Sodium benzoate is usually added as a preser-vative. The manufacture of both fresh-packedand refrigerated-style pickles is fast and easy andrequire few steps (Figure 7–4).

The only pickles that are fully fermented arethose referred to as salt-stock or genuine pick-les.They may also be referred to as processed,although this may be somewhat confusingsince non-fermented pickles can be made intorelishes and other processed pickle products.Fermented pickles are nearly as popular asfresh-packed pickles.They have a distinctly dif-ferent flavor and texture compared to fresh-packed or refrigerated pickles, and take muchlonger to make. Fermented or processed pick-les also have a very long shelf-life, about twoyears. Details regarding their manufacture andthe fermentation process are described below.

It should be noted that within these threegroups, pickles can be further distinguishedbased on the types of spices, herbs and flavor-ing agents used, the size or type of cucumber(gherkins and midgets), and the form or shapeof the pickle (i.e., whole, spear, slice, etc.). Forexample, dill pickles, the most popular, refer toany type of pickle to which dill weed (eitherthe seed or oil) is added. If the dill pickles arelabeled as genuine dill, it means the pickles areof the processed type and are dill-flavored.Oth-erwise,dill pickles are usually made from fresh-packed or refrigerated pickles. Other commonflavored pickle types include sweet, bread-and-butter, kosher (style), and garlic.

Manufacture of fermented pickles

The actual process steps used for the manufac-ture of fermented pickles are similar to thoseused for making sauerkraut. Both rely on salt,oxygen exclusion, and anaerobiosis to providethe appropriate environmental conditions nec-essary to select for growth of naturally-occurringlactic acid bacteria. There are, however, severaldifferences between pickle and sauerkraut fer-mentations. First, salt concentrations are higherthan those used for sauerkraut, resulting in thedevelopment of a less diverse microflora. In



addition, a brine, rather than dry salt, is used forpickle fermentations.Finally,the pickle fermenta-tion process, unlike sauerkraut, is amenable tothe use of pure starter cultures and a more con-trolled fermentation. Indeed, such cultures arenow available and some (but not many) picklemanufacturers have adopted controlled fermen-tation processes (Box 7–3).

The manufacture of fermented pickles startswith selection and sorting of cucumbers (Fig-ure 7–5). Only small or immature cucumbers,harvested when they are green and firm, areused for pickles.They are then washed, sorted,and transferred to tanks, and a brine solution isadded.

The brine typically contains at least 5% salt(or about 20° salometer, where 100° salometer� 26% salt). Because the cucumbers-to-brine ratio is nearly 1:1,the actual salt concentration isactually less. For so-called salt stock pickles,which may be held in bulk for long periods, theinitial brine may contain 7% to 8% salt, which isfollowed by the addition of more salt to raise thetotal salt concentration to above 12%. For gen-

uine dill-type pickles, the brine concentration isusually between 7.5% and 8.5%.Dill weed is alsoadded,usually in the seed or oil form.

Care must be taken when weighing downthe pickles, because the buoyancy of the cu-cumbers may cause those at the top to becomedamaged.The large tanks used by large picklemanufacturers are usually located outdoors,and temperatures may, therefore, vary between15°C to 30°C.The lower the temperature, thelonger it takes to complete the fermentation.Thus, in Michigan (the largest northern pro-ducer of pickles), fermentation may require upto two months,whereas in North Carolina (themain southern producer), fermentations maybe complete in three weeks.At the end of thefermentation, the pH will be about 3.5, withacidities between 0.6% and 1.2% (as lactic).

Pickle fermentation

As noted above, the high salt concentrationsused in pickle manufacturing cause the fer-mentation to proceed quite differently from

Fresh-packed

Cucumbers sized, cleaned,flowers removed

Blanch

Wash

Fill and package

Pasteurize

Refrigerated

Cucumbers sized, cleaned,flowers removed

Wash

Fill and package

Refrigerate

Figure 7–4. Non-fermented pickles and their manufacture. Adapted from Harris, 1998.



Box 7–3. Starter Cultures and Fermented Vegetables

The modern manufacture of most fermented vegetables, in contrast to cheese, sausage, andother fermented food products, still relies on a natural fermentation. In large part, this is be-cause vegetable fermentations occur as a succession, and duplicating this process with “con-trolled fermentations” using starter cultures has not been a viable option. Also, vegetable fer-mentations are often conducted in less than aseptic conditions, so adding a culture to a rawmaterial comprised of a complex, well-populated background flora is unlikely to be very effec-tive.Finally,paying for cultures to perform a step that ordinarily costs the manufacturer nothingmakes little economic sense.

Yet, for all of the same reasons that eventually drove other industries to adopt pure starterculture technology (i.e., consistency, control, safety, and convenience), the fermented vegetableindustries have indeed developed manufacturing procedures that depend on starter cultures,rather than the natural flora, to perform the fermentation. Research on pure culture technologyfor fermented vegetables actually began nearly fifty years ago, and cultures were developed inthe 1960s (Daeschel and Fleming,1987).Despite the availability of these cultures,however, theyhave not been widely used.

More recently, other factors have provided additional motivation for the use of starter cul-tures. One particularly relevant issue relates to the large volumes of high salt brines that aregenerated by pickle and olive fermentations.These salt solutions create significant environmen-tal problems.The most obvious way to reduce the discharge of this material into the environ-ment is to use less salt. However, since salt provides the major means for controlling the microflora, conducting a natural fermentation at low-salt concentrations is unlikely to be satis-factory. In contrast, less salt could be used if the background flora was controlled by othermeans (see below), and a starter culture was used instead to dominate the environment and tocarry out the fermentation.

The first requirement for performing a controlled fermentation is to remove and/or inacti-vate the endogenous microflora.This can be done with chemical agents that kill organisms atthe surface. Specifically, the raw product is washed first with dilute chlorine solutions, thenwith acetic acid. In reality, chemical pasteurization is possible only for cucumbers and olivesand not for shredded cabbage due to surface area considerations.Although olives can tolerate amodest heat treatment, cabbage and cucumbers suffer severe texture defects if heated.Finally, anitrogen purge drives out air and creates anaerobic conditions, and an acetate buffered brine isthen added, followed by the starter culture.

The organisms that are currently available or are being considered for use as starter culturesinclude many of the same species ordinarily isolated from vegetable fermentations, e.g., Lacto-bacillus plantarum, Leuconostoc mesenteroides, Pediococcus acidilactici, and Lactobacillusbrevis. Strain selection,however, is necessary and must be based on the specific application anddesired characteristics of the particular fermented food (Table 1). For example, strains to beused as starter cultures for fermented olives must resist the antimicrobial phenolic compoundsordinarily present in olives.

As noted above, another inherent problem in controlled fermentation technology is the diffi-culty in establishing a microbial succession, such that a heterofermentative phase always pre-cedes the homofermentative phase.That is, how can L. mesenteroides and L. plantarum bothbe added at the outset of a sauerkraut fermentation, yet have conditions controlled such thatgrowth of L. plantarum is delayed until the later stages of the fermentation? Researchers atUSDA and North Carolina State University in Raleigh,North Carolina,addressed this problem bydeveloping a clever bio-controlled fermentation process (Breidt et al., 1995; Harris et al., 1992).In this model system, the starter culture consisted of an L.mesenteroides strain that resisted thebacteriocin nisin.As hypothesized, in the presence of nisin (either added directly or producedby a companion nisin-producing strain of Lactococcus lactis), the L. mesenteroides strain grew

(Continued)



that in sauerkraut. Only those pickles made us-ing brines at less than 5% salt will allow forgrowth of L. mesenteroides. Although hetero-fermentative fermentations may promote morediverse flavor development, the formation ofCO2 is undesirable, because it may lead tobloater or floater defects (see below). More-over, low salt brines may also permit growth ofunwanted members of the natural flora, includ-ing coliforms, Bacillus, Pseudomonas, andFlavobacterium. At salt concentrations be-tween 5% and 8%,growth of Leuconostoc is in-hibited and instead the fermentation is initi-ated by Pediococcus sp. and L. plantarum.

Pickle fermentation brines typically containhigh concentrations of salt and organic acidsand have a pH less than 4.5.These conditions areespecially inhibitory to coliforms, pseudomon-ads, bacilli, clostridia, and other non-lactic acidbacteria that would otherwise cause flavor andtexture problems. This environment, in fact, ishard even on lactic acid bacteria. However, thelatter have evolved sophisticated physiologicalsystems that enable them to survive under veryuncomfortable circumstances (Box 7–4).

After fermentation, salt stock pickles can beheld indefinitely in the brine. However, thesepickles cannot be eaten directly, but rather

fine, but growth of nisin-sensitive L. plantarum was inhibited.Thus, it was possible to prolongthe heterofermentative (and flavor-generating) phase of the fermentation, while delaying thehomofermentative phase.

Ultimately, the use of starter cultures for the manufacture of fermented vegetables is likely toincrease as the size of the production facilities and the demand for speed, efficiency, andthroughput both increase. In addition, the starter culture industry is now able to develop strainsthat have specific physiological properties, satisfy specific performance characteristics, are sta-ble during storage, and are easy to use.

Table 1. Properties of starter cultures for fermented vegetables.

Minimum nutritional requirementsAble to grow at low temperaturesAble to ferment diverse carbohydrate substratesAble to compete against wide array of organismsAble to produce desirable flavorRapid growth and acid productionTolerant to acids and low pHTolerant to saltTolerant to antimicrobial phenolicsResistant to bacteriophageNon-pectinolyticUnable to produce dextrans or other polysaccharidesUnable to produce biogenic aminesMinimum loss of viability during storage

References

Breidt, F., K.A. Crowley, and H.P. Fleming. 1995. Controlling cabbage fermentations with nisin and nisin-resistant Leuconostoc mesenteroides. Food Microbiol. 12:109–116.

Daeschel, M.A., and H.P. Fleming. 1987.Achieving pure culture cucumber fermentations: a review, p. 141—148. In G. Pierce (ed.), Developments in industrial microbiology. Society for Industrial Microbiology,Ar-lington,Va.

Harris, L.J., H.P. Fleming, and T.R. Klaenhammer. 1992. Novel paired starter culture system for sauerkraut,consisting of a nisin-resistant Leuconostoc mesenteroides strain and a nisin-producing Lactococcus lac-tis strain.Appl. Environ. Microbiol. 58:1484–1489.

Box 7–3. Starter Cultures and Fermented Vegetables (Continued)



must be de-salted by transfer to water. Afterseveral changes (a process called refreshing),the salt concentration is reduced to about 4%.They are then used primarily for relishes andother processed pickle products.

Defects

Among the microbial defects that occur inpickles, the most common are bloaters andfloaters (Table 7–4).The defect is caused by ex-cessive gas pressure that subsequently results

in internal cavity formation within the pickles.The CO2 gas is mainly produced by heterofer-mentative lactic acid bacteria (some of whichmay produce CO2 via the malolactic fermenta-tion), although coliforms and yeasts may alsobe responsible. Floaters and bloaters can stillbe used for some processed products (i.e., rel-ish), however, they cannot be used for thewhole or sliced pickle market.The most com-mon way to control or minimize this defect isto remove dissolved CO2 by flushing or purg-ing the brine with nitrogen gas.

Cucumbers sized and cleaned,and flowers removed

Brine added

Wash

Fermentation

PackageDesalt

Refrigerate

Pasteurize

Process

Add acetic acid

Salt stock Genuine

Acidify with acetic acidPurge with N2

Add dill spice

Package

Figure 7–5. Manufacture of fermented pickles. Adapted from Harris, 1998.



Box 7–4. In a Pickle: How Lactic Acid Bacteria Deal with Acids and Salts

The key requirement to ensure the success of vegetable fermentations is to create a restrictive,if not inhospitable, environment, such that most indigenous microorganisms are inhibited orotherwise unable to grow.Typically, this is initially accomplished by adding salt, excluding oxy-gen, and maintaining a somewhat cool temperature.As lactic acid bacteria grow and produceorganic acids and CO2, the ensuing decreases in pH and Eh provide additional hurdles, espe-cially for salt-sensitive,neutrophilic,aerobic organisms.These conditions,however,may not onlyaffect resident enteric bacteria, pseudomonads, clostridia, fungi, and other undesirable microor-ganisms, but they can also impose significant problems for the lactic acid bacteria whosegrowth is to be encouraged.

In addition, there may be other ionic compounds present in the vegetable juice brine that in-terfere with growth of lactic acid bacteria. For example, acetate, lactate, and other buffer saltsare often added to brines, especially when pure cultures are used to initiate the fermentation.These acids and salts may have significant effects on cell metabolism and growth (Lu et al.,2002). Thus, how plant-associated lactic acid bacteria cope with these challenges may be ofpractical importance.

Lactic acid bacteria are prolific producers of lactic acid (no surprise there), and can toleratehigh lactic acid concentrations (�0.1 M) and low pH (�3.5), much more so than most of theircompetitors.At least several physiological strategies have been identified that enable these bac-teria to tolerate high acid, low pH conditions.

First, lactobacilli and other lactic acid bacteria can generate large pH gradients across the cellmembrane, such that even when the medium pH is low (e.g., 4.0), the cytoplasmic pH (the rel-evant pH for the cell’s metabolic machinery) is always higher (e.g, 5.0). For example, in onestudy (McDonald et al., 1990),Lactobacillus plantarum and Leuconostoc mesenteroides main-tained pH gradients of nearly 1.0 or higher,over a medium pH range of 3.0 to 6.0 (acidified withHCl). In the presence of lactate or acetate, however, somewhat lower pH gradients were main-tained.This is because organic acids diffuse across the cytoplasmic membrane at low pH (orwhen their pKa nears the pH), resulting in acidification of the intracellular medium. For somebacteria, e.g., L. mesenteroides, the pH gradient collapses at low pH (the so-called critical pH).At this point, the cell is in real trouble, as enzymes, nucleic acid replication,ATP generation, andother essential functions are inhibited. In general, these results reflect, and are consistent withthe observed lower acid tolerance of L.mesenteroides, as compared to L. plantarum.

If maintenance of a pH gradient is important for acid tolerance, then the next question toask is how such a gradient can be made.That is, how can the protons that accumulate insidethe cell and cause a decrease in intracellular pH be extruded from the cytoplasm? Althoughthere are actually several mechanisms for the cell to maintain pH homeostasis,one specific sys-tem, the proton-translocating F0F1-ATPase (H�-ATPase) is most important.This multi-subunit,integral membrane-associated enzyme pumps protons from the inside to the outside using ATPhydrolysis as the energy source (Figure 1A).This enzyme is widely conserved in bacteria (infact, throughout nature), but the specific properties of enzymes from different species showconsiderable variation.Thus, the H�-ATPases from lactobacilli have a low pH optima, account-ing, in large part, for the ability of these bacteria to tolerate low pH relative to less tolerant lac-tic acid bacteria.

Although the H�-ATPase system is the primary means by which lactic acid bacteria maintainpH homeostasis, other systems also exist (Figure 1A). For example, deamination of the aminoacid arginine releases ammonia, which raises the pH. Decarboxylation of malic acid, which iscommonly present in fermented vegetables, also increases the pH by conversion of a dicar-boxylic acid to a monocarboxylic acid. In fact, when lactobacilli and other lactic acid bacteriaare exposed to low pH, a wide array of genes are induced (Van de Guchte et al., 2002). Collec-tively, this adaptation to low pH is referred to as the acid tolerance response. Some of the in-duced genes code for proteins involved in the machinery used by the cell to deal with other



physical or chemical stresses.Thus, the acid tolerance response may not only protect the cellagainst low pH, but also heat and oxidative stress.

In vegetable fermentations, the other important stresses encountered by lactic acid bacteriaare high salt concentrations and high osmotic pressures. Salt concentrations in sauerkrautbrines are around 0.4 M,giving an osmolality of about 0.8 Osm.Pickle and olive brines may con-tain more than 1.0 M salt, or osmolalities above 2 Osm. Salt is an extremely effective antimicro-bial agent due to its ability to draw water from the cytoplasm, thereby causing the cell to be-come dehydrated, lose turgor pressure,and eventually plasmolyze (and die!).The ability of lacticacid bacteria to tolerate high salt conditions varies (just as it did for acid tolerance), dependingon the organism. Given that L. plantarum usually dominates high salt fermentations, it should

Box 7–4. In a Pickle: How Lactic Acid Bacteria Deal with Acids and Salts (Continued)

out

in

membrane

malate2-

malate2-

lactate-

lactate-

H+ CO2

arginine

arginine

ornithine

ornithine

NH3

citrullinecarbamoyl-P

NH3 + CO2

ADP

ATP

ADP + PiATP

H+

H+

A

out

in

membrane

betaine

S-BP

ATP-BP

ADP + PiATP

betaine

S-BP

ATP-BP

B

permease

Figure 1. pH and osmotic homeostastis in Lactobacillus plantarum. Panel A shows three main systemswhose function is to maintain pH homeostasis in L. plantarum. The F0F1-ATPase (left) is a primary protonpump that extrudes protons from the inside to the outside, using ATP as the energy source. In contrast, themalate and arginine systems rely on product efflux to drive uptake (no energy is required). In the malolactatesystem (center), proton consumption de-acidifies the medium and raises the pH. In the arginine diiminase sys-tem (right), medium pH is raised by virtue of the two molecules of NH3 that are released per mole of arginine.

Panel B shows the QacT system responsible for osmotic homeostasis in L. plantarum. The components ofthis putative opuABCD-encoded system (left) include membrane-associated substrate-binding proteins (S-BP),a betaine permease, and ATP-binding proteins (ATP-BP). When the osmotic pressure is high, betaine is boundby the S-BP and taken up by the permease (center). Transport is driven by an ATPase following ATP-binding.If the osmotic pressure is reduced, accumulated betaine is effluxed via membrane channels (right).

(Continued)



Destruction and softening of the pickle sur-face tissue is another serious defect.When thishappens, the pickle loses its crispness andcrunch and becomes slippery and soft. Thesepickles are neither edible nor can anything bedone to salvage them.The defect is caused bypectinolytic enzymes produced by microorgan-isms that are part of the natural cucumber mi-croflora. The organisms responsible include

mostly filamentous fungi, especially species ofPenicillium, Fusarium, Alternaria, Aschyta,and Cladosporium. Pectins are complex het-eropolysaccharides that serve as the main struc-tural component of plant cell walls. Their hy-drolysis requires the concerted action of threedifferent enzymes—pectin methylesterase,poly-galacturonase, and polygalacturonate lyase. Al-though Penicillium and other fungi are capable

be no surprise that this organism has also evolved physiological and genetic mechanisms thatmake it salt- and osmotolerant.

Like acid tolerance, the salt tolerance system depends on the activity of membrane pumps.However, in the latter case, the pumps are actually transport systems that take up a special classof molecules called compatible solutes. By accumulating these non-toxic solutes inside the cy-toplasm to high concentrations, cell water is retained and osmotic homeostasis is maintained.

Compatible solutes are also referred to as osmoprotectants because they not only maintainosmotic balance, they also protect enzymes, proteins, and other macromolecules from dehydra-tion and misfolding.Among the osmoprotectants accumulated by L. plantarum, the quaternaryamine, glycine betaine (or simply betaine) is the most effective. Potassium ion, glutamate, andproline are also accumulated, but to lower concentrations, at least in lactobacilli. In L. plan-tarum, betaine is preferentially transported by the quaternary ammonium compound or QacTtransport system (Glaasker et al.,1998).This transporter is a high affinity,ATP-dependent systemwhose activity is stimulated by high osmotic pressure (Figure 1B). In contrast, at low osmoticpressure, the efflux reaction is activated, and pre-accumulated betaine is released back into themedium.Analysis of the L.plantarum genome sequence indicates QacT may be encoded by theopuABCD operon,which is widely distributed in other Gram positive bacteria (Kleerebezem etal., 2003).The natural source of betaine, it is worth noting, is plant material, so perhaps it is nocoincidence that L. plantarum is unable to synthesize betaine and instead relies on a transportsystem to acquire it from the environment (Glaasker et al., 1996).

References

Glaasker, E.,W.N. Konings, and B. Poolman. 1996. Osmotic regulation of intracellular solute pools in Lacto-bacillus plantarum. J. Bacteriol. 178:575–582.

Glaasker, E., E.H. Heuberger,W.N. Konings, and B. Poolman. 1998. Mechanism of osmotic adaptation of thequaternary ammonium compound transporter (QacT) of Lactobacillus plantarum. J. Bacteriol.180:5540–5546.

Kleerebezem, M., J. Boekhorst, R. van Kranenburg, D. Molenaar, O.P. Kuipers, R. Leer, R.Tarchini, S.A. Peters,H.M.Sandbrink,M.W.Fiers,W,Stiekema,R.M.Lankhorst,P.A.Bron,S.M.Hoffer,M.N.Groot,R.Kerkhoven,M. de Vries, B. Ursing,W.M. de Vos, and R.J. Siezen. 2003. Complete genome sequence of Lactobacillusplantarum WCFS1. Proc. Natl.Acad. Sci. U.S.A. 100:1990–1995.

Lu, Z., H.P. Fleming, R.F. McFeeters, and S.S.Yoon. 2002. Effects of anions and cations on sugar utilization incucumber juice fermentation. J. Food Sci. 67:1155–1161.

McDonald, L.C., H.P. Fleming, and H.M. Hassan. 1990. Acid tolerance of Leuconostoc mesenteroides andLactobacillus plantarum.Appl. Environ. Microbiol. 56:2120–2124.

Van de Guchte, M., P. Serror, C. Chervaux,T. Smokvina, S.D. Ehrlich, and E. Maguin. 2002. Stress response inlactic acid bacteria. Antonie van Leeuwenhoek 82:187–216.

Box 7–4. In a Pickle: How Lactic Acid Bacteria Deal with Acids and Salts (Continued)



of secreting these enzymes, they may also beproduced by various yeasts,as well as by the cu-cumber flowers. Bacteria, however, do not ap-pear to be a major source of pectin-degradingenzymes.

The fungi responsible for the softening de-fect gain entry into the fermentation tank viatheir association with cucumber flowers.Thus,excluding these constituents from the fermenta-tion may reduce or minimize this defect. Otherpreventative measures include maintaining suf-ficient salt concentrations, acidity, anaerobiosis,and temperature.Another method used to con-trol adventitious microorganisms and to ensurea prompt fermentation (especially when the saltconcentration is at the low end) is to partiallyacidify the cucumbers with acetic acid to aboutpH 4.5 to 5.0.This practice of chemical pasteur-ization is also used when starter cultures andcontrolled fermentation methods are used toproduce pickles (Box 7–3).

Olives: Products and Markets

Olives refer not only to the usually salty, acidicproduct known as table olives, but to the fruitfrom which they are made. The main use ofraw olives, in fact, is as olive oil—more than90% of the total worldwide olive production isused for oil and only 7% to 10% are consumedas table olives.

Olive trees are native to the Middle Eastand, due to their hardy nature (some trees liveas long as 1,000 years), olives have long been amajor agricultural crop throughout the region.Olives and olive oil are among the most fre-quently mentioned foods in the Bible. Oliveproduction subsequently spread from the Mid-dle East across the Mediterranean, to Greece,Italy, France, Spain, and Northern Africa. Oliveswere not introduced to the Americas until the18th century. Currently, four countries—Italy,Spain, Greece, and Turkey—are responsible for75% of the total worldwide olive production(between 9 billion Kg and 15 billion Kg. Incontrast, the United States accounts for lessthan 1% of total world olive production, with

essentially all 90 million Kg (200 millionpounds) being produced in California (2003statistics).

Currently, about 90% to 95% of the olivesgrown in California are used in the manufac-ture of table olives (making California amongthe leaders in table olive production).Althoughfermented olives are common in Europe andother olive-producing regions,most of the tableolives produced and consumed in the UnitedStates are not fermented. In fact,more than 70%of the U.S. olive market consists of olives thatare simply brined and canned (hence, this typeis referred to as California-style olives).

There is also a small (but dedicated) marketfor fresh or raw tree-ripened olives.The famousProvence-style olives of France are of this type.More than thirty cultivars are grown world-wide; however, fermented olives are usuallyproduced using one of eight main varieties:Manzanillo, Gordal, Picholine, Rubra, Mission,Sevillano, Ascolano, and Barouni. These culti-vars differ widely with respect to their compo-sition, size, texture, color, and flavor. Oliveproperties also change during ripening; mostolives are picked when they are still green,straw-yellow, or cherry-red, even if later theyacquire a much darker appearance. Ultimately,their selection for table olive production isbased on the style of the olive produced (dis-cussed below).

Composition

Phenol and polyphenol compounds are com-mon to all olives. Some of these phenol com-pounds contribute to the color of the olive.Others have antimicrobial activity against awide variety of microorganisms (including lac-tic acid bacteria).The phenol fraction of olives(and olive oil) also may have positive dietaryand nutritional benefits, due to their antioxi-dant activities (Box 7–5).The most importantand abundant phenol is oleuropein (part of aclass of compounds called secoiridoids).

Structurally, oleuropein is a glucoside esterof 3,4-dihydroxytyrosol and elenic acid (Figure



7–6). Glucosidic phenols, and oleuropein inparticular, are important in olives owing to thepronounced bitter flavor they impart. Someolive varieties can contain as much as 14%oleuropein (on a dry basis) during the earlystages of growth, although most contain nomore than 3% to 6%.As the olives mature, theoleuropein concentration decreases, and atmaturation,about 1 mg to 2 mg per g of pulp ispresent. Still, between the remaining oleu-ropein and related derivatives, the olives aretoo bitter to consume. Thus, manufacturingprocesses for most olives, including some fer-mented varieties, include sodium hydroxide-treatment steps to remove the bitter oleu-ropein fractions.

Olives,of course contain a significant amountof oil (12% to 30%, depending on cultivar).Thefermentable carbohydrate concentration of ripe

olives generally ranges from 2% to 5%; most ofthis carbohydrate is glucose.However,when theolives are washed or treated to remove the bittercomponents, sugars are also lost.

Manufacture of fermented olives

There are three main styles or types of tableolives, based on their method of production(Figure 7–7). Spanish-style (or green Spanish-style) olives are treated with sodium hydroxide(lye) and fermented. Greek-style or naturally-black, ripe-style olives are not treated with lye,but are fermented. The fermentation for bothtypes is mediated by the natural microflora,much like that for other fermented vegetables(discussed below).The third type of olive is theripe black- or green-style.They are lye-treated,but are not fermented.They may also undergo

Box 7–5. Olives and the Mediterranean Diet

The Mediterranean diet has long been promoted as a model for healthy eating (Willet et al.,1995). This diet takes its name from the foods normally consumed by Greek, Italian, Spanish,and other populations that reside around the Mediterranean Sea.The diet advocates high con-sumption of fruits and vegetables, grains and pasta, legumes, olive oil, and fish; moderate intakeof wine and dairy products; and low intake of meat products.The epidemiological associationof this diet with reduced risks of cancer and heart disease and total mortality is highly signifi-cant (Keys, 1995 and Trichopoulou et al., 2003).

Perhaps the most prominent foods associated with the Mediterranean diet, and among thosethought to be responsible for many of the health benefits, are olives and olive oil. Studies onolive and olive oil components indicate that the phenolic fraction is particularly important (Visioli et al., 2002).As previously noted, phenolic compounds confer bitterness to olives (e.g.,oleuropein) and are involved in color development, and some have antimicrobial activity. Itnow appears that some of these same phenolic compounds also have pharmacological activityand may be responsible, in part, for the health-promoting properties of the Mediterranean diet(D’Angelo et al., 2001).

The main biological activities associated with oleuropein, hydroxytyrosol, tryosal, verbasco-side, and other phenolic compounds found in olives and olive oil are related to their antioxidantproperties (Soler-Rivas et al., 2000). They have been found to inhibit low-density lipoprotein(LDL) oxidation and accumulation of oxidized end products.Oxidation of lipoprotein and otherlipids is considered to be an important step in initiating coronary heart disease. Other cardioac-tive effects associated with these phenols include anti-arrhythmic and cardioprotective activi-ties. Olive phenols also protect human cells against oxidative stress and injury. In addition, ox-idative damage to DNA may be prevented by the free radical scavenging activity of olive phenols.

As noted earlier, most table olives are treated with lye as part of the de-bittering process (theexception being Greek-style, naturally black olives).Although this process hydrolyzes much ofthe bitter oleuropein, olives still contain appreciable amounts of the hydrolysis products,



a special aeration treatment that promotes oxi-dation of pigments and conversion of a greencolor to black. This is the type referred to asCalifornia-style olives.

Spanish-style

Spanish-style olives are harvested when theskin color is green or straw-yellow. They arethen treated with a lye solution for four totwelve hours at 15°C to 20°C to de-bitterizethe olives via hydrolysis of oleuropein. Thelye concentration may range from 0.5% to3.5%, depending on the size and type ofolive. Once the lye has penetrated to just out-side the pit (about two-thirds of the way fromthe skin to the center), the olives are washedin one to three rinse cycles of water to re-move the lye.The pH of the olives after wash-ing should be less than 8.0. Although there

should be little residual lye remaining withthe olives, a slight amount of bitterness maystill be present, which is characteristic ofthese olives. Following the washing, theolives are moved to tanks or barrels and abrine of varying salt concentrations is added.For some olives, 10% to 15% salt brines areused (giving an actual concentration of 6% to 9%), whereas others start with lower saltbrines (5% to 6%), and salt is added later to give comparable final concentrations.Glucose may be added to restore sugars lostduring lye treatment and washing steps.The brined olives are subsequently held at22°C to 26°C.

Although the endogenous microbial popula-tion is reduced by the lye and washing treat-ments, the olive production environment stillcontains a wide assortment of microorganisms.There is, in fact, an opportunity for growth of

elenoic acid glucoside and hydroxytyrosol, as well as other so-called biophenols. One recentstudy (Romero et al., 2004), for example, reported total polyphenol concentrations of 2.4 to11.0 mM in juice from Spanish- and Greek-style olives. Some of the Spanish-style olives con-tained up to 1 g per kg or 0.1%, which is even more than that found in virgin olive oil.The pres-ence of these compounds is now considered so important to human health that researchers inGreece have encouraged the table olive industry to consider modifying processing conditionsto increase polyphenol concentrations (Blekas et al., 2002).

References

Blekas,G.,C.Vassilakis,C.Harizanis,M.Tsimidou, and D.G.Boskou.2002.Biophenols in table olives. J.Agric.Food Chem. 50:3688–3692.

D’Angelo, S., C. Manna,V. Migliardi, O. Mazzoni, P. Morrica, G. Capasso, G. Pontoni, P. Galletti, and V. Zappia.2001. Pharmacokinetics and metabolism of hydroxytyrosol, a natural antioxidant from olive oil. DrugMetab. Dispos. 29:1492–1498.

Keys,A.1995.Mediterranean diet and public health:personal reflections.Am. J.Clin.Nutr.61:1321S–1323S.Romero, C., M. Brenes, K.Yousfi, P. García,A. García, and A. Garrido. 2004. Effect of cultivar and processing

method on the contents of polyphenols in table olives. J.Agri. Food Chem. 52:479–484.Soler-Rivas, C., J.C. Espín, and H.J. Wichers. 2000. Oleuropein and related compounds. J. Sci. Food Agric.

80:1013-1023.Trichopoulou,A.,T. Costacou, C. Bamia, and D.Trichopoulos. 2003.Adherence to a Mediterranean diet and

survival in a Greek population. N.E.J. Med. 348:2599–2608.Visioli, F.,A. Poli, and C. Galli. 2002.Antioxidant and other biological activities of phenols from olives and

olive oil. Med. Res. Rev. 22:65–75.Willett, W.C., F. Sacks, A. Trichopoulou, G. Drescher, A. Ferro-Luzzi, E. Helsing, and D. Trichopoulos.

1995. Mediterranean diet pyramid: a cultural model for healthy eating.Am. J. Clin. Nutr. 61 (Suppl. 6):1402S–1406S.

Box 7–5. Olives and the Mediterranean Diet (Continued)



O

OOCH3

O

HO

HO

CH3CHH O-b-D-glycopyranose

O

Figure 7–6. Structure of oleuropeinfrom olives.

Olives harvestedwhen purple-black

Greek-style

Sort and wash

Fermentation

Brine

Sort and size

Package

Olives harvested when green(with cherry-red blush)

California-style

Wash and sort

Treat with lye (NaOH)

Liquid storage

Wash aerate forblack olives

Pit

Sort and size

Package

Package

Olives harvested when greento straw-yellow

Spanish-style

Treat with lye (NaOH)

Wash

Brine

Fermentation

Pit

Sort and size

Stuff

Package

Package

Package

Figure 7–7. Types of olives and their manufacture. Adapted from Harris, 1998 and Romero et. al., 2004.



coliforms, Pseudomonas, Bacillus, Clostridiumand other Gram negative and Gram positivebacteria during the first two to four days, espe-cially when low salt brines are used, and beforea lactic fermentation can begin in earnest.Thisso-called first stage of the fermentation is thenfollowed by a second stage, initiated primarilyby L. mesenteroides and Pediococcus sp. Oncethe pH nears 5.0, non-lactics are inhibited, andthe brine consists only of lactic acid bacteria.After two to three weeks, L. plantarum, L. bre-vis, Lactobacillus fermentum, and other lacto-bacilli displace the leuconostocs,and eventuallyL. plantarum dominates the third stage of thefermentation. The appearance of facultativeyeasts at the end stage of the fermentation,how-ever, is not uncommon, nor undesirable, sincethey may produce ethanol, acetaldehyde, andother flavor compounds. In contrast, growth ofPropionibacterium acnes is to be avoided,since this organism can raise the pH by fer-menting lactic acid.

When the fermentation is complete, the fi-nal pH after three to four weeks (or longer forbarrel-fermented olives) should be within arange of 3.5 to 4.2, with a titratable acidity (aslactic) of 0.8 to 1.0. Following the fermenta-tion, Spanish-style olives can be pitted, pittedand stuffed (e.g., with pimentos), or packed di-rectly into jars. Pasteurization to promote ex-tended shelf-life is optional.

Greek-style

There are several significant differences be-tween Greek-style and other olive types. First,these olives are naturally black when they areharvested, in contrast to California-style blackolives that rely on oxidation to generate blackpigments. Second, Greek-style olives are notlye-treated, giving them a more bitter flavor.Third, the fermentation is mediated not just bylactic acid bacteria,but also by yeast,non-lacticacid bacteria, and even fungi. Some of thesenon-lactic organisms (e.g., Pseudomonas andEnterobacteriaceae) may actually remain in thebrine, albeit at low levels, for several weeks be-

fore they begin to decline (Figure 7–8).Thus,the fermentation end-products include not justlactic acid, but also acetic acid, citric acid,malic acid, CO2, and ethanol. This mixed fer-mentation may result in a less acidic productwith a final pH as high as 4.5 and an acidity lessthan 0.6%. Lower brine concentrations (5% to10%) may contribute to the more diverse florathat develop in these olives.

Another type of fermented olive that ismade in a very similar manner (i.e.,no lye-treat-ment) is the Sicilian olive.The main differencebetween these olives and the Greek olives isthat the Sicilian olives are green (like Spanisholives). Also, while lactobacilli are the mainbacteria involved in the fermentation, the dom-inant species appears to be Lactobacillus ca-sei, an organism not ordinarily associated withfermented olives.

Ripe- or California-style

California-style olives are the most popularolives consumed in the United States. Both theblack and green versions are produced fromthe same starting material: green olives (with abit of cherry-red blush). For both types, theolives are lye-treated, as described above forSpanish olives, except that several applicationsare used. In the case of green olives, the lye isremoved and the olives are washed in water,and then dilute brine is added. The olives arethen canned (after a pitting step, if desired) andthermally processed as a low-acid canned food.

For black olives, the lye-treated olives areheavily aerated between the lye applicationsto promote darkening reactions. Aeration canbe accomplished either manually, by stirring,or mechanically, by direct injection of air intothe tanks. The latter method is preferred be-cause it is faster and reduces opportunities forspoilage.Air, or more specifically, oxygen, pro-motes a chemical oxidation reaction in whichphenolic compounds are polymerized to formfirst brown, then black pigments. During airexposure, the concentration of two phenols,hydroxytyrosol and caffeic acid in particular,



decreases, as the black color increases. Ioniciron can form complexes with the oxidizedphenols; thus, iron salts are usually added tothese olives to fix the color and prevent fading.

Defects and spoilage

Olives are, generally, more susceptible to mi-crobial spoilage than other fermented vegeta-bles. Initially there is a diverse microflora pre-sent in raw olives and that flora is wellmaintained during the early stages of the fer-mentation. Reflective of this heterogenousmix, spoilage organisms include aerobic bacte-ria and fungi, facultative bacteria and yeasts,strict anaerobes, as well as sporeforming bacte-ria. If the lactic fermentation is delayed due toresidual lye or limiting glucose, or if the pH re-mains high (i.e., above 4.5), organisms capable

of causing spoilage may be given ample oppor-tunity for growth and production of spoilageproducts.

Several of the spoilage defects that occur inolives are not unlike those that occur in picklesand other fermented vegetables (Table 7–4). Ex-cessive gas production and tissue softening aremajor problems in olives, just as they are forpickles. In olive production, for example, gasproduced by coliforms early in the fermentationcan accumulate as gas pockets inside the olivesurface, causing an unsightly blister-like appear-ance called “fish-eyes.” Likewise, pectin hydroly-sis by pectinolytic Penicillium, Fusarium, As-pergillus, and other fungi or, less frequently, bycoliforms, results in a soft tissue defect.

Another defect that is mostly specific forolives is termed zapatera spoilage.This occurslate in the fermentation and is characterized by

BA

Figure 7–8. Fermentation of black olives. Panel A shows growth of total viable bacteria (�), lactic acid bac-teria (� �), Pseudomonas sp. (�), Enterobacteriaceae (�), and yeasts (�) during the fermentation of Greek-style black table olives. The decrease in pH is shown in Panel B. From Nychas et al., 2002.



foul, fecal, cheesy-like odors. It is caused byClostridium sporogenes, Clostridium bu-tyricum, and other putrefactive clostridiawhose presence in fermented foods is never agood thing. These bacteria, along with otheranaerobes, produce butyric acid, sulfur diox-ide, and other sulfur-containing compounds.Putrescine, cadaverine, and other putrefactiveand malodorous end products may also be pro-duced. Propionibacterium sp. also are associ-ated with zapatera spoilage, although they maybe indirectly involved.Growth of Propionibac-terium zeae and other related species can oc-cur at low salt concentrations (�5%), whichleads to an increase in the brine pH (as lacticacid is consumed). If the pH is high enough,clostridia are no longer inhibited.

Control of spoilage organisms can be ac-complished by making sure the salt concentra-tion is sufficiently high (�7.5%) and the pH issufficiently low (�4.0). Olives can be given aquick heat treatment or partially acidified withlactic acid (prior to fermentation) to reduceand control the background flora.Although sel-dom practiced, the addition of a lactic starterculture can provide additional assurance that aprompt fermentation will occur.

Fermented Vegetables and BiogenicAmines

Nearly all fermented foods can potentially con-tain biogenic amines, and fermented vegetablesare no exception.This is because the series ofevents that culminate in biogenic amine forma-tion are essentially the same for fermented veg-etables as they are for other fermented foods.Specifically, accumulation of amines occurswhenever there is an available pool of freeamino acids and a source of amino acid decar-boxylase enzymes.The amino acids are formedfrom protein in the food by the action of mi-crobial proteinases and peptidases. The decar-boxylases are also produced by microorgan-isms, including lactobacilli that may be present

in fermented vegetables. Thus, the amount ofbiogenic amines that are actually produced de-pends on both the concentration of the aminoacid substrates and the expression and activityof the relevant decarboxylases. The biogenicamines that have been found in fermented veg-etables (mainly sauerkraut, kimchi, and fer-mented olives) include histamine (from histi-dine), tyramine (from tyrosine), and putrescineand cadaverine (from lysine,arginine,and gluta-mine). In most cases, however, the concentra-tions that have been reported were less thanthe level ordinarily thought to cause food dis-ease symptoms (1 g/Kg).

BibliographyAyres, J.C., J.O. Mundt, and W.E. Sandine. 1980. Micro-

biology of Foods. Chapter 9, Lactic fermenta-tions, p. 198–230. W.H. Freeman and Company,San Francisco.

Fernández, A.G., M.J.F. Díez, and M.R. Adams. 1997.Table Olives: Production and Processing. Chap-man and Hall, London.

Harris, L.J. 1998.The microbiology of vegetable fer-mentations, p. 45–72. In B.J.B.Wood (ed.) Micro-biology of Fermented Foods, Volume 1, BlackieAcademic and Professional (Chapman and Hall).

Nout, M.R.J., and F.M. Rombouts. 1992. Fermentativepreservation of plant foods. J. Appl. Bacteriol.Symp. Supple. 73:136S–147S.

Nychas, G.-J.E., E.Z. Panagou, M.L. Parker, K.W. Wal-dron, and C.C. Tassou. 2002. Microbial coloniza-tion of naturally black olives during fermentationand associated biochemical activities in thecover brine. Lett.Appl. Microbiol. 34:173–177.

Pederson, C.S. 1960. Sauerkraut. In C.O. Chichester,E.M. Mrak, and G.F. Stewart (ed.), Advances inFood Research. 10:233–291.

Randazzo, C.L., C. Restuccia, A. D. Romano, and C.Caggia. 2004. Lactobacillus casei, dominantspecies in naturally fermented Sicilian greenolives. Int. J. Food Microbiol. 90:9–14.

Romero, C., M. Brenes, K.Yousfi, P. García,A. García,and A. Garrido. 2004. Effect of cultivar and pro-cessing method on the contents of polyphenolsin table olives. J.Agri. Food Chem. 52:479–484.


8

Bread Fermentation

“... cut for yourself, if you will, a slice of bread that you have seen mysteriously riseand redouble and fall and fold under your hands. It will smell better, and taste bet-ter, than you remembered anything could possibly taste or smell, and it will makeyou feel, for a time at least, newborn into a better world than this one often seems.”

From How to Rise Up Like New Bread by M.F.K. Fisher 1942

1960s. During the next thirty years, per capitagrain consumption slowly increased, to nearly70 Kg (Figure 8–1), only to fall back slightly in2000 to just under 65 Kg,presumably as a resultof the popularity of low-carbohydrate diets(more than one-third of Americans think thatbread is “fattening”).

Given the recommendations made in theUSDA’s revised Food Pyramid and 2005 DietaryGuidelines, one might expect that consump-tion of bread (and whole grain breads, in par-ticular) will begin to increase. Still, some coun-tries already have much higher per capitaconsumption of bread,compared to the UnitedStates (Figure 8–2), with annual consumptionof more than 140 Kg per person (more than0.8 pounds per day). Compared to Europeanand other high bread-consuming countries, theUnited States ranks near the bottom for breadconsumption, at only 25 Kg per person peryear (less than 2.5 ounces per day).

History

It is impossible to know exactly when hu-mans first made and ate bread. Ancient arti-facts and writings discovered in the MiddleEast suggest that bread-making had its originsin the eightieth century B.C.E., but it is possi-ble that bread may have been produced even

Introduction

The fermentation that occurs during breadmanufacturing is different from most otherfood fermentations in that the purpose is notto extend the shelf-life of the raw materials,perse,but rather is a means of converting the grainor wheat into a more functional and consum-able form. In fact, in contrast to dairy, meat,vegetable, or wine fermentations, where thestarting material is much more perishable thanthe finished product, the raw material forbread-making, i.e., cereal grains, are better pre-served than the bread that is ultimately pro-duced. It is also interesting to note that in thebread fermentation, again in contrast to otherlactic acid or ethanolic fermentations, essen-tially none of the primary fermentation endproducts actually remain in the food product.

In the United States, bread manufacturing isa $16 billion industry. Despite its commercialand dietary importance,however, consumptionof bread was substantially higher a century agothan it is currently.At the end of the nineteenthcentury, for example, Americans consumedmore than 100 Kg of wheat flour (used mostlyfor bread) per person per year.Consumption ofwheat, however, then began to decline steadilyfor nearly a hundred years, reaching an all-timelow of just 50 Kg per person per year in the

261





several thousands of years earlier. The firstbreads were probably flat breads, with little orno leavening (i.e., fermentation), similar tothose consumed even today in many parts ofthe world. It seems likely that humans beganmaking leavened bread at very near the sametime they began to make other fermentedproducts. Leavening must certainly have been

an accidental discovery made when a flour-wa-ter mixture became contaminated with wildairborne yeasts, leading to a spongy dough thatwas transformed by baking into a light andairy, aromatic, and flavorful product. Perhapsmore so than any other fermented food, breadhas played a major role in the history and cul-ture of human civilization (Box 8–1).

1964 1972 1980 1988 1996 2004

70

45

50

60

55

65

Kg

Year

0

40

80

120

160140

Argen

tina

69

Aust ir

a

47

Belgi

um

71

Denm

ark

60

Fran

ce

84

Ger

man

y

68

Italy

Nethe

rland

s124

Russia

59

Spain

146

Turk

ey

25

Unite

dSta

tes

KgUni

ted

Kingd

om

47

Finl

and

51

101

Bulga

ria

Norway

5055

Czech

Repub

lic

60

Figure 8–2. Worldwide consumption of bread (per person per year). Adapted from Association Interna-tionale de la Boulangeri Industrielle (www.aibi-online.org) and Federation of Bakers (www.bakersfederation.org), 2001-2002 statistics.

Figure 8–1. Wheat flour consumption (perperson per year) in the United States. Adaptedfrom the USDA Economic Research Service.


Bread Fermentation 263

Despite obvious increases in the size andscope of the bread-baking industry, the histori-cal record has revealed that ancient bread-making is not all that different from modernbread manufacturing practices. Given the sim-plicity of the bread manufacturing process,per-haps this observation is not surprising.After all,

bread-making requires only a few ingredients, afew simple mixing and incubation steps,and anoven for baking. In this chapter, the chemical,physical, and biological properties of these in-gredients and the processes used for convert-ing ingredients into doughs and breads will bedescribed.

Box 8–1. Civilization and Bread

It is no coincidence that the history of bread making parallels the history of human civilization.Bread, the oft-quoted “staff of life”is still one of the principle staple foods throughout the world,and has sustained human beings for thousands of years.

Bread may have been one of the earliest of all “processed” foods and was certainly one of thefirst foods to be produced on large scale.Archaeological evidence (recounted in Roberts, 1995and Wood, 1996) indicates that a rather sophisticated bread manufacturing “industry” had de-veloped in Egypt as long ago as 3,000 B.C.E., in large part to provide nourishment for the sub-stantial labor force necessary to build pyramids and other structures.

The very first breads were probably flat breads that had been made using barley, which doesnot work well for leavened breads.Therefore, the production of leavened breads likely coin-cided with the development of suitable ancestral wheat varieties, such as emmer and kamut.Given the open environments in which these factories operated, the dough fermentation wasundoubtedly due to the combined effects of yeasts and bacteria, suggesting that these earlybreads were probably sourdough breads (discussed later). Rather than baking the breads onpans in ovens, the doughs were apparently developed and baked in clay pots arrayed aroundcharcoal-fired pits. However, the use of enclosed ovens eventually became more common asthis early industry evolved.

The importance of grain and bread during the course of human history is evident in theBible, in literature and mythology, and throughout the historical record.Osiris, the Egyptian godof the underworld, was also responsible for vegetation and agriculture; Neper was the Egyptiangod of grain. During the ancient Greek spring festival of Thargelia, bread is offered to themythological gods Artemis and Apollo, symbolizing the first fruits of the year and thanking themfor providing fertile soil.

According to the Old Testament story, when Joseph, son of the Hebrew patriarch Jacob, pre-dicted that a famine was soon to occur in Egypt, the Pharaoh ordered that grain be stored, en-suring a sufficient supply of wheat when the famine eventually occurred. Later, when Hebrewslaves fled from Egypt more than 4,000 years ago, their hasty exodus left no time for theirdough to rise. Instead, they had to eat unleavened bread, a form of which is eaten even today asa symbol of the Jewish holiday of Passover.The Eucharist, one of the sacraments of Catholicism,requires consecration and consumption of bread during Holy Communion. Communion in theProtestant tradition also includes bread, typically unleavened. A blessing over bread is com-monly recited as part of many religious rites.

During the Roman era, bakers had an elevated status and were widely respected. Bread wasso important to the citizenry that it was either provided free or was heavily subsidized. Romansoldiers were even equipped with portable bread-making equipment. Milling technologieswere also developed that could separate wheat bran from endosperm (discussed in more detaillater in this chapter), which made it possible to produce white flour bread.

However, genuinely effective technologies for producing refined white flour were not devel-oped until the 1800s,when the roller mill was invented.For many centuries, this more expensive

(Continued)



Wheat Chemistry and Milling

The most common starting material for mostbreads is wheat flour.Breads are also commonlymade from a wide variety of other cereal grains,including rye, barley, oats, corn, sorghum, andmillet. However, as will be discussed in moredetail below, gluten, a protein complex thatgives bread its structure and elasticity and isnecessary for the leavening process, is poorlyformed or absent in most non-wheat flours.

Thus, although accommodations can some-times be made to account for the absence ofgluten in wheat-free doughs, most commercialbreads contain at least some wheat.

Wheat is one of several cereal grasses in thefamily Poaceae (or Gramineae).The main varietyused for bread is Triticum aestivum.The wheatplant contains leaves, stems and flowers. It iswithin the flowers (referred to as spikelets) thatthe wheat kernels are formed.The kernels are

white flour was prized by society’s upper class, while dark, whole grain breads were left to thepeasantry. In the Middle Ages, baker’s guilds, the forerunners of trade unions, were formed inBritain and Europe to set quality standards and fair prices and to protect their overall interests.Atthe same time, laws were enacted that also established prices and weights for bread.

These laws were apparently necessary because some bread manufacturers resorted to ques-tionable, if not dangerous, practices, such as adding potentially toxic whitening agents to flour.These unscrupulous tactics led some bakers to garner a rather poor reputation. In GeoffreyChaucer’s the Miller’s Tale, one of the great early pieces of literature (and one that all collegestudents no doubt remember), the miller is hardly portrayed as a pillar of the community.

Because bread is such an important food, wheat has become one of the most important agri-cultural crops produced on the planet. For thousands of years, and even today, when wheat harvests are poor, due to weather or plant diseases, famines and starvation conditions often re-sult.Political unrest, trade embargos,and other economic factors have also contributed to short-ages of flour and bread. History books are replete with instances in which rebellions and upris-ings, and even revolutions, occurred when a steady supply of affordable bread becameunavailable.

Queen Marie-Antoinette’s infamous remark about the lack of bread for the hungry children ofFrance (“let them eat cake”) proved to be her undoing, figuratively as well as literally. In themodern era, it was only as recently as the 1980s when citizens of the former Soviet Union hadto wait in long lines just to buy a loaf of bread.This situation no doubt led to significant publicdiscontent and a loss of confidence in the government, both of which may have even con-tributed to the fall of the USSR.

Reliance on cereal grains for bread making is so crucial that during times of famine and war,even mold-infested grains have been used to make bread.Such was the case in the Soviet Unionduring the 1930s and 1940s and through World War II, when farmers were unable to promptlyharvest cereal grains, which then over-wintered and became infected with mycotoxin-produc-ing field fungi. Consumption of bread made from infected flour led to thousands of deaths, andmany more were sickened. In more modern times, wheat has even been used as a “weapon,” asgrain embargoes against several U.S.adversaries are still in place even today. Ironically,Europeanand other countries have threatened to boycott U.S. wheat, if genetically-engineered varietiesare produced.

References

Roberts, D. 1995. After 4,500 years rediscovering Egypt’s bread-baking technology. National Geographic187:32–35.

Wood, E. 1996. World Sourdoughs from Antiquity.Ten Speed Press. Berkeley, California, U.S.A.

Box 8–1. Civilization and Bread (Continued)



essentially seeds, and are surrounded by a hardcovering material designed to protect thewould-be plant from the external elements.Theoriginal or wild wheat seed that grew manythousands of years ago had a tough, hard-to-break husk which made it difficult to use. How-ever, the development of crude hybrids (themixing of different wheat varieties) led to the“domestication” of wheat with more manage-able properties. Moreover, the wheat kernel de-velops in the husk,but the husk is removed dur-ing harvest, unlike barley or oats, where thehusk remains attached to the kernel.

The modern day wheat kernel consists ofthree main constituents: the germ,bran,and en-dosperm (Figure 8–3). The germ portion con-tains the embryonic plant, along with oils, vita-mins, and other nutrients. It represents 2% to3% of the total kernel weight.The bran portion,or coat, represents 12% to 13% of the kernel,and is comprised of multiple distinct layers.Thebran contains mostly cellulose and other fi-

brous carbohydrates, along with proteins, min-erals, and vitamins. The remaining portion,about 85%, is called the endosperm and is themain constituent of wheat flour. It consists ofmostly protein and starch, along with some wa-ter (12% to 14%) and a small amount of lipid(1%). Kernels that are crushed or milled with-out regard to separation of the individual frac-tions yield a “whole”wheat flour containing thegerm, bran, and endosperm. Usually, however,these three constituents are separated, eitherpartially or nearly completely, in the millingstep,yielding refined flours of endosperm,withlittle or no germ or bran (Box 8–2).

Flour Composition

The specific composition of flour is criticallyimportant because it has a major influence onthe fermentation as well as the physical struc-ture of the dough and finished bread. Wheatflour, after all, is the primary ingredient in most

germ

endosperm

bran

Figure 8–3. Structure of a wheat kernel. Adaptedfrom the Montana Wheat & Barley Committee.



Box 8–2. Principles of Milling Wheat

The object of wheat milling is to extract the flour from the wheat kernel.This is such an integralstep in the manufacture of bread that an entire milling industry has developed whose purposeis to do just that—convert wheat into flour.Originally, some thousands of years ago,milling wasdone in a single step, as the kernels were ground or pounded manually by a crude mortar andpestle or between stones.

Even today, single-pass milling is performed to make whole wheat flour, although the millingtechnologies have obviously improved and are now mechanized and automated. Still, stone-ground flours are widely available and are prized for their quality. Functionally, however, stone-grinding has no advantages over other size reduction techniques, such as a hammer mill.

The modern milling operation begins with various wheat handling steps, including sampling,inspection, and testing; screening; and conditioning (Catterall, 1998). Following these steps,which ensure that the wheat has the appropriate composition and quality and is free of debrisand impurities, the wheat is ready for milling.Although the whole wheat flours produced by asingle-pass grinding step have wide appeal, most breads are made using flours that contain re-duced levels (or even none) of the germ and bran. It is the job of the wheat miller to separatethese wheat constituents from the main component, the endosperm.

In principle, this can be done simply by successive grinding and separation steps, such thatthe resulting fractions can either be removed or further processed. In actual practice, the wheatkernels are passed through a series of paired-rollers or stones whose gap distances betweeneach pair is successively more narrow. Modern grinding devices consist of cast iron breakrollers, constructed with flutes or ridges to enhance the grinding or shearing action (Figure 1).The rolls run at different speeds to give a shearing, rather than a crushing, action. In betweeneach so-called break steps the ground wheat is passed through sieves that retains the coarse ma-terial (containing bran still stuck to the endosperm), while permitting the fine flour to pass.Along the way, an air purifying device is also used to further separate the different fraction.Eventually, the germ and bran-free white flour that is the most common type used for breadmanufacture is produced.

Figure 1. A simplified diagram of wheat milling. Adaptedfrom www.fabflour.co.uk.



breads. As noted above, however, most of theflour used for bread manufacture in the UnitedStates is refined white flour, containing only theendosperm portion. It consists of two maincomponents—protein and starch—with a smallamount of hemicellulose and lipid.

Protein

About 8% to 15% of wheat flour is protein.This wide range reflects the various types orclasses of wheat used as a source of the flour(Table 8–1). Protein content, as well as pro-tein quality of a given flour, dictates the use ofthat flour. High protein flours, derived fromhard wheat, generally contain more than 11%protein and are best used for bread. In con-trast, low-protein flours, derived from softwheat, contain 9% or less protein, and are

most often used for cakes, cookies, and pas-tries. It is the protein than provides the sup-port matrix necessary to retain the carbondioxide made during fermentation.Therefore,in bread making, protein content has a majorimpact on dough expansion and loaf volume.

The protein fraction actually contains sev-eral different proteins. The most important ofthese are gliadin and glutenin, which accountfor 85% of the total protein.When gliadin andglutenin are hydrated and mixed, they form a

The milling operation yields dozens of different product streams (Calvel, 2001). Flour re-moved from the early and late separations is called clear or common flour. It is generally less refined, higher in protein (�14%), and somewhat darker in color. Patent flour is obtained fromthe intermediate separation steps. It has the least amount of germ and bran, and, therefore hasthe whitest appearance. Patent flour is still high in protein (13%) and is considered to have thebest bread-making quality (at least in the United States and Canada).

In many parts of the world, it is a standard practice to re-combine the different flour fractionsto give flours of varying composition, ranging from a straight flour containing all of the sepa-rated fractions, to different patent flours containing 60% to 80% of the original wheat.This al-lows baker to choose a flour based on desired strength (i.e., protein content) or functionalproperties (i.e., dough elasticity, bread flavor, and oven spring).

It is important that a high yield be achieved during the milling process.The yield is given asthe extraction rate, and is expressed as the amount of flour that is obtained from the wheat. Ingeneral, extraction rates of about 72% are achieved, meaning that 72 Kg of straight flour can beobtained from 100 Kg of wheat.The balance (about 28%) includes mainly bran and germ. How-ever, since 85% of the wheat is endosperm, some of the remaining material consists of en-dosperm that adheres to the bran layers (Pyler, 1988).

References

Calvel, R. 2001. The Taste of Bread. (R.L.Wirtz, translator and J.J. MacGuire, technical editor). Aspen Pub-lishers, Inc. Gaithersburg, Maryland.

Catterall, P., 1998. Flour milling, p. 296–329. In S.P. Cauvin and L.S.Young, (ed.), Technology of breadmak-ing. Blackie Academic and Professional (Chapman and Hall). London, UK.

Pyler, E.J., 1988. Wheat flour, p. 300–377. In Baking Science and Technology. Volume 1, 3rd edition,Sosland Publishing Co. Merriam, Kansas.

Box 8–2. Principles of Milling Wheat (Continued)

Table 8.1. Major types of wheat and their applications.

Type % Protein Application

Hard winter 10 – 13 BreadHard spring 13 – 16 BreadSoft winter 8 – 10 Cakes, cookies, pastriesDurum 12 – 14 Pasta



complex called gluten, which is a key compo-nent of bread dough. The remaining proteins(15% of the total) consist of other globulins andalbumins. Several enzymes are also present, inparticular, �-amylase and �-amylase, that bothplay an important role (discussed below).

Carbohydrates

Carbohydrates represent the main fraction offlour, accounting for up to 75% of the totalweight. This fraction is largely comprised ofstarch, although other carbohydrates are alsopresent, including a small amount (about 1%)of simple sugars, cellulose, and fiber.The maincarbohydrate component, however, is starch,which consists of amylose,an �-1,4 glucose lin-ear polymer (about 4,000 glucose monomersper molecule) and amylopectin,an �-1,4 and �-1,6 glucose branched polymer (about 100,000glucose monomers per molecule). About 20%to 25% of the starch fraction is amylose and70% to 75% is amylopectin. Properties of thesetwo fractions are described in Table 8–2.

In its native state (i.e., before milling),wheat starch exists in the form of starch gran-ules. The amylose and amylopectin are con-tained within these spherical granules in arigid, semi-crystalline network. Native starchgranules are insoluble and resist water pene-tration. However, some of the starch granules

(3% to 5%) are damaged during milling, whichenhances absorption of water and exposes theamylose and amylopectin to hydrolytic en-zymes, such as �-amylase.

Yeast Cultures

The yeast used for bread manufacture is Sac-charomyces cerevisiae, often referred to assimply bakers’ yeast.Although ale (Chapter 9),wine (Chapter 10), and various distilled alco-holic beverages are also made using S. cere-visiae, the cultures for each of these productsare not interchangeable. The bakers’ yeaststrains of S. cerevisiae clearly have propertiesand performance characteristics especiallysuited for bread manufacture (Table 8–3). Forexample, bread strains are selected, in part, ontheir ability to produce CO2, or their gassingrate. It is the CO2,evolved during fermentation,that is responsible for leavening. Yeast strainsshould also produce good bread flavor. In addi-tion,stability and viability during storage is alsoimportant in selecting suitable yeast strains.

Modern industrial production of bakers’yeaststarts with a pure stock culture, which is thenscaled up through a series of fermentors until alarge cell mass is produced (Figure 8–4).The cul-ture is propagated initially in small (1 L to 5 L)seed flasks, and then in progressively larger fer-mentors of about 250 L to 1,000 L. Eventuallythe culture is inoculated into large (250,000 L)production fermentors. The growth mediumusually consists of molasses or another inexpen-sive source of sugar and various ammonium salts(e.g.,ammonium hydroxide,as a cheap source ofnitrogen). Other yeast nutrients include ammo-nium phosphate,magnesium sulfate,calcium sul-

Table 8.2. Physical and chemical properties of amylaseand amylopectin.

Property Amylose Amylopectin

Monomers per 300 — 1,000 � 5,000molecule

Linkage �-1,4 �-1,4 and �-1,6

Conformation Linear Branched% in wheat 25 75

starchRetrogradation Fast Slow

rate

Products of:�-amylase Dextrins Limit dextrins�-amylase Maltose Maltose, limit

dextrins

Table 8.3. Desirable properties of bakers’ yeasts.

Gassing powerFlavor developmentStable to dryingStable during storageEasy to dispenseEthanol tolerantCryotolerant



flask

pureculturefermentor

beetmolasses

canemolasses

water

sterilizer

clarifierseed

yeast

nutrients

productionfermentor

pH andfoam control

cooling water

filtered air

water

cooler

centrifuge

yeastbroth

yeast creamstorage tank

cream yeast

salt, starch

water

rotaryvacuumfilter

bag yeastpackaging

granularyeast

granulatorextruderand cutter

yeast cooler

palletizers

cakeyeast

cake yeastpackaging

distribution

sterilemashtank

emulsifieror oil

seedfermen.

semi-seedfermentor

vial

Figure 8–4. Industrial production of bakers’ yeast. From Lallemand Baking Update, Volume1/Number 9 (with permission).



fate, and lesser amounts of the trace mineralszinc and iron. Since the goal is to produce cellmass as the end product (rather than metabolicproducts), growth occurs under highly aerobicconditions by providing continuous agitationand oxygen (or air). In addition, the yeast propa-gation step is performed at optimum tempera-ture and under pH control (usually around 30°Cand at a pH of 4.0 to 5.0), with nutrients pro-vided on a continuous basis. At the end of thegrowth fermentation, the yeast is collected bycentrifugation, resulting in a yeast “cream” thatcontains about 20% total yeast solids and a yeastconcentration of about 1010 cells/g.

Bakers’ yeast preparations are available inseveral forms.The yeast cream can be used di-rectly, although this form is highly perishable.Most commercial bakers use compressed yeastcultures.These are produced by pumping theyeast cream through a filtration press or vac-uum filter to remove more of the water. Theyeast is collected in the form of moist cakes,separated by wax paper. Compressed yeastcakes (about 30 cm x 30 cm x 2 cm) still havea high moisture content (70% to 75%), requirerefrigeration, and last only a few weeks. How-ever,because the cells are metabolically active,once they are introduced into the dough, fer-mentation can occur very quickly.

Compressed yeast can be further dried toabout 90% solids to provide dry active yeast.This is the form that is familiar to consumerswho make homemade bread, but small manu-facturing operations or those located wherecompressed yeast is not available also use dryyeast preparations. Dry active yeast prepara-tions last six months or longer, even at roomtemperature.They do require a hydration step,and in general, are not as active as compressedyeast, although improved drying technologieshave greatly enhanced the activity of driedyeast. In addition, dry active yeasts can be “in-stantized”such that they rehydrate quickly.

Bread Manufacturing Principles

In general, bread manufacture involves just afew steps. First the ingredients are assembled,

weighed, and mixed to make a dough and the“bulk” dough is allowed to ferment. The fer-mented dough is then portioned and shaped,given a second opportunity to ferment, andthen baked, cooled, sliced, and packaged. Ofcourse, the actual procedures used by modernbakeries are often more involved, as will be de-scribed later.Described below is what is essen-tially the “straight dough”system.

Ingredients

Most ingredient statements on a loaf of com-mercial bread are a paragraph long and listthirty or more ingredients. However, only fouringredients—flour, water, salt, and bakers’syeast—are actually required to make a per-fectly acceptable, if not exceptional, bread.Al-though the other ingredients do indeed pro-vide important functional properties, not onlyin the finished product, but also during thedough development steps, they are strictly op-tional. Of course, one could argue that con-sumer preferences for breads with a softcrumb texture; long,mold-free shelf-life; and re-sistance to staling have led to such widespreaduse of some of these ingredients that they canhardly be considered “optional.” Moreover,large scale manufacturing operations that de-mand high throughput could not functionnearly as well without these extra ingredientsand process aids.

Wheat flour is the main ingredient of bread,representing about 60% to 70% of a typical for-mulation.The flour contains the proteins thatare essential for dough formation and thestarch that absorbs water and serves as an en-ergy source for the yeasts. Water, added atabout 30% to 40%, acts as the solvent neces-sary to hydrate the flour and other ingredients.Salt, at 1% to 2%, toughens the gluten, controlsthe fermentation, and gives a desirable flavor.Finally, the yeast, also added at about 1% to 2%,provides the means by which leavening andflavor formation occur.

Many optional ingredients are often in-cluded in bread formulations, depending onthe needs and wants of the manufacturer.



• Sugars. In the United States, many large-scale bakeries add either sucrose or glucose(about 2% to 3%) as an additional source ofreadily fermentable sugars.They also supplyflavor and, when the dough is baked, color.Wheat flour contains only modest amountsof maltose and glucose, and although yeastsdo express amylases that can release fer-mentable sugars from starch, sugar availabil-ity may still be growth-limiting.

• Enzymes. Another way to increase theamount of free sugars in the dough is to add�- and �-amylases, enzymes that specificallyhydrolyze the �-1,4 glucosidic bonds of amy-lose and amylopectin (Figure 8–5). Flour nat-urally contains both of these enzymes, but �-amylase, in particular, is present at very lowlevels. These enzymes are also available inthe form of microbial preparations or in theform of malt. Not only do these enzymes col-lectively increase the free sugar concentra-tion by hydrolyzing amylose and amy-lopectin to maltose, but they also canenhance bread quality. Controlled hydrolysisof starch—especially the amylopectin por-tion—by �-amylase tends to increase loafvolume and,by virtue of softening the crumbtexture, staling is delayed. Of course, if toomuch enzyme is added, the dough will besticky and unmanageable and the excess freesugars that are formed could contribute toover-browning during baking.

One other important factor that must beconsidered when enzymes are used asprocess aids is their temperature inactivationprofile. Heat-resistant amylases, such as thoseproduced by bacteria,may survive baking andcontinue to hydrolyze starch, which could re-sult in a soft and sticky bread.Fungal enzymesare generally heat labile and are inactivatedduring baking.

Increasing the amylase activity in dough,and thereby increasing the free sugar con-centration, can also be accomplished byadding malt powder. Malt, an essential ingre-dient in beer manufacture (see Chapter 9), isprepared from germinated grains, usuallybarley and wheat. Malt contains both �- and

�-amylases that are both somewhat moreheat stable than fungal enzymes.

Finally, the addition of proteolytic en-zymes has been advocated as a means ofachieving partial hydrolysis of the gluten. Intheory, this would reduce mixing time andwould make a softer dough.

• Fat.The addition of 0.1% to 0.2% fat,as eithera shortening or oil, is now commonplace inmost commercial breads. Whereas a genera-tion or two ago, animal fats with excellent“shortening” properties were used in breads,the use of such fats has about disappeared inthe United States.Now most breads are madewith partially hydrogenated vegetable oils.The composition of these oils is not that dif-ferent from animal fats, in that they containmostly long chain,saturated fatty acids,with ahigh melting point. Not surprisingly, they im-part similar properties as the animal fats, giv-ing a soft,cake-like texture.The precise mech-anism, however, for how these functionaleffects occur is not clearly known. Fats withhigh-melting points (i.e., containing long, sat-urated fatty acids, such as palmitic and stearicacids) clearly provide discernable improve-ments, especially for short-proof breads.There is currently a movement to use non-hydrogenated oils to eliminate trans fattyacids.This does change the dough propertiescompared to the use of hydrogenated veg-etable shortenings.

• Yeast nutrients. Various nutrients can beadded to the dough mixture to enhancegrowth of the yeast, including ammoniumsulfate, ammonium chloride, and ammoniumphosphate, all added as sources of nitrogen.Phosphate and carbonate salts may also beadded to adjust the acidity or alkalinity, andcalcium salts can be added to increase min-eral content and water hardness.

• Vitamins. There has been a flour and breadenrichment program in the United States formore than sixty years. Flour currently is forti-fied with four B vitamins—thiamine, ribo-flavin, niacin, and folic acid—and one min-eral, iron. However, non-enriched flour is alsoavailable and bread manufacturers can enrich



OC

H2O

HO

CH

2OH

OC

H2O

HO

CH

2OH

OC

H2O

H

OO

OO

O

OC

H2O

HO

CH

2OH

OC

H2O

H

OO

O

OC

H2O

HO

CH

2OH

OC

H2O

HO

CH

2OH

OC

H2O

H

OO

OO

O

a-a

myl

ase

a-a

myl

ase

OC

H2O

HO

CH

2OH

OC

H2O

H

OO

OC

H2O

H

O

OC

H2O

HO

CH

2OH

OC

H2O

HO

CH

2OH

OC

H2O

H

OO

OO

OC

H2O

HO

CH

2OH

OOam

ylos

e

amyl

ose

OC

H2O

HO

CH

2OH

O

OC

H2O

HO

CH

2OH

OC

H2O

H

OO

O

OC

H2O

HO

CH

2OH

OC

H2O

HO

CH

2OH

OC

H2O

H

OO

OO

O

OC

H2O

HO

CH

2OH

OC

H2O

HO

CH

2OH

OC

H2O

HO

CH

2OH

OO

OO

OO

b-a

myl

ase

Fig

ure

8–5

.H

ydro

lysi

s o

f am

ylo

se a

nd

am

ylo

pec

tin

by

�-

and

�-a

myl

ases

. Th

e to

p a

nd

mid

dle

pan

els

(A a

nd

B)

sho

w r

eact

ion

pro

du

cts

of

�-a

myl

ase

and

�-a

myl

ase

on

am

ylo

se.

The

bo

tto

m p

anel

(C

) sh

ow

s re

acti

on

pro

du

cts

of

�-a

myl

ase

and

�-a

myl

ase

on

amyl

op

ecti

n.



OC

H2O

HO

CH

2OH

O

OC

H2O

HO

CH

2OH

OC

H2O

HO

CH

2

OC

H2O

H

OO

OO

OC

H2O

HO

CH

2OH

OO

O OC

H2O

HO

CH

2OH

OC

H2

OC

H2O

HO

CH

2OH

OO

OO

O

O

OC

H2O

HO

CH

2OH

O

O

O

OC

H2O

HO

CH

2OH

OC

H2

OC

H2O

HO

CH

2OH

OO

OO

OC

H2O

HO

CH

2OH

OO

O

OC

H2O

HO

CH

2OH

OC

H2

OC

H2O

HO

CH

2OH

OO

OO

OO

O

OC

H2O

HO

CH

2OH

OO

OC

H2O

HO

CH

2OH

OC

H2

OC

H2O

HO

CH

2OH

OO

OO

O

O

OC

H2O

H

O

b-a

myl

ase

a-a

myl

ase

a-a

myl

ase

OC

H2O

H

amyl

opec

tin

Fig

ure

8–5

.(C

ontin

ued)



bread by adding these nutrients directly tothe dough.

• Dough improvers. Mixing the ingredientsand making an elastic dough can often be aslow process, due to the sticky gluten proteinthat forms when the flour is hydrated. Reduc-ing agents, such as cysteine, are used to speedup dough mixing, by decreasing the numberof disulfide cross-links that make gluten sohighly elastic (as described in the next sec-tion).In other words,these agents weaken thedough structure and decrease mixing times.Alternatively,other bread dough improvers in-clude oxidizing agents, such as ascorbic acidand potassium bromate, that can improvedough structure by increasing the number ofdisulfide bonds (and cross-linkages) in thegluten network.Their effect is to increase elas-ticity and gas retention. Bromates have longbeen considered as the most effective oxidiz-ing agent; however, recent questions regard-ing their safety has led some government au-thorities to discourage or prohibit the use ofthis additive in bread manufacture.

• Biological Preservatives. Biological spoil-age of bread (in contrast to staling or otherchemical-physical causes) is invariably causedby fungi. Thus, mold inhibitors are routinelyadded to most breads. The use of these in-hibitors is somewhat less common today thantwenty years ago, however, due to the in-creased consumer demand for natural foodsthat are free of chemical preservatives. By far,the most common anti-fungal agents are thesalts of weak organic acids; they includepotassium acetate, sodium diacetate, sodiumpropionate, and calcium propionate. The lat-ter is the more effective and is the mostwidely used bread preservative in the UnitedStates.As illustrated in Box 8–3, the mode ofaction of these agents against target cells issimilar and depends on the form of the acidand the pH of the food. They are typicallyadded at 0.15% to 0.2% of the flour weight;the maximum amount allowed for acetatesand propionates in the United States is 0.32%and 0.40%, respectively (less in Europe). It isimportant to note that, despite their effective-

ness against fungi, these agents are only in-hibitory and are not cidal.They also are not in-hibitory to bakers’ yeast, and, therefore, haveno effect on the fermentation.

• Emulsifiers. Although bread dough does notordinarily require emulsification, emulsifyingagents may improve the functional propertiesof the dough by increasing water absorptionand gas retention, decreasing proofing times,and reducing the staling rate. As such, breademulsifiers are perhaps best considered asdough conditioners. By far, the most com-monly added emulsifiers are mono- and di-glycerides, derived from hydrogenated veg-etable oils.In the United States,they are addedat a concentration of 0.5% (flour weight).

• Gluten. Gluten, as discussed below, isformed naturally during dough hydrationand mixing steps. However, it is also a com-mon practice to add dried gluten, in the formof vital wheat gluten, as an ingredient to thedough mixture. Not only does the addedgluten increase the protein content of thebread, but it also increases loaf volume andextends shelf-life. In addition, high glutenbreads may have improved texture proper-ties by virtue of strengthening the doughand bread structure. In general, vital gluten ismost commonly added to flour in crop yearswhen protein quality or quantity are low; it isalso added to whole grain and specialtybreads to increase loaf volume.

Hydration and Mixing

Depending on the nature of the ingredients(solid or liquid), they are either weighed ormetered into mixing vats. Once all of the in-gredients are combined, they are vigorouslymixed. This is a very important step becausethe flour particles and starch granules are hardand dense and water penetrates slowly.Mixing,therefore, is the primary driving force for thewater molecules to diffuse into the wheat par-ticles. Mixing also acts to distribute the yeastcells, yeast nutrients, salt, air, and other ingredi-ents throughout the dough. Mixing, in otherwords, is necessary to develop the dough.



Box 8–3. How Propionates Preserve Bread

Calcium propionate (Figure 1) is one of the most widely used antimicrobial preservatives in thefermented foods industry, due to its widespread application as an anti-fungal agent in bread. Italso inhibits rope-producing species of Bacillus.Although it is the propionic acid moiety that isthe active agent, the calcium salt of this weak organic acid is used commercially since it is read-ily soluble and is easier to handle. Calcium propionate is a GRAS substance (generally recog-nized as safe), and occurs naturally in various cheeses (e.g., Swiss cheese).

The inhibitory activity of calcium propionate against fungi is due to its ability to penetratethe cytoplasmic membrane and gain entry inside the fungal cell. However, propionate can onlytraverse the membrane when it is in the acid form.This is because the lipid-rich, non-polar, hy-drophobic cytoplasmic membranes of cells are largely impermeable to charged, highly polarmolecules, such as the anion or salt form of organic acids. Rather, only the uncharged, undisso-ciated, lipophilic acid form can diffuse across the membrane.

Ca++H

C

H

H

C

H

H

C

OH C

H

H

C

H

H

C -O

O

-O

(Continued)

Figure 1. Structure of calcium propionate.

How then, does calcium propionate exerts its anti-fungal activity? Recall that weak organicacids can exist in either the acid (undissociated) or salt (dissociated) form (Figure 2, upperpanel).When the pH is equal to the pKa (the dissociation constant) of the acid, the concentra-tion of undissociated and dissociated forms (at equilibrium) will be the same.When the pH isabove the pKa, the equilibrium shifts to the salt or dissociated form.Conversely,when the pH isbelow the pKa, the acid or undissociated form predominates. It is this uncharged, undissoci-ated, lipophilic form of the acid that can diffuse across the cytoplasmic membrane. Of course,once the acid reaches the cytoplasmic environment,where the pH is usually near neutral, it dis-sociates, in accord with its pKa, forming the anion and releasing a proton. Not only does the or-ganism have to spend valuable energy (in the form of ATP) to expel these protons, but their ac-cumulation in the cytoplasm eventually will result in a decrease in the intracellular pH to apoint that is inhibitory to the cell.



The pKa of propionic acid is 4.87,and,according to the Henderson-Hasselbalch equation,cal-cium propionate would be in the acid form below this pH and in the salt form above this pH. Itshould be noted that the ordinary pH range of bread is between 5.0 and 6.0 (i.e., above the pKaof propionic acid).However,enough of the acid form will still be present to diffuse into the celland cause inhibitory effects (Figure 2, middle and lower panels).

Box 8–3. How Propionates Preserve Bread (Continued)

pH pKa= + log[A-]

[HA]1.

5.20 4.87= + log[A-]

[HA]2.

2.14 =[A-]

[HA]4.

0.33 = log[A-]

[HA]3.

4.40

pH [A-] [HA]

4.80

5.20

[A-] / [HA]

5.60

6.00

0.34

0.85

2.14

5.37

13.49

0.05

0.09

0.13

0.17

0.19

1.15

0.11

0.07

0.03

0.01

outpH < pKa

inpH > pKa

C

O

CC OH

H H

H H

H C

O

CC O- + H+H H

H H

H

C

O

CC OH

H H

H H

H C

O

CC O- + H+H H

H H

H

Figure 2. Mode of action of propionate.The cell membrane is permeable only to theacid form (top panel). The ratio of propionate(A�) to propionic acid (HA) at a typical breadpH of 5.2 can be calculated from the Hender-son-Hasselbalch equation (middle panel). If0.2% propionate is added to bread, the con-centrations (%) of propionate and propionicacid at other pH values can be similarly deter-mined (lower panel).



In the home, a good strong pair of hands cando the mixing and kneading, but commercialbread manufacturers rely on large mixers thatknead and tumble the dough.There are actuallymany different types of mixers, ranging fromlow- or slow-speed mixers to high-speed devicesthat require as little as five minutes to completethe job.The latter are used for processes, such as no-time bread manufacturing methods (dis-cussed later) that rely on mechanical develop-ment of the dough and shorter fermentations.This is in contrast to traditional low-speed mix-ing methods in which dough development (de-fined loosely as the cumulative changes that givethe dough its necessary physical and chemicalproperties) depends on an extended period ofyeast activity and fermentation.

Several events occur during mixing. Uponhydration, gliadin and glutenin form a visco-elastic gluten complex. The sulfur-containingamino acid cysteine is present in both gliadinand glutenin, where it forms disulfide bonds orcrosslinks between poypeptide chains. How-ever, in gliadin the disulfide crosslinks occurwithin a single protein molecule, whereas it isgenerally believed that the disulfide crosslinksin glutenin form between different proteinstrands (according to most cereal chemists).The more crosslinking, the more elastic thedough. Dough mixing and elasticity are af-fected by oxidizing and reducing agents, as de-scribed above.Eventually,as the dough is mixedand the gluten complex is formed,a continuousnetwork or film will surround the starch gran-ules. Naturally occurring or exogenous amy-lases then begin to hydrolyze the starch, gener-ating maltose, maltodextrins, and other sugars.

Fermentation

Yeast growth is initiated as soon as the flour,wa-ter, yeasts, and other ingredients are combinedand the dough is adequately mixed.A lag phaseusually occurs, the duration of which dependson the form of the yeast and the availability offermentable sugars. Bakers’ yeast (S. cerevisiae)has a facultative metabolism,meaning that it can

use glucose by either aerobic (i.e., via the tricar-boxylic acid or TCA cycle) or anaerobic path-ways (Figure 8–6, upper panel). The formerpathway yields much more cell mass and moreATP per glucose than the anaerobic pathway.However, despite the incorporation of oxygeninto the dough during the mixing step, oxida-tive metabolism of carbohydrates occurs onlybriefly, if at all. Instead,carbohydrates are metab-olized by the glycolytic fermentative pathway.This is due, in part, to the presence of glucose,which inhibits synthesis of TCA cycle enzymesvia catabolite repression, but also because thedough quickly becomes anaerobic due to theevolved CO2. Also, as shown in Figure 8–6(lower panel), the ethanol-forming reduction re-action generates oxidized NAD, which is neces-sary to maintain glycolysis.

Sugar metabolism by bakers’ yeast

Because several sugars are ordinarily present inthe dough (mainly glucose and maltose), andothers may be added (sucrose, high fructosecorn syrup), the yeast has a variety of metabolicsubstrates from which to choose. In addition,more maltose may be formed in the dough dur-ing the fermentation step via the successive ac-tion of endogenous �- and �-amylases presentor added to the flour and that act on damagedstarch granules.The order in which these differ-ent carbohydrates are fermented by S. cere-visiae is not random, but rather is based on aspecific hierarchy, with glucose being the pre-ferred sugar.

For the most part, regulation is mediatedby catabolite repression, acting at early stepsin various catabolic pathways. Thus, in moststrains of S. cerevisiae, glucose repressesgenes responsible for maltose transport andhydrolysis, as well as the invertase that hy-drolyzes sucrose to glucose and fructose(Figure 8–7). Consequently, in a dough con-taining glucose, sucrose, and maltose, the dis-accharides will be fermented only when theglucose is consumed. Moreover, maltose re-presses invertase expression, so sucrose



would be the last sugar fermented in thismixture. Depending on how the cells hadpreviously been grown, however, it is possi-ble that invertase had already been induced,resulting in rapid formation of sucrose hy-drolysis products. Finally, there exist strainsof S. cerevisiae whose expression of cata-bolic genes is constitutive, meaning that theyare not subject to catabolite repression. Suchstrains may be particularly useful, since con-stitutive expression of genes coding for malt-ose permease and maltase means that thesestrains will ferment maltose even in the pres-ence of glucose. In most doughs, however,

glucose is metabolized rather quickly, andthe maltose utilization genes would soon bede-repressed and induced, even in non-con-stitutive strains.

Sugar transport

As shown in Figure 8–7, glucose and fructoseare both transported by common hexose trans-porters. Genetic evidence has revealed thatthere may actually be as many as twenty suchtransporters in S. cerevisiae. Although it is notknown how different these transporters are toone another, they all transport their substrates

½ glucose pyruvate

ethanol acetaldehyde

NADH CO2NAD

ATP

glucose CO2aerobic

(+ biomass and ATP)

glucose pyruvateanaerobic

ethanol + CO2 (+ ATP)

1.

2.

Figure 8–6. Biochemical reactions in Saccharomyces cerevisiae. Aerobic metabolism (upper panel, reac-tion 1) occurs via the tricarboxylic acid cycle, resulting in CO2 and cell mass, and an ATP yield of 38 molesper mole of glucose oxidized. Under anaerobic conditions (reaction 2), the Embden-Meyerhoff-Parnas gly-colytic pathway yields ethanol and CO2, an ATP yield of 2 moles of ATP per mole of glucose fermented, andlittle cell mass. The regeneration of oxidized NAD by acetaldehyde dehydrogenase is essential (lower panel)to maintain glycolytic flux.



via facilitated diffusion. No energy is requiredand instead the driving force is simply the os-motic gradient. It is also clear that they varywith respect to their substrate affinities.For ex-ample, there are hexose transporters that havehigh affinity for glucose (with affinity con-stants or Km values �5 �M) or very low affin-ity (Km � 1 M). The presence of low-affinityand high-affinity systems provides the cell withthe versatility necessary to transport sugars un-der a wide range of available concentrations.Thus,yeasts are especially suited to grow in en-vironments containing glucose and fructose,since their uptake costs are cheap (in terms ofenergy expenditure), and they have carriersthat function even when substrates are notplentiful (for example, in lean doughs whereno exogenous source of sugar is added).

Glycolysis

Following transport, the accumulated mono-saccharides are rapidly phosphorylated byhexokinases or glucokinases.As for the hexosetransport systems, hexokinase enzymes (thereare at least two, hexokinase PI and PII) havebroad substrate specificity, being able to phos-phorylate glucose and fructose, as well as man-nose.The sugar phosphates then feed into theglycolytic or Embden-Meyerhoff (EM) pathway(Figure 8–8), with the eventual formation ofpyruvate. Reducing equivalents, in the form ofNADH, are formed from oxidized NAD by theglycolytic enzyme,glyceraldehyde-3-phosphatedehydrogenase.To maintain glycolytic metabo-lism during active growth, therefore, it is nec-essary to restore the NAD. As described inChapter 2, lactic acid bacteria reduce pyruvate

out

in

H+

H+maltose

maltosepermeasemembrane

hexosetransporter

maltosefructose glucose

fructose glucose+sucroseinvertase

maltase

fru-6-P glu-6-P

fru-1,6-BP

EMP pathway

hexokinaseATP

ADP

ATP

ADPphosphofructokinase

Figure 8–7. Transport and metabolism of glucose, fructose, maltose, and sucrose by Saccharomyces cere-visiae. Fructose and glucose are transported by one of several different hexose transporters via facilitated dif-fusion. Sucrose is hydrolyzed by a secreted invertase. Maltose transport occurs via a proton symport-mediatedmaltose permease. Once inside the cell, maltose is hydrolyzed by maltase to give free glucose. The accumu-lated monosaccharides are phosphorylated to glucose-6-phosphate (glu-6-P) and fructose-6-phosphate (fru-6-P) by hexokinases, then to fructose-1,6-bisphosphate (fru-1,6-BP) which feeds directly into the Embden-Meyerhoff-Parnas (EMP) pathway.



glucose

hexokinaseATP

ADP

glucose-6-phosphate

fructose-6-phosphate

phosphoglucose isomerase


phosphofructokinaseATP

ADP

aldolase

glyceraldehyde-3-phosphate dihydroxyacetone phosphate

glyceraldehyde-3-phosphate dehydrogenase

NAD

NADH +H

P

1,3-diphosphoglycerateADP

ATP3-phosphoglycerate kinase

3-phosphoglycerate

2-phosphoglycerate

phosphoglyceromutase

phosphoenolpyruvate (PEP)

enolaseH2O

ADP

ATPpyruvate kinase

pyruvic acid

NADH + H

NAD

pyruvate decarboxylaseCO2

acetaldehyde

ethanol

alcohol dehydrogenase

Figure 8–8. The Embden-Meyerhoff-Parnas glycolytic pathway used by Saccharomyces cerevisiae



and re-oxidize NAD via the lactate dehydroge-nase reaction. In contrast, in S. cerevisiae, thisis done via a two-step series of reactions, inwhich pyruvate is first decarboxylated bypyruvate decarboxylase, generating acetalde-hyde and CO2.

Next, the acetaldehyde is reduced by one ofseveral NADH-dependent alcohol dehydroge-nases to form ethanol and oxidized NAD.Thenet effect, then, of glycolytic metabolism bybakers’ yeast is the formation of two moles ofCO2 and two moles of ethanol per mole of glu-cose fermented (Figure 8–6).The oxidized andreduced forms of NAD are in balance, and thecell obtains a net gain of two moles of ATP.Yeast growth itself is quite modest during gly-colysis, because most of the glucose carbon isused for generating energy rather than for pro-ducing cell mass.Therefore, the initial popula-tion will increase only a few generations dur-ing the entire fermentation process.

End products

The most obvious and most important endproduct of the bread fermentation is CO2. It is,after all, the CO2 that is mainly responsible fortransforming the stiff, heavy, dough into aspongy, elastic, and airy material. Some of theCO2 dissolves in the water phase (eventhough CO2 solubility is generally low), caus-ing a slight decrease in pH. Eventually, the CO2

saturates the aqueous environment surround-ing the yeast microcolonies.At that point, theCO2 evolves into the dough. The CO2 causesthe gluten proteins to stretch, and some es-cape, but most of the gas is retained and istrapped within the matrix. This process isknown as leavening.

The retained CO2 ordinarily collects in thedough to form large irregular gas cells that are dispersed in a non-uniform heterogenousmanner. This is one reason why doughs are“punched down” during the course of the fer-mentation.This step re-disperses the gas, form-ing smaller, more regular, and evenly dispersedgas cells.About 45% of the gas is lost.However,re-mixing causes the yeast cells and sugars to

be re-distributed in the dough such that moresubstrate is made available to the yeast and ad-ditional CO2 can be formed.

Other metabolic products are also pro-duced during the yeast fermentation, includingvarious acids, that give the dough a slightlyacidic, but pleasant flavor and aroma.The yeastitself can produce acids, as well as other or-ganic compounds. Lactic acid bacteria, whichare inevitably present, either in the yeast cakeor in the flour, also begin to grow, ferment sug-ars, and produce acids. Ultimately, the pH ofthe dough will drop from about 6.0 to 5.0.Thisdecrease in pH, however, has little effect onyeast growth and metabolism; in fact, pH 5 isvery near the optimum for S. cerevisiae. Onlywhen there is enough acid produced in thedough to lower the pH to 4.0 or less (as is thecase for sourdough breads, discussed below)will inhibition of bakers’ yeast strains occur.Despite their modest accumulation in dough,the acids produced by yeasts and lactic acidbacteria make important contributions to theflavor, as well as rheology of the dough. For ex-ample, low pH improves the water-binding ca-pacity and swelling of gluten, making it moreelastic and pliable.

Factors affecting growth

Temperature and relative humidity have an im-portant effect on yeast growth and fermenta-tive capacity. Most strains of S. cerevisiae usedfor baking have an optimum temperature ofabout 36°C to 39°C, although seldom are suchhigh temperatures used during the dough fer-mentation. Rather, doughs are ordinarily held attemperatures of about 25°C to 28°C (andsponges slightly lower). Although higher tem-peratures can accelerate fermentation andgassing rates (defined as the amount of CO2

produced per unit time),elevated temperaturesalso can enhance growth of microbial contami-nants, including wild yeasts and mold.Also, thetemperature at which yeast activity declines isonly a few degrees higher than its optima. Rela-tive humidity must also be controlled,usually atabout 70% to 80%. Lower levels of moisture in



the air may cause the dough to dry at the sur-face, leading to formation of a crust-like mater-ial,as well as inhibiting the fermentation withinthe areas near the surface.

Dividing, Rounding, and Panning

The fermentation described above for astraight dough or bulk fermentation (see be-low) usually occurs in troughs. Once the fer-mentation is considered complete, the doughis then ready to be divided, on a volumetric ba-sis, into loaf-sized portions.Dividing is usually asimple process that involves cutting extrudeddough at set time intervals, such that eachpiece has very near the same weight. Other devices, driven by suction, are also common.Despite the differences in design, the main re-quirement of the dividing step is that it bedone quickly, since the fermentation is still on-going and the weight-to-volume ratio (i.e., den-sity) of the dough pieces may change.

The divided dough is then conveyed to arounding station, where ball-shaped pieces areformed.At this point the dough is given a short(less than twenty minutes) opportunity to re-cover from the physical strains and stressescaused by being cut,compressed,and bouncedabout. A portion of the gas is also lost duringthe dividing and rounding steps.Thus, not onlydoes the dough have a chance to rest (struc-turally, that is), but the fermentation also con-tinues, adding a bit more gas into the dough.Following this so-called intermediate proofingstep, the dough pieces are delivered into amolding system that first sheets the dough be-tween rollers, then rolls the dough into a cylin-der shape via a curling device, and finally,shapes the dough into the desired final form.The shaped loafs are then transferred to pansfor proofing and baking.

Proofing

Most of the CO2 is expelled during the sheet-ing step, and it is during the final proofing stepwhere the dough is re-gassed and the fermen-

tation is completed. For some bread produc-tion systems (e.g., no-time processes describedbelow), the entire fermentation takes placeduring the proofing step. Proofing is usuallydone in cabinets or rooms between 35°C and42°C,giving dough temperatures near the opti-mum for S. cerevisiae (35°C to 38°C). Proofingrooms are also maintained at high relative hu-midity (�85%). Total proofing times vary, de-pending on temperature, but are usually lessthan one hour.The increase in dough volume,measured as the height of the loaf, is oftenused to determine when the dough is suffi-ciently proofed.

Baking

A remarkable confluence of physical, chemi-cal, and biological events occurs when thedough loaves are placed into a hot oven. Intothe oven goes a glutenous, sticky, spongymass, with a pronounced yeasty aroma andinedible character, and out comes an airy,open-textured material, with a unique aromaand complex flavor. It is a transformation likenone other in food science.

Most modern commercial bakers use con-tinuous conveyer-type ovens, rather thanbatch-type, constant temperature ovens. Mostof these tunnel-like ovens are configured assingle or double-lap types, in which the doughloaves are loaded and unloaded (after baking)at the same end. Alternatively, in other ovens,the dough is fed at one end and the bakedbreads are discharged from the other end. Itusually takes about twenty-five to twenty-eightminutes for a loaf to traverse through the oven.

The temperature with these ovens is notconstant, but rather increases in several stagesalong the route, starting at about 200°C for sixto eight minutes, and then increasing to about240°C for the next twelve to fourteen minutes.Finally, the temperature is reduced slightly toabout 220°C to 235°C for the remaining four toeight minutes. It should be evident that thetemperature of the dough itself will always beless than the oven temperature.The actual tem-



perature of the loaf during baking depends onthose general factors affecting heat transfer ki-netics, including dough composition, moisturecontent, loaf size and shape, and the type orform of heating applied. Although ovens maytransfer heat by convection currents (i.e., viathe air and water vapor), heat transfer withinthe dough is via conduction.Thus, there is a de-cided temperature gradient within the loaf,with the interior being much cooler than thesurface.Moreover,many of the readily observedphysical changes that occur in the bread duringheating are localized, the formation of crustcolor being the most obvious example.

Several events rapidly occur when the doughis placed into the first stage or zone of the oven.First, as the interior dough temperature in-creases (about 5°C per minute), there is an im-mediate increase in loaf volume (“oven spring”)due to expansion of the heated CO2.When thedough temperature reaches 60°C, all of the dis-solved CO2 is evolved, causing further expan-sion of the dough.Ethanol also is volatilized andlost to the atmosphere. At very near the sametime, there is a transitory increase in yeast andamylase activity, at least up to dough tempera-tures of 55°C to 60°C, at which point the yeastsare killed. Starch swelling and gelatinization in-creases, and the gluten begins to dehydrate anddenature, causing the bread to become morerigid and more structured.All of these events oc-cur within about six to seven minutes.

In the meantime, the surface temperaturewill have reached close to 100°C, and the earlystages of crust formation will have begun. Dur-ing the next heating stage,evaporation of water,gelatinization of starch, and denaturation andcoagulation of gluten continue. Remaining en-zymes and microorganisms are inactivated asthe internal temperature approaches 80°C.Finally, when the temperature reaches 95°C,the dough takes on a crumb-like texture and thecrust become firmer, due to dehydration at thesurface.A brown crust color is formed as a resultof caramelization and Maillard non-enzymaticbrowning reactions.The latter reaction also gen-erates volatile flavor and aroma compounds.

Cooling and Packaging

Once out of the oven, the baked bread is sus-ceptible to microbial spoilage. Therefore,cooling must be performed under conditionsin which exposure to airborne microorgan-isms, particularly mold spores, is minimized.The bread also must be cool enough so thatcondensate will not form inside the package,a situation that could also lead to microbialproblems. Various cooling systems are used,including tunnel-type conveyers in whichslightly cool air passes counter-current to thedirection of the bread, as well as forced air,rack-type coolers. For sandwich breads, thecooled loaves are sliced by continuous slicingmachines. Ultimately, commercial breads arepackaged in moisture impermeable polyethyl-ene bags.

Modern Bread Technology

Although hundreds of variations exist, mostleavened breads are manufactured by one offour general processes (Figure 8–9). Theseprocesses vary in several respects, includinghow the ingredients are mixed, where and forhow long the fermentation occurs,and the over-all time involved in the manufacturing process.Of course,how the bread is made also has a pro-found effect on the quality characteristics of thefinished product.

Straight dough process

The homemade, one-batch-at-a-time methoddescribed above is generally referred to as thestraight method or straight dough process. Ba-sically, the overall procedure (outlined in Fig-ure 8–9) involves mixing all of the ingredientsand then allowing the dough to ferment forseveral hours (with intermittent punchingdown).The developed dough is then divided,formed into round balls, given a brief (inter-mediate) proof, shaped into loaves, and placedin baking pans. Finally, after a final fermenta-tion or proof, the bread is baked.The straight



dough method is little used by the baking in-dustry, the exceptions being mostly small orspecialty bakeries. The process results in abread that is chewy, with a coarse cell struc-ture and a moderate flavor.Although the qual-ity of these breads (Box 2) is quite good, thelimitation of this process is that it lacks flexi-bility and is sensitive to time. In other words,when the fermentation is sufficiently com-plete, the dough must soon be baked. Other-wise, if fermentation is prolonged, the breadwill be yeasty, excess air cells will be formed,and the structure will be weak.

Sponge and dough process

The most commonly used method in thebread industry is the sponge and doughprocess. It is used by small, medium, and mostlarge bakeries in the United States.The basic

principle of this process relies on the use of a“sponge,”a partially concentrated portion of aflour-water dough that is allowed to fermentand then is mixed with the remaining doughingredients. The main advantage of thismethod is it that it is tolerant to time. In otherwords, once the sponge is developed, it doesnot have to be used immediately, but rathercan be used over a period of time.Also, breadmade by this method has a fine cell structureand well developed flavor.As outlined in Fig-ure 8–9, the process involves making asponge that consists of part of the flour (60%to 70%), part of the water (40%), and all of theyeast and yeast nutrients.The sponge is mixedand allowed to ferment for three to six hoursat 16°C to 18°C (80°F).The remaining ingredi-ents are then added and mixed, and thedough is allowed to develop. After a brieffloor time period, the dough is then divided,

Figure 8–9. Major bread manufacturing processes. From Lillemand Bakery Update, Volume 1/Number 10(with permission).



molded, panned, given a final proof, andbaked.

Liquid sponge process

The traditional methods of bread making, asdescribed above for the straight dough andsponge and dough methods, require sufficienttime for the initial or bulk fermentation to oc-cur and for the dough to develop. In the pastseveral decades, an emphasis on speed andeconomy of scale has led to newer methods ofbread manufacture.These methods rely less onbulk fermentation and natural dough develop-ment, and more on mechanical dough devel-opment and a relatively short fermentation period.The straight dough method, for exam-ple, can be shortened considerably by addingmore yeast and giving the dough just a fewminutes of floor time, or perhaps even elimi-nating the bulk fermentation step altogether.The sponge and dough method can be simi-larly modified by reducing the flour-to-waterratio in the sponge (i.e., adding more water),thereby shortening the pre-fermentation timeand producing a pumpable sponge.

One of the best examples of the quick typeof process is the continuous bread-making pro-cedure, sometimes called the liquid spongeprocess (Figure 8–9). Developed in the 1960s,it was adopted by large-scale bakeries in theUnited States to enhance throughput and tomass produce loaves in as short a time as pos-sible. The method saved time and labor andwas very economical.As much as 35% to 50%of the industry used this method, but its popu-larity eventually waned,and now more bread isprobably made by sponge-and-dough methods.In principle, the process is simply an extensionof the sponge-and-dough rationale.

Instead of making a thick sponge (i.e.,mostly flour), a thin, liquid sponge (mostly wa-ter) is made.This mixture of water, yeast, and avery small portion of the flour is allowed toferment in a tank to form a liquid “pre-ferment.” This pre-fermented material is thenmetered into high speed mixers, where it israpidly mixed with the remaining dough ingre-

dients. The dough is then divided, panned,given a final proof, and baked.

Although the process is fast—three hoursversus up to six hours for sponge and doughmethods—and produces bread with a uniformstructure and texture, many experts have ar-gued that bread quality is not nearly as good asthat made by other methods.By growing mainlyin the liquid material and not in the actualdough, the yeasts have less time to generate fla-vor compounds or to react with flour compo-nents.Moreover,yeast has an oxidizing effect ongluten proteins during fermentation in thestraight dough or sponge and dough methods,leading to increased elasticity and strongerdough structure.This is lost in the continuousmix system. Thus, these breads are typicallysofter and spongy, less elastic, and have a weak,cake-like texture.The flavor is rather bland andnot well-developed.One noted bread researchersuggested that these breads are simply wrap-pers for meats, cheeses, and jams.Today, contin-uous mix systems are used almost exclusivelyfor hot dog and hamburger buns.

Chorleywood process

In the United Kingdom, the Chorleywoodprocess is also a rapid,high throughput methodand is probably the main commercial system forbread manufacture (Figure 8–9). It is little used,however, in the United States.This is a no-timeprocess in which rapid mixing is a critical fea-ture. Whatever fermentation there is, prior tothe final proof, occurs during the brief floortime (at temperatures near 40°C).Thus,the start-to-finish process takes only two hours.

Sourdough Fermentation

Most yeast breads have a slightly acidic pH,due, in part, to acids produced by the yeastand, to a lesser extent, to dissolved CO2 in thedough. In addition, acids also are produced bylactic acid bacteria that ordinarily are presentin the flour, in the yeast preparations, or thatsimply naturally contaminate the dough. How-ever, the presence of endogenous lactic acid



bacteria has, under most circumstances, only amodest effect on dough and bread acidity.Theperception of sourness for most consumersdoes not occur unless the pH is near 4.0,which is far lower than the normal pH range of5.0 to 6.0 for most yeast breads.However, if lac-tic acid bacteria are deliberately added and al-lowed to grow, enough acids can be producedto lower the pH to below 4.0 and to cause a

distinct sour but appealing flavor. Not only dothese breads have a unique flavor, but they arealso better preserved due to their high acidityand low pH. In fact, there are several functionaladvantages to the sourdough fermentation,particularly for breads made from rye andother flours (Box 8–4).

Despite the fact that sourdough breads haveonly recently been studied, they actually have

Box 8–4. Beyond Sourness: Functional Advantages of the Sourdough Fermentation

Although a desirable sour flavor is likely the most noticeable property of sourdough breads, theuse of sourdough cultures in bread manufacture offers other important advantages (Thiele etal.,2004; Table 1). Arguably, the most important benefit of the sourdough fermentation is due tothe organic acids that are produced by the sourdough bacteria and the ensuing low pH that re-sults in enhanced preservation and increased shelf-life of sour dough breads.However, the sour-dough fermentation also has a major impact on dough functionality, in particular, when rye orother whole grain flours are used as an ingredient.

Table 1. Functional advantages of sourdough cultures.

Property References

Preservation Messens and De Vuyst, 2002.Production of anti-fungal agents Gänzle and Vogel, 2002.Flavor Thiele et al., 2002.Reduced staling rate Armero and Collar, 1998.Increased loaf volume Collar Esteve et al., 1994.Lower glycemic index Liljeberg and Björck, 1996.Increased gluten tolerance Di Cagno et al., 2004.Increased mineral availability Lopez et al., 2003.Improved texture Korakli et al., 2001.

Next to wheat, rye is the second most common cereal grain used to make bread. Rye, how-ever, has properties that pose particular challenges when used in bread making. First, rye con-tains a high concentration of pentosans, a heterogenous mixture of pentose-containing poly-saccharides consisting mostly of xylose and arabinose.They constitute as much as 10% of ryeflour, which is four to five times more than that found in wheat.

In bread making, the pentosans may have both positive and negative effects. For example,pentosans have high water-binding capacity and may, therefore,decrease retrogradation and de-lay staling. On the other hand, pentosans may interfere with gluten formation, giving an inelas-tic dough that retains gas poorly.Perhaps the major problem with rye is that the rye proteins donot form a viscoelastic dough.As a result, breads made with rye as the main grain typically havea small loaf volume and a dense crumb texture. In addition, rye flour contains more �-amylasethan is present in wheat, and this amylase is particularly active at the temperature at whichstarch gelatinizes.This results in excessive starch hydrolysis in the dough and bread, giving apoor texture and further reducing loaf volume.

The addition of sourdough cultures to rye doughs can compensate, to a large extent, for thesecomplications. First, as the pH decreases due to the lactic fermentation, the pentosans becomemore soluble.They also begin to swell and form a gluten-like network that enhances dough elas-



ticity and gas retention. In other words, at low pH, the pentosans assume the role normally per-formed by gluten. In addition, the sourdough organisms are stimulated in rye flours by the avail-ability of free sugars liberated from starch via the �-amylase. Moreover, this enzyme begins tolose activity at the low pHs achieved during the sourdough fermentation, so excessive hydroly-sis is avoided. Some sourdough bacteria also have the ability to ferment pentoses released frompentosans, producing heterofermentative end products, including acetic acid (Gobbetti et al.,2000). Finally, the subsequent acidic conditions enhance the water-binding capacity of thestarch granules, which further decreases the rate at which staling occurs.

Another potential problem in breads made from rye and other whole grain flours is due tothe presence of phytic acid. Phytic acid (also called myoinositol hexaphosphate) is a highlyphosphorylated, negatively charged molecule that is capable of binding zinc, iron, calcium, andother divalent metal cations, preventing their absorption. Thus, phytic acid, by reducing thebioavailability of essential minerals,has “anti-nutritional”properties.Cereal grains can contain asmuch as 4% phytic acid, and although levels in whole grain bread are usually less than 0.2%, thisis still enough to be a nutritional concern.

Degradation of phytic acid in bread ordinarily occurs via the enzyme phytase, which is pre-sent in flour and is also produced by yeasts.Recently, it has been shown that Lactobacillus san-franciscensis, Lactobacillus plantarum, and other sourdough bacteria also produce this en-zyme (De Angelis et al.,2003;Lopez et al.,2000).Moreover,phytase activity,and in particular, thephytase produced by L. sanfranciscensis, has an optima at low pH (pH 4 in the case of L. san-franciscensis).Thus, the sourdough fermentation also enhances the nutritional quality of ryeand other whole grain breads.

References

Armero, E., and C. Collar. 1998. Crumb firming kinetics of wheat bread with anti-staling additives. J. CerealSci. 28:165–174.

Collar Esteve,C.,C.Benetido de Barber,and M.Martinez-Anaya.1994.Microbial sour doughs influence acid-ification properties and breadmaking potential of wheat dough. J. Food Sci. 59:629–633, 674.

De Angelis,M.,G.Gallo,M.R.Corbo,P.L.H.McSweeney,M.Faccia,M.Giovine,and M.Gobbetti.2003.Phytaseactivity in sourdough lactic acid bacteria: purification and characterization of a phytase from Lacto-bacillus sanfranciscensis CB1. Int. J. Food Microbiol. 87:259–270.

Di Cagno, R., M. De Angelis, S.Auricchio, L. Greco, C. Clarke, M. De Vincenzi, C. Giovannini, M. D’Archivio, F.Landolfo, G. Parrilli, F. Minervini, E.Arendt, and M. Gobbetti. 2004. Sourdough bread made from wheatand nontoxic flours and started with selected lactobacilli is tolerated in celiac sprue patients.Appl. En-viron. Microbiol. 70:1088–1096.

Gänzle,M.G., and R.F.Vogel. 2003.Contribution of reutericylin production to the stable persistence of Lac-tobacillus reuteri in an industrial sourdough fermentation. Int. J. Food Microbiol. 80:31–45.

Gobbetti, M., P. Lavermicocca, F. Minervini, M. De Angelis, and A. Corsetti. 2000.Arabinose fermentation byLactobacillus plantarum in sourdough with added pentosans and �-L-arabinofuranosidase:a tool to in-crease the production of acetic acid. J.Appl. Microbiol. 88:317–324.

Korakli,M.,A.Rossmann,M.G.Gänzle,and R.F.Vogel.2001.Sucrose metabolism and exopolysaccharide pro-duction in wheat and rye sourdoughs by Lactobacillus sanfranciscensis. J.Agric.Food Chem.49:5194–5200.

Liljeberg, H.G.M., and I.M.E. Björck. 1996. Delayed gastric emptying rate as a potential mechanism for low-ered glycemia after eating sourdough bread: studies in humans and rats using test products with addedorganic acids or an organic salt.Am. J. Clin. Nutr. 64:886–893.

Lopez, H.W.,V. Duclos, C. Coudray,V. Krespine, C. Feillet-Coudray,A. Messager, C. Demigne, and C. Remesy.2003. Making bread with sourdough improves mineral availability from reconstituted whole wheatflour in rats. Nutrition 19:524–530.

Box 8–4. Beyond Sourness: Functional Advantages of Sourdough Fermentation (Continued)

(Continued)



been made for many years. In fact, it seemslikely that the very first breads made thousandsof years ago were sourdough breads. Until theadvent and availability of bakers’ yeast culturesjust a century ago, it also is probable that sour-dough breads were the main type of bread con-sumed throughout Europe and North America.This is because doughs were naturally fer-mented by wild cultures, which inevitably con-tained lactic acid bacteria, including heterofer-mentative strains that could produce sufficientCO2 to cause the dough to rise. In addition,most wild sourdough cultures also containedwild yeasts that would further contribute to theleavening process (see below). These mixedbacteria-yeast wild type cultures could easily bemaintained and used for many years, if notdecades.

Only in the past thirty years, however, havethe lactic acid bacteria and yeasts that partici-pate in the sour dough fermentation beenidentified. In the early 1970s, researchers at theUSDA Western Regional Research Laboratoryin Albany, California (located about ten milesfrom San Francisco) isolated a strain of Lacto-bacillus from a commercially-manufactured,locally-produced sourdough sponge. The iso-late appeared to be a unique species, based onits physiological properties and DNA-DNA hy-bridization, leading the discoverers to proposethe name, Lactobacillus sanfrancisco (nowcalled Lactobacilllus sanfranciscensis). Thisorganism, which has become known as thesourdough bacterium,was also present in sour-dough sponges from other area bakeries. Inter-estingly, this strain appears to be identical toLactobacillus brevis var. lindneri, a Lacto-

bacillus culture used in European sourdoughbreads.

In addition to having isolated L. sanfrancis-censis from sourdoughs, the same USDA re-searchers discovered that the sourdoughsponge also contained a unique yeast strain.This strain was identified as Torulopsisholmii (the imperfect form of a Saccha-romyces species that was once classified asSaccharomyces exiguus), and was later re-classified as Candida milleri. This yeast wasfound in other L. sanfranciscensis-containingsourdough sponges from the local area, andalways in a yeast:bacteria ratio of about 1:100.The same yeast-bacteria duo has subsequentlybeen found in sourdoughs from around theworld.That these organisms are found consis-tently in sourdough breads from such distantgeographical locations reflects their remark-able adaptation to the dough environment(Box 8–5). This also accounts for the durabil-ity and stability of sourdough sponges, someof which have reportedly been carried formore than 100 years.

Many traditional sourdough bread manufac-turers continue to maintain their own particu-lar sponge, which contains a mixture of unde-fined, yet distinctive strains. In fact, someartisanal bakeries consider their sourdoughstarters to be uniquely responsible for the qual-ity of their bread and guard their cultures asvalued assets. Analyses of these cultures hasshown that they contain many differentspecies of Lactobacillus and other lactic acidbacteria, as well as several species of yeast(Table 8–4).This diversity of sourdough starterstrains likely accounts for the variety of sour-

Lopez, H.W.,A. Ouvry, E. Bervas, C. Guy,A. Messager, C. Demigne, and C. Remsey. 2000. Strains of lactic acidbacteria isolated from sour doughs degrade phytic acid and improve calcium and magnesium solubilityfrom whole wheat flour. J.Agric. Food Chem. 48:2281–2285.

Messens W., and L.De Vuyst.2002. Inhibitory substances produced by Lactobacilli from sour doughs—a re-view. Int. J. Food Microbiol. 72:31–43.

Thiele, C., M. G. Gänzle, and R.F.Vogel. 2002. Contribution of sourdough lactobacilli, yeast, and cereal en-zymes to the generation of amino acids in dough relevant for bread flavor. Cereal Chem. 79:45–51.

Thiele, C., S. Grassl, and M. Gänzle. 2004. Gluten hydrolysis and depolymerization during sourdough fer-mentation. J.Agric. Food Chem. 52:1307–1314.

Box 8–4. Beyond Sourness: Functional Advantages of Sourdough Fermentation (Continued)



Box 8–5. Happy Together: Sourdough Bacteria and Yeast

Almost all wild sourdough cultures, especially those that have been maintained for a long time,contain both lactic acid bacteria and yeasts.Although many species of lactic acid bacteria andyeast have been isolated from sourdoughs, the most common appear to be Lactobacillus san-franciscensis (formerly Lactobacillus sanfrancisco) and Candida milleri (Kline and Sugihara,1971 and Sugihara et al., 1971). It is remarkable that the interaction between these organisms isnot only stable, but that both organisms derive benefit from this unique ecological association.Moreover, despite their saccharolytic and fermentative metabolism, these organisms appear toobserve a non-compete clause, and instead share the available sugars present in the dough.

Several physiological and biochemical properties are now known to be responsible for thestable, symbiotic relationship between L. sanfranciscensis and C. milleri and their ability togrow together in bread dough.First,L. sanfranciscensis is heterofermentative,producing aceticand lactic acids,ethanol, and CO2.The pH of the dough can drop to as low as 3.5, a pH that is in-hibitory to competing organisms, including bakers’ yeast strains of Saccharomyces cerevisiae.In contrast, C. milleri (and related species) is acid-tolerant and benefits by the lack of competi-tion from other acid-sensitive yeasts.The yeast, in turn, releases free amino acids and other nu-trients needed by the Lactobacillus.

There is also a fascinating metabolic component to the L. sanfranciscensis-C. milleri symbi-otic partnership. Most of the fermentable carbohydrate found in sourdough is in the form ofmaltose.Whereas S. cerevisiae, the ordinary bakers’ yeast, readily ferments maltose, C. milleri isunable to ferment this sugar, leaving this organism without an obvious substrate for growth. Incontrast, L. sanfranciscensis not only prefers maltose as a carbohydrate source, but fermentsthis sugar in a particularly relevant manner.

Ordinarily, one might expect that metabolism of the disaccharide maltose would result inboth glucose moieties being fermented.However,when L.sanfranciscensis grows on maltose ituses the cytoplasmic enzyme maltose phosphorylase to hydrolyze accumulated maltose, yield-ing free glucose and glucose-1-phosphate. The latter is isomerized to glucose-6-phosphate,which then feeds directly into the heterofermentative phosphoketolase pathway (Chapter 2).The glucose generated by the reaction is not further metabolized,but rather is released into theextracellular medium, or in this case, the dough. Glucose excretion occurs, in part, becausemaltose is transported and hydrolyzed so rapidly that more glucose is formed than the cell canhandle (Neubauer et al., 1994).

Weak expression of hexokinase also contributes to slow glucose use.The net effect of thisunique metabolic situation is that maltose is available for the sourdough bacteria, glucose isavailable for the sourdough yeast, and neither organism competes against each other for fer-mentation substrates.

References

Gobbetti, M., 1998.The sourdough microflora: interactions of lactic acid bacteria and yeasts.Trends FoodSci.Technol. 9:267–274.

Kline, L., and T.F. Sugihara. 1971. Microorganisms of the San Francisco sour dough bread process. II. Isola-tion and characterization of undescribed bacterial species responsible for the souring activity. Appl.Microbiol. 21:459–465.

Neubauer,H.,E.Glaasker,W.P.Hammes,B.Poolman,and W.N.Konings.1994.Mechanisms of maltose uptakeand glucose excretion in Lactobacillus sanfrancisco. J. Bacteriol. 176:3007–3012.

Sugihara, T.F., L. Kline, and M.W. Miller. 1971. Microorganisms of the San Francisco sour dough breadprocess. I.Yeasts responsible for the leavening action.Appl. Microbiol. 21:456–458.

dough breads produced by different bakeries.However, sourdough cultures, comprised ofone or more defined strains, are now commer-

cially available. In addition to L. sanfranciscen-sis, available strains include L. brevis, L. del-brueckii, and L. plantarum. Some of these



for a sourdough fermentation. Although suchbreads are quite common in the retail market,they certainly lack the complex flavor andfunctional characteristics associated with realsourdough bread. Alternatively, some bakersadd bakers’ yeast to the dough to accelerateproduction of CO2. While this practice mayshorten the time necessary for dough develop-ment and leavening, it likely results in breadwith a weak sourdough flavor.

Bread Spoilage and Preservation

Despite the fact that bread has a moderatelylow water activity and pH and contains few mi-croorganisms when it leaves the oven, it is still,relative to other fermented foods, a highly per-ishable product.This is because the shelf-life ofbread depends not only on microbial activities,but also on physical-chemical changes in thebread. Specifically, it is the phenomenon calledstaling that most frequently causes consumersto reject bread products. For some fresh bakedproducts stored under ambient conditions,shelf-life can be as short as just two or threedays, due to the onset of staling. However, thelonger the bread is held, the more likely it isthat microorganisms, and fungi, in particular,will grow and produce visible and highly ob-jectionable appearance defects.Thus, maintain-ing bread in a fresh condition remains a majorchallenge for the bread industry.

Staling

It would be an understatement to say that stal-ing is a complicated phenomenon. Entirebooks and extensive review articles have beenwritten on this topic alone. Recently, however,thanks to the application of modern molecularmethods, including infrared, near infrared, andnuclear magnetic resonance spectroscopy, x-ray crystallography, and various microscopictechniques, a clearer understanding of the de-tails involved in staling has emerged. Simplystated, staling refers to the increase in crumbfirmness that makes the bread undesirable toconsumers. In addition, staling is associated

Table 8.4. Microorganisms isolated from sourdoughs1.

Bacteria Yeast

Lactobacillus Saccharomyces alimentarius exiguus2

Lactobacillus Saccharomyces brevis cerevisiae

Lactobacillus casei Candida milleriLactobacillus curvatus Candida humilisLactobacillus

delbrueckii Issatchenkia orientalisLactobacillus fermentumLactobacillus hilgardiiLactobacillus pentosusLactobacillus plantarumLactobacillus pontisLactobacillus panisLactobacillus reuteriLactobacillus

sanfranciscensisPediococcus acidilactici

1Data from Okada et al., 1992; Pulvirenti et al., 2004; and Vogel

et al., 19942This species is now classified as Saccharomyces cerevisiae

strains are obligate homofermentors (e.g., L.delbrueckii), and although the sourdoughbreads made from these cultures are indeedsour, they often lack the characteristic aceticacid flavor and have a somewhat less complexflavor profile.

In actual practice, authentic sourdoughbreads are almost always made via a sponge anddough process.The initial sponge is maintainedby regular (daily or even more frequent) addi-tions of flour-water mixtures. Portions are re-moved and are successively built up by addingwater-flour mixtures to the sponge followed byan incubation period, until the appropriate sizeis achieved. Incubations can be long (twenty-four hours) or short (three hours), dependingon the temperature, culture activity, and thepresence of salt.The sponge can be held for aperiod of time,and is eventually mixed with theremaining dough ingredients at a rate of about10% to 20% of the total dough mixture.

It should be noted that a type of sourdoughbread can be made simply by adding vinegar ora combination of acetic and lactic acids di-rectly to the dough. This eliminates the need



with an increase in crust softness and a de-crease in fresh bread flavor.

Staling is basically a starch structure andmoisture migration problem. The reactionsthat eventually lead to staling actually startwhen the bread is baked, as starch granules inthe dough begin to adsorb water, gelatinize,and swell. The amylose and amylopectinchains separate from one another and becomemore soluble and less ordered.Then, when thebread is cooled, these starch molecules, andthe amylopectin, in particular, slowly begin tore-associate and re-crystallize (see below).Thisprocess, called retrogradation, results in an in-crease in firmness due to the rigid structuresthat form. Amylose retrogrades rapidly uponcooling, while amylopectin retogrades slowly.It is the slow retrogradation of amylopectinthat is now thought to be primarily associatedwith staling of the crumb. Furthermore, mois-ture migration from starch to gluten and fromcrumb to crust make the crumb dryer andmore firm. Although staling is an inevitableprocess, a number of strategies have beenadopted to delay these reactions and extendthe shelf-life of bread (Box 8–6).

Biological spoilage

Microbiological spoilage of bread is most oftenassociated with fungi, and occurs when fungalmycelia are visible to the consumer. Somestrains of Bacillus subtilis, Bacillus mesenteri-cus, and Bacillus licheniformis can spoil high-moisture breads via production of an extracel-lular capsule material that gives the infectedbread a mucoid or ropy texture.There are alsowild yeasts capable of causing flavor defects inbread after baking;however,bacterial and yeastspoilage of bread is relatively rare, and it isfungi that are, by far, the most common micro-bial cause of bread spoilage (Table 8–5). Inlarge part, this is because the water activity ofbread is usually less than 0.96, which is belowthe minimum for most spoilage bacteria. In ad-dition,baking ordinarily kills potential spoilagebacteria, the exception being spore-formingbacilli mentioned above.

The baking process kills fungi and theirspores.Thus, when molds are present in bread,it is invariably a result of post-processing con-tamination. Fungal spores are particularlywidespread in bakeries due to their presencein flour and their ability to spread throughoutthe production environment via air movement.When the baked breads leave the oven, theirtransit through the cooling,slicing,and packag-ing operations leave plenty of opportunity forinfection, either indirectly by airborne sporesor directly by contact with contaminatedequipment. Fungi are aerobic and ordinarilygrow only on the surface of loaf bread. How-ever, slicing exposes the internal surfaces tomold spores, enabling growth within the loaf.Once packaged, moisture loss in the bread isnegligible; thus, the water activity remainswell within the range necessary for growth offungi. Bread that is packaged while still warmis very susceptible to mold growth, due to lo-calized areas of condensate that form withinthe package.

Although there are many different speciesof fungi associated with bread spoilage (Table8–5), the most common are species of Penicil-lium, Aspergillus, Mucor, and Rhizopus. Breadserves as an excellent growth substrate forthese fungi; visible mold growth may appearwithin just a few days.What one actually sees isa combination of vegetative cell growth (themycelia),along with sporulating bodies (Figure8–10).The latter are responsible for the charac-teristic blue-green or black color normally as-sociated with mold growth.The ability of fungito grow on bread and which species predomi-nate depend on several factors, includingbread pH and water activity, and storage tem-perature and atmosphere. Finally, it is impor-tant to recognize that some mold strains notonly can grow on bread, causing spoilage andeconomic loss, but, under certain conditions,specific strains can also produce mycotoxins.Fortunately, visible mold growth ordinarily pre-cedes mycotoxin formation, so in the very un-likely event that a mycotoxin were present,most consumers would reject the product be-fore ingesting it.



Box 8–6. Fresh Ideas for Controlling Staling of Bread

The precise mechanisms responsible for bread staling are complex.According to current mod-els (Figure 1), staling occurs when unstructured and gelatinized starch (mainly amylopectin)that is formed during baking reorganizes or retrogrades, first after cooling and continuing dur-ing aging, into rigid, crystalline structures (Gray and Bemiller, 2003).As starch retrogrades, in-tramolecular hydrogen bonding between amylopectin branches occurs, causing additionalstructural rigidity. In addition, the outer chains of adjacent amylopectin molecules may formdouble helices. Efforts to reduce staling, therefore, necessarily involve reducing the rate atwhich these crystallization and hydrogen-binding reactions occur.

Because staling is influenced by many factors (Table 1),no single approach aimed at delayingor preventing staling is likely to be completely effective. Rather, bread manufacturers haveadopted a more multifaceted strategy that considers the entire bread-making process, from in-gredient selection and product formulation to processing and packaging steps (Zobel and Kolp,1996). In the discussion below, the role of ingredients, enzymes, and processing conditions oncontrolling staling are described.

Table 1. Factors affecting staling.

Factor Effect

IngredientsFlour High protein flours maintain crumb softnessFats Delay staling by increasing loaf volumeSurfactants Increase crumb softness

Enzymes Amylases (mainly �) hydrolyze starch and reduce retrogradationMoisture Low moisture increases stalingPackaging Prevents moisture loss and crumb firmingManufacturing Long fermentation times increase loaf volume and softnessBaking temperature High temperature, short time baking increases staling ratesStorage temperature Refrigeration increases staling, freezing decreases staling rates

Ingredients

Several of the ingredients that are added to bread dough either retard staling rates and/or softenbread crumb and crust. Included are surfactants and other emulsifying agents, dextrins, and se-lected oligosaccharides. Mono- and diglycerides are probably the most common surfactant-typeagents used by the bread industry and the most effective as anti-staling agents.They appear tofunction by binding to amylose and amylopectin and forming complexes with lipids, thereby re-ducing starch retrogradation rates (Knightly, 1996).

Enzymes

Enzymes are widely used by the bread industry.Amylases, in particular,are often used to hydrolyzestarch and to increase the concentration of fermentable sugars in the dough.These sugars can alsocontribute to desirable flavor and color changes via the Maillard reaction. However, specific �-amylases are now also used for their anti-staling properties.

How amylases actually delay staling is not clearly known, although several explanations havebeen proposed (Bowles, 1996). It could be that the simple sugars released by �-amylases couldact directly by interfering with crystal formation; however, this seems unlikely since sugar addi-tion to breads has no effect on retrogradation. Instead, it appears that the anti-staling function of�-amylases is due to the hydrolysis of amylopectin, especially near branch-points.This would re-duce the number of cross-links that contribute to crumb and crust firmness.



The commercially-available �-amylases used as anti-staling agents are derived from either bac-terial or fungal sources. Most are produced by species of Bacillus and Aspergillus, and consistof either pure enzymes or mixtures of enzymes.These enzymes have optimum activity within amoderately high temperature range (50°C to 70°C), and are most active during the early stagesof baking.However,bacterial �-amylases are generally thermal-resistant and are inactivated onlyat high baking temperatures (�80°C).

If too much hydrolysis occurs during dough formation and baking, the dough will be gummyand sticky, and the baked bread may also have a similar texture. Depending on the specific ap-plication, residual activity may remain in the bread even after baking, resulting in an increase in

Box 8–6. Fresh Ideas for Controlling Staling of Bread (Continued)

amylose(amorphous) amylopectin

(gelled)

amylose(crysallline)

amylopectin(crystalline)

Figure 1. Model of bread staling. In fresh bread, amylose and amylopectin exist in amorphous or gelledforms. During storage, moisture is lost, the amylose and amylopectin retrograde, and crystalline forms ap-pear, leading to firmness and staling. Adapted from Zobel and Kulp, 1996 and Lallemand Baking Update,Volume 1/Number 6.

(Continued)



Preservation

Given that biological spoilage of bread iscaused primarily by molds, it is not surprisingthat preservation strategies have focused oncontrolling fungi, both in the production envi-ronment and in the finished product.As notedpreviously, mold and mold spores are presentin flour and other raw materials and may bewidespread in bakeries.Therefore, rigorous at-tention to plant design and sanitation is essen-tial. The post-production environment (i.e.,baked products) should be separated from pre-production environments. Air handling sys-tems should be designed (including the use offilters or ultraviolet lamps and positive air pres-sure) such that airborne mold spores cannotgain entry to the product side.

Several different approaches have beenadopted to control mold growth in bread. In

starch hydrolysis and free sugars during storage.Thus, there is interest in identifying �-amylasesthat are less thermal-resistant.Although the fungal-produced �-amylases are more heat-labile (orhave intermediate heat tolerance) than bacterial enzymes, they typically have less activity at thebaking temperatures normally used.

Processing

The manner in which the dough is mixed and fermented has considerable impact on bread stal-ing. In general, bread freshness may be poorly retained during storage if the dough is over-mixed or under-mixed. Likewise, if the fermentation time is too short or too long, bread fresh-ness is reduced. Thus, all other factors being equal, bread made from no-time doughs orlong-time (i.e., straight) doughs usually will be less soft than bread made using continuous orsponge and dough processes.

References

Bowles, L.K. 1996.Amylolytic enzymes, p. 105—129, In Hebeda, R.E., and H. F. Zobel. (ed.). Baked goodsfreshness. Marcel Dekker, Inc. New York.

Gray, J.A., and J.N. Bemiller. 2003. Bread staling: molecular basis and control. Comp. Rev. Food Sci. FoodSafety 2:1–21.

León,A.E.,E.Durán and C.B.de Barber.2002.Utilization of enzyme mixtures to retard bread crumb firming.J.Agri. Food Chem. 50:1416–1419.

Knightly,W.H. 1996. Surfactants. p. 65—103, In Hebeda, R.E., and H. F. Zobel. (ed.). Baked goods freshness.Marcel Dekker, Inc. New York.

Zobel,H.F.,and K.Kulp.1996.The staling mechanism,p.1—64, In Hebeda,R.E.,and H.F.Zobel. (ed.).Bakedgoods freshness. Marcel Dekker, Inc. New York.

Box 8–6. Fresh Ideas for Controlling Staling of Bread (Continued)

Table 8.5. Spoilage organisms of bread.

Organism Appearance or defect

FungiAspergillus niger BlackAspergillus glaucus Green, grey-greenAspergillus flavus Olive greenPenicillium sp. Blue-greenRhizopus nigricans Grey-blackMucor sp. GreyNeurospora sitophila Pink

YeastsSaccharomyces Alcoholic-estery

cerevisiae off-odorsPichia burtonii White, chalky appearance

BacteriaBacillus subtilus RopyBacillus licheniformis RopyBacillus mesentericus Ropy

Adapted from Pateras, 1998



the United States, the most common means ofbread preservation is to use propionate salts(Box 8–3). Calcium propionate is especially ef-fective against most of the molds associatedwith bread,and is widely used in commercially-produced products. Sorbate and acetate saltsare also used as anti-mycotic preservatives inbread. However, these acids also have inhib-itory activity against baking yeasts, whereaspropionates are much less inhibitory.Moreover,calcium propionate has no flavor or toxicity, isactive even against rope-forming bacteria, andis effective in most varieties of bread.There are,however, some fungal strains (e.g., Penicilliumroquefortii) that are associated with rye breadsand that are insensitive to propionates.

Due to interest in chemical additive-freebreads, alternative approaches for bread preser-vation have been considered. Bread can be ex-posed to ultraviolet, infrared,or microwave radi-ation to inactivate mold and mold spores orpackaged in modified or vacuum atmospheresto inhibit their growth. These methods, how-ever, are not widely used. In contrast, one effec-tive method used for bread preservation that isvery popular and does not involve addition ofchemical preservatives instead relies on growthof lactic acid bacteria in the dough.As discussedabove, these bacteria produce lactic and aceticacids and lower the pH to levels inhibitory tomost fungi.

Another indirect way to extend the shelf-lifeof bread is via freezing.Many bread manufactur-

ers freeze the baked and packaged breads as ameans of preserving the bread prior to delivery.Another freezing method that has been adoptedis to freeze un-baked breads (Box 8–7). Thisprocess is gaining popularity, since it allows re-tailers to do the baking and then sell “fresh”-baked breads (aroma and all) at the retail level.

Bread Quality

Assessing the quality of bread is a very subjec-tive process. Cultural, ethnic, and personal atti-tudes certainly influence the sort of bread anindividual prefers.For example, the hard,crustybaguette eaten in France bears little resem-blance to the soft, doughy, plastic-wrappedFrench bread preferred by many U.S. con-sumers.Whole grain peasant breads—so-namedbecause centuries ago expensive refined flourswent to the upper classes, while the poor wereleft with unrefined, whole flours—are nowpopular in specialty bakeries serving an afflu-ent, upscale clientele.

The criteria commonly used to assess breadquality are based primarily on appearance, tex-ture, and flavor attributes. Appearance refersboth to external (i.e., crust), as well as internalappearance (i.e., crumb). Depending on thetype of bread, the intact loaf should have a par-ticular volume, or in some cases, height andlength.The color of the crust also is an impor-tant property, and is judged mainly on the ba-sis of the expected color established by the

Figure 8–10. Spoilage of bread by fungi. Shown are Penicillium commune (A), Aspergillus niger (B), Pencil-lium roqueforti (C), and Rhizopus stolonifer (D). Photos courtesy of A. Bianchini and L. Bullerman, University ofNebraska.



Box 8–7. Frozen Doughs and Yeast Cryotolerance

One of the hottest trends in the baking industry is the use of frozen doughs.These doughs areespecially appealing to small retail bakery operations (like those in grocery stores), becausethey eliminate the need for dough production equipment and labor, while at the same timemaking it possible to offer fresh-baked bread products to customers.

Frozen doughs are particularly convenient for the end-user.The frozen dough is simply re-moved from the freezer, thawed overnight in the refrigerator, given a final proof, and baked. De-spite these advantages, however, the quality of breads made from frozen doughs can be quitevariable, due to the loss of yeast viability during storage. Doughs made using cold-sensitiveyeasts not only require longer proof times,but the bakers’yeast strains ordinarily used for breadmanufacture are so cryosensitive that they may fail to provide any leavening at all after thedough-thawing step.Therefore, for frozen dough manufacture, it is essential that the yeast strainbe cryotolerant.

What makes some yeast strains resistant to the effects of freezing? And what can be done toimprove cryotolerance? These questions are now receiving much attention, and the answersare, at least in part, inter-related.

First, it appears that cryotolerance in Saccharomyces cerevisiae is due the ability of the or-ganism either to transport extracellular cryoprotectant solutes from the environment or to syn-thesize them within the cytoplasm.These solutes, much like osmoprotectants, are then accu-mulated to high intracellular concentrations, without causing detrimental effects on theenzymatic or reproductive machinery within the cell.

At a molecular level, these agents prevent dehydration, presumably by re-structuring thebound water in the cytoplasm. Microbial cryoprotectants are typically small, polar moleculesand include several amines (e.g.,betaine,carnitine,proline, and arginine), as well as various sug-ars. In S.cerevisiae, the disaccharide trehalose (which consists of two glucose moieties linked �-1,1; Figure 1, upper panel) is probably the most important cryoprotectant (although someamino acids also have cryoprotective activity).When the temperature decreases (or when otherenvironmental stresses are applied), the cell responds by increasing de novo synthesis of tre-halose. Extracellular trehalose, if available, may also be transported via a high affinity trehalose(�-glucoside) transport system. Ultimately, intracellular levels can reach up to 20% of the totalweight of the cell (on a dry basis).

The trehalose concentration in yeasts is not always maintained at such a high level. Rather,there is a balance between trehalose synthesis and metabolism, such that under non-stress con-ditions, the enzymes used to synthesize trehalose (the trehalose synthetase complex) areturned off and the hydrolytic enzymes (trehalases) are turned on (Figure 1, lower panel).Tre-halose, it is important to note, can also serve as a storage carbohydrate in S. cerevisiae, provid-ing a readily metabolizable source of energy for when exogenous sugars are in short supply.

The coordination of the catabolic and anabolic enzymes, combined with the trehalose trans-porter, therefore, allows the cell to modulate the intracellular trehalose concentration, depend-ing on the particular circumstances in which it finds itself.This system,however,can also be ma-nipulated genetically for very practical purposes. If, for example, the trehalose degradationpathway is blocked, then higher levels of trehalose can be maintained, making the yeast morecryotolerant. One recently described strategy involved inactivation of genes encoding for anacid trehalase (Ath1p) and a neutral trehalase (Nth1p).When yeasts containing these defectivegenes were used to make frozen doughs, cryotolerance was enhanced, as observed by the in-crease in gassing power (CO2 produced per minute) in the thawed doughs.

Despite its importance in cryotolerance, trehalose is not the only cryoprotectant moleculeaccumulated by S. cerevisiae. As noted above, arginine can also serve this role. Thus, an ap-proach similar to that described for trehalose has recently been used to enhance crytolerancein a bakers’ yeast strain of S. cerevisiae. In this case, the CAR1 gene, encoding for the arginine-hydrolyzing enzyme arginase was inactivated.This led to increased intracellular concentrations



(Continued)

of arginine and an increase in freeze tolerance and gassing power compared to the parentstrain.

Despite the promising experimental results with genetically modified strains, such strains arenot used commercially for frozen bread manufacture. Instead, the industry relies more on ma-nipulating manufacturing conditions to enhance yeast performance. It is essential, for example,

Box 8–7. Frozen Doughs and Yeast Cryotolerance (Continued)

OHOH2O

O

CH2OH

O

out

in

H+

H+trehalose

trehalosepermeasemembrane

trehalose

glucose

glucose-6-P

trehalose-6-P synthase

trehalose-6-P

hexokinase

trehalose-6-P phosphatasePitrehalase

Figure 1. Trehalose metabolism by Saccharomyces cerevisiae. Structure of trehalose (�-D-glucosyl-1→1-�-D-glucose) is shown in the upper panel. Transport is mediated by a symport system, but synthesis fromglucose-6-phosphate (P) may be the major route for trehalose accumulation (lower panel). Trehalose notonly alleviates stress, but it can also serve as a carbon source following its hydrolysis to glucose by the en-zyme trehalase.


manufacturer. Internal appearance is alsobased, in part, on color, specifically, the ab-sence of streaks or uneven color.The size, fre-quency, and distribution of air cells are impor-tant determinants of bread quality, especially ifthere are large, irregular-shaped holes that areoften considered as defects.

In contrast to the appearance properties,which are difficult to assess objectively, meth-ods do exist to measure texture properties. In-strumental devices, such as the Instron Uni-versal Testing Machine, can be used tomeasure crumb softness, resistance, and com-pressibility. Other compression-type instru-ments also exist, such as the Baker Compres-simeter and the Voland Stevens TextureAnalyzer, which can be used to measure firm-ing changes and staling rates. Various spec-trophotometric methods, including near in-frared and infrared spectroscopy can detectchanges in starch crystallization and confor-mation, and are especially useful for measur-ing retrogradation and staling rates. However,X-ray diffraction is now generally consideredthe most definitive technique for detectingchanges in crystallization.

Bread flavor is no less easy to describe orquantify. It can also be difficult to separate fla-vor from texture and other sensory attributes.In general, bread should have a fresh, some-

what yeasty flavor and aroma, with acidic andwheaty notes. Salt also provides an importantflavor in bread, as do the alcohols, aldehydes,and other organic end products produced dur-ing yeast metabolism.

Many of the important flavor compoundsof bread are formed during baking and are de-rived from caramelization and Maillard reac-tions; the Maillard reaction alone results inmore than 100 volatile aroma compounds. Im-portantly, these reactions also generate pig-ments and are responsible for crust color for-mation. Although both types of reactionsrequire heat, are non-enzymatic, and use sug-ars as the primary reactant, they differ in thatcaramelization reactions occur at higher tem-peratures than Maillard reactions and only re-ducing sugars react in the Maillard reaction.The latter also requires primary amino acids.The Maillard reaction is probably predomi-nant during baking, since the temperaturesachieved generally are not high enough to in-duce extensive caramelization.

BibliographyCalvel, R. 2001. The Taste of Bread. (R.L.Wirtz, trans-

lator and J.J. MacGuire, technical editor). AspenPublishers, Inc. Gaithersburg, Maryland.

Cauvain, S.P. 1998. Breadmaking Processes, p. 18—44. In S.P. Cauvain and L.S.Young (ed.), Technol-


that naturally cryotolerant strains or mutant strains derived from wild type strains be used(Codón et al., 2003). Of course, increasing the amount of yeast is also important to compensatefor any loss of activity. Finally, minimizing the extent of fermentation prior to freezing appearsto improve yeast viability after thawing.

References

Codón,A.C.,A.M Rincón, M.A. Moreno-Mateos, J. Delgado-Jarana, M. Rey, C. Limón, I.V. Rosado, B. Cubero, X.Peñate, F.Castrejon, and T. Benítez. 2003. New Saccharomyces cerevisiae baker’s yeast displaying en-hanced resistance to freezing. J.Agri. Food Chem. 51:483–491.

Shima, J.,A. Hino, C.Yamada-Iyo,Y. Suzuki, R. Nakajima, H.Watanabe, K. Mori, and H.Takano. 1999. Stress tol-erance in doughs of Saccharomyces cerevisiae trehalose mutants derived from commercial baker’syeast.Appl. Environ. Microbiol. 65:2841–2846.

Shima, J.,Y. Sakata-Tsuda,Y. Suzuki, R. Nakajima, H.Watanabe, S. Kawamoto, and H.Takano. 2003. Disruptionof the CAR1 gene encoding arginase enhances freeze tolerance of the commercial baker’s yeast Sac-charomyces cerevisiae. Appl. Environ. Microbiol. 69:715–718.

Box 8–7. Frozen Doughs and Yeast Cryotolerance (Continued)



ogy of Breadmaking. Blackie Academic and Pro-fessional (Chapman and Hall). London, UK.

Gobbetti, M., and A. Corsetti. 1997. Lactobacillussanfrancisco a key sourdough lactic acid bac-terium: a review. Food Microbiol. 14:175–187.

Gray, J.A., and J.N. Bemiller. 2003. Bread staling: mo-lecular basis and control. Comp. Rev. Food Sci.Food Safety. 2:1–21.

Kline, L., and T.F. Sugihara. 1971. Microorganisms ofthe San Francisco sour dough bread process. II.Isolation and characterization of undescribedbacterial species responsible for the souring ac-tivity.Appl. Microbiol. 21:459–465.

Lallemand Baking Update, Lallemand Inc., Montreal,QC, Canada.

Okada, S., M. Ishikawa, I. Yoshida, T. Uchimura, N.Ohara, and M. Kozaki. 1992. Identification andcharacteristics of lactic acid bacteria isolatedfrom sour dough sponges. Biosci. Biotech.Biochem. 56:572–575.

Pateras, I.M.C. 1998. Bread spoilage and staling, p.240—261. In S.P. Cauvain and L.S. Young (ed.),Technology of Breadmaking. Blackie Academicand Professional (Chapman and Hall).

Pulvirenti, A., L. Solieri, M. Gullo, L. De Vero, and P.Giudici. 2004. Occurrence and dominance of

yeast species in sourdough. Lett.Appl. Microbiol.38:113–117.

Pyler,E.J.,1988.Baking Science and Technology.Vol-umes 1 and 2, 3rd edition. Sosland PublishingCo. Merriam, Kansas.

Röcken, W., and P.A. Voysey. 1995. Sour-dough fer-mentation in bread making. J. Appl. BacteriolSymp. Supple. 79:38S-48S.

Stear, C.A. 1990. Handbook of Breadmaking Tech-nology. Elsevier Applied Science, New York.

Stolz,P.,G.Böcker,R.F.Vogel, and W.P.Hammes.1993.Utilization of maltose and glucose by lactobacilliisolated from sourdough. FEMS Microbiol. Lett.109:237–242.

Sugihara,T.F., L.Kline,and M.W.Miller.1971.Microor-ganisms of the San Francisco sour dough breadprocess. I.Yeasts responsible for the leavening ac-tion.Appl. Microbiol. 21:456–458.

Vogel, R.F., G. Böcker, P. Stolz, M. Ehrmann, D. Fanta,W. Ludwig, B. Pot, K. Kersters, K.H. Schleifer, andW.P. Hammes. 1994. Identification of lactobacillifrom sourdough and description of Lactobacil-lus pontis sp. nov. Int. J. System. Bacteriol.44:223–229.


9

Beer Fermentation

“I would give all my fame for a pot of ale.”King Henry V, Act 3, Scene 2, William Shakespeare

“I would kill everyone in this room for a drop of sweet beer.”Homer Simpson

continued to grow throughout Europe duringthe Middle Ages.

During that time, beer making was consid-ered an art, performed by skilled craftsmen.Many of the early breweries were locatedwithin various European monasteries, whichcreated a tradition of brewing expertise and in-novation.The use of hops in beer as a flavoringagent, for example, was first practiced bymonks, as was the use of bottom-fermentingyeasts (discussed below).Although beer manu-facture was practiced throughout Europe, itwas of particular cultural and economic impor-tance in Great Britain and Germany, which be-came the epicenters for brewing technology.

By the late eighteenth century, and espe-cially by the mid-1800s, beer making was oneof the first food processes to become industri-alized. Although monastery-based brewerieswere still common, many small breweries be-gan to form, serving their product directly onthe premises (much like the brew pubs thatare popular even today). Eventually, some ofthe larger breweries contained not only pro-duction facilities, but also the beginning ofwhat we might now consider to be qualitycontrol laboratories.And although these brew-eries were located mainly in Europe and Eng-land, beer making had also spread to NorthAmerica and the American colonies.

Introduction

In 1992, a research team from the Applied Sci-ence Center for Archaeology at the Universityof Pennsylvania made a discovery that was pub-licized within archaeology circles (and collegecampuses) around the world.These researchershad analyzed a small amount of organic residuefrom inside an ancient pottery vessel that hadbeen retrieved from the Zagros Mountains ofWestern Iran. When the residue from this claypot (which itself was dated circa 3500-3100B.C.E.) was analyzed, the results revealed thepresence of oxalate ion. Since oxalate, and cal-cium oxalate in particular, collect in only a fewplaces, this finding could mean only onething—this pot had been used to brew beer.The 5,000-year-old residue, the archaeologistsconcluded, was the earliest chemical evidencefor the origins of beer in the world.

Although this discovery attracted its shareof headlines, it was but one of many events inthe remarkable historical record of beer mak-ing (Figure 9–1). Beers were widely preparedand consumed in Egypt and other Middle East-ern counties from ancient days and thenspread west into Europe and the British Isles.Despite the introduction of Islam in the eighthcentury, and its prohibition against alcoholconsumption, beer making and consumption

301





Reinheitsgebot Beer Purity Law1516

1876 Pasteur publishes “Etudes sur la Biere”

1620Mayflower (with beer in the cargo) lands atPlymouth Rock, Massachusetts

Brewers begins adding hops to beer736

Carlsberg Laboratory established in Denmark1875

Beer making established in the Middle East3500 B.C.

18TH Amendment (Prohibition) ratified by the U.S.Congress, most U.S. breweries close1919

Prohibition repealed1933

Miller Lite introduced by Miller Brewing Co. 1975

U.S. minimum drinking age raised to 211984

1637Beer prices fixed in Massachusetts (“not morethan one penny per quart”)

1876 Anheuser and Co. (later named Anheuser-Busch)introduces Budweiser Lager Beer

Ice Beer introduced by Labatt Brewing Co. 1994

Brewing strain of Saccharomyces cerevisiae purifiedby Hansen from Carlsberg Brewery

1883

Saccharomyces cerevisiae genome sequenced1996

First canned beer (Krueger's Finest Beer and Krueger'sCream Ale, Richmond, Virginia)

1935

Schwann recognizes the genus, Saccharomyces 1837

Lager strain smuggled from Bavaria to Pilsen 1842

van Leewenhoek, using a hand-made microscope,observes yeasts in beer

1677

Figure 9–1. Milestones in the history of beer.


Beer Fermentation 303

In fact, English ale-style beer had beenamong the provisions carried by Pilgrims onthe Mayflower, and there are reports, written inthe actual travel logs, of the ship running lowon beer and the worries that caused the pas-sengers. (“We could not now take time for fur-ther search . . . our victuals being much spent,especially our beer. . .”) During the latter halfof the nineteenth century, the number of brew-eries in the United States increased three-fold,from 400 to 1,300.Most of these breweries pro-duced German lager-style beer (see below), re-flecting the huge immigrant population fromGermany during that era (including the brew-ers themselves, e.g.,Adolphus Busch, EberhardAnheuser, Adolph Coors, Frederick Miller,Joseph Schlitz).

Beer Spoilage and the Origins ofModern Science

Most fermentation microbiology students areaware that cheese, sausage, and other fer-mented foods evolved, in part, because theseproducts had unique and desirable sensorycharacteristics.Likewise, they might also appre-ciate the many pleasant attributes of maltedand hopped beverages.However, it is importantto recognize that,while our ancestors undoubt-edly enjoyed beer for many of the same reasonsas today’s consumers, they also understood thatbeer, like other fermented foods,was somehowbetter preserved than the raw materials fromwhich they were made. For hundreds of years,for example, early trans-Atlantic voyagers (likethe Pilgrims mentioned above) relied on beerbecause it was a well-preserved form of nour-ishment. In fact, until relatively recently, beerwas often safer to drink than water, was lesslikely to cause water-borne disease, and wasless susceptible to spoilage. Of course, we nowknow that the microbiological stability and“safety” of beer is due, in part, to its ethanolcontent, as well as other anti-microbial con-stituents and properties (discussed, in more de-tail, later in this chapter).

Despite the stability and general acceptanceof beer as a well preserved product, it can in-

deed become spoiled. As beer manufacturingin Europe grew from small, craft-oriented pro-duction into a large brewing industry in themiddle of the nineteenth century, the preven-tion of beer spoilage became an important goalof industrial brewers,who were often troubledby inconsistencies in product quality.These dif-ficulties attracted the attention of chemists andother scientists, who were enlisted to solvesome of these technical problems.

Beer manufacture was, in fact,one of the firstindustrial fermentations to be studied and char-acterized, and was the subject of scientific in-quiry by early microbiologists and biochemists.The very development of those scientific disci-plines coincides with the study of beer andother fermented foods (Box 9–1). In particular,the science of beer making was revolutionizedin 1876 by Pasteur, who not only showed thatyeasts were the organisms responsible for thefermentation, but also that the presence of spe-cific organisms were associated with specifictypes of spoilage. Pasteur also developedprocesses to reduce contamination and pre-serve the finished product. It is worth notingthat even today,preventing beer spoilage by mi-croorganisms is still an important challengefaced by the brewing industry.

Scientific interest in brewing extendedthroughout Europe, leading to establishmentof research laboratories in Copenhagen (theCarlsberg Laboratory) and Bavaria (the Facultyof Brewing), as well as laboratories locatedwithin several breweries. Perhaps the mostnoteworthy discoveries of the late nineteenthand early twentieth centuries were made byEmil Christian Hansen at the Carlsberg Labora-tory. He developed pure culture techniques foryeasts,which eventually led to methods (still inuse today) for propagation and production ofyeast starter cultures free of contaminatingbacteria and wild yeasts.

The Modern Beer Industry

Today, beer has one of the largest dollar valuesof all fermented food products, with U.S. retailsales in 2002 of more than $65 billion dollars.



Box 9–1. Pasteur, the Origins of Microbiology, and Beer

It is hard to imagine, given the current age of scientific specialization, that one person couldhave been as accomplished in so many fields as was Louis Pasteur in the latter half of the nine-teenth century. He was trained as a chemist and, at the age of only 26, made important discov-eries in stereochemistry and crystallography (specifically, while examining tartaric acid crystalsin wine). Pasteur then became interested in the not yet named field of microbiology, and de-voted nearly two decades of his life disproving the theory of spontaneous generation and es-tablishing microorganisms as the causative agent of fermentation and putrefaction.

His work on vaccines against animal anthrax and human rabies, as well as the establishmentof the germ theory of disease, brought Pasteur worldwide recognition and fame (Schwartz,2001). During his career, Pasteur addressed practical agricultural, food, and medical problems,but did so by relying on basic science (“There is no such thing as applied science, only applica-tions of science,”as quoted in Baxter,2001).He was a meticulous researcher,who rigorously de-fended the scientific method, saying,“In the field of observation, chance favors only the pre-pared mind.”

There were certainly other well-known scientists of his era, including Schwann, Koch, andLister, whose works also helped establish the field of microbiology. But the collective contribu-tions of Pasteur led to his recognition as one of the “most distinguished microbiologists of alltime” (Barnett, 2000), or as stated by Krasner (1995), the “high priest of microbiology.”

Pasteur was already an accomplished scientist, having completed his work on the stereo-chemistry of tartaric acid, when he turned his attention to solving spoilage problems plaguingthe French wine industry. In reality,he was “commissioned”by Emperor Napoleon III in 1863 toserve as a “consultant” and to study these so-called wine diseases. Pasteur correctly diagnosedthe spoilage conditions,which he called tourne (mousy),pousse (gassy),and amertume (bitter),as being caused by bacteria. He then showed that the responsible organisms could be killed,and the wine stabilized, by simply heating the wine to 50°C to 60°C.These findings were pub-lished in 1866, as “Études sur le Vin.”

Nearly ten years later, Pasteur began studying the fermentation of beer, another industrially-important product that was also suffering from spoilage defects.This work led to a second trea-tise on alcoholic fermentations,“Études sur la bière,” published in 1876 (the English-translatedversion is available and well worth reading). His motivation this time had less to do with his in-terest in beer (he apparently was not a beer drinker), than it did for his animosity for all thingsGerman (Baxter, 2001).

In the early 1870s, France and Germany were at war (the Franco-Prussian War), whose out-come resulted in France having to cede the hop-producing region of Alsace-Lorraine to Ger-many.Already Germany produced the best beer and was the dominant producer in Europe.Ger-man beers were better from most English and other beers for several reasons, but mainlybecause these beers were fermented at a low temperature (6°C to 8°C). The resulting “lowbeers” (so-named because of the low fermentation temperature, but also because the yeastwould sink to the bottom of the tank) were lighter-colored and less heavy than “high beers,”butmost importantly, they were also better preserved.

Low beers (which we now refer to as lagers) could be produced in the cool winter months,and, provided they were stored cold (in cellars or by ice), be consumed during the summer.High beers (or ales) had to be consumed shortly after manufacture and did not travel well.Thissituation, in which French beers were at a commercial disadvantage, was intolerable to Pasteurand he was determined to improve French-made beers or as he himself stated, to make the“Beer of the National Revenge.”

Pasteur started out by building a brewing “pilot plant” in the basement of his laboratory andby visiting commercial breweries.Then, using microscopic techniques, he showed that specificyeasts were necessary for a successful beer fermentation. He also observed that, based on mi-croscopic morphology, certain contaminating microorganisms were associated with specific



In 2002, U.S. consumers drank nearly 190 mil-lion barrels (1 barrel � 117 L � 31 gallons),with per capita consumption of 81 L or 21.5gallons per person per year (or nearly forty six-packs per man, woman, and child!).That’s alsomore, on a volume basis, than wine (8 L), juice(34 L), milk (72 L), and bottled water (42 L).Al-though one might assume that this is a lot ofbeer, beer consumption in the United States ishalf as much as that of several other Europeancountries (Figure 9–2). For example, per capitaconsumption (for 2002) in Germany and Ire-land was 121 L and 125 L per person per year,respectively, and Austria (109 L) and the U.K.(100 L) are not far behind.The Czech Republic,

however, leads all other countries, with percapita consumption of 162 L (2003).

The importance of beer to the world econ-omy is also considerable. In the United Statesalone, the overall economic impact is esti-mated to be more than $144 billion, and the in-dustry pays more than $5 billion directly in ex-cise taxes to federal and state government(according to the U.S. Beer Institute). The in-dustry also spends about $1 billion per year foradvertisements and promotions.

The beer industry, like many other segmentsof the food industry,both in the United States aswell as internationally, has become highly con-solidated. In the past fifty years, many medium

types of spoilage conditions (or “diseases”). For example, long rod-shaped or spherical-shapedorganisms in chains (probably lactic acid bacteria) were responsible for a sour defect. In Pas-teur’s own words,“Every unhealthy change in the quality of beer coincides with a developmentof microscopic germs which are alien to the pure ferment of beer”(Pasteur,1879).The corollarywas also true:“The absence of change in wort and beer coincides with the absence of foreignorganisms.”These foreign organisms had gained entrance into the beer either by virtue of theirpresence in the production environment or via the yeast (which were used repeatedly andwere subject to contamination).

To solve these problems,Pasteur offered several recommendations.He showed that a modestheating step (55°C to 60°C) would destroy spoilage bacteria and render a beer palatable evenafter nine months. He also emphasized that elimination of contaminants, both environmentaland within yeast preparations, would also be effective. Oxygen, Pasteur realized, enhancedgrowth of spoilage organisms, thus, he devised a brewing protocol that precluded exposure ofthe wort to air (described in a U.S. patent; Figure 1).

Pasteur demonstrated through actual industrial scale experiments that these strategies wouldwork,and indeed these practices were readily adopted.And although the German beer industrywas hardly affected by his personal “revenge,” Pasteur’s influence on brewing science and mi-crobiology is considerable even today, more than 125 years later.

References

Barnett, J.A., 2000. A history of research on yeasts 2: Louis Pasteur and his contemporaries, 1850–1880.Yeast. 16:755–771.

Baxter,A.G. 2001. Louis Pasteur’s beer of revenge. Nature Rev. 1:229–232.Krasner, R.I. 1995. Pasteur: high priest of microbiology.ASM News. 61:575–579.Pasteur,L.1879. Studies on fermentation.The diseases of beer, their causes, and the means of preventing

them.A translation,made with the author’s sanction,of “Études sur la bière”with notes, index, and orig-inal illustration by Fran Faulkner and D. Constable Robb. Macmilan and Co. 1879, London and KrausReprint Co., New York, 1969.

Schwartz, M. 2001.The life and works of Louis Pasteur. J.Appl. Microbiol. 91:597–601.

Box 9–1. Pasteur, the Origins of Microbiology, and Beer (Continued)

(Continued)



Box 9–1. Pasteur, the Origins of Microbiology, and Beer (Continued)

Fig

ure

1.

“Im

prov

emen

t in

bre

win

g be

er a

nd a

le,”

a U

.S. p

aten

t (N

o. 1

35,2

45) i

ssue

d to

Lou

is P

aste

ur o

f Pa

ris, F

ranc

e in

187

3. T

he p

aten

t is

not

for

“pa

s-te

uriz

atio

n,”

but

rath

er d

escr

ibes

a m

etho

d fo

r ex

clud

ing

air

from

the

boi

led

wor

t pr

ior

to in

ocul

atio

n w

ith y

east

. The

res

ultin

g be

er h

as im

prov

ed s

tabi

lity,

or,

acco

rdin

g to

the

pat

ent,

“po

sses

ses

in a

n em

inen

t de

gree

the

cap

acity

for

unc

hang

eabl

enes

s.”



sized breweries were either acquired by largerbreweries or were unable to compete and wentout of business.Where there were once severalhundred small,privately owned breweries oper-ating in the United States as recently as 1950,now more than 80% of U.S.beer is produced byjust three companies.Despite this centralizationof the beer industry, during the past twentyyears there has been a remarkable increase(from zero to more than 1,000) in the numberof small breweries (small enough to earn thedesignation “microbreweries”) that are now op-erating in all regions of the United States.Thesebreweries are important to mention here, notonly because they have captured a small but no-ticeable share of the market, but also because

they have adopted many of the traditional brew-ing practices that will be described later in thischapter.

Beer Manufacturing Principles

Only four ingredients are necessary to makebeer: water, malt, hops, and yeast. Despite itsancient origins and long history, and this seem-ingly short list of ingredients, the manufactureof a quality beer remains a rather challengingtask. In part, this is because beer making con-sists of several different and distinct processesthat are not always easy to control. In addition,some steps taken to improve one aspect of theprocess—for example, filtering the finished

0

20

40

60

80

100

120

140

L

Aust

ralia

Belg

ium

Den

mar

kFi

nlan

d

Ger

man

yG

reec

eIre

land

Italy

Luxe

mbo

urg

Portu

gal

Net

ehrla

nds

Nor

way

Switz

erla

nd

Spai

nSw

eden

Uni

ted

Kin

gdom

Uni

ted

Sta

tes

28

99

79

5459

73

56

100

52

83

125122

35

97

80

9096

Fran

ce

39

162

160

Cze

chR

epub

lic

51

Rus

sia

Aust

ria

109

Figure 9–2. Worldwide per capita beer consumption (L per person per year). Adapted from Brewers ofEurope 2002 Statistics and Beer Facts.



beer to enhance clarity—may at the same timeremove desirable flavor and body constituents.

The actual brewing process involves notonly the well-studied yeast fermentation, butalso entails other biological, as well as chemi-cal and physical reactions. It is, therefore, con-venient to consider the beer manufacturingprocess as consisting of several distinct phasesor steps.

The primary purpose of the first phase,called mashing, is to transform non-fer-mentable starch into sugars that the yeast canferment.The process,which is enzymatic in na-ture, involves a series of biochemical eventsthat starts with the conversion of cereal grain,usually barley, into malt.The malt is then usedto make a mash, ultimately resulting in the for-mation of a nutrient-rich growth medium,called wort. Although other grains, such assorghum,maize, and wheat can also be malted,barley is by far the most frequently malted ce-real grain.

In the second, or fermentation, phase,which is microbiological in nature, sugars,amino acids, and other nutrients present in thewort are used to support growth of yeasts.Yeast growth, under the anaerobic conditionsthat are soon established, is accompanied byfermentation of sugars and formation of theend products ethanol and CO2. Technicallyspeaking, the fermentation step results in beerthat could then be consumed. However, at thispoint, the beer contains yeast cells, insolubleprotein-complexes, and other materials thatcause a cloudy or hazy appearance. In addition,carbon dioxide is lost during the fermentation.Beer at this stage is also microbiologically un-stable and susceptible to spoilage. Therefore,additional measures are almost always taken toremove yeasts and other microorganisms andany other substances that would otherwise af-fect product quality and shelf-life and to pro-vide carbonation in the final product.

In this chapter, therefore, a third phase, con-sisting of important post-fermentation activi-ties, also will be discussed.These latter steps ofthe beer-making process, some might argue,are among the most important, since they have

a profound effect on the appearance, flavor,and stability of the finished product.The gen-eral beer manufacturing process is outlined inFigure 9–3. As the reader will soon be aware,many beer-specific terms, such as mashing, areused to describe the brewing process. Some ofthese terms are defined in Box 9–2.

Enzymatic Reactions: Malting and Mashing

The first part of beer manufacture, the enzy-matic steps, actually begins far from the brew-ing facility, in the malting houses that convertbarley into malt. It is the malt that serves as thesource of the amylases, proteinases, and otherenzymes necessary for hydrolysis of largemacromolecules, such as starch and protein.For most beers, the malt also serves as the sub-strates for those enzymes (i.e., malt containsthe starch and protein hydrolyzed by malt en-zymes). In addition, malt is the primary deter-minant of color and body characteristics inbeer, and influences flavor development. Con-sidering the critical functions that malt con-tributes to the beer-making process, it is evi-dent that malt quality has a profound influenceon beer quality.

The goal of the maltster is to convert barleyto malt such that enzyme synthesis is maxi-mized and enzyme activity is stable and well-preserved.The process starts with selection ofbarley. Several different barley cultivars areused to produce malt. In North America, six-row barley is generally preferred, whereas inEurope and the U.K., two-row barley is used(however, some six-row is also used in Europeand the U.K. and some two-row is used in theUnited States). Six-row barley contains moreprotein and less starch than two-row barley.The former also contains a higher level ofstarch-degrading enzymes, making it more suit-able for beers containing adjuncts (discussedbelow). The harvested barley is then cleaned,graded, and sized. Unless the barley is to beused right away (which is usually not thecase), it is dried from about 25% moisture to10% to 12% so that it can be safely stored.



In the next step, the barley is allowed togerminate or to develop the beginning of aroot system.Barley is essentially a seed—if it isheld under warm, moist conditions, it will be-gin to sprout or root, just as it would if it weregrowing in nature. For malting purposes, ger-mination is done by steeping the barley incool water at 10°C to 20°C for two to threedays, or long enough to increase the moisturefrom 10% to 12% to about 45%.The steep wa-ter is usually changed every twelve hours, inlarge part to minimize the impact of potentialspoilage organisms. During this step, the mate-rial is well aerated to promote the germinationprocess. The moist barley is then removedfrom the steep water and incubated in trays ordrums under cool, humid, and well-aeratedconditions for two to eight days. The barleyinitially swells, then germinates, such that

“rootlets”or sprouts appear.Germination is ac-companied by the synthesis of myriad en-zymes that the barley grain theoretically needsfor subsequent growth into new barley plants.The barley has reached its maximum enzymeactivity when the sprout reaches a length ofone-third the size of the grain.

Next, it is necessary to arrest further germi-nation and to stabilize and preserve the enzy-matic activity.This is done by a step-wise dry-ing process, in which the germinated “green”barley is dried incrementally within a tempera-ture range of 45°C to 60°C.The purpose of thisslow drying process is to remove water with-out inactivating enzymes.The dried malt, there-fore, becomes a stable source of enzymes.Since moist heat is more detrimental to en-zymes than dry heat, it is important to dryslowly at the start (when the grains are still

Malting Mashing Fermentation Post-Fermentation

Barley

Malt

Steeping andGermination

Drying andKilning

Fermentation

Pitch

Wort

Conditioningand

Maturation

Beer

Clarificationand

Filtration

Packagingand

Pasteurization

Beer

Beer

WaterAdjuncts

Milling

Wort

Kettle Boil

Hops

Malt

Heating

Figure 9–3. Manufacture of beer. The beer manufacture process consists of four distinct stages: malting,in which barley is converted to malt; mashing, in which enzyme and substrate extraction and reactions occurand a suitable growth medium is prepared; fermentation, in which wort sugars are fermented to beer; andpost-fermentation, in which the beer is made suitable for consumption.



wet), and only later, when the moisture is re-duced, to dry at higher temperatures.

During the drying step, the moisture dropsfrom 45% at the start to about 15% to 18% atthe end.The dried malt is then further dried—or “cured”or “kilned”—at temperatures as highas 80°C, and the moisture decreases to lessthan 5%.At this point the water activity of themalt is usually reduced to 0.3 or lower, so en-zymes are well preserved and no microbialgrowth is expected to occur.At this point, the

malt contains mostly carbohydrate and protein(Table 9–1). Finally, the malt is placed in largetrucks or train cars for large breweries or pack-aged in bags and shipped to smaller brewers.

Although the main purpose of the dryingand kilning steps is to arrest germination andpreserve enzyme activity, another importantseries of reactions also takes place. It is duringkilning that non-enzymatic browning and asso-ciated heat-generated flavor reactions occur,which is readily apparent in the malt and espe-

Box 9–2. Beer Terminology 101

The manufacture of beer, like that for wine making, bread making, and other food processingtechnologies, has evolved its own peculiar vocabulary to describe many of the manufacturingsteps.Thus, the brewing jargon contains words and phrases unique to beer making, but not sofamiliar to the casual reader. Given that the origins of modern beer manufacture arose in Ger-many, it should not be surprising that many of the terms are derived from German. In fact,muchof the overall beer vocabulary used by brewers even today contain German expressions. Beloware some of the more common terms used in brewing.

Coppers—Vessels used for wort-boiling, called coppers because of the construction materialused in their manufacture; even though these kettles are now made from stainless steel, theyare still referred to as coppers.

Diastase, diastatic power—Refers to the overall starch hydrolyzing activity present in themalt.

Fass—The process of pumping beer into kegs.Kiln—The oven used for drying and cooking malt.Krausen—The carbon dioxide layer that forms at the top of the fermentation tank; “high

krausen”refers to the period at which the krausen reaches its maximum.Krausening—The addition of krausen (containing very active, log phase cells) to beer to pro-

mote a secondary fermentation.Lagering—The act of storing or conditioning the beer at low temperature to promote matu-

ration.Lauter tun—A tank, containing a false bottom, used to promote clarification and separation of

the spent grains from the wort.Mash—The heated malt-water mixture, consisting of malt-derived enzymes and substrates.Mash off—At the end of mashing, when the temperature is raised to about 170°C to inactivate

enzymes.Mash tun—Where the mashing step takes place; may contain a false bottom and be used for

wort separation.Oast house—The facility used for drying whole hops.Pitch—The step when wort is inoculated with yeasts.Trub—The precipitated material obtained from the wort after boiling; it is rich in protein and

hop solids.



cially later in the finished beers. Dark beers(e.g., stouts, porters) use darker, more flavorfulmalt, whereas paler beers use lightly-colored,less roasted malt (Figure 9–4).

The flavors contributed by dark malts are of-ten described as coffee- or chocolate-like, withtoasty nutty notes. It would be a mistake, how-ever, to conclude that darker beers, because oftheir stronger, more pronounced flavor and

color, contain more ethanol than pale-coloredbeers. In fact, the darker the malt, the less amy-lolytic (or diastatic) activity is present. Ulti-mately, less sugar substrates are generated fromstarch, and less ethanol can theoretically beproduced from the mashes (see below) madeusing dark malts. For this reason, stouts andother dark beers are inevitably made with acombination of dark, highly-roasted malt, toprovide color and flavor,and lighter base malts,to supply enzymatic activity.

Although the malted and kilned barley con-tains starches, proteins, and the enzymes nec-essary for their degradation, these materials arenot readily extracted from the intact malt ker-nel. Rather, to facilitate extraction, the maltmust first be milled to break the kernel and ex-pose the interior portion to the aqueousmedium. Most breweries purchase whole orunground malt and do the milling themselves,according to their particular specifications.Thekilned malt is most commonly milled in rollermills or hammer mills that can be set to deliver

Table 9.1. Approximate composition of malt.

Component % dry weight

Starch 58Sucrose 4Reducing Sugars 4Hemicellulose 6Cellulose 5Lipid 2Protein 12Amino acids/peptides 1Minerals 2Other 6

Figure 9–4. Types of malt used in brewing.Shown (from left to right) are four types of malt,ranging from light to dark. These malts are usedto make beers with similar colors. Figure cour-tesy of Rich Chapin (Empyrean Ales, Lincoln, Nebraska) and John Rupnow (University of Nebraska).



products ranging from a very fine to verycoarse grind.

Prior to milling, the malt is surrounded by atough husk, which separates from the grainduring milling, but generally is only partiallydegraded.The fineness of the malt and the ex-tent of husk degradation are extremely impor-tant and depend on the preference of thebrewer.The finer the malt, the more easily en-zymes and other materials are subsequently ex-tracted during the next step, the mashingphase. However, the insoluble residue remain-ing after mashing is difficult to separate, due toits small particle size. In contrast, the spentmalt material is more easily separated from themixture if coarse malt, with minimum huskdamage, is used. In other words, extraction ofenzymes, starches, protein, nutrients, and fla-vors is greater with finely milled malts, but thefiltration rates are slow. Conversely, good filtra-tion rates but poor extraction occurs withcoarsely milled malts.

Finally,the ground malt or “grist”is ready to beused in the step called mashing.The purpose ofmashing is two-fold. First, the enzymes, starches,proteins,and other substances are extracted andsolubilized in a hot aqueous solution. Second,the extracted enzymes, starches, and proteins react to form products that ultimately serve as substrates to support growth of yeasts duringthe fermentation phase. The precise steps in-volved in mashing—in particular, the composi-tion of the water, the temperature, and the heat-ing rate—are critical factors, subject to thepreferences of the individual brewer.

Mashing begins when the malt is mixedwith brewing water (usually at ambient tem-perature) in a specialized tank called a “mashtun.” About 20% to 30% malt is usually added.Water quality has a profound influence onthe quality of the beer (it is, after all, the mainingredient), thus geographical considerationscan be very important.

The reputations of many beers are based, inpart, on the source, location, and properties ofthe water used to make the beer. For example,the famous Pilsener beers were made from thevery soft water in the Pilsen region of the

Czech Republic, whereas the great English-style pale ales developed in Burton-on-Trent inEngland, where the relatively hard water ishigh in minerals and bicarbonate.And althoughmany beer manufacturers continue even todayto make quality claims (and television com-mercials) based on the water from which thebeer is made, others argue that by making ad-justments for mineral content, pH, and ionicity,water variability is much less a factor in deter-mining beer quality.

In general, brewing water should havemedium hardness, containing about 100 ppmeach of calcium and magnesium salts and 50ppm or less of bicarbonate/carbonate.The pHshould be between 6 and 7.The specific func-tions of these components are described inTable 9–2.

The malt-water mixture, or mash, is gradu-ally heated from 20°C to 60°C to 65°C by infu-sion, decoction, or a combination of the two.Infusion heating is a simple step-wise processin which the mash is heated to an incrementaltemperature,held for a period of time, then thetemperature is raised again, held, and so on un-til the final temperature is reached.

In decoction heating, a portion of the mashis separated and boiled in a separate kettle,then added back to the main mash, therebyraising the combined mash temperature. Themain advantage of decoction mashing is an in-crease in starch extraction. Decoction mashingis used mainly for lager beers; in the UnitedStates a combination of infusion and decoctionheating is most common.

One might ask why the heating process dur-ing mashing is done in a particular order, ratherthan simply heating the mash to some giventemperature and then holding it there for apre-determined time. Remember that the pur-pose of mashing is to extract and react. In real-ity, there are many enzymes in the malt,not justa single amylase or a single proteinase. Like-wise, there are many sugar-containing poly-mers and starches with varying degree ofbranching, as well as heterogenous proteins,that will eventually serve as substrates. Pig-ments, flavors, and other substances are also



extracted during mashing.The important pointis that each of the enzymatic reactions haveoptimum temperatures that are likely differentfrom one another. By extracting the reactantsin step-wise fashion, it is possible to coordinatethe optimum or near-optimum temperature ofa particular enzyme with the appearance of itssubstrate. For example, a particular �-amylasemay hydrolyze its substrate at 45°C, but a dif-ferent amylase (perhaps a �-amylase) may havea temperature optimum of 60°C. If the mashwas heated directly to 60°C, then the �-amy-lase would not have had enough time at 45°Cto perform its job.

Malt enzymology

The carbohydrate fraction of malt is mostly inthe form of starch.Approximately one-fourth ofmalt starch is amylose, a linear polymer consist-ing of glucose, linked �-1, 4. The remainingstarch, three-fourths of the total, is amylopectin,which contains not only linear glucose,but alsoglucose in branched �-1, 6 linkages. The mainstarch-degrading enzymes synthesized duringmalting are �-amylase and �-amylase. The for-mer is an endoenzyme, acting primarily at in-

tramolecular �-1, 4 glucosidic bonds. Productsformed by �-amylase are dextrins (short chainglucose-containing �-1, 4 linear oligosaccha-rides) and limit dextrins (short chain glucose-containing �-1, 4 and �-1, 6 branched oligosac-charides). In contrast, �-amylase acts at the endor near the end of amylose and amylopectinchains. The main products are maltose andsmall branched dextrins.

Adjuncts

The mash, as described so far, contains only themalt and brewing water. The malt, as statedabove, supplies the enzymes and the enzymesubstrates. Many of the European beers, as wellas the micro-brewed beers produced in theUnited States, are made using 100% maltmashes. However, the most widely-consumedbeers in the United States, those made by thelarge breweries, contain an additional source ofstarch-containing material in the form of corn,rice flakes, or grits. These materials are calledadjuncts, and can account for as much as 60%of the total mash solids.Adjunct syrups, whichcontain sucrose, glucose, or hydrolyzed starch,are also commonly used.Adjuncts are allowed

Table 9.2. Components of brewing water and their function in beer1.

Concentration Component (ppm) Function

Calcium 10—250 When wort is heated, insoluble calcium phosphate complexes are formed,resulting in a decrease in pH during mashing and an increase in the activity of enzymes active at low pH (e.g., �-amylase). Calcium may alsoincrease thermal stability of �-amylase.

Magnesium 5—100 Required for activity and stability of enzymes produced by the yeasts.At high concentrations, magnesium may compete with calcium forphosphate, reducing calcium phosphate levels.

Bicarbonate 10—250 High bicarbonate concentrations increase the pH of the mash or wort.High pH then causes a decrease in the activity and stability of amylasesand proteinases.

Sulfate 5—500 Required for biosynthesis of sulfur-containing amino acids, which are essential for yeasts.At high concentrations, sulfates may be reduced tosulfur dioxide and hydrogen sulfide, which confer undesirable aromaand flavor characteristics in beer.

1Adapted from www.rpi.edu/dept/chem-eng/Biotech-Environ/beer/water2.htm



in most of the world, but their use is actuallyforbidden in some countries, including Ger-many, where centuries-old beer purity laws arestill enforced (Box 9–3).

Adjuncts have several functions. First, theydilute the strongly flavored, dark-colored,“heavy” characteristics of the malt. Althoughthese properties are preferred by some beerdrinkers, most U.S. consumers favor the palercolor and milder flavor associated with adjunct-containing beers. Second, adjuncts increase thecarbohydrate content and provides the amylaseenzymes with an additional source of substrate,and, ultimately, more fermentable sugar for the yeast.Third, they reduce the carbohydrate-to-protein ratio, such that less haze-formingproteins are present in the finished beer.Finally,adjuncts are less expensive than malt as asource of carbohydrate and, therefore, reducethe ingredient costs. In the United States, 70%of the adjuncts are corn or rice grits.The twobrands that dominate the U.S. market, Bud-weiser and Miller, both incorporate adjuncts intheir formulations.

It is important to recognize that adjunctsare not essential, since malt contains as muchas 70% soluble carbohydrate, which is morethan enough to satisfy the energy and carbonrequirement of the yeast. As noted above, ad-juncts lower the protein:carbohydrate ratio, soif too much adjunct is added, relative to malt(e.g.,�1:1 ratio), there may not be enough pro-tein for the yeast.When adjuncts are used, theyare usually added to brewing water in a sepa-rate tank (cereal cooker). The mixture isbrought to a boil to pre-gelatinize the starchand is then added to the mash tun.

Wort

Eventually, the temperature of the mash israised to 75°C, which effectively inactivatesnearly all enzymatic activity.The insoluble ma-terial in the mash liquid is then separated byone of several means (discussed below).This isan important step because the mash still con-tains grain solids and insoluble proteins, car-bohydrates, and other materials. In addition,

within the mash solids is a reasonable amountof soluble materials, including fermentable sug-ars, that would otherwise be discarded withthe spent grains.

For some breweries, the mash tun can alsoprovide filtration. These tanks contain a falsebottom that allows most of the mash liquid toflow out of the tank and into a collection ves-sel, while at the same time retaining a portionof the mash containing the spent grains. Thespent grains form a filter bed that enhancesflow rate (depending on the fineness of themalt).This material is then stirred and spargedwith hot liquid to extract as much of the solu-ble material as possible.The liquid is added tothat already removed from the mash tun.An al-ternative process involves pumping the mashinto a separate tank called a lauter tun, whichalso contains a false bottom and sparging sys-tem and operates much like the mash tun, ex-cept it provides greater surface area and fasterand more efficient filtration. Because the initialliquid material obtained from the lauter tun (orthe mash tun) often still contains solids, it maybe recycled back until a better filter bed is es-tablished and the expected clarity is achieved.

Finally, the mash separation step can be per-formed using plate and frame type filtrationsystems. These systems consist of a series ofconnected vertical plates,housing filters of var-ious composition and porosities.The mash liq-uid is pumped horizontally through the sys-tem.As for the mash and lauter tuns, the filtratecan be recycled and the entrapped materialcan be sparged to enhance extraction.

The liquid material or filtrate that is collectedat the end of the mash separation step is called“wort.” Since it is the wort that will be thegrowth medium for the yeast,and which will ul-timately become beer, its composition is veryimportant (Table 9–3).The main component ofwort (other than water) is the carbohydratefraction (90%).Most (75%) of the carbohydratesare in the form of small, fermentable sugars, in-cluding maltose, glucose, fructose, sucrose, andmaltotriose.The rest, about 25% of the total car-bohydrate fraction, are longer, nonfermentableoligosaccharides that include dextrins (�-1-4



Box 9–3. The Reinheitsgebot

Among the most ancient of all laws related to foods is the German Purity Law known as Rein-heitsgebot. These brewing laws were established in 1516 by Bavarian dukes Wilhelm IV andLudwig X in response to the frequent occurrences of what we would now call adulteration.During this time, brewers had begun to add questionable, if not dangerous, substances to beerin an effort to disguise defects and deceive consumers as to the quality. Brewers had beenknown to add tree bark,various grains,herbs,and spices to make the beer more palatable.Theseunscrupulous beer traders were giving the legitimate brewers of Bavaria a rather bad name.Thus, these laws were written to protect the brewing industry and perhaps to protect con-sumers. It is interesting to note that the ingredients clause (in bold) is but a part of the Rein-heitsgebot—most of the law deals with the price that brewers can charge for beer.

Below is an English translation of the Reinheitsgebot, taken from the article “History of Ger-man Brewing”by Karl J. Eden, published in Zymurgy,Vol. 16, No. 4 Special 1993.

“We hereby proclaim and decree,by Authority of our Province,that henceforth in the Duchyof Bavaria, in the country as well as in the cities and marketplaces, the following rules applyto the sale of beer: From Michaelmas to Georgi, the price for one Mass1 or one Kopf2, is notto exceed one Pfennig Munich value,and From Georgi to Michaelmas, the Mass shall not besold for more than two Pfennig3 of the same value, the Kopf not more than three Heller4. Ifthis not be adhered to, the punishment stated below shall be administered.

Should any person brew, or otherwise have, other beer than March beer, it is not to besold any higher than one Pfennig per Mass. Furthermore,we wish to emphasize that in fu-ture in all cities, markets and in the country, the only ingredients used for the brewing ofbeer must be Barley, Hops and Water. Whosoever knowingly disregards or transgressesupon this ordinance, shall be punished by the Court authorities’ confiscating such barrelsof beer, without fail. Should, however, an innkeeper in the country, city or markets buytwo or three pails of beer (containing 60 Mass) and sell it again to the common peasantry,he alone shall be permitted to charge one Heller more for the Mass of the Kopf, than men-tioned above. Furthermore, should there arise a scarcity and subsequent price increase ofthe barley (also considering that the times of harvest differ,due to location),We, the Bavar-ian Duchy, shall have the right to order curtailments for the good of all concerned.”

Note that, in contrast to the often-quoted statement that only four ingredients are permitted, theReinheitsgebot actually restricts beer making to just three ingredients: barley, hops, and water.This is because the fourth ingredient, the yeast, had not yet been “discovered.” Rather, earlybrewers either relied on a natural fermentation (i.e.,with wild yeast initiating the fermentation)or else used a portion of a previous batch to start the fermentation (i.e., backslopping).

The Reinheitsgebot is still in effect today in Germany, although it has been re-written and isnow more explicit.The current laws also have been modernized, such that hop extracts and fil-ter aids are now permitted. In addition, there is now a distinction made between lagers and ales,with exceptions regarding adjuncts and coloring agents allowed for the latter.

Notes

1. One Mass�one mug, or about a liter2. Kopf is a bowl-shaped container for fluids, not quite one Mass3.Two Pfennig�two pennies4. One Heller�one-half Pfennig



linked glucose molecules) and limit dextrins (�-1,4 and �-1,6 glucose molecules). A smallamount of �-glucans (�-1,4 and �-1,6 linked glu-cose molecules), derived from the cell walls ofthe barley,may also be present.

In addition to the carbohydrate fraction,wort also contains 3% to 6% nitrogenous mat-ter, including proteins, peptides, and freeamino acids. Proportionally less protein, rela-tive to the total solids, will be present if ad-juncts are used.About half of the amino acidswill eventually be used to support yeastgrowth, and the other half remaining in thewort contributes to flavor and browning reac-tions.Various ions (e.g., ionic calcium, magne-sium, and carbonates) are also present.The fi-nal pH of the wort is around 5.2.

The average molecular weight of the pro-tein fraction is very important, because it de-termines the “palate fullness” or mouth feel,and also affects the physical stability (cloudi-ness), color, and the foamability. In general, thegreater the protein hydrolysis, the less cloudythe beer. However, foam stability also is re-duced. In contrast, the less hydrolyzed the pro-tein, the more likely there will be cloudinessproblems.

Hops

In the final step prior to fermentation, the wortis pumped into a special heating tank calledthe brew kettle. It is here that the wort isboiled and other important reactions occur.Be-fore the wort is heated, however, one more es-sential beer ingredient, hops, is added to thewort. Hops are derived from the plant Homu-lus lupulus (in the family, Cannabinaceae), andalthough they were not part of the “original”beer formula, they have been added to beersince the Middle Ages. Why hops came to beused in beer making is not known,but it seemslikely that they were added as flavoring agents,then later additional benefits were realized.

Hops provide two main sensory character-istics to beer—flavor and aroma.A large num-ber of different hop cultivars or varieties arecommercially available, and they are distin-guished mainly on the basis of their aroma andflavor properties. By far, the predominant fla-vor is bitterness, which is due primarily to thepresence of �-acids, such as humulone and co-humulone, that are contained within the resinfraction. Although another group of relatedcompounds, the �-acids, are also found inhops and also contribute some bitterness, the�-acids are the more important group. In fact,brewers usually select hops based on their �-acid content (as a function of the whole hopweight), which can range from 5% (low bitter-ness) to as high as 14% (very bitter).While theintact �-acids do impart some bitterness, theyactually serve as precursors for isomerizationreactions (see below) that result in formationof the main iso-�-acids, isohumulone and iso-cohumulone, which are considered the realbittering compounds.

The amount of bitterness ultimately con-tributed by the hops is expressed in terms ofInternational Bitterness Units (IBU).One IBU isapproximately equal to 1 mg of iso-�-acids perliter of beer. Most U.S. lager beers have bitter-ness intensities of less than 15 IBU, comparedto nearly 50 for some of the ales produced inthe United Kingdom (hence, the name “bitters”for a typical British ale).

The other major component of hops is theessential oil fraction.This fraction,which is com-

Table 9.3. Approximate composition of mash, wort, andbeer (g/100 ml)1,2.

Component Mash3 Wort4 Beer

Water 75 88 92Carbohydrates 15 10 2.3

Starch 10 �0.1 �0.1Maltose 1 5 �0.1Maltotriose 0.5 1 0.2Dextrins 0.2 2 2.0Sucrose 0.5 0.3 �0.1Glucose 1 1 �0.1Fructose 0.5 0.2 �0.1Others 3 0.2 �0.1

Proteins � peptides 3 0.3 0.3Amino Acids 0.2 0.2 �0.1Lipid �0.1 �0.1 �0.1Ethanol �0.1 �0.1 3.5pH 5.6 5.4 4.2Specific Gravity (°Plato) �0.1 12 3

1Assumes an all-malt mash (no adjuncts)2From multiple sources3At start of mashing4After kettle boiling



prised of various terpenoids,esters,ketones,andother volatiles, is responsible for the aroma orbouquet properties of hops. Most of these sub-stances are volatile and, if added at the begin-ning of the kettle boil step,would mostly be lostby the end of the step.Therefore, when high-oilhops are used (when the brewer desires a beerwith a strong hoppy aroma), the hops are usu-

ally added near the end of the kettle boil step.Inaddition to providing flavor and aroma charac-teristics, hops also enhance preservation and increase shelf-life of beer. This is because the iso-�-acids formed during isomerization haveconsiderable antimicrobial activity and inhibitlactic acid and other bacteria capable of causingbeer spoilage (Box 9–4).

Box 9–4. Antimicrobial Activity of Hop �-acids and Mechanisms of Resistance

It has long been known that beer is less prone to microbial spoilage than many other aqueousbeverages.Thus, in some situations, beer has historically often been considered better preservedthan even water (e.g., the Pilgrims on the Mayflower). It is now known that the preservationproperties of beer are due to the presence of several components or constituents that are in-hibitory to a wide variety of microorganisms.Ethanol concentrations above 3% and pH levels be-low 4.6 are particularly inhibitory to spoilage organisms.Carbon dioxide, low oxygen levels, andlimiting nutrient concentrations also restrict growth of microorganisms.The hops are, perhaps,one of the more potent antimicrobial constituents in beer, and are, in large part, responsible forits long shelf-life. In particular,hops inhibit many of the lactic acid bacteria that spoil beer due totheir production of acids, diacetyl, and other products that confer flavor and aroma defects.

Although the antibacterial properties of hops, and the �-acid fraction, specifically, were recog-nized more than seventy-five years ago, the physiological means by which hop acids inhibitspoilage bacteria in beer has only been determined recently (Sakamoto et al.,2002,Simpson,1993).As shown in the model below (Figure 1A),the mechanism is actually similar to how other weak or-ganic acid preservatives, such as benzoic and propionic acids, inhibit microorganisms in foods.

Hop acids, like other weak acid preservatives,have ionophoric activity and are inhibitory dueto their ability to decrease intracellular pH and to disrupt proton gradients.Given the low pH ofbeer (�5) and their low pKa (3.1), isomerized iso-�-acids are in their undissociated or acid form(abbreviated as HA), and hence, are able to diffuse across the lipophilic cell membranes ofspoilage lactic acid bacteria. Since the cell cytoplasm of these bacteria is near-neutral, the iso-�-acids dissociate, releasing the anion (A�) and a free proton (H�).The latter then lowers the cy-toplasmic pH.The cell may respond by activating ATP-dependent pumps that efflux the accu-mulated protons, but the continued proton cycling is expensive (energy-wise) and also maylead to dissipation of the pH gradient portion of the proton motive force (the PMF or protongradient) that the cell uses to drive various transport systems.Thus, energy depletion, cytoplas-mic acidification, and reduced nutrient uptake all occur as a result of hop-derived �-acids andaccount for their inhibitory activity.

However, it is possible that a second mechanism also may be involved,because it appears thatthe dissociated form (A�) of the �-acids binds to cytoplasmic divalent cations, such as Mg2�,and the entire complex is effluxed from the cell.This results in the loss of a necessary nutrientthat must then be re-transported, with the concurrent cost of energy.

Despite the effectiveness of hop �-acids as antimicrobial agents, some bacteria have been iso-lated from spoiled beer that appear insensitive or resistant to hops. At least three resistancemechanisms have been identified (shown in Figure 1B as dashed lines). In Lactobacillus brevis,a common spoilage organism, resistance is mediated, in part, by two multiple drug resistance(MDR) systems. Both consist of membrane-integrated proteins.The HorA system is classified asan ATP-binding cassette (ABC) transport system, using ATP hydrolysis to drive efflux of iso-�-acids.The MDR system also pumps out iso-�-acids,but uses the PMF as the driving force.A thirdsystem relies on increased activity of the proton translocating ATPase (H�-ATPase) to pump outaccumulated protons.

(Continued)



Box 9–4. Antimicrobial Activity of Hop �-acids and Mechanisms of Resistance (Continued)

MDR

ATP ADP + Pi

A- + H+

AH

AH

AH

H+

ATP ADP + Pi

OUT

IN

membrane

H+

H+

H+-ATPase

B

HorA

AH

AH

AH

OUT

IN

membrane

H+

ATPADP + Pi

H+-ATPase

Mg2+

Mg AA

Mg AHAH

H+

A

H+H+ H+

H+ + A-

Figure 1. Inhibition of lactic acid bacteria by hops. In A, the means by which hop iso-�-acids inhibitsensitive organisms is shown. A model for hop resistance is shown in B. Adapted from Sakamoto et al.,2002 and Sakamoto and Konings, 2003.

References

Sakamoto, K., H.W. van Veen, H. Saito, H. Kobayashi, and W. N. Konings. 2002. Membrane-bound ATPase con-tributes to hop resistance of Lactobacillus brevis.Appl. Environ. Microbiol. 68:5374–5378.

Sakamoto,K., and W.N.Konings.2003.Beer spoilage and hop resistance. Int. J. Food Microbiol. 89:105–124.Simpson, W.J., 1993. Ionophoric action of trans-isohumulone on Lactobacillus brevis. J. Gen. Microbiol.

139:1041–1045.



Hops are usually obtained from specialtysuppliers in one of several forms, including liq-uid hop extracts,hop pellets,or as dried wholeflowers or cones. Although the latter are stillused, hop extracts and pellets have gainedpopularity due to their consistency and ease ofuse. Hop products can also be obtained as iso-merized extracts or pellets, containing iso-�-acids.This allows the brewer to add the hopsto the wort during kettle boil, as well as laterstages, so that the desired hops flavor can moreeasily be achieved.

Kettle Boil

After the hops are added to the wort, the mix-ture is boiled for one to one-and-a-half hours.Boiling accomplishes at least seven functions.First, it kills nearly all of the microorganisms re-maining after mashing, making the wort, for allpractical purposes, sterile. Second, boiling inac-tivates most of the enzymes still active aftermashing or reduces their activity to barely de-tectable levels.Third, the boiling step enhancesextraction of oils and resins from the hops andaccelerates isomerization of hop acids. Fourth,proteins,tannins,and other materials that wouldordinarily cause clarity and cloudiness problemsprecipitate during the boiling step. Removingthis precipitate (known as “hot break” or “hottrub”) helps to prevent such defects. Fifth, wortboiling enhances color development by catalyz-ing formation of Maillard reaction products.Sixth, undesirable volatile components, such assulfur-containing aroma compounds, are re-moved. Finally, prolonged boiling causes evapo-ration of water and concentration of the wort.Because the mash or wort is diluted by spargewater, evaporation returns the density or spe-cific gravity back to the desired level.

As noted above, a trub or precipitate formsduring the kettle boil step. In addition, the wortmay also contain insoluble hop debris, espe-cially if whole hops were used.The latter are re-moved via strainers or screen-type devices thatmay also allow for sparging and re-circulation.For hot trub removal, several separation sys-tems can be used.The simplest method is sedi-

mentation, but this is not a very efficientprocess. A much more efficient method is tocentrifuge the hot wort using continuous cen-trifuges. Such units,however, are expensive andrequire frequent maintenance. The most com-mon means of separating out the hot trub is thewhirlpool separator, which is typically a low-height, conical-shaped (cone pointing down-ward) tank whose geometry promoteswhirlpool-type movement of the wort.The trubor precipitate collects in the center of the tank.

After the trub is removed, the wort is cooledin plate-type heat exchangers. During the cool-ing stage, when the wort reaches about 50°C,more precipitated material forms (known asthe “cold break”), which is separated by cen-trifugation or filtration.When the wort temper-ature is reduced to 10°C to 15°C, it is pumpedinto fermentation vats, where it is, at last, readyto be made into beer.

The Beer Fermentation

All of the steps described above are performedto provide a suitable growth medium for theethanol-producing yeast. Before yeast is added,however, the wort is first aerated or spargedwith sterile air.Even though the beer fermenta-tion eventually becomes an anaerobic process,creating an aerobic environment at the outsetjump-starts the yeast and promotes a morerapid entry into a logarithmic growth phase.Importantly, oxygen also is necessary for bio-synthesis of cell membrane lipids that are es-sential for growth of yeast in wort.

Fermentors

The fermentation of the wort occurs in fer-mentor vessels of varying composition, sizeand configuration, and in either batch or con-tinuous modes.Although most modern fermen-tors are now constructed of stainless steel, thetraditional materials were wood, concrete, orcopper. Size and shape depend on a number ofconsiderations, but especially on whether thefermentation is top fermenting, as for ales, orbottom fermenting, as for lagers (see below).



For example, lager fermentation vessels aretypically cylindric, with cone-shaped bottoms(cylindroconical), so that when the yeasts floc-culate and settle, the cells collect within theconical region. Traditional fermentors, at leastfor ales, were also open or uncovered, so thatthe CO2 foam that developed during fermenta-tion provided the only protection against theelements (e.g., oxygen in the air, airborneyeasts and bacteria, and other contaminants).Thus, although the yeast can be skimmed fromthe top and re-used, the CO2 is lost.

Currently, enclosed, pressurized, cylindro-conical fermentors are used for lagers, as wellas ales, with capacities of nearly 106 L. En-closed fermentors obviously have the advan-tage of reducing exposure to air and airbornecontaminants; however, it is not possible inthese fermentors to recover the evolved CO2.Whereas open fermentors, whether round orrectangular, are necessarily more horizontal(i.e., width � height), enclosed fermentors areusually constructed in a vertical orientation,so less floor space is required. Regardless ofshape, however, most modern fermentors arejacketed to provide efficient cooling.

Inoculation

After the wort is cooled and aerated, the yeastculture is, at last, added to the wort, in a stepcalled pitching. Brewing yeast strains are dra-matically different from the Saccharomycescerevisiae strains used in the laboratory inmany important respects (Box 9–5). In addi-tion, there are two types or strains of brewingyeasts that are used, depending on the type ofbeer being produced, and these strains also dif-fer physiologically, biochemically, and geneti-cally. Many of these differences are highly rele-vant to the beer fermentation (Table 9–4).

Specifically,ale beers are made using selectedstrains of Saccharomyces cerevisiae, otherwiseknown as the “ale” or “top-fermenting” yeast.Lager beers are fermented by Saccharomycespastorianus (formerly called Saccharomycescarlsbergensis), also known as the “lager” or“bottom-fermenting” yeast. Despite this termi-

nology, growth of these yeasts in wort is notnecessarily confined to the top or bottom re-gions of the fermentor. Rather, top-fermentingyeasts, as they grow in the wort, tend to formlow density clumps or flocs that trap CO2 andrise to the surface.In contrast,when lager yeastsflocculate, the flocs sediment or settle to thebottom.The ability of brewing yeast to floccu-late and the point during the fermentation atwhich flocculation occurs are very importantfactors, as will be discussed later. It should alsobe noted that the distinction between top- andbottom-fermenting yeast is beginning to be lessrelevant, as the use of enclosed cylindroconicalfermentation vessels results in even ale yeastsdropping to the bottom of the fermentor (dis-cussed below).

Another major difference between ale andlager beer concerns the temperature at whichthe fermentation occurs. Ales are normally fer-mented at fairly high temperatures, from 18°Cto as high as 27°C, well within the range atwhich S. cerevisiae grows. It is no coincidence,therefore, that ales were more commonly pro-duced in warmer climates that are typical in theBritish Isles; hence, many of the classic English-style beers are ales. In contrast, lager style beersare fermented by yeast capable of growing attemperatures below 15°C. Not surprisingly,these beers evolved from Germany and othernorthern European areas, where the ambientclimates were cooler. Interestingly, when Ger-man immigrants moved to the United States inthe late 1800s and early 1900s, they broughtwith them German beer-making technology.The American beer industry subsequently be-came dominated by large breweries makinglager beers (e.g., Schlitz,Anheuser-Busch,Coors,Miller, Stroh,Heileman,Pabst, and others).

Although beer yeast cultures are commer-cially available, most breweries use their ownhouse cultures. Strains can be bred, much likeplants or seeds,to provide hybrids with specifictraits.These cultures are maintained and propa-gated in the laboratory and, when needed,grown in volumes necessary for inoculationinto the fermentation tanks. Some breweriesmay also use the yeast slurry left over from the



Box 9–5. Classification of Brewing Yeasts

Yeast classification can be confusing for microbiologists who are often more familiar with clas-sification systems used for bacteria. Part of the difficulty in understanding yeast taxonomy isdue to the significant biological differences between prokaryotic and eukaryotic organisms.Structurally,yeasts contain a nuclear membrane,whereas this membrane,by definition, is absentin prokaryotic bacteria.Although yeasts mostly exist, like bacteria, as unicellular organisms, theycan also develop mycelia and grow as multicellular organisms.

The means by which yeasts replicate provides another major difference.Yeasts can divide,like bacteria, by binary fission, but they can also reproduce by sporulation or by budding.Whereas bacteria contain a single chromosome and are monoploid,yeasts, in their diploid state,contain sixteen pairs of chromosomes.Actually, as discussed in Box 9-10, most brewing yeaststrains are polyploid, containing more than two chromosomal copies. Finally, in addition togenes located on the chromosome, yeasts contain mitochondrial DNA, and most strains of Sac-charomyces cerevisiae also contain multiple copies of a large, extrachromosomal elementcalled 2 �m DNA.

Although S. cerevisiae is among the most well-studied of all organisms, the laboratory strainsfor which genome sequences and other biological information are known are quite differentfrom the strains used to make beer (Table 1). For example, the life cycle of S. cerevisiae ordinar-ily includes a sexual or sporulation phase, with diploid cells undergoing meiosis and leading toformation of asci-containing spores (Figure 1).The haploid spores are of opposite mating types(a and �) and can mate, forming new diploid cells (a/�). Brewing strains of Saccharomyces, incontrast, rarely sporulate, and when they do, the spores are usually not viable nor are they ableto mate.Rather, strains used for brewing reproduce primarily by multilateral budding,with budsforming across the entire surface of the cell (i.e., in contrast to strictly polar budding).

Table 1. Differences between lab and brewing strains of Saccharomyces1.

Lab strains Brewing strains

Haploid and diploid Polyploid and aneuploidSporulating Sporulate poorlySpores viable Spores mostly non-viableAble to mate (a and � mating types) Mating rare

1Adapted from Dufour et al., 2003

Until relatively recently,yeast classification was based on much the same criteria used for fun-gal taxonomy (yeast, after all, are unicellular fungi and are grouped in the same family, Eu-mycetes). Thus, the main classification criteria were based on spore formation and the pres-ence/absence of ascospores. However, as noted above, many yeasts (especially brewing andwine strains) do not sporulate, and classifying yeasts on this basis was not very informative.

Instead, yeast classification is now based on morphology, biochemical, physiological, and ge-netic properties.Brewing yeasts, in contrast, share similar morphological, biochemical, and phys-iological properties and distinguishing between different strains is not easy (Table 2).This hasnot been a serious problem for brewers, who routinely classify yeast in a rather practical way asbelonging to one of two species. Ale, or top-fermenting yeasts, belong to the Saccharomycescerevisae group. Lager, or bottom-fermenting yeasts, belong to the Saccharomyces pastorianusgroup (formerly classified as Saccharomyces carlsbergensis, then as Saccharomyces uvarum).

When a more exact identification is necessary (i.e., to the strain level) or when the goal is todifferentiate between strains, genetic methods are the most reliable. DNA probes that hybridizespecific genes or regions can distinguish between ale and lager strains, but other methods haveproven to be easier and more powerful. For example, techniques based on the polymerase

(Continued)



chain reaction (PCR), restriction fragment length polymorphism (RFLP), amplified fragmentlength polymorphism (AFLP), pulsed field gel electophoresis (PFGE), and karyotyping all pro-vide unique genetic fingerprints that distinguish not only between ale and lager yeasts, but alsobetween strains of the same group.

Table 2. Characteristics of Saccharomyces species used in brewing1.

Fermentation Assimilation Fructose Assimilation Growth of: of: transport of: at:

Species Suc2 Raf Tre Suc Mal Raf Man Eth activity Lys Cad Ety 30°C 37°C

S. bayanus3 � � � � � � v � � � � � � �

S. cerevisiae � � � � � � � � � � � � � vS. exiguus � s � � � � � s u � � � � �

S. paradoxus4 � � � � � � � � � � � � � �

S. pastorianus5 v � � � � � � � � � � � � �

1Adapted from Vaughan-Martini and Martini, 19982Abbreviations: Suc, sucrose; Raf, raffinose;Tre, trehalose; Mal, maltose; Man, mannose; Eth, ethanol; Lys, lysine; Cad, cadaverine; Ety, ethyl-

amine; s, positive, but slow; v, variable; u, undetermined3synonyms: S.uvarum4synonyms: S. cerevisiae5synonyms: S. carlsbergensis

Box 9–5. Classification of Brewing Yeasts (Continued)

a

a

a/

a/

type and a spores

ascospore

germination

mating

zygote

mitosis

budding

sporulation

hybridization

a a

a

a

a

a

a

a

aa

a

a

a

a

a

a

a

Figure 1. Yeast life cycle. Most lab strains form ascospores containing spores of opposite mating types, aand �. These spores can mate, hybridize, and form diploidal cells capable of budding or sporulating. In con-trast, brewing yeasts generally only reproduce by budding.



previous fermentation to pitch the next batch.Only when the fermentation appears slow orsluggish or when quality attributes suffer will anew culture be prepared.

Yeast Metabolism

The usual starting inoculum gives an initialyeast population of abut 5 � 106 cells per mlof wort. Depending on the activity of the yeastinoculum, a lag period of six to eighteen hoursmay occur. Although no increase in cell num-ber is observed during the lag phase,metabolicactivity is well under way. The yeasts are syn-thesizing sugar and amino acid transport sys-tems, as well as enzymes necessary for theirmetabolism. As noted earlier, the wort issparged with oxygen prior to pitching, thusthe wort medium is initially highly aerobic.This oxygen-rich environment stimulates syn-thesis of membrane-associated sterols and un-saturated fatty acids. Sterols, and in particular,

ergosterol, are essential for formation of yeastcell membranes; likewise, the unsaturated fattyacids provide membrane fluidity. Because ofthese requirements, yeasts may grow poorly inoxygen-deprived worts.

Following the lag phase, cells enter into thelogarithmic growth phase, and a primary eth-anolic fermentation period begins. Despite theaerobic conditions initially established bysparging,sugar metabolism by yeast occurs notby aerobic respiration (i.e., the tricarboxylicacid or Kreb’s cycle), but rather via the Emb-den-Meyerhoff-Parnas (EMP) glycolytic path-way. Although Saccharomyces are facultativeanaerobes, enzymes of the Kreb’s cycle are noteven expressed during growth in wort.The re-pression of these genes under aerobic condi-tions occurs only for as long as the sugar con-centration is high. This phenomenon, calledthe glucose effect, is due to catabolite repres-sion by glucose. When the availability of glu-cose and other fermentable sugars is nearly

References

Dufour, J.-P., K. Verstrepen, and G. Derdelinckx. 2003. Brewing yeast, p. 347—388. In T. Boekhout and V.Robert (ed.), Yeasts in food.Woodhead Publishing Limited, Cambridge, England and CRC Press, BocaRaton, Florida.

Hammond, J.R.M., 2003.Yeast genetics, p. 67—112. In F.G. Priest and I. Campbell (ed.),Brewing Microbiol-ogy, 3rd ed. Kluwer Academic/Plenum Publishers. New York.

Kurtzman, C.P., and J.W. Fell. 1998. Definition, classification and nomenclature of the yeasts. P. 3—5. In C.P.Kutrzman and J.W. Fell (ed.), The Yeast: A Taxonomic Study, 4th Edition. Elsevier Science B.V. TheNetherlands.

Vaughan-Martini,A., and A. Martini. 1998. Saccharomyces Meyen ex Reess, 358–371. In C.P. Kurtzman andJ.W. Fell (ed.), The Yeast:A taxonomic study, 4th Edition. Elsevier Science B.V.The Netherlands.

Box 9–5. Classification of Brewing Yeasts (Continued)

Table 9.4. Differences between ale and lager yeasts.

Ale (Saccharomyces cerevisiae) Lager (Saccharomyces pastorianus)

Flocculated yeast rises to the top Flocculated yeast settles to the bottomOptimum growth temperature � 30°C Optimum growth temperature � 30°CMinimum growth temperature � 15°C Minimum growth temperature � 7°CMaximum growth temperature � 40°C Maximum growth temperature � 34°CCannot metabolize melibiose Able to metabolize melibioseSlow assimilation of maltotriose Efficient assimilation of maltotrioseSporulating (at low frequency) Non-sporulating



diminished (i.e., � 1 g/L), however, metabo-lism in an aerobic environment shifts from fer-mentation to respiration (Box 9–6).

Catabolite repression by glucose also affectsthe specific biochemical processes that occurduring the fermentation. Although brewingstrains of S. cerevisiae and S. pastorium canuse all of the major sugars normally present inwort (glucose, fructose, maltose, and mal-totriose), metabolism of these sugars does not

occur concurrently. In fact,wort contains moremaltose and maltotriose than glucose, yet glu-cose is still metabolized first.This preferentialuse of glucose occurs because: (1) at least oneof the glucose transporters is constitutively ex-pressed; and (2) the gene coding for the �-glucoside (maltose and maltotriose) trans-porter (AGT1) is repressed by glucose. Onlywhen the glucose is fermented (after aboutone to two days), therefore, will expression of

Box 9–6. Physiology of Brewing Yeasts

Given the importance of yeast to the brewing industry, it is not surprising that so much atten-tion has been devoted to understanding the physiological properties of Saccharomyces cere-visiae. Most brewing strains, as noted elsewhere, are very different from ordinary laboratorystrains.This is because the beer and the brewing environment in which these strains have beengrown for hundreds of years is quite unlike the docile conditions in the laboratory.Thus, mostbrewing strains have been screened and selected on the basis of their ability to grow well inwort and on their overall beer-making performance.Strains also have been modified by classicalbreeding programs such that specific traits are (or are not) expressed. More recently, the appli-cation of molecular techniques have made it possible to directly and specifically alter yeastphysiology by introducing genes that encode for specific properties.

Although the primary function of the yeast during brewing is to ferment sugars to ethanoland CO2, it is also responsible for performing many other important duties.The yeast must notonly produce adequate amounts of end products,but it must also metabolize all of the availablefermentable carbohydrate to fully attenuate the beer. The main end products produced bybrewing yeasts are ethanol and CO2; however, small amounts of higher alcohols, esters, aldehy-des, phenolics, and other organic compounds also are produced and may affect beer flavor andquality. Some of these compounds have a negative impact, and their production is discouraged.Of course, the ability of brewing yeast to perform specific metabolic functions depends on howwell it tolerates the high osmotic pressure,high ethanol concentrations,and other physiologicalstresses found in wort and beer.Thus, there are considerable physiological demands on yeastsduring the brewing operation.

When the brewer first adds yeast to the wort, the environment is highly aerobic, due to theoxygen-sparging step that occurs just after the wort is boiled and cooled.Thus, metabolism bythe facultative yeast might be expected to occur via a respiratory route (i.e., the tricarboxylicacid or TCA cycle). In fact, Pasteur demonstrated more than a century ago that when facultativeorganisms, such as brewers’ yeast, are placed in an aerobic environment, aerobic metabolism ispreferred (the so-called Pasteur effect).Aerobic metabolism,after all, is a more efficient process,yielding considerably more ATP than anaerobic metabolism (36 moles versus 2 moles ATP perglucose).

However, in the wort situation just described, yeast growth occurs via fermentative metabo-lism.Why would glycolysis occur in this environment? In part,because when the concentrationof available sugar (i.e., glucose) is high (�1% or 50 mM), the reverse of the Pasteur effect oc-curs. That is, even under aerobic conditions, provided there is excess glucose, expediency,rather than efficiency,dictates how metabolism will occur. In fact, the enzymes of the TCA cycleare actually repressed,a situation referred to as the “glucose effect.”Another reason,perhaps, forwhy the fermentation route proceeds under aerobic conditions is that brewing yeast strainsare, in general, highly accustomed to fermentative metabolism.



this gene, as well as the �-glucosidase genecoding for maltose and maltotriose hydrolysis,occur.

Sucrose, in contrast, is hydrolyzed by an ex-tracellular invertase, yielding its componentmonosaccharides, glucose and fructose. Theregulation of sugar metabolism becomes espe-cially relevant during the beer fermentationwhen high glucose adjuncts are used. Underthese circumstances, repression of maltose andmaltotriose metabolism may exist throughoutthe entire fermentation, and the beer will bepoorly attenuated.

Following transport, intracellular metabo-lism of wort sugars via the EMP pathwayyields mostly pyruvic acid and reducedNADH. Pyruvate is subsequently decarboxy-lated by pyruvate decarboxylase, forming CO2

and acetaldehyde. The latter is then reducedby NADH-dependent alcohol dehydrogenase,forming ethanol and re-oxidizing NAD. Be-cause a small amount of the glucose carbonmust be used to support cell growth, some ofthe pyruvate is diverted, via pyruvate dehydro-genase, to the Kreb’s cycle, where biosyn-thetic precursors are formed. Re-oxidation ofNAD then requires an alternative source ofelectron acceptors, leading to the formation of glycerol, higher alcohols, and other minorend-products found in beer.

The logarithmic growth phase period con-tinues for two to three days for ales, and up tosix or seven days for lagers, at which point allof the mono- and disaccharides, and most ofthe maltotriose, will have been fermented. Cellmass usually increases by less than two logs.

Importantly, the fermentation is exothermic,meaning that heat is generated, so the fermenta-tion tanks must be cooled and maintained.Either internal cooling coils or external jacketsare used to maintain temperatures of 8°C to15°C for lagers and 15°C to 22°C for ales.

For ale fermentations, yeast cells, as notedearlier, rise to the surface, along with theevolved CO2.The yeast can then be skimmed offfor re-use later. For lager fermentations, cells remain suspended, but there is enough CO2

evolved during the early stages of the fermenta-

tion to form cauliflower-shaped clumps. Thisthick foam layer is called krausen,and as growthand CO2 formation become more rapid, corre-sponding to the maximum growth rate of thecells, a period known as “high krausen” isreached. A portion of this material can be col-lected and re-used to initiate a secondary fer-mentation, as described below.

Although it is possible to monitor theprogress of the fermentation simply by observ-ing CO2 formation, the preferred means is tomeasure the specific gravity.As the fermentablesugars are consumed, the specific gravity, ex-pressed as °Plato, decreases.When the specificgravity no longer is decreasing, most of the fer-mentable sugars will have been depleted, andthe fermentation is complete.The beer is thenconsidered to be “fully attenuated.”At this point,usually four to seven days after the beginning ofthe primary fermentation period, the fermenta-tion vessel is quickly cooled to 4°C or less.How-ever, some brewers, prior to cooling, maintain ahigh temperature to promote use of diacetyland related vicinal diketones that are them-selves produced by yeast and which are respon-sible for off-flavors in beer (discussed later).Thisso-called “diacetyl rest” may reduce the neces-sary duration of the post-fermentation condi-tioning phase, when diacetyl reduction wouldordinarily occur.

Flocculation

Flocculation is the ability of yeast cells to ag-glomerate or adhere to one another in theform of clumps. When lager yeasts flocculate,the clumps have a density greater than that ofthe beer and settle to the bottom.Ale yeasts, incontrast, form clumps or flocs that entrap CO2

bubbles and have a lower density, and, there-fore, rise to the surface.

The ability of yeast cells to clump or floc-culate, and the time at which flocculation oc-curs, are very important properties in beermanufacture. In most cases, flocculationshould occur at the end of the fermentation,when all of the monosaccharides (glucoseand fructose), disaccharides (sucrose and



maltose), and trisaccharides (maltotriose) havebeen fermented. The yeast will have done itsjob, fermentable sugars are depleted, and thebeer is considered to be fully attenuated.

Importantly, the beer will be clear or“bright,” even in the absence of additional clar-ification steps. If, however, flocculation occursprematurely, before the end of the fermenta-tion, fermentable sugars will remain in thebeer (a situation brewers refer to as a “hangingfermentation”).These residual sugars can affectmaturation and flavor. In particular, these unat-tenuated beers have a lower than normalethanol concentration and are relatively sweetwhich, depending on the intent of the brewer,may or may not be desirable.

Of course, the presence of sugars in the beeralso provides growth substrates for other or-ganisms. In contrast, yeast cells that fail to floc-culate, and instead remain in the beer, are diffi-cult to remove, causing cloudiness problems.These yeasts may later autolyze in the beer, re-leasing enzymes that contribute to yeasty flavordefects.The onset of flocculation appears to betriggered by several factors, including entryinto stationery phase, low pH,low temperature,low sugar concentration, and high ethanol con-centrations. Hops and calcium (see below) alsopromote flocculation activity.

Despite the importance of flocculation in thebrewing process, the physical-chemical basis ofthis property is not well understood.Ale yeasts,in general, have been reported to be more hy-drophobic than lager yeasts, due to differencesin cell surface charges,but it is not clear that thisproperty affects flocculation. Rather, it now ap-pears that flocculation occurs as a result oflectin-like domains,contained within cell surfaceproteins, that bind in the presence of ionic cal-cium to mannans (mannose-containing chains)located on the surface of adjacent cells. Floccu-lation is also a heritable property,meaning it hasa genetic basis. Several genes have been identi-fied in S. cerevisiae that code for proteins in-volved in flocculation,and efforts are now underway to manipulate floc gene expression duringbeer making (Box 9–7).

Post-fermentation Steps

As noted earlier, beer manufacture, at firstglance, would appear to be a rather simpleexercise: barley is converted to malt, which isconverted to wort, which is then fermentedby yeast.The problem is that at this point thebeer still contains yeasts, other microorgan-isms, and other insoluble and non-dissolvedmaterials that give it an undesirable cloudyand hazy appearance. In addition, this so-called “green” or immature beer may containchemical constituents that impart off-flavors.Finally, little or none of the carbon dioxideproduced during the primary fermentation isretained in the beer, leaving the productwithout one of its key components.Thus, thechallenge for the brewer is to promote ac-ceptable flavor development and maturation,to remove the undesirable cloud- and haze-forming materials, and to introduce carbona-tion into the beer. Indeed, it can be arguedthat the post-fermentation steps are as impor-tant as any of the preceding activities.

Flocculation should occur as the fermenta-tion ends, or more precisely, when the fer-mentable sugars are depleted and the beer isfully attenuated. Once the yeasts have floccu-lated, they are usually collected promptly andremoved from the beer. It is important to recog-nize that, even though most of the yeast cellsare removed by the sedimentation or skimmingstep, the beer still contains yeast cells. In addi-tion, this green beer typically contains undesir-able flavor compounds, including sulfur diox-ide and diacetyl, as well as proteins and tanninsthat can potentially form haze complexes.Therefore, it is essential that the beer be “condi-tioned” to enhance sedimentation of the re-maining yeasts and haze-forming proteins, pro-mote dissipation of off-flavors, and producebeer with a mature or finished flavor. Sinceyeast growth also occurs during the condition-ing period, a secondary fermentation may takeplace, resulting in formation of CO2.Dependingon the specific conditions, enough CO2 may beproduced to naturally carbonate the beer.



Box 9–7. Flocculation—A Case of Beer Yeasts Sticking Together

The ability to flocculate, consistently and at the right time during the fermentation, is one of themost important traits of a good brewing yeast. If flocculation occurs too early or too late,beer fla-vor, clarity, and overall quality are bound to suffer.Yeast floccuation is now thought to be medi-ated via a lectin-like mechanism in which the N-terminal region of a cell wall-associated proteinrecognizes and binds, in a calcium-dependent manner, to mannose or trimannoside oligosaccha-rides located on the surface of other cells (Figure 1;Kobayashi et al., 1998; Javadekar et al., 2000;Verstrepen et al., 2003). Although both flocculating and non-flocculating cells contain this man-nan receptor, only flocculating cells produce the binding protein. Importantly, flocculation is in-fluenced by several physiological and environmental factors, including growth phase, ethanoland sugar concentrations, and calcium ion availability (Table 1).These observations suggest thatflocculation depends not only on physical interactions, but is also subject to genetic regulation.

Five genes (FLO1, FLO5, FLO9, FLO10, and FLO11) currently have been identified in Saccha-romyces cerevisiae that encode for proteins involved in adhesion (Halme et al., 2004). At leasttwo of these genes, FLO5 and FLO9, have high sequence homology to FLO1. However, several

(Continued)

Figure 1. Model for yeast flocculation. Floccu-lating cells produce cell wall-attached lectin-likeproteins that recognize and bind to mannan re-gions on the surface of receptor cells. The man-nan receptors are present on most cells; thus, itis the expression of the lectin-like proteins thatis responsible for the flocculation phenotype.Adapted from Verstrepen et al., 2003.

Table 1. Factors affecting flocculation.

Factor Effect

Fermentable sugars InhibitoryNitrogen, other nutrients Little effectTemperature Strain-dependent, with broad rangepH Optimal flocculation at pH 3.5—5.8Oxygen No direct effectEthanol Strain-dependent, affects cell surfaceCell age Old cells flocculate more than young cellsInoculum handling Ambient temperature increases flocculation



Conditioning

Conditioning can occur by one of severalmethods. In traditional ale manufacture, thebeer is pumped from the fermentor intowooden casks ranging in size from 20 L tomore than 200 L. Stainless steel casks have re-placed many of the wood casks, although thelatter are still available. Sugar is added to in-duce the secondary fermentation, and addi-tional hops and fining agents (see below) mayalso be added. The casks are held at 12°C to18°C for up to seven days. Following the matu-ration period, this cask-conditioned or drought

beer is ready for consumption, without any ad-ditional clarification steps (thus, the yeast sedi-ment is still present).

For lager beers, conditioning or lagering oc-curs in tanks held at lower temperatures for alonger time. Typically, temperatures as low as0°C, for as long as three months, can be used.Aspecial variation of the lager method involvesadding a portion of wort obtained at highkrausen (i.e., cells in their most active growthphase) to the green beer. Since the krausenalso contains wort sugars, this step, calledkrausening, essentially serves as a source ofboth fast-growing cells, as well as fermentable

different adhesion phenotpyes exist, including cell-to-cell adhesion (i.e. flocculation), as wellas adherence to other surfaces.

The expression of these dominant genes is complex and unstable, and appears to be con-trolled by both genetic and epigenetic mechanisms (Halme et al., 2004).Whereas FLO11 is or-dinarily expressed (in a lab strain), the other four genes are silent. Expression of FLO11 gives anon-flocculating (but adhesion positive) phenotype, whereas expression of the silent FLOgenes (via a heterologous promoter) results in flocculating adherence. Epigenetic silencing ofFLO11 also causes a morphological change in the appearance of the cells, from yeast cells in afilamentous form to yeasts that grow as discreet cells.

Depending on growth and environmental conditions, four different flocculation phenotypeshave been described for brewing strains of S. cerevisiae (Soares et al., 2004). In cells with aFLO1 phenotype, flocculation is inhibited by mannose. The NewFlo phenotype is inhibited by mannose and glucose.The MI phenotype is mannose insensitive. Finally, flocculation can beethanol-dependent. While flocculation by most FLO1 yeasts is constitutive, flocculation byNewFlo cells occurs during stationary phase.Thus, most lab strains have the FLO1 phenotypeand most brewing strains have the NewFlo phenotype (Javadekar et al., 2000).

Brewing yeasts ordinarily do not flocculate in the presence of glucose and other fermentablesugars.Furthermore,when fermentable sugars are added to flocculant cells, the flocs will disperseand subsequent flocculation is lost (Soares et al., 2004). It also appears that, under experimentalconditions, the rate of fermentation-dependent loss of flocculation correlates with growth rateand sugar use.That is, the faster the cells grow, the faster the flocs come apart.Despite these find-ings,however, it is not clear at a molecular level how the cells sense sugar availability and then re-spond by inducing or repressing flocculation. Recently, it was suggested that the onset of floccu-lation occurs as a result of the combined effects of nutrient shortage (of either fermentable sugarsor nitrogen) and the presence of ethanol (Sampermans et al., 2005).

Efforts to modify or improve flocculation properties of brewing yeast are being pursued ac-tively as part of current strain improvement programs. Originally, plasmid vectors were used todeliver flocculation genes (e.g., FLO1), but because expression was constitutive, these trans-formed strains had limited application for brewing. In one recent study, however, researcherssought to manipulate the onset of flocculation by using an inducible promoter (Verstrepen et al.,2001).A non-flocculating wild-type strain,harboring a genomic copy of the FLO1 gene (encoding

Box 9–7. Flocculation—A Case of Beer Yeasts Sticking Together (Continued)



substrates. Since the vats are closed to the at-mosphere, the CO2 that evolves is trapped, andthe beer becomes naturally carbonated.

Traditional conditioning of lagers and alesis time-consuming and expensive, and there-fore, used mainly for premium beers that canbear higher production costs. Most modernbreweries have adopted faster methods, re-ferred to as accelerated lagering or brewery-conditioning. In these systems, the fully atten-uated beer is pumped into storage tanks at0°C to 2°C and held for one week or less.Small wood chips (less than 3 cm � 30 cm)may be layered on the bottom of the tank to

promote maturation (e.g., “beechwood ag-ing”), as well as serve a clarification function(see below). During the fermentation and laterduring conditioning, many volatile flavor com-pounds are formed,which have a major impacton the overall quality of the beer (Table 9–5).

Clarification

Whether the beer has been conditioned viacasking or lagering, suspended yeast cells willstill be present.Because consumers have cometo expect beers to be haze- or cloud-free, withexcellent clarity, clarification and filtration

for the lectin-like protein, FLO1p), was used as the host organism.The expression of FLO1 wasmade inducible by replacing the FLO1 promoter with an inducible promoter derived from theHSP30 gene.

The latter is a heat shock gene that responds not only to heat, but also to nutrient depletionand high ethanol concentrations. Importantly, the HSP30 promoter (HSP30p) is also induced bystationary phase (i.e., a late-fermentation).Thus, expression of FLO1 behind HSP30p would beexpected to occur in the transformant only at the end of the fermentation. Indeed, this strategywas successful, as flocculation was evident in the transformants during stationary phase (atabout 95% of the wild-type level).As expected, heat shock or ethanol additional also inducedflocculation. The phenotype was stable for 250 generations, and when inoculated into wort,about 90% of the transformed cells sedimented at the end of the fermentation, indicating thatflocculation had occurred.

These results indicate that it is now possible to improve the flocculation properties of yeaststrains. It is noteworthy that the actual strategy used to modify gene expression in this organisminvolved a simple self-cloning or benign approach.Thus, the transformants contained no foreignDNA that might warrant legal or public scrutiny.

References

Halme,A., S. Bumgarner, C. Styles, and G.R. Fink. Genetic and epigenetic regulation of the FLO genefamilygenerates cell-surface variation in yeasts. Cell 116:405-415.

Javadekar,V.S., H. Sivaraman, S.R. Sainkar, and M.I. Khan. 2000.A mannose-binding protein from the cell sur-face of flocculent Saccharomyces cerevisiae (NCIM 3528): its role in flocculation.Yeast. 16:99–110.

Kobayashi,O.,N.Hayashi,R.Kuroki,and H.Sone.1998.Region of FlO1 proteins responsible for sugar recog-nition. J. Bacteriol. 180:6503–6510.

Sampermans, S., J. Mortier, and E.V. Soares. 2005. Flocculation onset in Saccharomyces cerevisiae: the roleof nutrients. J.Appl. Microbiol. 98:525–531.

Verstrepen, K.J., G. Derdelinckx, F.R. Delvaux, J. Winderickx, J.M. Thevelein, F.F. Bauer, and I.S. Pretorius.2001. Late fermentation expression of FLO1 in Saccharomyces cerevisiae. J. Am. Soc. Brew. Chem.59:69–76.

Verstrepen, K.J., G. Derdelinckx, H. Verachtert, and F.R. Delvaux. 2003. Yeast flocculation: what brewersshould know.App. Microbiol. Biotechnol. 61:197-205.

Box 9–7. Flocculation—A Case of Beer Yeasts Sticking Together (Continued)



steps are almost always performed. Clarifica-tion can occur via physical separation meth-ods, such as centrifugation, or by the additionof fining agents.

Fining agents act by promoting aggregationor flocculation of yeast cells, and include woodchips, gelatin, and isinglass.The latter, which isderived from fish bladders (really), is morecommonly used for ales, especially in the U.K.,whereas gelatin is frequently used in theUnited States for lager-style beers. The sus-pended cells and other insoluble debris thensettle to the bottom of the tank or cask. Theclarified,or bright,beer that remains is not cell-free. Therefore, the beer is filtered to removeresidual yeast or bacterial cells that still re-main.

Several filtration configurations exist, in-cluding plate and frame systems, leaf filters,and cartridge filters.The filters or filter sheetsdo not necessarily exclude cells on their own,that is, the pores within these filters are usuallynot so small that they directly restrict passageof cells.Yeast and bacterial cells have approxi-mate diameters of about 10 �m and 2 �m, re-spectively, and typical cellulose filters havemuch larger pore sizes.To remove cells by fil-tration, then, filtration aids are used.

Filtration aids are inert, insoluble materials,such as cellulose or diatomaceous earth, thatact to form a filter bed that traps cells withoutcausing the filter to become plugged orfouled with a layer of cells. Cell counts of lessthan 10 per ml can be achieved in the filteredbeer. Other filter aids, such as silica gel and

polyvinylpolypyrrolidone (PVPP), can also beused, but mainly to remove haze-forming pro-teins and polyphenolic materials, rather thancells (see below).

Despite the effectiveness of these bulk fil-tration systems, however, there are now manybreweries that further filter the beer throughmembrane filters with pore sizes less than 0.45�m. Since bacteria and yeasts cannot passthrough these filters, the beer becomes sterile.Many of the large breweries now produce fil-ter-sterilized beer. Because filtration is done atlow temperature, this process is also referredto as cold-pasteurized beer (which also has themarketing advantage of distinguishing thisprocess from heat-pasteurized beer).

If a sterile (or polish) filtration step is to beperformed, the beer must first be pre-filtered(as described above), to remove cells and othermaterial that would otherwise foul the mem-branes.Although most brewers recognize thatfiltration is necessary to produce beer with theexpected clarity and stability properties, theyalso realize that filtration may remove flavor,color, body characteristics, and other desirablecomponents. Thus, filtration systems must becarefully designed to achieve the desired ap-pearance without sacrificing flavor and body.

Process aids

Various approved additives are frequentlyused during the post-fermentation steps.Theseinclude agents to improve flavor, color, appear-ance, and stability. Hop extracts are sometimes

Table 9.5. Important volatile flavor compounds in beer.

Esters Alcohols Acids Carbonyls Sulfur compounds

isoamyl acetate propanol acetic acid diacetyl hydrogen sulfide2-phenylethyl acetate 2-methylpropanol caproic acid 2,3-pentenedione sulfiteethyl acetate 2-methylbutanol caprylc acid acetaldehyde methanethiolethyl caproate 3-methylbutanol capryc acid dimethyl sulfideethyl caprylate isoamyl alcohol lactic acid

amyl alcoholisobutanolethanol2-phenylethylethanol

Adapted from Dufour et al., 2003 and Hough et al., 1971



added late in the process to provide additionalhop flavor. Another important additive is pro-teolytic enzymes, which are used as “chill-proofing” agents. These enzymes hydrolyzeproteins that would otherwise precipitate andform complexes with tannins and otherpolyphenolic compounds at low temperaturesand give a cloudy or hazy appearance whenthe beer is chilled. Hydrolysis of these pro-teins prevents this “haze” or cloudiness.

Several commercially available types of en-zymes are used.Papain,obtained from papaya,is commonly used, as are fungal- and bacterial-derived proteinases. The main requirementsare that the enzyme be rather specific forhaze-forming proteins (and not the foam-forming proteins), and that they are inacti-vated by a subsequent heating step, so resid-ual activity in the finished beer is absent.Although widely used in the United States,chill-proofing enzymes are not allowed inmany European countries.

Non-enzymatic, chill-proofing agents arealso used, including tannic acid, bentonite, sil-ica gels, and PVPP. Again, the use of theseagents is permitted in some countries, and pro-hibited in others. In the United States, regu-lations regarding permitted additives andprocess aids are described in the U.S. Code ofFederal Regulations, Title 21, Part 173. Othergeneral regulations on beer manufacturing arecontained in Title 27, Part 25.

Carbonation

Of all the post-fermentation steps, perhapsnone is as important as carbonation. Carbona-tion provides sensual appeal by enhancingmouth feel,flavor,body,and foam (or head).TheCO2 preserves the beer by reducing pH and theoxidation-reduction potential (Eh), such thatvarious aerobic, acid-sensitive spoilage organ-isms are inhibited. Carbonation of beer can oc-cur naturally, via a secondary fermentation, ormechanically, by directly adding CO2 after theconditioning and filtration steps.

As described earlier, beer can be condi-tioned via several procedures. In the case of

traditional ales, conditioning is commonlydone in casks, whereas traditional lagers areconditioned, or lagered, in enclosed tanks.Some specialty beers (and those made byhome-brewers) are conditioned directly in bot-tles. In all of these processes, the beers becomenaturally carbonated due to a secondary fer-mentation that yields CO2 (as well as addi-tional ethanol).

Several requirements are necessary to in-duce a successful secondary fermentation.First, there must be sufficient fermentable sugarpresent in the beer to serve as substrate for theyeast.The cask, tank, or bottle must be able towithstand the pressure that accumulates as aresult of CO2 gas formation. Finally, either natu-rally present yeasts must remain in the beer,even after flocculation, or more yeast must beadded. For lagers, in particular, the krauseningtechnique can be used to initiate a secondaryfermentation. Recall that the krausen is wortobtained during the most active phase of thefermentation and that it contains sugars as wellas log phase yeasts.The CO2 bubbles that formby krausening are believed to be “finer” com-pared to those produced by other procedures.However, the krausening technique is now onlyinfrequently used.

The amount of CO2 that forms during a sec-ondary fermentation and is retained in the beerdepends on several factors.The amount of avail-able sugar present or added to the beer directlyinfluences CO2 production, but rarely is thesugar a limiting factor. Rather, the correctamount of carbonation is more a function of thecounter-pressure and temperature in the tank.Obviously, the more back pressure applied, thegreater the dissolved CO2 in the beer. Likewise,as the temperature decreases, solubility of CO2

increases. Proper adjustment of temperatureand back pressure, therefore, is critical toachieving the desired level of carbonation in thefinished beer.

Typically, ales have less carbonation thatlagers. In the United States, for example, lager-style beers contain about 2.6 volumes of CO2

per volume of beer (that is, if the dissolvedCO2 was completely dissipated and collected,



it would have a volume 2.6 times that of the“flat”beer).Traditional ales, in contrast, containabout half as much CO2.

It should be noted that beers that have un-dergone a secondary fermentation still need tobe clarified or filtered (or both) to remove cellsand haze-forming material, but without losingany of the accumulated CO2. Traditional cask-conditioned beer can be treated with finingagents to enhance sedimentation, leaving thesediment in the bottom of the cask (where itremains even during dispensing). In contrast,lagers or tank-conditioned beer must be pro-cessed under pressure to retain the CO2.

In the absence of a secondary fermentation,mechanical carbonation provides a convenientand more controllable means for introducingCO2 into the beer. Mechanical carbonation isnow widely used in the beer industry, not justby large breweries, but by small and even micro-breweries. There are essentially twosources of CO2. As noted earlier, the CO2 thatevolves during the primary fermentation canbe collected, cleaned and purified, and thenadded back to the beer.This process is not in-expensive, however.Therefore, the more com-mon means of mechanical carbonation is sim-ply to pump CO2 from pressurized tanksdirectly into the beer.As for when a secondaryfermentation occurs, the CO2 back pressureand temperature dictate the necessary volumeof CO2 pumped into the beer. In some cases,CO2 is even added to beer that has undergonea secondary fermentation to achieve the de-sired level of carbonation.

Packaging and pasteurization

Following carbonation, the beer is ready to bepackaged.Perhaps the simplest form of packag-ing is to fill the beer directly into kegs.Kegs areconstructed of aluminum or stainless steel andvary in size between 50 L and 100 L. Kegs notonly mimic the traditional cask-style beers (interms of perceived quality), but provide a con-venient means of delivering non-bottled orcanned product to the consumer.Thus,kegs arewidely used by bars and restaurants for servingdraught (otherwise known as draft) beer.

Beer destined for kegging is processed likeother beers, in that it is clarified and filtered.Most kegged beer is filter sterilized, althoughsome is heat pasteurized (see below). There-fore, its shelf-life is usually considerably longerthan traditional casked beer (about threemonths versus one month).Kegged beer that isneither filter- or heat-pasteurized has a shelf-life similar to that of casked beer.

Importantly, since kegs are re-usable con-tainers, breweries must pay particular atten-tion to the keg cleaning and sanitizing opera-tion before the kegs are re-filled.The interior ofthe keg, including the valve housing assemblywhere microorganisms may collect, is typicallysteam sterilized.And although it is outside thecontrol of the brewer, the dispensing systemused to pump beer from the keg to the glass isalso a source of potential spoilage organisms.

The most common package for beer in theUnited States is the bottle or can (sometimesreferred to as “small packs”).Bottles are usuallyconstructed of glass, although plastic (e.g.,polyethylene terephthalate or PET) bottleshave become popular. Most cans are madefrom aluminum. Beer that is to be bottled orcanned is pasteurized, either before or follow-ing packaging.

If the beer is pasteurized prior to filling, twooptions exist.The beer can be filter-sterilized,as described above, or it can be flash pasteur-ized via a plate-type, regenerative heat ex-changer (similar to that used for pasteurizationof milk). In fact, the time-temperature condi-tions for heat pasteurization of beer, 71°C to75°C for fifteen to thirty seconds, are very nearthat used for milk. In beer, these conditionsachieve a very high level of microbial killing(�5 logs), such that the product is stable atroom temperature for at least six months.

Whether the beer is filter-sterilized or flashpasteurized, the packaging steps must be per-formed under rather stringent aseptic condi-tions to prevent post-processing contamina-tion. Bottles (whether glass or plastic) andfilling and capping (i.e., crowning) equipmentmust be sterile, and surroundings protectedsuch that microorganisms are excluded. SterileN2 or CO2 can be used to flush the environ-



ment, and sterile water used to ensure that therelevant equipment,especially the areas aroundthe fillers, remain free of microbial contami-nants.The beer filling and packaging operationis essentially no different from that used formilk, juice, or other fluid products.

Beer can also be heat pasteurized after ithas been filled into cans or bottles. In fact,most beer consumed in the United States isprocessed via tunnel pasteurization systemsin which filled and sealed bottles or cans areheated by hot water. Tunnel pasteurizers op-erate in a continuous mode, with the temper-ature being gradually raised to as high as 62°Cand held for up to twenty minutes.The collec-tive effect of longer heating at lower tempera-ture has the same kinetic effect on the de-struction of microorganisms in beer asflash-pasteurization.

Pasteurizing the beer after it is already inthe sealed package prevents post-pasteuriza-tion contamination; however, the longer expo-sure to heat may promote undesirable flavorchanges in the beer, including formation ofcooked and oxidized flavors. Of course, somebeer manufacturers eschew any type of heattreatment and instead rely on filtration systemsfor preservation of beer. However, even filtra-tion is not benign,as flavor and other beer con-stituents may be removed during the filteringprocess.Thus, the challenge for modern brew-ers is to produce not only a beer with desirableflavor, body, and appearance characteristics,but also one that meets the preservation re-quirements necessary to operate successfullyin a highly competitive market.

Beer Defects

Despite the low pH, high ethanol content, andhop antimicrobials ordinarily present in beer,microorganisms are responsible for many (butcertainly not all) of the defects that occur inbeer. Chemical and physical defects are alsocommon and can cause significant problemsfor brewers. However, preventing or minimiz-ing entry and growth of microbial contami-nants throughout the beer-making process isabsolutely essential for consistent manufacture

of high quality beer. This is no simple matter,because fungi,wild yeasts, and bacteria are nat-urally present as part of the normal microfloraof the raw ingredients, the brewery environ-ment, and the brewing equipment. Moreover,even if a heat or filtration step is included atthe end of the beer-making process, the dam-age may have already been done.

Although the water activity of stored barley istoo low to support the growth of fungi, yeasts,and bacteria, those organisms begin to grow assoon as the barley is steeped, with increases ofabout 1 to 3 logs. Kilning eventually inactivatessome microorganisms; however, the finishedmalt may still contain as many as 106 bacteria,103 fungi,and 104 yeast per gram.Various speciesof lactic acid bacteria that account for the major-ity of bacteria found in malt and that may causeseveral serious defects in the finished beer areparticularly important (discussed below). How-ever,even fungi that are present at relatively lowlevels can cause problems. For example, somestrains of Fusarium can infect barley in the fieldand produce the toxin deoxynivalenol. Asidefrom its possible toxicity to humans, deoxyni-valenol is correlated with the “gushing”defect inpackaged beer. Gushing is characterized by ex-cessive gassing or over-foaming when packagedbeer is opened. Although the exact cause ofgushing is not clear, it appears that deoxyni-valenol or another mold-produced product (usu-ally from Fusarium, but other fungi may also beinvolved) serves as a seed or nucleation site, atwhich carbon dioxide microbubbles are formed.

Even though the beer fermentation requiresyeast, it is also true that wild yeasts can causeseveral different types of product defects. Wildyeasts, including species of Zygosaccharomyces,Kluyveromyces, and other Saccharomyces, maycompete with brewing yeast during the fermen-tation, and, in some cases, reach high levels.Since these yeasts typically do not flocculate asdo brewing yeasts,they remain in the beer,creat-ing cloudiness and filtration problems. Some ofthese yeasts also produce killer toxins that canpotentially inhibit the brewing yeast culture anddominate the fermentation. Other yeasts, re-ferred to as diastatic yeasts (of which Saccha-romyces cerevisiae var. diastaticus is the most



well-known), can hydrolyze and ferment dex-trins, and are particular post-fermentation prob-lem contaminants for two main reasons. First,they often produce phenolic off-flavors in thebeer, often characterized as medicinal. Second,by fermenting the body-forming dextrins, theymake the beer “thinner.” Finally, aerobic yeasts,which include members of the genera Pichia,Debaryomyces, and Brettanomyces, can pro-duce acetic acid,volatile esters and alcohols,andother oxidized compounds, and cause turbidityproblems.

Although high mashing temperatures (alongwith low pH) restrict growth of most bacteria,some heat-tolerant bacteria can survive andgrow during the mashing step. However, farmore serious are those bacteria that contami-nate the wort after the kettle boil step andgrow in the beer during the fermentation andpost-fermentation steps. Lactic acid bacteriarepresent the most bothersome organisms inthe brewing operation and are the cause of themost common defects. Of particular concernare species of Lactobacillus, including Lacto-bacillus brevis, Lactobacillus casei, Lacto-bacillus plantarum, and Lactobacillus del-brueckii.

During growth in wort, these bacteria fer-ment wort sugars and produce several unde-sirable end-products, especially lactic acid. Ofthese bacteria, L. brevis is heterofermenta-tive, meaning that it produces acetic acid,CO2, and ethanol, in addition to lactic acid.Some lactobacilli produce 3-hydroxypropi-onaldehyde, which serves as a precursor forsynthesis of acrolein, a compound that im-parts a bitter flavor. Importantly, species andstrains of Lactobacillus, as well as other lac-tic acid bacteria—most notably Pediococcusacidilactici, Pediococcus inopinatus, and Pe-diococcus damnosus—are responsible forperhaps the most objectionable defect inbeer, namely, the production of diacetyl.

Diacetyl is a 4-carbon molecule that impartsa buttery flavor and aroma that, while pleasantin cultured dairy products, is highly offensive inbeer. Pediococci (and P. damnosus, in particu-lar) produce appreciably more diacetyl than

other lactic acid bacteria and are perhaps themost serious microbial contaminants that affectbeer quality.They are capable of growth over awide range of temperatures, including thoseused during both the ale and lager fermenta-tions.These bacteria are also tolerant of low pH,ethanol, and,depending on strain,hop �-acids.

The diacetyl defect caused by these bacteria(referred to as “sarcina sickness” due to theoriginal classification of pediococci as belong-ing to the genus, Sarcina) is detectable at con-centrations as low as 0.2 ppm (or 0.00002%).In addition to flavor defects, some pediococciand lactobacilli are also capable of producingextracellular polysaccharides that cause a ropi-ness defect. This ropy or “slime” material in-creases beer viscosity and gives the beer an un-desirable mouth feel.

Although the lactic acid bacteria are arguablythe most serious spoilage organisms in beer,Gram negative bacteria also can contaminatebeer and cause spoilage problems. In fact, thespoilage organisms identified by Pasteur wereprobably species of the genus Acetobacter,Gram negative bacteria that oxidize ethanol toacetic acid. Strains of Acetobacter are used inthe manufacture of vinegar, thus, they give beera highly objectionable vinegary flavor. It is nowonder that the French beer industry had en-listed the technical advice of Pasteur to solvethis problem. Acetobacter and the relatedgenus, Gluconobacter, consist of aerobic rodsthat grow well in the presence of high ethanolconcentrations and are resistant to hop �-acids.Because these bacteria grow poorly, if at all, un-der the anaerobic conditions established duringthe beer fermentation, they are more likely tobe a problem only when the beer is exposed tooxygen, such as during storage of beer in casks.

Other Gram negative bacteria may also oc-casionally contaminate and spoil beer. In-cluded are Zymomonas, ethanol-tolerant, fac-ultative, or obligate anaerobes associated withale spoilage, and Citrobacter, Klebsiella, andother bacteria belonging to the family Enter-obacteriaceae.The former are prolific ethanolproducers that also make lesser amounts oflactic and acetic acid, acetaldehyde, and glyc-



erol. They may also produce sulfur-containingcompounds, such as dimethyl sulfide and di-methyl disulfide. As spoilage organisms, theycause a fruity flavor defect that occurs mostfrequently in casked beer.

Among the Enterobacteriaceae that arefound in beer, the most common is Obesum-bacterium proteus. However, its role in beerspoilage is not clear. It reportedly interfereswith normal growth of the yeasts, leading topoor attenuation.The anaerobic Gram negativerods include species from the genera Pectina-tus, Selenomonas, and Zymophilus. They arecapable of producing organic acids, H2S, andturbidity in packaged beer.They are easily killedby heat, thus they probably gain entrance intobeer during post-pasteurization packaging.

Beer spoilage does not occur only by mi-croorganisms; a number of chemical andphysical defects are also common in beer.One of the more frequent defects (althoughless so in recent years) is referred to as“skunky”beer flavor or light-struck flavor. It iscaused by sunlight-induced photooxidation.This defect, which can range from slight tonauseous, is most noticeable in beer bottledin clear glass. The most likely mechanism bywhich this defect occurs is via ultraviolet orsunlight-induced cleavage of side chains onhop-derived iso-�-acids. These reactive sidechains then react with sulfur-containing com-pounds to form products with a skunkyaroma. One thiol in particular, 3-methyl-2-butene-1-thiol or MBT, has a low flavor thresh-old (just a few ppb) and is characteristic ofthis defect. Given the potential for beer to be-come skunky, how is it that there are stillmany popular beers packaged in clear glass?In general, these beers are “skunk-proofed” toprevent the light-induced reactions describedabove. Specifically, hop extracts (obtained viaeither liquid or supercritical CO2 extraction)are used that are processed such that the iso-merized iso-�-acids are converted to a re-duced form. Depending on the extent of thereduction (i.e., how many hydrogen atoms areincorporated), dihydro-, tetrahydro-, or hexa-hydro-iso-�-acids are formed. Importantly,

these modified iso-�–acids are very light-stableand non-reactive.

Another flavor defect is referred to as staleor oxidized. Stale beer is variously described ashaving cardboard-like, rotten apple, cooked, ortoffee-like flavors. It is invariably caused by au-tooxidation reactions, not unlike the staling re-actions that occur in other foods.Staling is gen-erally time-dependent and increases duringstorage. Therefore, it is often the main factorthat determines shelf-life of beer.The best wayto control staling is simply to keep oxygen outof the finished beer.

Among the physical defects in beer,perhapsthe most common is haze, which, as discussedabove, occurs during cooling, and is caused primarily by proteins that form complexeswith polyphenols and carbohydrates. Treat-ment with proteinases can reduce this prob-lem. In addition, �-glucan-containing polysac-charides can also cause haze, gel, and filtrationproblems, as well as increasing viscosity. Bacte-rial �-glucanases can be added to the mash orwort to degrade these materials. Foam instabil-ity problems are also common and result inloss of foaming or head. Adding proteins, hopcomponents, and gums increases viscosity andsurface tension, and can therefore stabilizefoaming.

Waste Management in the Brewing Industry

Although waste management affects all seg-ments of the fermented foods industry, the is-sue is particularly important in the brewing in-dustry.This is because beer manufacturing usesnearly a half-billion tons of grains each yearand, although some components of that grainare fermented,much of the grain is left behind.These spent grains represent a considerabledisposal problem. Currently, this material isused in one of several ways. It can be furtherprocessed and used as a specialty food ingredi-ent (e.g., as a fiber supplement), however thismarket is way too small to make much of adent in the supply of spent grains. It is morecommon for breweries (small and large) to



contract with local farmers, who use the spentgrains as livestock feed. Spent grains can alsobe used as fertilizer.The grains can be dried toextend shelf-life and reduce shipping cost, butdrying requires energy and is expensive. Still,both wet and dry grains are used. Anotherwaste product that poses problems for the in-dustry is the spent filter material used as filteraids. Alternative filtration systems that do notrequire filter aids or that use materials that canbe recycled are now available.

Finally, brewing uses large volumes of wa-ter.As much as 5 L to 10 L of water are usedfor every liter of beer. Much of this water con-tains biological material, and its disposal intomunicipal water supplies may be restrictedand expensive.Therefore, water handling sys-

tems, such as anaerobic digesters that treatingand/or recycle this water are critical.

Recent Developments in the Beer Industry

The beer industry is one of the most competi-tive segments of the food and beverage indus-try.This competition has led to new technolo-gies, new innovations, and new products. Atthe same time, there has been remarkablegrowth in the microbrewing industry, and areturn to traditional or craft brewing prac-tices. It is now possible to find nearly everytype of beer at the local pub or retail outlet(Box 9–8). Thus, the beer industry, from thesmallest to the largest brewer, continues to

Box 9–8. Beer from A to Z

There are two general types of beer, ales and lagers. However, while it is certainly true that alesand lagers represent the major beer categories, this simple distinction is not really adequate todescribe the many varieties that are produced throughout the world. In reality, there are somany different versions of these two styles, that to refer to a particular beer simply as an ale ora lager provides little useful information about the actual nature of that beer. In addition, beersmade in a similar manner, but from different geographical regions, often have quite differentcharacteristics, and are classified accordingly.These differences in beer characteristics are due,in part, to the manner in which the beer is produced, but perhaps more importantly, to the spe-cific ingredients used in the particular recipe. For example, one of the key ingredients that dis-tinguishes one beer from another is the water used in the brewing process.

Of course, hops and malt have a major impact on the flavor, aroma, and color properties ofthe beer.Thus, beers are often classified based on their malt and hop type and content. Forexample, the Saaz hops used in Belgian pale ales and Czech Pilsner impart a flowery andfruity flavor to the beer. Listed below are some of the major categories or styles of beer andbrief descriptions for each.

Ale—Top-fermented beer,produced throughout the world,but most commonly in the U.K.Alesgenerally have more hop aroma and bitter flavor than lagers, and are usually less carbonated.

Altbier—A German style ale, historically made in Düsseldorf. It is copper-colored, with a mod-est bitter flavor, and only slight bitterness.Altbiers are lagered (and served) at lower temper-atures than typical ales.

Bitter—A typical British-style ale, with a strong bitter flavor.Bock—A German lager with strong malt flavor and sweetness, but little hop bitterness and little

or no hop aroma.The color ranges from light to dark brown,and these beers generally are full-bodied.

Brown Ale—An ale originated from Newcastle upon Tyne (in northeastern England), that ischaracterized by its full body, dark color, malty and generally sweet flavor, and low to moder-ate hop aroma and bitterness.



develop new and innovative products andprocesses, many of which involve biotechnol-ogy and bioprocessing.

Low-calorie beer

One of the most successful and influentialproducts developed by the beer industry was

low-calorie beer. Introduced in the 1970s, this“new” type of beer had a dramatic impact, notonly in the beer industry, but throughout thefood industry, as “light” became one of themost widely used descriptors for reduced calo-rie foods. In fact, low-calorie beers had beenaround for at least ten to twenty years prior tothe introduction of Miller Lite in 1975, but

Cream Ale—An ale made in the United States; it has a mild sweet flavor.Dunkel—The classic Munich lager;characterized by its dark brown color, sweet and nutty malt

flavor, and full body. Bitterness and hop aroma are usually mild.Lager—Bottom-fermented beer that originated in Bavaria, probably in the fifteenth century.

Lagers generally are less hoppy and are more carbonated than ales, although exceptions ex-ist. In the United States, lagers are amber or golden in color, with medium body and flavor.

Lambic—A specialty beer that originated in Belgium that is noted for several unique character-istics. First, it contains unmalted wheat as a major ingredient (as much as 30% or more). Sec-ond, aged hops, having decreased iso-�–acids, are used, minimizing the antimicrobial activityand bitter flavor.Third, microbial growth (including lactic acid bacteria, Enterobacteriaceae,and various yeast), and subsequent acid development is encouraged. Finally, the fermentationis entirely natural or spontaneous—no yeast is added, and it may take place over a period ofseveral months.Lambic beer variously described (even by its followers) as “sour, thin, leathery,fruity, cheesy.”

Malt Liquor—An American-style, pale colored lager, with low hop bitterness or aroma, thinbody, and little flavor, but with a higher-than-normal alcohol content (�4.5%)

Pale Ale—Typical British ale, with a medium body and malty flavor and light bronze color. Paleales are moderately dry and bitter, and have a hoppy aroma. Carbonation is generally lightand ethanol content high.Adjuncts are often added. India Pale Ale is similar in most respectsto conventional pale ale,except it contains more hops,which promotes shelf-life (hence, thisstyle was developed for export from England to India).

Pilsner—A classic lager that originated in Pilzen, a city in the Bohemia region of what is nowthe Czech Republic. Pilsners are among the most popular lagers in Germany (Bohemia wasonce part of the Austrian Empire and has a heavy German influence), as well as other regionsin Europe. Pilsner (or Pils) has a strong malt flavor and hop aroma, medium body and sweet-ness, and a golden color. German Pilsners generally are lighter in color and body, and are lesssweet compared to Czech Pilsner.

Porter—A British ale that has characteristics of both stouts and pale ales. Porters have light-to-medium body and malty flavor, are well-bittered, have dark color, and usually containadjuncts.

Steam beer—Introduced (and trademarked) by the Anchor Brewery Co. in San Francisco in thelate 1800s. Steam beers have both ale and lager properties, in that they are bottom fer-mented, but at a high ale-like temperature.They are fermented in wide, shallow tanks.

Stout—A type of ale that is noted for its very dark color (due to heavily roasted malt), rich maltand hop flavor, and moderate to high bitterness. Stouts have a creamy head, and vary fromsweet to dry.The Irish stout, Guinness, is the most well-known example.

Zwickelbier—A German beer that is unfiltered.

Box 9–8. Beer from A to Z (Continued)



these products were not very popular. Un-doubtedly, highly successful advertising andmarketing campaigns, along with more calorie-conscience consumers, led to what is now amajor part of the beer industry. Currently, low-calorie beers have about 45% of the total U.S.beer market, and three of the top five sellingbrands of all beer are in this category (Bud-weiser Light, Miller Lite, and Coors Light).When first introduced for beer, “light” had noofficial meaning. The widespread use of thisterm, however, eventually led the FDA to de-fine light foods as those that are significantlyreduced in either fat,calories,or sodium. As ap-plied to beer,“light” means that there must beone-third fewer calories.

The most obvious way to reduce the caloriccontent of beer would be to simply dilute regu-lar beer with water. A typical American-stylelager beer ordinarily contains about 150 calo-ries in a 340 ml (12 ounce) serving, so adding25% water would result in a beer with only 112calories. Doing so, however, would obviouslynot only reduce calories,but would also reduceflavor, color, body, and the ethanol content. In-stead,to make a “light”beer with fewer calories,but with a minimal loss of flavor and body char-acteristics, manufacturers had to consider beercomposition and the components that con-tributed calories.

In general, a typical U.S. beer contains (on aweight basis) about 3.7% ethanol, 1% protein,and 0.5% fat.Thus, in 340 ml of beer, the caloriccontribution of those components is about 90(from ethanol) � 15 (from protein) � 15 (fromfat) � 120 calories. What is the source of theadditional 30 calories (i.e., 150 to 120)?

As discussed earlier in this chapter, the beerfermentation is considered complete or fullyattenuated when the fermentable carbohy-drates have been depleted.However, that is notto say that there are not carbohydrates in beer.In fact, wort contains a mixture of fermentablesugars (i.e., glucose and maltose), as well asmore complex, non-fermentable sugars, in par-ticular, dextrins and limit dextrins. The latterare a result of the incomplete hydrolysis ofstarch during the mashing step. Whereas the

simple sugars are readily fermented, the com-plex carbohydrate fraction is not. Importantly,however, these carbohydrates are caloric,meaning they can be hydrolyzed to glucoseduring digestion, adsorbed into the bloodstream, and either used as energy or stored asfat. Calories can be reduced by somehow re-ducing or removing the nonfermentable butcaloric carbohydrates from beer.

The strategy that was adopted, and which isstill used today, adds enzymes that hydrolyze aportion of the nonfermentable carbohydratesto form free sugars that are then fermentedduring the primary fermentation step.The en-zymes are commercially available and inexpen-sive glucoamylases (derived from Aspergillusniger and other fungi).Added to the wort dur-ing mashing, the exogenous enzymes work inconcert with the endogenous enzymes fromthe malt. Ideally, the fungal enzymes should betemperature labile so there is no residual activ-ity after kettle boil.

Because more fermentable carbohydrate inthe wort necessarily results in more ethanol inthe beer (and at 7 calories per gram, would de-feat the purpose, as well as add to the alcoholexcise tax), the extra ethanol is ordinarily re-moved. The carbohydrate content of thesebeers is reduced from 9 g per serving to lessthan 3 g, resulting in a calorie reduction of 25or more.Thus, a typical light beer contains 100to 120 calories. Recently (in 2003), even lowercalorie beers, containing fewer carbohydratethan conventional light beers,were introduced.These so-called “low-carb beers”are targeted todieters on carbohydrate-restrictive diets. Onepopular brand contains only 2.6 g of carbohy-drate and 95 calories per 340 ml serving.

An alternate strategy to the use of enzymesreplaces even more of the malt or starch ad-juncts with adjuncts consisting of simple sug-ars. For example, if sucrose, fructose, or glu-cose syrups were used as adjuncts, at theexpense of malt, the dextrin fractions wouldalso be reduced. Both of these approaches,however, provide thin-bodied beers, since it isthe dextrins that contribute mouth feel andbody properties to beer.



Several other approaches for reducing thedextrin (and caloric) content of beer havebeen considered, and are now being studied.Malt preparations with greater amylolytic ac-tivity can be used so hydrolysis of starch ismore complete. Alternatively, brewers’ yeaststrains with greater amylolytic activity can bedeveloped,either via conventional yeast breed-ing programs or by molecular techniques.Thegoal is to increase activity of amyloglucosidase,which breaks down �-1,4 and �-1,6 glucosidiclinkages in dextrin.

Finally, there is a perception that low-calorieor low-carbohydrate beers also have a muchlower alcohol concentration than normal beer.In fact, the alcohol content for most of thesebeers is somewhat lower, generally by 0.5% to1.0% (or about 20% less). For example, three ofthe most popular brands of light beer in theUnited States each contain 4.2% alcohol (onvolume basis), compared to 4.7% to 5% fortheir conventional counterparts. For otherbrands, the differences are less than 0.5%

Ice and dry beers

Two other types of beer that gained a followingin the 1990s, but whose popularity has sincefaded somewhat, are ice beers and dry beers.Al-though made by different processes, both icebeer and dry beer contain more ethanol thanconventionally-processed beers and are thoughtto impart a smoother flavored beer with less af-tertaste. Ice beers are manufactured accordingto the freeze concentration principle—namely,by cooling the beer to temperatures as low as�4°C, ice crystals will form, which can then beremoved. The actual technology for the manu-facture of these beers is not new—a Germanversion called Eisbock has been made for manyyears.The resulting beer contains less water andmore ethanol, and is generally sweeter, withgood body and color. Dry beers, in contrast, areless sweet (hence the term “dry”), and are madeby using less malt and more readily fermentableadjuncts.These beers are rather similar to low-calorie beers, in that they have less flavor, color,and body compared to conventionally-made

products. They are also produced in a similarmanner, in that enzymes are used to hydrolyzedextrins. This results in a more complete fer-mentation, less residual sugars, and moreethanol.

Non-alcoholic beer

Low- or non-alcoholic beers were first pro-duced in the United States more than eightyyears ago (during Prohibition), and have beenavailable ever since. However, the relativelylow demand for these products did not drivethe industry to devote very much research ef-fort into new technologies.Due to a marked in-crease in the consumer demand for low- ornon-alcohol products, the technology for mak-ing these beers has improved dramatically inthe last decade.The quality of these beers, notsurprisingly, has also improved (although thereare still many detractors who would argue oth-erwise),as has the availability of many differentdomestic and imported products.

Non-alcoholic beers must contain less than0.5% ethanol. Historically, several differentprocesses have been used to produce non-alcoholic beer. The earliest methods generallyinvolved removing the ethanol from normalbeer by evaporation or distillation. Althoughthese methods are still used, various filtrationconfigurations, in particular, dialysis and re-verse osmosis, are now more widely used.Beerquality is arguably much better by the latterprocesses because they operate at low temper-atures (�10°C) and the heat-generated reac-tions that affect flavor do not occur.

In reverse osmosis systems, the beer ispumped or circulated at high pressure (morethan 5,000 Pascals) through small pore mem-branes with average molecular weight cut-offs of less than 0.01 �m. The ethanol mole-cules pass through the membrane as part ofthe permeate, along with a portion of the wa-ter. The retentate contains nearly all of thebeer solids and most of the water, but little ornone of the ethanol. Water can either beadded back to its normal level, or the beer can be diluted beforehand to account for the



water lost from the retentate. Dialysis oper-ates in a similar manner, except that the dri-ving force is simply the concentration differ-ence across the membrane. Despite the lessintrusive nature of these processes, however,flavor compounds may still be lost.

Another entirely different way to makethese beers relies not on ethanol removal, butrather on modifying the fermentation so thatethanol is not made as an end product.The fer-mentation can simply be abbreviated,such thatyeast growth is stopped or curtailed beforemuch ethanol has been produced. Alterna-tively, the fermentation can be conducted at alow temperature (�5°C) that restricts yeastgrowth and ethanol formation. However, thesemethods may result in a beer that is too sweetand microbiologically unstable,due to high lev-els of residual sugars. Modifying the wort com-position by removing fermentable sugars priorto fermentation is another way to limit ethanolproduction. Finally, metabolic engineering ofthe yeast, such that the ethanol pathway isblocked, may be an ideal way to reduce theethanol content in beer, but without leavingbehind fermentable sugars (Box 9–9).

Wheat beer

Wheat beers have long been popular in Eu-rope,but have only recently become known toU.S. consumers. They are made using wheatmalt combined with barley malt in ratios vary-ing from 1:3 to as high as 3:1.These beers aregenerally thinner and more sour than barleymalt beers, and are usually high in phenoliccompounds.They also often have a somewhatcloudy appearance. However, their unique fla-vor, due in part, to vinyl guaiacol generatedfrom ferulic acid, is particularly appreciated bysome consumers.

High-gravity fermentations

Another technology that has captured interestamong brewers, especially large, high-volumeproducers, involves developing a wort with aspecific gravity considerably higher than that

used ordinarily. The concentrated wort, oncemade, can be diluted before or, more com-monly, after fermentation.The net effect is anincrease in brewing capacity and productthroughput without having to incur additioncapital expenses for equipment (e.g., mashtanks, lauter tanks, and fermentors). Moreover,the production of high-gravity worts reducesenergy, labor, and effluent costs.Although high-gravity fermentations place extra demands on the yeast (e.g., the osmotic pressure andethanol concentrations are higher), the qualityof the beer is considered to be comparable toconventionally brewed beers.

Biotechnology and the BrewingIndustry

For the first 5,000 years that humans madeand consumed beer, little was known aboutthe actual scientific principles involved in itsmanufacture. Beer making was an art, prac-ticed by craftsmen. Only in the last 150 yearshave biochemists and microbiologists identi-fied the relevant organisms and metabolicpathways involved in the beer fermentation.In the past ten years alone, the entire genomeof Saccharomyces cerevisiae (albeit, a labstrain, not an actual brewing strain) has beensequenced, with nearly half of the genes nowhaving an assigned function.

This sequence information is now beingused to understand yeast physiology,especiallyas it relates to brewing (Box 9–10). However,despite this knowledge, and the thousands ofyears of “practice,” the brewing process is stillfar from perfect, and producing consistent,high-quality beer is still a challenge, even forlarge,highly sophisticated brewers. In addition,economic pressures, quality concerns, and perhaps most importantly, new market de-mands, have led the industry to consider newways to improve the beer-making process.Ad-vances in molecular genetics and biotechnol-ogy have made it possible to address thesechallenges via development of new brewingstrains with novel traits and tailored to per-form in a specific manner.



Box 9–9. Metabolic Engineering Approaches in the Manufacture of Non-alcoholic Beer

Two general processing approaches have been used to make non-alcoholic beer. One involvesremoving or separating ethanol from beer by physical means (e.g., distillation, evaporation, andreverse osmosis).The other approach has been to curtail the fermentation such that little or noethanol is produced. Both of these approaches have their pitfalls, as already described.

The availability of molecular tools has now made it possible to consider biological strategiesto produce non-alcoholic beer. Although the selection of brewers’ yeasts has been based, forhundreds of years,on their ability to ferment wort sugars to ethanol, recent studies have shownthat it is possible to redirect metabolism away from ethanol production and toward synthesis ofother end products.

During the beer fermentation, several end products, other than ethanol, are ordinarily pro-duced.Glycerol, in particular,can reach concentrations of more than 2 g per liter.Lesser but stillrelevant amounts of acetaldehyde, 2,3-butanediol, and acetoin are also formed.These productsare derived from the glycolytic intermediates,dihdyroxyacetone phosphate (DHAP) and glycer-aldehyde-3-phosphate (G-3-P), at the expense of ethanol (Figure 1).

Glycerol, for example, is formed from G-3-P via glyceraldehyde-3-phosphate dehydrogenase(GPD) and glycerol-3-phosphatase (GPP). By diverting even more of the glucose (or maltose)carbon to these alternative pathways, less ethanol would theoretically be produced. In the ap-proach adopted by Nevoigt et al., 2002, the GDP1 gene encoding for GPD was cloned and over-expressed in an industrial lager yeast strain.The transformants produced more than four timesthe amount of GPD, compared to the parent strain, and more than five times more glycerol. Im-portantly, the ethanol concentration in a typical brewing wort during a simulated beer fermen-tation was reduced by 18%, from 37 g/L to 30 g/L.This reduction, however, was less than halfthat achieved previously when GPD was over-expressed in a laboratory strain of Saccha-romyces cerevisiae (Nevoigt and Stahl, 1996).

There were also large increases in the concentrations of acetaldehyde and diacetyl during theprimary fermentation, and although these levels decreased during a subsequent secondary fer-mentation, they were still high enough to affect the flavor in a negative way.Thus, more meta-bolic fine-tuning of the competing pathways will be necessary to engineer a yeast capable ofproducing nonalcohol beers.

A completely different strategy for making non-alcoholic beer also was described by Navrátilet al. (2002).Their approach was based on the knowledge that: (1) non-alcoholic beer is sensi-tive to microbial spoilage; (2) low wort pH is inhibitory to contaminating microorganisms; and(3) addition of lactic acid or lactic acid bacteria to the wort stabilizes the beer. Because acidifi-cation of wort with lactic acid or lactic acid bacteria is either not allowed or is difficult to con-trol, another way to promote acidification was needed.

Therefore, strains of S. cerevisiae defective in enzymes of the tricarboxylic acid (TCA)pathway and known to produce elevated concentrations of organic acids were used undersimulated batch or continuous fermentation conditions (and using free or immobilized cells).In all cases, the beer pH was 3.25 or less when the mutant cells were used (compared to pH4.1 to 4.2 for the control strain).Although the mutations were not located within ethanol pro-duction genes, the test strains produced very low amounts of ethanol (�0.31% for free cellsand �0.24 for immobilized cells).The latter result presumably was due to the inhibition ofethanol formation, specifically, pyruvate decarboxylase and alcohol dehydrogenase, at lowpH. Although other end products, including diacetyl, were also produced, informal sensoryanalysis suggested that the beer compared favorably to conventionally-produced non-alcoholic beer.

(Continued)



References:

Navrátil, M., Z. Dömény, E. sturdík, D. smogrovicová, and P. Gemeiner. 2002. Production of non-alcoholicbeer using free and immobilized cells of Saccharomyces cerevisiae deficient in the tricarboxylic acidcycle. Biotechnol.Appl. Biochem. 35:133–140.

Nevoigt, E., R. Pilger, E. Mast-Gerlach, U. Schmidt, S. Freihammer, M. Eschenbrenner, L. Garbe, and U. Stahl.2002. Genetic engineering of brewing yeast to reduce the content of ethanol in beer. FEMS Yeast Res.2:225–232.

Nevoigt, E., and U. Stahl. 1996. Reduced pyruvate decarboxylase and increased glyceraldehyde-3-phos-phate dehydrogenase [NAD�] levels enhance glycerol production in Saccharomyces cerevisiae.Yeast12:1331–1337.

Box 9–9. Metabolic Engineering Approaches in the Manufacture of Non-alcoholic Beer (Continued)

glucose


glyceraldehyde-3-phosphate dihydroxyacetone phosphate

pyruvate glycerol-3-phosphate

acetaldehyde

GPD

glycerol

CO2

NADH + H

NAD

ethanolacetoin

NADH + H

NAD

2,3-butanediol

NADH + H

NAD

acetate

NADH + H

NAD

Pi

Figure 1. Formation of glycerol from glucose by Sac-charomyces cerevisiae. The glycolytic reactions from glu-cose to pyruvate are not shown. Once acetaldehyde isformed, end-products other than ethanol can be formed(acetoin, acetate, and 2,3-butanediol). Over-expressionof glycerol-3-phosphate dehydrogenase (GPD) results inglycerol production from dihydroxyacetone phosphate.



Strain improvement strategies

Since the beginning of beer making, all the wayto the present, brewing strains have been usedcontinuously,being passed down from batch tobatch.These strains are highly adapted to wortand beer and are not very amenable to classicalstrain improvement strategies.That is, trying toselect strains,either spontaneously or followingmutagenesis, with improved fermentative, fla-vor-producing, or other relevant properties, isnot an easy proposition. In addition, whereaslaboratory strains are usually diploid and capa-ble of sporulating and forming haploid spores

or asci, brewing strains are polyploid, contain-ing multiple alleles for many genes. Brewingstrains also sporulate poorly.Thus, even matingand hybridization techniques, the standardmeans for modifying yeast in the laboratory, areoften unsuccessful in brewing strains. More-over, changes in one trait often have pleitropiceffects and lead to undesirable changes in otherperformance characteristics.

Although strain improvement programs stillrely on these classical approaches, the abilityto target specific genes or traits is now possi-ble using recombinant DNA and genetic engi-neering techniques.However, such approaches

Box 9–10. Beer-omics

Of all the microorganisms used in food and beverage fermentations, Saccharomyces cerevisiaewas the first whose genome was sequenced (Goffeau et al., 1996). The sequenced strain(S288C) is considered a lab strain, as opposed to the industrial strains used for ale or lager man-ufacture.Although there is only limited sequence data for the latter strains, genome structureanalysis has revealed significant differences between the laboratory and brewing strains.

Most importantly, whereas most lab strain genomes exist as diploids (i.e., containing twocopies of each chromosome), beer yeasts are poly- or alloploidal, containing multiple chromo-somal copies (i.e., more than one “genome” per cell). Moreover, it appears that the genome ofthe lager yeast, Saccharomyces pastorianus (or Saccharomyces carlsbergensis), is a hybrid,consisting of one genome that evolved from an S288C-like strain of S. cerevisiae and anotherwhose ancestral progenitor is still unknown (Casaregola et al., 2001).Despite these differences,however,an incredible amount of practical information has been generated from the original se-quence data, as well as that subsequently made available via the web-based SaccharomycesGenome Database (SGD, located at www.yeastgenome.org).

The S. cerevisiae S288C genome consists of sixteen chromosomes and a total of 12,068 kilo-bases of DNA.Based on the original computer predictions, there were 5,885 genes encoding forproteins and another 455 that encode for ribosomal, small nuclear, and transfer RNA molecules.However, a more recent analysis of the S. cerevisiae genome indicates that there are probablyabout 500 fewer genes for a revised total of 5,538 (Kellis et al.,2003).Also, the genomes of threeother species of Saccharomyces (Saccharomyces paradoxus, Saccharomyces mikatae, andSaccharomyces bayanus), as well as the related yeast,Schizosaccharomyces pombe, have beensequenced, providing even more information on the structure and function of yeast genomes(Kellis et al., 2003;Wood et al., 2002). Numerous computational tools for comparing these se-quences with those obtained from industrial strains have been described and are now availablevia the SGD (Christie et al., 2004).

While genome information can predict the genes that are present in the chromosome(s) of anorganism,it is often more informative to identify genes that are actually transcribed.The so-calledtranscriptome can be determined using DNA macro- or microarray technology. It also may be in-teresting to determine how temporal gene expression is influenced by environmental conditionsor growth phase.For example, in the case of beer strains of S.cerevisiae,one might wish to iden-tify genes expressed during growth on maltose or that are induced when ethanol is present. Infact, it is possible to monitor gene expression throughout the entire beer fermentation process,

(Continued)



generating an expression profile of all the proteins produced at any given time during the courseof the fermentation (the proteome, commonly determined by two-dimensional gel elec-trophoresis).

Finally, the quantitative, global collection of small metabolic products (the metabolome) thatappears in the medium during beer manufacture (as measured by NMR, mass spectroscopy, orother analytical methods) provides yet another molecular picture of the beer making process.Several recent reports provide excellent examples of how “omic” approaches can be used togain a better understanding of the genetic, cellular, and metabolic events that occur during thebeer fermentation.

Beer transcriptomics

As reported in two independent investigations, one using a macroarray representing 6,084ORFs (Olesen et al., 2002) and the other using a microarray representing 6,300 ORFs (James etal., 2003), the gene expression patterns of brewing yeasts (lager strains) change during growthin wort. However, the number of differentially-expressed genes was relatively modest (about20% of the total ORFs). Furthermore, functions for the majority of the induced or repressedgenes could not be assigned.

It should also be noted that the mRNA used in these studies was from beer strains and thatthe arrays had been derived from a lab strain of S.cerevisiae.Thus, there may have been inducedor repressed genes that would not hybridize with any of the orfomers on the arrays.Still, in bothstudies, there were several interesting results. For example, there was a significant increase inthe expression of genes that encode for putative enzymes involved in sterol and lipid biosyn-thesis.This is consistent with the well-known observation that yeast cells synthesize sterols andmembrane unsaturated fatty acids following aeration of the wort just prior to the pitching step.Similarly, there was a marked increase in the expression of mitochondrial- and peroxisome-related genes, which may reflect the necessity of anabolic pathways to generate products re-quired for anaerobic growth.

Not surprisingly, genes encoding for glycolytic enzymes were also induced after inoculationof the yeast into the wort. However, the most unexpected result was the finding by James et al.(2003) that glycolytic genes were subsequently repressed after about forty-eight to seventy-twohours and respiration genes were activated. In addition, James et al., (2003) showed that proteinsynthesis and most stress response genes were repressed after the first day of fermentation, de-spite the increase in the ethanol concentration. Based on these collective reports, the lageryeast transcriptome clearly changes during the beer fermentation, reflecting the catabolic andanabolic adjustments the cell evidently makes in response to chemical and physical changes inthe growth environment.

Beer proteomics

The yeast transcriptome, as described above, provides a means of identifying the mRNA tran-scripts that are produced at a given time during growth. In contrast, the proteome representsthe complete set of actual functional gene products (i.e., the proteins) that are synthesized dur-ing growth. Proteome maps of industrial lager and other yeast strains during growth in syn-thetic medium revealed that about 1,200 polypeptides are produced, although many appear tobe duplications (Joubert et al., 2000).There is also a high degree of similarity between the pro-teins produced by lager strains and lab strains, supporting the notion that lager strains are hy-brids and contain a genome derived from S. cerevisiae. However, as many as thirty-two otherproteins are made by lager yeasts that do not appear in the S. cerevisiae proteome (Joubert etal., 2001).Analysis of these non-S. cerevisiae proteins by peptide mass fingerprinting and massspectroscopy techniques revealed that many are involved in maltose metabolism, glycolyticpathways, and production of ethanol.

Box 9–10. Beer-omics (Continued)



In another recent report, induction of protein synthesis by lager yeasts was studied duringthe lag and early log phase of growth (Brejning et al., 2005). Interestingly, the induced proteinexpression pattern differed from the gene expression patterns (at least for the genes investi-gated), indicating that post-transcriptional regulation was involved.Among the early expressedproteins were those involved in amino acid and protein synthesis, glycerol metabolism, glycoly-sis, and ergosterol biosynthesis. Moreover, the expression profile for cells grown in minimalmedium was consistent with those grown under brewing conditions.

Beer metabolomics

It would be understatement to suggest that wide variations exist in the chemical compositionof different beers.After all, differences in composition account, to a large extent, for the manydifferent types of beer that are produced around the world. Still, determining on a quantitativebasis what the actual chemical composition is for a given beer can provide considerable infor-mation regarding the brewing process, yeast metabolism,and overall beer quality.This chemicalprofile, or the metabolome, which can be determined by principal component analysis, hasbeen shown to distinguish between ales and lagers, as well as between different manufacturedbeers (Duarte et al., 2002; Duarte et al., 2004). For example, the presence or absence of particu-lar beer constituents containing aliphatic and aromatic regions (as determined by NMR spec-tra), were associated with ales or lagers.

References

Brejning, J.,N.Arneborg,and L. Jespersen.2005. Identification of genes and proteins induced during the lagand early exponential phase of growth of lager brewing yeasts. J.Appl. Microbiol. 98:261–271.

Casaregola, S., H.-V. Nguyen, G. Lapathitis,A. Kotyk, and C. Gaillardin. 2001.Analysis of the constitution ofthe beer yeast genome by PCR, sequencing and subtelomeric sequence hybridization. Int. J. Syst. Evol.Microbiol. 51:1607-1618.

Christie,K.R.,S.Weng,R.Balakrishnan,M.C.Costanzo,and 19 other authors.2004.Saccharomyces GenomeDatabase (SGD) provides tools to identify and analyze sequences from Saccharomyces cerevisiae andrelated sequences from other organisms. Nucl.Acids Res. 32 (Database issue):D311–D314.

Duarte, I.,A. Barros, P.S. Belton, R. Righelato, M. Spraul, E. Humpfer, and A.M. Gil. 2002. High-resolution nu-clear magnetic resonance spectroscopy and multivariate analysis for the characterization of beer. J.Agric. Food Chem. 50:2475–2481.

Duarte, I.F.,A. Barros, C.Almeida, M. Spraul, and A.M. Gil. 2004. Multivariate analysis of NMR and FTIR dataas a potential tool for the quality control of beer. J.Agric. Food Chem. 52:1031–1038.

Goffeau,A., B.G. Barrell, H. Bussey, R.W. Davis, B. Dujon, H. Feldmann, F. Galibert, J.D. Hoheisel, C. Jacq, M.Johnston, E.J. Louis, H.W. Mewes,Y. Murakami, P. Philippsen, H.Tettelin, and S.G. Oliver. 1996. Life with6000 genes. Science 274:546–567.

James,T.C., S. Campbell, D. Donnelly, and U. Bond. 2003.Transcription profile of brewery yeast under fer-mentation conditions. J.Appl. Microbiol. 94:432–448.

Joubert, R., P. Brignon, C. Lehmann, C. Monribot, F. Gendre, and H. Boucherie. 2000.Two-dimensional gelanalysis of the proteome of lager brewing yeast.Yeast 16:511–522.

Joubert, R., J.-M. Strub, S. Zugmeyer, D. Kobi, N. Carte,A.V. Dorsselaer, H. Boucherie, and L. Jaquet-Gutfreund.2001. Identification by mass spectrometry of two-dimensional gel electrophoresis-separated proteinsextracted from lager brewing yeast. Electrophoresis 22:2969–2982.

Kellis, M., N. Patterson, M. Endrizzi, B. Birren, and E.S. Lander. 2003. Sequencing and comparison of yeastspecies to identify genes and regulatory elements. Nature 423:241–254.

Olesen, K.,T. Felding, C. Gjermansen, and J. Hansen. 2002.The dynamics of the Saccharomyces carlbergen-sis brewing yeast transcriptome during a production-scale lager yeast fermentation. FEMS Yeast Res.2:563–573.

Wood,V., R. Gwilliam, M.-A. Rajandream, M. Lyne, and 130 other authors. 2002.The genome sequence ofSchizosaccharomyces pombe. Nature 415:871–880.

Box 9–10. Beer-omics (Continued)



must be carefully considered,due to the publicperception (often supported by regulatory au-thorities) that genetically modified organismspose risks to the consumer or the environ-ment.Rather, it seems that commercial applica-tions can still be realized, but will require theuse of more benign techniques (i.e., no foreignDNA used, no antibiotic resistance markers)for strain construction.

Examples of modified brewing strains

In the brewing industry, several traits have at-tracted the most research attention (Table 9–6).As noted earlier, brewing strains of S. cerevisiaegenerally do not hydrolyze dextrins and limitdextrins during the beer fermentation. Thesesugars remain in the beer, and although theymay have a positive influence on body andmouth feel characteristics, they also contributecalories. Reducing these dextrins provides thebasis for making low-calorie or light beers.Mostbrewers use commercially available enzymes,usually fungal glucoamylases, that are added tothe wort during mashing.Another approach isto use yeast strains that express glucoamylaseand that degrade these dextrins during the fer-mentation. Such strains have been isolated(e.g.,S.cerevisiae var.diastaticus),but,unfortu-nately, when used for brewing, the beer qualityis poor, due, in part, to the production of phe-

nolic off-flavors. Even hybrid strains, obtainedby mating brewing yeasts with diastatic yeasts,may still produce off-flavors.

Therefore, a better alternative directly intro-duces glucoamylase genes (encoding for bothglucose-1,4 and glucose-1,6 hydrolysis activi-ties) into brewing strains. Accordingly, STA2genes (from S. cerevisiae var. diastaticus) andGA genes (from Aspergillus niger) have beencloned into brewing strains, and the geneswere expressed and the enzymes functionedas expected (i.e., dextrins and limit dextrinswere used). Instability problems occurred ini-tially, since the cloned genes were first locatedon plasmids, but this problem has largely beencircumvented either by using more stable,yeast-derived plasmids or by using integrationvector systems that result in the cloned genesbeing integrated within the host chromosome.

Another problem that occurs in beer pro-duction is due to the �-glucans that are derivedfrom the cell walls of barley malt.Consisting of�-1,4 and �-1,6 linked glucose polymers, thesematerials form gels and foul filters, and in-crease beer viscosity and haze formation.Theycan be digested by commercially available en-zymes, but, like the glucoamylase example de-scribed above, genes coding for bacterial orfungal �-glucanases can be cloned into brew-ing strains such that they are expressed duringfermentation.

Table 9.6. Traits targeted for strain improvement.

Trait Relevant gene1 Function

Dextrin utilization STA2 GlucoamylaseDiacetyl utilization ALDC Acetolactate dehydrogenaseGlucan utilization EG1 �-glucanaseMaltose utilization MALT Maltose permeaseH2S reduction MET25 SulfhydrylaseDimethyl sulfide reduction mxr12 Sulfoxide reductaseEnhanced flocculation FLO1 FlocculationSO2 production MET3 ATP sulphurylaseEster production ATF1 Alcohol acetyltransferase

1Except where indicated, the goal is increased expression2Lowercase indicates the gene is inactivated and not expressed



BibliographyBamforth, C.W. 2004. Beer: Health and Nutrition.

Blackwell Science Ltd., Oxford, United Kingdom.Campbell, I. 2003. Microbiological aspects of brew-

ing, p. 1–17. In F.G. Priest and I. Campbell (ed.),Brewing Microbiology, 3rd Edition, Kluwer Academic/Plenum Publishers. New York.

De Keukeleire, D., 2000. Fundamentals of beer andhop chemistry. Química Nova 23:108–112.

Dufour, J.-P.,K.Verstrepen, and G.Derdelinckx.2003.Brewing yeasts, p. 347–388. In T. Boekhout and V.Robert (ed.), Yeasts in Food.Woodhead Publish-ing Limited, Cambridge, England and CRC Press,Boca Raton, Florida.

Flannigan, B. 2003. The microbiota of barley andmalt, p. 113–180. In F.G. Priest, and I. Campbell(ed.),Brewing Microbiology.3rd Edition,KluwerAcademic/Plenum Publishers. New York.

Goldammer,T. 1999. The Brewers’ Handbook. KVPPublishers, Clifton,Virginia.

Hammond, J.R.M., 2003.Yeast genetics, p. 67–112. InF.G. Priest, and I. Campbell (ed.), Brewing Microbiology. 3rd Edition, Kluwer Academic/Plenum Publishers. New York.

Hornsey, I.S. 1999. Brewing. Royal Society of Chem-istry. Cambridge, U.K.

Hough, J.S.,D.E.Briggs,and R.Stevens.1971.Maltingand Brewing Science. Chapman and Hall Ltd,London, U.K.

Stewart, G.G. 2004.The chemistry of beer instability.J. Chem. Ed. 81:963–968.

Van Vuuren, H.J.J., and F.G. Priest. 2003. Gram-nega-tive brewery bacteria, p. 219–245. In F.G. Priest,and I.Campbell (ed.),Brewing Microbiology.3rdEdition, Kluwer Academic/Plenum Publishers.New York.


10

Wine Fermentation

“I do like to think about the life of wine, how it is a living thing. I like to thinkabout the year the grapes were growing, how the sun was shining that summer orif it rained . . . what the weather was like. I think about all those people whotended and picked the grapes, and if it is an old wine, how many of them must bedead by now. I love how wine continues to evolve, how every time I open a bottleit’s going to taste different than if I opened it on any other day.Because a bottle ofwine is actually alive—it’s constantly evolving and gaining complexity—like your’61—and it begins a steady, inevitable decline.”

Maya, from the movie Sideways, screenplay by Alexander Payne and Jim Taylor

and true manufacturing practices are still im-portant, and in many cases, necessary to pro-duce high quality products.Thus, wine makingserves not only as an example of value-addedprocessing, but also as an example of an an-cient technology that has adopted twenty-firstcentury science.

History

The history of wine is nearly as old as the his-tory of human civilization.The earliest writingsdiscovered on the walls of ancient caves and inburied artifacts contain images of wine andwine-making instruments. Wine is mentionedmore than 100 times in both the Hebrew andChristian bibles and many of the most well-know passages involve wine. The very firstvines, for example, were planted by Noah, whopresumably was the first wine maker; later Jesus performed the miracle of turning waterinto wine.Wine also was an important part ofGreek and Roman mythology and is describedin the writings of Homer and Hippocrates. Forthousands of years, even through the presentday, wine has had great ritual significance in

Introduction

The conversion of raw food materials into fin-ished fermented products is often consideredto be one of the best examples of “value-added”processing. If this is the case, then per-haps no other process or product exemplifiesthis more than the fermentation of grapesinto wine. To wit, consider the following: InOctober, 2004, at a wine auction in Los Ange-les, a single bottle of Chateau d’Yquem 1847(a Sauterne wine from the Bordeaux region inFrance) sold for $71,675, making it the mostexpensive bottle of wine ever sold in theworld. At the same auction, two bottles ofCabernet from the Inglenook Winery in NapaValley sold for $24,675 each (which was arecord for a California wine). Although thesewines are certainly the exception, high qual-ity California wines still average nearly $30per bottle.

Given these examples, one may assume thatwine making must be a highly complicatedprocess, involving sophisticated technologies.In fact, although wine manufacture does in-deed rely heavily on modern microbiology andbiochemistry, traditional techniques and tried

349





many of the world’s major religions and cul-tures, and it is an important part of the worldeconomy and commerce.

Grape cultivation (viticulture) and winemaking appears to have begun in the ZagrosMountains and Caucasus region of Asia (northof Iran,east of Turkey).Domestication of grapesdates back to 6000 B.C.E., and large-scale pro-duction, based on archaeological evidence, ap-pears to have been established by 5400 B.C.E.A fermented wine-like beverage made fromhoney and fruit appears to have been producedin China around 7000 B.C.E., and rice-basedwines, similar to modern day sake, were pro-duced in Asia a few thousand years later (Chap-ter 12).Wines were imported into France, Italy,and other Mediterranean countries by sea-faring traders sometime around 1000 B.C.E.,and vines and viticulture techniques werelikely introduced into those regions severalcenturies later.

Wine is also one of the oldest of all fermentedproducts that has been commercialized, mass-produced, and studied. In fact, many of the earlymicrobiologists and chemists were concernedwith wine making and wine science. Less than150 years ago, when the very existence of mi-croorganisms was still being debated, Pasteurshowed that not only did microorganisms exist,but that they were responsible for both produc-tion and spoilage of wine. Of course, winepreservation has been important since ancientdays,when early Egyptian and Roman wine mak-ers began using sulfur dioxide (in the form ofburned sulfur fumes) as perhaps the first appli-cation of a true antimicrobial agent.

Production and Consumption

Most wines are made in temperate climates,particularly those areas near oceans or seas.About 75% of all wine is made in the Mediter-ranean areas of Europe (Table 10–1). France,Italy, and Spain are the largest producers, andare responsible for more than half of the nearly27 billion liters of wine produced from aroundthe world.Not surprisingly, these countries alsodevote the most acreage to grape production.

However, several relatively new entrants intothe global wine market have made a significantimpact. In particular,Australia, South Africa, andChile are now responsible for more than 10% ofworldwide wine production. Given that thesethree countries consume less than half the vol-ume of wine that they produce, they representa significant part of the export market. TheUnited States is also one of the leading produc-ers of wine (more than 2 billion liters per year),with most production coming from California(which alone accounts for about 90% of all U.S.wine). In fact, according to wine industry statis-tics (www.wineinstitute.org), the economic im-pact of the California wine industry is esti-mated to be $33 billion.

On a volume basis, the top five wine con-suming countries are France, Italy, the UnitedStates, Germany, and Spain (Table 10–1). On aper capita basis, consumers in Luxembourg,France,and Italy drink the most wine,more than50 liters per person per year (Figure 10–1).Thiscompares to the world per capita average ofabout 3.5 liters. In the United States, per capitaconsumption is about 9 liters. However, in con-trast to many of the European countries wherewine consumption has either been the same ordecreased in recent years,wine consumption inthe United States has increased (about 8% since1997). On average, most of the wine consumedin the United States costs only about $7 per bot-tle (0.75 L).

Wine Basics

Although wine making, like beer manufacture,involves an alcoholic fermentation, the similar-ities, for the most part, end there. The wine fermentation requires different yeasts and sub-strates and yields distinctly different products.And whereas beer is best consumed fresh,most wines improve markedly during an agingperiod that can last for many years. Finally, al-though most wines, like most beers, start withrelatively inexpensive raw materials, the qual-ity of premium wines depends, more so thanany other fermented product, on the quality ofthe raw material used in their manufacture.


Wine Fermentation 351

Although about 99% of the wine producedthroughout the world is made from grapes,juice from other fruits can also be made intowine. Berries, including raspberries, boysen-berries, and strawberries, are common sub-strates. Many wines also are made from treefruits, including apples and pears.The only re-quirement is that the fruit and the juice fromthat fruit must contain enough free sugars tosupport growth of the yeast and to yield a suf-ficient ethanol concentration (usually �12%by volume).

Viticulture and Grape Science

The starting material for most wines, as notedabove, is grapes.The main wine grape grown intemperate zones throughout the world is Vitis

vinifera.Another grape, Vitis labrusca, growswell in northern regions in the United Statesand is frequently used for Concord varieties. Itis important to note that, despite the existenceof only a few major grape species, there aremany different grape cultivars grown through-out the world. For example, Cabernet Sauvi-gnon, Chardonnay, Gamay, Mission, Gewurz-traminer, Grenache, and Sangiovese all refer todifferent varieties or cultivars of the V. viniferagrape. These grapes not only have differentcompositions, sugar contents, and pigmenta-tion, they also grow better in different climatesand soils and are used for different types ofwine. Thus, most Bordeaux wines (those pro-duced in the Bordeaux region of France) aremade from the grapes that grow well in that re-gion, namely Cabernet Sauvignon.Those same

Table 10.1. Global production and consumption of winea.

Acreage Production ConsumptionCountry(acres � 103) (L � 106) (L � 106)

Argentina 506 1,584 1,204Australia 366 1,016 398Austria 121 253 248Brazil 156 297 308Bulgaria 273 226 153Canada 22 45 280China 888 1,080 1,095Chile 440 566 225France 2,259 5,339 3,391Germany 257 889 2,004Greece 321 348 284Hungary 230 541 320Italy 2,244 5,009 3,050Japan 53 110 278Luxembourg 3 14 26Portugal 613 779 470Romania 610 509 471Russia 173 343 500South Africa 292 647 397Spain 3,052 3,050 1,383Switzerland 37 110 308Turkey 34 27 34United Kingdom 2 2 1,010United States 2,417 2,130 2,417Uruguay 27 100 98Totalb 19,586 26,683 22,785

aAdapted from 2001 statistics from the Wine Institute (www.wineinstitute.org)bTotal based on 69 countries



grapes, however, can be grown in California oranywhere else in the world, for that matter,andused to make Bordeaux-style or other types ofwine.

Although it is common throughout Europeto name a wine after a region, in other coun-tries,wines are ordinarily named after the grapevariety from which the wine was made. How abottle of wine is actually labeled is itself an im-portant issue. Labels not only must complywith local regulations, but they may also revealuseful information on the precise source of theraw materials,how the wine was made,and theexpected quality attributes of the finished pro-duct (Box 10–1).

Assuming that it is possible to successfullygrow Cabernet Sauvignon grapes, say, in cen-tral Minnesota, which is on the same generallatitude as central France, could a vintage bot-tle of Bordeaux be produced? Or to phrasethe question another way, what distinguisheswine made from the same grape but from dif-

ferent geographical locations? Several an-swers are evident. First, different areas havedifferent growing soils and climates. For ex-ample, in the famous grape-growing regionsin France, different districts within the sameregion may yield grapes with markedly differ-ent composition and properties due to differ-ences in soil composition, temperature, mois-ture, and sunlight. Even different vineyardswithin the same district may yield slightly dif-ferent grapes. In addition, grapes cultivated inthe same vineyard but harvested at differenttimes within the same season,or from two dif-ferent growing seasons,may be quite differentand yield distinctly different wines.

Climatic factors have such an importanteffect on grape quality and maturity that so-phisticated computer-generated meteorologi-cal models have been developed recently toallow viticulturists to predict how grapes willgrow and mature in a given geographical en-vironment.

31

21

9

29

57

2428

12

53

2

59

21

47

35

4237

16 17

9

30

0

10

20

30

40

50

60

70

Aus

tria

Aus

tralia

Can

ada

Den

mar

kFr

ance

eG

rman

yr

Gee

ceIre

land

Italy

Japa

nLu

xem

bour

gN

ethe

rland

sP

ortu

gal

Spa

inS

witz

erla

ndS

pain

Sw

eden

Uni

ted

Kin

gdom

Uni

ted

Sta

tes

Ura

guay

L

Figure 10–1. World-wide per capita wine consumption (L per person per year). Adapted from the WineInstitute 2001 Statistics (www.wineinstitute.org).



Box 10–1. Understanding Wine Labels

Few, if any, food labels reveal or imply as much information on the source of the raw materials,when and where the product was made, the manufacturing procedures used, and the overallquality of that food,as do the labels on wine bottles.Wine nomenclature can be so exact that in-formed and experienced consumers can even predict with some certainty how a particularwine should taste even before the cork is removed from the bottle.

Unfortunately, the lack of international conventions on wine standards and labeling has madeit difficult to develop a uniform and consistent set of rules and definitions. For example, Frenchwineries are subject to strict regulatory authority by virtue of Appellation d’Origine Contrôlée(AOC) laws, and, as a result, French wine labels can be extremely informative (at least for thosewho understand what the labels mean).Moreover,European Union labeling laws prevent use ofgeneric labels (i.e., naming a wine after a specific wine region, when the wine was producedfrom somewhere else).That is, just as cheese labeled as Parmesan must have been made in theParma region of Italy, so must a bottle of Champagne have been produced in the Champagne re-gion of France.Also, because authentic Champagne (and Parmesan cheese, for that matter) ismade using traditional methods, labeling a bottle of wine as Champagne necessarily tells theconsumer how that wine was made.

For the most part, no such rule exists in the United States, making comparisons betweenwines from different countries, based on labels, a somewhat challenging exercise.What does itmean, for example, for a California-made wine to be labeled as Bordeaux or Champagne? Whatfollows, then, is a general discussion on the terminology and “appellations” used for labelingwines.

In general, there are three types of information on a wine label that provide some indicationas to the nature or quality of that wine.Although different countries vary with respect to thespecific details (see below), most labels indicate the type of grape (varietal), where the grapeswere grown (the appellation), and the owner and/or bottler of the wine (proprietor).

Varietal refers simply to the cultivar (e.g., Chardonnay, Riesling, Cabernet Sauvignon) used tomake the wine.For many French wines, the variety is not stated,but can be inferred by virtue ofwhere it was made (Figure 1). In the United States (as well as most other countries), the varietalname can be used only when that variety comprises more than 75% of the grapes used in itsmanufacture.

Appellations are essentially geographical designations and refer to the location where thegrapes were grown, such as Champagne, Chablis, Chianti, Sherry, and Napa Valley. In France andother European countries, even small sub-regions and villages within the larger regions can bedesignated. In fact,protecting the geographical integrity of wine is one of the main reasons whylabeling laws evolved.Although no law prohibits an American wine manufacturer from adopt-ing a European geographical designation, this so-called generic labeling is merely a marketingdevice and provides no real information.This practice is now much less of a problem than itwas previously. To account for the geographical and varietal properties, most countries haveadopted a form of the French AOC labeling regulations.

In France,most wines fall into one of three main categories:Appellation d’Origine Contrôlée,Vins de Pays,and Vin de Table.Wines bearing an AOC declaration are those that are (1) producedin one of the designated regions, (2) comprised of specific cultivars, and (3) made according totraditional manufacturing practices. In theory, wines bearing the AOC designation are of thehighest quality.

Wines produced from grapes from different regions,outside the established AOC boundaries,are referred to as Vins de Pays (wine of the country).These wines are generally considered to beof somewhat lower quality, but in many cases, they are of similar quality as AOC wines. Finally,wines bearing the Vin de Table (or Vin Ordinaire) designation are of the lowest quality.The labelmakes no mention of the region or area where the wine was produced or of the grape variety.

(Continued)



Other than declaring the wine as “vin blanc,”“vin rouge,”or “vin rosé,” the label simply states theproprietor and the country.

In the United States, regulations regarding wine labeling are established by the Alcohol andTobacco Tax and Trade Bureau (TTB), a division within the Department of the Treasury (until2003, this division was called the Bureau of Alcohol,Tobacco, and Firearms). In addition to ahealth warning on alcoholic beverages and a sulfite warning, there are a minimum of five otherTTB requirements that must also be on the label: (1) the brand name or owner name; (2) the

Box 10–1. Understanding Wine Labels (Continued)

This wine is made in avillage called Pommard.Although not noted on thelabel (but certainly knownto informed consumers),Pommard is locatedwithin the Burgundy AOC.This village is part of Coted'Or, one of five regions inBurgundy (other well-known regions includeChablis and Beaujolais).Note that there is also nomention of the grapesused to make this wine.While varietal labels arecommon in the U.S., onlya few AOC wines(including Rieslings andGewurztraminers from theAlsace region) containthis information. Forother wines, the variety isimplicit in the AOC (e.g.,Pommard is made fromPinot Noir, Beaujolais isfrom Gamay).

The proprietor's namemust be listed on the label.For this bottle, thecomplete family name isgiven, a gesture meant toadd prestige to the wineand to instill confidence tothe customer. Also, it isnoted that the wine wasmade and bottled in thiswinery from grapes grownin this vineyard (i.e., thegrapes were not obtainedfrom another vineyard).Finally, in very small print(bottom center), there isa required code or "serialnumber" that indicates thelot or barrel from whichthe wine was obtained.

Another traditional way toclassify AOC wines is viathe "cru classé" system.This system is basedmostly on geographical,viticultural, and historicalconsiderations, and theactual quality of the wineis somewhat secondary.The highest class inBurgundy is Grand Cru(the terminology variesaccording to the region).This particular wine,although classified onelevel lower, as PremierCru, is still considered tobe of very high quality

The labeling regulations require that thealcohol content and the volume bedeclared on the label. In France, thealcohol level for AOC wines vary;Pommard wines must contain 13%alcohol (by volume). The bottle volumeis 750 ml, and the stamp at the upperright indicates that the producer has paidthe required recycling tax. Note thatneither sulfite, nor health warning labelsare required in France.

Wine labels should be informative, butalso distinctive and unique. A sketch ofthe vineyard or an image of the familycrest are common. Also, there is noyear stated on this label. This particularproducer indicates the vintage year on aseparate label. However, to include avintage year, at least 85% (higher insome regions) of the grapes must havebeen harvested from that year.

Not only does the labelindicate the village withinthe AOC region, theactual vineyard in whichthe grapes were grown isalso stated (VineyardDesignates). For this wine,the grapes were grown ina walled or enclosedvineyard called ClosMicault.

Figure 1. Understanding wine labels. The label in the above panel is from a French Burgundy made in the Cote d’Orregion. The label on the right panel is from a bottle of Cabernet Sauvignon, made in the Sonoma County region ofCalifornia. Labels were provided courtesy of the François-Xavier de Vaux and the Ernest and Julio Gallo wineries.



class or type;e.g.,Table Wine,Sparkling Wine,Carbonated Wine,Citrus Wine,etc., (3) the alcoholcontent, by volume; (4) the bottler’s name and address; and (5) the fluid volume, in metric mea-surements.

If the wine is made from grapes harvested and fermented during a particular year and from aparticular region, a vintage year may be indicated on the label. Similarly, an appellation or geo-graphical designation may be allowed, provided that more than 85% of the grapes were grownwithin the appellation boundaries.The boundaries for these appellations, called American Viti-culture Areas (AVA), are defined by the TTB. Most of the approved AVA designations are in Cali-fornia (examples include Napa Valley, Sonoma Coast, and Carmel Valley), but twenty-four otherstates also have AVA regions.

Finally, although some consumers are intimidated by wine labels (especially when they arenot written in English), the relevant information can be readily understood, provided oneknows a few simple basics (Figure 1).

Box 10–1. Understanding Wine Labels (Continued)

In contrast to most European wines, where appellation is emphasized, the grape variety is usually a moreprominent feature of U.S.-made wines. To use a varietal label, the wine must have been made from at least 75%of that variety. Although varietal labeling is optional, labels must indicate the type or class of wine. Thisparticular wine is a Class 1 wine, as signified by the “Red Wine” declaration. U.S. federal regulations recognize9 classes; other examples include sparkling wines (Class 2), fruit wines (class 5), and apertif wines (Class 7).

The vintage year (e.g., 2000) can alsobe stated, provided at least 95% of thegrapes were grown and harvested inthat calendar year.

In addition to statingthe wine type or class,other requirements forall U.S. wines includethe proprietor nameand address; thealcohol content (if 14%or higher), and themetric volume. If thewine was fermented

Many U.S. wines include optionalinformation regarding the region wherethe grapes were grown and details onthe fermentation and aging conditions.

The U.S. version ofAOC regions are calledAmerican ViticulturalAreas. The BarrelliCreek Vineyard, wherethis wine is made, islocated in AlexanderValley, one of 11 AVAsin Sonoma County,California. The notationat the top of the label “single vineyardselection” signifies that the grapescame from one vineyard located withinthis region, comparable in meaning tothe enclosed vineyards of Burgundy.

and bottled by the same producer, aphrase such as the one used here“vinted and bottled by" can be used.Otherwise, the producer name would beprefaced by the statement “bottled by."

There are two other label requirements for U.S. wines (although not shown here): the “contains sulfites”declaration and the Surgeon General’s warning about alcohol consumption and risks to pregnant women andoperators of cars and machinery.

Figure 1. (Continued)



Vineyard management practices also affectgrape maturation. Important factors includevine spacing and density, pruning and thin-ning, training and trellising, use of canopies,and application of pesticides. Although thecombined effects of soil, climate, and seasoncertainly affect grape quality, the overall im-portance of these factors, the so-called “ter-

roir,” on wine quality, is frequently debated(Box 10–2). Certainly, wines made from pre-mium grapes can sometimes be very ordinary,indicating that attention to proper wine-making techniques is also critical. On theother hand, high quality wines can hardly beproduced from inferior or immature grapes.The bottom line is that the basic raw material

Box 10–2. The Terroir Theory

No other issue divides European and American wine makers as much as the issue of “terroir.” Ter-roir, translated loosely from the French as “soil” or “earth,” refers, at its most simple level, to theground or location in which wine grapes are grown. In France, the entire appellation control sys-tem that forms the basis for how French wines are labeled is based on terroir.Thus, each of theten AOC regions has its own terroir (e.g., the Bordeaux terroir, the Burgundy terroir, and so on).

In the broader sense, however, terroir theory holds that wine quality, apart from the actualmanufacturing steps, is a function of a host of environmental and agronomic factors. Relevantfactors include the nutrient composition of the soil in which the grapes were grown,the cultivarof the grapes, the topography of the vineyard,watering conditions,climatic conditions,exposureto sunlight, vine density, and training and pruning practices. Some wine makers might suggestthat the specific yeasts that are present on the grapes also contribute to terroir (Box 10–3).

On the surface, one might wonder why controversy regarding terroir exists. Is it not unrea-sonable to expect that wine quality might be influenced by where the grapes were grown? Infact, it is well established that sugar concentration, phenol content, other grape constituentsand overall grape maturity are affected by climate, sunlight, and other growing conditions.Thus,on a general level, there is probably not much argument—environmental conditions have a pro-found effect on grape composition which ultimately affects wine quality.

Why the issue of terroir raises concern among enologists, it seems, relates to two other is-sues. First, there is the view held by terroir advocates that the soil itself contributes to wine fla-vor by somehow translocating minerals and other soil molecules directly into the grapes.Thus,the wine has the “goût de terroir or “taste of the soil.”This claim is generally disputed by manyviticulturists, who would argue that grape vines and other plants adsorb water, minerals, nutri-ents, and other inorganic and organic molecules from the soil, but not flavor compounds.

The second issue is more philosophical in nature. Terroir, according to the popular winewriter, Jamie Goode, is considered by many to be an “ethos...a unifying theory encapsulating acertain approach to wine that encompasses the almost metaphysical circle of soil, nature, ap-pellation, and human activity.” In other words, it is terroir, and not the skills of the wine-maker,that is responsible for the overall sensory characteristics of a wine. Ultimately, the argumentgoes, terroir makes a particular wine unique.

By definition, therefore, it would not be possible to replicate a particular wine produced fromgrapes from one region with a wine made from the same cultivar, but grown even a few metersaway. Further, and this is probably what really annoys “new world” wine makers, terroir impliesor reinforces the perception that “old world” (and French, in particular) wines are superior tothose produced elsewhere.Despite whether or not there actually is a scientific basis for terroir,there are clearly considerable market advantages for French wines bearing AOC designations(which are essentially synonymous with terroir).And even though wines produced in the NapaValley of California or the Finger Lakes region of New York, or anywhere else for that matter,could also make terroir claims, there is little question that it is difficult to match the history andreputation of Bordeaux, Loire Valley, Champagne, or any of the other terroir regions.


The Wine Fermentation 357

for wine—the grape—is prone to consider-able variability, which may have profound ef-fects on the physical, chemical, sensory, andother properties of the finished wine.

Grape Composition

Given that the two major constituents of wine,water and ethanol, have no flavor, color, oraroma, it is not surprising that the other grapecomponents contribute so much to the organ-oleptic properties of wine. Some of these sub-stances can be problematic, causing a varietyof defects. In addition, the composition ofgrapes changes during growth and maturationon the vine, such that the time of harvest influ-ences the chemical constituents of the grapeas well as the wine.For example, the sugar con-centration increases as the grape ripens on thevine, due to increased biosynthesis, and to alesser extent, to water evaporation and subse-quent concentration of solutes. In contrast,acid concentrations decrease during matura-tion. Finally, in discussing the composition ofwine, it is often more useful to consider the liq-uid juice just after the grapes have beencrushed as the starting material, rather thanthe intact grape. As listed in Table 10–2, thejuice, or “must,” consists of several major con-stituents and many other minor componentsthat are important, but which are present atrelatively low concentrations.

Sugars

Other than water, which is 70% to 85% of thetotal juice volume, simple sugars represent thelargest constituent of grapes or must. Depend-ing on the maturation of the grape at harvest,must usually contains equal concentrations ofglucose and fructose,with the latter increasingsomewhat in over-ripened grapes. Sucrose isusually present at very low concentrations(less than 1%), except for musts from V. lab-rusca grapes, which can contain as much as10% sucrose. In general, most grape cultivarscontain about 20% sugar (i.e., 10% glucose and10% fructose), but the actual amount of total

sugar may vary, depending mostly on maturity.Other sugars also may be present, but at verylow concentrations, including the sugar alco-hol—sorbitol—and the pentoses—arabinose,rhamnose, and xylose. It is common practiceamong wine makers to refer to the total sugarconcentration in units of Brix or °Brix. The°Brix value is actually a measure of density orspecific gravity and is easily and quickly deter-mined using a hydrometer. Juice from maturegrapes at 20% sugar is ordinarily about 21°Brixto 24°Brix.

As will be discussed later in more detail, glu-cose and fructose (and to a lesser extent, su-crose) serve as the major growth substrates forthe fermenting yeasts. In fact, the main wineyeast, Saccharomyces cerevisiae, ferments onlythese simple sugars, and all of the ethanol thatis produced is derived from sugar fermentation.In addition, a small amount of the sugar is con-verted to esters, aldehydes,higher alcohols, andother volatile organic compounds (formed alsofrom metabolism of fatty acids and aminoacids) that contribute important flavor andaroma characteristics. When nearly all of thesugar is fermented, and only residual (0.1% to0.2%) amounts remain, the wine is considered“dry.” In contrast, when the residual sugar con-centration at the end of fermentation is 10 g/L

Table 10.2. Constituents of juice and red table wine(g/100 ml)1.

Compound Juice Wine

Water 70—80 80—90Carbohydrates 15—25 0.1—0.3Glucose 8—12 0.5—0.1Fructose 8—12 0.5—0.1Other2 1—3 0.1—0.2Organic acids3 0.5—2.0 0.3—0.6Inorganic salts4 0.3—0.5 0.2—0.4Nitrogenous 0.1—0.4 0.01—0.1Phenolics 0.1—0.2 0.2—0.3Ethanol 0.0 8—15Other alcohols 0.0 0.01—0.04

1Adapted from Amerine et al., 1980 and Boulton et al., 19962Mainly sucrose and various pentoses3Mainly tartaric, malic, citric, lactic, and acetic4Mainly potassium, magnesium, and calcium



or higher, a “sweet” wine is produced. Verysweet wines, such as those consumed asdessert wines, can contain as much as 100 g/Lto 200 g/L (10% to 20%). Finally, it should benoted that when the sugar concentration in themust is low, as might occur in grapes grown incool climates, it is permissible (in some coun-tries, but not in California) to add sugar to themust, a process known as chaptalization.

Organic Acids

Organic acids comprise the second most plen-tiful non-water constituent in must. Althoughthe organic acids are present at relatively mod-est concentrations, typically ranging from 1% to2%, their effect on wine quality is extremely im-portant.These acids are responsible for the lowand well buffered pH of the must and the wine(usually between 3.0 and 3.5).That wine is sowell preserved is due not only to the ethanoland low pH,but also to the presence of organicacids that have powerful antimicrobial activi-ties. At these pH values, pathogenic and othermicroorganisms are inhibited, including Salmo-nella, Escherichia coli, and Clostridium spp.The exceptions are aciduric and acidophilic or-ganisms such as lactic acid bacteria, acetic acid-producing bacteria, and some fungi and yeast.In addition to antimicrobial effects, low pH alsostabilizes the anthocyanins that give red wineits color, inhibits oxidation reactions, and con-tributes desirable flavor.

Wine chemists generally categorize wineacids into two general groups.Volatile acids arethose that are volatilized or removed by steamtreatment; those that remain are consideredfixed acids.There are little, if any, volatile acidsin must. During the ethanolic fermentation,small and usually inconsequential amounts(�0.5 g/L) of acetic acid and other short- andmedium-chain fatty acids may be produced.However, if higher concentrations are pro-duced, either by oxidation of ethanol by bacte-ria or as an end product of microbial metabo-lism, the wine will suffer serious defects andmay be unsalable (as discussed later).

The main fixed acids, depending on thegrape, condition, and maturity, are tartaric acidand malic acid.The ratio of these two acids isusually about 1:5 (tartaric:malic), but it can bereversed in some grapes. These acids are im-portant in wine for several reasons.Since malicacid contains two carboxylic acid groups, itcontributes more protons in solutions, andmakes the must more acidic. If the malic acidconcentration is too low, as might occur inoverly mature grapes grown in warm climates,the wine pH will be too high. The wine willlack the desirable acid flavor and may be moreprone to spoilage by bacteria. In contrast, al-though a minimum acidity is desirable in wine,excess acidity is also a defect and results in aninferior sour-tasting wine. Musts obtained fromgrapes grown in cool climates may containhigh levels of malic acid, and are, therefore,problematic. A natural, biological method fordeacidifying wine is commonly used for suchmusts (see below). Finally, it should be notedthat other important fixed acids may be pre-sent in wine, including succinic acid and lacticacid, both which are produced during thewine fermentation by yeast or bacteria.

Nitrogenous Compounds

Grapes contain both inorganic and organicsources of nitrogen.Total nitrogen concentra-tions in grapes (or musts) range from about0.2 g/L to 0.4 g/L.The ammonium nitrogen isless than 0.1 g/L. Despite their relatively lowconcentration in juice, the nitrogen content ofmost musts is generally adequate for rapidgrowth of yeasts. In fact, the primary role of ni-trogen in wine appears to be as a nutrientsource for the yeasts, rather than affecting anyof the organoleptic or other properties of thewine, per se. Moreover, wine yeasts can assimi-late free ammonia into amino acids and can,therefore, use ammonia directly as a source ofnitrogen. Nonetheless, nitrogen deficiency canoccasionally occur, and is one of the maincauses for sluggish or “stuck” fermentations.Some wine makers, therefore, routinely add



ammonium salts to the must, in the form of ayeast food (especially for those occasionswhen the grapes are deficient in nitrogen).

The main organic nitrogen-containing com-pounds are amino acids, amides, amines, andproteins.The free amino acids are not only syn-thesized into proteins (following transamina-tion reactions), but several can also be used asan energy source.Among the proteins found inmust, the most important are the enzymes.Some grape enzymes serve useful functions,whereas others can present problems. Thepectinases and other hydrolases that enhanceextraction of juice from the grapes duringcrushing are especially important. In contrast,phenol oxidases participate in the well-knownenzymatic browning reaction. In the presenceof oxygen, these enzymes form undesirablebrown pigments that seriously discolor thewine. This reaction, however, is effectively in-hibited by sulfur dioxide.

In recent years, the presence of potentialbioactive nitrogen-containing compounds inwine has attracted concern and interest. Whenthe amino acids histidine and tyrosine are decar-boxylated by decarboxylase enzymes producedby wine bacteria, the amines histamine and tyra-mine are formed.Depending on the dose and in-dividual sensitivity, these biogenic amines cancause headaches, nausea, and allergic-type re-actions. Another compound, ethyl carbamate,can be formed via a chemical reaction betweenurea and ethanol. Ethyl carbamate is suspectedof being a carcinogen,and its presence in wine isincreased by heating steps, such as pasteuriza-tion,and by high urea concentrations.

Polysaccharides

The main polysaccharide in grapes or must ispectin, a structural carbohydrate that providesstructural integrity to the plant cell walls. Thepectin concentration in the must can be as highas 5 g/L,which could potentially cause the wineto become cloudy. However, most of the pectinis either precipitated out during fermentationor is hydrolyzed to soluble sugars by exogenous

microbial pectinases. The latter are commer-cially available and are used not only to enhancemaceration and pressing, but also to improvewine clarity. Other polysaccharides, such as cel-lulose and hemicellulose, are not soluble in thejuice and are removed along with the other non-soluble solids in the form of pomace.

Sulfur Compounds

Several sulfur-containing substances are foundor are formed in grape juice that have a pro-nounced affect on the wine fermentation andwine quality. Hydrogen sulfide (H2S) and vari-ous organic forms of sulfur, especially the mer-captans which are formed from H2S, imparthighly offensive odors in the wine. They areproduced in trace amounts by grape yeastsduring fermentation.

The other major group of sulfur compoundsfound in wine are sulfur dioxide (SO2) and re-lated aqueous forms that exist as sulfite ions.These substances are produced naturally byyeast, and are invariably present in wine, albeitat concentrations usually less than 50 mg/L.However, sulfur dioxide and bisulfite salts arenow commonly added to must due to theirstrong antimicrobial, antioxidant, and anti-browning properties. It is important to recog-nize that these activities occur only when theSO2 is in its free, un-bound form.When boundor fixed with other wine compounds, such asacetaldehyde, reducing sugars, and sugar acids,SO2 activity is diminished. How SO2 specifi-cally functions in wine and its important rolein wine making will be further discussed later.

Phenols, Tannins, and Pigments

Among the most important naturally occurringsubstances in grapes and musts are the pheno-lic and polyphenolic compounds. Some phe-nols can also be introduced into the wine fol-lowing aging in wooden casks or via yeast andbacterial metabolism.These chemically diversecompounds contribute color,flavor, aroma,and



mouth feel to the wine. They can also reactwith other grape components and can eitherimprove or diminish wine quality. Finally, manyof the phenolic compounds found in wine arethought to be responsible for the putative

health benefits associated with wine consump-tion (Box 10–3).

The phenols important in wine are groupedaccording to their chemical structure (Figure10–2). Phenols containing a single phenolic

Box 10–3. L’chaim! Healthful Properties of Wine

However one says it—L’chaim! Salute! a Votre Santé! Cheers!—all are usually said just prior toconsuming wine, and all meaning the same thing—to life and good health. But is wine simply asymbol for expression of this sentiment or might there be an actual scientific basis to supportthe healthful properties of wine? Let’s examine some of the epidemiological, as well as mecha-nistic, evidence collected in recent years.

Epidemiology

If drinking wine promotes health or well-being, one would expect that residents of countrieswhere wine consumption is high would live longer. Rates of heart disease and cancer wouldalso be expected to be lower in those countries compared to countries where wine consump-tion is less. Indeed, there has long been the suggestion, based on international population com-parisons, that moderate wine drinking is good for one’s general health (Figure 1).

France

FinlandIreland

NorwayDenmark

Sweden

GermanyAustria

SwitzerlandItaly

PortugalSpain

Greece

UK

Australia

USA

CanadaNetherlands

Belgium

50 100 150 2000

100

200

300

400

eh yranoroc yb ytilatroM

)000,001 rep( esaesid tra

Consumption of wine (calories per day)

0

Figure 1. Correlation between wine consumption in eighteen industrialized countries and mortality bycoronary heart disease. Adapted from Renaud and de Lorgeril, 1992.



Several recent studies have provided additional epidemiological evidence to support this hy-pothesis. In the study by Klatsky et al (2003), mortality data of nearly 130,000 California adultswho initially had been interviewed between 1978 and 1985 regarding their drinking habits,wasobtained through 1998. Compared to groups that either did not drink wine or that drank beeror liquor,wine drinkers had a significantly lower mortality risk.This result was due mostly to thereduced risks of coronary disease and respiratory deaths. Mortality risk reduction did not de-pend on the type of wine (i.e., red or white) and was consistent for most population subsets(e.g., men versus women, age, and education). Moreover, the risk of coronary disease decreasedas wine drinking frequency increased, from less than once per week to daily or almost daily. Fi-nally, it has also been recently suggested (Pinder and Sandler, 2004) that wine consumptionmight provide a protective effect against dementia and the onset of Alzheimer’s disease.

In another epidemiological study, the mortality of 24,000 Danish adults was followed fornearly thirty years (Grønbæk et al., 2000).Although moderate consumption of beer or liquorhad little effect on mortality, a significant reduction in mortality risk was associated with wineconsumption, even up to twenty-one glasses per week.As for the California study, reduced inci-dence of heart disease accounted for most of the overall reduction in mortality risks, althoughreduced cancer risks were also observed. It is important to note that these and other epidemio-logical studies that show positive associations of wine consumption and reduced mortality donot account for other factors that may influence mortality, such as exercise, diet, other healthhabits, and genetics. Finally, it should be emphasized that heavy wine drinking may increasemortality risk (due, for example, to cirrhosis or automobile accidents).

Effect of wine consumption on heart disease

High rates of coronary heart disease within a population are associated with those groups thatconsume high saturated fat diets.Yet, in some countries, such as France and Italy, where satu-rated fat consumption is high, heart disease rates are significantly lower than in other countrieswhere fat consumption is also high.This so-called French Paradox is thought to be due to thehigh levels of wine consumed in these countries (Renaud and de Lorgeril, 1992).

Given the strong epidemiological evidence showing an inverse relationship between coro-nary disease and wine consumption, what biochemical mechanisms would account for thisfinding? Several possibilities have been suggested (da Luz and Coimbra, 2004). First, alcoholconsumption, in general, is known to increase the high-density lipoprotein (HDL) cholesterolconcentration in blood, a factor known to reduce heart disease risks.Alcohol may also increasefibrinolytic activity and decrease platelet aggregation,which similarly reduce plaque formation.Wine, however, contains specific components (other than alcohol) that may also account forthe reduction in cardiovascular disease.Among the compounds thought to be most importantare polyphenols and flavonoids (Stoclet et al., 2004).These compounds have antioxidant activ-ity and can scavenge superoxide and other reactive oxygen species.Thus, formation of oxidizedlow-density lipoprotein (LDL), which is known to be toxic to endothelial cells and impair va-sorelaxation, are inhibited. Similarly, the antioxidant activity of wine flavonoids can prevent thereaction of superoxide and nitric oxide (NO).The latter has antithrombotic activity, preventsatherosclerosis, and inhibits platelet adhesion and aggregation, and, therefore, is necessary fornormal endothelial function.

Another explanation for the role of wine in reducing coronary disease and endothelium-dependent constriction, in particular, was recently described (Corder et al., 2001).These in-vestigators reported that polyphenols in red wine extracts suppress aortic endothelial syn-thesis of endothelin-1 (ET-1), a peptide that has vasoconstriction activity and that is intimatelyinvolved in development of atherosclerosis.This inhibition of ET-1 synthesis was specific for

Box 10–3. L’chaim! Healthful Properties of Wine (Continued)

(Continued)



red wine, because red grape juice was only 15% as inhibitory and white or rosé wine had lessthan 5% inhibition.The mechanism by which ET-1 synthesis is suppressed was not due to an-tioxidant activity, but rather to the specific inhibition of the tyrosine kinase cell signaling machinery involved in the initiation of ET-1 synthesis.

Effect of wine consumption on cancer

Based on epidemiological studies, consumption of alcohol, in the form of beer or spirits, gener-ally increases cancer risk mortality, whereas moderate wine consumption appears to lowerrisks of some cancers (Bianchini and Vainio,2003;Grønbæk et al.,2000;Klatsky et al.,2003).Sev-eral wine polyphenolic compounds, such as quercetin and catechin, have been shown to haveinhibitory effects, in vitro, against tumor and cancer cells.Two other compounds found in redwine, resveratrol and acutissimin A, have recently attracted considerable attention as moleculesthat may have anti-cancer properties.

Resveratrol is a polyphenol that may contribute cardiological benefits (as do other wine phe-nols), in addition to its protective effects on carcinogenesis.Specifically, resveratrol is thought tointerfere with the cell signaling cascades that result in the synthesis of NF-�B and AP-1.The lat-ter are transcription factors that activate expression of genes involved in promotion of cell cy-cle progression, inflammation, anti-apoptosis, and tumorigenesis.Although the precise mecha-nisms responsible for the anti-carcinogenic and anti-inflammatory effects of resveratrol are notyet established, it appears that by blocking or modulating NF-�B and AP-1 synthesis, resveratroldown-regulates genes necessary for carcinogenesis.

Another component of red wine that has recently been recognized as having anti-cancer ac-tivity is acutissimin A (Quideau et al., 2003).Acutissimin A is a polyphenol with an ellagitanninstructure (galloyl units linked to glucose) that was originally isolated from the bark of specifictypes of oak trees (although not from the oak used to make wine barrels). Its activity as a possi-ble anti-cancer agent is due to its ability to inhibit DNA topoisomerase II, an enzyme involved ingrowth of cancer cells. In fact, acutissimin A is 250 times more inhibitory to this enzyme (invitro) than etoposide, a drug used clinically to treat some cancers. Finally, it appears that wine,but not grapes,contains acutissimin A.Rather, the presence of acutissimin A in wine is due to itssynthesis, during the aging step, from polyphenolic precursors found in grapes. Evidently, theconditions that exist during the aging of wine (i.e., low pH, ethanol, and mildly aerobic atmos-phere) are necessary for the synthesis of acutissimin A.

Despite the mostly positive attention that the wine-health connection receives in the popularpress, as well as in the scientific literature-gathering headlines, it would be misleading not tonote that there are many detractors.Some investigators have argued that dietary diversity, ratherthan wine, is responsible for the French Paradox, that statistical biases underestimate coronarydeaths in France, and that the risk/benefit ratio does not support wine consumption as a meansof promoting cardiovascular health (Bleich et al., 2001; Criqui and Ringel, 1994).

References

Bianchini, F., and H. Vainio, 2003. Wine and resveratol: mechanisms of cancer prevention? Eur. J. CancerPrev. 12:417–425.

Bleich,S.,K.Bleich,S.Kropp,H.-J.Bittermann,D.Degner,W.Sperling,E.Rüther,and J.Kornhuber.2001.Mod-erate alcohol consumption in social drinkers raises plasma homocysteine levels: a contradiction to the‘French Paradox?’Alcohol Alcohol. 36:189–192.

Corder, R., J.A. Douthwaite, D.M. Lees, N.Q. Khan, A.C. Viseu dos Santos, E.G. Wood, and M.J. Carrier.2001. Endothelin-1 synthesis reduced by red wine. Nature 414:863–864.

Criqui, M.H., and B.L. Ringel. 1994. Does diet or alcohol explain the French paradox? Lancet 344:1719–1723.




ring and functional groups at carbon 1 or 3 arereferred to as nonflavonoids. In contrast, com-pounds containing two or more phenolic ringsconnected by pyran ring structures are calledflavonoids. Both nonflavonoids and flavonoidscan be polymerized to form another importantclass of phenolic compounds called tannins.Tannins consisting of flavonoid phenols are es-pecially important in wine due to their bitterflavor and color stabilization properties.

Within the flavonoid group are several dif-ferent types of compounds, including theflavonols, catechins, and anthocyanins. Theseare extracted from the grape skins and seedjust after the grapes have been crushed. Ingeneral, the longer the wine is in contact withthe seed and skin and the higher the tempera-ture, the more of these substances will be ex-tracted. In the manufacture of red wine, thismaterial is left in contact with the pressedjuice for several days, even during fermenta-tion, whereas it is removed almost immedi-ately after crushing for white wine manufac-ture (see below).Thus,although both red wineand white wine contain phenolic compounds,the concentration in red wines is usually atleast four times higher than that of whitewine. Importantly,many of these phenolics areresponsible for pigmentation of grapes and,

subsequently, the color of the wine. Antho-cyanins, for example, have red or blue-redcolor and make red wines red. Most of the an-thocyanins are attached, via ester linkages,to one or two glucose molecules, which en-hances their solubility and stability.

Wine Manufacture Principles

Making wine, as far as the actual steps are con-cerned, looks to be a rather simple andstraightforward process (Figure 10–3). Grapesare harvested and crushed, the crushed mater-ial or juice is fermented by yeasts and bacte-ria, the organisms and insoluble materials areremoved, and the wine is aged and bottled. Inreality, the process is far from easy, and each of these pre-fermentation, fermentation, andpost-fermentation steps must be carefully exe-cuted if high-quality wine is to be consistentlyproduced.

Harvesting and Preparing Grapes forWine Making

According to both viticulturists and enologists,the first step in wine making is considered tobe one of the most important. Grapes must beharvested at just the right level of maturity.This

da Luz, P.L., and S.R. Coimbra. 2004.Wine, alcohol and atherosclerosis: clinical evidences and mechanisms.Braz. J. Med. Biol. Res. 37:1275–1295.

Klatsky,A.L., G.D. Friedman, M.A.Armstrong, and H. Kipp. 2003.Wine, liquor, beer, and mortality.Am. J. Epi-demiol. 158:585–595.

Kundu, J.K., and Y.-S. Suhr. 2004. Molecular basis of chemoprevention by resveratrol: NF-KB and AP-1 as po-tential targets. Mut. Res. 555:65–80.

Grønbæk,M.,U.Becker,D.Johansen,A.Gottschau,P.Schnohr,H.O.Hein,G.Jensen,and T.I.A.Sørensen.2000.Type of alcohol consumed and mortality from all causes, coronary heart disease, and cancer.Ann. In-tern. Med. 133:411–419.

Quideau, S., M. Jourdes, C. Saucier,Y. Glories, P. Pardon, and C. Baudry. 2003. DNA topoisomerase inhibitoracutissimin A and other flavano-ellagitannins in red wine.Angew. Chem. Int. Ed. 42:6012–6014.

Pinder, R.M., and M. Sandler. 2004.Alcohol, wine and mental health: focus on dementia and stroke. J. Psy-chopharmacol. 18:449–456.

Renaud,S.,and M.de Lorgeril.1992.Wine,alcohol,platelets,and the French paradox for coronary heart dis-ease. Lancet 339:1523–1526.

Stoclet, J.-C.,T. Chataigneau, M. Ndiaye, M.-O. Oak, J.E. Bedoui, M. Chataigneau, and V.B. Schini-Kerth. 2004.Vascular protection by dietary polyphenols. Eur. J. Pharmacol. 500:299–313.




means that the concentrations of sugars andacids (and the sugar/acid ratio), pH, the totalsoluble solids, and even the phenolic con-stituents must be at just the right level for theparticular cultivar and the type of wine beingmade. In addition, berry size and weight alsoinfluence the time at which grapes are har-vested. In general, grapes should be sampledsometime before their expected harvest timeand their composition assessed (at minimum°Brix and pH should be measured) to makesure that over-ripening does not occur.

Unfortunately, there is no exact or objectiveset of rules to ensure or predict the optimumtime for harvesting grapes. Rather, grapes arefrequently harvested based on more subjectivecriteria.As grapes ripen on the vine, the sugarconcentration, as well as flavor and color com-ponents, increase, and acids usually decrease,so identifying the correct moment for harvest-ing can be a real challenge. It is possible, more-over, for grapes to over-ripen, such that the har-vested grapes contain too much sugar or toolittle acid or be too heavily contaminated with

COOH

trans - cinnamic acid

a.

OH

HO

HO

trans - resveratrol

b.

catechin

OH

HO O

OHOH

OH

c.

COOH

HO

HO

HO

gallic acid

d.

quercetin

OH

HO O

OH

OH

OH

O

e. f.

malvidin

OH

HO O

OH

OH

OCH3

OCH3

Figure 10–2. Representative phenolic compounds in wine. Examples include: a. cinnamic acid, a phenolicacid; b. resveratrol, a derivative of cinnamic acid; c. catechin, a flavan-3-ol; d. gallic acid, a tannin; e. quercetin,a flavonol; and f. malvidin, an anthocyanin.



wild yeast and molds. Once the grapes havebeen deemed properly mature, it is essentialthat they be picked and harvested quickly,since the composition can continue to change.

Even in this twenty-first century, when somuch of modern agriculture has become auto-mated and mechanized, a sizable portion ofgrapes for wine making is still harvested manu-ally. Only recently has mechanical harvestingbegun to displace manual harvesting. In theUnited States, the majority of grapes are nowharvested by mechanical means; however,

manual picking of grapes is still done for pre-mium quality American wines and in much ofEurope. Manual harvesting is more gentle onthe grapes, and bruising and breaking of thegrapes is minimized. For certain wines, such assweet wines made from noble rot grapes, orwines for which grape harvesting methods areregulated (e.g., French Champagne), manualharvesting is required. Anyone who has seenthe odd contours and steep terrains of some ofEuropean vineyards will also appreciate thenecessity of manual grape picking. On the

sulfite

grapes

stemming

crushing

white wine

press

juice

yeast fermentation

racking

bottling

aging

clarification

malolacticfermentation

red wine

yeast fermentationand maceration

aging

clarification

press

racking

bottling

malolacticfermentation

Figure 10–3. Flow chart of wine manufacture.The main difference between the manufacture ofred and white wine is that the grapes for redwines are fermented in the presence of skins andseeds (maceration), whereas for white wines,these materials are removed after crushing. Al-though optional, sulfiting agents are added tomost U.S. wines. Whether the malolactic reac-tion is induced or encouraged depends on thegrape composition, the style of wine, and themanufacturer’s preferences.



other hand, mechanical harvesters are fasterand cheaper, and,unlike hired laborers,deploy-able on short notice and available around theclock.

Once the grapes are removed from thevines, they must be transported to the winery.It is important that the grapes not be bruised,crushed, or otherwise damaged either duringharvesting or transport, since this encouragesgrowth of microorganisms prior to the actualstart of the fermentation. For the same reason,transportation time is also important.

Crushing and Maceration

The purpose of crushing is to extract the juicefrom the grapes. Before the grapes are crushed,however, leaves, large stems, and stalks are re-moved. Some wine makers may not remove allof the stems to increase the concentration oftannins and other phenolic compounds thatare present in the stems and extracted into thejuice. Once the extraneous material is sepa-rated and removed, the grapes are crushed byone of several types of devices. Roller crushersconsist of a pair of stainless steel cylinder-shaped rollers. Another type of crusher, calledthe Garolla crusher, not only performs thecrushing step, but also removes stems. It con-sists of a rotating shaft contained within a largehorizontal stainless steel cylinder or cage.Armson the shaft are attached to paddles or bladessuch that when the shaft turns, the grapes aremoved and pressed against the side of the cylin-der. Perforations on the walls of the cylinder al-low for the juice (along with the skin, seeds,and pulpy material) to pass through into collec-tion vats, whereas the stems gather at the end.

The crushed grape material, as noted above,contains juice, seeds, and skins. Pigments, tan-nins, and other phenolic compounds are lo-cated in the skins and seeds, and their extrac-tion into the juice takes time. Endogenouspectinases and other hydrolytic enzymes withinthe grapes enhance extraction and must also begiven time to work.This extraction step, wherethe crushed grape material is allowed to sit, isreferred to as maceration.

Maceration conditions are not the same forall wines. For red wines, where pigment ex-traction is especially important, long macera-tion times at high temperatures are usually employed. In general, maceration is done ataround 28°C for up to five days. The shorterthe maceration times and the lower the tem-perature, the less material will be extracted.Thus, lighter red wines, such as Beaujolais, aremacerated for just a few days at no higher than25°C. In contrast, deeper red wines, such asBordeaux, are macerated for up to twenty-eight days at 30°C. Since fermentation beginsshortly after the grapes are crushed, macera-tion and fermentation essentially occur at thesame time. In fact, the ethanol made by fer-menting yeasts enhances extraction.This situa-tion only occurs, however, if the musts are nottreated with sulfur dioxide (see below).

As noted above, maceration at low tempera-tures (�15°C) ordinarily results in only moder-ate pigment extraction and little fermentation.However, if the must is macerated at a low tem-perature (between 5°C and 15°C),but for longertime, extraction of anthocyanins and aroma andflavor compounds can be enhanced.This tech-nique, called cold maceration, simulates the nat-ural conditions in cooler wine-producing areas,such as the Burgundy region of France. Forwhite wines, the maceration step is done atlower temperature and for much less time.Typi-cally, only a few hours at 15°C is sufficient. Formost white wines, the producers remove theseeds and skins immediately after crushing. Asfor red wine, the maceration conditions used forwhite wines influence the amount of pigmentsand tannins that are extracted.Wines made fromSauvignon blanc grapes where little macerationoccurs typically have a low phenolic concentra-tion, whereas Riesling and Chardonnay musts,which are often macerated in the cold,may con-tain appreciable amounts.

Sulfur Dioxide Treatment

This is now a convenient point at which to dis-cuss the use of sulfur dioxide and pure culturetechnique for wine making. As soon as the in-



tegrity of the grapes has been compromised bythe crushing step, the sugars in the juice are lib-erated and made available for whatever mi-croorganisms happen to be present. Ordinarily,the must is populated by epiphytic yeasts (thatis, yeasts that reside on the surface of thegrapes) and by yeasts that have “contaminated”the crushers, presses, and other wine-makingequipment. Although the surface of a singlegrape may contain only about 102 to 104 yeastcells, after the grapes have been exposed to thecontaminated equipment, the number of cellsincreases about 100-fold, to about 104 to 106

cells per ml. Whether this resident microfloraactually commences a fermentation, however,depends on the intent of the wine maker.

Two options exist. First, a spontaneous ornatural fermentation may be allowed to pro-ceed (“natural” simply means that pure startercultures are not added). In this case, except fortemperature control, essentially no other re-strictions are placed on the fermentation, and

yeast (and bacterial) growth occurs with just arelatively short lag phase. The other option isto start the fermentation, under controlled cir-cumstances, with a defined yeast starter cul-ture selected by the wine maker. The latter option usually requires that the indigenous microflora be inactivated, so that it does notcompete with and possibly interfere with theadded culture.This is not, however, always thecase, because it is still possible to add a starterculture even in the presence of the back-ground flora. The use of wine starter cultureshas now become commonplace, even amongmany traditional wine manufacturers. More-over, recent research on yeast ecology suggeststhat there may be less diversity in wines madeby spontaneous fermentations than previouslythought and that these natural fermentationsinvolve relatively few strains (Box 10–4).

When starter cultures are used, the naturallyoccurring or so-called wild yeasts are inacti-vated in one of two ways. First, the must can be

Box 10–4. Culture Wars and the Role of Diversity in Wine Fermentations

Like all of the fermented products described in this text,wines were made for thousands of yearsbefore scientists recognized that microorganisms were responsible for the fermentation. In con-trast to dairy, meat, and other food fermentations, where first backslopping, and later pure cul-ture techniques were applied, wine makers have long relied on the indigenous microflora to ini-tiate and complete the wine fermentation.Even today,many of the wines produced by traditionalmanufacturers, especially those in Europe, are made by a “spontaneous” or natural fermentationcarried out by the wild yeasts that are present on the grapes, in the must,or on equipment.Mostlarge, modern wineries, however, have adopted yeast starter culture technology, and many haveabandoned the traditional practice of allowing the wild yeasts to commence the fermentation.

The debate

Advocates of wine starter culture technology make several good arguments. First, they contendthat using defined yeast strains results in wine having a “pure” or cleaner flavor. In addition, theavailability of yeast with particular physiological characteristics such as osmotolerance or theability to grow at low temperature allows the wine maker to select yeasts particularly suited tothe grape characteristics one is using or the wine type one is making (Table 1).

Importantly, the wine maker can expect culture performance to be consistent in terms offermentation times, flocculation properties, and flavor and end-product formation. In addition,culture yeasts can also be customized to provide traits and performance characteristics thatsuit the particular needs of the wine producer. Finally, there may be conditions that make purecultures almost essential, such as when the grapes or musts contain very high sugar or acidconcentrations.

(Continued)



Table 1. Starter cultures versus natural fermentation for wine-making.

Starter cultures Natural

Cleaner flavor More complex flavorGreater consistency Unique qualitiesFaster SlowerLow frequency of stuck fermentations Greater frequency of stuck fermentationsCan customize strains Cannot customizeImmune to killer yeasts May be sensitive to killer yeasts

Despite these advantages, it is undoubtedly true that naturally fermented wines have an ex-cellent track record and many are highly regarded by wine experts. Even if one were to con-cede that the flavor, aroma, and other organoleptic properties are more variable in naturally fer-mented wines, is that such a bad thing? Certainly, one could argue that it is the variability itselfthat makes it possible to produce truly exceptional wines. Proponents of natural wine fermen-tations claim that such wines are more “complex” due to the more complicated metabolicprocesses that occur within a complex yeast flora. In other words, it is the diverse metabolicflora that leads to a more complex distribution of flavor components in the wine.

To address this issue of metabolic diversity on a rational basis, several questions must first beraised. How many different strains are actually present during a natural wine fermentation? Aresome strains more dominant than others? What are the phenotypic properties of these strains, rel-evant to wine flavor generation? Several recent studies have attempted to answer these questions.In the report by Cappello et al. (2004), between thirty and forty naturally-occurring yeasts wereisolated from each of twelve different musts.Although the musts were made from different culti-vars (each obtained aseptically), the grapes were grown within the same vineyard.All of the yeastisolates were identified by classical methods and by genetic analyses as Saccharomyces cere-visiae, and, based on PCR amplification and mitochondrial restriction analyses, all of the strainswithin each of the twelve groups appeared to be genetically homogenous. Comparative analysisof representative strains from the twelve groups revealed that there were three genetically dis-tinct groups.However, there were only minor differences in their physiological properties.

This study points out several important features. First, it suggests that the indigenous yeastspresent in wine reflect the yeast population within the vineyard, rather than the grape. Second,it would appear that only a few strains, rather than a diverse collection of strains, dominate thewines produced within a vineyard. Finally, these dominant yeasts were well-adapted to the par-ticular local environmental conditions and shared similar physiological properties relevant towine making. However, because the yeasts were isolated at a single point in time (at the end ofthe fermentation), it is possible that other strains may have been present at the beginning or atsome earlier time during the fermentation.

Another recent study addressed similar questions, but in these experiments the musts wereanalyzed at the beginning and end of the fermentation (Demuyter et al., 2004).Thus, strain di-versity could be determined at two different stages.Over a three-year period,grapes from Alsace(used to produce Gewurztraminer wines) were obtained, crushed, and fermented under threedifferent conditions. For each year, 100 yeast isolates obtained from the beginning and end ofthe three independent fermentations were identified and classified by polymerase chain reac-tion and pulsed field gel electrophoresis techniques. Grape and must handling conditions wereslightly different for each of the three fermentations, such that the origins of the isolated yeastscould have been from the grapes, the press equipment, or the vat.

The results for the first-year grapes showed that nearly all of the grape-associated strains,bothfrom the beginning and end, belonged to a single homogenous group of S. cerevisiae. In con-

Box 10–4. Culture Wars and the Role of Diversity in Wine Fermentations (Continued)



heat treated, usually via a high-temperature,short-time (or flash) pasteurization process. Al-though very effective against most organismsfound in musts, even moderate heating is oftendetrimental to the juice and to the wine.Thus,this process is rarely used. The preferredmethod is to chemically pasteurize the must byadding sulfites. It should be noted that even nat-urally fermented wines, especially white wines(see below), are often sulfite-treated to controlundesirable organisms.The most common sulfit-ing agents are SO2 gas and potassium bisulfitesalts.Sulfites are cheap,effective,and multi-func-

tional. In addition to their effectiveness againstwild yeast, these agents also inhibit growth ofacetic acid-producing microorganisms, malolac-tic bacteria, and various fungi. Sulfites can thusbe considered as serving a preservative functionin wine. Importantly, they also control severaldeleterious chemical reactions, particularly oxi-dation and browning reactions.

The amount of sulfite added and when it isadded varies depending on the condition ofthe grapes, the microbial load, must pH andacidity, and the type of wine (red or white).Musts from mature grapes that often contain

trast, the microflora from grapes exposed to crushing equipment or to the vat environment wasmore heterogenous and contained relatively few S. cerevisiae isolates. Instead, these musts con-tained mostly Saccharomyces bayanus subsp. uvraum. Results for the second-year grapesshowed that S. bayanus subsp. uvraum was the dominant organism from all three environ-ments at all sampling times.Third-year results for grapes or grapes exposed to crushing equip-ment indicated that S. cerevisiae was dominant at the beginning and end of the fermentation,but grapes exposed to the vat environment contained S. bayanus subsp. uvraum as the domi-nant organism (as was the case for the year one results).

Collectively, this study on Alsatian wine shows that a homogeneous group of S. cerevisiaestrains appears to dominate the grape surface, yet other non-S. cerevisiae yeasts still show up inthe wine.The latter are associated with both the crushing equipment or the vat environment.Moreover, under the conditions established in this study, those yeasts present in the vat envi-ronment may actually be the dominant yeasts in the wine fermentation,since they were presentat the beginning and end of all three fermentations conducted over a three-year period.

Although results represent only two limited studies, they nonetheless suggest that the yeastpopulation responsible for the wine fermentation had only modest diversity and that domina-tion by a particular strain or strains may be common. Indeed, these findings are consistent withother ecological studies (Versavaud et al., 1995). Certainly, these reports also confirm previousreports that suggest wine crushing equipment and the wine environment serve as a majorsource of yeasts involved in the wine fermentation.Thus, it may well be that natural fermenta-tions do not owe their special properties to a diverse population of wild yeasts that reside inthe vineyard or on the surface of grapes, but rather to only a few yeasts that contaminate thewine-making equipment and that live in the confines of the winery environment.

References

Cappello, M.S., G. Bleve, F. Grieco, F. Dellaglio, and G. Zacheo. 2004. Characterization of Saccharomycescerevisiae strains isolated from must of grape grown in experimental vineyard. J. Appl. Microbiol.97:1274–1280.

Demuyter, C., M. Lollier, J.-L. Legras, and C. Le Juence. 2004. Predominance of Saccharomyces uvarum dur-ing spontaneous alcoholic fermentation, for three consecutive years, in an Alsatian winery. J. Appl.Microbiol. 97:1140–1148.

Versavaud,A., P. Courcoux, C. Roulland, L. Dulau, and J.-N. Hallet. 1995. Genetic diversity and geographicaldistribution of wild Saccharomyces cerevisiae strains from the wine-producing areas of Charentes,France.Appl. Environ. Microbiol. 61:3521–3529.

Box 10–4. Culture Wars and the Role of Diversity in Wine Fermentations (Continued)



high levels of wild yeast require more SO2, butin general, about 80 mg/L is sufficient.Also, thelower the pH, the less sulfite is necessary forantimicrobial activity. Due to human healthconcerns, however, there are also regulationsthat dictate how much sulfite can be present inwine. In France, most red and white winesmust contain less than 160 mg/L or 210 mg/L,respectively. In the United States, the limits are350 mg/L for both red and white wines. Usu-ally, SO2 or sulfite salts are added to the mustjust after crushing.

When pure yeast cultures are used, they areusually obtained from commercial sources. So-called house strains, although common forbeer, are infrequently used for wine making,due primarily to the expertise and equipmentrequired for strain maintenance and propaga-tion. In contrast, commercial cultures are easyto use and require only modest technical ex-pertise.The main advantage of commercial cul-tures, however, is that they provide the winemanufacturer with many culture options.Thisis because a variety of strains is available, andthe choice of strain can be selected based onthe needs of the customer (Table 10–3).For ex-ample, grapes with consistently high sugarconcentrations may require an osmotolerantyeast culture. Yeast used to perform the sec-ondary fermentation that occurs in Cham-pagne manufacture must sediment well to fa-cilitate its removal during the disgorgementstep (described later in this chapter).

Yeasts are usually supplied in the form of ac-tive dry yeast,not unlike that used by the breadindustry (Chapter 8). Culture suppliers ordi-narily recommend inoculum levels sufficientto provide a starting cell density of about 106

cells per ml of must.An increase of only abouttwo log-cycles occurs during the fermentation.The manner in which the yeast cultures areproduced and preserved is important, how-ever, because how the cells are handled duringthe culture production steps may influencehow they grow later in the juice.

When culture companies produce yeast cul-tures, cells are grown under highly aerobic con-ditions to achieve the maximum amount of bio-mass. These high biomass fermentations alsoelicit a general stress response that enhancessurvival and viability during the subsequent dry-ing steps. It appears that resistance to drying ispromoted by the synthesis of the disaccharidetrehalose by wine yeast (similar to the cryoresis-tance mechanism by bakers’ yeast, described inChapter 8).Thus, following culture production,yeast cells are induced for aerobic metabolismand anabolic activities, whereas, when thesecells are inoculated into the must, fermentativemetabolism, with a minimum of biomass, is de-sired. Therefore, rapid adaptation to the newwine environment is also important.

Other Pre-treatments

It is permissible to add other materials, be-sides sulfites, to the must to enhance extrac-tion, modify the composition, or promote fer-mentation. For example, pectic enzymes canadded during crushing to facilitate extractionof juice from skins and later during pressingto improve clarification.This practice is quitecommon for white wines manufactured in theUnited States.As noted above, it is not uncom-mon for some grapes, and juices from thosegrapes, to have either too low or too high of apH.Thus, either acids, such as tartaric acid, orneutralizing salts, such as calcium carbonate,can be added to adjust must pH. Finally, nutri-ents that enhance yeast growth and fermenta-tion can be added.These yeast growth factors

Table 10.3. Desirable properties of wine cultures.

Able to produce high levels of ethanolDoes not produce off-flavorsCapable of producing unique flavorsEthanol tolerantOsmotolerantFerment sugars to completenessSediments or flocculates well (especially for

Champagne yeasts)SO2 and sulfite resistantCold resistantRapid fermentation ratesAble to grow at temperatures below 10°CPredictable and consistent and genetically stable



usually are added to white wine juices, sinceshorter extraction times result in lower nutri-ent concentrations. Nutrients added to juicesinclude mainly ammonium salts and variousvitamins.

Microbial Ecology and SpontaneousWine Fermentations

In the absence of SO2 addition, the indigenousmicroflora is relied upon to initiate and thencarry out a spontaneous or natural fermenta-tion.This is one of the most well studied of allfermentations, and much is now known aboutthe ecology of wine and the yeasts that partici-pate in the wine fermentation. In reality, how-ever, the yeast fermentation is but one of two distinct fermentations that occur in wine making. Yeasts, of course, ferment sugars toethanol, CO2, and small amounts of other endproducts. A second fermentation, called themalolactic fermentation, is carried out by spe-cific lactic acid bacteria that are either natu-rally present or added for this purpose. Themalolactic fermentation, to be discussed later,is now regarded as nearly as important to winequality as the ethanolic fermentation.

As noted above, the surface of grapes usuallycontain less than 104 yeast cells per grape (orper ml of juice).This number may increase dur-ing ripening on the vine, especially if the tem-perature is warm. Although ten or more yeastgenera may be represented, the primary organ-ism most frequently isolated from grape sur-faces and the fresh must is Kloeckera apicu-lata. In contrast, S. cerevisiae, the yeast mostresponsible for the wine fermentation, is rarelyobserved on grapes. Rather, S. cerevisiae andother related strains are introduced into themust during grape handling and crushing stepsdirectly from the equipment.The must is inocu-lated, in other words, by the yeasts originatingfrom the grape surface as well as by those re-siding on the winery equipment.

Despite the large amount of available carbo-hydrates and other nutrients, the must is actu-ally quite a selective environment.The pH is typ-ically below that which many organisms can

tolerate, and the organic acids present in themust have considerable antimicrobial activity.The high sugar concentration and resultinghigh osmotic pressure can also inhibit many ofthe indigenous organisms. Eventually, the CO2

formed during the early stages of fermentationmakes the environment anaerobic, restrictinggrowth of aerobic organisms. Likewise, enoughethanol is produced to provide selection againstethanol-sensitive organisms. Finally, sulfites, ifadded,have antimicrobial activity against a widerange of yeasts and bacteria.Thus, as the envi-ronment changes,yeast species that are numeri-cally dominant may be displaced by otherspecies or strains better suited to the environ-ment at any particular time.Microbial ecologistsrefer to such arrangements as successions. Onereason why the outcome of a natural wine fer-mentation is so difficult to predict is that suc-cessions rarely occur exactly the same way in different musts, since the wild flora is neverentirely the same.

Members of the genus Saccharomyces repre-sent less than 10% of the initial yeast popula-tion.The early yeast population is dominated byK. apiculata.This organism produces up to 6%ethanol from glucose and fructose; however, ithas a low ethanol tolerance, being inhibited by3% to 4% ethanol.Thus, even though K. apicu-lata is among the first to grow in the must, itsnumbers are not usually maintained beyond thefirst several days.However, it is thought that thisorganism does produce small amounts of acids,esters, glycerol, and other potential flavor com-ponents, some of which may or may not be de-sirable. As the ethanol concentration increasesand the Eh (or redox potential) is reduced, dueto CO2 formation, the environment begins to se-lect for S. cerevisiae and other various ethanol-tolerant Saccharomyces spp.

According to current yeast taxonomy (Chap-ter 2), most wine yeasts are now classified simply as Saccharomyces cerevisiae. Thus, al-though different Saccharomyces strains maybe present in musts during the course of a fer-mentation or in musts of varying composition(e.g., high sugar, low pH, etc.), these strains areall S. cerevisiae. It is interesting to note that



S. cerevisiae is especially well-adapted to thewine environment and appears to have beenthe predominant wine strain for thousands ofyears (Box 10–5).

Separation and Pressing

After the maceration step, or in the case ofmost white wines, almost immediately aftercrushing, the juice is separated from the seeds,skins, and pulp (collectively referred to as thepomace). For red wines, some fermentationwill have already occurred prior to the separa-tion step, whereas for white wine, fermenta-tion follows the separation and clarificationsteps. The juice that separates from the po-mace simply by gravitational forces is calledthe “free run.” Screens are typically used tocatch any large particles.The free run juice ispumped into vats or barrels.

Since the free run juice contains less than75% of the total juice volume and the rest ispresent within the pomace, the latter is usuallypressed to recover the remaining juice. Severaltypes of presses and configurations are used.Hydraulic or pneumatic wine presses squeezethe juice from the pomace. Screw- or auger-type devices force the juice against perforatedcylinder walls and have the additional advan-tage of being continuous. The so-called firstpress juice can be collected and either addedback to the free run juice or kept as a separateportion.The free run fraction is considered tohave an appreciably higher quality and is usedfor premium wines. Juices containing mixturesof free run and pressed fractions are used forlower quality wines. Finally, for white wine, thejuice is clarified to remove any remainingsolids. Clarification is done via either settlingand decantation, filtration, or centrifugation.

Fermentation

The wine fermentation begins as soon as thegrapes are crushed.However,when a starter cul-ture is used and SO2 is added to control the in-digenous organisms, limited ethanol fermenta-tion will occur prior to addition of the culture.

In the case of white wine production, the cul-ture is added to the must after pressing and clar-ification, whereas for red wine, culture additionis done prior to seed and skin removal.Thus, forred wines, fermentation occurs during macera-tion, just as it would for a natural fermentation.The amount,concentration,and form of the cul-ture depends on the type of wine being pro-duced, the composition of the grapes,and otherconsiderations specific to the wine manufac-turer. Of course, the culture’s main responsibil-ity is to produce ethanol from sugars (see be-low), but the criteria for culture selectionactually includes many other properties (Table10–3). For most wines, the culture inoculum,whether in a rehydrated dried or active liquidform, (Chapter 3), should provide about 106

cells per ml of must.Traditionally, fermentations were performed

in open barrels or vats with a capacity of 500 Lor less. Such barrels still are used today; how-ever, enclosed stainless steel tanks are nowmore common. The latter have several advan-tages.They are easy to clean and disinfect andoften can be sterilized. Airborne microorgan-isms are less likely to contaminate the wine.Various control features, including tempera-ture control and mixing and pumping activi-ties (see below), are easily incorporated intothe design, and are usually computerized. Fi-nally, modern tanks can be quite large, with ca-pacities of more than 250,000 L.

The temperature of incubation depends onthe type of wine being produced. In general,white wines are fermented at lower tempera-tures than red wines. For example, manywineries control the temperature between7°C and 20°C for white wine and between20°C and 30°C for red wines. Some wineriesprefer low incubation temperatures for allwines, because less ethanol is lost to evapora-tion and fewer of the volatile flavors are lost.In addition, low temperature incubations re-sult in overall higher ethanol concentrationsand less sugar remaining at the end of the fer-mentation (assuming time is not a factor).Although S. cerevisiae has a lower growth rateas the temperature decreases, the diversity of



metabolic end products may actually increase,enhancing flavor development. In addition,lower temperatures may favor growth of K.apiculata and other wild yeasts (at least whenthey are not inactivated by SO2) that producevarious volatile compounds, making the winearoma and flavor appear more complex.

It is critical to recognize that the wine fer-mentation is exothermic and a considerableamount of heat may be generated. Some of thisheat is gradually lost or dissipated into the envi-ronment without ill effect. However, in largevolume fermentations, in particular, much ofthis heat is retained, raising the temperature of

Box 10–5. Molecular Archaeology and the Origin of Wine Yeast

How is it that the fermentation of beer, wine and bread all require yeasts with specific physio-logical and biochemical traits, yet all are made with Saccharomyces cerevisiae? Was there aprogenitor S. cerevisiae strain that somehow evolved over thousands of years into specializedstrains that were especially suited for these different fermented foods? If so, where did thatprogenitor strain arise and what were its properties? These are obviously questions that cannotbe answered directly. However, it is possible, using molecular archaeology analyses, to make in-ferences regarding the evolution of these important fermentative yeasts (Cavalieri et. al., 2003).

As noted previously, wine making appears to have developed in the Caucasus and ZagrebMountain areas of the Near East (the “Fertile Crescent”) more than 7,000 years ago.Ancient clayjars (estimated date about 3150 B.C.E.) from these regions once contained wine, according tochemical and instrumental analyses. Furthermore, DNA residues were found when organicresidues (presumably the lees, comprised of dead yeast cells) from these artifacts were ana-lyzed. In fact, the amount of DNA present in the residue was high enough (it could even be seenin a gel) to eliminate the possibility that it had come from a stray microbial contaminant. Giventhe low water activity, low relative humidity (essentially zero), and overall ideal conditions forstoring a biological material, it was also not surprising that the DNA was in such good shape.

The DNA was subsequently extracted and used as a template for PCR.Primers were based onS. cerevisiae rDNA sequences, such that the spacer regions between the 18S and 28S rRNAgenes would be amplified.Three PCR products (540, 580, and 840 base pairs) were initially se-quenced.Based on a BLAST search,all of the sequences had homologies to existing GenBank se-quences.The sequence of the 580 bp PCR product was similar to various fungi, but remarkably,there was strong homology (nearly 90% identity) between the 540 bp sequence and a similarlyderived sequence obtained from a fungal clone isolated from the clothing of “The Iceman Ötzi.”The latter dates back to 3300 B.C.E.The investigators suggest that both of the ancient fungi hadbeen “buried”along with the wine.

Perhaps the most interesting finding from this analysis, however, concerned the 840 bp PCRproduct.This fragment, as well as smaller pieces obtained using internal PCR primers, had veryhigh sequence similarity with a 748 bp region from chromosome 12 (part of the 5.8S rDNA)from contemporary strains of S. cerevisiae,Saccharomyces bayanus, and Saccharomyces para-doxus. Across this entire region there were only four nucleotide mismatches and no deletionsor insertions between S. cerevisiae and the ancient wine sequence. This level of similarityshould not be so surprising, perhaps, since the DNA regions represented by the PCR productsare known to be quite stable.Thus, this stability could account, in part, for the lack of sequencealterations between the ancient and modern strains during the past 5,000 years of yeast evolu-tion. Importantly, these findings suggest the yeast used to carry out the wine fermentation in an-cient days was S. cerevisiae, just as it is today.

References

Cavalieri, D., P.E. McGovern, D.L. Hartl, R. Mortimer, and M. Polsinelli. 2003. Evidence for S. cerevisiae fer-mentation in ancient wine. J. Mol. Evol. 57:S226–S232.



the wine.For example, if the initial temperaturestarts at 20°C, the temperature can increase10°C or more. If the temperature were to riseabove 30°C, the yeast may become inhibited orstop growing altogether.The wine will containless ethanol and more residual sugar.

Such fermentations are said to be “stuck.”Although other factors may cause a wine fer-mentation to become stuck (see below), hightemperature is the most common reason. It is,therefore, essential that the appropriate tem-perature is maintained. For fermentationsconducted in modern, stainless steel, jacketedvats, coolant solutions can easily be circu-lated, externally.Alternatively, internal coolingcoils can also achieve the same effect. Thecooling requirement can also be met, espe-cially for white wines, by simply locating fer-mentation barrels in cold rooms or cellars.However, the need for adequate cooling hasled even some traditional wine manufacturersto abandon oak barrels and casks in favor ofstainless steel vats.

The actual fermentation period is not long.After culture addition, the yeasts enter a shortlag phase (from a few hours up to a day or two)that is then followed by a period of activegrowth (log phase) that lasts for three to fivedays. If the fermentation is conducted at lowertemperatures (10°C to 15°C), the lag and logphases can be extended for several days. Con-versely, if the yeast culture is highly active at theoutset, by virtue of having been previouslypropagated under ideal growth conditions, thecells will almost immediately enter log phase.Al-though one might expect that a natural fermen-tation would take longer, in fact, growth of theindigenous yeast begins so soon after crushing,that the lag phase is barely noticeable.

During the log phase of growth, when anactive fermentation is occurring, a layer ofCO2 forms across the surface. In red wine pro-duction, some of the pomace will float to thetop and be trapped within this CO2 layer,forming a dense blanket or cap. Since the pig-ments and tannins are present in this thickcap layer, a mixing step is required to returnthese substances back into the fermenting

must.The temperature in the cap can also be-come elevated, supporting growth of undesir-able thermophilic bacteria; thus,mixing servesto maintain a more uniform temperature.Vari-ous techniques exist for this mixing step(called pigeage). In the “pumping over” tech-nique, a portion of the must is periodically re-moved from the vat and pumped onto thecap. Alternatively, the cap can be “puncheddown,” either manually or via mechanicalmeans. In modern wineries, automated pump-ing and punching systems are used to mix thepomace cap into the fermenting must.

The fermentation of white wine occurs af-ter the must is pressed and clarified. Afterabout the seventh or eighth day of fermenta-tion, cell numbers may begin to decline, repre-senting the end of the primary ethanolic fer-mentation. Some yeast strains will flocculate orclump together, causing them to settle to thebottom of the tank.This is a desirable propertythat enhances their removal later during theracking and clarification processes. The fer-mentation is considered complete when all ormost of the sugars are depleted, as determinedby a decrease in the Brix value. For red winesthis may take as long as five to six weeks. If lessthan about 0.5% sugar is present, and there isno apparent perception of sweetness, the wineis considered to be dry.Of course,not all winesare intended to be dry. If some sugars are pre-sent in the wine after fermentation (or if sug-ars are added), a sweet wine results. Special-ized techniques for the manufacture of sweetwines will be described later.

For red wines, much (if not all) of the fer-mentation occurs in the presence of the po-mace. When the extraction of pigments, tan-nins, flavor compounds, and other materials isconsidered sufficient, or when the desiredethanol concentration is reached, the free runjuice is separated from the pomace and movedinto another tank. The fermentation is thencompleted (if not already). In the meantime,the pomace is pressed and is either fermentedseparately from the free run or is mixed withthe free run for the final fermentation. Sincethe pressed wine is rich in pigments and tan-



nins, adding a portion back to the free runwine makes the final product richer in colorand flavor.

Yeast Metabolism

It should be evident by now that the main jobof the yeasts during wine manufacture is to pro-duce ethanol from the sugars present in thejuice. However, if ethanol was the only productformed and if sugars were the only substratesmetabolized by the yeast, then wine flavor andaroma would be sorely lacking. In fact, yeastgrowth and fermentation results in a myriad ofmetabolic end products that contribute, for bet-ter or worse, to the organoleptic properties ofthe finished wine.

The main wine yeast, S. cerevisiae, as wellas other species found in natural fermenta-tions, are facultative anaerobes.They have thefull complement of enzymes necessary to oxi-dize sugars via the Kreb’s or tricarboxylicacid (TCA) cycle, as well as enzymes of theEmbden-Meyerhoff (EM) glycolytic pathway.Metabolism of sugars by wine yeasts is not,however, simply a choice of the TCA versusthe EM pathway, but rather depends on theenvironmental conditions, particularly the re-dox potential or Eh.

At the start of the wine fermentation, the Eh of the must is positive, and metabolism byeither the natural flora or added culture organ-isms is mostly by the TCA cycle, which isstrongly favored under aerobic conditions.Sugar metabolism by the TCA cycle yields CO2

and H2O as end products, with a large increasein cell mass.The consumption of oxygen andproduction of CO2, however, quickly reducesthe Eh such that anaerobic conditions eventu-ally develop. Metabolism then shifts to the gly-colytic pathway, yielding equimolar concen-trations of ethanol and CO2, and a relativelymodest increase in cell number.Both pathwaysalso generate ATP, but aerobic metabolismyields a net of 36 ATP molecules per glucosecompared to only 2 molecules of ATP madeper glucose during anaerobic glycolysis.A thirdroute for sugar metabolism, the pentose phos-

phate pathway, also exists, but it is used mainlyin the anabolic direction, for biosynthesis ofpentoses.

The biochemical details for sugar metabo-lism by S. cervisiae are now well known. Asnoted above, environmental conditions dictatewhich metabolic pathway is chosen. Regula-tion occurs both at the enzyme level, as well asgenetic level. Moreover, the absolute require-ment for maintaining redox balance within thecell also has a major influence on specificmetabolic reactions and the end products thatare formed.

For both the TCA and EM pathways, glucoseand fructose metabolism begins with theirtransport into the cell via specific transportsystems. For glucose, at least two such systemsexist. One is a low affinity (high Km), facili-tated transport system, driven simply by theconcentration gradient. It operates when theglucose concentration is high (i.e., at the be-ginning and during much of the fermentation).A second system has a low Km (high affinity)for glucose, is energy-requiring, and functionsonly at low substrate concentrations (i.e., atthe end of the fermentation, when most of theglucose is depleted).

Next, the accumulated glucose is phospho-rylated by one of several kinases to form glucose-6-phosphate, which is then isomer-ized to fructose-6-phosphate. In the case of fruc-tose, transport is also via a facilitated system(not energy-requiring), and the intracellularfructose is directly phosphorylated to fructose-6-phosphate by a hexokinase. The fructose-6-phosphate is phosphorylated a second time toform fructose-1,6-bisphosphate (FDP), which issplit by an aldolase to form triose-phosphates,and eventually pyruvate is formed.

During the glycolytic pathway, a net of twomolecules of ATP are synthesized per hexoseby the substrate level phosphorylation reac-tions. Since the glyceraldehyde-3-phosphatedehydrogenase reaction generates NADH fromNAD, there must be some means of replenish-ing the oxidized form of NAD. Otherwise, notenough NAD would be available for this reac-tion, and metabolism would bottleneck. Under



aerobic conditions, the pyruvate is decarboxy-lated by pyruvate dehydrogenase and the prod-uct, acetyl CoA, feeds directly into the TCA cy-cle. The latter pathway generates reducingequivalents in the form of NADH and FADH,which are subsequently oxidized via the elec-tron transport system.The electron recipient ismolecular oxygen, NAD is re-formed, and ATPis subsequently synthesized by oxidative phos-phorylation via the ATP synthase reaction.When all is said and done, each molecule ofglucose yields 36 of ATP.

In contrast, during anaerobic metabolism,the enzymes of the TCA cycle are not ex-pressed because the absence of oxygen causesrepression of the genes that code for those en-zymes. However, since glucose and fructoseare still metabolized to pyruvate by the gly-colytic pathway, generating NADH, an alterna-tive means of regenerating NAD is necessary.Recall that lactic acid bacteria, facing the samesituation, use pyruvate itself as the electron ac-ceptor, generating lactic acid via the enzymelactate dehydrogenase. The route taken by S.cerevisiae involves two reactions. In the first,pyruvate is decarboxylated via pyruvate decar-boxylase, in a reaction that requires the co-factor thiamine pyrophosphate.Carbon dioxideand acetaldehyde are formed. The acetalde-hyde then serves as the electron acceptor, in areaction catalyzed by NADH-dependent alco-hol dehydrogenase, resulting in production ofethanol and NAD.No additional ATP is made bythese two reactions, indicating their function issolely to replenish NAD.

Factors Affecting Yeast Metabolism

S. cerevisiae has the genetic capacity to metab-olize sugars via either the glycolytic or respira-tory (i.e., TCA) pathways. Although oxygenavailability affects expression of genes encod-ing enzymes of these two pathways and is,therefore, an important determinant of whichway metabolism will occur, gene expression is also regulated by substrate availability. Al-though one might expect that in the presence

of oxygen,metabolism would always be via therespiratory pathway, this is not the case. If theglucose concentration is sufficiently high, me-tabolism will be fermentative. This is becausetranscription of catabolic genes, includinggenes coding for some of the TCA enzymes, isrepressed by glucose, a phenomenon knownas the glucose or Crabtree effect.Thus, in thewine fermentation, where sugar concentra-tions are especially high, metabolism of glu-cose yields mostly ethanol and CO2, and oxida-tive metabolism of sugars is unlikely to occur.Only when the substrate concentration is low(less than 2 g/L), will O2 repression of glycoly-sis occur (the so-called Pasteur effect).

Since the wine fermentation, as notedabove, is anaerobic, one might reasonably ex-pect that the only end products formed fromglucose and fructose would be ethanol andCO2. If that were the case, then how couldone explain the appearance of glycerol, suc-cinic acid, acetaldehyde, acetic acid, and otherproducts, that appear during the wine fer-mentation? In addition, some of the glucosecarbon (albeit only about 1%) is used to formbiomass (i.e., cell constituents). Moreover,even after accounting for evaporative effects,the theoretical 50% yield (on a weight basis)of ethanol from glucose is never reached dur-ing the fermentation, due to byproduct forma-tion. These byproducts can account for asmuch as 4% of the total products formed.Synthesis of these byproducts occurs, in part,in response to demands on the cell to main-tain Eh balance and to salvage ATP. For exam-ple, when the demand for NAD is high, a por-tion of the dihydroxyacetone phosphateformed from FDP via the aldolase reaction isreduced to glycerol-3-phosphate, and NAD isgenerated. The glycerol-3-phosphate is subse-quently dephosphorylated to glycerol.

Aerobic metabolism,at least at the very startof the fermentation, may also result in prod-ucts other than ethanol, in particular, TCA in-termediates, such as succinic acid. Even whenconditions are anaerobic, TCA products maystill be formed to provide the cell with carbon



skeletons (e.g., �-ketoglutarate) necessary forbiosynthesis of amino acids.Another factor in-fluencing sugar metabolism is SO2, added tocontrol the indigenous yeast population. Sulfurdioxide can bind acetaldehyde, preventing itsreduction to ethanol by alcohol dehydroge-nase.As a result, NADH accumulates inside thecells and is diverted to other NAD-generatingreactions, forming glycerol and other non-ethanol end products. Finally, end product for-mation is also influenced by the yeast strainsnaturally present or added to the must.

Sulfur and Nitrogen Metabolism

Although metabolism of carbohydrates is obvi-ously critical to the outcome of the wine fer-mentation, metabolism of other must compo-nents is also important. How wine yeastsmetabolize sulfur-containing compounds thatare present in the must as normal grape con-stituents is particularly important. Most of thesulphur in grapes is in the form of elementalsulfur, sulfates, or as sulfur-containing aminoacids. Since the range of sulfur-containingmetabolic end products includes various sul-fides, mercaptans, and other volatile com-pounds, sulfur metabolism can have a pro-found influence on wine quality.Yeasts can alsoproduce sulfites, which, as already mentioned,have antimicrobial activity. In fact, even if sul-fur dioxide or sulfite salts are not intentionallyadded, wine invariably contains sulfite due toits production by wine yeast.

Grapes usually contain sufficient ammonia,ammonium salts, and free amino acids to sup-port good growth of most wine yeasts. In addi-tion, wine yeasts can synthesize amino acidsand purine and pyrimidine nucleotide fromammonia, and, therefore, have no essential re-quirement for amino nitrogen or pre-formednucleotides. During the course of the fermen-tation, total nitrogen decreases. However, nitro-gen-deficient grapes can result in an inade-quate amount of nitrogen in the must, and, ifnot supplemented with ammonium salts, canlead to reduced fermentation rates.

Stuck Fermentations

Despite the apparent simplicity of the wine fer-mentation, as evidenced by so many successfuloutcomes, there are occasions when the fer-mentation fails. Such wines typically contain asignificant amount of residual, unfermentedsugar and an insufficient concentration ofethanol. The fermentation may actually still beoccurring, just more slowly, or it may be at acomplete standstill. These slow or halted fer-mentations are referred to as sluggish or stuckfermentations,respectively.Although they occurinfrequently, they represent a significant prob-lem and source of economic loss for the wineproducer, since wine quality will inevitably bepoor. Stuck wine is especially susceptible tospoilage, since the low ethanol concentrationand the availability of fermentable sugar maypromote growth of undesirable bacteria andyeasts. Under more severe conditions, the winemay simply have to be discarded.Therefore, de-spite their rare occurrence, it is important to un-derstand the causes and to know how to pre-vent sluggish or stuck wine fermentations.

Among the possible causes of a stuck fer-mentation are those that are due to the mustcomposition, the handling of the must andwine, or the presence of wild yeasts that inhibitdesirable wine yeast.The must may contain, forexample, an insufficient level of nitrogen orother nutrients necessary to support adequateyeast growth. The sugar concentration in thegrapes or must may be too high, resulting in os-motic pressures that inhibit the yeasts. Someyeasts also are inhibited by high ethanol con-centration.As noted previously, the ethanol fer-mentation is exothermic,and if the temperatureis not controlled or cooling is inadequate,the re-sulting high temperature (�30°C) may causethe fermentation to come to an abrupt halt. Incontrast, too cool an incubation temperature(e.g.,�10°C),as might occur during white wineproduction, can also result in a stuck fermen-tation. Most of these situations, however, are easily corrected, either by supplementing thejuice with appropriate yeast nutrients, using



osmotolerant or ethanol-tolerant yeast strains,orby proper temperature control.

Finally, some wild yeast strains secrete pro-teins called killer toxins that inhibit or killother indigenous or starter culture yeasts.There are more that five types of killer toxins,but the most common are K1 and K2. Theseproteinaceous toxins first attach to cell wall re-ceptors, then integrate and form pores withinthe cytoplasmic membranes of sensitive cells,thereby disrupting ion gradients and interfer-ing with energy-transducing reactions. Pro-ducer strains resist the toxin they produce,butare sensitive to those toxins produced byother strains. Some yeast strains do not pro-duce killer toxins, but are nonetheless resis-tant. Several yeast genera, including species ofHansenula, Pichia, and Saccharomyces, areable to produce killer toxins. Importantly, so-

called killer yeasts are found throughout thewine environment, and even many of theKloeckera and S. cerevisiae strains isolatedfrom natural wine fermentations have the“killer” property. In fact, since yeast strainswith killer toxin activity can potentially inacti-vate competing strains, this trait may be desir-able for yeast starter cultures (Box 10–6). Cer-tainly, starter culture strains that are immune tothese toxins would not be affected by otherkiller strains and would not be the cause of astuck fermentation.

Adjustments, Blending, andClarification

After the fermentation is complete, the winewill contain little or no sugar and about 12% to14% ethanol. Still, because of differences in

Box 10–6. Killer Yeasts

Wine spoilage is often mediated by growth of wild yeasts that contaminate wine while it isaging in wooden barrels. In particular, species of the Brettanomyces and Dekkera groups arefrequently involved in wine spoilage and are difficult to control.This is because these yeastsare naturally present not only on grapes and in musts, but also in the barrels in which thewine is aged (Comitini et al., 2004).

Growth of Dekkera and Brettanomyces may result in formation of ethylphenolic compoundsthat have unpleasant, “barny” or “wet-dog” off-odors. They are also a possible cause of the“mousy” off-odor defect (the smell of which, one can imagine). Control measures ordinarily in-clude attention to cleaning and sanitation at the harvest and crushing steps and adequate sulfit-ing of the grapes. It is more difficult, however, to control Brettanomyces and Dekkera once thewine reaches the wooden barrels, which cannot be effectively sterilized.

Many yeast strains can produce and secrete exotoxins that can kill other yeast cells.There arenumerous differences, depending on the producer, between the many so-called killer toxinsthat have been characterized (Magliani et al., 1997).They vary with respect to their genetic ba-sis, how they are synthesized and exported, their size and structure, and their mode of action.Genes encoding for killer toxins, for example, may be located on linear plasmids, nuclear generegions, or double-stranded, virus-like RNAs. In addition, the killer activity in sensitive cells maybe due to pore formation and ion leakage, the arrest of G1 or G2 phases of cell growth, and theinhibition of cell wall synthesis (Magliani et al., 1997).

Among the yeasts that are known to produce killer toxins are several genera associated withthe wine fermentation, including Saccharomyces, Kloeckera, Pichia, and Kluyveromyces. Al-though killer yeast strains are immune to the cognate toxin they produce, they may be sensitiveto toxins produced by other strains. If wild killer yeast strains are present during the wine fer-mentation, and the wine starter yeast (whether part of a pure starter culture or are naturally-occurring) are susceptible to that toxin, a failed or stuck fermentation may result.Thus, it is es-sential that the killer sensitivity phenotype for wine cultures be known, since, in most naturalwine fermentations, killer yeasts can be found (Vagnoli et al., 1993).



In contrast to the detrimental effects killer yeasts may have during wine fermentations, theycan also be exploited in wine fermentations. In fact, killer yeasts are now routinely used in thewine industry, just as they have been used in other industrial processes to control wild yeasts.Killer toxins have also been shown to be effective in clinical applications to treat fungal andbacterial infections (Buzzini et al., 2004; Conti et al., 2002; Palpacelli et al., 1991;Walker et al.,1995).Yeast cultures with a killer phenotype have the potential to dominate a fermentation.Provided that the killer strains have other desirable wine-making traits, the use of such cultureswould be expected to produce a consistent quality wine, due to the inhibition of wild yeastcompetitors. Killer toxin genes can be introduced into non-killer wine-making strains by mat-ing, protoplast fusion, transformation, or micro-injection techniques.

A wine preservation strategy based on the application of killer toxins also was recently re-ported (Comitini et al., 2004b).As noted above, spoilage of wine by Brettanomyces and Dekkeraduring barrel aging is a serious and difficult-to-control problem. In the study by Comitini et al.(2004b), however, it was shown that these yeasts can be controlled by specific killer toxins.Twotoxins in particular, Pikt and Kwkt (produced by Pichia anomala and Kluyveromyces wicker-hamii,respectively), inhibited species of Brettanomyces and Dekkera.Their activities were stablewithin a typical wine pH range of 3.5 to 4.5.When the producer strains were grown in co-culturewith Dekkera bruxellensis (at a 1:10 ratio), growth of the latter was delayed; however, when in-oculated at higher rate (1:1),complete inhibition occurred.Finally, these researchers showed thatdirect addition of the killer toxins to wine effectively inhibited growth of D. bruxellensis for atleast ten days.

Because not all killer toxin-producing yeasts grow well in wine (e.g.,P.anomala and K.wick-erhamii), and it is unlikely that killer toxin preparations could be added directly to wine, othermeans of applying killer toxin technology must be considered. As noted above, killer toxingenes could be genetically introduced into other wine-adapted yeast strains, but it would benecessary for these strains be present during the aging step. It is also conceivable that the tox-ins could be used to treat the interior surfaces of wine barrels and thereby inactivate the Bret-tanomyces and Dekkera that reside there.

References

Bizzini,P.,L.Corazzi,B.Turchetti,M.Bratta,and A.Martini.2004.Characterization of the in vitro antimycoticactivity of a novel killer protein from Williopsis saturnus DBVPG 4561 against emerging pathogenicyeasts. FEMS Microbiol. Lett. 238:359–365.

Comitini, F., N. Di Pietro, L. Zacchi, I. Mannazzu, and M. Ciani. 2004a. Kluyveromyces phaffii killer toxin ac-tive against wine spoilage yeasts: purification and characterization. Microbiol. 150:2535–2541.

Comitini, F., J.I. De, L. Pepe, I. Mannazzu, and M. Ciani. 2004b. Pichia anomala and Kluyveromyces wicker-hamii killer toxins as new tools against Dekkera/Brettanomyces spoilage yeasts. FEMS Microbiol. Lett.238:235–240.

Conti, S.,W.Magliani, S.Arseni,R.Frazzi,A.Salati, L.Ravanetti, and L.Polonelli.2002. Inhibition by yeast killertoxin-like antibodies of oral streptococci adhesion to tooth surfaces in an ex vivo model. Mol. Med.8:313–317.

Magliani,W., S. Conti, M. Gerloni, D. Bertolotti, and L. Polonelli. 1997.Yeast killer systems. Clin. Microbiol.Rev. 10:369–400.

Palpacelli,V., M. Ciani, and G. Rosini. 1991.Activity of different ‘killer’ yeasts on strains of yeast species un-desirable in the food industry. FEMS Microbiol. Lett. 84:75–78.

Vagnoli, P., R.A. Musmanno, S. Cresti,T. di Maggio, and G. Coratza. 1993. Occurrence of killer yeasts in spon-taneous wine fermentation from the Tuscany region of Italy.Appl. Environ. Microbiol. 59:4037–4043.

Walker, G.M.,A.H. McLeod, and V.J. Hodgson. 1995. Interaction between killer yeasts and pathogenic fungi.FEMS Microbiol. Lett. 127:213–222.

Box 10–6. Killer Yeasts (Continued)



grape composition,microflora,and wine manu-facturing practices, variations in wine compo-sition and sensory quality are to be expected.Therefore, adjusting the wine after fermenta-tion (and sometimes before) is a normal step.The pH and acidity, in particular, can varymarkedly,as can the color and flavor.Therefore,some wineries adjust the acidity of wine byacidification or deacidification steps (e.g., byadding acids or neutralizing agents).

When acidity of wine or must is due to malicacid, deacidification is managed via the malo-lactic fermentation,which is discussed in detaillater in this chapter.Adjustment to wine colorand flavor can also be done, if legally permitted,by filtration and enzyme treatments, respec-tively. Filtration techniques, for example, aremost often used to “decolorize”wine by remov-ing undesirable pigments.

Except for very small wineries, which mayhave only a few vats of wine, most modernwineries have many individual vats of wine.Each one is unique, in that a particular vat maycontain wine made from grapes harvested at atime or place different from the grapes in aneighboring vat.Wines within a single winerymay be made from different grape varieties.Therefore, another common procedure, espe-cially for premium wines, is to blend differentwines to optimize or enhance the organolepticproperties. Blending also produces wines withconsistent flavor, aroma, and color from year toyear. Perhaps more so than any other wine-making step,however,blending is a tricky busi-ness,and is a highly subjective process.Successrelies on the imagination, creativity, and skill ofthe wine blending specialist.

At the end of the fermentation, the winecontains non-soluble proteins and protein-tannin complexes, as well as living and deadmicroorganisms.These materials give the winea cloudy, hazy, undesirable appearance. Theclarification step removes these substancesfrom the wine without removing desirable fla-vor and aroma components. It is particularlyimportant that the cells are removed. If left inthe wine, these cells can lyse, releasing en-zymes that may catalyze formation of off-

flavors and odors (although an exception tothis rule exists, as described below). Inducingprecipitation of tartrate salts and tannin-protein complexes is also commonly done tofacilitate their removal before they precipitatelater during aging. It is important to recognizethat, in some cases, cell lysis may be a goodthing. Intracellular constituents released dur-ing cell lysis include amino acids and nu-cleotides, providing nutrients that are laterused by bacteria in secondary fermentations orthat contribute to the sensory properties.

According to traditional practices, wine isclarified by simply allowing the sediment, con-taining the yeasts and bacterial cells, as well asprecipitated material, to settle naturally in bar-rels or vats.The wine could then be removedfrom the sediment (or “lees”) by decantation.This process, called “racking”, is usually donefor the first time after three to six weeks fol-lowing the end of the fermentation. Rackingcan be repeated several times over a period ofweeks or months until the wine is nearly crys-tal clear. During the racking step, the wine isalso aged (see below). Racking can now bedone in enclosed tanks using automated trans-ferring systems.

Filtration is another method used to clarifywine.This can be especially effective if finingagents, such as bentonite, albumin, or gelatin,are used as filtration aids. If micropore filtra-tion membranes are used, it is even possible tosterilize wines. Clarification may occur afterracking or after aging.

In contrast to removing the sediment shortlyafter the fermentation has ended, some winesare intentionally left in contact with the lees foran extended time before the first racking oc-curs.This traditional maturing practice, knownas “sur lies,” enhances the flavor, character,mouth feel, and complexity of the wine. It ismore common for white wines than red.

Aging

Aging actually begins just after fermentation.Thus, aging occurs when the wine is racked,as well as beyond.Aging conditions vary con-



siderably. Some wines are aged for severalyears, whereas others are “aged” for only a fewweeks. Some wines are aged in expensive oakbarrels, others in stainless steel, and yet othersdepend on bottle-aging, or a combination ofall of the above.Whether a wine is aged for along time in oak barrels or is quickly bottledand sent to market depends, in part, on mar-keting considerations, but also on the originalcomposition of the grapes and how they aremade into wine. Thus, long, careful agingshould be reserved for only premium winesmade from high quality grapes. By analogy,Cheddar cheese manufactured for the processcheese market cannot be expected to developinto a flavorful, two-year Cheddar, no matterhow carefully it may have been aged.

Of course, some excellent quality wines,like excellent cheeses, are meant to be con-sumed in a “fresh” or un-aged state, so whetheror not a wine is aged does not distinguish winequality, per se. For example, Beaujolais nou-veau, a popular wine from the Burgundy areain France, is meant to be drunk after only a fewweeks after the grapes are harvested. Thesewines are fruity and “gulpable”; no amount ofaging will lead to their improvement.

To say that the actual events that occur dur-ing aging are complicated would be quite theunderstatement. Hundreds of enzymatic, mi-crobiological, and chemical reactions occur,and as many as 400 to 600 volatiles, includingesters, aldehydes, higher alcohols, ketones,fatty acids, lactones, thiols, and other com-pounds are formed (Table 10–4).The wine in-teracts with the wood and wood constituents

in the barrel, oxygen in the air, and even thecork. It is important to recognize that not all ofthese reactions are beneficial in terms of winequality, and some wines may actually deterio-rate during aging. In fact, long aging is notgood for most wines.

Ordinarily, in large wineries, fermentationsoccur in large tanks (exceeding 250,000 liters),and then the wine is moved into wooden 200liter barrels for aging.However wine can also beaged, at least initially, directly in tanks, and thenlater moved into oak barrels for final aging.

In many European and other traditionalwineries, in contrast, the entire aging period isconducted in oak barrels. The oak barrel or“cooperage” is so important to wine qualitythat entire industries devoted to oak tree pro-duction and cooperage construction have de-veloped.This is because the oak barrels are notinert containers used simply to store wine, butrather they are a source of important flavorand aroma compounds. In fact, one of the ma-jor steps in barrel construction involves heat-ing or “toasting” the barrels to promote pyro-lysis. This generates a number of flavor andaroma volatiles. In the presence of wine, thesecompounds, along with tannins, phenolics,lignins, and lactones, are extracted from thewood and solubilized in the wine. Some ofthese compounds impart unique flavor notes,including vanilla and coconut. Aging wine inoak cooperage is not, however, without adownside.Oak barrels are expensive, and,evenif carefully maintained, do not last forever. Lossof wine volume (and hence profit) due toevaporation can also occur. Thus, alternative

Table 10.4. Effects of aging on wine.

Reaction or step Effects

Tannin precipitation Color darkens; astringency increases initially,then decreases

Wood cooperage Phenolic and other flavors extractedEster hydrolysis Fruitiness decreasesOxidation Browning and flavor reactions inducedEvaporation Concentration of nonvolatile solutes; color and flavor intensifies,

but aroma volatiles decrease



materials, in particular, stainless steel, have dis-placed oak cooperage in many wineries.Whileit certainly does not contribute flavor andaroma compounds, stainless steel is less expen-sive, easier to clean and maintain, and can befabricated to accommodate size and shapepreferences. It is still possible for wine aged instainless steel cooperage to obtain desirableoak-derived flavors by adding oak shavings orchips to the aging wine.

Malolactic Fermentation

A certain amount of acidity is expected and de-sirable in wine. Red wines typically have a pHof 3.3 to 3.6; white wines are usually slightlymore acidic. Some grapes, and the musts madefrom those grapes, however, may contain highlevels of organic acids, such that the pH is toolow (i.e., �3.5).Wines made from those grapeswill suffer from excess acidity, a serious andreadily noticeable flavor defect.

Of the organic acids ordinarily present ingrapes, malic acid is particularly important be-cause of its ability to influence pH.This is be-cause malic acid is a four carbon dicarboxylicacid, meaning it contains two carboxylic acidgroups and can release or donate two protons.Thus, musts containing high concentrations(0.8% to 1.0%) of malic acid are acidic andhave a low pH. High malic acid concentrationsare especially common in grapes grown incooler, more northern climates, such as thosein Oregon, Washington, northern California,and New York.Although many of the vineyardsin Europe are located in warmer regions andproduce grapes with less malic acid (a situa-tion that may lead to the opposite problem—too little acidity), grapes from Germany,Switzerland, and even some regions in Francecan still contain significantly high malic acidlevels. Also, some grape cultivars ordinarilycontain more malic acid than others.

One way to reduce the malic acid levels andto “deacidify” the wine is to promote the bio-logical decomposition of malic acid. Thisdeacidification process occurs via the malolac-tic fermentation pathway that is performed by

specific species and strains of lactic acid bacte-ria.These bacteria may be naturally present inwine and may, therefore, initiate the fermenta-tion on their own.It has now become commonto add selected malolactic strains, in the formof a pure culture, directly to the must.

The malolactic fermentation has been thesubject of extensive research in the last twodecades. The pathway was initially thought toinvolve two separate reactions (Figure 10–4A),in which malate is first decarboxylated by anNAD-dependent decarboxylase yielding pyru-vate and reduced NADH.In the second reaction,pyruvate serves as the electron acceptor yield-ing lactic acid.Is it now known that the reactionis catalyzed instead by a single malolactic en-zyme that decarboxylates malic acid directly tolactic acid (Figure 10–4B).Although this enzymerequires NAD (and manganese), no intermedi-ate is formed.The net effect of the malolactic re-action is that malic acid, a dicarboxylic acid, isconverted to lactic acid,a monocarboxylic acid,thereby reducing the acidity of the wine.

Several lactic acid bacteria have the enzy-matic capacity to perform this fermentation.The most well-studied are species of Oenococ-cus, particularly Oenococcus oeni (formerlyLeuconostoc oenos). Several species of Lacto-bacillus also have malolactic activity.Althoughsome of these bacteria are found naturally inmusts, commercial cultures are now availableand are commonly used.

The malolactic fermentation was one thatinterested microbiologists, not only because ofits industrial importance, but also becausethere seemed to be no obvious reason for bac-teria to perform this conversion. In otherwords, how do malolactic bacteria benefit, ormore to the point, how do they gain energy,from the conversion of malic acid to lacticacid? The pathway, after all, contains no sub-strate level phosphorylation step that wouldlead to ATP formation, nor is there a change inthe redox potential. As it turns out, there is ameans of generating ATP via the malolacticpathway, but it is indirect.

As shown in Figure 10–5, malic acid is trans-ported into the cell via one of two ways. In



Oenococcus, the malate permease is a uni-porter, whereas in lactococci and lactobacilli,uptake of malate is mediated by an antiporterthat exchanges an incoming malic acid for anoutgoing lactic acid. No energy is spent formalate transport in either system. The ex-change reaction, however, is not electroneu-tral.This is because one extracellular moleculeof malic acid, carrying a net electric charge of-2, is exchanged for one of lactic acid that car-ries a net charge of only -1.This charge differ-ence arises as a result of the decarboxylationreaction, which consumes an intracellular pro-ton.The bottom line is that the cell is able toextrude a proton (or its equivalent) withouthaving to spend energy to do so. That is, thecell conserves energy, in the form of ATP, that it

would ordinarily spend to pump protons fromthe cytoplasm to the extracellular medium.Theproton gradient that forms, or the proton mo-tive force, can then either perform other work(e.g., nutrient transport) for the cell or be usedto drive ATP synthesis by the proton-translocat-ing F0 F1 ATPase.

It is important to note that in low-acidgrapes, the malolactic fermentation is undesir-able, since some acidity is desired in wine.Thus, under some circumstances, the presenceof naturally occurring malolactic bacteria is un-desirable and the source of potential defects(see below). In contrast, the malolactic fermen-tation not only is performed for deacidifica-tion, but also to promote flavor stability andbalance. Moreover, malolactic bacteria often

H

H

C

OH

C

H

CHO

OC

OH

OCO2+

NAD NADH + H

NADH + H NAD

A

B

H

H

C

O

C

H

CHO

OH

H

H

C

O

C

H

CHO

OH

H

H

C

OH

C

H

CHO

OH

H

H

C

OH

C

H

CHO

OH

H

H

C

OH

C

H

CHO

OC

OH

OCO2+

malolacticenzyme

Figure 10–4. The Malolactic Reaction. The conversion of malic acid to lactic acid was originally thought tobe catalyzed by two separate reactions, with pyruvic acid formed as an intermediate, and with NAD requiredas a co-factor (A). It is now well established that the malolactic reaction occurs as a single step, with lacticacid formed directly from malic acid via a decarboxylation reaction catalyzed by malolactic enzyme (B). Nointermediate is formed.



produce diacetyl from citrate, which may, atthe appropriate concentration (generally be-tween 1 mg/L and 4 mg/L), be desirable insome wines.

Types of Wine

Aside from the rather broad distinction of clas-sifying wines based on their color—red,white,or rose—there are obviously more descriptivemeans for grouping different types of wine.Similarly, the procedures used for manufactureof wine have, so far,been described in a mostlygeneric manner. In this section, the manufac-ture of several well-known categories or stylesof wine will be discussed, including sweetwines, fortified wines, sparkling wines, and dis-tilled wines.

Sweet wines

Sweet wines are simply those that contain un-fermented sugar (either fructose, glucose, orsucrose).There are several ways to produce asweet wine. The easiest (and least expensive)way is to simply add sugar (up to 2% to 4%) orsucrose syrup to a dry wine. This results in a

wine that is definitely sweet, but not necessar-ily one that would be acceptable to many con-sumers.A similar and more common approachis to add unfermented juice, preferably fromthe same grapes used to make the wine.Alter-natively, sweet wines can be made by stoppingthe fermentation before all of the glucose andfructose have been fermented. In some cases,sugar may be added to the juice prior to fer-mentation, such that when the fermentation iscomplete, residual sugar (and sweetness) re-main. This is one of the more common prac-tices in the United States, and many of thesweet wines from New York state are pro-duced this way.

In either case, arresting the fermentation isthe key step. Generally, this is done by rapidlycooling the wine, and then filtering out theyeast. One recent innovation is to encapsulatethe yeasts within alginate beads, which areplaced inside permeable bags. At the desiredtime, the bag (containing the yeast) is simplyremoved (see Chapter 3 for encapsulated cul-tures). Culture activity can also be stopped byadding alcohol spirits, as is the case for forti-fied wines (see below). Finally, sweet winescan be manufactured via one of several tradi-

malate2-

malatepermease

out

in

lactate-

malate2- lactate-

CO2H+

A

ME

malate2-

malatepermease

out

in

lactate-

malate2- lactate-

CO2H+

B

ME

Figure 10–5. Malate transport in wine lactic acid bacteria. Transport of malate by Oenococcus oeni is me-diated via a uniport malate permease (A). The accumulated malate is then decarboxylated directly to lactateby malolactate enzyme (ME); lactate efflux occurs via passive diffusion. The malolactic reaction also consumesa proton, such that the cytoplasm becomes more alkaline and more negatively charged. Thus, the electo-chemical gradient across the membrane (the proton motive force) is increased, and ATP is conserved. In lac-tobacilli and lactococci, a similar process occurs (B), except that malate permease acts as an antiporter, suchthat lactate efflux drives malate uptake. Adapted from Konings, W.N. 2002.



tional techniques.These are based on concen-trating the sugar in the juice or grapes. Thejuice, for example, can be partially concen-trated by heating. In contrast, the grapes can bedried via atmospheric drying, or frozen so thatthe ice can be removed. However, the mostwell-known traditional technique for concen-trating sugars in grapes is to dehydrate thegrapes while they are still on the vine. Theseso-called botrytized wines are made through-out Europe, with the Sauternes, white winesfrom the Loire Valley of France, being the mostwidely prized.

The traditional process for making sweetbotrytized wines starts in the vineyard.Grapes left on the vine past their optimumharvesting will invariably be infected byfungi, including the ubiquitous fungal organ-ism Botrytis cinerea. Depending on climaticconditions, there can be two outcomes to thisinfection. If the climate is sunny and moder-ately dry during the day but humid during theevening, B. cinerea will grow to just the rightextent on the surface of the grape, resulting inwhat is referred to as “noble”rot (Figure 10–6).If conditions are not “just right,” overgrowthof B. cinerea or infection by other fungi canoccur, resulting in simply rotten grapes that

have no value.Thus, the manufacture of thesewines carries considerable risk.

Ultimately, noble rot growth of B. cinerearesults in dehydration of the grapes and con-centration of the sugars. Dehydration occursdue to secretion of pectinolytic enzymes by B.cinerea that degrade the pectin-containingcell walls of the grapes. Water evaporationthen occurs, provided the atmosphere is suffi-ciently dry.The decrease in moisture is impor-tant not only because it concentrates thesolids, but also because it controls growth ofundesirable fungi. Eventually, the mold-cov-ered grapes begin to take on the appearanceof raisins (moldy ones at that). The moisturelevel of the grapes will be reduced by about50% and, even though the mold metabolizessome of the sugars in the grapes, the sugarconcentration in the juice is still increased by20% to 40%.After the grapes are harvested andgently crushed, portions of the free run andthe pressed juice are combined.The fermenta-tion is then generally the same as for otherwhite wines, with a temperature range of18°C to 20°C.At the end of the fermentation,the ethanol concentration will range from 9%to 13%, and the wine will contain as much as10% total unfermented sugar.

Fortified wines

Fortified wines are those to which distilledspirits (containing as much as 95% ethanol) areadded to raise the total ethanol concentrationto 15%.Not only do these wines contain higherconcentrations of ethanol, the source of theethanol (e.g., brandy) is also important sincethey may contribute unique flavor compoundsto the finished product. Aside from this com-mon feature, however, a variety of quite differ-ent fortified wines exist. Included are whitesand reds, dry and sweet. Fortification usuallyoccurs during or just after the fermentation. Insome wines, the added ethanol may inhibit theyeast and prevent the complete fermentationof sugars, resulting in sweet dessert wines.Themost well-known of the fortified wines aresherry and port.

Figure 10–6. Grape clusters infected with Botrytiscinerea, the so-called noble rot. Photo courtesy of DavidMills, University of California, Davis.



Sherry originated in Spain around 200 yearsago, and is now produced in the United Statesand throughout the world.One of the main fea-tures used in the traditional manufacture ofSpanish sherry (but rarely outside of Spain) is ablending technique known as the solera sys-tem.According to this technique, wines of dif-ferent ages are progressively and uniformlyblended to achieve wines of consistent quality.Another unique feature of most types of sherryis the presence of film yeasts that grow on thesurface of the wine while it is aged in specialcasks called “butts.” The S. cerevisiae strainsthat comprise this film, called a flor, arethought to be the same yeasts involved in theprimary alcoholic fermentation,although it hasalso been suggested that a separate group ofyeasts actually comprise the flor yeast popula-tion. In either case, the secondary yeast growthoccurs under aerobic conditions, resulting inoxidation of some of the ethanol and subse-quent formation of unique flavor and aromacomponents (e.g., esters, aldehydes, higher al-cohols, and lactones).

In contrast to the traditional (and costly) sol-era system for manufacture of sherry,other tech-niques are now widely used in the United Statesand other countries. In the submerged flor pro-cedure, the wine is aerated and mixed in fer-mentors to enhance growth of the flor yeast.Baked sherries are made by heating fortified butnon-aged wine to 50°C to 60°C for ten totwenty weeks. A number of flavor and aromacompounds, including acetaldehyde and fur-fural,are generated by baking due to carameliza-tion and non-enzymatic browning reactions.

Port, first produced in Portugal, is anotherfortified wine. It is made by adding distilledwine (i.e.,brandy) to the red base wine, therebyraising the ethanol concentration to nearly 20%.Since fortification occurs well before all of thesugars have been fermented, and this muchethanol inhibits yeast, this step effectively endsthe fermentation.Thus,Port contains as much as10% sugars.Port and other fortified sweet winesmade in this manner (e.g.,Madeira and Marsala)are often referred to as dessert wines. Becausefortification not only arrests the fermentation, it

also reduces the time available for extraction ofanthocyanin pigments and phenolic flavor com-pounds. Mixing pressed wine (pomace) withthe free run wine, therefore, is necessary to pro-vide adequate color and flavor in the finishedwine. In the United States, heat (e.g., steam) issometimes used to extract pigment from thegrapes. Finally, as for sherry, blending and agingare also important steps for port production.Some ports are aged for as many as fifty years.

Sparkling wines

Sparkling wines are those which contain car-bon dioxide, providing bubbles and efferves-cence. For some sparkling wines, CO2 pres-sures as high as 600 kPa atmospheres can bereached (by comparison, the pressure inside acan of soda pop is less than 200 kPa).Althoughsparkling wines are made throughout theworld, there are several manufacturing meth-ods that are used to produce the CO2, andthese methods define, to a certain extent, thetype of sparkling wine being produced.

Clearly, the most well known sparkling wineis Champagne, which is traditionally made, notsurprisingly, via the Champagne method. Forother sparkling wines, CO2 is introduced byother methods, as described below. It is worthmentioning that Champagne, like Parmesanand Roquefort cheeses, enjoys a protected sta-tus in the European Union. Only wine pro-duced in the Champagne region of France—just to the southeast of Paris—and made ac-cording to the centuries-old process and agedin the caves in that region can be called Cham-pagne. All other carbonated wines, regardlessof how CO2 is introduced, must be referred toas sparkling wine. In contrast, wine makers inthe United States, which are under no obliga-tion to follow EU rules, can label any sparklingwine as a champagne.

The manufacture of sparkling wines startsout no differently than other wines.The maindifference is that at some point carbon dioxideis introduced into the “still” (i.e., not carbon-ated) wine.There are four possible ways to per-form this step. The first is to simply add CO2



directly to the wine, followed by bottling.Thisis not unlike soda pop production and is defi-nitely the most economical means.At about $5a bottle, this “champagne,”although bearing lit-tle resemblance to the real thing, is certainlythe favorite of budget-minded celebrants.

The other processes all involve a secondethanolic-CO2 fermentation performed by yeasts(the exception, however, are those wines thatcontain CO2 as a result of the malolactic fermen-tation). However, in contrast to the primary fermentation, which is done in ambient atmos-phere (i.e., open), the second fermentation oc-curs in an enclosed environment such that theCO2 is trapped and becomes supersaturated inthe wine. One seemingly simple way to initiatesuch a fermentation is to stop the primary fermentation before all of the sugar has been fer-mented, and then to bottle the wine and allowthe remaining sugar to be fermented by theresidual yeast. The difficulty with this method is that stopping and starting the fermentation isnot always so easy. Still, this method has longbeen practiced in France,Germany,and Italy.

Rather than rely on the endogenous sugarand yeast from the primary fermentation, analternative and more common means of initi-ating the second fermentation involves the di-rect addition of an additional source or “dose”of sugar, as well as an inoculum of yeast, to thestill wine.The yeast need only ferment a mod-est amount of substrate (usually sucrose) togenerate enough CO2 to make the wine bub-bly. The secondary fermentation can be per-formed either in enclosed vats or directly inbottles. For either method, however, there remains the problem of the residual yeast,which, if not removed, will make the winecloudy or turbid. Sparkling wines, especiallychampagnes, are prized for their absolute clar-ity and perfect appearance.Thus, removing theyeast following the second fermentation is ab-solutely necessary. Several methods have beendevised to perform this step.

In the bulk or Charmat method, the secondfermentation occurs in a pressurized vat, thenthe wine is filtered to remove residual yeastsand filled into bottles under pressure. In the

transfer process the wine is placed into thick-walled bottles, sugar and yeasts are added, andthe bottle is corked.After the second fermenta-tion in the bottle is complete, the wine is trans-ferred into tanks, filtered to remove yeast andthen re-bottled. In the Champagne method, theprocess is similar to the transfer system,exceptthat no transfer occurs. Rather, the second fer-mentation occurs in the bottle, where thewine remains. How then, if the wine does notleave the bottle, can it be clarified and madefree of residual yeasts? The answer to this ques-tion and other unique features of Champagnemanufacture are described below.

Champagne

Champagne is made from Chardonnay, PinotNoir, and Pinot Meunier grapes grown in theChampagne district. This is a northern grape-growing region and the still wines made fromindividual cultivars (the Pinots make red winesand the Chardonnay is used for white) are notparticularly remarkable (some might call theminsipid).However,when the base wines are ap-propriately blended (a skill first perfected bythe monk Dom Pérignon), the wine assumesthe best qualities of each individual cultivar.Manufacture starts, as for other white wines,with a rather fast pressing of the grapes, suchthat pigment extraction is minimized.The firstpress juice, called the cuvée, is then inoculatedwith selected yeasts (these are almost alwaysproprietary strains), and the juice is fermentedat about 18°C in oak barrels (stainless steelvats are also used).Because the base wines canbe rather acidic (�pH 3.0), a malolactic fer-mentation may be desirable and the appropri-ate cultures can be added (or growth of indige-nous malolactic strains encouraged). However,promotion of the malolactic fermentation inChampagne is not a universally accepted prac-tice, because some bacteria-generated flavorproducts may affect the delicate flavor andaroma balance of the finished wine.

After fermentation to about 12% ethanol,the critical blending step is performed (usinghigh quality wines, free of defects). Base wines



from a single year are blended to create vintageChampagne, whereas wines from differentyears are more often used to create non-vintage Champagne. Next, the blended wine isbottled, the “tirage”or sugar solution (about 20g to 25 g of sucrose per liter) and a yeast mix-ture is added, and a cork closure is inserted.Afastener or clamp is placed over the cork toprevent its displacement.The bottles are thenheld at about 10°C to 12°C.The yeast inoculumfor the secondary fermentation usually con-tains strains of S. cerevisiae (formerly Saccha-romyces bayanus) that are selected based ontheir ability to grow at high ethanol concentra-tions and low pH and temperature.The bottlesare also constructed differently from ordinarywine bottles (thicker, with a tapered neck),since they must be able to withstand the highCO2 pressure.

When the secondary fermentation is com-plete (after about seven weeks), the wine maybe allowed to age in the bottles for up to threeyears.Aging or maturation occurs while the leesare still present in the wine, a critical step thatdistinguishes Champagne from other sparklingwines fermented by the bulk method. Next, thetime to remove the yeast cells (most of whichare now dead) and sediment material has ar-rived. This step, called riddling (or remuage),involves gradually shifting the position of thebottle, starting from near horizontal and endingup near vertical.At the end of this daily or every-other-day jolting, turning, and twisting process(about one to three months), the sediment willhave collected within the neck of the bottle,specifically settling on the inside of the cork.Although the riddling process is still done man-ually in several of Champagne houses, mechani-cal riddling is now common.

Once the sediment has settled on the cork, itis removed by a process called disgorgement.First, the now vertical bottles (neck down) arecooled to about 7°C, then conveyed through afreezing solution (at �20°C) such that the mate-rial in the neck, not more than about 2 cmabove the cork,is quickly frozen.The bottle is in-verted to a 45° angle (neck up), and the frozen

plug, containing the cork, sediment, a smallamount of slushy wine, is removed.About 10 to50 ml of a sugar-wine solution (called thedosage) is then quickly added to replace thewine lost in the disgorgement process, and afresh cork is applied and fastened with a wire“crown.” The amount of sugar in the dosage,from less than 1.5% to more than 5%, deter-mines whether the champagne is dry (brut),medium-dry (sec) or sweet (doux).The bottle isgently shaken to mix the wine and dosage, andthe bottles are then stacked horizontally for upto three months. Champagne does not improvemuch beyond this aging period, and is essen-tially ready for consumption.

Brandy

Brandy is produced by distilling wine.The winecan be made from other fruits, but when madefrom grapes, white wines are used as the base.The most well known brandy is Cognac, madefrom the Cognac district of France. In theUnited States, brandy must conform to a stan-dard of identity that describes the starting fruitor juice, the ethanol concentration, the durationof aging, and other compositional and manufac-turing details. Most American beverage-typebrandies contain less than 50% ethanol (100°proof). In contrast, brandy used for fortificationpurposes usually contains 70% to 95% ethanol(140° proof to 190° proof). Due to evaporation,long aging can substantially reduce the ethanolconcentration.

In general, the base wine production fol-lows that for other white wines. Skins andseeds are removed immediately after crushingand pressing to minimize pigment extraction.Distillation can occur as soon as the wine isfermented. Two types of distillers or stills areused, pot stills and continuous stills. The potstill, used for Cognac, consist of a single pot orboiler that is directly heated. A portion of thevapor is condensed in a reflux condenser andreturned to the boiler, and the remaining vaporcondenses and is collected. After distillation,



the brandy is aged in oak barrels, which resultsin extraction of oak flavor, aroma, and colorcompounds. Most American brandies are agedfor two to three years; however, some Cognacsand other premium brandies are aged for morethan twenty years.

Wine Spoilage and Defects

Wine making is fraught with risks. It can rea-sonably be said that as many things that can goright during wine manufacture, just as manycan go wrong. The risk is exacerbated by theconsiderable investment that must be made toproduce the wine (thus the old joke:How doesone make a million dollars in the wine busi-ness? Start with $2 million). Grapes must becultivated over several seasons before a reason-able crop can be harvested.Disease,climate, in-sects, and other factors can cause serious prob-lems even before the first grape has beenharvested and crushed. Once wine making be-gins, growth of desirable yeast and bacteriaand inhibition of others is not always easy tomanage. Finally, aging of wine may result in aproduct ranging from truly spectacular to to-tally undrinkable.

Spoilage of wine, like other fermented (andeven non-fermented) foods, is often due tochemical and physical activities.Oxidation reac-tions, whether induced by oxygen, sunlight, ormetals, can be especially damaging to wine,leading not only to development of off-flavorsand aromas, but also to discoloration and pig-mentation defects. High temperatures are alsodetrimental to wine quality, catalyzing chemicaland biochemical reactions that result in car-amelization, browning, and oxidation reactions.Physical phenomena, such as precipitation ofproteins and tannins, are responsible for cloudi-ness and haze formation. A particular type ofhaze, called casse, occurs in white wines and iscaused by precipitation of metallic salts.Despitethese various chemical-physical defects, how-ever, wine spoilage is most commonly causedby microorganisms. In fact, fungi, yeasts, andbacteria can all be responsible for spoilage at

some point during the wine manufacturingprocess.

Spoilage by fungi

Fungal growth and spoilage rarely occurs dur-ing the wine fermentation, since most fungiare aerobic and sensitive to ethanol. Rather,molds are most important before and after thewine is made. Fungal growth on grapes is oneof the most serious problems encountered ingrape viticulture, causing considerable loss ofcrop. If not controlled, rots, mildews, and otherfungal diseases can wipe out an entire vine-yard. As noted earlier for sweet botryizedwines, a fine line sometimes separates spoiled,rotten grapes from desirable, noble rot growthof B. cinerea. Some fungi, such as Penicillium,Aspergillus,Mucor,and Rhizopus,can grow onfreshly harvested grapes during transport tothe winery. Pesticides, sulfiting agents, andother antimycotics can be applied to help con-trol this problem, but care must be exercisedto minimize their impact on the wine and dur-ing fermentation.

Post-fermentation problems with mold areusually due to contaminated cork closures.Cork, the bark from the cork tree, can con-tribute so-called cork taints. This is one of themost serious defects in bottled wine and onewhich has attracted wide attention among corkproducers, wineries, and consumers (Box 10–7).The defect is now thought to occur as a re-sult of growth of various fungi (including Peni-cillium, Aspergillus, and Trichoderma).Visiblemold growth is rarely evident (those corkswould not be used), so cork taints occur earlierin the cork making process, when exposure tofungi is common. These fungi then producemusty- or mushroom-smelling compounds thatdiffuse into the wine after the corks are insertedinto the bottles.Wine corks can now be treatedto remove the offending taint. However, therehas been a strong trend toward the use of plas-tic and other alternative wine enclosures, theuse of which may significantly reduce the inci-dence of cork-related defects.



Box 10–7. Cork Taints in Wine: Are Corks the Cause?

Imagine you’ve just purchased an expensive bottle of wine and are anticipating the fine bou-quet about to be released from the soon-to-be opened bottle. Upon removing the cork, how-ever, you discover that the wine has an obnoxious musty and moldy aroma that renders thewine undrinkable.As unfortunate as this scenario might be, it is by no means rare. In fact, corktaint, the most likely cause of this particular defect, is considered to be the most common causeof wine spoilage, affecting as much as 12% of all wine (this spoilage rate has been disputed,however, as discussed below).

How often consumers actually return or even reject cork-tainted wine is not known, but itseems likely they will remember their experience with that particular brand when they pur-chase their next bottle of wine. Not surprisingly, given the frequency with which this problemoccurs, cork taint has become one of the most important (and perhaps most controversial) is-sues facing the wine industry.Annual economic losses of more than $10 billion have been esti-mated (Peña-Neira et al.,2000).Therefore,understanding what causes cork taint and developingstrategies to prevent it are the focus of considerable academic and industry-sponsored research.

Cork technology and microbiology

Corks have been used as closures for wine bottles at least since the 1600s. Not only are corkstoppers effective at preventing leaking, they also permit a very small amount of gas exchangethat is believed to promote flavor and aroma development in aged wine. Cork is obtained fromthe bark of the Quercus suber oak tree, which grows in regions around the western Mediter-ranean Sea. Portugal is the main producer of cork (and has the biggest stake in the cork taintproblem).

The manufacture of cork follows a traditional process that starts with its harvest (Jackson,2000).The first,or virgin bark,stripped from the tree when it is about twenty to thirty years old,is not used; neither is the second growth obtained after another seven to ten years. Bark is notcork-worthy for another nine years, when it is called reproduction bark (subsequent strippingscan occur every nine years).The harvested cork slabs are then boiled, cut into strips, and thecylindric cork units are punched out and washed in water.Next, the corks can be treated eitherwith a hypochlorite solution or with peracetic acid. Both of these treatments inactivate mi-croorganisms, although they are not without important side reactions (see below).The corksare dried, sorted, and packaged.The name of the winery or vintage can also be etched onto thecorks.

Among the microorganisms associated with cork, fungi, and Penicillium species, in particu-lar, appear to be the most common (Lee and Simpson,1993).Other fungi include Trichoderma,Aspergillus, Mucor, and Monillia.Yeast may also be present.The boiling and bleaching steps ef-fectively reduce the microbial load.However, fungi growth is possible prior to these steps or af-ter, if re-contamination occurs. In fact, mold growth on cork has long been thought to be nec-essary for cork maturation and is, therefore, encouraged as part of traditional cork processing(Silva Pereira et al., 2000). Still, fungi have long been suspected of being involved in wine taintformation, although direct evidence establishing this link has only recently emerged.

Cork Taints

The actual chemical agent responsible for cork taint has been identified as 2,4,6-trichloroanisole or TCA (Figure 1). Not only does TCA have a musty, disagreeable odor, but itsthreshold for detection is extremely low. Sensitive tasters can detect TCA at levels as low as afew parts per trillion, although for most individuals, higher concentrations must be present be-fore TCA is noticed (Peña-Neira et al.,2000).Although other chloroanisoles,as well as other phe-nol- and pyrazine-containing compounds, may also contribute to cork taint (Peña-Neira et al.,2000; Simpson et al., 2004),TCA is invariably present in tainted wine (Coque et al., 2003).



Biochemical evidence suggests that TCA is synthesized from chlorophenol precursors by amethylation reaction (Álvarez-Rodríguez et al., 2002). Chlorophenols are highly toxic, and thismethylation reaction has been thought to be involved in detoxification (TCA, although odor-ous, is non-toxic). It has recently been reported that cork-associated fungi can perform this re-action and produce TCA in cork, provided 2,4,6-trichlorophenol was present as a substrate(Ílvarez-Rodríguez et al., 2002). In this particular study, a strain of Trichoderma longibrachia-tum produced the highest level of TCA (nearly 400 ng/g of cork), and several other fungi pro-duced more than 100 ng/g of TCA. It is important to emphasize that TCA synthesis required2,4,6-trichlorophenol as a substrate and that no TCA was formed in these experiments whenother chlorinated phenols were present.Thus, these investigators concluded that fungi are re-sponsible for TCA and cork taint formation.When TCA is actually made during cork processingis not known, nor is it clear from where the trichlorophenol originates. Chlorinated polyphe-nols were once used as agricultural herbicides and fungicides and were likely applied in cork-producing areas. It has been suggested that chlorine treatment of harvested cork may further in-crease the trichlorophenol content, and subsequently, the TCA level.

Despite these recent findings, there is still considerable debate regarding the true incidenceof cork-taint in wine, as well as the source of the tainted aroma. Cork producers and some corkresearchers have argued that cork taint occurs far less often than the reported frequencies(Silva Pereira et al., 2000).They claim that TCA can arise from other sources—not just cork—and that other non-cork taints continue to be referred to as cork taint. Poor storage and han-dling of corks can indeed result in other off-flavors; however, these defects are controllable.Cork stoppers, their proponents contend, are still the best way to seal bottles and preserve thequality of wine (Silva Pereira et al., 2000). Finally, some wine writers have suggested that thecork taint problem has been exaggerated, in part by the industry itself and also by other winewriters (Casey, 1999).

Remedies and alternatives

Regardless of whether the frequency of cork taint is 1% or 12% or somewhere in between (2%to 5% is the figure most often quoted), this can hardly be considered an acceptable defect rate.No industry could long tolerate a situation where even one unit out of 100 fails to meet qualitystandards.The wine industry is no different and that is why non-cork alternatives have becomeso popular in the past ten years. Efforts by the cork industry to improve cork and prevent corktaint are also under way. For example, supercritical CO2 treatment can reportedly extract TCAfrom cork to undetectable levels (by sensory analysis).Analytical methods have also been de-veloped to screen corks for TCA so that tainted corks are not used.

Box 10–7. Cork Taints in Wine: Are Corks the Cause? (Continued)

Cl

OCH3

ClCl

Cl

OHClCl

2,4,6-trichlorophenol 2,4,6-trichloroanisole (TCA)

Figure 1. Conversion of 2,4,6-trichlorophenol to 2,4,6-trichloroanisole (TCA).

(Continued)



The main alternatives to corks for bottled wine include cork agglomerates and hybrids, plas-tic enclosures, and metal screw caps (bag-in-the-box packages might also qualify,but will not befurther discussed). Agglomerates consists of cork granules (also called “dust”) stuck togetherwith glue and extruded into cork shapes.They have long been used by sparkling wine manu-facturers and have many of the desirable features of cork, but there are reports they may alsocontain TCA.A new generation of agglomerates are now available that reportedly are TCA-free(via supercritical extraction). Hybrid corks are similar to the agglomerates, but contain plasticparticles mixed with natural, ground cork. Plastic or synthetic corks are comprised of ethylenevinyl acetate.They have resilience similar to cork, but contain no cork material (and no TCA).Detractors have suggested that plastic corks impart plastic flavors to the wine (i.e., plastictaint).

Metal screw caps for wine were introduced in the 1990s.They are similar to those used forcarbonated beverages. Screw caps are easy to open and re-close, and are TCA-free.They have allof the attributes of an ideal enclosure with no real technical downside. Wines from screw-capped bottles always score well in taste tests (although rubber-like flavors, from the enclosureseal, have been reported to migrate into the wine).

Ultimately, the type of enclosure used by a particular winery depends only in part on the per-formance characteristics of the particular material. Rather, perhaps the more important deter-mining factor is the attitude or perception consumers have toward specific types of wine en-closures. Surveys indicate that consumers prefer cork, that they expect to hear a “pop” whenthe bottle is opened, and that tradition is a key feature of the wine-drinking experience. Manywine drinkers may not even recognize off-flavors in wines, including cork and other taints.Moreover, plastic corks, and especially screw cap enclosures, are often associated with econ-omy wines, regardless of what is actually in the bottle.

Despite these attitudes, non-cork enclosures continue to gain popularity, and now are usedfor as much as 50% or more of the wine produced in some countries.For example,Australia andNew Zealand have championed the use of screw cap enclosures. Many California wineries usecork only for their premium wines, and synthetic composite or screw caps for their otherwines. Some experts have predicted that the trend away from cork will continue.

References

Álvarez-Rodríguez, M.L., L. López-Ocaña, J.M. Lípez-Coronado, E. Rodríguez, M.J. Martínez, G. Larriba, and J.J.R.Coque. 2002. Cork taints of wines: role of the filamentous fungi isolated from cork in the formation of2,4,6-trichloroanisole by O–methylation of 2,4,6-trichlorophenol.Appl.Environ.Microbiol.68:5860–5869.

Casey, J. 1999.T’aint necessarily so.Wine Industry Journal. 14:(6)49–56.Coque, J.-J.R., M.L. Álvarez-Rodríguez, and G. Larriba. 2003. Characterization of an inducible chlorophenol

O-methyltransferase from Trichoderma longibrachiatum involved in the formation of chloroanisolesand determination of its role in cork taint of wines.Appl. Environ. Microbiol. 69:5089–5095.

Lee,T.H., and R.F. Simpson. 1993. Microbiology and chemistry of cork taints in wine, p. 353–372. In Fleet,G.H. (ed.), Wine Microbiology and Biotechnology. Harwood Academic Publishers. Chur, Switzerland.

Jackson, R.S. 2000. Wine Science: Principles, Practice, Perception, 2nd Ed.,Academic Press. San Diego, Cal-ifornia.

Ribéreau-Gayon, P., D. Dubourdieu, B. Donèche, and A. Lonvaud. 2000. Handbook of Enology, Volume 1:The Microbiology of Wine and Vinifications. John Wiley and Sons, Ltd.West Essex, England.

Peña-Neira,A., B.F. de Simón, M.C. García-Vallejo,T. Hernández, E. Cadahía, and J.S. Suarez. 2000. Presence ofcork-taint responsible compounds in wines and their cork stoppers. Eur. Food Res. Technol. 211:257–262.

Silva Pereira, C.S., J.J.F. Marques, and M.V.S. Romão. 2000. Cork taint in wine: scientific knowledge and pub-lic perception—a critical review. Crit. Rev. Microbiol. 26:147–162.

Simpson, R.F., D.L. Capone, and M.A. Sefton. 2004. Isolation and identification of 2-methoxy-3,5-dimethylpyrazine, a potent musty compound from wine corks. J.Agric. Food Chem. 52:5425–5430.

Box 10–7. Cork Taints in Wine: Are Corks the Cause? (Continued)



Spoilage by yeasts

Yeasts represent a major cause of wine defectsand spoilage. Moreover, since yeasts are an ex-pected part of the natural flora of grapes andmust, their growth before, during, and after thewine fermentation is difficult to control. Forexample, Kloeckera apiculata, one of theyeasts involved in the early stages of a naturalfermentation, can produce high enough levelsof various esters (mainly ethyl acetate andmethylbutyl acetate) to cause an ester taint,which has a vinegar-like aroma. Once vigorousgrowth of S. cerevisiae begins, other yeasts aregenerally unable to compete and grow. How-ever, if S. cerevisiae does not become well-established (i.e., during natural fermentation),other yeasts, including Zygosaccharomycesbailii, can grow and produce acetic and suc-cinic acids. Growth of this organism is espe-cially a problem in sweet wines, due to its abil-ity to tolerate high osmotic pressure and highethanol concentrations.

Growth of yeasts during aging of wine, ei-ther in barrels or bottles, is a particularly seri-ous spoilage problem. The main culprits arespecies of Brettanomyces/Dekkera, and Bret-tanomyces bruxellensis in particular (note thatBrettanomyces/Dekkera exists in one of twoforms, where Brettanomyces is the asexual,nonsporulating form and Dekkera is the sex-ual, sporulating form of the same yeast).Theseyeasts are common contaminants of wineriesand the oak barrels used for aging. Growth ofthese organisms may lead to volatile phenol-containing compounds that give the wine a dis-agreeable “mousy” aftertaste (although mousytaints may also be caused by bacteria).

Brettanomyces may produce a variety ofother taints, variously described as “barnyard,”“horse sweat,” “band-aid,” and “wet dog.”A common marker or signature chemical for“Brett” spoilage is 4-ethyl phenol, whose pres-ence in suspect wines can be routinely moni-tored.There are also various film yeasts, includ-ing species of Pichia,Candida,and Hansenula,that grow on the surface of wine during barrelaging, much like the flor yeast whose growth insherry making is desirable. Under ordinary cir-

cumstances, however, growth of film yeast canlead to oxidation of ethanol and formation ofacids, esters, acetaldehyde, and other undesir-able end products. Most of these spoilage yeastcan be controlled or managed by SO2 addition,maintenance of proper anaerobic conditions(topping off of barrels),barrel management,andgood sanitation practices, or in extreme cases,by sterile filtration.

Spoilage by bacteria

Probably the most common and most disas-trous types of microbial spoilage of wine arethose caused by bacteria.Two distinct groupsare of importance: the acetic acid bacteria andthe lactic acid bacteria, both of which containspecies able to tolerate the low pH, highethanol conditions found in wine.These bacte-ria are responsible acidic and other end prod-ucts that seriously affect wine quality.

The acetic acid bacteria that are most im-portant in wine spoilage belong to one ofthree genera: Acetobacter, Gluconoacetobac-ter, and Gluconobacter. The main species in-volved in wine spoilage are Acetobacter aceti,Acetobacter pasteurianus, and Gluconobac-ter oxydans.They are Gram negative, catalase-positive rods capable of oxidizing alcohols toacids. These bacteria also are considered asobligate aerobes; however, it now appearsthat limited growth and metabolism can oc-cur even under the mostly anaerobic condi-tions that prevail during wine making. Al-though acetic acid bacteria are generallyfound at relatively low levels in vineyards andin must (�100 cells per g), moldy or bruisedgrapes can contain appreciably higher levels.If the ethanol fermentation occurs soon afterharvesting and crushing, then growth of theseorganisms, especially G. oxydans, is inhibitedand numbers may actually decline. When thefermentation is complete and the wine isdrawn off and subsequently transferred andracked, aeration inevitably occurs, activatinggrowth of these bacteria.

Aging in oak barrels may also promotegrowth of acetic acid bacteria, in part becausebarrels may actually be contaminated with



these bacteria, but also because oxygen diffu-sion into small oak barrels can be significant.Under these conditions, then,acetic acid bacte-ria can oxidize ethanol in the wine, producingenough acetic acid (�0.7 g/L) to give the winea pronounced vinegar flavor and aroma. Al-though lesser amounts can sometimes be toler-ated,other end products resulting from growthof acetic acid bacteria, including ethyl acetate,acetaldehyde, and dihydroxyacetone, may alsocontribute to spoilage defects. Low sulfur di-oxide levels will further enhance growth ofacetic acid bacteria.

The other group of bacteria associated withmicrobial spoilage of wine are the lactic acidbacteria. This cluster of Gram positive, faculta-tive rods and cocci consists of twelve genera,however, only four, Lactobacillus, Oenococcus,Leuconostoc, and Pediococcus, are involved inwine spoilage.Due to their saccharolytic metab-olism,it is not surprising that lactic acid bacteriaare found on intact grapes, albeit at low popula-tions (102/g to 103/g) that eventually increaseten- to 100-fold during harvesting and crushing.Although most lactic bacteria do not grow dur-ing the ethanolic fermentation (the exceptionare the malolactic bacteria), some strains are tol-erant of high ethanol concentrations and areable to grow later during post-fermentationsteps. High temperature and pH and low SO2

concentrations favor growth of lactic acid bac-teria. Growth of these bacteria in wine can re-sult in several spoilage conditions, includingacidification and deacidification, as well as pro-duction of various metabolites that cause off-flavors, aromas, and other defects.

As described in Chapter 2, lactic acid bacte-ria have either a homofermentative or hetero-fermentative metabolism, or in some strains,the metabolic capacity for both. When ho-mofermentative lactic acid bacteria, such asPediococcus and Lactobacillus plantarum,grow in wine and ferment glucose, they pro-duce lactic acid.This acid causes the wine tobecome excessively sour. In contrast, metabo-lism of pentose sugars, present in the grapesas well as extracted from oak barrels, occursvia the pentose phosphate pathway (also

known as the phosphoketolase pathway),yielding acetic acid, ethanol, and CO2. Thesesame end-products are also produced by sev-eral heterofermentative lactic acid bacteriafound in wine, including Leuconostoc mesen-teroides, Oenococcus oenus, and Lactobacil-lus brevis. Although these bacteria produceconsiderably less acetic acid than Acetobacteror Gluconobacter, enough may be producedto impart a detectable vinegar-like flavor tothe wine.

Several lactic acid bacteria have the capac-ity to convert malic acid to lactic acid via themalolactic pathway. In musts containing highmalic acid concentrations and having low pH,the malolactic bacteria perform a desirable,even essential function, by deacidification ofthe wine. However, if the must acidity is al-ready low and the pH high, the malolactic fer-mentation may result in an increase in wine pHsuch that too little acidity remains. The winealso will be more susceptible to spoilage sincea major barrier to microbial growth, low pH, isdiminished.

Spoilage by lactic acid bacteria can also becaused by other metabolic end products. Me-tabolism of fructose by heterofermentative L.brevis, for example, can cause mannitol toform, and generate acetic acid as a side reac-tion. In a fermentation analogous to the malo-lactic pathway, in terms of effect, L. brevis canmetabolize tartaric acid, causing an increase inwine pH and formation of volatile acids.Mousytaints, similar to those produced by thespoilage yeast Brettanomyces, can also be pro-duced by Lactobacillus sp., Leuconostoc sp.,and other heterofermentative lactic acid bacte-ria.A flowery off-odor referred to as geraniumtaint also may be produced by lactobacilli.

Glycerol oxidation by Leuconostoc mesen-teroides, Lactobacillus brevis, and other lacto-bacilli generates acrolein,which reacts with tan-nins and anthocyanin phenolics to form bittercompounds. This defect, referred to as amer-tume, is more common in red wines, due to thehigher tannin and phenolic levels in red wines.Its microbiological cause was first noted by Pas-teur. Several lactic acid bacteria, including



strains of Pediococcus and Lactobacillus, pro-duce diacetyl,which imparts a buttery aroma inwine that, at high concentrations (generallyabove 5 mg/L), is considered undesirable.Finally, the formation of glucose-containingpolysaccharides, such as dextrins and glucans,can give an oily, viscous and objectionablemouth feel. Ropiness usually occurs only insweet wines and is caused by Pediococcus,Oenococcus, and Leuconostoc spp.

BibliographyAmerine,M.A.,H.W.Berg,R.E.Kunkee,C.S.Ough,V.L.

Singleton, A.D. Webb. 1980. The Technology ofWine Making, 4th Ed.,Avi Publishing Company,Inc.,Westport, Connecticut.

Bartowsky,E.J.,and P.A.Henschke.2004.The ‘buttery’attribute of wine-diacetyl-desirability, spoilageand beyond. Int. J. Food Microbiol. 96: 235–252.

Boulton, R.B.,V.L. Singleton, L.F. Bisson, and R.E. Kun-kee. 1996. Principles and Practices of Winemak-ing. Chapman and Hall, New York, New York.

Coates, C. 2000. An Encyclopedia of the Wines andDomaines of France. University of CaliforniaPress. Berkeley, California.

Fleet, G.H. 2003.Yeast interactions and wine flavour.Int. J. Food Microbiol. 86:11–22.

Fugelsang, K. C. 1997. Wine Microbiology. Chapmanand Hall, New York, New York.

Giudici,P., L. Solieri,A.M.Pulvirenti, and S.Cassanelli.2005. Strategies and perspectives for genetic im-provement of wine yeasts. Appl. Microbiol.Biotechnol. 66:607–613.

Jackson, R.S. 2000. Wine Science: Principles, Prac-tice, Perception, 2nd Ed., Academic Press, SanDiego, California.

Loureiro,V., and M. Malfeito-Ferreira. 2003. Spoilageyeasts in the wine industry. Int. J. Food Microbiol.86:23–50.

Konings, W.N. 2002. The cell membrane and thestruggle for life of lactic acid bacteria. Antonievan Leeuwenhoek 82:3-27.

Pretorius, I.S., and F.F. Bauer. 2002. Meeting the con-sumer challenge through genetically customizedwine-yeast strains. Trends Biotechnol. 20:426–478.

Ribéreau-Gayon, P., D. Dubourdieu, B. Donèche, andA.Lonvaud.2000.Handbook of Enology,Volume1: The Microbiology of Wine and Vinifications.John Wiley and Sons, Ltd.,West Essex, England.

Ribéreau-Gayon, P., Y. Glories, A. Maujean, and D.Dubourdieu. 2000. Handbook of Enology, Vol-ume 2:The Chemistry of Wine Stabilization andTreatments. John Wiley and Sons,Ltd.West Essex,England.

Romano, P., C. Fiore, M. Paraggio, M. Caruso, and A.Capece. 2003. Function of yeast species andstrains in wine flavour. Int. J. Food Microbiol. 86:169–180.


11

Vinegar Fermentation

“Vinegar, son of wine.”Hebrew proverb

vegetables.Thus, vinegar can be considered asthe first biologically-produced preservative.However, in addition to its use as a so-calledpickling agent, it was also consumed directly asa flavoring agent, and, in a diluted form, as abeverage.

Vinegar consumption is first noted in theBible (Numbers 6:1 and Ruth 2:14),and,accord-ing to the New Testament, was given to Jesusduring the crucifixion ( John 19:23). Dilutedvinegar was a popular beverage throughout theGreek and Roman eras, where it gained favor asa therapeutic beverage.The production of vine-gar was not confined to Europe; it was also pro-duced throughout Asia.Although the substrateswere different, the processes were remarkablysimilar to those that evolved in Europe. Finally,in addition to its use as a food or food ingredi-ent, vinegar has also long been used as a topicaldisinfectant,as a cleaning agent,and as an indus-trial chemical due to its strong demineralizationproperties.

In the food industry, vinegar is used mainlyas an acidulent, a flavoring agent, and apreservative, but it also has many other foodprocessing applications. It is found in hun-dreds of different processed foods, includingsalad dressings, mayonnaise, mustard andketchup, bread and bakery products, pickledfoods, canned foods, and marinades and

History

Although it is not possible to know preciselywhen human beings first began to produceand consume fermented foods, the origin forsome products can reasonably be estimated.Vinegar, for example, was likely discoveredshortly after (like about a week) the advent ofthe first successful wine fermentation. As ex-cited as that first enologist must have been tohave somehow turned grape juice into wine,one can imagine the disappointment that musthave followed when the wine itself subse-quently turned into a sour, unpalatable liquid,seemingly devoid of any redeeming virtues.Even though that sour wine, vin aigre inFrench, obviously couldn’t be drunk with thesame enjoyment or enthusiasm as the wine, itwas not, it turned out, without value. Indeed,vinegar has a long history of use, and is nowone of the most widely used ingredients in thefood industry, with world-wide production ofabout 1 million L per year.

The actual history of vinegar consumptiondates back several thousand years. Althoughother fermented foods, such as wine, beer, andcheese, evolved, in part, because of their en-hanced preservation status, vinegar was likelyused for its ability to preserve other non-fermented, perishable foods, such as meats and

397





sauces. It is used in almost every culture andis part of nearly every cuisine. And althoughmany of the vinegars produced around theworld are made from ordinary substrates andoften have rather nondescript sensory prop-erties, others are produced from premiumwines, carefully aged, and prized (and priced)based on their unique organoleptic attributes.

Returning to our early wine maker, there is abiological reason, of course, to account for theunfortunate fate of that ancient wine.Wine, wenow know, turns into vinegar because natu-rally-occurring bacteria exist, namely speciesof Acetobacter, that oxidize the ethanol in thewine to form acetic acid. In fact, any ethanol-containing material can serve as a substrate forthe vinegar fermentation, and, as will be dis-cussed below, there are numerous types ofvinegars that are produced from a variety ofethanolic substrates. While wine producerstake special measures to prevent contamina-tion by acetic acid-producing bacteria and toavoid the oxygen supply that is necessary forthe ethanol-to-acetic-acid conversion, manufac-turers of vinegar do just the opposite. In otherwords, the vinegar fermentation is conductedin the presence of Acetobacter and under con-ditions that favor growth and oxidative metab-olism by this organism.

Definitions

There is no standard of identity for vinegar inthe United States, but there are guidelines thatdefine the starting material, the finished specifi-cations, and the labeling declaration. First, aswill be described in more detail later, vinegarmust be made from one of various types ofethanol-containing solutions. The most com-mon starting materials, whose identity must beindicated on the label, are grape and rice wine,fermented grain or malt mashes, and fermentedapple cider. Distilled ethanol is also permittedas a substrate for vinegar manufacture. Impor-tantly, vinegar must, by definition, result fromthe “acetous fermentation” of ethanol. In otherwords, acetic acid made via chemical synthesis

cannot be labeled as vinegar.Vinegar must con-tain at least 4% acetic acid or at least fortygrains (where one grain � 0.1% acid). Usually,the ethanol concentration is less than 0.5%,andthe pH is between 2.0 and 3.5. In countrieswhere identity standards do exist, they gener-ally are consistent with the U.S. definition.

As noted above, the raw material determinesthe name of the vinegar (e.g., red wine vinegar,apple cider vinegar,malt vinegar).However, thestarting material also has a profound influenceon the flavor and overall quality attributes ofthe vinegar. Although the predominant flavor of all vinegars is due to acetic acid, other fla-vors, specific to the ethanol source or themeans of its manufacture, may also be present.In addition, some vinegars may also containherbs (e.g., tarragon), added before or after thefermentation.

Vinegar Manufacturing Principles

The manufacture of vinegar consists of two dis-tinct processes. The first step is an ethanolicfermentation performed mostly by yeasts. Inthe second step, an acetogenic fermentation iscarried out by acetic acid bacteria.Whereas theethanol fermentation is anaerobic, the latter isconducted under highly aerobic conditions. Infact, many of the technological advances in thevinegar fermentation have focused on ways tointroduce more air or oxygen into the fermen-tation system.

Microorganisms

Aside from those bacteria that produce aceticacid as an overflow or side reaction fromsugar metabolism (e.g., lactic acid bacteria),there are several genera of bacteria that pro-duce acetic acid as the primary metabolic endproduct. It is important, however, to distin-guish between those that produce acetic acidfrom one-carbon precursors and those thatproduce acetic acid via oxidation of ethanol.The former group are referred to as acetogensand include species of Clostridium, Eubac-


Vinegar Fermentation 399

terium, Acetobacterium, Peptostreptococcus,and other Gram positive anaerobic rods andcocci. Acetogens rely on the acetogenesispathway, in which energy is derived fromother sources and CO2 merely serves as theelectron acceptor. The latter becomes re-duced to acetic acid under strict anaerobicconditions. In contrast, in the ethanol oxida-tion pathway, acetic acid bacteria actually usethe electrons from ethanol, with oxygen ser-ving as the electron acceptor. Thus the sameproduct is formed, but for two totally differ-ent biochemical reasons.The acetic acid bac-teria are represented by four genera: Aceto-bacter, Gluconobacter, Gluconoacetobacter,and Acidomonas (a fifth genus, Asaia, has re-cently been described, but oxidizes ethanolweakly or not at all). These genera belong tothe Phylum Proteobacteria and the familyAcetobacteraceae. The phylogeny of theseand related bacteria is illustrated in Figure11–1.Acetic acid bacteria are Gram negative,obligate aerobes, motile or non-motile, andhave an ellipsoidal to rod-like shape that ap-pear singly, in pairs or in chains.They have aG�C base composition of 53% to 66%.

Acetic acid bacteria are widely distributed inplant materials rich in sugars, such as fruits andnectars. They are also common inhabitants ofalcohol-containing solutions, including wine,beer, hard cider, and other ethanolic beverages.Not surprisingly, acetic acid bacteria oftenshare habits with ethanol-producing yeasts.Although some acetic acid bacteria, notablyAcetobacter, are more efficient at oxidizingethanol, rather than glucose, to acetic acid,Glu-conobacter oxidizes and grows especially wellon glucose compared to ethanol.

Three genera, Acetobacter, Gluconobacter,and the recently named Gluconoacetobacter,contain most of the species used industriallyfor vinegar manufacture (see below).AlthoughAcetobacter aceti has long been considered tobe the primary organism involved in the vine-gar fermentation, many other species havebeen identified in vinegar and in production facilities, including Acetobacter pasteurianus,

Gluconoacetobacter europaeus (formerly Ace-tobacter europaeus), Gluconacetobacter xyli-nus (formerly Acetobacter xylinus), and Glu-conobacter oxydans (Table 11–1).

Many vinegar fermentations are conductedusing mixed or wild cultures, accounting forthe large number of strains isolated from differ-ent production facilities. For example, Glu-conobacter entanii was the predominant or-ganism isolated from a single high-acid vinegarproduction facility in Germany, and A. pas-teurianus was identified as the main speciesinvolved in production of rice wine vinegarsproduced in Japan. In Germany,A.europaeus isnow considered to be the most common or-ganism isolated from vinegar factories that usesubmerged-type fermentation processes (seebelow). It seems clear, therefore, that no singlespecies can be considered to be solely respon-sible for the vinegar fermentation.

Although all species of Acetobacter andGluconobacter produce acetic acid fromethanol, they differ, metabolically, in several re-spects (Table 11–2). First, whereas Acetobac-ter can, under certain conditions, completelyoxidize ethanol to CO2, metabolism of ethanolby Gluconobacter stops at acetic acid (dis-cussed below). In contrast, oxidative metabo-lism of glucose and other sugars is muchgreater in Gluconobacter compared to Aceto-bacter, which produces little acid from sugars.Finally, these organisms can be distinguishedon the basis of the quinone co-factor systemsused as electron acceptors (see below) duringthe ethanol oxidation reactions, because Glu-conobacter possesses a G10 type quinone sys-tem and Acetobacter has a Q9 system.

As noted above, several species of Aceto-bacter are used in industrial production ofvinegar. In general, the criteria used to distin-guish between species have been based onthose biochemical and physiological proper-ties most relevant to the acetic acid fermenta-tion (Table 11–3). The phylogeny and taxon-omy of acetic acid bacteria have undergonesignificant changes in recent years. Until 2000,eight species of Acetobacter were recognized,



Acetobacter lovaniensisAcetobacter peroxydans

Acetobacter pasteurianus Acetobacter pomorum

Acetobacter aceti Acetobacter estunensis

Acetobacter orleanensisAcetobacter cerevisiae Acetobacter malorum

Acetobacter tropicalisAcetobacter indonesiensis

Acetobacter orientalisAcetobacter cibinongensis

Gluconoacetobacter intermediusGluconoacetobacter oboediens

Gluconoacetobacter europaeus

Gluconoacetobacter xylinus

Gluconoacetobacter hanseniiGluconoacetobacter entanii

Gluconoacetobacter diazotrophicusGluconoacetobacter liquefaciens

Gluconoacetobacter azotocaptans Gluconoacetobacter johannae Gluconobacter oxydans Gluconobacter cerinus

Gluconobacter frateurii 0.01

Figure 11–1. Phylogenetic relationship of Acetobacter, Gluconobacter, and Gluconoacetobacter sp. basedon 16S rRNA sequences. The tree was generated using the UPGMA clustering method. For most of the nodes,bootstrap values were 90% to 100%.



based on the characteristics listed in Table 3,as well as on DNA homology and G�C con-tent. Recently, however, it was proposed, onthe basis of additional DNA-DNA and phyloge-netic relatedness, that several Acetobacterspecies be moved to the genus Gluconoaceto-bacter, and that several other new species beadded (Table 11–1).

Metabolism and Fermentation

The acetic acid pathway used by Acetobacter,Gluconobacter and other acetic acid bacteriais an example of what is referred to as an in-complete oxidation.Whereas in most oxidativepathways (e.g., the Krebs or citric acid cycle),organic substrates are ordinarily oxidized allthe way to CO2 and H2O, in the vinegar fer-mentation, acetic acid bacteria usually oxidize

the substrate, ethanol,only to acetic acid.How-ever, as described below, exceptions existwhere complete oxidation to CO2 can occur.

The actual pathway consists of just twomain steps (Figure 11–2). First, ethanol is oxi-dized to acetaldehyde and then the acetalde-hyde is oxidized to acetic acid.An intermedi-ate step, in which acetaldehyde hydrate isformed, may also occur. Oxygen is required asthe terminal electron acceptor (see below). Itshould also be noted that the conversion ofethanol to acetic acid occurs on an equimo-lar basis, such that after the reactions arecomplete, the final concentration of aceticacid will be equal to that of the ethanol in thestarting material (assuming negligible lossfrom evaporation). Only a minor amount ofthe carbon from ethanol is used for biomassor converted to other products.

Table 11.1. Species of Acetobacter and Gluconoacetobacter isolated from vinegar1.

Acetobacter aceti Gluconoacetobacter xylinus2

Acetobacter cerevisiae Gluconoacetobacter hansenii2

Acetobacter cibinongensis Gluconoacetobacter europaeus2

Acetobacter estunensis Gluconoacetobacter intermedius2

Acetobacter indonesiensis Gluconoacetobacter liquefaciens2

Acetobacter lovaniensisAcetobacter malorumAcetobacter orleanensisAcetobacter orientalisAcetobacter pasteurianusAcetobacter peroxydansAcetobacter pomorumAcetobacter tropicalis

1From Cleenwerck et al. 2002; Garrity et al., 2004; and Lisdiyanti et al., 20002These species were previously classified as Acetobacter

Table 11.2. Physiological properties and characteristics of acetic acid bacteria.

Property Acetobacter Gluconobacter Gluconoacetobacter

Temperature optimum (°C) 25–30 25–30 25–30pH optimum 5.4–6.3 5.5–6.0 5.4–6.3Acetic acid oxidation � � �/�Lactic acid oxidation � � �/�Acid from glucose � � �

Ubiquinone type Q–9 Q–10 Q–10

Adapted from Yamada et al., 2000 and Holt et al., 1994



Tab

le 1

1.3.

Bioc

hem

ical

and

phy

siol

ogic

al p

rope

rtie

s of

spe

cies

of

acet

ic a

cid

bact

eria

.1

Glu

con

o-

Glu

con

o-

Glu

con

o-

Glu

con

o-

Glu

con

o-

Pro

per

tyA

ceto

ba

cter

Ace

toba

cter

ace

toba

cter

ace

toba

cter

ace

toba

cter

ace

toba

cter

ace

toba

cter

Aci

dom

on

as

ace

tipa

steu

ria

nu

sdia

zotr

oph

icu

seu

ropa

eus

ha

nse

nii

liqu

efa

cien

sxy

lin

us

met

ha

noli

cus

Gro

wth

on

3%

eth

ano

l �

��

��

��

�

in 4

-8%

ace

tic

acid

Gro

wth

in a

bse

nce

��

u2

��

��

�

of

acet

ic a

cid

Gro

wth

on

ace

tate

at

��

��

��

��

pH

2.5

in a

bse

nce

of

eth

ano

lG

row

th o

n 1

.7 M

glu

cose

��

��

��

��

Gro

wth

in p

rese

nce

of:

lact

ic a

cid

�d

w3

d4

d�

u�

eth

ano

l�

d�

��

��

w30

% g

luco

se�

��

��

��

�

Bio

syn

thes

is o

f ce

llulo

se�

��

d�

��

�

DN

A h

om

olo

gy (

%)

wit

h:

G.e

uro

peu

s�

10�

25u

70–1

00�

15�

1530

�10

G.x

ylin

us

u�

10u

37u

u10

0u

A.p

ast

euri

an

us

u10

0u

17u

u�

10u

G �

C c

on

ten

t (%

)56

–60

53–6

361

–63

56–5

858

–63

62–6

555

–63

62

1A

dap

ted

fro

m S

ieve

rs a

nd

Teu

ber

,199

5 an

d S

oko

llek

et a

l.,19

982u

nd

eter

min

ed3w

eak

4so

me

stra

ins

po

siti

ve



The biochemical and physiological processesinvolved in the acetic acid fermentation arequite unlike those described for the lactic andethanolic fermentations.The acetic acid fermen-tation is performed by obligate aerobes,and me-tabolism occurs not in the cytoplasm,but ratherwithin the periplasmic space and cytoplasmicmembrane. The acetic acid pathway yields en-ergy, but not by substrate level phosphoryla-tion. Instead, the oxidation reactions are cou-pled to the respiratory chain which thengenerates ATP via electron transport and oxida-tive phosphorylation reactions. Finally, both thesubstrate and product are toxic, which nodoubt contributes to the relative lack of com-petitors during what is essentially an open fer-mentation. In fact,acetic acid bacteria are rather

remarkable for their ability to tolerate low pHand high acetic acid concentrations (Box 11–1).

The two main enzymes involved in theacetic acid fermentation, alcohol dehydroge-nase (AlDH) and acetaldehyde dehydrogenase(AcDH), have been characterized, and thegenes coding for their expression have beencloned and sequenced (Figure 11–3). Theseenzymes are the “primary dehydrogenases,”responsible for nearly all of the detectable ox-idation activity. They are located within thecytoplasmic membrane and face into theperiplasm (Figure 11–4). This means that the ethanol and acetaldehyde oxidation reac-tions occur in the periplasm and that ethanolcan be used without having to be accumu-lated within the cytoplasm. Both AlDH and

CCHH

HH

OH

H

ethanol acetaldehyde

CO

HCH

H

Halcoholdehydrogenase

PQQH2PQQ

CO

OHCH

H

H

acetic acid

PQQH2PQQ

aldehydedehydrogenase

N O

O

NC

O HHO

HOOC

COOH2 H+

OH

OH

N

NC

O HHO

HOOC

COOH

PQQ (oxidized) PQQ (reduced)

2 H+

Figure 11–2. Formation of acetic acid from ethanol by Acetobacter aceti. The oxidation of acetic acid toethanol involves two enzymatic reactions (upper panel). In the first reaction, ethanol is oxidized to acetalde-hyde by alcohol dehydrogenase. Next, acetaldehyde dehydrogenase further oxidizes acetaldehyde to aceticacid. An intermediate hydration reaction may also occur, in which acetaldehyde hydrate is formed (notshown). Both alcohol dehydrogenase and acetaldehyde dehydrogenase enzymes require pyrroloquinolinequinone (PQQ) as a co-factor, which serves as an electron acceptor (lower panel). The reduced PQQ then sup-plies electrons to the electron transport chain, eventually leading to the formation of ATP via oxidative phos-phorylation. Adapted from Fuchs, G., 1999.



Box 11–1. Acetic Acid Tolerance in Acetobacter

Microorganisms that produce peroxides, alcohols, acids,or other inhibitory agents must also beable to tolerate such products. In the case of acetic acid bacteria, tolerance to both ethanol andacetic acid is a necessary feature of their physiology.

During the vinegar fermentation, cells typically encounter ethanol at concentrations near12% (more than 2.5 M). Later, as ethanol is oxidized to acetic acid, vigorous growth is main-tained even when the concentration of acid reaches more than 7%. Some strains of Acetobactercan produce (and tolerate) as much as 20% acetic acid (3.3 M).To provide some perspective,Es-cherichia coli is inhibited by 1% or less acetic acid (Lasko et al., 2000). Thus, the means bywhich acetic acid bacteria deal with acetic acid stress is not only of fundamental interest, but italso has practical implications. Strains that are more acid tolerant, for example, would be ex-pected to be more productive during commercial vinegar fermentations.

To fully appreciate the challenge faced by these cells, consider first the actual problem thatexists when Acetobacter grows during the acetic acid fermentation. In a 6% (1 M) acetic acidsolution at a pH of 4.5, most of the acetic acid will be in the undissociated or acid form (giventhat the pKa of acetate is 4.76). Since the cell membrane is permeable to the acid form, cells atthat pH could accumulate, in theory, more than 0.6 M acetic acid within the cytoplasm (viasimple diffusion).Accumulation of the acid will continue, in fact, until the inside and outsideconcentrations are equal.Of course,because the physiological pH within the cytoplasm is gen-erally higher than the external pH (at least for most bacteria), the accumulated acid will immediately dissociate to form the anion and a free proton (in accordance to the Henderson-Hasselbach equation).This will further shift the [acidin]/[acidout] equilibrium such that evenmore acid will be accumulated.

At some point one of several outcomes are possible.First,unless the accumulated protons areexpelled from the cytoplasm, their accumulation will eventually cause the intracellular pH todecrease to an inhibitory level. Second, if the cells relied on proton extrusion to maintain amore conducive intracellular pH, then substantial amounts of ATP (or its equivalent) would berequired to drive the proton pumping apparatus, leading to a major drain on the energy re-sources available to the cell. Finally, if a high cytoplasmic pH is maintained, the acetate anioncould, in theory, reach concentrations of 1 M or higher (under the conditions stated above).Thiswould create severe toxicity and osmotic problems for the cell, which may be even worse thanacidity problems (Russell, 1992).

Despite these apparent hurdles, researchers, primarily in Switzerland and Japan, have re-vealed that there may be several mechanisms that account for the ability of Acetobacter andother acetic acid bacteria to tolerate high acetic-acid and low-pH environments.These mecha-nisms involve both physiological and genetic responses (Table 1).As noted above,acetic acid ac-cumulation depends on the pH gradient across the cell membrane.The higher the gradient (i.e.,low pH outside and high pH inside), the greater will be the acetate that will be trapped inside.If, on the other hand, the cytoplasmic pH is allowed to drop (decreasing the gradient), then lessacetate is accumulated.

Table 1. Summary of mechanisms contributing to acetic acid tolerance.

Mechanisms Possible effect

Low pH gradient Reduced anion accumulationgroESL operon induced General stress responseExpression of Aar proteins Stimulate acetate assimilationExpression of Aap proteins Stimulate transport systemsExpression of Asp proteins UnknownAconitase expression Stimulate acetate assimilationAcetic acid efflux pump Stimulate acetic acid efflux



In fact, this indeed seems to be the case for Acetobacter aceti, because this organism main-tains only a very small pH gradient (Menzel and Gottschalk, 1985).Thus, the cell is spared theproblem caused by anion (acetate) accumulation (of course, this implies that Acetobacter muststill be physiologically able to tolerate a low intracellular pH).That acetate accumulation mightcontribute to cell inhibition is supported by the observation that resistant strains accumulateless acetate than wild-type strains that are only moderately resistant (Steiner and Sauer, 2003).The response to high acetate concentrations at a near neutral pH may, however, be quite differ-ent.These same researchers showed, for example, that A.aceti at pH 6.5 accumulated more than3 M acetate, at an outside concentration of less than 0.4 M. However, these high-pH conditionsclearly do not reflect those that this organism would likely experience.

As indicated above, intracellular acetic acid also could, in theory,be pumped out directly,pro-vided the energy sources were available. Recently, the presence of an acetic acid efflux systemwas reported in Acetobacter aceti (Matsushita et al., 2005). The system was sensitive toprotonophores and ionophores, agents that dissipate proton and ion gradients across the cellmembrane and by a respiration inhibitor.Therefore, it appears that acetic acid efflux is ener-gized by a proton motive force that is itself dependent on respiration activity.

Transcriptional analyses of A. aceti also have revealed the presence of genes whose expres-sion is induced by acetic acid and that may be associated with acetic acid resistance. Examplesinclude the groESL and the aar (acetic acid resistance) operons (Fukaya et al., 1990; Fukaya etal., 1993; Okamoto-Kainuma et al., 2002).The GroES and GroEL proteins (encoded by groESL)are part of the chaperonin family of stress-induced proteins that protect cells by stabilizing pro-teins and preventing their mis-folding. In A. aceti, groESL was induced by heat as well as aceticacid and ethanol. Furthermore, a groESL over-expressing strain was shown to be even more re-sistant to these stresses, suggesting that these gene products contribute to overall acetic acid re-sistance (Okamoto-Kainuma et al., 2002).

The aar operon consists of three genes,aarA,aarB, and aarC, that encode for proteins withhomology to enzymes involved in the citric acid pathway. Specifically,aarA encodes for the cit-rate-forming enzyme citrate synthase and aarC appears to encode for coenzyme A transferase,an enzyme also involved in acetate assimilation.Thus, assimilation of acetic acid (i.e., via its oxi-dation) would be one way to not only de-toxify the acid, but also enable the cell to use acetateas an energy source.

Recently, another assimilation or acetate oxidation pathway was suggested to be responsiblefor acetate resistance (Nakano et al., 2004).These researchers used two-dimensional (2-D) gelelectrophoresis to show that expression of the enzyme aconitase by A. aceti was significantlyincreased in response to an acetic acid shock (Nakano et al., 2004). Further, a strain harboringmultiple copies of the aconitase gene (and that over-expressed aconitase) produced moreacetic acid and was more acetic acid resistant compared to the parent strain.Apparently, the in-creased expression of aconitase,an enzyme involved in both the citric acid cycle and glyoxylatepathway, confers resistance by stimulating consumption and detoxification of intracellularacetic acid.Although effective for the cells, these strategies (the aar and aconitase) are not onesthat would likely be exploited for vinegar production, since the product is consumed and theoverall yield would not be enhanced.

Finally,as shown by proteome analyses (using 2-D gels),a number of proteins are synthesizedby Acetobacter and Gluconobacter in response to both short-term and long-term exposure toacetic acid (Lasko et al., 1997;Steiner and Sauer,2001).These proteins are distinct from the gen-eral stress-response proteins induced by heat shock and are thus referred to as either acetateadaptation proteins (Aaps) for adapted cells or acetate-specific stress proteins (Asps) for un-adapted cells.Although the function of most of these proteins has not yet been established, itappears (based on partial sequence analyses) that at least several of the Aaps are associated with

Box 11–1. Acetic Acid Tolerance in Acetobacter (Continued)

(Continued)



AcDH also contain pyrroloquinoline quinone(PQQ) as a prosthetic group (Figure 11–2).The latter serves as the primary electron ac-ceptor during the oxidation reactions, and isresponsible for the transfer of electrons to thecytochromes of the respiratory chain.

Protein analyses of purified PQQ-dependentAlDH and AcDH from several Acetobacter andGluconobacter species have shown that theseenzymes consist either of two or three subunits.These subunits contain heme-binding moietiesthat mediate intramolecular transport of elec-trons, in addition to the PQQ prosthetic groups.Acetic acid bacteria also have NAD- (or NADP)-dependent dehydrogenases present within thecytoplasm. However, the specific activities ofthese enzymes are up to 300 times lower thanthe membrane-bound,PQQ-dependent dehydro-genases. In addition, the pH optima of the latter

enzymes in Acetobacter is between 4.0 and 5.0,which is much closer to the normal physiologi-cal pH than the NAD-dependent dehydrogenaseenzymes, whose pH optima is above 7.0. Al-though the function of these enzymes has notyet been established, it has been suggested thatthey are involved in acetaldehyde and acetate assimilation. Finally, it should be noted that alco-hols and aldehydes other than ethanol and ac-etaldehyde can serve as substrates for both ofthese dehydrogenases. Thus, primary alcoholssuch as propanol, and secondary alcohols andpolyols, such as isopropanol and glycerol, all ofwhich can be present in mashes used for vinegarproduction, can be oxidized and converted intoacids, ketones, and other organic end products.Many of these compounds make important con-tributions to the aroma and flavor characteristicsof vinegar (see below).

membranes and may be involved in transport processes. Indeed, acetate transport and mem-brane-associated respiratory proteins could both be envisioned to promote acetate resistance(Steiner and Sauer, 2001).

References

Fukaya, M., H.Takemura, H. Okumura,Y. Kawamura, S. Horinouchi, and T. Beppu. 1990. Cloning of genes re-sponsible for acetic acid resistance in Acetobacter aceti. J. Bacteriol. 172:2096–2104.

Fukaya M., H.Takemura, K.Tayama, H. Okumura,Y. Kawamura, S. Horinouchi, and T. Beppu. 1993.The aarCgene responsible for acetic acid assimilation confers acetic acid resistance on Acetobacter aceti. J.Ferm. Bioeng. 76:270–275.

Lasko, D.R., N. Zamboni, and U. Sauer. 2001. Bacterial response to acetate challenge: a comparison of toler-ance among species.Appl. Microbiol. Biotechnol. 54:243–247.

Lasko, D.R., C. Schwerdel, J.E. Bailey, and U. Sauer. 1997.Acetate-specific stress response in acetate-resistantbacteria: an analysis of protein patterns. Biotechnol. Prog. 13:519–523.

Matsushita,K., T. Inoue,O.Adachi, and H.Toyama.2005.Acetobacter aceti possesses a proton motive force-dependent efflux system for acetic acid. J. Bacteriol. 187:4346–4352.

Menzel, U., and G. Gottschalk. 1985.The internal pH of Acetobacterium wieringae and Acetobacter acetiduring growth and production of acetic acid.Arch. Microbiol. 143:47–51.

Nakano, S., M. Fukaya, and S. Horinouchi. 2004. Enhanced expression of aconitase raises acetic acid resis-tance in Acetobacter aceti. FEMS Microbiol. Lett. 235:315–322.

Okamoto-Kainuma,A.,W.Yan, S. Kadono, K.Tayama,Y. Koizumi, and F.Yanagida. 2002. Cloning and charac-terization of groESL operon in Acetobacter aceti. J. Biosci. Bioeng. 94:140–147.

Russell, J.B.1992.Another explanation for the toxicity of fermentation acids at low pH:anion accumulationversus uncoupling. J.Appl. Bacteriol. 73:363–370.

Steiner,P., and U. Sauer.2001.Proteins induced during adaptation of Acetobacter aceti to high acetate con-centrations.Appl. Environ. Microbiol. 67:5474–5481.

Steiner, P., and U. Sauer. 2003. Long-term continuous evolution of acetate resistant Acetobacter aceti.Biotechnol. Bioeng. 84:40–44.

Box 11–1. Acetic Acid Tolerance in Acetobacter (Continued)



As noted earlier, the acetic acid pathway isusually considered an incomplete oxidation,since the substrate, ethanol, is only partially ox-idized and the acetic acid that is formed is notoxidized further. However, while this is true forsome acetic acid bacteria, such as Gluconobac-ter (a member of the so-called suboxydansgroup), most species of Acetobacter can oxi-dize acetic acid, provided conditions are suit-able.The latter include bacteria of the oxydansgroup, represented by the common vinegar-producing species A.aceti and A.pasteurianus.The absence of ethanol in the medium is themain condition necessary for acetic acid oxida-tion. Ethanol apparently represses synthesis ofcitric acid cycle enzymes; when ethanol is ab-sent, those enzymes are induced and completeoxidation of acetic acid to CO2 and H2O can

occur. Of course, in actual vinegar production,this so-called over-oxidation of acetic acid isparticularly undesirable, because it causes theliteral disappearance of the end product to theatmosphere.Thus, this is likely one reason whyvinegar fermentations have historically beenconducted in a semi-continuous mode (see be-low), in which a minimum amount of ethanol isalways present.

Although vinegar-producing cultures arefrequently used repeatedly without loss of via-bility or performance, genetic instability is notuncommon. Spontaneous mutations, at rela-tively high frequencies, have been reported forseveral Acetobacter sp., resulting in defects inseveral important functions. For example, mu-tants having lost the ability to oxidize ethanol,tolerate acetic acid, and form surface film (see

aldF aldG aldH

AldG AldH

49.0 kDa 16.7 kDa 84.1 kDa(773)(157)(444)

gdhA

Glucosedehydrogenase

AldF

B

A

AdhS

adhS

20 kDa(205)

adh gene cluster

Dehydrogenase Cytochrome C77.6 kDa 47.4 kDa

(157)(444)

Figure 11–3. Organization of genes coding for ethanol oxidation in Acetobacter. The alcohol dehydroge-nase complex (A) consists of three subunits. The dehydrogenase subunit and the cytochrome c subunit areencoded by two co-transcribed genes. The third gene, adhS, is not located near the adh gene cluster. Thefunction of its gene product, AdhS, is not known, although it is required for enzyme activity (except for thosestrains that apparently contain only subunits I and II). The aldehyde dehydrogenase gene cluster (B) also con-tains three genes coding for three subunits of aldehyde dehydrogenase. Based on sequence analysis, thegene products all contain domains that bind co-factors involved in electron transport. AldF and AldH areheme-binding proteins, and AldG binds an iron-sulfur cluster. A gene encoding glucose dehydrogenase isupstream of the aldFGH promoter (arrow). The predicted molecular mass (and amino acid residues) areshown for each protein. From Takemura, et al., 1993; Kondo, et al., 1995; and Thurner, et al., 1997.



below) have been obtained. Most industrialstrains of Acetobacter contain plasmids,whoseloss could conceivably account for phenotypicchanges. However, except for the presence ofantibiotic resistance genes, functions for otherplasmid-borne genes have not been identified,and most plasmids are considered to be cryp-tic and not essential for acetic acid formation.

In contrast, the presence of insertion se-quences (IS) in Acetobacter has been associated

with several loss of function mutations. Inser-tion sequences are small DNA sequences, usu-ally less than 3 kb, that contain inverted repeatsand other characteristic features.They can trans-pose or integrate, randomly or at specific siteswithin the genome, resulting in the disruptionof gene encoding regions. For example, in A.pasteurianus, A.aceti,G.xylinus,and other Ace-tobacter sp., the high-copy number IS element,IS1380, was found to integrate within the gene

Outermembrane

Cytoplasmic membrane

Periplasm

AcDHAlDH

ethanol acetaldehyde acetic acid

Cytoplasm

ethanol acetic acid

ethanol acetaldehyde acetic acid

NAD NADH2 NAD NADH2 acetyl CoA

2PQQ

2PQQ-H2

ethanol

aceticacid O2

H2O

2H+

Figure 11–4. Membrane biology, oxidation of ethanol, and acetate assimilation by Acetobacter aceti. Theoxidation reactions (upper panel), catalyzed by the PQQ-dependent enzymes alcohol dehydrogenase (AlDH)and acetaldehyde dehydrogenase (AcDH) occur within the periplasm. Ethanol can also be oxidized by cyto-plasmic NAD-dependent dehydrogenases, but at very low rates. Acetic acid assimilation to acetyl Co-A andsubsequent oxidation to CO2 via the citric acid cycle occurs only when ethanol is absent. The oxidation reac-tions are accompanied by transfer of electrons from PQQ to cytochromes of the respiratory chain (lowerpanel), leading to formation of a proton motive force and synthesis of ATP via oxidative phosphorylation. Oxy-gen serves as the terminal electron acceptor. Adapted from Matsushita, K., H. Toyama, and O. Adachi. 1994.Respiratory chains and bioenergetics of acetic acid bacteria, p. 247–301. In A.H. Rose and D.W. Tempest(ed.), Advances in Microbial Physiology, vol. 36. Academic Press, Ltd., London.



coding for cytochrome c, which, along with al-cohol dehydrogenase,is essential for ethanol ox-idation. In addition, transposition of another in-sertion sequence, IS1031, in A. xylinum, led tothe appearance of mutants that were unable tosynthesize the cellulose-containing surface filmnecessary during surface type fermentations(see below).Thus, it appears that the genetic in-stability associated with acetic acid bacteria islikely due to inactivation of genes necessary forgrowth and acetic acid formation by transpos-able IS elements.

Vinegar Technology

The first step in the manufacture of vinegar, asnoted earlier, is the production of an ethanolicsubstrate. For the most part, the relevant fer-mentations (i.e., wine and beer) are discussedseparately (Chapters 9 and 10). However, it isworth noting that ethanol from musts, ciders,or malt mashes are usually prepared with theeventual end product (i.e., vinegar) in mind.Thus, the starting materials do not necessarilyhave to be of the highest quality. However, thisdoes not mean that poor quality materialsshould be used, since many of the flavor andaroma properties of the finished vinegar are

derived from the starting material. Of course, ifdistilled ethanol is used as the substrate, thenthe resulting vinegar will lack most of these fla-vor and aroma characteristics.

The second step of vinegar production, con-version of ethanol substrate to acetic acid, canbe performed by one of several methods, all ofwhich rely on oxidative fermentation of ethanolby acetic acid bacteria.These processes are gen-erally referred to as the open vat method, thetrickling generator process, and the submergedfermentation process. The latter two are nowthe more widely used, especially for larger vine-gar manufacturers, since they can be performedin continuous mode and can reduce the fer-mentation time from several weeks to a matterof days. However, even the traditional open vatmethod can be run in a semi-continuous fashionand is still used in many parts of the world.Moreover,the open vat process is the method ofchoice for the premium types of vinegar, includ-ing various wine vinegars, such as Balsamic(Box 11–2).

Open vat process

The open vat process relies on surface growthof acetic acid bacteria in vats, barrels, jars, or

Box 11–2. The Special Appeal of Specialty Vinegars (or, “You Paid How Much for That Vinegar?”)

By far, most of the vinegar produced in the United States is used by the food industry in themanufacture of salad dressings, pickles, catsup, mustard, and a variety of other processed foods.In general, the vinegar used for these products should be inexpensive and plain-tasting. Eventhe more flavorful wine, malt, and cider vinegars available at the local grocery store and used astable vinegar carry rather modest retail prices.A visit to the gourmet section, however, will reveal that some specialty vinegar products are asexpensive as fine wines.Among the most well-known of the specialty vinegars are the sherryand balsamic vinegars.They are made, respectively, in the Jerez region of Spain and the Modenaregion of Italy.Although versions are now made in the United States and elsewhere, these areprotected names in Europe, where their production and authenticity are highly regulated.

Balsamic vinegar is striking in several respects. It has a very dark brown appearance, a strongcomplex aroma, and a sour, but slightly sweet flavor. Importantly, the fermentation of balsamicvinegar is quite different than traditional vinegars.The starting material is musts obtained froma variety of grapes,mainly Trebbiano,but also Lambrusco and Sauvignon.The musts are concen-trated two-fold by raising the temperature to near boiling, followed by simmering until the mix-ture is syrup-like.Thus, step-wise process can take as long as two to three days and results in a

(Continued)



sweet solution with a sugar concentration of 20% to 24%. Next, the material is inoculated witha “mother culture” (from a previous batch) and transferred into premium quality wooden bar-rels (examples include oak, chestnut, mulberry, and cherry wood).These wild cultures containvarious osmophilic yeasts, including strains of Saccharomyces and Zygosaccharomyces,and theacetic acid bacterium Gluconobacter.An alcoholic fermentation and the acetic acid fermenta-tion essentially occur at very near the same time, since Gluconobacter can oxidize sugars di-rectly to acetic acid.

Another unique feature of the balsamic fermentation process concerns the manner in whichthe vinegar is aged. Rather than aging the product in bulk in a single barrel, balsamic vinegar isaged in a series of steps in which the product is transferred, via decantation, from barrel to bar-rel (Figure 1).Each barrel is usually constructed from different woods,which contributes to theflavor and aroma of the finished product. During each transfer step, about half of the volumefrom the preceding barrel is moved to the next barrel.

The barrels (usually in sets of five or seven) are held in attics that are exposed both to verywarm and very cool temperatures, depending on the season.Warm temperatures stimulate fer-mentation, whereas cool temperatures promote sediment formation (which improves clarity).The rich and aromatic flavors associated with balsamic vinegar are due, in part, to the grapeconstituents and the volatile and non-volatile components extracted from the wood,but also tothe acids, esters and other metabolic end products produced by the microflora.To achieve De-nominazione di Origine Controllate (DOC) status and Protected Designation of Origin (PDO)certification, the vinegar must be aged for at least twelve years, but some balsamic vinegars areaged for as long as twenty-five to fifty years. Less than 12,000 liters of authentic balsamic vine-gar (Aceto Balsamico Tradizionale) are released for sale per year, thus it is not surprising that re-tail prices for 90 ml bottles can be more than $150.While more moderately priced versions ex-ist (Aceto Balsamico di Modena), they usually are diluted with wine vinegars or aged less.

Sherry vinegars are produced via the “solera” process in a manner similar to that used forSherry wines (Chapter 10).Specifically, a portion (less than one-third of the total volume) of thevinegar is transferred from cask (or “butt”) to cask. Each successive portion is more aged thanthe preceding material, such that the final product has a uniform and consistent quality.Agingtypically occurs for a minimum of six months, to as long as two years or more (to earn “Reserva”status). Flavors are, as for Balsamic vinegar,derived from fermentation, as well as the wood used

Box 11–2. The Special Appeal of Specialty Vinegars (or, “You Paid How Much for That Vinegar?”) (Continued)

Figure 1. Sequential barrel formation used during manufacture and aging of balsamic vinegar. Successivebarrels are made from different woods, and as aging progresses, the size of the barrels decreases.



trays. This was undoubtedly the first methodused for vinegar production, since it can beeasily performed and involves naturally-occur-ring or wild cultures.Although many variationsexist, barrel-type fermentations are typical ofthe processes that evolved in Europe and thatare still used today. The most well-known ex-ample is the Orleans process, which takes itsname from the French city where this particu-lar method was originally developed. Similarprocesses also evolved independently in theFar East,but the latter were based on rice wineand other types of alcohol-containing sub-strates, rather than on grape wine or cider.Theprinciple of the process is simple and straight-forward.The ethanolic substrate is placed in asuitable vessel, and the fermentation is initi-ated either by acetic acid bacteria that natu-rally contaminate the vessel or by a portion ofvinegar from a recent batch.The material is ex-posed to the atmosphere, but is otherwise leftundisturbed (hence, this process is sometimesreferred to as the “let alone”method).

In the open vat process, acetic acid bacte-ria grow only at the surface. Growth is accom-panied by production of a polysaccharide-containing film or pellicle that forms at theliquid-air interface. Although most strains ofAcetobacter are capable of pellicle formation,this property can occasionally be lost if cellsare grown under non-static (i.e., shaking) con-

ditions (see above). This is important duringopen vat processes, because an intact film isessential to the success of the fermentation.Any disturbance or disruption of the film may delay the fermentation. Thus, open vatprocesses may take several weeks before thefermentation is established, and severalmonths before the fermentation is complete.

Traditional open vat or surface film fermen-tations ordinarily operate in batch mode.Afterthe fermentation is complete, the vessel isemptied, then refilled with substrate. Theacetic acid bacteria, left over or added, mustthen re-establish surface growth. It is possible,nonetheless, to perform this fermentation in asemi-continuous mode, provided steps aretaken to maintain the surface film during sub-strate addition and product removal steps.

In the Orleans process, for example, the bar-rels are constructed such that a filling device(i.e.,a pipe) extends from just outside the top ofthe barrel all the way to the inside bottom (Fig-ure 11–5). Aeration is provided by holes (cov-ered with cheesecloth) drilled into the sides ofthe vessel.A tap is positioned at the bottom endthat is used to withdraw the product.The barrelis filled with wine to about 60% to 70% capacityand inoculated with a fresh vinegar culture(“mother of vinegar”).When the acetic acid fer-mentation is completed, it is then possible to re-move a portion (from one-third to as much as

during aging.The latter contribute important phenolic compounds associated with aged sherryvinegar.

Finally, it should be noted that distinctive specialty wine vinegars are not just produced inModena, Jerez, or Orleans, but also in the United States. A growing market for varietal vine-gars, made from single variety wines, has emerged, especially in the wine-producing regionsin California. These vinegars are fermented in barrels; some producers have adopted tradi-tional Orleans-style fermentors and long aging periods.

References

Cocchi, M., P. Lambertini, D. Manzini,A. Marchetti, and A. Ulrici. 2002. Determination of carboxylic acids invinegars and in Aceto Balsamico Tradizionale di Modena by HPLC and GC methods. J.Agric.Food Chem.50:5255–5261.

García-Parilla, M.C., F.J. Heredia, and A.M. Troncoso. 1999. Sherry wine vinegars: phenolic compositionchanges during aging. Food Res. Int. 32:433–440.

Box 11–2. The Special Appeal of Specialty Vinegars (or, “You Paid How Much for That Vinegar?”) (Continued)



two-thirds) of the fermented vinegar from thetap end, and to replace it with fresh wine orcider, through the filling device, without dis-rupting the film.The process takes two to threeweeks, or longer, for each completed cycle, de-pending on the temperature and the rate of vol-ume exchange.Although other methods requiresignificantly less time,the quality of vinegar pro-duced by the Orleans and similar barrel meth-ods is considered to be far superior to thoseproduced by more rapid methods.

Trickling generator processes

In the vinegar fermentation, the rate at whichethanol is oxidized to acetic acid depends onthe presence and availability of oxygen and thesurface area represented by the air-liquid inter-face. In other words, the ability of acetic acidbacteria to perform the acetic acid fermenta-tion is limited primarily by the diffusion ortransport of oxygen from the atmosphere tothe cell surface. It is possible to significantly ac-celerate the oxidation of ethanol by increasingthe surface area to which oxygen is exposed.This observation, which was originally de-scribed nearly 300 years ago by the famousDutch physician-chemist Hermann Boerhaave,forms the basis for the trickling generatorprocesses, which are sometimes referred to assimply quick vinegar processes.

The earliest examples of trickling genera-tors used on an industrial scale were theSchutzenbach and Ham processes, introducedin 1823 and 1824, respectively. In these sys-tems, which are still widely used, ethanolicsubstrates are circulated or trickled throughcylindrical fermentation vessels or vats con-taining inert packing materials, such as curledwood shaving, wood staves, or corn cobs (Fig-ure 11–6). Many other materials can be used,provided they do not contribute flavor or aredegraded. The main requirement is that theyhave a large surface area.

As the inoculated substrate or feedstockpasses from the top to the bottom of the ves-sel, the total surface area of the liquid materialincreases as it moves around and between theparticulate packing material. Growth of aceticacid bacteria will then occur at the air-liquidinterface, such that the ethanol concentrationdecreases and the acetic acid concentration in-creases during the transit of substrate from topto bottom. Holes can be drilled into the side ofthe vessel to ensure that aeration is adequate.When the substrate reaches the bottom sec-tion (the collection chamber), it may be re-turned to the top until the effluent is suffi-ciently acidic to be called vinegar.Alternatively,a second tank can be used to increase the oxi-dation rate and product throughput.The inocu-lated mash is fermented first in Tank 1, then

Figure 11–5. Schematic illustration of vinegarbarrel used for the Orleans process. Adaptedfrom Adams and Moss, 1995.



passed,with aeration, into Tank 2,and back andforth until all of the ethanol is converted tovinegar.About three days are required to con-vert a 12% (v/v) ethanol solution to a vinegarcontaining 10% to 12% acetic acid.

Many modern trickling generator-type sys-tems have been developed over the past sev-enty years based on the same principle as de-scribed above, but with more control andoperational features. The Frings generator, in-troduced in the 1930s, and still widely used today in various modified formats, provides ameans for incorporating air, in a counter-current direction,directly into the fermentationvessel.These systems can also accommodate aheat-exchange step, so that a constant tempera-ture can be maintained. In addition, the feed-stock is ordinarily dispensed via a sparging armthat breaks up the liquid into smaller droplets

that are more easily distributed within the inte-rior matrix.

Submerged fermentation

Advances in biotechnology, and in particular,the fermentation industry, have led to the de-velopment of modern industrial fermentors ca-pable of rapid, high throughput bioconversionprocesses. Although submerged fermentationsystems are now widely used for many indus-trial fermentation processes, they were actu-ally developed by the vinegar industry (and theHeinrich Frings Company, in particular) morethan fifty years ago. The Acetator (the Fringsfermentor), Cavitator, Bubble fermentor, andother similar units are constructed of stain-less steel, can be easily cleaned and sanitized,and can operate in batch, semi-continuous, or

pump

exhaust air

sparge

vinegar

inertpackingmaterial

air

XX

Figure 11–6. Trickling generator system.



continuous mode. Most of the vinegar pro-duced worldwide is now made using sub-merged fermentation systems.

The most important feature of the sub-merged fermentation systems is their ability toprovide rapid and efficient aeration. For exam-ple, the Frings Acetator is equipped with tur-bines that mix the liquid with air or oxygenand deliver the aerated mixture at very highrates inside the fermentor. The aeration sys-tem’s ability to break up air bubbles and facili-tate transfer of oxygen molecules from the gasphase to the liquid phase is essential, since thesuccess of submerged fermentation systems relies on the transfer of oxygen from themedium to the bacteria.The viability of Aceto-bacter cells can be quickly compromised ifaeration is even momentarily lost.

In addition to the aeration system, processcontrols for propeller speed (up to 1,750rpm), nutrient feed, foam-handling, and tem-perature further provide for consistent, semi-continuous operation. Temperature controlmay be of particular concern in large-scalefermentations because ethanol oxidation is anexothermic reaction and heat is released.Themedium temperature, if not controlled, can in-crease from an optimum of 28°C to 30°C to ashigh as 33°C. Even this modest increase intemperature can result in a serious decreasein fermentation productivity.

The size and design of submerged fermenta-tion systems vary considerably. Production ca-pacities of 25 liters per day of 10% acetic acidfor very small-scale fermentors, up to 30,000liters per day for large-scale reactors, are possi-ble. In general, the starting material containsboth ethanol and acetic acid (from a prior fer-mentation). After a single cycle of sixteen totwenty-four hours, the ethanol concentrationwill be reduced from about 5% to less than0.5%, and the acetic acid will increase fromaround 7% to 12%. Up to half of the vinegar isthen removed and replaced with fresh ethanolfeedstock.Cell growth during a typical fermen-tation cycle is modest, with only a single dou-bling of the initial population.

Post-fermentation processing

Depending on the manner in which the fer-mentation occurs, the vinegar may be rela-tively clear, in the case of barrel-aged product,or very turbid, in the case of submerged fer-mentation-produced product.Thus, vinegar ob-tained by the Orleans method or after agingand subsequent transfer in barrels may requirelittle, if any, filtration. In contrast, removing sus-pended cells from vinegar produced by sub-merged fermentation generally requires moreelaborate filtration treatments, including theuse of inert filtration aids.Vinegar produced bytrickling fermentation processes may also re-quire a filtration treatment.

Although it may seem counter-intuitive topasteurize vinegar (after all,what can grow in a1 M acetic acid solution?), heat treatments (upto 80°C for thirty to forty seconds) are com-mon in the vinegar industry.Usually the targetsare acetic acid bacteria, lactic acid bacteria,and wild yeasts and fungi. As noted earlier,some vinegars are aged, usually in wooden bar-rels, although this practice is usually reservedfor high quality wine vinegars (Box 11–2).

Spoilage, bacteriophages,and other problems

Spoilage of vinegar by microorganisms is rare,although aceto-tolerant fungi, such as Monil-iella acetoabutens, have occasionally beenknown to grow in raw vinegar.The more com-mon problem, at least for open vat or tricklingtype processes, is the occasional contaminationwith mites and flies.The vinegar eel, in particu-lar, was once quite common, especially in tradi-tional open vat or trickling fermentation sys-tems, but is now infrequently present.The eel isactually a small worm (i.e., a nematode classi-fied as Anguillula aceti) whose main effect isaesthetic rather than influencing the fermenta-tion or product quality.Vinegar eels are readilyremoved by filtration and easily killed by heat.

Given that vinegar fermentations are con-ducted in the presence of ethanol and acetic



acid, one might expect that bacteriophageswould not be a serious problem. However, theabsence of a heating step for the substrate, therepeated use of the same culture, and the gen-erally open manner in which vinegar fermenta-tions are conducted, certainly provides theright opportunities for phage proliferation.Even submerged fermentations in enclosedstainless steel fermentors do not precludephage infections, unless pasteurized substratesare used or other phage control measures areapplied. Nonetheless, it appears that althoughphage problems have occasionally been re-ported, the occurrence is apparently low, andthere are only a few published accounts ofphage infections in vinegar production facili-ties (although the incidence of unpublishedphage problems may be higher).

In general, phages are more commonly asso-ciated with trickling generator processes, be-cause those fermentations are usually conductedby a diverse mixture of strains where phageswould likely proliferate. In fact, trickling pro-cesses are considered to be the main reservoirfor Acetobacter phages. However, submergedfermentation systems are likely more sensitive tophage problems,since they often rely on a singlestrain whose infection could conceivably resultin a slow or failed fermentation.

Perhaps the most frequent microbial prob-lem associated with vinegar spoilage is causedby oxidation of acetic acid by Acetobacter. Asdescribed earlier, these bacteria have the meta-bolic capacity to oxidize the acetic acid to CO2

and water. However, the genes that encode forcitric acid cycle enzymes necessary for aceticacid oxidation are not induced in the presenceof ethanol. Moreover, acetic acid oxidation isapparently repressed when the acetic acidconcentration is below 6% to 7%.Thus, underthe ordinary conditions in which the aceticacid fermentation occurs (i.e., when theethanol concentration ranges between 2% and10% and the acetic acid concentration is above6%), the oxidation pathway is not active. Onlyunder prolonged fermentations, when theethanol is completely dissipated and when the

acetic acid concentration is low, would over-oxidation occur.

Vinegar Quality

At least two criteria must be considered to as-sess the quality of vinegar. In contrast to otherfermented foods, where authenticity is rarelydisputed, vinegar can be adulterated, eitherwith less expensive vinegar or with otheracidic agents. In rare instances, adulteration iscaused by even more dubious means.For exam-ple, acetic acid produced via chemical synthe-sis can be diluted and then marketed (illegally)as fermentation-derived vinegar. Thus, the firstquality criterion is based on establishing thatthe vinegar is actually vinegar (and not just di-luted acetic acid) and that the type of vinegarindicated on the label accurately representswhat is in the product. Satisfying this require-ment,however, is no easy matter, and requires astatistical multivariate approach (reviewednicely by Tesfaye et. al., 2002). Discriminatinganalyses that are used to distinguish samples of vinegar involve measuring selected chemicalconstituents, including minerals, alcohols, acids,phenolics, and other volatile compounds.Usingsuch an approach, it is even possible to differ-entiate the fermentation process used to pro-duce the vinegar (e.g., traditional versus quickacetification methods).

The other main characteristics on whichvinegar quality is based are those that relate toflavor,aroma,and other organoleptic properties.Vinegar flavor is particularly influenced by theraw ethanolic material from which it was made.Thus, wine vinegars contain a mixture of phe-nolic compounds that are ordinarily present ingrapes. And although acetic acid is by far thepredominant flavor present in vinegar, othervolatile flavor compounds are also present andcontribute to the overall flavor profile of vine-gar.The enzymes—alcohol dehydrogenase andaldehyde dehydrogenase—that oxidize ethanoland acetaldehyde, respectively, also oxidizeother alcohols and aldehydes.Thus, if the must,mash, or other starting material contains these



substrates, a diverse array of end products canbe obtained.These products include formic,pro-pionic, and butyric acids, and glycerol. Othervolatile compounds commonly found in vinegarinclude diacetyl, acetoin, and ethyl acetate.

Of course, the ultimate measure of vinegarquality,as for other fermented food products, isbased on sensory analysis. Performing theseanalyses for vinegar, however, is particularlychallenging since acetic acid has a strong,over-powering flavor.Therefore, samples can be di-luted, mixed with lettuce, or even served insmall amounts in wine glasses. In fact, sensoryanalysis of vinegar is not unlike that used forwine analysis, in that both generally requiretrained tasters and both rely on similar descrip-tors (e.g., woody, fruity, etc.).

BibliographyAdams, M.R.Vinegar. p. 1–44. In B.J.B.Wood (ed.) Mi-

crobiology of Fermented Foods, Volume 1,1998. Blackie Academic and Professional (Chap-man and Hall). London, United Kingdom.

Adams, M.R., and M.O. Moss. 1995. In Food Microbi-ology. Royal Society of Chemistry. Cambridge,U.K.

Beppu,T. 1993. Genetic organization of Acetobacterfor acetic acid fermentation. Antonie VanLeeuwenhoek 64:121–135.

Cleenwerck, I., K. Vandemeulebroecke, D. Janssens,and J. Swings. 2002. Re-examination of the genusAcetobacter, with descriptions of Acetobactercerevisiae sp. nov. and Acetobacter malorum sp.nov. Int. J. Syst. Evol. Microbiol. 52:1551–1558.

Coucheron, D. 1991.An Acetobacter xylinum inser-tion sequence element associated with inactiva-tion of cellulose production. J. Bacteriol. 173:7523–5731.

Fuchs, G. 1999. Oxidation of organic compounds, p.187–233. In J. W. Lengeler, G. Drews, and A. S.Schlegel (ed.), Biology of the Prokaryotes. Black-well Science, Stuttgart, Germany.

Garrity,G.M., J.Bell,and T.G.Lilburn.2004.TaxonomicOutline of the Prokaryotes. Bergey’s Manual ofSystematic Bacteriology, 2nd ed. Release 5.0,Springer-Verlag. DOI: 10.1007/bergeysonline.

Holt, J.G.,N.R.Krieg,P.H.A.Sneath, J.T. Staley, and S.T.Williams. 1994. Gram-negative aerobic/micro-aerophilic rods and cocci, p. 71–174. In Bergey’s

Manual of Determinative Bacteriology, 9th Ed.,Williams and Wilkins Co. Baltimore, Maryland.

Kondo, K., T. Beppu, and S. Horinouchi. 1995.Cloning, sequencing, and characterization of thegene encoding the smallest subunit of the three-component membrane-bound alcohol dehydro-genase from Acetobacter pasteurianus. J. Bacte-riol. 177:5048–5055.

Krahulec, J., M. Kretová, and J. Grones. 2003. Charac-terization of plasmids purified from Acetobacterpasteurianus 2374. Biochem. Biophys. Res.Comm. 310:94–97.

Lisdiyanti, P., H. Kawasaki, T. Seki, Y. Yamada, T.Uchimura, and K. Komagata. 2000. Systematicstudy of the genus Acetobacter with descriptionsof Acetobacter indonesiensis sp. nov., Acetobac-ter tropicalis sp. nov., Acetobacter orleanensis(Henneberg 1906) comb. nov., Acetobacter lo-vaniensis (Frateur 1950) comb. nov., and Aceto-bacter estunensis (Carr 1958) comb. nov. J. Gen.Appl. Microbiol. 46:147–165.

Matsushita,K.,H.Toyama,and O.Adachi.1994.Respi-ratory chains and bioenergetics of acetic acidbacteria, p. 247–301. In A.H. Rose and D.W.Tem-pest (ed.), Advances in Microbial Physiology,vol. 36.Academic Press, Ltd., London.

Sievers, M., and M. Teuber. 1995. The microbiologyand taxonomy of Acetobacter europaeus in com-mercial vinegar production. J. Appl. Bacteriol.Symp. Suppl. 79:84S–95S.

Sokollek, S.J., C. Hertel, and W.P. Hammes. 1998. De-scription of Acetobacter oboediens sp. nov. andAcetobacter pomorum sp.nov., two new speciesisolated from industrial vinegar fermentations.Int. J. Syst. Bacteriol. 48:935–940.

Stamm, W.W., M. Kittelmann, H. Follmann, and H.G.Trüper. 1989.The occurrence of bacteriophagesin spirit vinegar fermentation. Appl. Microbiol.Biotechnol. 30:41–46.

Takemura, H., S. Horinouchi, and T. Beppu. 1991.Novel insertion sequence IS1380 from Acetobac-ter pasteurianus is involved in loss of ethanol-oxidizing ability. J. Bacteriol. 173:7070–7076.

Takemura, H., K. Kondo, S. Horinouchi, and T. Beppu.1993. Induction by ethanol of alcohol dehydro-genase activity in Acetobacter pasteurianus. J.Bacteriol. 175:6857–6866.

Tesfaye,W.,M.L.Morales,M.C.García-Parilla,and A.M.Troncoso. 2002. Wine vinegar: technology, au-thenticity and quality evaluation. Trends FoodSci.Technol. 13:12–21.

Thurner, C., C.Vela, L.Thöny-Meyer, L. Meile, and M.Teuber. 1997. Biochemical and genetic character-



ization of the acetaldehyde dehydrogenase com-plex from Acetobacter europaeus. Arch. Micro-biol. 168:81–91.

Yamada,Y., K. Katsura, H. Kawasaki,Y. Widyastuti, S.Saono, T. Seki, T. Uchimura, and K. Komagata.

2000.Asaia bogorensis gen. nov., sp. nov., an un-usual acetic acid bacterium in the �-Proteobacte-ria. Int. J. Syst.Evol. Microbiol. 50:823–829.


12

Fermentation of Foods in the Orient

“Chiang (fermented soybean paste) is to food, what a general is to an army.”Yen Shiu-Ku, seventh century writer

One of the main differences between the fer-mented foods that evolved in West and thosethat evolved in the Far East was that the latterpopulations depended far less on animal prod-ucts. Several reasons likely accounted for thisdifference. First, religions practiced in the Ori-ent often excluded meat products and insteadpromoted diets based on plant proteins. Bud-dhism, for example, which developed in thesixth century B.C.E., prohibited animal foods inthe diet. Second, that the Far East was generallymore densely populated meant that less spacecould be devoted to animal agriculture sogreater emphasis was placed on growing plantfoods for human consumption. In addition, eco-nomic pressures led to greater reliance on inex-pensive food commodities, such as cereals,grains,and legumes.Finally, the ready availabilityof fish and seafood provided an abundantsource of inexpensive,high-quality protein.

The other major differences between fer-mented foods of the West and East relate di-rectly to the fermentation substrates, the man-ner in which the fermentations occur, and thetypes of microorganisms that are involved.The substrates used for the fermented foodsof the West, for example, are usually simplesugars, whereas in many of the Asian-type fer-mentations, the substrates consist largely ofcarbohydrates in the form of starch and other

History

Like the bread, wine, dairy, and other fermenta-tions described in earlier chapters, Asian-typefermented foods also evolved thousands ofyears ago. In addition, these Asian fermentedfoods had many of the same general character-istics and properties as those that developed inMiddle Eastern and Western cultures. That is,the products were comprised of ingredientsnative to their geography, they had enhancedfunctional properties, and they were well pre-served. Likewise, Asian fermented foods werealso subject to cultural, economical, and reli-gious influences.

Despite these similarities, the starting mate-rials, the specific types of products producedin the Far East, and the means by which fer-mentation occurs are dramatically differentfrom products that evolved in the West. Also,whereas the fermented foods industry in West-ernized countries, and especially in the UnitedStates, have become large, mechanized, andtechnologically well defined, production meth-ods for fermented foods in the Orient varygreatly, with many small manufacturers still inoperation. The trend toward large-scale pro-duction, however, has now become evident inChina, Japan, Malaysia, Thailand, Korea, andmany other Far East countries.

419





polysaccharides.There are, for example, littleif any free fermentable sugars present in riceand soybeans. Therefore, the manufacture ofthe Asian fermented products must start witha step that converts the starch to sugars thatcan be used by fermentative organisms. In theWest, there is one product that shares thisproblem—beer.

As described in Chapter 9, the conversion ofstarch to free sugars is catalyzed by various amy-lolytic enzymes present in malt during a stepcalled mashing.The free sugars, mainly maltoseand glucose, are then readily fermented toethanol and carbon dioxide by yeasts. In theproduction of Asian fermented products, a verysimilar mashing process occurs,with one signif-icant difference. The enzymes necessary for saccharification (i.e., starch hydrolysis) are sup-plied not by malted cereals,but rather by fungalorganisms, previously grown on cereal matter.Like malt, the production of this enzyme-ladenfungal material is itself an important part of theoverall manufacturing process.

Another important difference between fer-mented foods of the East and West, as notedabove, concerns the very nature of the fermen-tation. For the most part, dairy, meat, vegetable,and ethanolic fermentations are conducted un-der predominantly anaerobic conditions.In con-trast, aerobic conditions are necessary (at least,for certain critical parts of the process) to pro-duce many of the Asian fermented products.Moreover, whereas pure starter cultures, con-taining defined strains, are now well acceptedand widely used for most fermented food prod-ucts in the West, defined cultures are infre-quently used (except by large manufacturers) inthe production of Asian fermented foods. Fi-nally, although fungi are used in the manufac-ture of a few Western-type fermented foods,such as some cheeses and sausages, they are essential in most of the Asian fermented foods.In fact,the food products described in this chap-ter are often referred to simply as fungal-fermented foods.

It should be noted that the consumption ofAsian fermented or fungal-derived foods is no

longer confined to Asia or Asian populations.Chinese foods have long been part of the West-ern cuisine, but in the past two decades, Japan-ese, Thai, Vietnamese, Indonesian, Korean, andother cuisines from the Far East are nowwidely consumed throughout the world. It isremarkable, however, that few consumers real-ize that so many of these foods contain, as pri-mary ingredients or constituents, fermentedfoods products. Perhaps even more surprisingis that many of the Asian fermented foods havecaptured a large part of the U.S. ethnic foodsmarket, such that many of these products, in-cluding tempeh, miso, kimchi, and fish sauce,are available in American grocery and specialtyfood stores. Due to the savory, meat-like flavorsome of the products impart, they have also be-come popular as vegetable-based meat substi-tutes (as “faux” meats). There are now severalmanufacturing facilities in the United Statesthat produce these products, and their popu-larity is predicted to increase in coming years.

Types of Asian Fermented Foods

There are hundreds of different types of fer-mented foods produced in China, Japan, thePhilippines, and throughout Asia (Beuchat,2001). However, there are two general types offermented foods that are associated with or areindigenous to Eastern or Asian cuisines: thosethat are plant-based and those that are fish-based.The former are made using primarily soyand rice as substrates, but other grains andlegumes are also used. For the most part, thesesoy-based fermentations have been industrial-ized and the products are now produced on alarge scale. In addition, many of the microbio-logical and biochemical events that occur dur-ing the fermentation have been studied and de-fined. The fish-based products use varioustypes of fish and seafood that are indigenous tothe Orient. However, in contrast to the soy- orplant-based fermentations, the fish fermenta-tions are generally not well defined, nor hasthere been,until recently,very much publishedresearch (at least in English-language journals)


Fermentation of Foods in the Orient 421

on the microorganisms and biochemical reac-tions involved in their manufacture.

Plant-based Fermentations

Soy and rice are the most frequent substrates forAsian fermented foods.Wheat flour is also oftenincluded as an ingredient in many of these prod-ucts. With few exceptions, however, there is acommon starting material—koji—that is essen-tial for most Asian fermented foods. As will bedescribed in more detail below, koji is simply amoldy mass of grain, and is derived from theChinese word meaning “moldy grain.” In somecases, the koji mold is added to a portion of theraw material which is later added to the remain-ing substrate (analogous, in a way, to a bulk cul-ture or bread sponge). In other applications, thekoji mold is added to the entire raw material.Koji is used not only for production of the manysoy sauce-type products that will be describedlater, but also for sake and related rice wines.Because of the important role koji plays in somany of the products that will described, itsmanufacture and its microbiological and enzy-matic properties are treated separately.

Koji and Tane Koji Manufacture

First, it is necessary to recognize there aremany types of koji used in the Far East,and thateach fermented food requires a specific type ofkoji.Thus, Japanese soy sauces generally use akoji that is different from the one used to makeChinese-style soy sauces,and both are differentfrom the koji used for sake manufacture. Gen-erally, koji can be referred to by its intendedproduct (e.g., sake koji or shoyu koji) or by thesubstrate from which the koji is prepared (e.g.,rice koji, barley koji, or soybean koji).

Despite the type of koji or how it is made,the purpose and function are always the same.Namely, the koji provides a source of enzymesnecessary to convert solid, raw materials con-taining complex and non-fermentable sub-strates into soluble, simple, metabolizable prod-ucts that can be easily fermented by suitable

microorganisms. It is also important to realizethat koji contains not only amylolytic enzymes,but an array of enzymes capable of hydrolyzingproteins and peptides, lipids, cellulose, pectin,and other complex substrates. The koji alsoserves as a source (often the main source) ofsubstrates for the enzymes produced by thekoji fungi.

Raw materials preparation

The manufacture of koji starts by treating theraw materials.When soybeans are used, they arefirst soaked for about twelve hours in severalchanges of water. For traditional koji-making,they are then cooked, usually with steam underpressure, for one hour—a process called puff-ing. For some products, soybean koji also con-tains wheat, which is prepared by roastingwheat kernels (or wheat flour) at a high tem-perature, followed by a crushing step. Rice kojifor sake manufacture is prepared from polishedrice (rice minus the bran) that is steeped andsteamed, similar to soybean koji.

Despite the seemingly simple process in-volved in preparing the koji substrate, numer-ous innovations have been designed to enhancethe digestibility of the raw soy beans and ulti-mately improve yield and product quality. Forsoy sauce, digestibility or yield is based, in largepart, on the amount of total protein nitrogenthat is converted to amino nitrogen during themashing step. However, how the soy beans areinitially heat processed influences the rate andextent that the soy proteins are hydrolyzed.Thisis because only denatured soy proteins are hy-drolyzed by fungal proteinases, whereas nativeproteins are not.Thus, a cooking step is essen-tial.Yields of about 82% can be achieved when atraditional cooking process (e.g., 118°C at 0.9kg/cm2 for 45 minutes) is used.In contrast,high-temperature, short-time cooking not only en-hances protein extraction, but also gives aminoacid yields greater than 90%. It is now possible,for example, to perform the cooking step in lessthan two minutes, using temperatures and pres-sures above 150°C and 5 kg/cm2, respectively.



Microorganisms

Once the koji substrates are prepared, there areessentially two means by which they can be in-oculated. One is simply to use a pure culturecontaining spores of Aspergillus oryzae and/or Aspergillus sojae. Alternatively, the mixturecan be inoculated with 0.1% to 0.2% of a seedculture called tane koji.Tane koji is made by in-oculating soaked, steamed, and cooled rice(usually brown rice) with spores of A. oryzaeor A. sojae. The inoculated rice is incubatedovernight, then transferred and distributedevenly into shallow trays and incubated at 30°Cfor five to seven days. Intermittent mixing pro-vides aeration necessary for fungi growth.Even-tually, the moldy rice mixture is dried and pack-aged. This tane koji material is then used toinoculate other koji mixtures.

Ordinarily, multiple strains are used to maketane koji, depending on the final product thatis manufactured. In general, koji mold strainsshould produce proteolytic, peptidolytic, andamylolytic enzymes with high activity,and theyshould sporulate and grow well on their sub-strate. Some species of Aspergillus are knownto produce aflatoxin and other mycotoxins.However, none of the koji strains that havebeen isolated and studied produce these tox-ins under production conditions. Finally, kojistrains should be genetically stable and pro-duce products having consistent flavor andcolor properties.

After the koji substrate (rice, soybeans,wheat,barley) has been inoculated with either apure spore culture or the tane koji, it is mixedand incubated in large rectangular trays or

boxes (5 � 12 meters) with a depth of about 30to 40 cm.Very large producers may use vats thatcan accommodate even greater quantities. Per-forations in the trays enhance air and moisturecirculation. Temperature control is particularlyimportant because some proteolytic enzymestend to be produced at lower temperatures(25°C to 30°C), and amylases are produced athigher temperatures (�30°C to 35°C). Modernmanufacturers now rely on temperature pro-grams to maximize enzymatic activities. In fact,the incorporation of continuous cookers, auto-mated mixers, and other modern devices hasmade it possible to automate the entire koji-making operation so that as much as 3,000 kg ofkoji can be produced per hour.After two to fourdays at 30°C, the mass should be completelycovered throughout with mold growth.

Manufacture of Soy Sauce andRelated Products

Soy sauce is one of the most widely consumedproducts in Asia. In Japan, per capita consump-tion is more than 10 liters per person per year,or more than 30 g per day. Dozens of differenttypes of soy sauces are manufactured in Asia. Infact, even within the same country, there maybe several distinct products, each having theirown particular qualities and each made ac-cording to specific manufacturing procedures(Table 12–1).

Moreover, quality standards further distin-guish one product type from another. For ex-ample, in Japan (where the name for soy sauceis shoyu), three shoyu production methodsand five types of shoyu products are recog-

Table 12.1. Types of soy sauces.

Product Country Description

Shoyu Japan Five types based on soybean and wheat contentChiang-yiu China Similar to shoyu tamari (mostly soybeans)Kecap Indonesia Two types, manis (sweetened) and asin (salty)Kicap Malaysia Tamari typeKanjang Korea Tamari typeSee-iew Thailand Two types, sweetened and saltyToyu mansi Philippines Lemon-like flavored



nized by the Japan Agricultural Standard(Table 12–2). In addition, there are three qual-ity grades for each Japanese shoyu—Special,Upper, and Standard—that are assigned basedon manufacturing methods, composition, fla-vor, and color.The most widely consumed vari-ety of shoyu in Japan is koikuchi, which con-tains 50% wheat. In contrast, tamari, anotherJapanese shoyu, contains little or no wheat.Similar versions of these soy sauce products,containing either part wheat or no wheat, ex-ist in China,Taiwan, and other Asian countries.

Koji

The manufacture of traditional soy sauce orshoyu starts with preparation of the raw mate-rials and addition of the koji (Figure 12–1).Thesoy fraction consists of either whole beans, soymeals, or soy flakes. It is now common to usede-fatted flakes rather than whole beans orflakes to improve yield and reduce fermenta-tion times. When beans are used, they arewashed, sorted, and soaked overnight, thencooked under pressure and cooled rapidly.Flakes are simply soaked and cooked (as forwhole beans). At the same time, whole wheatkernels, if included, are roasted and crushed.The wheat adds flavor and color, reduces themoisture so that growth of undesirable bacte-ria is minimized, enhances mold growth, andcontributes glutamic acid-rich proteins. Whenwheat is used, the soy-wheat ratio depends onthe manufacturer’s preference and can rangefrom 50:50 to 67:33.

The soy bean-wheat mixture is then inocu-lated with either tane koji or a pure culture ofsuitable fungal strains of A.oryzae or A.sojae tostart the koji fermentation.As described above,the inoculated material is incubated in largetrays, boxes, or vats that are perforated to allowair and moisture to circulate throughout the ma-terial and to enhance fungal growth. The tem-perature is maintained at about 30°C, whichmeans that the incubation rooms or vats musthave cooling capacity, since fungal growth gen-erates heat. Once the koji is mature (i.e., cov-ered completely with mold), it is fully devel-oped and ready to be used for the fermentation.

Mashing

It is during the next step, mashing, where thekoji enzymes begin to hydrolyze proteins, poly-saccharides,and other substrates,and where mi-croorganisms begin to use the products of thesereactions. First, however, a high salt brine con-taining 20% to 25% sodium chloride is added tothe solid material.The volume added may vary,depending on the manufacturer’s specifica-tions,but a ratio of about 1:1.2 to 1:1.5 (solid tobrine) is normal.The high salt concentration inthe mash (ranging from 16% to 19%) restrictsgrowth of microorganisms to only those thatare especially halo- or osmotolerant.The mashmaterial, referred to as “moromi”or the “moromimash,”is allowed to ferment in large (300,000 L)tanks for up to a year. Whereas wooden tankswere commonly used, these have largely beenreplaced first by concrete and then resin-coated

Table 12.2. Compositions of different types of shoyu1.

Total N2 Reducing sugars3 EthanolProduct(g/100 ml) (g/100 ml) (v/v) pH Color

Koikuchi 1.6–2.0 2.8–5.5 2.1–2.5 4.7–4.9 Dark brownUsukuchi 1.2 4.0–5.0 2.1–2.6 4.8 Light brownTamari 1.8–2.6 2.4–5.3 0.1–2.4 4.8–5.0 Dark red-brownShiro 0.5 14–20 0–0.1 4.5–4.6 Yellow/tanSaishikomi 2.0–2.4 7.5–9.0 0–2.2 4.6–4.8 Dark brown

1 Adapted from Yokotsuka and Sasaki, 1998 and Fukushima, 19892 Total nitrogen, comprised of amino acids, small peptides, and ammonium salts3Consisting of about 85% glucose, 10% galactose � mannose, and 5% pentoses



iron tanks. Moromi held at lower temperatures(15°C) are thought to produce a better qualityproduct,but take longer. In contrast,higher tem-peratures (as high as 35°C) can be used to com-plete the fermentation in as little as three to fourmonths.Agitation and forced aeration generallyenhances enzymatic activity and microbialgrowth during mashing.

Moromi enzymology

The koji molds, A. oryzae and A. sojae, producea wide assortment of enzymes. They are espe-cially prolific at secreting proteinases, pepti-dases, cellulases, and amylases as they grow inthe koji. During the mashing step, these en-zymes degrade soy (and wheat) proteins,

Soy beans

Inoculate with tane koji orAspergillus oryzae /Aspergillus sojae

Incubate aerobically at 28 - 32°C for 3 days to make koji

Pasteurize (and re-filter)

Soak 16 hours

Steam/autoclave for 1 hour

Cool to 25°C and mix (1:1 ratio)

Roast

Wheat

Crush

Mix koji with salt brine (23%) in a ratio of 1:1.5 to make moromi or mash

Ferment

Press (filter)

Bottle

Figure 12–1. Shoyu flow chart.



starches, and other macromolecules to endproducts that are either necessary as nutrientsor that can be fermented by the mash mi-croflora.As noted above,the actual amounts andspecific types of enzymes produced by koji As-pergillus can vary,depending on the incubationtemperature, strain, and substrate used duringkoji manufacture. In general, the lower the tem-perature, the greater will be the enzyme activity(within a range of 20°C to 35°C).

Protein hydrolysis occurs via one of severaldifferent fungal proteinases and peptidases,each having specific pH optima and substratepreferences. In terms of enzyme activity, alka-line proteinase has more than ten times the ac-tivity of other proteinases. This enzyme is ac-tive between pH 6.0 and 11.0, although its pHoptima is about pH 10.0. A semi-alkaline pro-teinase (pH optima 8.3), as well as at least twoneutral- proteases and three acid-proteases (pHoptima around 7.0 and 3.0, respectively), allcontribute to hydrolysis of soy proteins. Thepeptides that are generated are further hy-drolyzed by peptidases that also have individ-ual pH optima and are specific for particularamino or carboxy amino acid residues. En-zymes that release glutamic acid or that con-vert glutamine to glutamic acid are especiallyimportant, because this amino acid is responsi-ble for much of the flavor-potentiating charac-teristics of the soy sauces. Salt-tolerance is another important property of koji pro-teinases, since they will still retain activity inthe presence of high salt concentrations. Anearly complete hydrolysis of soy proteins isdesirable, since the subsequent lactic acid andyeast fermentation, color and flavor develop-ment, and overall organoleptic quality of the

finished soy sauce depend on amino acid for-mation during the mashing step.

Most of the amylolytic activity present dur-ing mashing is due to fungal �- and �-amylasesthat effectively convert starch to simple sug-ars.Although some sugars are used to supportmold growth, most remain in the mash wherethey will later serve as substrates during thefermentation stage of the process.This simplesugar fraction is comprised of several reducingsugars, including glucose and maltose, as wellas xylose, galactose, and arabinose that are re-leased from other polysaccharides present inthe soy or wheat mixture (discussed below).The reducing sugars play an especially impor-tant role in color and flavor development,sincethey react with free amines during non-enzymatic browning reactions.

The other main group of enzymes importantduring mashing are the cellulases, pectinases,hemicellulases, and other tissue-degrading en-zymes. They enhance extraction of substratesfrom the soy bean and wheat tissues, therebyincreasing yield and nutrient availability.Forma-tion of pentoses and other sugars also provideadditional substrates for microorganisms andfor browning reactions.

Fermentation

Immediately after the addition of the salt brineto the koji, the mash will contain not only theAspergillus strains, but also bacteria and yeaststhat were present as part of the natural koji microflora (Table 12–3). The fungi are salt-sensitive and quickly lyse and die (Figure 12–2).Other organisms initially present includemostly Micrococcus and Bacillus species, with

Table 12.3. Microorganisms important in soy sauce.

Fungi Bacteria Yeast

Aspergillus oryzae Tetragenococcus halophilus Zygosaccharomyces rouxiiAspergillus sojae Lactobacillus delbrueckii Zygosaccharomyces soyaMucor sp. Leuconostoc mesenteroides Candida versitalisRhizopus sp. Torulopsis sp.

Hansenula sp.



starting levels as high as 108 cells per gram. Incontrast, lactic acid bacteria and various yeastsare present at much lower initial cell concen-trations (101 to 104 per gram).

Although the initial pH of the mash (6.5 to7.0) is perfectly fine for all of these organisms,the high salt concentration poses a substantialbarrier.A mash containing 18% salt will have anosmolality of more than 6 Osm and a water ac-tivity (aw) of about 0.88, which is lower thanmost of the moromi organisms can tolerate.Thus, the salt-sensitive organisms, includingthe Micrococcus, wild yeasts, and most of theBacillus, will not survive for long and will beundetectable in the mash within a month ortwo. In contrast, the restrictive conditions willselect for those organisms capable of tolerat-ing a high salt, high osmotic pressure, low aw

environment. Among the lactic acid bacteriathat can tolerate these conditions and predom-

inate the early stages of the mash fermentationare strains of Lactobacillus delbrueckii andTetragenococcus halophilus (formerly Pedio-coccus halophilus).Although they may be pre-sent initially at low levels (�103/ml) in themash, they can reach cell densities of 107

cells/ml or higher after six to eight weeks. It isnow common to add selected strains of theseorganisms to the moromi, rather than rely ontheir natural occurrence.

Salt-tolerant yeasts, Zygosaccharomycesrouxii and Candida versitalis, are also typi-cally added to the mash.These yeasts are evenmore salt- and acid-tolerant than the lactic acidbacteria. When the mash pH falls below 5.0,growth of L. delbrueckii and T. halophilus (inthe presence of more than 18% salt) will be in-hibited, and yeasts will dominate the rest ofthe fermentation. Eventually, when the pHreaches 4.0, even Z. rouxii will stop growing.

104

107

106

105

5

4

7

6

0 4 8 12 16 20 24 3228

pH

Weeks

)thgiew yrd(

marg/rebmun llec goL

108

pH

Yeast

Lactic acid bacteria

Figure 12–2. Idealized model for successive growth of lactic acid bacteria and yeast during the soy saucefermentation.



The manner in which the lactic acid bacteriaand yeasts grow during the soy sauce fermen-tation,therefore,resembles that of a succession-type fermentation.

A wide variety of fermentation end prod-ucts accumulates in the mash and is responsi-ble for the complex flavor of the finished soysauce (Table 12–4). Growth of the homofer-mentative T.halophilus on glucose results pri-marily in lactic acid formation. Fermentationof other sugars, particularly pentoses, how-ever, may yield heterofermentative productsincluding ethanol, acetate, and CO2, as well as other organic acids and alcohols. Acids,amines, ammonia, and other products may beproduced from amino acid metabolism. Thesubsequent sugar fermentation by the yeastsalso generates ethanol and CO2 via theethanolic pathway, as well as many otherhigher alcohols, esters, furanones, and otherflavor volatiles.

Pressing and refining

After the fermentation is complete and the mo-romi is adequately aged, the soluble soy saucemust be separated from the solid residue.This isdone using hydraulic filter presses that force the

moromi through multi-layered sheets of cloth ornylon filters. Pressing can last for several days,with pressures increasing incrementally, up to100 kg per cm2.About 90 liters or more of liquidmaterial can be obtained from 100 liters of mo-romi.The solid filter cake byproduct ordinarilyis used as animal feed. Greater filtration effi-ciency and overall shoyu yield can be achievedby enhancing protein hydrolysis and pectinaseactivities during mashing.

Refining is an important step,since the liquidmaterial that is collected after pressing still con-tains components that need to be removed. Inparticular, an oil layer that appears in the upperphase, and a sediment material that collects inthe bottom phase, are both undesirable and areseparated. The middle phase, containing theshoyu, can be re-filtered and standardized to adesired composition (based on salt and nitrogenconcentrations and color and flavor).

Pasteurization and packaging

The refined, raw soy sauce is usually heat pas-teurized, either in bulk or, more commonly, ina plate-type heat exchange system (at 70°C to80°C). Flash pasteurization at 115°C or higherfor three to five seconds may also be used.

Table 12.4. Concentration of important flavor compounds in shoyu1

Compound Concentration (ppm) Compound Concentration (ppm)

Alcohols AldehydesEthanol 31,501 Isovaleraldehyde 233Methanol 62 Isobutyraldehyde 152-phenolethanol 4 acetaldehyde 5n-propanol 4 FuranonesIsobutyl alcohol 12 HMF2 256Isoamyl alcohol 10 HEMF2 232Furfuryl alcohol 12

Acids PolyolsLactic acid 14,347 Glycerol 10,209Acetic acid 2,107 Acetoin 10Esters 2,3-butanediol 238

Ethyl lactate 24 PhenolsEthyl acetate 15 4-EG2 3Methyl acetate 14 4-EP2 0.3

1Adapted from Nunomora and Sasaki, 19922See text for chemical name



Aside from inactivating enzymes and micro-organisms and thereby increasing shelf-lifeand stability, pasteurization contributes otherimportant properties (Table 12–5). In particu-lar, heating promotes color and flavor devel-opment by accelerating the non-enzymaticbrowning reaction. In addition, heat inducesprecipitation of insoluble proteins and othercomponents that can then be removed by fil-tration to give greater product clarity. Finally,the heat step concentrates acids, phenols,and other compounds that either contributedesirable flavors or that have antimicrobialactivity. Heating, however, may cause loss ofvolatile components.

Most soy sauces are packaged in glass bot-tles varying in size from 0.5 liters to 2 liters.Plastic bottles and 20 liter pails (for institu-tional use) are also used. Various chemicalpreservatives, including benzoic acid,benzoatesalts, and ethanol, are often added.

Product characteristics

The final chemical composition of soy saucesdepends on the specific type being produced(Table 12–2). In general, the greater the pro-portion of soy beans, relative to wheat, thegreater will be the total nitrogen concentra-tion. Conversely, products made using higherlevels of wheat will contain less nitrogenousmaterial, but more reducing sugars. Thus,shoyu tamari, which is produced mainly fromsoy beans, contains nearly four times more to-tal nitrogen, but six times fewer reducing sug-

ars, than shiro, a type of shoyu made fromwheat with very few soy beans.The amount ofnitrogenous material and reducing sugars, asreflected by the wheat-to-soy bean ratio, has amajor influence on color. Soy sauce productscontaining greater levels of wheat are gener-ally more light-colored and products contain-ing more soy beans are dark-colored. Shiro,therefore, has a more tan appearance whereastamari has a dark red-brown color.

The final composition of soy sauce dependson the specific type and the manufacturer’sspecifications. In general, soy sauce contains(on a weight/volume basis) about 1.5% total ni-trogen, 1% sugars, 1% lactic acid, 2% to 2.5%ethanol, and 14% to 18% salt.The pH is usuallybetween 4.5 and 4.8. Due to the association ofhigh salt diets with hypertension and other hu-man health problems, there has been a consid-erable effort to reduce the salt concentrationin soy sauce products. Among the strategiesthat have been considered are the use of saltsubstitutes, salt removal systems, and manufac-turing modifications (Box 12–1).

The Flavor of Soy Sauce

Considering that most soy sauces contain up to18% NaCl, the most immediate and obvious fla-vor one detects is saltiness. However, the flavorof soy sauce is far more complex than simplysaltiness. In fact, the shoyu products listed inTable 12–2 all contain between 16% and 19%salt, yet their flavor profiles, and the concentra-tions of volatile flavor compounds,can vary con-siderably.Nearly 200 volatile flavor componentshave been identified in shoyu, using GC orGC/MS analysis. Several of these, in particular,various furanones and phenolic compounds,areconsidered to be the most important, based ontheir relatively high concentrations and theircharacteristic soy sauce-like flavor (Table 12–4).

Furanones include 4-hydroxy-2-(or 5) ethyl-5 (or 2) methyl-3-(2H) furanone (HEMF) and4-hydroxy-5-methyl-3-(2H) furanone (HMF),which confer sweet or roasted flavor notes,respectively. HEMF has been considered to bethe main “character-impact” flavor of most

Table 12.5. Beneficial effects of pasteurization on soysauce.

• Enhances color formation via the Maillard non-enzy-matic browning reaction

• Inactivates enzymes• Kills microorganisms• Accelerates precipitation of protein complexes• Concentrates anti-fungal agents (phenols and or-

ganic acids)• Concentrates flavor compounds (phenolic com-

pounds, aldehydes, pyrazines, ketones, and other organic compounds)



Box 12–1. Taking Salt out of Soy Sauce and Other Soy-fermented Foods

Salt is an essential ingredient in many of the Asian-type fermented foods. For example, shoyuand other soy sauces contain between 14% and 18% salt. Likewise, miso contains as much as13% to 14% salt. Even when used in moderation, these high-salt foods may still account for a relatively high intake of salt.A single 1 tablespoon serving of soy sauce (about 17 g) containsabout 950 mg of sodium or 2.4 grams of salt, and a cup of miso soup contains nearly 2 g of salt.

Not surprisingly, those countries that regularly consume these products have high salt con-sumption rates. In Japan, for example, per capita salt consumption has averaged (from 1975 to1997) around 13 g per person per day,30% more than the 10 g per day that is recommended asa maximum. In contrast, salt consumption in the United States is somewhat less, about 10 g perperson per day, but still nearly twice that recommended by health authorities.

The problem with high-salt diets is that there is a strong association between sodium or saltconsumption and various disease states, including hypertension, stroke, and gastric cancer.Thishas led public health agencies to recommend that salt intake be reduced, especially for at-riskindividuals. Since soy sauce and related products account for as much as 20% of total salt intakein Japan and other Asian countries, decreasing their salt content could contribute to a signifi-cant overall reduction in salt consumption.

There are a number of possible ways to reduce the sodium content in soy sauce types ofproducts.The most common means is to manufacture the product in the usual way (i.e., addingthe same amount of salt to the mash as would be done for a conventional fermentation), fol-lowed by removal of a portion of the salt via one of several separation technologies.The latterinclude ion exchange chromatography using de-salting resins, ion-exchange membrane pro-cessing, and electrolysis. Potassium chloride can then be added to satisfy many of the functionsplayed by sodium chloride (e.g., preservation and flavor).

A second way to make low-salt soy products is to replace some of the sodium chloride withpotassium chloride prior to the fermentation. The potassium salt will still influence the microflora, as does sodium chloride, by selecting for salt-tolerant lactic acid bacteria and yeast.The downside to both of these approaches, however, is that potassium imparts an undesirablebitter or metallic flavor that consumers generally dislike.Another approach is simply to reducethe salt content and add another antimicrobial agent, such as ethanol, to control the microflora(Chiou et al., 1999).

A quite different route to reduce the salt content has been to identify other compoundsthat contribute “saltiness” (without affecting the overall soy sauce flavor) and then to usethose substances as salt mimics. Researchers in Japan recently discovered that several aminoacid derivatives and peptides, such as ornithyltaurine, glycine ethyl ester hydrochloride, andlysine hydrochloride, were effective substitutes for potassium chloride (Kuramitsu, 2004).

In addition to affecting flavor, reducing the salt content of soy sauce and other fermented soyproducts may also have an effect on the microbiological stability of these products. However,pathogen challenge studies on low-salt miso revealed that even when salt levels were reducedto less than 3%, Staphylococcus aureus, Clostridium botulinum, Salmonella typhimurium,and Yersinia enterocolitica were unable to grow or survive, and toxin production was absent(Tanaka et al., 1985). Still, as an additional barrier, some manufacturers add ethanol to the prod-uct to control spoilage or pathogenic organisms.

References

Chiou, R.Y.-Y., S. Ferng, and L.R. Beuchat. 1999. Fermentation of low-salt miso as affected by supplementa-tion with ethanol. Int. J. Food Microbiol. 48:11–20.

Kuramitsu, R., 2004. Quality assessment of a low-salt soy sauce made of a salty peptide or its related com-pounds.Adv. Exp. Med. Biol. 542:227–238.

Tanaka, N., S.K. Kovats, J.A. Guggisberg, L.M. Meske, and M.P. Doyle. 1985. Evaluation of the bacteriologicalsafety of low-salt miso. J. Food Prot. 48:435–437.



Japanese-type soy sauce (Figure 12–3). How-ever, although it has been suggested that fura-nones are derived from fermentation, theirpresence and concentration in shoyu de-pends mostly on the composition of the rawmaterial. For example, HEMF occurs inkoikuchi, but not in tamari, indicating thatwheat is necessary for its synthesis. Amongthe phenolic compounds that contribute tosoy sauce or shoyu flavor are 4-ethylguaiacol(4-EG) and 4-ethylphenol (4-EP).They are alsoproduced during fermentation by yeasts, butare derived from precursors formed by kojimolds grown on wheat. They confer typicalsoy sauce-like flavors.

In addition to the furanone and phenols, alarge number of acids, aldehydes, alcohols, andesters are produced during the shoyu fermen-tation.Although the concentrations of some ofthese compounds, i.e., ethanol and lactic andacetic acids, and 2,3-butanediol, can reach veryhigh levels, most are in the order of parts permillion.They are generally produced by lacticacid bacteria and yeasts. Other important fla-vors are generated, along with color, via theMaillard reaction. Although the formation ofMaillard reaction products is accelerated byheating (i.e., pasteurization conditions), manyend products are also formed during the mash-ing and moromi aging steps.

Another group of flavor constituents ex-ists that is very important in soy sauce prod-ucts, but whose flavor is not so easily de-

scribed. Included in this group are several ni-trogenous compounds, especially the aminoacid glutamic acid, and the 5� nucleotides in-osine 5�-monophosphate (IMP) and guano-sine 5�-monophosphate (GMP).The flavor ortaste of these compounds (and their sodiumsalts) fall outside what are ordinarily consid-ered to be the four basic flavors—sweet,sour, bitter, and salty. Rather, sodium gluta-mate, sodium inosinate, and sodium guany-late impart a totally unique flavor known asumami, a Japanese term meaning “delicious-ness.” Umami has been described as confer-ring a meaty, savory, brothy flavor to foods.The glutamate salt monosodium glutamate isalso considered to be a flavor enhancer. Andalthough umami is usually associated withAsian cuisines, it is present in a wide array offoods, including cured ham, Parmesancheese, and shitake mushrooms.

Non-fermented soy sauce

Given the many steps involved in preparingthe raw materials, making the koji, and per-forming the fermentation, it is not surprisingthat an alternative, non-biological process formaking soy sauce exists.This process is basedon an acid hydrolysis of soy beans, and al-though the finished product is less expensiveto make and takes much less time, it lacks theflavor and complexity of fermented or brewedsoy sauce.

OH3C

O

C2H5

HO

4-hydroxy-2-ethyl-5-methyl-3-(2H) furanone

OH3C

OH

C2H5

O

4-hydroxy-5-ethyl-2-methyl-3-(2H) furanone

Figure 12–3. Chemical structure of HEMF. Adapted from Nunomura and Sasaki, 1992.




Although the low pH and high salt concentra-tion inhibits most spoilage organisms, ben-zoate is sometimes added as a preservative,mainly against fungi. Ethanol may also beadded for preservation.As noted above, exces-sive browning during mashing or aging mayalso be considered a defect, especially whenthe soy sauce is destined for use as a flavor in-

gredient in light-colored products (Box 12–2).In addition, some undesirable flavors, such asisobutyric acid and isovaleric acid, may formduring prolonged storage.

Miso

Miso is another popular fermented soy prod-uct in the Orient.Although miso originated inChina and Korea more than a thousand years

Box 12–2. Reducing the Dark Color in Soy Sauce

The dark brown color of shoyu and other types of soy sauce is ordinarily an expected property.A brown color is certainly a major part of the overall appearance of many of these products.However, for many applications—in particular, when soy sauce is used as an ingredient inprocessed foods—a lighter, amber-like color is more desirable.Although it is certainly possibleto produce such products by adding more wheat to the formulation (e.g., shiro has a lightercolor than tamari), there is much interest in developing manufacturing processes that producetraditional soy sauce products with a lighter, less brown appearance.

Strategies that have been devised to “lighten” the color of soy sauce are based mainly on con-trolling the Maillard reaction responsible for brown pigment formation, which occurs primarilyduring the mashing and pasteurization steps.The two reactants in the Maillard, non-enzymaticbrowning reaction are reducing sugars and primary amino acids. Due to the extensive hydrolysisof soy protein and wheat starches and polysaccharides, both are present at very high concentra-tions in the moromi mash.For example,the amino acid and reducing sugar concentrations may beas high as 1.5% and 7%, respectively (although average levels are usually somewhat less).Amongthe total free sugars present in the mash are various pentoses, including arabinose (30 mM) andxylose (20 mM).Both of these sugars are very reactive in the Maillard reaction.In fact, the order inwhich these sugars “brown”is xylose � arabinose � hexoses.Thus,reducing the concentration ofxylose in the mash could have a significant impact on brown pigment formation.

One way to decrease xylose levels would be to block or inhibit the xylan hydrolysis reactionsthat release free xylose from the xylan polysaccharides that are present in the wheat and soy-bean cell walls. Xylan hydrolysis during mashing occurs by the action of one or more of severalxylanases and �-xylosidases produced by the koji mold Aspergillus oryzae.Xylanases hydrolyzexylan to xylobiose and xylan-containing tri- and tetra-saccharides,which are then hydrolyzed by�-xylosidases to give free xylose.

Researchers in Japan identified and cloned the xynF1 gene encoding for the major xylanasein an industrial strain of A.oryzae (Kitamoto et al.,1998).Then,by re-introducing (in trans) mul-tiple copies of the xynF1 promoter region back into the parent strain, transcription of xynF1was reduced simply by a titration effect.As expected, expression of the xylanase was also de-creased.The same group also identified the xylA gene that encodes for the major �-xylosidaseproduced by this organism (Kitamoto et al., 1999). An antisense mRNA strategy was subse-quently used to reduce expression of the enzyme by 80%. In theory, then, if this modified A.oryzae strain were used for koji-making, xylose formation and browning reactions would besignificantly reduced.

A second, altogether different, approach was adopted by another group of Japanese re-searchers (Abe and Uchida, 1989). They observed that, despite the metabolic capacity ofTetragenococcus halophilus (formerly Pediococcus halophilus) to metabolize free xylose,nearly

(Continued)



all of the xylose present in the moromi mash remained unfermented.As noted above, the mashcontains plenty of glucose, which is the preferred energy and carbon source for this organism.During the soy sauce fermentation, however, use of xylose and other sugars is catabolite re-pressed by glucose.That is, the xyl genes coding for transport (xylE), isomerization (xylA), andphosphorylation of xylose (xylB) are present (Figure 1),but are not induced as long as glucose isavailable.These researchers reasoned,however, that T. halophilus would ferment xylose, even inthe presence of glucose, if catabolite repression could be lifted or de-repressed. Spontaneous

Box 12–2. Reducing the Dark Color in Soy Sauce (Continued)

A xylose

xylose xyulose xyulose-Pxyulose kinase

ATP ADP

xylose isomerase

xylosetransporter

PK Pathway

out

in

B

xylBxylE xylA xylR

kinasetransport isomerase

CRE

FBP

CcpA-HPr(Ser)-P

P

PP

PP

HPrHPr Kinase

ATP ADP

+ CcpAHPr(Ser)-PP

FBP

(+)

glucose

FBP FBP

(-)

(-)

+ xylose- glucose

(+)

xylose

repressor+ glucose-6-P

Figure 1. Xylose metabolism (A) and catabolite repression (B) of the xylose operon in Tetragenococcushalophilus. The xylose metabolic pathway consists of a transporter, an isomerase, and a kinase, encoded byxylE, xylA and xylB, respectively. The product of this pathway, xyulose phosphate, then feeds directly into thepentose phosphoketolase (PK) pathway. Expression of the xyl operon is mediated by positive (�) and nega-tive (�) regulation. During active growth on glucose, the glycolytic metabolite fructose-1,6-bisphosphate(FBP) accumulates inside the cell. FBP then activates the enzyme HPr kinase, leading to the phosphorylationof the phosphotransferase protein HPr at a specific serine residue. HPr(Ser)-P then forms a complex with an-other protein called Catabolite-control protein A (CcpA). This complex, along with FDP, binds to a DNA se-quence called catabolite repression element (CRE) located upstream of the promoter region of the xyloseoperon, preventing its transcription (indicated by the negative sign). In the absence of glucose (and itsmetabolites), CcpA repression is lifted. In addition, a xylose repressor that is expressed during growth on glu-cose binds to a putative operator region positioned near the xyl promoter, further repressing transcription.However, when xylose is present and glucose is absent, xylose can bind directly to the repressor, causing itsrelease from the operator and allowing the operon to be transcribed (�). Adapted from Abe and Higuchi,1998 and Takeda et al., 1998.



ago, Japan is the leading producer and con-sumer, with per capita consumption at about 5Kg per person per year or 14 g/day. Miso has aflavor similar to soy sauce, except instead ofbeing a liquid, it is paste-like,with a texture likethat of a thick peanut butter. It can be used likesoy sauce as a seasoning or flavoring agent, butis more commonly used, at least in Japan, to

make soups and broths. Similar products areproduced in Korea (doenjang),China (jang), In-donesia (taoco), and the Philippines (taosi).

In Japan, there are three general types ofmiso Each is based mainly on composition,specifically the raw ingredients used to makethe miso koji (Table 12–6). Rice miso is madeusing a rice koji, barley miso is made using a

catabolite repression mutants were subsequently isolated that fermented xylose concurrentlywith glucose. Biochemical analyses revealed that the mutant strain was defective in glucosetransport. Specifically, the mannose phosphotransferase system (PTS),which is primarily respon-sible for glucose transport in this strain, was found to be non-functional.Thus, if a product orcomponent of the mannose PTS is necessary for catabolite repression (as is the case for otherGram positive bacteria), then a non-functional PTS would result in the xylose-fermenting pheno-type these researchers observed.

Indeed, detailed analysis of the xylose operon revealed that catabolite repression of the xy-lose pathway was mediated, in part, by the phosphorylated form of the PTS protein HPr.Thus,repression required an intact PTS, which would explain why a PTS mutant would be de-re-pressed for xylose use. Furthermore, these researchers also identified an xylR gene, encoding axylose repressor, whose expression (and effectiveness as a repressor) was activated by glucose-6-phosphate (Figure 1).

References

Abe, K., and K. Uchida. 1989. Correlation between depression of catabolite control of xylose metabolismand a defect in the phosphoenolpyruvate:mannose phosphotransferase system in Pediococcushalophilus. J. Bacteriol. 171:1793–1800.

Kitamoto,N., S.Yoshino,M. Ito,T.Kimura,K.Ohmiya, and N.Tsukagoshi.1998.Repression of the expressionof genes encoding xylanolytic enzymes in Aspergillus oryzae by introduction of multiple copies of thexynF1 promoter.Appl. Microbiol. Biotechnol. 50:558–563.

Kitamoto, N., S.Yoshino, K. Ohmiya, and N.Tsukagoshi. 1999. Sequence analysis, overexpression, and anti-sense inhibition of a �-xylosidase gene, xylA, from Aspergillus oryzae KBN616.Appl. Environ. Micro-biol. 65:20–24.

Takeda,Y., K.Takase, I.Yamoto, and K.Abe. 1998. Sequencing and characterization of the xyl operon of aGram-positive bacterium, Tetragenococcus halophila.Appl. Environ. Microbiol. 64:2513–2519.

Box 12–2. Reducing the Dark Color in Soy Sauce (Continued)

Table 12.6. Types and properties of different types of miso1.

Product Moisture (%) Protein (%) Reducing sugar (%) Fat (%) pH Color

Rice misoSweet 43–49 10–11 15 3–5 5.4 White or redSalty 45 10–13 12 6 5.3 Yellow or redBarley miso

Sweet 44–47 10 17 4–5 5.2 YellowSalty 46–48 13 11 5–6 5.1 Red

Soy bean miso 45–46 17–20 4 11 5.0 Red or brown

1Adapted from Ebine, 1986 and Baens-Arcega et al., 1996



barley koji, and soybean miso is made using asoybean-only koji. There are, however, varia-tions of these three miso types, depending onthe salt content,color,and flavor.Because colordevelopment occurs mostly as a function ofhow much rice (or barley) is added and howlong the product is aged, lighter colored misostypically contain more rice (or barley) and areaged for only a few months.They also have amore mild flavor, due to shorter fermentationtimes. For example, white or shiro miso ismade using a rice koji, contains soy beans andsalt, and is usually light in appearance. Thistype of miso has a mild, sweet flavor. Red misoor “akamiso” is made from a barley koji and hasa stronger, more robust flavor. Finally, hatchomiso contains only soybeans, is aged for as longas three years, and has a very dark brown colorand a strong, complex flavor.

Manufacture of miso

Miso and related products are manufacturedmuch like soy sauce, except for one major dif-ference. In miso manufacture, dry salt, ratherthan a brine, is added directly to the koji-soybean mixture.Therefore, the resulting producthas approximately twice the total solids of soysauce (50% to 60% versus 24% to 28%). Theprocess starts with the manufacture of a koji(Figure 12–4). As noted above, the koji sub-strate can be rice, barley, or soy beans.The rice(usually polished rice is used) or barley issoaked in water overnight at 15°C, thensteamed in a batch or continuous cooker.Aftercooling, a tane koji (or a spore culture), con-taining strains of A. oryzae and A. sojae withdefined properties, is used as the inoculum (at0.1%). The koji is then incubated at 30°C to40°C for forty to forty-eight hours in fermenta-tion chambers equipped with aerating devices.

The soy beans (usually yellow soy beans areused) are similarly prepared,first by sorting andsoaking and then by cooking under pressure(0.5 kg/cm2 to 1.0 kg/cm2) for about fifteen toforty-five minutes. At this point in the process,the cooked soy beans are ground or extruded,then mixed with the koji and salt in automated

mixing machines. Alternatively, the wholecooked soybeans, koji, and salt can be mixed,and then the entire mixture is mashed. Thismashing step is performed using an extrusion-like device that grinds the mixture to a chunkyhomogenous paste. The mixture is transferredto large tanks or fermentors (stainless steel orepoxy resin-lined steel) with capacities of 1,000kg to more than 100,000 kg. Depending on thetype of miso being made, the salt concentrationmay range from 6% to 13%.

Fermentation

The miso fermentation occurs in a mannersimilar to that of soy sauce, in that the hydroly-sis of complex macromolecules to form simplenutrients and the subsequent fermentation andmetabolism of those nutrients occurs essen-tially at the same time. As it does with soysauce, the koji serves both as the source of pro-teolytic and amylolytic enzymes, as well as asubstrate source for those enzymes.About 50%of the total protein and 75% of the polysaccha-rides are completely hydrolyzed to amino acidsand monosaccharides, respectively. Becausemiso is made from whole soy beans, the lipidportion (about 3% to 10%) is also present inthe miso mash. Fungal lipases (from the koji)hydrolyze triglycerides to di- and monoglyc-erides, glycerol, and free fatty acids.The lattermay accumulate to high levels (as much as 1%to 3% in the miso). Collectively, the aminoacids, sugars, and fatty acids provide a richsource of nutrients for fermentative organisms.

In traditional miso manufacture, a miso seedculture, obtained from previous raw misoproduct, is used to initiate the fermentation.Al-though many small manufacturers continuethis practice, miso starter cultures are now of-ten used, especially by large modern manufac-turers. These cultures ordinarily contain yeaststrains of Z. rouxii and C.versatilis, along withT. halophilus, P. acidilactici, L. delbrueckii, andother lactic acid bacteria. Fermentation of sug-ars present in miso results in the use of glucoseand other free sugars and the formation of sev-eral organic acids, including lactic, acetic, and



succinic acids. The longer the miso ferments,the lower the sugar concentration. However,the total acid concentrations in the finishedmiso generally are less than 1% and the pH isusually not less than 5.0. About 0.1% to 1%ethanol and 1% to 2% glycerol may be formedduring the yeast fermentation.

Miso fermentations are conducted in tanksat 25°C to 30°C.Temperature control and me-chanical stirring are common features in manyfactories. Fermentation can be as short as onemonth or less to more than two years.Transfer-ring the miso from one tank to another is oftendone to promote a homogenous fermentation.When the fermentation and aging period is

Whole soy beans

Inoculate with 0.1% tane kojior Aspergillus oryzae

Incubate aerobically at 30 - 38°Cfor 3 days to make koji

Pasteurize

Soak 16 hours

Steam/autoclave for 0.5 hour

Cool to 30°C

Rice (or barley)

Steam and cool

Mix soy beans with koji and salt

Mash

Package

Soak (overnight)

Ferment (28 - 30°C)

Package

Age

Lactic/yeast starter cultureMash

Figure 12–4. Miso flow chart.



complete, the miso is re-mashed and packagedinto small plastic tubs or bags (about 400 g to500 g capacity). The miso can then either beleft in its raw state or subjected to a short-timeheat treatment. While pasteurization does en-hance shelf-life by preventing package swelling(due to gas formation), this is strictly an op-tional practice.To many consumers, raw miso ismuch preferred, based on the putative healthproperties conferred by the live microorgan-isms. In this respect, miso is viewed much likeyogurt, in that any heat step that inactivates mi-croorganisms is tantamount to blasphemy.


Despite the high salt concentration and rela-tively low pH, growth of spoilage organisms inmiso can occur, resulting in gas, off-odors,over-acidification,and surface slime.Yeasts andbacteria responsible for these defects includeHansenula and Z. rouxii, Pediococcus acidi-lactici, and Bacillus sp. Spoilage is more likelyto occur when salt levels are reduced (�12%)or when the koji molds are inhibited. Pasteur-ization, either before or after packaging, inacti-vates these organisms. However, as indicatedabove, many consumers prefer raw, nonheat-treated miso.

Natto

Natto is another soybean-fermented productconsumed mainly in Japan, but similar productsare also produced in China, Thailand, and thePhilippines. Per capita consumption in Japan isabout 1.2 Kg per person per year or 3 g per day.Natto is used as a condiment or flavoring agent,usually for rice and vegetables or as an ingredi-ent in sushi.Nutritionally,natto is comparable toother fermented soybean products. It contains16% to 18% protein (45% on a dry basis), withgood digestibility and biological value, com-pared to cooked soybeans.It is considered to bean excellent source of isoflavones and readilyadsorbed Vitamin K.The latter is thought to con-tribute to bone-strength and reduced rates offractures among natto-consuming populations.

Manufacture of natto

The manufacturing procedure for natto beginslike that for miso; however, the organisms in-volved in the fermentation are different andthe final product bears little resemblance tomiso. Natto is made from whole, somewhatsmall-sized soybeans that are cleaned, soakedfor twelve to twenty hours at ambient temper-ature, and steamed at 121°C for twenty to fortyminutes. The thoroughly cooked and cooledbeans are then inoculated with about 106 to107 spores per Kg of Bacillus subtilis var.natto (formerly Bacillus natto), and the mater-ial is well mixed.The incubated beans are thendivided into 100 g portions and placed intopackages. According to traditional practices,bundled rice straw was used as the container(some manufactures still use straw); however,polyethylene bags are now common. In eithercase, the beans are moved into aerobic incuba-tors at 40°C at 85% relative humidity for six-teen to twenty hours.Growth of B. subtilis var.natto occurs primarily at the surface and is ac-companied by a change in color (from yellowto white) and synthesis of a highly viscouspolysaccharide material. The latter can com-pletely cover the entire bean surface, account-ing for nearly 1% of the total dry weight ofnatto.After incubation, the natto is held at 2°Cto 4°C to minimize further growth.

The final product is quite different frommiso and shoyu. Its main flavor attribute issweetness, due to the presence of the polysac-charide material, which is composed of a fruc-tose-containing polymer. The polysaccharidealso confers a definite sticky and stringy tex-ture to the product, which, although character-istic of natto, nonetheless causes some con-sumers to consider other menu options.

Tempeh

Tempeh is a mold-fermented soy bean productthat originated many centuries ago in Indone-sia,where it remains a major food staple and aninexpensive source of dietary protein. Unlikeother fermented soy products, tempeh produc-



tion has spread to only a few other countries,including Malaysia, the Netherlands (Indonesiawas once under Dutch rule), Canada, and theUnited States. However, Indonesia is by far themain producer and consumer of tempeh. Cur-rent per capita consumption in Indonesia isabout 15 grams per person per day. Althoughtempeh production has been industrialized, asubstantial amount of the more than 500 mil-lion kg of tempeh produced in Indonesia peryear is still made in the home or in small “cot-tage-sized”production facilities.

It is remarkable, given the fact that tempehwas not discovered by American consumersuntil the last twenty to thirty years, that it hasbecome quite popular (relatively speaking) inthe United States. Undoubtedly, this suddenpopularity is due, in part, to interest in vegetar-ian cuisine and non-meat alternative food prod-ucts (“faux”meats). However, the popularity oftempeh in the United States may also be due toits nutritional properties. In particular, tempehis a rich source of protein (19%) and, in addi-tion, is one of only a few plant-based foods thatcontains vitamin B12 (discussed below). Sincethis vitamin is often lacking in vegetarian diets,tempeh serves as a modest source of this es-sential nutrient (a 100 g serving provide 6% ofthe 2.4 �g of B12 recommended for a healthyadult, according to USDA estimates).

Another reason, perhaps, why tempeh hasattracted interest in the United States relatesto its versatile applications and organolepticproperties. In its raw state, tempeh has abland,mushroom-like flavor.However,cookingtransforms this plain-tasting material into apleasant,nutty,flavorful product. If one can getpast the fact that tempeh consists entirely ofmoldy beans, its flavor, especially when it issauteed or fried, resembles that of cooked orroasted meat (sort of). After all, tempeh andmuscle protein (i.e., meat) both derive muchof their flavor from the Maillard reaction prod-ucts that result when amino acids and reduc-ing sugars are heated at high temperature.Thelipid component of soy beans may also serveas a precursor for meaty flavor and aroma de-velopment. Finally, the development of a tem-

peh “industry” in the United States and con-sumer interest in tempeh as a food have likelybeen driven by academic researchers, particu-larly Keith Steinkraus at Cornell Universityand C. W. Hesseltine and H.L. Wang at theUSDA, who played key roles in understandingmany of the microbiological and technical is-sues related to tempeh manufacture.

Manufacture of tempeh

The industrial manufacture of tempeh is quitesimple, although numerous variations exist, de-pending largely on the scale of production, geo-graphical and climatic considerations, and man-ufacturer preferences.The only raw material issoy beans, the fermentation time is short, andthere is no aging or ripening period involved. Infact, the entire start-to-finish process is less thanforty-eight hours.Tempeh can be considered assolid-state fermentation in that it consists of soybeans that are essentially held together by themold mycelium that grows throughout and inbetween the individual beans. It should benoted that although other legumes, cerealgrains, vegetables, and even seafood can be in-corporated into tempeh (and are commerciallyavailable), soy beans are by far the most com-mon starting material.

Substrate preparation

The process starts by sorting the soy beans toremove damaged or moldy beans (Figure 12–5).The beans are usually given a quick heat treat-ment (about 70°C to 100°C for ten to thirtyminutes), and the hulls are removed either man-ually or mechanically.The former method, prac-ticed by very small traditional manufacturers, in-volves rubbing the beans with hands (or evenfeet!) and then separating the hulls by floata-tion. Manual de-hulling, however, has beenlargely replaced by the use of various types ofmills, which can be used on dry or wet beans.Some manufacturers omit the first boiling step,although this is considered by some tempeh experts to be an essential part of the process,facilitating hydration and hull removal.



The de-hulled beans are then soaked in wa-ter, a seemingly simple step,but one which hasvery important implications for tempeh qual-ity. It is during this steeping period (sixteen totwenty-four hours at warm ambient tempera-

ture), that endogenous lactic acid bacteriagrow and produce acids that lower the pH.Thelow pH generally restricts growth of undesir-able spoilage bacteria, as well as potentialpathogens. It should be noted, however, that

Soybeans

Clean, sort, and dry

Wash and soak for 16- 24 hours at 22 - 26°C

Inoculate with spores of Rhizopus oligosporus

Boil for 30 - 60 minutes

Drain and cool

Pack into forms or trays

Incubate at 25 - 30°C, 70 - 85% RH for 24 - 48 hours

Store at 4°C or freeze

Wash and boil for 30 minutes

Drain, de-hull, and rinse

Figure 12–5. Tempeh flow chart.Adapted from Shurtleff and Aoyagi,1979.



some of these undesirable bacteria can surviveduring the soaking step even if acid conditionsare established. To promote acidification, thesteep water can be acidified directly with lac-tic or acetic acid. Finally, potential mold in-hibitors may also be extracted from the soybeans during the steeping period.

After the soaking step, the remaining hullsare removed from the beans and the dehulledbeans or cotyledons are heated a second time,either by steaming or boiling (for anywherefrom thirty minutes to two hours). This heattreatment enhances extraction of soluble nutri-ents and inactivates microorganisms that mightotherwise interfere with the subsequent fer-mentation.This step is also necessary to dena-ture trypsin inhibitor, a native soy protein thatacts as an anti-nutritional factor.

Inoculation

After cooking, the beans are drained and dried.Next, they are inoculated with either a portionfrom a previous batch of fully developed tem-peh, a wild, mixed strain culture called “usar,”or, as is more common in modern manufac-ture, with a spore culture of Rhizopus micro-sporus var. oligosporus (hereafter referred toas R. oligosporus). Typical culture inocula levels range from 107 to 108 spores (about 1 g)per kg of beans. Pure R. oligosporus culturescan be added directly to the beans, or thespores can be inoculated onto steamed rice,incubated until well-grown, and then used toinoculate the soy beans.

Fermentation

According to traditional Indonesian manufac-turing practices, the inoculated beans are thenshaped into cakes and wrapped in bananaleaves.The use of banana leaves as the packingchamber is not merely for natural aesthetics;they provide a moist, microaerophilic environ-ment that supports rapid growth of R. oligo-sporus. However, for large-scale manufactureof tempeh (and as conducted in the UnitedStates), the inoculated beans are distributed ontrays 1 cm to 3 cm deep and ranging in length

and width from 1 meter to several meters.Afterone to two days of incubation in a warm room(ranging from 25°C to 37°C), the beans arecovered with white mycelium,and the fermen-tation is considered complete.The mycelia alsowill have grown in between the individualbeans such that solid soy bean cakes will haveformed. Detailed analyses have revealed thatthe fungal hyphae actually penetrate nearly 1 mm into the bean, or about 25% of the diam-eter of the cotyledon. It is important that thefermentation ends, however, before the moldbegins to sporulate, since the appearance ofthe dark-colored black or grey sporangia is gen-erally unattractive to consumers (see below).

Tempeh Microbiology

The surface of raw soy beans contains an as-sortment of Gram positive and Gram negativebacteria, including Lactobacillus casei andother lactic acid bacteria; enterococci, staphy-lococci; streptococci; bacilli; Enterobacter,Klebsiella, and other coliforms.Yeasts, such asPichia,Saccharomyces, and Candida,may alsobe present. During the soaking step, sucrose,stachyose, and raffinose diffuse out of thebeans and into the water.Their subsequent hy-drolysis by invertases and glucosidases releasesglucose and fructose, which can then be usedto support growth of the resident microflora.Consequently, at the end of the soaking period(twenty to twenty-four hours at 20°C), the totalmicroflora may reach levels of 109 cfu per mlor higher. Most of the organisms isolated aftersoaking are lactobacilli, enterococci, and strep-tococci. However, the specific species that pre-dominate appears to depend on the tempera-ture and the pH of the soak water (i.e., in thoseapplications where lactic or acetic acid isadded to the water). Importantly, the pH valuesof the soak water, whether acidified or not,generally will decrease to 4.5 to 5.0 by virtueof lactic and mixed acid fermentations.

Although the manufacture of tempeh clearlydepends on the growth of R. oligosporus (dis-cussed below), the fermentation that occursduring the soaking of the soy beans is also es-sential.This is because the formation of organic



acids and the decrease in pH are necessary tocontrol pathogens, including Salmonella ty-phimurium, Yersinia enterocolitica, Staphylo-coccus aureus, and Clostridium botulinum,that might otherwise grow in non-acidified soybeans. Low pH also inhibits Bacillus, Enter-obacter, and other microorganisms capable ofcausing spoilage effects. It is important to notethat even if the beans are heated prior to soak-ing, the lactic fermentation will still occur, al-though the rate and extent of the fermentationmay be affected.

Tempeh cultures

As noted above, the primary fermentation ismediated by growth of R. oligosporus, whichcan be added to the soy beans in one of severaldifferent forms. First, it can be added as a purespore culture. Recommended strains includeNRRL 2710 and DSM 1964, both isolated fromIndonesian tempeh and both available frompublic culture collections.Like the commercialstrains used for other fungal-fermented prod-ucts, tempeh starter cultures should be se-lected based on specific phenotypic traits(Table 12–7).Alternatively, a backslop materialconsisting of a dried tempeh culture can beused. Finally, a third form, used in traditionaltempeh manufacture, is called usar, and ismade by inoculating wild Rhizopus sporesonto the surface of leaves obtained from the in-digenous Indonesian Hibiscus plant.After twoto three days of incubation, the leaves containa dense spore crop that can be dried and usedto inoculate the soy beans. Species other thanR. oligosporus may be present when wild cultures are used. Other species isolated from

tempeh include Rhizopus oryzae, Rhizopusstolonifer, and Rhizopus microsporus var.chinensis.

Regardless of source or strain, tempeh cul-tures have a limited shelf life,as little as three tofour months. This is because the Rhizopusspores enter into a dormancy phase duringstorage that reduces viability and germability(the ability of dormant spores to become acti-vated and produce biomass). Moreover, even ifthe soy beans are inoculated with a pure sporeculture of R.oligosporus, the tempeh is unlikelyto be maintained in a pure state for too long,since the substrate itself will likely be contami-nated with an array of different organisms. Infact, as discussed below, other microorganismsmay play important nutritional and organ-oleptic roles during the tempeh fermentation.

Tempeh biochemistry

In addition to producing the mycelia mass thatliterally holds the soy beans together, R.oligosporus is also responsible for causing ma-jor biochemical changes in the composition ofthe soy bean substrate (Table 12–8). In particu-lar, lipids and proteins serve as substrates forfungi-excreted lipases and proteinases, respec-tively. During the incubation period, about athird of the lipid and a fourth of the proteinfractions are degraded. Lipid hydrolysis resultsmainly in mono- and diglycerides, free fattyacids, and only a small amount of free glycerol.

Table 12.8. Composition of soybeans and tempeh1.

Soybeans2 Tempeh3Constituent

(g/100 g) (g/100 g)

Moisture 7–9 60–65Protein 30–40 18–20Soluble Nitrogen �1 2–4Carbohydrate 28–32 10–14Fiber 4–6 1–2Fat 18–22 4–12pH 6–7 6–7

1Adapted from Hachmeisterand Fung, 1993 and Winarno and

Reddy, 19862Whole raw soybeans (prior to soaking)3Fresh (wet) weight basis

Table 12.7. Properties of Rhizopus microsporus var.oligosporus.

Unable to metabolize major soy carbohydrates (sucrose, stachyose, or raffinose)

AerobicRapid growth and mycelia production at 30–42°CProteolytic and lipolyticUses free fatty acids, derived from lipids, as an energy

source



Most of the free fatty acids are subsequentlyoxidized by R. oligosporus, resulting in a 10%decrease in the total dry matter in the finishedtempeh. In fact, the relative inability R.oligosporus to metabolize the available soycarbohydrates (mainly stachyose, raffinose, andsucrose) means that fatty acids serve as the pri-mary energy and carbon source.

In contrast, only about 10% of the releasedamino acids and peptides are oxidized by R.oligosporus. Of the remaining portion, about25% are assimilated into biomass, and the restis left in the tempeh.The soluble nitrogen con-centration, therefore, increases four-fold, fromabout 0.5% to 2%. Despite the limited metabo-lism of proteins and amino acids, ammonia ac-cumulates during the tempeh fermentation,such that the pH rises from 5.0 to above 7.0.This increase in pH during the fungal fermen-tation stage underscores the importance ofacidification during bean soaking, since the lat-ter step is responsible for inactivating potentialpathogens that may have been present in thestarting material. Once the low pH barrier nolonger exists, neutrophilic pathogens couldtheoretically grow and cause problems.

Polysaccharide-hydrolyzing enzymes arealso produced by R. oligosporus. These en-zymes attack pectin, cellulose, and other fiberconstituents, releasing various pentoses (xy-lose, arabinose) and hexoses (glucose, galac-tose). However, the activity of these enzymesduring the tempeh fermentation is modest,andonly minor amounts of free sugars are found inthe final product.

Tempeh nutrition and safety

Among the most important changes that occurduring the tempeh fermentation are those thataffect the nutritional quality of tempeh. Asnoted above, the concentration of the majormacronutrients (i.e., protein, fat, and carbohy-drates) decreases as the soybeans are con-verted to tempeh, due to enzymatic hydrolysis.These changes may account, in part, for an im-provement in nutritional quality. For example,it has been suggested that protein hydrolysis

makes tempeh more digestible, compared tosoybeans, although the protein efficiency ratio(used as a measure of protein quality) of tem-peh is no higher than an equivalent amount ofcooked soy beans. There is also a decrease inthe amount of soy oligosaccharides (mainlystachyose and raffinose) during the conversionof soybeans into tempeh.These sugars, whichare quite undesirable due to their ability tocause flatulence, are removed from soybeansnot by fermentation, but rather by diffusionduring the soaking and cooking steps.

Probably the most important nutritional im-provement that occurs during tempeh manu-facture is the increase in vitamin content. Ofparticular interest is vitamin B12, whose con-centration in cooked tempeh generally rangesfrom 0.1 �g to 0.2 �g per 100 g. Since soy-beans contain negligible levels of this vitamin,it is evident that its presence in tempeh mustoccur as a result of biosynthesis by microor-ganisms in the tempeh. What is surprising,however, is that vitamin B12 is made not by R.oligosporus nor by lactic acid bacteria, butrather by bacteria that are essentially chancecontaminants in the tempeh-making process(Box 12–3). For example, Klebsiella pneumo-niae, a Gram negative member of the familyEnterobacteriaceae, is a known B12 producer,and is also present in Indonesian tempeh.Other vitamins, including riboflavin (B2),niacin, pyroxidin (B6), biotin, pantothenic acid,and folic acid, also increase in concentrationby up to five-fold during tempeh production.The concentration of thiamine (B1), however,decreases, and there is no effect on the level offat-soluble vitamins.

In addition to the increase in vitamin con-tent, the processing and conversion of soybeansinto tempeh also results in a decrease in severalanti-nutritional factors ordinarily present in soy-beans. These compounds are anti-nutritional because they either interfere with digestion(trypsin inhibitor), reduce protein quality (tan-nins), reduce mineral adsorption (phytic acid),cause blood to form clumps (hemagglutenins),or cause metabolic disturbances (goitrogens).Some of the water-soluble, low-molecular-



Box 12–3. Vitamin B12 in Tempeh

Tempeh is one of only a few non-meat or non-animal-containing foods that contain vitamin B12.However,despite considerable research interest, the actual means by which synthesis of B12 oc-curs has yet to be resolved. In part, this is because measuring the actual amount of this vitaminand its biological activity in tempeh have not been easy tasks. Reported concentrations rangefrom 0.002 �g/100 g to 8 �g/100 g, a 4,000-fold difference (Nout and Rombouts, 1990), de-pending on both the particular assay method (radio-immunoassay versus bioassay) and the sam-ple extraction method. In general, tempeh is considered to contain 0.1 �g to 0.2 �g of B12/100g.According to the USDA Nutrient Database, the Vitamin B12 content is 0.14 �g per 100 gramsof cooked tempeh.

The main questions regarding the B12 content of tempeh, however, concern the organismsand pathways involved in B12 biosynthesis.The tempeh mold, Rhizopus oligosporus, does notproduce any B12. In fact, the vitamin B12 (or cyanocobalamin) biosynthetic pathway is absent ineukaryotic organisms (Keuth and Bisping, 1993; Rodionov, 2003). In contrast, production of vit-amin B12 by several bacteria, including aerobes (e.g., Pseudomonas), anaerobes (e.g., Propioni-bacterium and Clostridium), and facultative anaerobes (Citrobacter, Lactobacillus) has beenreported (Martens et al., 2002;Taranto et al., 2003). Furthermore, a recent genomic analysis hasrevealed that the biosynthetic machinery for making B12 is widely distributed in bacteria (Rodi-onov, 2003).

In tempeh, the main organisms responsible for vitamin B12 biosynthesis are Klebsiella pneu-moniae and Citrobacter freundii (Keuth and Bisping,1994).Although these organisms are pre-sent in the soy beans and soaking water, growth and vitamin production do not occur untillater, after the overnight soaking step and during the primary fungal fermentation. Interestingly,there are reports that B12 production by C. freundii may be up to three times higher whenthese organisms are grown in the presence of R. oligosporus (Keuth and Bisping, 1993;Wieselet al., 1997).

The reader may be wondering about the presence of K. pneumoniae in tempeh and its sug-gested use as a starter culture.This organism is an opportunistic pathogen and is capable of pro-ducing several toxins, including three enterotoxins. However, not all strains are toxigenic; in-deed it does not appear that strains isolated from tempeh contain enterotoxin genes (Keuthand Bisping, 1994).

References

Keuth, S., and B. Bisping. 1994.Vitamin B12 production by Citrobacter freundii or Klebsiella pneumoniaeduring tempeh fermentation and proof of enterotoxin absence by PCR. Appl. Environ. Microbiol.60:1495–1499.

Keuth, S., and B. Bisping. 1993. Formation of vitamins by pure cultures of tempe moulds and bacteria dur-ing the tempe solid substrate fermentation. J.Appl. Bacteriol. 75:427–434.

Martens, J.-H., H. Barg, M.J.Warren, and D. Jahn. 2002. Microbial production of vitamin B12.Appl. Microbiol.Biotechnol. 58:275–285.

Nout, M.J.R., and F.M. Rombouts. 1990. Recent developments in tempe research. J.Appl. Bacteriol. 69:609–633.

Rodionov, D.A.,A.G.Vitreschak,A.A. Mironov, and M.S. Gelfand. 2003. Comparative genomics of the vitaminB12 metabolism regulation in prokaryotes. J. Biol. Chem. 278:41148–41159.

Wiesel, I., H.J. Rehm, and B. Bisping. 1997. Improvement of tempe fermentations by application of mixedcultures consisting of Rhizopus sp. and bacterial strains.Appl. Microbiol. Biotechnol. 47:218–225.

Taranto, M.P., J.L.Vera, J. Hugenholtz, G.F. De Valdez, and F. Sesma. 2003. Lactobacillus reuteri CRL1098 pro-duces cobalamin. J. Bacteriol. 158:5643–5647.



weight compounds, such as phytic acid andpolyphenolic tannins, are removed during soak-ing and washing. In contrast, proteinaceouscompounds, including trypsin inhibitor andother protease inhibitors, and hemaggluteninsare inactivated by heating steps. It is also possi-ble that some of these compounds are degradedby enzymes produced during fermentation. Forexample, phytic acid can be hydrolyzed by phy-tases produced by R.oligosporus.

Finally, there has long been concern aboutthe microbiological safety and the possiblepresence of mycotoxins in tempeh. As is thecase with other mold-fermented foods (e.g.,soy sauce and miso), the wild or pure culturestrains used for tempeh production do not pro-duce mycotoxins.


In Indonesia, where tempeh is consumed on anear-daily basis, spoilage is not much of an is-sue, provided the product is eaten within a dayor two of manufacture. However, the shelf-lifeof tempeh held at room temperature is veryshort, owing to the continued growth of themold and bacteria. Once R. oligosporus beginsto sporulate and produce colored sporangia,the product’s shelf-life is essentially finished.Even when stored at refrigeration tempera-tures, mold growth is slowed but not stopped.Therefore, some form of preservation is neces-sary. In the United States, tempeh is most oftenvacuum packaged in oxygen impermeable plas-tic to restrict growth of aerobic fungi and bac-teria.Another effective way to preserve tempehis freezing, which halts fungal growth. Finally,tempeh can be dehydrated or cooked prior topackaging or made into various processedproducts, such as vegetarian meat-like foods.

Manufacture of Sake and Rice Wines

As noted in Chapter 10, most wines producedthroughout the world rely on grapes as the start-ing raw material. Grapes used for wine makingordinarily contain an ample amount of glucoseand fructose, which are readily fermented by

the endogenous yeasts or added pure yeast cul-tures. In contrast, when starchy substrates, suchas rice, are used as the raw material, the com-plex polysaccharides (mainly amylose and amy-lopectin) must first be hydrolyzed to producefermentable sugars.When rice wines were firstdeveloped, this hydrolysis step was done bychewing and masticating the rice. As will be discussed below, saccharification is now per-formed by a rice koji,not unlike the koji used inthe soy sauce and miso fermentations.

The most well-known rice wine is sake(rhymes with hockey), a Japanese rice winethat likely originated in China,probably severalmillennia ago. Many other rice-derived winesand alcoholic beverages exist throughout theFar East. Examples include shaosing (Chineserice wine), awamori (a Japanese product dis-tilled from rice), makgeolli (Korean rice wine),and ruou (Vietnamese rice wine). Historically,many of these wines have long been associatedwith Shintoism and Buddhism religious rites.For many years, until relatively recently, ricewines were the most popular alcoholic prod-ucts consumed in Japan,China,and other Asiancountries. Current (year 2000) per capita con-sumption of sake in Japan is 8.2 L, nearly 50%less than was consumed in 1970.Consumptionof rice wine in China is now (2003) only 1 Lper person per year (of course, that is still a lotof sake, considering there are nearly 1.3 billionpeople living in China). The decrease in sakepopularity is due mainly to increased beer con-sumption (with per capita consumption ofabout 20 L in both Japan and China). In con-trast, in the United States, sake is on the up-swing, with several modern sake manufactur-ing facilities, although the total U.S. market isstill very small.

Manufacture of sake

Sake is different from other wine fermenta-tions in at least two main respects. First, asnoted above, fermentable sugars are absent inrice, the sake substrate.Thus, it is necessary toprovide exogenous enzymes, in the form of akoji, that can hydrolyze starch to simple sugars



that the yeasts can ferment. This part of thesake manufacturing process, therefore, sharessimilarity with the beer-brewing process, inwhich malt is used to convert the starch (inbarley) to simple sugars. The other major dif-ference that distinguishes sake productionfrom wine is that the saccharification step justdescribed and the actual ethanolic fermenta-tion step occur simultaneously or in parallel. Inother words, nearly as soon as sugars are madeavailable by action of koji enzymes, they arequickly fermented by sake yeasts.The implica-tions of these parallel processes will be dis-cussed below.

The sake-making process starts by preparinga rice koji (Figure 12–6).There are actually sev-eral different types of sake produced in Japan,

based in part on whether alcohol is added, butmainly on how the rice used to make the koji ismilled (or polished). In general, the whole orbrown rice is milled to remove 25% to 50% ofthe surface material (containing the germ andbran), which is necessary since the fat and pro-tein components are undesirable.Next, the riceis rinsed and soaked for several hours toachieve about 30% moisture.The moist rice isthen briefly steamed (�1 hour) and cooled to30°C to 35°C.About three-fourths of this rice isremoved and cooled further to 5°C to 10°C forlater use (see below). The remaining fourth isused for making koji.

To make the koji, the rice is inoculated withabout 0.01% of a tane koji (see above) or an A.oryzae spore culture.The material is incubated

Brown rice

Mill

Rinse and soak

Steam

Cool

Water

Moto

Sake yeast

Ferment

Add tane koji

Koji

Koji, rice, water

Moromi

Ferment

Filter

Pasteurize AgeBottle Blend

Figure 12–6. Sake flow chart. Adapted from Yoshiwawa and Ishikawa, 2004.



for forty to forty-eight hours at 30°C to 35°Cunder high humidity.The koji is mixed at inter-vals to re-distribute the growing fungi, main-tain aerobiosis, and prevent excessive heatbuild-up. When finished, the rice should becovered with fungal mycelia and should con-tain high amylolytic and proteolytic activities.

Next, the koji is moved to a tank, and thesteamed rice (from above) and water areadded.This material, called moto, is essentiallya pre-culture whose purpose is to increase thepopulation of endogenous yeast and lactic acidbacteria (see below) and to initiate a fermenta-tion. According to traditional sake brewingpractices, this moto seed culture incubates forabout two weeks at 15°C, and then definedsake yeast strains are added. It is now morecommon,however, to add sake yeast at the out-set, rather than rely on wild yeast for the initialfermentation. As the moto incubates, enzy-matic hydrolysis of rice starch and proteins oc-cur, releasing sugars and nutrients that supportgrowth of the yeast. The moto is eventuallytransferred to large tanks (capable of accom-modating more than 10,000 kg of rice) con-taining more koji, steamed rice, and water forcombined mashing and fermentation.

The fermenting mixture, now called mo-romi, is then diluted several times with anequal amount of koji-rice-water, at two-to-threeday intervals, such that the now thick ,moromi

mash increases in volume three- to four-fold.This process ensures that the yeast populationremains high throughout the fermentation andprovides, on a step-wise basis, adequateamounts of substrates to support an extended,semi-continuous fermentation.This unique fer-mentation process also serves to maintain lac-tic acid levels and a moderately high solidscontent, such that contaminating organismsare inhibited.Thus, even though the sugar con-centration never reaches much more than 6%to 8%, the ethanol concentration in the fin-ished product may be 18% or higher. Toachieve such a high yield of ethanol by a sim-ple batch fermentation (i.e., where all of thefermentable sugar was initially present) wouldrequire a mash containing as much as 30% glu-cose.This is far more than could be osmoticallytolerated by the yeast and would result in astuck fermentation.

The S. cerevisiae strains used for sake manu-facture are different from typical wine andbeer strains, in that they generally have higherosmotic, acid, and ethanol tolerance.They alsoproduce copious amounts of foam, which oc-cupies about a third of the volume in the fer-mentation tank, reducing the efficiency of thefermentation. Strains that are defective in mak-ing foam (i.e., foamless mutants) could have in-dustrial advantages (Box 12–4). In addition toyeast, other organisms are also involved in the

Box 12–4. Improving Sake Yeasts

Although the origin of sake manufacture, like most other fermentations, dates back thousandsof years, ways to improve the fermentation are now actively being explored. In particular, therehas been interest in modifying sake yeasts to enhance fermentation efficiency, flavor, and theoverall quality of the finished product. One specific approach, described below, involves isolat-ing a particular class of foam-defective mutants that have been used successfully for more thanthirty years in the Japanese sake industry. The molecular basis for this phenotype of thesestrains, however, has only recently been studied.

Growth of sake strains of Saccharomyces cerevisiae results in formation of a thick froth orfoam layer.The foam occupies so much volume in the fermentation tank (as much 50%) that lessspace is available for the moromi. High throughput processes are required in modern, largescale sake manufacturing facilities. High-foaming fermentations decrease production capacity.Fermentation efficiency could be improved if foam formation could be reduced using non-foaming yeast strains.

(Continued)



sake fermentation. In particular, endogenouslactic acid bacteria, including Lactobacillussake and other Lactobacillus and Leuconostocsp., grow and produce acid early in the fer-mentation. This early acidification evidentlyhelps to control other adventitious organisms,

although the lactics are eventually inhibited bythe high ethanol, high acid, and high osmoticpressure that accumulates in the moromi. Lac-tic acid can be added directly to promote acid-ification, a now-common practice in modernsake facilities.

This rationale led researchers in Japan to establish screening protocols so that non-foamingmutants could be isolated (Ouchi and Akiyama, 1971).The screening strategies were based onthe observation that froth or foam is formed either by strains that agglutinate to one another viaa capsular material or by strains that have affinity to the surface of gas bubbles. One method in-volved separating agglutinating wild-type cells from spontaneous non-agglutinating mutantsduring a series of successive fermentations, thereby concentrating the non-agglutinating, non-foaming cells.Alternatively, a similar approach was used to isolate S. cerevisiae cells that did nothave bubble affinity. In this case, cells were grown in broth in the presence of sparged air (toform bubbles).Wild-type cells,having affinity for bubbles, stuck to the bubbles and accumulatedin the froth layer.The non-foaming mutants could then be retrieved from beneath the froth layerand re-inoculated into fresh broth, with the entire cycle repeated several times.The mutantswere then isolated by plating, characterized to ensure other desirable properties had not beenlost, and then used successfully in commercial production (and are still in use).

A more recent investigation on the molecular differences between foaming and non-foamingsake strains has led to the identification of a gene that confers the foaming phenotype (Shimoiet al., 2002).The AWA1 gene (“awa” is Japanese for foam) encodes for a cell wall protein that iscovalently attached to glucans located within the cell wall. Insertion mutants that did not ex-press Awa1 did not form foam in a sake mash. Interestingly, this gene is not present in the S.cere-visiae genome database (built from laboratory strains), indicating it is likely specific to sakestrains.Using a method that measures cell surface hydrophobicity (i.e.,adherence ability), it wasalso shown that wild-type cells were significantly more “hydrophobic” than non-foaming cells.Thus, the latter cells are less likely to attach to other cells or to the surface of gas bubbles thatform during the sake fermentation.

A second protein involved in the foaming phenotype was also recently identified by proteomeanalysis using two-dimensional electrophoresis.The protein,YIL169c,was qualitatively and quan-titatively different between the foaming parental strain and a non-foaming mutant. Protein se-quence comparison of YIL169c with the S. cerevisiae genome data base revealed that this pro-tein was similar (71.5% identity) to YOL155c, a cell wall protein with glucosidase activity.Thus,although the function of YIL169c has not been determined, it appears that this protein likely pro-trudes from the cell wall, where it can interact with bubbles or other hydrophobic surfaces.

References

Ouchi, K., and H. Akiyama. 1971. Non-foaming mutants of sake yeasts: selection by cell agglutinationmethod and by froth flotation method.Agri. Biol. Chem. 35:1024–1032.

Ouchi, K., and Y. Nunokawa. 1973. Non-foaming mutants of sake yeasts: their physio-chemical characteris-tics. J. Ferment.Technol. 51:85–95.

Shimoi, H., K. Sakamoto, M. Okuda, R.Atthi, K. Iwashita, and K. Ito. 2002.The AWA1 gene is required for thefoam-forming phenotype and cell surface hydrophobicity of sake yeast. Appl. Environ. Microbiol.68:2018–2025.

Yamamoto,Y., K. Hirooka,Y. Nishiya, and N.Tsutsui. 2003.The protein which brings about foam-forming bythe sake yeast. Seibutsu-Kogaku Kaishi 81:461–467 (Abstract only).

Box 12–4. Improving Sake Yeasts (Continued)



After about three weeks, the fermentation isended, and the moromi is separated by settlingand filtration, yielding a sake cake and a veryclear, light yellow liquid, which is then calledsake.The sake can be aged (usually only a fewmonths), adjusted to a desirable ethanol con-tent, pasteurized by heat (65°C) or ultrafiltra-tion, and bottled. Like grape-derived wine, thefinished sake can be sweet or dry. However,this classification is not entirely based on theamount of residual sugar that is present, butrather on a combination of the sugar, acid, andalcohol content. Depending on the type ofsake and consumer preference, sake can beserved warmed (35°C to 40°C) or slightlychilled. In general,higher quality sake is usuallyserved at the lower temperature to retain moreof the aroma and flavor volatiles.

Fermented Fish-type Foods

Although fish and shrimp sauces and pastes aremostly unknown to Western consumers, theyare staple items in much of Southeast Asia, in-cluding Thailand, Vietnam, Malaysia, and thePhilippines (from where such products werethought to have evolved). Although fresh fishhas long been widely available throughout thisregion, refrigeration has not. Thus, fish wouldlikely spoil before it could be consumed.In con-trast,fish-type sauces and pastes not only have along shelf-life, they also serve as an inexpensivesource of high quality protein and other nutri-ents. In addition, these products significantly enhance the flavor of rice, noodles, and otherbland-tasting foods. Per capita consumptionranges from 10 ml to 15 ml per day in Thailandto about 1 ml per day in the Philippines.

The manufacture of fish sauces in SoutheastAsia is a major industry; over 40 million litersof fish sauce are produced annually in Thailandalone. The popularity of cuisines from thosecountries in the United States (in food serviceas well as in processed foods) has undoubtedlyled to consumption of fish sauces by U.S. con-sumers, even if they aren’t aware of what actu-ally contributes to the unique flavors of thosefoods. In regions where these products are pro-

duced and consumed, they are considered in-dispensable flavoring agents, much like salt inthe United States or shoyu in Japan.

There are many types of fermented fishproducts, ranging from light- or dark-coloredpourable sauces to very thick pastes (Table12–9). In some cases, the same manufacturingprocess is used to make both a liquid and apaste. For example, in the Philippines, a saltedand fermented fish mixture, when allowed tosettle, yields a supernatant liquid called patisand a sediment that, when dried, gives a pastecalled bagoong. Just as with soy sauce andother fermented foods from Asia and the FarEast, there are countless versions of fish saucesand pastes, many of which are made in thehome or on a very small cottage scale using tra-ditional techniques. Thus, only rather genericdescriptions will be given (see below).

Although fish sauces are considered to befermented foods, this is true only in a ratherbroad sense. Microorganisms do, indeed, growand produce end products during the manu-facture of these products (see below). How-ever, most of the reactions that are responsiblefor flavor and aroma development occur as aresult of endogenous fish enzymes, releasedduring autolysis of fish tissue, rather than frommicrobial activities.

Manufacture of fish sauces and pastes

The general procedure for the production offish sauces is not complicated.The starting ma-terial can be small (�15 cm in diameter) fish,

Table 12.9. Types of fish sauces and pastes.

Product Country Description

Nam-pla Thailand Fish sauceBudu Malaysia Fish saucePatis Philippines Fish or shrimp sauceIshiru Japan Fish sauceNouc-mam Vietnam Fish sauceNam-pa Laos Fish sauceBagoong Philippines Fish or shrimp pasteMam Vietnam Fish pasteTrassi Indonesia Shrimp or fish pasteBelachan Malaysia Shrimp paste



such as sardines (that otherwise have minimalcommercial value), small shrimp, squid, or oys-ters (Figure 12–7). Fish is usually used wholeand uneviscerated, although de-headed, evis-cerated, ground, or cut-up pieces can also beused. The only other ingredient necessary tomake these products is salt.The fish-to-salt ra-tio varies, depending on the product, but usu-ally ranges from 3:1 to 5:1. Fish sauces do notundergo a lactic fermentation, per se, and arepreserved mainly by salt and low water activ-ity, rather than by pH.Thus, high salt concen-trations are necessary.

After the salt is added to the fish (on woodenor concrete floors), the mixture is moved into

tanks (often built into the ground) and covered.The material is held for about six months (orlonger) at ambient temperature. At varioustimes, the mixture may be uncovered, stirred,and exposed to air and sunlight,all of which arethought to improve flavor and color and accel-erate enzyme activity. During this incubation period, the solid fish material is transformed,or more precisely, liquefied by the action of endogenous fish enzymes. These enzymes, pri-marily trypsin-like acid-proteases and variousendo- and exo-peptidases, are ordinarily presentwithin the intact cells of various fish tissues.However, in the non-living animal, the cells soonautolyze and those enzymes are released, result-

Salt

Ferment

Settle

Liquid

Filter

"Air"

Sediment

Fish or Shrimp

Bottle(1st Grade)

Brine

Fillter

Steep

Bottle(2nd Grade)

Paste

Figure 12–7. Fish sauce flow chart.



ing in extensive hydrolysis of muscle tissue. Infresh fish, autolysis and proteolysis result in tis-sue softening and spoilage; in fish sauce produc-tion, the result is liquefaction.

In addition to the physical transformationfrom solid to liquid, proteolytic digestion of thefish substrates results in formation of free aminoacids and peptides. In intact tissue, for example,the soluble nitrogen concentration is essentiallynil,but in nam-pla and nuoc-mam (Thai and Viet-namese fish sauce, respectively) there is morethan 2% soluble nitrogen (mostly amino nitro-gen). Glutamic acid, which, like in soy sauceproducts,is responsible for flavor enhancement,is among the amino acids that accumulate infish sauce. Likewise, 5�-nucleotides may also beformed, providing a source of umami or meaty-like flavors (as described above).

Further hydrolysis of peptides and aminoacids by enzymes that are either endogenousor microbial in origin (see below) eventuallyresults in a large number of volatile aroma andflavor products. Among those that are mostprominent and that confer “fish sauce flavor”are ammonia, triethylamine, and various alco-hols, aldehydes,ketones, and lactones.Lipolysisalso occurs during fish sauce manufacture, re-sulting in formation of volatile fatty acids, in-cluding acetic, butanoic, and propanoic acids.These compounds are particularly characteris-tic of fish sauce flavor, which is sometimes de-scribed as “cheesy.”

After several months of enzymolysis and fer-mentation, the liquid is separated from the sed-imented material by decanting or filtering theliquid directly through the fish solids.This “firstrun”product has the highest quality. Additionalbrine can be added to the solid material, themixture aged for several more weeks (or sim-ply boiled), and then a second, lower quality,liquid is obtained. The remaining solids canthen be recovered and used as a paste. Somefish sauces and pastes are aged in the open(and exposed to the sun) for several moreweeks to allow partial dissipation of the strongfish aroma.The sauce or paste is then bottled.The final composition (weight basis) of fishsauces is usually about 60% moisture, 30% salt,

10% protein (including amino nitrogen),with afinal pH of about 6.5. Pastes contain about 30%moisture, 20% salt, 30% protein, and 20% ash.

Fish sauce microbiology

While it is evident that microorganisms arepresent during the production of fermentedfish sauces, it is not clear to what extent theseorganisms contribute to the finished product.The microbial population in raw, unsalted fishand shellfish is high in number and rich in di-versity. Considering the fact that whole unevis-cerated fish (guts and all) are usually used tomake fish sauces, the initial load of organismsis significant.

In addition, the manufacturing environmentis not aseptic, and even the salt (usually ob-tained by solar drying of sea water) may con-tribute microorganisms. However, the high saltconcentration established early on in the fer-mentation provides strong selective pressurefor halotolerant organisms. Not surprisingly,therefore, there is a shift in the initial bacterialpopulation, from a wide variety of aerobic andanaerobic organisms to a more narrow micro-flora consisting mainly of salt-tolerant species of Bacillus, Halobacillus, Staphylococcus, andMicrococcus. Lactic acid bacteria are alsowidely present in fish sauce (Table 12–10), in-cluding some strains that are capable of grow-ing in media containing 12% (2.1 M) salt.

Table 12.10. Lactic acid bacteria isolated from fishsauces1.

Organism

Lactobacillus plantarumLactobacillus farciminisLactobacillus acidipiscisLactobacillus pentosusWeissella thailandensisLeuconsotoc citreumLactococcus lactic subsp. lactisTetragenococcus muriaticus

1From Tanasupawat et al., 1998,Tanasupawat et al., 2000,

Paludan-Muller et al., 1999, and Kimura et al., 2001



Establishing the function of the microfloraduring the fish sauce fermentation is not easy.It has been suggested that there is a succes-sion that occurs during the fermentation, lead-ing to the eventual dominance of the moresalt- and acid-tolerant organisms. While thismay be true for some products, in which con-siderable microbial growth is evident, forother products, microbial growth hardly oc-curs. In either case, the population eventuallydecreases, such that at the end of the fermen-tation (six to twelve months), there are usuallyless than 103 cells per ml. Many of the organ-isms isolated from fermented fish productshave proteolytic activity and likely contribute,at least in part, to the overall proteolysis of fishprotein. Lactic acid bacteria and other anaer-obes also can metabolize amino acids, produc-ing volatile fatty acids, amines, ammonia, andother volatile end-products.

Safety of Fungal Fermented Foods

There are two main reasons why the safety ofAsian, fungal-fermented foods has been ques-tioned.First, the Aspergillus sp.used in the pro-duction of soy sauce, miso, sake, and relatedproducts are taxonomically similar to the my-cotoxigenic aspergilli that produce aflatoxins,ochratoxin,and other toxins.Despite these sim-ilarities, however, surveys in which these foodshave been analyzed for the presence of myco-toxins indicate that mycotoxins are not pre-sent. Recent studies have shown that very sub-tle genetic differences exist between industrialstrains and mycotoxin-producing strains thatwould account for these findings (Box 12–5).

A second reason for the concern is due tothe suggested epidemiological relationship be-tween consumption of indigenous foods fromthe Far East and the development of certain

Box 12–5. Mycotoxins in Fungal Fermented Products: Cause for Concern?

In nature,growth of fungi is often accompanied by production of mycotoxins,produced as sec-ondary metabolites. Some of these mycotoxins—and aflatoxins, in particular—are extremelycarcinogenic and mutagenic. In fact, aflatoxins are considered to be among the most toxic of allnaturally occurring compounds found on the planet.

There are as many as sixteen different types of aflatoxins,of which B1,B2,G1,and G2 are mostcommonly found in human and animal foods.They are produced mainly by Aspergillus flavusand Aspergillus parasiticus, members of the Aspergillus section Flavi (or the A. flavus group).This taxonomic group also includes Aspergillus sojae and Aspergillus oryzae, species that areused in the manufacture of koji, soy sauce, miso, sake, and other Oriental fermented products.

Given the morphological and genetic similarities between these four different species, ques-tions have been raised regarding the potential ability of A. sojae and A. oryzae to produce my-cotoxins in foods in which these fungi are used. Indeed, several recent studies have revealedthat some of the genes (or homologs) that encode for proteins involved in the aflatoxin biosyn-thesis pathway may be present in these food production strains (see below).Despite the appar-ent metabolic potential of A. sojae and A. oryzae to produce aflatoxins, however, there are noreports of aflatoxins being produced by these fungi or for the presence of aflatoxins in fer-mented soy products (Watson et al., 1999).These observations have led investigators in Japan,the United Kingdom, and the United States to explore, in detail, why A. sojae and A. oryzae donot synthesize aflatoxins.

The biochemical pathway for aflatoxin biosynthesis consists of at least twenty-three enzy-matic reactions and more than fifteen intermediates (Yu et al, 2004A and 2004B). In 2004, thecomplete gene cluster encoding for aflatoxin biosynthesis in A. flavus and A. parasiticus wasidentified and sequenced (Yu et al., 2004A; Figures 1 and 2).There are twenty-five genes withinthis 70 Kb cluster (four additional sugar utilization genes are located immediately downstream).

Because enzymatic activities associated with the aflatoxin pathway are generally not detectablein A.sojae, it would appear that the relevant genes must also be absent.However,according to one



Box 12–5. Mycotoxins in Fungal Fermented Products: Cause for Concern? (Continued)

(Continued)

Figure 1. Linear arrangement of the afl gene cluster from Aspergillus parasiticus. Individual gene namesare abbreviated by a single letter (where aflR � R), and are named according to the scheme proposed by Yuet al., 2004B. Genes described in the text include pksA (aflC), aflJ (aflS), norA (aflE), avfA (aflI), and omtB(aflO). Four genes at the 3’ end of the afl cluster encode for a sugar utilization pathway, followed by a fifthgene whose function is unknown. Adapted from Yu et al., 2004A and 2004B.

F U T C D B A R S H J E MN G L I O P Q K V W X Y

report (Matsushima et al., 2001B), three aflatoxin genes (aflR, nor1, and pksA) were present in acommercial soy sauce strain (strain 477),based on Southern hybridization analysis.Another studyshowed that omtA, nor1, and ver1 were also present in A. sojae and A. oryzae and that thesegenes had high sequence similarity to those from aflatoxin-producing strains of A. parasiticus(Watson et al., 1999). However, none of these genes were transcribed, even when cells weregrown in alfatoxin-conducive media.Two other genes, avfA and omtB, were also reported to bepresent in A. sojae and were 99% identical to those found in A. parasiticus (Yu et al., 2000). Fur-thermore,complementation assays showed that the A.sojae avfA gene was fully functional.Thus,if aflatoxin genes are present in A. sojae and A. oryzae, and they appear to be functional, whydon’t these fungi produce aflatoxin in foods?

The answer to this question can now be revealed.The aflatoxin pathway in A. parasiticus isregulated mainly by the aflR gene product AflR.An adjacent gene,aflS, appears to modulate aflRexpression, but its exact role is not known.AflR is a 47 kDa protein that acts as a positive regu-lator by binding to the promoter region of the afl gene cluster, activating transcription of thestructural genes. Mutant strains of A.flavus, defective in aflR, do not express alfatoxin genes, in-dicating that AflR is required for aflatoxin biosynthesis.Although a homolog of the aflR gene ispresent in A. sojae and A. oryzae, afl genes are still not transcribed in these strains.The non-aflatoxin-producing phenotype is not due to the absence of a promoter-binding region, sincethat region has been shown to be functional (Takahashi et al.,2002).Rather, it is AflR that is non-functional, not only in the native A. sojae and A. oryzae backgrounds, but also in an A. parasiti-cus aflR deletion mutant transformed with aflR from A. sojae (Takahashi et al., 2002). Also,whereas introduction of an additional copy of the A. parasiticus aflR gene into A. parasiticusincreased aflatoxin production, no such effect was observed when the aflR from A. sojae wasintroduced into A.parasiticus (Matsushima et al., 2001A).

Sequence analysis has shown that there are two mutations located within the A. sojae aflAgene (Watson et al.,1999).One mutation consists of a duplication of histidine (H) and alanine (A)residues at amino acid positions 111 to 114, resulting in a HAHA motif (no joke).The second mu-tation consists of a C– T transition, and subsequent conversion of an arginine codon at position385 to a premature stop codon.This results in a truncated protein that is short sixty-two aminoacid residues at the carboxy end, which is where the transcription-activating domain is located.The HAHA mutation, it turns out, however, does not appear to affect AflR activity or synthesis ofaflatoxin,because cells harboring a hybrid gene containing the intact carboxy-terminal region—but with the HAHA motif—have a normal aflatoxin phenotype (Matsushima et al., 2001A).Thus,the defect in aflR must be due to the missing sixty-two residue region.The truncated AflR is thenunable to activate transcription of alf genes, resulting in the lack of aflatoxin biosynthesis.




A

0 01 02 03

norB

cypA

aflT

pksA

nor-

1

fas-

2

fas-

1

aflR

aflJ

adhA

aflF

aflU

aflT

aflC

aflD

aflB

aflA

aflR

aflS

aflH

0 01 02 030 01 02 03

norB

cypA

aflT

pksA

nor-

1

fas-

2

fas-

1

aflR

aflJ

adhA

aflF

aflU

aflT

aflC

aflD

aflB

aflA

aflR

aflS

aflH

aflF

aflU

aflT

aflC

aflD

aflB

aflA

aflR

aflS

aflH

HO

OH

OO

HO

OH

O

OH

OHO

O O

HO

HO

Me

O

OH

HO

O

O

OM

eH

O

Me

OO

OO

H

OH

HO

HO

O O

HO

OO

H

OH

OH

O

OH

O

HO

OH

O

OO

HO

HH

O

HO

OO

H

OH

OH

O

OH

OH

NO

R

AV

N

AV

NN

AV

F

VH

A

VA

L

HA

VN

?

Poly

ketid

e sy

ntha

se

Red

ucta

se

Fatty

aci

d sy

ntha

se ??

NO

R-r

educ

tase

Fatty

aci

d sy

ntha

se ??

Deh

ydro

gena

se

Oxi

dase

Est

eras

e

P450

mon

ooxy

gena

se

Alc

ohol

deh

ydro

gena

se

aflA

(fa

s-2)

aflB

(fa

s-1)

aflC

(pk

sA)

aflD

(no

r-1)

af

lE (

norA

)af

lF (

norB

)

aflG

(av

nA)

aflH

(ad

hA)

aflI

(av

fA)

aflJ

(es

tA)

stcJ

stcK

stcA

stcF

stcG

stcO

stcI

stcV

stcE

OA

VN

BC

D

ACET

ATE

POLY

KETI

DE

Fig

ure

2.

The

aflat

oxin

bio

synt

hetic

pat

hway

in t

oxig

enic

Asp

ergi

llus.

The

82

kb a

flato

xin

gene

clu

ster

fro

m A

. pa

rasi

ticus

and

A.

flavu

sis

show

n in

pan

el A

(new

nam

es o

n th

e le

ft a

nd o

ld n

ames

on

the

right

). Th

e ge

ne c

lust

er c

odin

g fo

r su

gar

utili

zatio

n is

als

o sh

own.

The

dire

c-tio

n of

gen

e tr

ansc

riptio

n is

sho

wn

alon

g th

e ve

rtic

al li

ne.

The

enco

ded

aflat

oxin

bio

synt

hetic

enz

ymes

are

sho

wn

in p

anel

B,

alon

g w

ith t

heafl

atox

in (m

iddl

e co

lum

n) a

nd s

terig

mat

ocys

tin (f

ar r

ight

) pat

hway

s.A

bbre

viat

ions

: N

OR,

nor

solo

rinic

aci

d; A

VN

, av

eran

tin;

HA

VN

, 5’

-hyd

roxy

-ave

rant

in;

OA

VN

, ox

oave

rant

in;

AV

NN

, av

eruf

anin

; A

VF,

aver

ufin

; V

HA

, ve

rsic

onal

hem

iace

tal a

ceta

te;

VAL,

ver

sico

nal;

VER

B, v

ersi

colo

rin B

; V

ERA

, ve

rsic

olor

in A

; D

MST

, de

met

hyls

terig

mat

ocys

tin;

DH

DM

ST,

dihy

drod

emet

hyls

terig

mat

ocys

tin;

ST,

ster

igm

atoc

ystin

; D

HST

, di

hydr

oste

rigm

atoc

ystin

; O

MST

, O

-met

hyls

terig

mat

ocys

tin;

DH

OM

ST,

dihy

dro-

O-m

ethy

lste

rigm

atoc

ystin

; A

FB1,

afla

toxi

n B 1

; A

FB2,

afla

toxi

n B 2

; A

FG1,

afla

toxi

n G

1;

and

AFG

2,

afla

toxi

n G

2.

From

Yu

etal

., 20

04, w

ith p

erm

issi

on.




(Continued)

04 05 06 07 bk 08

estA

norA

ver-

1ve

rA

avnA

verB

avfA

omtB

omtA

ordA

vbs

cypX

mox

Yor

dB

nadA

hxtA

glcA

sugR

orf

hypA

aflJ

aflE

aflM

aflN

aflG

aflL

aflI

aflO

aflP

aflQ

aflK

aflV

aflW

aflX

aflY retsulc ragus

04 05 06 07 bk 0804 05 06 07 bk 08

estA

norA

ver-

1ve

rA

avnA

verB

avfA

omtB

omtA

ordA

vbs

cypX

mox

Yor

dB

nadA

hxtA

glcA

sugR

orf

hypA

aflJ

aflE

aflM

aflN

aflG

aflL

aflI

aflO

aflP

aflQ

aflK

aflV

aflW

aflX

aflY

aflJ

aflE

aflM

aflN

aflG

aflL

aflI

aflO

aflP

aflQ

aflK

aflV

aflW

aflX

aflY retsulc ragus

DH

ST

DH

OM

STO

MSTST

OO

O

OM

eO

O

O

AF

B1

AF

G1

AF

B2

AF

G2

OH

O

O

OO

OM

e

O

O

OM

e

OM

eO

O

OO

O

OM

eO

O

OO

O

OM

eO

O

OM

eO

O

O

O

OM

e

OM

eO

O

O

O

OH

OO

OH

O

OH

HO

O

OO

O

OM

eO

O

O

HO

HO

OO

H

OO

O

VE

RA

VE

RB

DH

DM

STD

MST

OH

O

O

OO

OH

OH

OO

O

O

OH

VE

RB

syn

thas

e

Des

atur

ase

O-m

ethy

ltran

sfer

ase

B

O-m

ethy

ltran

sfer

ase

A

dehy

drog

enas

e

Oxi

dore

duct

ase

Mon

ooxy

gena

se

aflK

(vb

s)

aflL

(ve

rB)

aflM

(ve

r-1)

aflN

(ve

rA)

aflO

(om

tB)

aflP

(om

tA)

aflQ

(or

dA)

stcN

stcL

stcS

stcP

stcU

Fig

ure

2.

(Con

tinue

d)



cancers, including gastric cancer andesophageal cancer. Indeed, although somecancer rates are higher in China, Japan, andother Far East countries, the evidence suggeststhat dietary habits other than consumption offungal-fermented foods may be more likely tobe responsible. For example, the Asian diet ishigh in salt, which may predispose individualsto these cancers. In specific provinces inChina, where cancer rates are especially high,consumption of vegetables is very low.

Finally, it is also noteworthy that positive nu-tritional factors have been associated with fun-gal fermented foods. kIn particular, consump-tion of soy beans (and fermented soy products)

have been promoted for contributing manyhealth benefits, including reduced rates of coro-nary heart disease and mortality due to heartdisease. Soy beans contain several flavones,isoflavones, flavanols, and other flavonoids thatmay have anticarcinogenic activity and whoseconsumption may lower the risk of certain can-cers, including breast cancer and prostate can-cer. Although not all of these substances arepresent in fermented soy products, several spe-cific compounds have been shown to have anti-cancer activity. For example, the flavor fura-nones HEMF and HMF (discussed previously)that are found in soy sauce inhibit carcinogene-sis in laboratory animals.

According to another recent investigation, mutations in a second regulatory gene, aflJ (seeabove), may also be responsible, in part, for non-expression of afl genes in A. sojae. In A. para-siticus, aflJ gene product, AflJ, upregulates synthesis of aflatoxin. Mutations in aflJ decreasedtranscription of both afl structural genes and aflatoxin synthesis, although alfR expression wasnot affected (Chang, 2003).Thus, it appears that AflJ is a co-activator that interacts with AflR toactivate gene transcription. However, analogous to the aflR situation, the aflJ gene in A. sojae isdefective, owing to two amino acid substitutions that reduce this interaction and reduce tran-scription of afl genes (Chang, 2004).

References

Chang, P.-K. 2004. Lack of interaction between AFLR and AFLJ contributes to nonaflatoxigenicity of As-pergillus sojae. J. Biotechnol. 107:245–253.

Chang, P.-K. 2003.The Aspergillus parasiticus protein AFLJ interacts with the aflatoxin pathway-specificregulator AFLR. Mol. Gen. Genomics 268:711–719.

Matsushima, K., P.-K. Chang, J.Yu, K.Abe, D. Bhatnagar, and T.E. Cleveland. 2001A. Pre-termination in aflR ofAspergillus sojae inhibits aflatoxin biosynthesis.Appl. Microbiol. Biotechnol. 55:585–589.

Matsushima, K., K.Yashiro,Y. Hanya, K.Abe, K.Yabe, and T. Hamasaki. 2001B.Absence of aflatoxin biosyn-thesis in koji mold (Aspergillus sojae).Appl. Microbiol. Biotechnol. 55:771-776.

Takahashi,T.,P.-K.Chang,K.Matsushima, J.Yu,K.Abe,D.Bhatnagar,T.E.Cleveland,and Y.Koyama.2002.Non-functionality of Aspergillus sojae aflR in a strain of Aspergillus parasiticus with a disrupted aflR gene.Appl. Environ. Microbiol. 68:3737–3743.

Watson,A.J., L.J. Fuller, D.J. Jeenes, and D.A.Archer. 1999. Homologs of aflatoxin biosynthesis genes and se-quences of aflR in Aspergillus oryzae and Aspergillus sojae.Appl. Environ. Microbiol. 65:307–310.

Yu, J., D. Bhatnagar,T.E. Cleveland. 2004A. Completed sequence of aflatoxin pathway gene cluster in As-pergillus parasiticus. FEBS Lett. 564:126–130.

Yu, J.,P.-K.Chang,K.C.Ehrlich, J.W.Cary,D.Bhatnagar,T.E.Cleveland,G.A.Payne, J.E.Linz,C.P.Woloshuk,andJ.W. Bennett. 2004B. Clustered pathway genes in aflatoxin biosynthesis. Appl. Environ. Microbiol.70:1253–1262.

Yu, J.,C.P.Woloshuk,D.Bhatnagar, and T.E.Cleveland.2000.Cloning and characterization of avfA and omtBgenes involved in aflatoxin biosynthesis in three Aspergillus species. Gene 248:157–167.




BibliographyBaens-Arcega,L.,A.G.,Ardisher,C.G.Bellows, and 31

other authors. 1996. Indigenous amino acid/peptide sauces and pastes with meatlike flavor.p. 509–654. In K.H. Steinkraus (ed.), Handbookof Indigenous Fermented Foods, 2nd ed. Mar-cel Dekker, Inc. New York, New York.

Beddows, C.G. 1998. Fermented fish and fish prod-ucts, p. 416–440. In B.J.B. Woods, (ed.) Micro-biology of Fermented Foods, Volume 1. BlackieAcademic and Professional. London.

Beuchat, L.R. 2001. Traditional fermented foods,p. 701–719. In M.P. Doyle, L.R. Beuchat, and T.J.Montville (eds.). Food Microbiology: Funda-mentals and Frontiers, 2nd Ed. ASM Press,Washington, D.C.

Ebine, H. 2004. Industrialization of Japanese misofermentation, p. 99–147. In K.H. Steinkraus (ed.),Industrialization of indigenous fermentedfoods, 2nd ed. Marcel Dekker, Inc., New York,New York.

Ebine, H. 1986. Miso, p. 47–68. In N.R. Reddy, M.P.Pierson, and D.K. Salunkhe (eds.), Legume-basedfermented foods. CRC Press. Boca Raton, Florida.

Fukushima, D. 2004. Industrialization of fermentedsoy sauce production centering around Japaneseshoyu, p. 1–98. In K.H. Steinkraus (ed.), Industri-alization of indigenous fermented foods, 2nded. Marcel Dekker, Inc., New York, New York.

Hachmeister, K.A., and D.Y.C. Fung. 1993.Tempeh: amold-modified indigenous fermented food madefrom soybeans and/or cereal grains. Crit. Rev. Mi-crobiol. 19:137–188.

Keuth, S., and B. Bisping. 1993. Formation of vita-mins by pure cultures of tempe moulds and bac-teria during the tempe solid substrate fermenta-tion. J.Appl. Bacteriol. 75:427–434.

Kimura, B.,Y. Konagaya, and T. Fujii. 2001. Histamineformation by Tetragenococcus muriaticus, ahalophilic lactic acid bacterium isolated fromfish sauce. Int. J. Food Microbiol. 70:71–77.

Kiuchi, S., and S.Watanabe. 2004. Industrialization ofJapanese natto. p. 193–246. In K.H. Steinkraus(ed.), Industrialization of Indigenous Fer-mented Foods, 2nd ed. Marcel Dekker. NewYork, New York.

Mulyowidarso, R.K., G.H. Fleet, and K.A. Buckle.1989.The microbial ecology of soybean soakingfor tempe production. Int. J. Food Microbiol.8:35–46.

Nout, M.J.R., and F.M. Rombouts. 1990. Recent devel-opments in tempe research. J. Appl. Bacteriol.69:609–633.

Nout, M.J.R., and J.K. Kiers. 1995. Tempe fermenta-tion, innovation and functionality: update intothe third millennium. J.Appl. Microbiol. 98:789–805.

Nunomura, N., and M. Sasaki. 1992. Japanese soysauce with emphasis on off-flavors, p. 287–312.In G. Charalambous (ed.). Off-flavors in foodsand beverages. Elsevier Science Publishers.Ams-terdam, Netherlands.

Ohta, T. 1986. Natto, p. 85–93. In N.R. Reddy, M.P.Pierson, and K. Salunkhe (eds.), Legume-basedfermented foods. CRC Press. Boca Raton, Florida.

Paludan-Muller, C., H.H. Huss, and L. Gram. 1999.Characterization of lactic acid bacteria isolatedfrom a Thai low-salt fermented fish product andthe role of garlic as substitute for fermentation.Int. J. Food Microbiol. 46:219–229.

Ruiz-Terán, F., and J.D. Owens. 1996. Chemical andenzymic changes during the fermentation of bac-teria-free soya bean tempe. J. Sci. Food Agri.71:523–530.

Schipper,M.A.A.,and J.A.Stalpers.1984.A revision ofthe genus Rhizopus, II. The Rhizopus mi-crosporus-group. Stud. Mycol. 25:20–34.

Shurtleff, W., and A. Aoyagi. 1979. Tempeh Produc-tion. The Book of Tempeh: Volume II. New AgeFoods, Lafayette, California.

Sparringa, R.A., and J.D. Owens. 1999. Protein utiliza-tion during soybean tempe fermentation. J.Agric.Food Chem. 47:4375–4378.

Tanasupawat, S., S. Okada, and K. Komagata. 1998.Lactic acid bacteria found in fermented fish inThailand. J. Gen.Appl. Microbiol. 44:193–200.

Tanasupawat, S.,O.Shida,S.Okada,and K.Komagata.2000. Lactobacillus acidipiscis sp. nov. andWeisella thailandensis sp. nov. isolated from fer-mented fish in Thailand. Int. J. Syst. Evol. Micro-biol. 50:1479–1485.

Thanh,N.V., and M.J.R.Nout.2004.Dormancy,activa-tion and viability of Rhizopus oligosporus spo-rangiospores. Int. J. Food Microbiol. 92:171–179.

Winarno, F.G., and N.R. Reddy. 1986. Tempe. p. 95–117. In N.R. Reddy, M.P. Pierson, and K. Salunkhe(eds.), Legume-based Fermented Foods. CRCPress. Boca Raton, Florida.

Yokotsuka,T. and M. Sasaki. 1998. Fermented proteinfoods in the Orient: shoyu and miso in Japan, p.351–415. In B.J.B. Woods (ed.), Microbiology ofFermented Foods, Volume 1. Blackie Academicand Professional. London.

Yoshizawa, K., and T. Ishikawa. 2004. Industrializa-tion of Sake manufacture p. 149—192. Industri-alization of indigenous fermented foods, 2nded. Marcel Dekker, Inc., New York, New York.


Index

Aflatoxin genes, 453AFLP (amplified fragment length polymorphism), 322AflR gene, 453Aged cheese, 158Agglomerates, 392Aging

cheese, 165–166wine, 380–382

�-glucoside transporter (AGT1), 324AGT1 (�-glucoside transporter), 324Airborne microorganisms, 372Air handling systems, 294Akamiso, 434Alcaligenes, 203Alcohol and Tobacco Tax and Trade Bureau (TTB), 354Alcohol dehydrogenase (AlDH), 325, 403, 415Alcoholic fermentations, 4Aldehyde dehydrogenase, 415AlDH (alcohol dehydrogenase), 325, 403, 415Ale, 303, 336Alsace-Lorraine, Germany, 304Alsatian wine, 369Altbier, 336Alternaria, 252American Society of Microbiology, 63American-type cheeses, 145American Viticulture Areas (AVA), 355Amino acid decarboxylase enzymes, 259Amino acids, 117, 323Aminopeptidases, 190, 195Amplified fragment length polymorphism (AFLP),

322Amylase activity, 273Amyloglucosidase, 339Amylolytic activity, 339Amylolytic enzymes, 421, 434Amylopectin, 269, 290Amylose, 269Amylose retrogrades, 290Anabolic enzymes, 297Anabolic pathways, 344Anaerobic pathways, 277–278Anaerobic pockets, 210Anguillula aceti, 414

�-acetolactate synthase, 60, 137�-acids, 316�-amylase, 313Aaps (acetate adaptation proteins), 405ABC transport systems in lactic acid bacteria, 53–56Abortive infection systems (Abi), 200Accelerated ripening, 191AcDH (acetaldehyde dehydrogenase), 403Acetaldehyde, 119, 141, 401Acetaldehyde dehydrogenase (AcDH), 403Acetaldehyde-TPP, 137Acetate adaptation proteins (Aaps), 405Acetate oxidation pathway, 405Acetate-specific stress proteins (Asps), 405Acetic acid bacteria, 393–394, 411, 412, 414Acetic acid fermentation, 403Acetic acid formation, 409Acetic acid oxidation, 415Acetic acid pathway, 403Acetic acid resistance (aar) operons, 405Acetobacter, 37, 334, 393, 394, 398, 399, 401, 404, 405,

406, 407, 408, 411, 414Acetobacteraceae, 37, 399Acetobacter aceti, 37, 75, 140, 393, 405, 408Acetobacterium, 399Acetobacter orleanensis, 37Acetobacter pasteurianus, 393, 399, 408Acetobacter pasteurianus subsp. pasteurianus, 37Acetous fermentation, of ethanol, 398Acid-coagulated cheeses, 166–169Acidification, 155, 439Acidomonas, 399Acidophilic organisms, 358Acid-producing bacteria, 135Acid-tolerance, 33Actinobacteria, 19, 20, 39Adapted cells, 405Adjuncts, 313–314Aeration, 257, 411Aerobes, 210Aerobic chemical reactions, 7–8Aerobic metabolism, 324Aerococcus, 25Aflatoxin biosynthesis, 450

457

INDEX_BW_Hutkins_277165 4/12/06 8:26 AM Page 457



458 Index

Antibiotics, in milk, 163Anti-botulinum agents, 223Anticarcinogenic constituents, 244Anti-fungal agents, 274Antimicrobial activity, 234Antimicrobial compounds, 215“Anti-nutritional”properties, 287Antioxidant properties, in olives and olive oil, 254AOC (Appellation d’Origine Contrôlée) laws, 353Appellation d’Origine Contrôlée (AOC) laws, 353ArcD (arginine transporter), 51Archaea, 16Arginine transporter (ArcD), 51Arrest germination, 310Arthrobacter, 182Aschyta, 252Asci-containing spores, 321Ascomycetes, 17Ascomycota, 17, 41Asian foods. See Oriental foodsAspergillus, 41, 43, 62, 220, 258, 292, 294, 389, 390, 425,

450Aspergillus flavus, 450Aspergillus nidulans, 154Aspergillus niger, 154, 338, 346Aspergillus oryzae, 43, 76, 422, 423, 424, 431, 434, 444,

450, 453Aspergillus parasiticus, 450, 453Aspergillus sojae, 43, 76, 422, 423, 424, 433, 450,

453Asps (acetate-specific stress proteins), 405ATP-binding cassette (ABC) systems, 54, 55–56, 317ATP synthase, 21ATP synthase reaction, 376AVA (American Viticulture Areas), 355avfA gene, 453AWA1 gene, 446A.xylinum, 409

Bacillus, 20, 37, 104, 121, 130, 235, 248, 257, 275, 292,425–426, 436, 440, 449

Bacillus cereus, 131Bacillus licheniformis, 293Bacillus mesentericus, 293Bacillus subtilis, 19, 37, 293Bacillus subtilis var. natto, 436Backslopping, 67–68, 163, 210–211, 367Bacteria, 18

surface-ripening of cheese by, 181–182used in manufacture of fermented foods, 19–20wine spoilage by, 393–395

Bacterial starter cultures, 74–75Bacteriocin nisin, 247Bacteriocins, 105, 215, 217Bacteriophage genomics, 71–72Bacteriophage problem, 95Bacteriophages, 71, 94–96, 116, 194–202, 239

Bacterium Güntheri, 77Bacterium lactis, 19, 77Bacterium lactis acidi, 77Baked sherries, 386Bakers’ yeast, sugar metabolism by, 278–280Baking bread, 282–283Balsamic vinegar, 409�-amylase, 313Barley, 12, 309Barley malt, 340Base wines, 387–388Basidiomycota, 17, 18Beaujolais nouveau, 381Beef fat, 222Beer, 301–347

defects, 333–335enzymatic reactions, 308–316

adjuncts, 313–314malt enzymology, 313overview, 308–313wort, 314–316

fermentation, 319fermentors, 319–320flocculation, 325–326hops, 316–319industry, 303–307

and biotechnology, 340–347recent developments in beer, 336–340waste management in, 335–336

inoculation, 320–323kettle boil, 319manufacturing principles, 307–308overview, 301–303post-fermentation steps, 326–333

carbonation, 331–332clarification, 329–330conditioning, 328–329overview, 326–328packaging and pasteurization, 332–333process aids, 330–331

spoilage and origins of modern science, 303yeast metabolism, 323–325

Bergey’s Manual of Systematic Bacteriology, 17, 27–28,130

Berlin, Irving, 233�-galactosidase, 12, 53, 127, 134�-glucanases, 335, 346Bifidobacterium, 20, 37–39, 104, 109Bifidobacterium adolescentis, 37Bifidobacterium bifidum, 37Bifidobacterium bifidus, 130Bifidobacterium breve, 37Bifidobacterium infantis, 37Bifidobacterium lactis, 37Bifidobacterium longum, 37Binary fission, 321Biogenic amines, 221, 259


Index 459

Biological oxygen demand (BOD), 204Biological preservatives, in dough, 274Biological spoilage, 294Biomarkers, 132Biomass fermentations, 370Biotechnology, and brewing, 340–347

examples of modified brewing strains, 346–347overview, 340–342strain improvement strategies, 343–346

Bitter peptides, 192–193Bloaters, 249Blue-mold ripened cheese, 182–184Bock beer, 336BOD (biological oxygen demand), 204Boerhaave, Hermann, 412Bordeaux wines, 351Botrytis cinerea, 385Bottom-fermenting yeast, 320Bovine milk, 155Brandy, 388–389Bread, 261–299

baking, 282–283biological spoilage, 291–294bread quality, 295–299cooling, 283dividing, 282fermentation, 277–282

end products, 280–282factors affecting growth, 282glycolysis, 280overview, 277–278sourdough fermentation, 285–290sugar metabolism by bakers’ yeast, 278–280sugar transport, 280

flour composition, 265–269carbohydrates, 268–269overview, 265protein, 265–268

history of, 261–263hydration, 277ingredients, 272–277manufacturing principles, 272milling, 263–265mixing, 277overview, 261packaging, 283panning, 282preservation, 294–295proofing, 282rounding, 282staling, 290–291technology, 283–285

Chorleywood process, 285liquid sponge process, 284–285overview, 283–284sponge and dough process, 284straight dough process, 284

wheat chemistry, 263–265yeast cultures, 269–272

Breast-fed infants, 110Brettanomyces, 334, 378, 379, 393, 394Brettanomyces bruxellensis, 393Brevibacillus, 130Brevibacterium, 20, 39Brevibacterium linens, 39, 75, 181, 195Brewing. See BeerBrie, 184, 185Brine, 233, 246Brining, 175Brix, 357Brochothrix, 216Brochothrix thermosphacta, 231Bromates, 274Brown Ale, 336B. thermosphacta, 232Bulk culture tank, 196–197Bulk fermentation, 282, 284Bulk starter cultures, 91–93Buttermilk. See Cultured buttermilkButts, 386

Cabbage, 233Cabernet Sauvignon grapes, 352Calcium carbonate, 370Calcium-induced coagulation, 150Calcium phosphate, 179Calcium propionate, 275, 294Calf chymosin, 152California-style olives, 257–258Camembert, 184, 185Candida, 235, 393, 439Candida milleri, 288, 289Candida versitalis, 426CAR1 gene, 297Carbohydrates, 278, 335, 338, 419

in beer, 314in flour, 268–269

Carbon aldehyde, 119Carbonation, in beer, 331–332Carbon dioxide, 175, 181, 317, 420Carbon dioxide-evolving fermentation, 79Carbon-utilization pathways, 64Carboxylic acid, 358Carlin, George, 121Carlsberg Brewery, 69Carnobacterium, 23Casein, 58, 113, 149, 150, 170, 189, 191Caseinate salts, 149Casings, 225Catabolic enzymes, 297Catabolic pathways, 278Catabolite control protein A (CcpA ), 21, 56,

432Catabolite repression element (CRE), 56, 324, 432


460 Index

CcpA (Catabolite control protein A), 21, 56, 432cDNA (complimentary DNA), 153Cell-containing gel beads, 103Cell densities, 223Cellulose filters, 330Champagne, 353, 370, 386, 387–388Charmat method, 387Cheddar cheese family, 169–172Cheddar-like cheeses, 94Cheese, 145–205

blue-mold ripened cheese, 182–184curds, 148–149Dutch-type cheeses, 181factories, 9hard Italian cheese, 180–181making, 155–166

aging, 165–166coagulation, 163–164curd handling, 164–165cutting and cooking, 164milk, 155–161overview, 155salting, 165starter cultures, 161–163

manufacturing principles, 147–155converting liquid into solid, 148–154overview, 147–148squeezing out water, 154–155

microbial defects, preservation, and food safety, 202–204

mold-ripened cheese, 182Mozzarella and pasta filata, 176–180overview, 145–147pickled, 185–186process and cold pack cheese, 186–187ripening, 187–190

overview, 187–188proteolysis in cheese, 188–190

surface-ripened by bacteria, 181–182Swiss, 172–176technology, 190–202

bacteriophages, 194–202bitterness and accelerated ripening, 191–194overview, 190–191

types of, 166–172acid-coagulated cheeses, 166–169cheddar family, 169–172overview, 166

whey utilization, 204–205white-mold ripened cheese, 184–185

Chemical oxidation reaction, 257Chemical pasteurization, 247Chillproofing agents, 331Chlorinated polyphenols, 391Chorleywood process, 285Chymosin, 150–151, 185Chytridiomycota, 17

CitP (citrate permease), 60, 137Citrate fermentation, 60–61, 137, 181Citrate permease (CitP), 60, 137Citrate salts, 93–94Citrate-to-diacetyl pathway, 61Citric acid cycle enzymes, 415Citrobacter, 334Citrobacter freundii, 442Cladosporium, 252Clarification, in beer-making, 329–330Classification, microbial. See Microbial classificationClostridium, 257, 358, 398Clostridium botulinum, 31, 210, 211, 219, 223, 224, 429,

440Clostridium butyricum, 259Clostridium sporogenes, 259Clostridium tyrobutyricum, 202Coagulation, in cheese-making process, 163–164Cognac, 388Colby cheese, 171Cold pack cheese, 186–187Cold-pasteurized beer, 330Complimentary DNA (cDNA), 153Compressed yeast, 269Computer-generated meteorological models, 352Concord grapes, 351Conditioning, in beer-making, 328–329Conidia, 17Consejo Superior de Investigaciones Cientificas, 128Cooking

cheese, 164sausage, 229–230

Coppers, 310Corks, 389, 390, 391Coronary disease, 360Corynebacterium, 182Coryneform bacteria, 39Cottage cheese, 166, 168Crabtree effect, 376CRE (catabolite repression element), 56, 324, 432Cream Ale, 337Cream cheese, 166, 168–169Crude hybrids, 263Cryotolerance, 87Cucumbers, 245, 246Cultured butter, 69–70Cultured buttermilk, 135

factors affecting diacetyl formation in, 138manufacturing of, 135–138

Cultured dairy products, 107–144consumption of, 107–109cultured buttermilk, 135

factors affecting diacetyl formation in, 138manufacturing of, 135–138

defects, 121–127fermentation principles, 109–114kefir manufacturing, 140–141


Index 461

other cultured dairy products, 141–144overview, 107and probiotic bacteria, 108–109sour cream, 138–140yogurt

flavor and texture, 119–121frozen, 134–135manufacturing of, 114–119styles, 127–134

Culture inoculum, 166, 372Culture production, 370Curds, 164–165, 168Curing agents, in fermented meats, 223C.versatilis, 434Cyd genes, 22Cylindrical fermentation vessels, 412Cylindroconical fermentation vessels, 320Cylindroconical fermentors, 320Cytoplasmic membranes, 50, 275, 378, 403

Dairy products. See Cultured dairy productsDairy starter cultures, 70Dark beers, 311Dark malts, 311Debaryomyces, 334Debaryomyces hansenii, 230De-bittering of cheese, 195Decarboxylase enzymes, 359Decarboxylating enzymes, 221Decarboxylation, 250Decoction heating, 312Decoction mashing, 312Dehydrogenases, 406Dekkera, 378, 379Dekkera bruxellensis, 379Demineralization, 397De novo synthesis, 215Deuteromycetes, 17Deuteromycota, 17Dextrins, 346DHAP (dihdyroxyacetone phosphate), 341Diacetyl, 136, 141, 334Diacetilactis, 93Diacetyl formation, in cultured buttermilk, 138Diacetyl rest, 325Diacetyl synthesis, 138Diastase, 310Diastaticus, 333Diastatic yeasts, 333Dihdyroxyacetone phosphate (DHAP), 341Di Origine Controllate (DOC), 410Disgorgement, 388Distillation, 388Djordjevic, G.M., 200DNA-DNA hybridization, 19DNA microarray technology, 19DNA residues, 373

DNA-RNA hybridization, 19DOC (di Origine Controllate), 410Douglas, Loudon, 107Dried malt, 310Dry acid, 170Dry active yeast, 272Dry beer, 339Dry curd, 168Dunkel beer, 337Dutch-type cheeses, 181

Economic value, 13Eden, Karl J., 315Egyptian pottery vessels, 4Elastic dough, 274Elenoic acid glucoside, 255Embden-Meyerhoff (EM) pathway, 45, 280, 375Embden-Meyerhoff-Parnas (EMP) pathway, 279, 323,

325Emulsification, 187, 277Emulsifiers, 277Emulsifying salts, 186Encapsulated cells, 101–104Endogenous flora, 182Endogenous lactic acid bacteria, 285Endogenous microbial population, 255Endogenous yeasts, 443Endopeptidases, 59Endosperm, 265Endothelin-1 (ET-1), 361Enterobacter, 235, 238, 242, 439, 440Enterobacteriaceae, 221, 334, 441Enterococcus, 23, 26, 33, 78, 235, 240Enterococcus solitarius, 33Enterotoxins, 442Enzymatic activities, 422Enzymatic hydrolysis, 441Enzymatic reactions, in beer, 308–316

adjuncts, 313–314malt enzymology, 313overview, 308–313wort, 314–316

Enzyme-catalyzed reactions, 194Enzyme chymosin, 69, 149Enzyme hydrolyzes, 149Enzyme nitrate reductase, 213–214, 218Enzymes, 166, 272–273Enzymes hydrolyze proteins, 331Enzyme synthesis, 308Epigenetic silencing, 328Epiphytic flora, 235Epiphytic yeasts, 367EPS (exopolysaccharides), 121epsD gene, 125Ergosterol, 323Erwinia, 238Escherichia, 235


462 Index

Escherichia coli, 71, 104, 130, 153, 160, 219, 220, 235, 238,358, 404

ET-1 (endothelin-1), 361Ethanol, 283, 317, 339–340, 341, 371, 372, 375, 398, 401,

407, 414Ethyl carbamate, 359Eubacterium, 398Eukarya, 16, 17Eukaryotic organisms, 321Exogenous enzymes, 443Exons, 152Exopeptidases, 59Exopolysaccharides (EPS), 121Extracellular invertase, 325Extrachromosomal element, 321

Facultative yeasts, 257Fass, 310Fat

in beef, 222in cheese, 171in dough, 274

Fatty acids, 193FDP (fructose-1,6-bisphosphate), 375Fermentable carbohydrate, 324Fermentable sugars, 291, 331Fermentation acids, 231Fermentation-derived vinegar, 415Fermentation temperature, 304Ferridoxin, 224Feta cheese, 185Filtration, 330, 380Firmicutes, 19, 22Fisher, M.F.K., 261Fish-eyes, 258Fish sauces, 447Flash pasteurization, 332–333, 427Flavi, 450Flavobacterium, 235, 242, 248Flavonoids, 363Flavor-generating substrate, 193Flavor-producing bacteria, 135Floaters, 249Flocculation, 91, 325–326, 330Flocs sediment, 320Flour, 12, 265–269

carbohydrates, 268–269overview, 265protein, 265–268

Food poisoning, 11, 221Food safety

cheese, 202–204fungal fermented foods, 450–455tempeh, 441–443

Fortified wines, 385–386FOS (fructooligosaccharide) molecules, 112Free fatty acids, 183Free sugars, 273

French bread, 297French Paradox, 361Frings generator, 413Frozen cultures, 223Frozen dough, 295Frozen liquid cultures, 86Fructooligosaccharide (FOS) molecules, 112Fructose-1,6-bisphosphate (FDP), 375Fungal cultures, 76Fungal fermented foods, safety of, 450–455Fungal spores, 293Fungi, 17–18, 389–392Furanone, 429Fusarium, 252, 258, 333

G-3-P (glyceraldehyde-3-phosphate), 341Galactooligosaccharides (GOS), 112Galactose-6-phosphate, 53, 118, 174Galactose efflux, 119Galactose metabolism, 174Galactose phosphorylation, 174Galactosyltransferase, 112Garolla crusher, 366Gaulle, Charles De, 155Gene cluster, 124Gene expression patterns, 345Generally Regarded As Safe (GRAS), 78Genetically engineered organisms, 102Genetically modified organisms (GMOs), 101, 102, 103Genomes, 63, 343Geotrichum candidum, 141German beer industry, 304German Purity Law, 315Gliadin, 268Glomeromycota, 18Glucoamylase genes, 346Glucokinases, 280Gluconacetobacter europaeus, 399Gluconacetobacter xylinus, 399Gluconoacetobacter, 37, 393, 399Gluconoacetobacter europaeus, 37Gluconoacetobacter xylinus, 37Gluconobacter, 37, 334, 393, 394, 399, 401, 406, 407, 410Gluconobacter entanii, 399Gluconobacter oxydans, 393, 399Glucose, 53, 57, 222, 268, 272, 323, 324Glucoside ester, 253Glucosidic phenols, 254Glutamic acid, 423, 425Gluten, 277Glutenin, 268Glyceraldehyde-3-phosphate (G-3-P), 341Glyceraldehyde-3-phosphate dehydrogenase (GPD), 341Glycerol, 341Glycerol-3-phosphatase (GPP), 341Glycerol-3-phosphate, 376Glycolysis, in bread fermentation, 280Glycolytic Embden-Meyerhoff pathway, 45


Index 463

Glycolytic flux, 49Glycolytic pathway, 53, 375GMOs (genetically modified organisms), 101, 102, 103GMP (guanosine 5’-monophosphate), 430Goat milk, 155Goode, Jamie, 356Gorgonzola, 182GOS (galactooligosaccharides), 112GPD (glyceraldehyde-3-phosphate dehydrogenase), 341GPP (glycerol-3-phosphatase), 341Grape cultivation, 350Grapes, 351–357

crushing, 366domestication of, 350harvesting, 363–366

GRAS (Generally Regarded As Safe), 78Greek-style olives, 257“Green”barley, 309Green beer, 326Grist, 312GroESL, 405Guanosine 5�-monophosphate (GMP), 430G.xylinus, 408

H�-ATPase system, 250H2S (Hydrogen sulfide), 359HAACP (Hazard Analysis Critical Control Points), 203, 224Halobacillus, 449Hansen, Christian Ditlev Ammentorp, 69Hansen, Emil Christian, 69, 303Hansenula, 235, 378, 393, 436Hard Italian cheese, 180–181Hazard Analysis Critical Control Points (HAACP), 203, 224Haze-forming proteins, 331HDL (high-density lipoprotein) cholesterol, 361Hemagglutenins, 443Heme iron, 231Hemiascomycetes, 17Hesseltine, C.W., 437Heterofermentation, 45–48Heterofermentative leuconostocs, 241Heterofermentative phosphoketolase pathway, 289Heterofermentative products, 212Heteropolysaccharides, 122Hexokinases, 280, 289Hexoses, 53Hibiscus plant, 440High beers, 304High-density lipoprotein (HDL) cholesterol, 361Highgravity fermentations, 340High krausen, 325High-moisture cheeses, 164Histamine, 221Homofermentation, 45Homogenization, 169Homogenous fermentation, 435Homulus lupulus, 316Hops, 301, 316–319, 336

HorA system, 317HSP30 gene, 329Husk degradation, 312Hybrid corks, 392Hydraulic filter presses, 427Hydrogenase, 224Hydrogenated vegetable oils, 274Hydrogen peroxide, 163, 212, 215, 231Hydrogen sulfide (H2S), 359Hydrolysis, 293Hydrolysis products, 231Hydrolytic enzymes, 296, 366Hydrolyze dextrins, 339Hydrolyze milk proteins, 97Hydrolyzing enzymes, 118Hydrophilic polysaccharides, 120Hydroxyl radicals, 231Hydroxytyrosol, 255

IBU (International Bitterness Units), 316Ice beer, 339Immobilized cells, starter cultures, 101–104Immobilized lactic acid bacteria, 104Immunoglobulins, 163IMP (inosine 5�-monophosphate), 430Indigenous microflora, 367Industry

beer, 303–307brewing

and biotechnology, 340–347recent developments in, 336–340waste management in, 335–336

fermented foods, 10–11meat fermentations, 208–210starter cultures, 105–106

Inoculum, 211Inosine 5’-monophosphate (IMP), 430International Bitterness Units (IBU), 316Intracellular metabolism, 325Introns, 152Ionic compounds, 250Iso-�-acids, 317Isocohumulone, 316Isohumulone, 316Isomerized extracts, 319Isomerized iso-�-acids, 335Italian cheese

hard, 180–181Mozzarella, 176–180

Italian-type cheeses, 145

James,T.C., 344

Kappa casein, 149Kefir, 104, 140–141, 142Kegging, 332Kegs, 332Kettle boil, in beer-making, 319


464 Index

Killer toxins, 378–379Kilning, 333Kimchi, 234, 242–244Klaenhammer group, 199Klatsky, 360Klebsiella, 235, 334, 439Klebsiella pneumoniae, 441, 442Kloeckera, 378Kloeckera apiculata, 371, 373, 393Kluyveromyces, 41, 333, 378Kluyveromyces lactis, 154Kluyveromyces wickerhamii, 379Kocuria, 20, 39, 213, 214, 218Kocuria varians, 39Koji, 423, 433–434

microorganisms, 422overview, 421raw materials preparation, 421

Koji-rice-water, 445Krasner, R.I., 304Krausen, 325Krausening, 310, 328

LAB exopolysaccharide biosynthesis, 124Lacobacillus sakei, 216Lac operon, 57LacR, 56–57Lactate, 62Lactate dehydrogenase, 49, 138Lactate salts, 93Lactic acid bacteria, 20–35, 98, 99, 113, 122, 188, 212, 215,

233, 235, 250, 285, 382, 393, 394genera of, 22–25Lactobacillus, 33–35Lactococcus, 25–26Leuconostoc, 29–30Oenococcus, 30–31overview, 20–22Pediococcus, 31–32Streptococcus, 26–28sugar transport by, 48–53

overview, 48–49phosphoenolpyruvate-dependent

phosphotransferase system, 49–53symport and ABC transport systems in, 53–56Tetragenococcus, 32–33

Lactic acid fermentation, 235–236Lactic acid starter cultures, 99Lactic culture, 114Lactic fermentation, 235, 286Lactobacillales, 22Lactobacilli, 203, 241, 334Lactobacillus, 33–35, 37, 74, 84, 104, 122, 131, 142, 190,

193, 194, 211–212, 218, 235, 239, 334, 382, 394,395, 446

Lactobacillus acidophilus, 35, 86, 88, 109, 130Lactobacillus brevis, 35, 240, 247, 257, 288, 317, 334, 394

Lactobacillus buchneri, 203Lactobacillus bulgaricus, 86, 135Lactobacillus casei, 35, 104, 109, 128, 190, 194, 257, 334,

439Lactobacillus causasicus, 141Lactobacillus curvatus, 202, 212Lactobacillus cuvatus, 240Lactobacillus delbrueckii subsp. bulgaricus, 35, 45, 55,

84, 108, 110, 114, 116, 117–118, 119, 121, 122, 128,133–134

Lactobacillus fermentum, 202, 257Lactobacillus helveticus, 35, 45, 86, 92, 104, 122, 162, 173,

177, 178, 194Lactobacillus kefir, 141, 142Lactobacillus kefiranofaciens, 142Lactobacillus kefirgranum, 142Lactobacillus paracasei, 190, 194Lactobacillus parakefir, 142Lactobacillus-Pediococcus, 23, 35Lactobacillus plantarum, 35, 190, 212, 215, 239, 240, 242,

244, 247, 248, 250, 251, 257, 287, 334Lactobacillus reuteri, 122Lactobacillus sake, 212, 231, 446Lactobacillus sakei, 215Lactobacillus sakei subsp.sakei, 35Lactobacillus sanfranciscensis, 35, 45, 75, 122, 287, 288,

289Lactobacillus sanfrancisco, 288Lactobacillus sporogenes, 130Lactococcal-derived genes (LlaIR), 200Lactococcal peptidases, 190Lactococci, 133, 197Lactococcus, 19, 23, 25–26, 74, 142, 166, 193, 194Lactococcus garviae, 25Lactococcus lactis, 18, 19, 21, 45, 52, 63, 74, 77, 78, 82, 89,

91, 103, 122, 135, 178, 183, 200Lactococcus lactis subsp. cremoris, 25–26Lactococcus lactis subsp. cremoris, 27–28, 45, 78, 82Lactococcus lactis subsp. cremoris, 93, 97Lactococcus lactis subsp. cremoris, 161, 169, 192Lactococcus lactis subsp. diacetilactis, 25Lactococcus lactis subsp. hordinae, 25Lactococcus lactis subsp. lactis, 27, 78, 82, 93, 96, 97, 135,

161, 169, 178, 188, 190, 192, 199Lactococcus piscium, 25Lactococcus plantarum, 25, 231Lactococcus raffinolactis, 25Lactoperoxidase reaction, 163Lactose, 12, 55, 127, 165Lactose:galactose exchange reaction, 55Lactose-phosphate, 53Lager beer, 337Lager fermentation vessels, 320Lagering, 310Lager-style beer, 303Lager yeast, 320Lambic beer, 337


Index 465

Lauter tun, 310L.curvatus, 231LDL (low-density lipoprotein), 254, 361Leavened bread, 262, 264Leavening, 262Leloir pathway, 55, 119Leuconostoc, 23, 29–30, 74, 84, 142, 181, 183, 239, 240,

248, 394, 446Leuconostocaceae, 29Leuconostoc cremoris, 135Leuconostoc fallax, 29Leuconostoc kimchii, 29, 45Leuconostoc lactis, 29, 45, 135, 162, 181Leuconostoc mesenteroides , 93, 122, 162, 181, 238, 242,

244, 247, 250, 257, 394Leuconostoc mesenteroides subsp. cremoris, 29, 135Leuconostoc mesenteroides subsp. mesenteroides, 29, 45Levine, Philip, 207Limburger cheese, 181Linearization, 170Lipases, 180Lipid fraction, 188Lipid hydrolysis, 440Lipid phase, 155Lipolytic, 214Liquid fermentations, 103Liquid sponge process, 284–285Listeria, 20, 216, 217Listeria-active non-lanthionine peptides, 215Listeria monocytogenes, 31, 160, 185, 211, 216, 219, 220LlaIR, 200Logarithmic growth phase, 319, 323, 325Low-calorie beer, 337–339Low-density lipoprotein (LDL), 254, 361Low-moisture cheeses, 164Lymphocytes, 128Lyophilization, 86, 87, 212Lysogenic infection, 73Lytic phages, 98, 201

Maceration, 366Maillard reaction, 178, 298, 319Major depressive disorder (MDD), 128Malic acid, 358, 382, 383Malolactic fermentation, 101, 380, 382–384, 386, 387Malt, 313, 336Maltase, 280Malting houses, 308Malt Liquor, 337Maltose, 280, 308, 325Maltotriose hydrolysis, 325Malt-water mixture, 312Mannose, 327Mash, 310Mash liquid, 314Mash off, 310Mash tun, 310

McKay, Larry, 98, 99–100MDD (major depressive disorder), 128MDR (multiple drug resistance) systems, 317Meat fermentations, 207–232

defects and spoilage of fermented meats, 231–232flavor of fermented meats, 230–231industry, 208–210ingredients, 222–224

culture, 223curing agents, 223meat, 222overview, 222salt, 222–223spices, flavoring and other ingredients, 223–224sugar, 222

meat composition, 210Micrococcaceae cultures, 218overview, 207–208principles, 210–211protective properties of cultures, 215–218sausage manufacture, 224–230

cooking, drying, and smoking, 229–230cutting and mixing, 224–225fermentation, 226–229mold-ripening, 230overview, 224principles of, 218–221stuffing, 225–226

starter cultures, 211–215Mechanical harvesting, 365Mediterranean diet, 254Mesophilic cultures, 162Mesophilic lactic starter culture, 183Metabolic engineering, 62–66Metabolism. See Microorganisms and metabolismMetabolome, 345Methyl ketones, 187Microbial classification, 15–19

bacteria, 18fungi, 17–18microbial taxonomy and methods of analysis, 19nomenclature, 18–19overview, 15–16three domains of life, 16–17

Microbial cryoprotectants, 295Microbiological quality, 159Microbiological stability, 303Micrococcaceae, 218, 221Micrococcus, 20, 39, 75, 213, 214, 218, 425–426, 449Micrococcus luteus, 39Microflora, 222, 235–236, 367, 450Microorganisms and metabolism, 15–66

Acetobacter, 37Bacillus, 37bacteria used in manufacture of fermented foods, 19–20Bifidobacterium, 37–39Brevibacterium, 39


466 Index

Microorganisms and metabolism, (continued)fermentation and metabolism basics, 43–44Gluconoacetobacter, 37Gluconobacter, 37Kocuria, 39lactic acid bacteria, 20–35

genera of, 22–25Lactobacillus, 33–35Lactococcus, 25–26Leuconostoc, 29–30Oenococcus, 30–31overview, 20–22Pediococcus, 31–32Streptococcus, 26–28Tetragenococcus, 32–33

metabolic engineering, 62–66microbial classification, 15–19

bacteria, 18fungi, 17–18microbial taxonomy and methods of analysis, 19nomenclature, 18–19overview, 15–16three domains of life, 16–17

Micrococcus, 39other metabolic systems, 60–62

citrate fermentation, 60–61metabolism of molds, 62overview, 60propionic acid fermentation, 61–62

overview, 15Propionibacterium, 39–40protein metabolism, 58–60

overview, 58peptidases, 59–60peptide transport systems, 58–59proteinase system, 58

regulation of transport systems, 56–57Staphylococcus, 39sugar metabolism, 44–48

heterofermentation, 45–48homofermentation, 45overview, 44–45by Saccharomyces cerevisiae, 57–58

sugar transport by lactic acid bacteria, 48–53overview, 48–49phosphoenolpyruvate-dependent

phosphotransferase system, 49–53symport and ABC transport systems in lactic acid

bacteria, 53–56yeasts and molds used in manufacture of fermented

foods, 41–43Aspergillus, 43overview, 41Penicillium, 43Saccharomyces, 41–43

Milkin cheese-making process, 155–161

fat in, 191–192in yogurt manufacture, 114

Milk-borne bacteria, 172Milk protein casein, 58Milling, 263–265, 312Miso, 431–436

fermentation, 434–436manufacture of, 434overview, 431–434spoilage and defects, 436

Modern fermented foods industry, 10–11Mold-fermented sausages, 230Mold-ripened cheese, 182–185

blue-mold ripened cheese, 182–184white-mold ripened cheese, 184–185

Molds. See Yeasts and moldsMold starter cultures, 76Molecular archaeological analyses, 4Moles, 280Moniliella acetoabutens, 414Monillia, 390Monosaccharides, 434Moromi enzymology, 424–425Moromi mash, 423Mousy (tourne), 304Mozzarella cheese, 176–180MRNA, 152Mucor, 17, 41, 294, 389, 390Mucor miehei, 152Mucor pusilus, 152Müller-Thurgau, Hermann, 69Multiple drug resistance (MDR) systems, 317Mycotoxigenic aspergilli, 450Mycotoxin formation, 294Myriad enzymes, 309

NAD-dependent decarboxylase, 382NADH-dependent lactate dehydrogenase, 56Natto, 436Natural milk flora, 194Neolithic pottery vessels, 5Neutralizing agents, 93NewFlo phenotype, 328Nitrate-reducing bacteria, 214Nitrate-to-nitrite reaction, 214Nitric oxide (NO), 224, 361Nitrite, 222, 223, 224Nitrite salts, 224Nitrogen, 358, 377, 428Nitrogenous compounds, in wine, 358–359NO (nitric oxide), 224, 361Non-alcoholic beer, 339–340, 341Non-digestible food ingredient(s), 110Non-enzymatic browning, 311Nonfermentable oligosaccharides, 314Non-fermented soy sauce, 430Nonflavonoids, 363


Index 467

Non-homogenized milk, 169Non-lactic microorganisms, 237N-terminal region, 327Nutrition, 12

Oak barrels, 381Oast house, 310Obesumbacterium proteus, 335Oenococcus, 23, 29, 30–31, 382, 383, 394, 395Oenococcus oeni, 29, 45, 75, 104, 382, 394Oiling off, 180Oleuropein, 253, 254Oligopeptide transport system (OPP), 58–59, 190Oligosaccharide-multiple sugar transport system, 56Olives, 253–259

composition, 253–254defects and spoilage, 258–259Greek-style, 257manufacture of fermented olives, 254–255overview, 253ripe- or California-style, 257–258Spanish-style, 255–257

OmtB gene, 453Open vat process, of vinegar-making, 409–412OPP (oligopeptide transport system), 58–59, 190Optimum growth temperatures, 31OpuABCD operon, 251Organic acids, 358, 382Organic nitrogen, 359Organoleptic properties, 12, 415Oriental foods, 419–455

fermented fish-type foods, 447–450fish sauce microbiology, 449–450manufacture of fish sauces and pastes, 447–449overview, 447

history, 419–420koji and tane koji manufacture, 421–422

microorganisms, 422overview, 421raw materials preparation, 421

miso, 431–436fermentation, 434–436manufacture of, 434overview, 431–434spoilage and defects, 436

natto, 436overview, 419plant-based fermentations, 421safety of fungal fermented foods, 450–455sake and rice wines manufacture, 443–447soy sauce manufacturing, 422–431

fermentation, 425–427flavor of soy sauce, 428–430koji, 423mashing, 423–424moromi enzymology, 424–425non-fermented soy sauce, 430

overview, 422–423pasteurization and packaging, 427–428pressing and refining, 427product characteristics, 428spoilage and defects, 431

tempeh, 436–443biochemistry, 440–441cultures, 440fermentation, 439inoculation, 439manufacture of, 437microbiology, 439–440nutrition and safety, 441–443overview, 436–437spoilage and defects, 443substrate preparation, 437–439

types of, 420–421Orla-Jensen, S., 145Orla-Jensen treatise, 77Osmophilic wine yeast, 101Osmotic homeostasis, 251Osmotic problems, 404Ovens, 283Over-acidification, 119Oxaloacetate decarboxylase, 137Oxidation, 215, 222, 389, 406, 407, 412, 415Oxidation-reduction potential (Eh), 331Oxidative fermentation, 409Oxidative pathways, 401Oxygen-sparging, 324

Packagingbeer, 332–333bread, 283sauerkraut, 242soy sauce, 427–428

Pale ale, 337Parmesan cheese, 13, 180Parmigiano Reggiano cheese, 81, 162–163, 180Pasta filata cheese, 176–180Pasteur, Louis, 9, 15, 63, 69, 303, 304–305, 324, 350Pasteurization, 121, 159, 436

of beer, 332–333of soy sauce, 427–428

Pasteurized milk, 114, 160Patent flour, 268Pathogenic microorganisms, 229Pathogen Reduction/Hazard Analysis Critical Control

Point system (HAACP), 227–228Pathogens, 11, 160, 210Payne,Alexander, 349PCR (polymerase chain reaction), 321–322PDO (Protected Designation of Origin), 410Pectic enzymes, 370Pectinatus, 335Pectinolytic enzymes, 252Pediococci, 203


468 Index

Pediococcus, 23, 31–32, 33, 45, 84, 122, 212, 215, 218, 235,248, 257, 394, 395

Pediococcus acidilactici, 31, 74, 212, 215, 216, 231, 240,247, 334, 434, 436

Pediococcus cerevisae, 212Pediococcus damnosus, 31, 334Pediococcus halophilus, 33Pediococcus inopinatus, 334Pediococcus pentosaceus, 31, 212, 240Pellets, 319Penicillium, 41, 43, 62, 202, 220, 252, 258, 294, 389, 390Penicillium camemberti, 43, 62, 76, 182, 184, 230Penicillium caseicolum, 184Penicillium chrysogenum, 230Penicillium nalgiovense, 221, 230Penicillium roqueforti, 43, 62, 76, 182, 183, 187Pentose-containing polysaccharides, 286Pentose phosphate pathway, 375, 394Pentose phosphoketolase (PK) pathway, 432PEP (phosphoenolpyruvate)PTS (-dependent

phosphotransferase system), 49, 117–118Peptidases, 59–60Peptide hydrolysis, 189–190Peptide transport systems, 58–59Peptostreptococcus, 399Per (phage-encoded resistance), 200Peroxide-forming oxidation-reduction reactions, 84–85PFGE (pulsed field gel electrophoresis), 19, 322P. freudenreichii, 61P. freudenreichii subsp. shermanii, 61Pgm gene, 125pH, in starter cultures, 93–94Phage-encoded resistance (Per), 200Phage genome, 199–200Phage infections, 95Phage-inhibitory function, 92Phage-resistant cultures, 96Phage-resistant phenotypes, 96, 197Phage-resistant systems, 82Phage titers, 201–202pH electrode, 93Phenol compounds, 253Phenolic compounds, 254, 359Phenols, 359–363, 429Phenotype, 105pH monitoring device, 93Phosphate salts, 93–94Phospho-�-galactosidase, 118Phosphoenolpyruvate (PEP)-dependent

phosphotransferase system (PTS), 49–53, 117–118Phosphoketolase pathway, 394Phosphorylation, 49Phosphotransferase system (PTS), 433Physicochemical properties, 122Phytic acid, 287Pichia, 235, 334, 378, 393, 439Pichia anomala, 379

Pickled cheese, 185–186Pickles, 234, 245–253

defects, 249–253fermentation, 246–249manufacture of fermented pickles, 245–246overview, 245

Pigment extraction, 366Pilsner beer, 312, 337Pimaricin, 183Pip-defective cells, 199Pip gene, 199Pizza, 177–178PK (pentose phosphoketolase) pathway, 432Plant-based fermentations, 421Plasmid-borne genes, 26PMF (proton motive force), 21, 54PMF-mediated citrate permease, 60Poaceae, 263Polymerase chain reaction (PCR), 321–322Polypeptides, 344Polyphasic taxonomy, 19Polyphenol compounds, 253Polyphenolic compounds, 359, 362Polyphenols, 335Polysaccharide-hydrolyzing enzymes, 441Polysaccharides, 203, 334, 359, 434Polyvinylpolypyrrolidone (PVPP), 330Porter beer, 337Port wine, 386Post-fermentation acidification, 90, 119, 121Post-pasteurization contamination, 140Post-processing contamination, 332Potassium chloride, 429PQQ (pyrroloquinoline quinone), 406Preservation, 11–12

of bread, 294–295of cheese, 202–204

Primary dehydrogenases, 403Primary ethanolic fermentation, 323Probiotic activity, 33Probiotic bacteria, 108–109, 128Probiotics, 104–105, 110Process cheese, 186–187Prokaryotic organisms, 321Properties of fermented foods, 11–13

economic value, 13functionality, 12nutrition, 12organoleptic, 12overview, 11preservation, 11–12uniqueness, 12–13

Propionibacteria, 175Propionibacterium, 39–40, 259Propionibacterium acidopropionici, 40Propionibacterium acnes, 40, 257Propionibacterium avidum, 40


Index 469

Propionibacterium freudenreichii, 15Propionibacterium freudenreichii subsp.

freudenreichii, 40Propionibacterium freudenreichii subsp. shermanii,

40, 173, 175, 176Propionibacterium jensenii, 40Propionibacterium shermanni, 75Propionibacterium zeae, 259Propionic acid fermentation, 60, 61–62Protected Designation of Origin (PDO), 410Protein, in flour, 265–268Proteinases, 58, 150, 179, 192Protein cascades, 58Protein hydrolysis, 188, 316, 425Protein metabolism, 58–60

overview, 58peptidases, 59–60peptide transport systems, 58–59proteinase system, 58

Protein synthesis, 345Proteobacteria, 19, 20, 399Proteolysis, 179, 188–190Proteolytic enzymes, 162, 273, 422, 434Proteome maps, 344Proton motive force (PMF), 21, 54Proton-translocating ATPase, 55Provolone cheese, 180Pseudomonas, 168–169, 203, 231, 235, 242, 248,

257PTS (phosphotransferase system), 433Pulsed field gel electrophoresis (PFGE), 19, 322Putative antimutagenic constituents, 244PVPP (polyvinylpolypyrrolidone), 330Pyrroloquinoline quinone (PQQ), 406Pyruvate, 49Pyruvate decarboxylase, 325Pyruvate-ferridoxin oxidoreductase, 224Pyruvate-formate lyase, 49Pyruvate reduction, 137

Quercus suber oak tree, 390

Racking, 380Randomly amplified polymorphic DNA (RAPD),

19rDNA sequences, 373Red miso, 434Red wines, 363, 372Reinheitsgebot, 315Relative humidity (RH), 229Relevant decarboxylases, 259Rennet paste, 185–186Restriction fragment length polymorphism (RFLP), 19, 79,

322Retrogradation, 290RFLP (restriction fragment length polymorphism), 19, 79,

322

RH (relative humidity), 229Rheological properties, 119Rhizopus, 17, 41, 43, 294, 389, 440Rhizopus microsporus, 76Rhizopus microsporus var. chinensis, 440Rhizopus microsporus var. oligosporus, 439Rhizopus oligosporus, 15, 43, 439, 440, 441,

442Rhizopus oryzae, 440Rhizopus stolonifer, 440Rhodotorula, 43, 235, 242Rice miso, 433–434Rice wines, 421, 443–447Ricotta cheese, 204Roller crushers, 366Rootlets, 309

Saccharification, 420, 443Saccharomyces, 41, 42, 57, 104, 235, 288, 323, 333, 343,

371, 378, 410, 439Saccharomyces bayanus, 41, 373Saccharomyces bayanus subsp. uvraum, 369Saccharomyces boulardii, 132Saccharomyces cariocanus, 42Saccharomyces carlsbergensis, 41, 69, 321, 343Saccharomyces cerevisiae, 18, 41, 63, 75, 90, 98, 101, 104,

269, 278, 280, 282, 289, 295, 297, 320, 321, 324,327, 333, 340, 341, 343, 346, 357, 368, 371, 373,375, 376, 388, 445

Saccharomyces cerevisiae var.diastaticus, 346Saccharomyces ellipsoideus, 41Saccharomyces kefir, 142Saccharomyces kudriavzevii, 42Saccharomyces mikatae, 42Saccharomyces paradoxus, 42, 373Saccharomyces pastorianus, 41, 320, 321,

343Saccharomyces uvarum, 41, 321Safety. See Food safetySake, 443–447Salmonella, 218, 219, 358Salmonella typhimurium, 429, 440Salt, 222–223

in cheese, 165, 171in sauerkraut, 236–237

Salt-tolerant yeasts, 237Sarcina sickness, 334Saturated fatty acids, 274Sauerkraut, 236–244

end products, 240–242fermentation, 238–240kimchi, 242–244mixing, 237–238overview, 236packaging and processing, 242shredding and salting, 236–237spoilage and defects, 242


470 Index

Sausage, 209, 224–230cooking, 229–230cutting, 224–225drying, 229–230fermentation, 226–229mixing, 224–225mold-ripening, 230overview, 224principles of, 218–221smoking, 229–230stuffing, 225–226

Schizosaccharomyces pombe, 343Schwann,Theodor, 7, 67Sedimentation, 326Selenomonas, 335Shelf-life, 11–12Sherman scheme, 77Sherry, 386Sherry vinegars, 410Shiu-Ku,Yen, 419Shoyu, 422, 431Sicilian olives, 257Simpson, Homer, 301Single-pass milling, 267Skimming, 326Skunky beer, 335Smoking sausage, 229–230SO2 (sulfur dioxide), 350, 359, 366–370Sodium, 429Soluble nitrogen concentration, 441Sour cream, 138–140Sourdough, 285–290Soy-based fermentations, 420Soybean koji, 421Soybeans, 15, 428, 434Soybean-wheat mixture, 423Soy oligosaccharides, 441Soy proteins, 421Soy sauce, 422–431

defects, 431fermentation, 425–427flavor of, 428–430koji, 423mashing, 423–424moromi enzymology, 424–425non-fermented, 430overview, 422–423packaging, 427–428pasteurization, 427–428pressing, 427product characteristics, 428refining, 427spoilage, 431

Spanish-style olives, 255–257Sparkling wines, 386–387Spectrophotometric methods, 298Spectroscopy techniques, 344Spent filter material, 336

Spent grains, 336Spices, in fermented meats, 223–224Spoilage, 394

fermented meats, 231–232miso, 436olives, 258–259sauerkraut, 242soy sauce, 431tempeh, 443vinegar, 414

Sponge and dough process, 284, 285Spontaneous phage-resistant mutants, 197Sporangia, 443Sporangiospores, 17Sporolactobacillus, 130Sporulating bodies, 294Sprouts, 309Stale beer, 335Staphylococci, 221Staphylococcus, 20, 39, 213, 214, 218, 449Staphylococcus aureus, 31, 211, 218, 219, 429, 440Staphylococcus carnosus, 39, 214Staphylococcus xylosus, 39, 214Starch-degrading enzymes, 308Starter culture-mediated pickle fermentations,

70–71Starter cultures, 67–106

bacteriophages and their control, 94–96in cheese-making process, 161–163composition, 81–84

defined cultures, 81–84mixed or undefined cultures, 81overview, 81

encapsulated and immobilized cells, 101–104engineered phage resistance, 96–98evaluating performance, 90–91

compatibility issues, 91overview, 90–91

history, 68–70how used, 91–94

bulk cultures, 91–93controlling pH, 93–94overview, 91

industry, 105–106manufacture of, 84–90math, 79–80meat fermentations, 211–215microorganisms, 70–79

bacterial starter cultures, 74–75mold starter cultures, 76overview, 70–74strain identification, 76–79yeast starter cultures, 75–76

overview, 67–68probiotics and cultures adjuncts, 104–105propagation environment, 196role of, 68technology in twenty-first century, 98–101


Index 471

Steam beer, 337Steinkraus, Keith, 437Sterols, 323, 344Sticky gluten protein, 274Stilton, 182Stonegrinding, 267Stout beer, 337Straight dough system, 272, 284Strain identification, 76–79Streptococcus, 23, 26–28, 74, 84, 142, 235Streptococcus acidi lactici, 77Streptococcus agalactiae, 78Streptococcus durans, 26Streptococcus faecalis, 26Streptococcus faecium, 26Streptococcus lacticus, 77Streptococcus lactis, 19, 77Streptococcus-Lactococcus, 23Streptococcus mutans, 122Streptococcus pneumonia, 78Streptococcus pyogenes, 78Streptococcus salivarius, 28, 122Streptococcus thermophilus, 26, 28, 64, 71, 74, 78, 84, 90,

108, 110, 114, 116, 117–118, 119, 121, 122, 124,128, 133–134, 135, 162, 173, 177, 178

Stress proteins, 88Stress-response proteins, 405Stuck fermentations, 374Submerged fermentation systems, 413–414Sucrose, 272, 280Sugar

in beer, 325–326in fermented meats, 222metabolism, 44–48, 325

by bakers’ yeast, 278–280heterofermentation, 45–48homofermentation, 45overview, 44–45by Saccharomyces cerevisiae, 57–58

transportin bread fermentation, 280by lactic acid bacteria, 48–53

in wine, 357–358Sulfites, 369Sulfur compounds, in wine, 359Sulfur-containing amino acid cysteine, 277Sulfur dioxide (SO2), 350, 359, 366–370Superoxide, 361Sweet wines, 384–385Swiss cheese, 15, 172–176Symport and ABC transport systems, in lactic acid

bacteria, 53–56Symport system transporters, 54Syneresis, 164

Tane koji, 421–422microorganisms, 422overview, 421

raw materials preparation, 421Tannins, 363Taxonomy, microbial, 19Taylor, Jim, 349TCA (tricarboxylic acid) cycle, 324, 341, 375TCA (trichloroanisole), 390Tempeh, 436–443

biochemistry, 440–441cultures, 440fermentation, 439inoculation, 439manufacture of, 437microbiology, 439–440nutrition and safety, 441–443overview, 436–437spoilage and defects, 443substrate preparation, 437–439

Temperature gradient, 283Tetragenococcus, 23, 32–33, 45Tetragenococcus halophilus, 33, 426, 427, 431,

434Tetragenococcus muriaticus, 33Tetragenococcus solitarius, 33Thamnidium elegans, 230Thermization, 160–161Thermophilic bacteria, 374Thermophilic cultures, 162, 173Thermophilic sporeforming bacteria, 121Thermophilic starter culture bacteria, 177Thermophilus, 26Thiamine pyrophosphate (TPP)-dependent pyruvate

decarboxylase, 60, 137Three domains of life, 16–17Titratable acidity, 117Torulopsis holmii, 288Tourne (mousy), 304Toxicity, 404Toxin deoxynivalenol, 333TPP (thiamine pyrophosphate)-dependent pyruvate

decarboxylase, 60, 137Transmembrane electric charge, 50Trehalose concentration in yeasts, 295Tricarboxylic acid (TCA) cycle, 324, 341, 375Trichinae, 229Trichinella-free pork, 220Trichinella spiralis, 220Trichloroanisole (TCA), 390Trichoderma, 390Trichoderma longibrachiatum, 391Trickling generator processes, of vinegar-making,

412–413Trimannoside oligosaccharides, 327Triticum aestivum, 263Trub, 310, 319TTB (Alcohol and Tobacco Tax and Trade Bureau),

354Tunnel pasteurization systems, 333Tyramine, 221


472 Index

Umami, 430Uniqueness, 12–13Unleavened bread, 9Unripened cheese, 158

Vagococcus, 23, 25Van Leeuwenhoek,Antonie, 6Vegetable fermentation, 233–259

and biogenic amines, 259olives, 253–259

composition, 253–254defects and spoilage, 258–259Greek-style, 257manufacture of fermented olives, 254–255overview, 253ripe- or California-style, 257–258Spanish-style, 255–257

overview, 233–234pickles, 245–253

defects, 249–253fermentation, 246–249manufacture of, 245–246overview, 245

production principles, 234–236products and consumption, 234sauerkraut manufacture, 236–244

end products, 240–242fermentation, 238–240kimchi, 242–244mixing, 237–238overview, 236packaging and processing, 242shredding and salting, 236–237spoilage and defects, 242

Vegetative cell growth, 294Vibrio cholera, 71Vin aigre, 397Vinegar, 75, 397–417

definitions, 398history, 397–398manufacturing principles, 398metabolism and fermentation, 401–409microorganisms, 398–401overview, 397quality, 415–417technology, 409–415

bacteriophages, 414–415open vat process, 409–412overview, 409post-fermentation processing, 414spoilage, 414submerged fermentation, 413–414trickling generator processes, 412–413

Vinyl guaiacol, 340Viruses, 199, 220Viscoelastic dough, 286Vitamins, in dough, 274

Viticulture, 351–357Vitis labrusca, 351Vitis vinifera, 351

Wang, H.L., 437Waste management, in brewing industry, 335–336Water phase, of cheese, 171Weissella, 25, 29, 239Wheat beer, 340Wheat chemistry, in bread, 263–265Wheat kernels, 263–265Wheat milling, 267Whey, 168, 204–205White-mold ripened cheese, 184–185White wine, 363Whole grain peasant breads, 297Wild flora, 371Wine, 349–395

adjustments after fermentation, 378–380aging, 380–382blending, 380clarification, 380consumption, 350crushing and maceration, 366defects, 389fermentation, 372–375grape composition, 357harvesting and preparing grapes, 363–366history, 349–350malolactic fermentation, 382–384manufacturing principles, 363microbial ecology and spontaneous wine

fermentations, 371–372nitrogen metabolism, 377nitrogenous compounds, 358–359organic acids, 358other pre-treatments, 370–371overview, 349phenols, 359–363pigments, 363polysaccharides, 359pressing, 372production, 350separation, 372spoilage, 389–395

by bacteria, 393–395by fungi, 389–392overview, 389by yeasts, 393

stuck fermentations, 377–378sugars, 357–358sulfur compounds, 359sulfur dioxide treatment, 366–370sulfur metabolism, 377tannins, 363types of, 384–389

brandy, 388–389


Index 473

Champagne, 387–388fortified wines, 385–386overview, 384sparkling wines, 386–387sweet wines, 384–385

viticulture and grape science, 351–357yeast metabolism, 375–377

Woese, Carl, 16Wort, 308, 314–316, 319

Xylan polysaccharides, 431xyl genes coding, 432Xylose-fermenting phenotype, 433xylR gene, 433xynF1 gene encoding, 431

Yeast, The, 41Yeasts and molds, 41–43

Aspergillus, 43metabolism of mold, 62overview, 41Penicillium, 43Saccharomyces, 41–43wine spoilage by yeast, 393yeast cultures, 269–272yeast metabolism

in beer, 323–325in wine, 375–377

yeast starter cultures, 75–76, 378Yersinia enterocolitica, 429, 440Yogurt, 12, 108

flavor and texture, 119–121frozen, and other yogurt products, 134–135manufacture, 114–119manufacturing of

culture metabolism, 117–119milk treatment, 114overview, 114post-fermentation, 119yogurt culture, 114–117yogurt styles, 114

styles, 127–134

Zapatera spoilage, 258Zwickelbier beer, 337Zygomycetes, 17Zygomycota, 17, 41Zygosaccharomyces, 41, 333, 410Zygosaccharomyces bailii, 393Zygosaccharomyces rouxii, 426, 434, 436Zymomonas, 334Zymophilus, 335


microbiology and technology of - wordpress.com · 2009-03-23 · fermented foods,by a.h. rose, was...

Documents