protein purification...protein purification principles, high resolution methods, and applications...

PROTEIN PURIFICATION

Principles, High Resolution Methods, and Applications

Third Edition

Edited by

JAN-CHRISTER JANSON

Innodata9780470940075.jpg9780470940075.jpg


Principles, High Resolution Methods, and Applications

Third Edition

Edited by

JAN-CHRISTER JANSON

Copyright # 2011 by John Wiley & Sons, Inc. All rights reserved

Published by John Wiley & Sons, Inc., Hoboken, New JerseyPublished simultaneously in Canada

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying,recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written per-mission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 RosewoodDrive, Danvers,MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to thePermissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representationsor warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability orfitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies containedherein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable forany loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at(800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002.

Wiley also publishes its books in variety of electronic formats. Some content that appears in print may not be available in electronic format. For more informationabout Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data:

Protein purification : principles, high resolution methods, and applications / edited by Jan-Christer Janson. — 3rd ed.p. cm.

Includes index.ISBN 978-0-471-74661-4 (cloth)1. Proteins–Purification. 2. Chromatographic analysis. 3. Electrophoresis. I. Janson, Jan-Christer.QP551.P69754 20115720.6—dc22

2010033316

Printed in the United States of America10 9 8 7 6 5 4 3 2 1

http://www.copyright.comhttp://www.wiley.com/go/permissionhttp://www.wiley.com

CONTENTS

PREFACE TO THE THIRD EDITION vii

PREFACE TO THE SECOND EDITION ix

PREFACE TO THE FIRST EDITION xi

CONTRIBUTORS xiii

PART I INTRODUCTION 1

1 Introduction to Protein Purification 3Bo Ersson, Lars Rydén, and Jan-Christer Janson

PART II CHROMATOGRAPHY 23

2 Introduction to Chromatography 25Jan-Christer Janson and Jan Åke Jönsson

3 Gel Filtration: Size Exclusion Chromatography 51Lars Hagel

4 Ion Exchange Chromatography 93Evert Karlsson and Irwin Hirsh

5 High-Resolution Reversed-Phase Chromatography of Proteins 135Sylvia Winkel Pettersson

6 Hydrophobic Interaction Chromatography 165Kjell-Ove Eriksson and Makonnen Belew

7 Immobilized Metal Ion Affinity Chromatography 183Lennart Kågedal

8 Covalent Chromatography 203Francisco Batista-Viera, Lars Rydén, and Jan Carlsson

9 Affinity Chromatography 221Francisco Batista-Viera, Jan-Christer Janson, and Jan Carlsson

v

10 Affinity Ligands from Chemical Combinatorial Libraries 259Enrique Carredano and Herbert Baumann

11 Affinity Ligands from Biological Combinatorial Libraries 269Per-Åke Nygren

PART III OTHER SEPARATION METHODS AND RELATED TECHNIQUES 279

12 Membrane Separations 281Joachim K. Walter, Zuwei Jin, Maik W. Jornitz, and Uwe Gottschalk

13 Refolding of Inclusion Body Proteins from E. coli 319Zhiguo Su, Diannan Lu, and Zheng Liu

14 Purification of PEGylated Proteins 339Conan J. Fee and James M. Van Alstine

PART IV ELECTROPHORESIS 363

15 Electrophoresis in Gels 365Reiner Westermeier

16 Conventional Isoelectric Focusing in Gel Slabs and Capillaries and Immobilized pH Gradients 379Pier Giorgio Righetti, Elisa Fasoli, and Sabina Carla Righetti

17 Two-Dimensional Electrophoresis in Proteomics 411Reiner Westermeier and Angelika Görg

18 Protein Elution and Blotting Techniques 441Reiner Westermeier

19 Capillary Electrophoretic Separations 451Wolfgang Thormann

PART V SEPARATION METHOD OPTIMIZATION 487

20 High Throughput Screening Techniques in Protein Purification 489Karol M. Lacki and Eggert Brekkan

INDEX 507

vi CONTENTS

PREFACE TO THE THIRD EDITION

Most will agree that the major achievement in biosciencesince 1998, when the second edition of this book was pub-lished, is sequencing of the human genome. Rather thandiminishing interest in proteins, this has led to a revivalin protein exploration and an intensive search for betterunderstanding of molecular processes in health and disease.During this time, industrial exploitation of proteins in health-care has hardly declined. The application of monoclonalantibodies targeted against rheumatoid arthritis and cancerhas been booming, many second- and third-generation bio-pharmaceuticals have been approved, and modern technol-ogies for vaccine production based on protein engineeringand cell culture are being developed on a wide front.

There are approximately 21,000 protein-encoding genes,and the human proteome is much larger than this. Althoughmapping the genome revealed what was in the box, thejigsaw puzzle is far from complete. Several major researchprojects exemplify the revitalized interest in proteins. Oneis the Protein Atlas initiative (www.protematlas.org), aimedat providing a comprehensive database of high resolu-tion microscopic images identifying proteins in normal andcancer tissues. Others involve an ever-widening range ofrefined tools exploiting protein profiling micro arrays, surfaceplasmon resonance, mass spectrometry, ELISA, quantitative2D electrophoresis, and so on. Many technologies are aimedat parallel processing of thousands of targets, and this isprofoundly changing the way structural biology projectsare managed. Streamlined, miniaturized, automated highthroughput (HTP) protocols are becoming the standard, butthere is still a fundamental need for protein expression andpurification, not least for X-ray structural studies. Many“proteomic” projects exploit high throughput purification oftagged proteins or antibodies.

On the industrial side, particular in healthcare, protein pro-duction is rapidly maturing. Platform technologies are beingapplied both upstream and downstream, allowing faster and

leaner implementation as well as better control. Expressionof monoclonal antibodies in mammalian cells is at the multi-gram per liter level, with cell densities of more than twentymillion per milliliter, specific productivity over 20 picogramsper cell per day, in bioreactors with capacities up to 20,000liters. This several-hundred-fold increase in productivityhas changed the pressures on downstream purification,resulting in the development of very high capacity chromato-graphy media for product capture and highly selective media(frequently “multimodal”) for polishing. Downstream puri-fication of biopharmaceuticals uses platform modules forassuring virus safety and for removal of host cell proteins,aggregates and critical contaminants. Regulatory agenciesare encouraging greater understanding and control of pro-duction processes, a quality by design (QbD) doctrine, andthe use of modern risk management techniques and exper-imental design—all of which is impacting the developmentof purification methods.

Compared to the second edition of this book, four chaptershave been deleted (Chromatofocusing, Affinity Partitioning,Immunoelectrophoresis, and Large-Scale Electrophoresis).Three chapters have been totally rewritten by new authors:Chapter 5 (High Resolution Reversed-Phase Liquid Chro-matography of Proteins), Chapter 15 (Electrophoresis inGels), Chapter 16 (Conventional Isoelectric Focusing inGel Slabs and Capillaries and Immobilized pH Gradients).Six new chapters have been added: Chapter 10 (AffinityLigands from Chemical Combinatorial Libraries), Chapter11 (Affinity Ligands from Biological CombinatorialLibraries), Chapter 12 (Membrane Separations), Chapter 13(Refolding of Inclusion Body Proteins from E. coli), Chapter14 (Purification of PEGylated Proteins), and Chapter 20(High Throughput Screening Techniques in Protein Puri-fication). These new chapters have been written by leadingexperts in their respective fields. All other chapters havebeen thoroughly revised and updated regarding recent

vii

applications. A new section on the history of protein chro-matography has been added to Chapter 2 (Introduction toChromatography).

It is my hope that the third edition will receive the sameoverwhelmingly positive response as the first and secondeditions, and I would like to express my appreciation to

all contributing authors and to Ms Anita Lekhwaniand her staff at John Wiley & Sons, Inc., Hoboken, NewJersey, for their patience and never-failing support of thisproject.

JAN-CHRISTER JANSON

viii PREFACE TO THE THIRD EDITION

PREFACE TO THE SECOND EDITION

Since 1989, when the first edition of this book was launched,the development of biosciences has meant a revival of proteinchemistry in the wake of the molecular biology revolutionand the HUGO project. The total genome of baker’s yeastis now sequenced, that of E. coli is not far behind, andwithin a not too distant future the feat of the total mappingof the human genome, which at the beginning seemed ficti-tious, is now within reach. This means that the attention ofthe world’s bioscientific community will again, as in the1960s and most of the 1970s, focus on the structure and func-tion of the proteins. The PROTEOME era has thus begun, andwith it follows the need of more efficient and more selectivetools for the separation, isolation, and purification of the geneproducts, the proteins.

The development of new chromatographic separationmedia since 1989 has mainly been focused toward improve-ments demanded primarily by process development engineersin the biopharmaceutical industry. This has resulted in mediawith higher efficiencies, leading to shorter cycle times, primar-ily based on suspension polymerized styrene-divinylbenzenepolymers with optimized internal pore size distributions,some allowing partial convective flow through the particles.This trend has received its ultimate solution in totally perfusivesystems based on stacked membranes, or continuous “mono-lithic” columns made of cross-linked polymers, derivatizedwith various kinds of protein adsorptive groups. New compo-site media have been introduced primarily to increase theindustrial applicability of size exclusion chromatography ofproteins but also to increase binding capacity in, for example,ion exchange chromatography. The concept of “solid diffu-sion” in highly ionic group substituted composite media isstill awaiting its physicochemical explanation.

The demand for systems allowing direct capture of targetproteins directly from whole cultures or cell homogenates,resulting in fewer process steps and concomitantly higheryields, has led to a revival of the fluidized bed concept.However, now optimized with regard to the design of bothmedia and columns by the introduction of the more efficientone cycle technique called expanded bed adsorption.

As long as scientists have been engaged in the isolationand purification of proteins from crude extracts, there hasbeen a demand for media with higher adsorptive selectivities.The extremely high variability in protein surface structure aswell as their wide range of functional stabilities, makes itnecessary for every protein chemist to have a stock of severaldifferent separation media, ion exchangers, hydrophobicinteraction media, and a variety of general affinity media.Literature survey data presented in some of the chapters ofthis book reveal that on average somewhere between three andfour steps are required to purify a protein to homogeneity. Thehope for one-step purifications raised by the introduction ofimmobilized monoclonal antibodies has not yet been ful-filled. However, there is a renewed opportunity at hand toincrease the selectivity of immobilized ligands in affinitychromatography and thus decrease the number of steps inthe purification process. This opportunity has been raisedby the recent rapid development in the design of a largevariety of chemical and biological combinatorial librariesand high-speed screening technologies. It is easy to predictthat over the next few years there will be an unprecedentednumber of new highly selective ligands, monospecific aswell as group specific, introduced for the synthesis of newprotein separation media.

Compared to the first edition of this book, there exists oneadditional chapter (Chapter 18) on large-scale electrophoreticprocesses. Three chapters (Chapters 15, 16, and 17) havebeen totally rewritten. Chapters 15 and 16 by new authors.Most other chapters have been thoroughly revised, and allhave been updated regarding recent applications.

It is our hope that this new edition will receive the sameoverwhelmingly positive response as the first edition, andwe would like to express our appreciation to Dr. EdmundH. Immergut and the staff of VCH Publishers, now JohnWiley & Sons, Inc., for their patience and never-failing sup-port of this project.

JAN-CHRISTER JANSONLARS RYDÉN

ix

PREFACE TO THE FIRST EDITION

Over the last two decades the scientific community haswitnessed an unprecedented expansion within the biosciencesand biotechnology. This expansion has been to a large extentdriven by advances in several key areas, most notably recom-binant DNA technology, hybridoma and cell culture tech-niques and, finally, in biochemical separation methods. Thisbook is a description of the current status of one of theseareas:modern techniques for protein purification and analysis.

The research on which the progress in separationtechniques is based has been conducted both in universitydepartments, devoted to basic research, and in industriallaboratories whose main concern is the development of newequipment and tools. In many cases the two communitieshave cooperated to their mutual benefit. In fact, a greatnumber of the products now available for the separationand purification of proteins, such as chromatographic mediawith a wide range of selectivities and efficiencies, as wellas equipment for electrophoretic separation and analysis,were originally developed in a university setting. This bookis also the result of a joint effort between university research-ers, in particular at Uppsala University, and the research staffof a company, Pharmacia LKB Biotechnology. Although it isthus a product of this condition of mutual benefit, the ambi-tion has not been to give a selective description of methodsor materials from a single commerical source, but rather togive an unbiased account of all key techniques in the field.

Today it is to a great extent possible to base the separationof proteins on knowledge of their molecular properties, struc-tural as well as functional. Suggestions on how to solve a sep-aration problem can best be made if data on protein structureand function, including particular structural details, is avail-able. Conversely, results from the application of a particularseparation method can often be interpreted in terms of mol-ecular properties of the protein under study. Throughout thetext of this book, separation results are related to protein prop-erties, often in a detailed manner. We are the first generationto be on the verge of rational protein management.

Starting with this general concept, we have aimed atproviding students, teachers and research workers in bio-medicine, bioscience and biotechnology with a concise andpractical treatise covering, in a single volume, all importantchromatographic and electrophoretic techniques used in pre-parative and analytical protein chemistry. The book containsa general introductory chapter on protein preparative work,Chapter 1, where the key concepts are introduced. Similarly, ageneral introduction to chromatography is given in Chapter 2and an introduction to analytical electrophoresis in Chapter12. The major chromatographic and electrophoretic techni-ques are presented in individual chapters, including one chap-ter on affinity partitioning in aqueous polymer two-phasesystems.

No single person can today be even close to acquiring theamount of experience necessary to describe with confidencethe wealth of techniques and methods which makes up thearsenal for protein separations. We have thus chosen to pro-duce a multi-author volume recruiting expertise from theentire field. All chapters have, however, been thoroughlyworked through by the editors to achieve a reasonable uni-formity of style and organization. Each chapter deals firstwith the theory and underlying principles of each separationtechnique, followed by a section on methodology, and endswith a number of representative application examplesdescribed in detail.

The preparation of this book has been a matter of severalyears. We would like to thank the authors for theircooperation, from the first planning stage to the last phaseof updating and addition. Wewould also like to thank our edi-tors at VCH Publishers in New York, in particular Dr.Edmund H. Immergut who took the first initiative and whofollowed the project up to its realization. The managementand staff of Pharmacia LKB Biotechnology are thanked fortheir cooperation and support which allowed the sellingprice to be considerably reduced. Many staff members havemade invaluable contributions to the final result, which are

xi

gratefully acknowledged. We also thank Elizabeth Hill andUrsula Snow for their contributions in the early phase ofthe project; Inger Galvér, Gull-Maj Hedén, Inga Johanssonand Madeleine de Sharengrad for secretarial work; BengtWesterlund for handling the computer programmes for thechemical structures; Uno Skatt and Lilian Forsberg for produ-cing a number of the illustrations; and David Eaker and JohnBrewer for keeping our freedom with the English languagewithin limits. Finally, we would like to add that we are wellaware that much of our own efforts, occasional achievementsand sometimes hardwon experience, as well as that of several

of the other authors of this book, spring from the tree plantedlong ago by The Svedberg and Arne Tiselius, and later keptalive by Jerker Porath and Stellan Hjertén and many oftheir colleagues and pupils through fifty years of separationscience at Uppsala University. We offer this book as thelatest fruit of this tree, hopefully to be enjoyed by many.

JAN-CHRISTER JANSONLARS RYDÉN

Uppsala, Sweden, June 21, 1989

xii PREFACE TO THE FIRST EDITION

CONTRIBUTORS

Francisco Batista-Viera, Cátedra de Bioquı́mica, Dpto. deBiociencias, Facultad de Quimica Gral. Flores 2124.Casilla de Correo 1157, Montevideo, Uruguay

Herbert Baumann, GE Healthcare Bio-Sciences AB,SE-751 84 Uppsala, Sweden

Makonnen Belew, GE Healthcare Bio-Sciences AB,SE-751 84 Uppsala, Sweden

Eggert Brekkan, GE Healthcare Bio-Sciences AB,SE-751 84 Uppsala, Sweden

Jan Carlsson, Department of Physical & AnalyticalChemistry, Uppsala University, Box 579, SE-751 23Uppsala, Sweden

Enrique Carredano, GE Healthcare Bio-Sciences AB,SE-751 84 Uppsala, Sweden

Kjell-Ove Eriksson, GE Healthcare Bio-Sciences AB,SE-751 84 Uppsala, Sweden

Bo Ersson, Medicago AB, Danmark-Berga 13, SE-755 98Uppsala, Sweden

Elisa Fasoli, Department of Chemistry, Materials andChemical Engineering, “Giulio Natta,” Politecnico diMilano, Via Mancinelli 7, 20131 Milano, Italy

Conan J. Fee, Biomolecular Interaction Centre andDepartment of Chemical and Process Engineering,University of Canterbury, Private Bag 4800, Christchurch8020, New Zealand

Angelika Görg, Department of Proteomics, TechnischeUniversität München, D-85350 Freising-Weihenstephan,Germany

Uwe Gottschalk, Sartorius Stedim Biotech GmbH, August-Spindler-Straße 11, D-37079 Göttingen, Germany

Lars Hagel, GE Healthcare Bio-Sciences AB, SE-751 84Uppsala, Sweden

Irwin Hirsh, Novo Nordisk AS, Nybrovej 80, 2820 Gen-tofte, Denmark

Jan-Christer Janson, Department of Physical and Ana-lytical Chemistry, Uppsala University, Box 579, S-75123 Uppsala, Sweden

Zuwei Jin, GE Healthcare Life Sciences, Building 1, No 1Huatuo Road, Zhangjiang Hi-Tech Park, Pudong NewArea, Shanghai 201203, China

Maik W. Jornitz, Sartorius Stedim North America Inc.,5 Orville Drive, Bohemia, New York 11716, USA

Jan Åke Jönsson, Center for Analysis and Synthesis,Department of Chemistry, Lund University, Box 124,S-22100 Lund, Sweden

Lennart Kågedal, GE Healthcare Bio-Sciences AB,SE-751 82 Uppsala, Sweden

Evert Karlsson, Department of Biochemistry and OrganicChemistry, Uppsala University, Box 576, SE-751 23Uppsala, Sweden

Karol M. Lacki, GE Healthcare Bio-Sciences AB,SE-751 84 Uppsala, Sweden

Zheng Liu, Department of Chemical Engineering, TsinghuaUniversity, Beijing 100084, China

Diannan Lu, Department of Chemical Engineering, Tsin-ghua University, Beijing 100084, China

Per-Åke Nygren, Division of Molecular Biotechnology,School of Biotechnology, Royal Institute of Technology(KTH), SE-106 91 Stockholm, Sweden

Sylvia Winkel Pettersson, Eka Chemicals AB/AkzoNobel, Bohus, Sweden

Pier Giorgio Righetti, Department of Chemistry, Materialsand Chemical Engineering, “Giulio Natta,” Politecnico diMilano, Via Mancinelli 7, 20131 Milano, Italy

xiii

Sabina Carla Righetti, Department of Chemistry, Materialsand Chemical Engineering, “Giulio Natta,” Politecnico diMilano, Via Mancinelli 7, 20131 Milano, Italy

Lars Rydén, Centre for Sustainable Development (CSD)Uppsala, Uppsala University, Villavägen 16, SE-752 36Uppsala, Sweden

Zhiguo Su, State Key Laboratory of Biochemical Engineer-ing, Institute of Process Engineering, Chinese Academy ofSciences, Beijing 100080, China

Wolfgang Thormann, Department of Clinical Pharma-cology, University of Bern, Murtenstraße 35, CH-3010Bern, Switzerland

James M. Van Alstine, GE Healthcare Bio-Sciences AB,751 84 Uppsala, Sweden

Joachim K. Walter, InnoBiologics Sdn Bhd, Lot 1,Persiaran Negeri BBN, 71800 Nilai, Malaysia

Reiner Westermeier, SERVA Electrophoresis GmbH,Carl-Benz-Strasse 7, D-69115 Heidelberg, Germany

xiv CONTRIBUTORS

PART I

INTRODUCTION

1INTRODUCTION TO PROTEIN PURIFICATION

BO ERSSONMedicago AB, Danmark-Berga 13, SE-755 98 Uppsala, Sweden

LARS RYDÉNCentre for Sustainable Development (CSD) Uppsala, Uppsala University, Villavägen 16, SE-752 36 Uppsala, Sweden

JAN-CHRISTER JANSONDepartment of Physical and Analytical Chemistry, Uppsala University, Box 579, S-751 23 Uppsala, Sweden

1.1 Introduction 4

1.2 The Protein Extract 41.2.1 Choice of Raw Material 41.2.2 Extraction Methods 51.2.3 Extraction Medium 5

1.2.3.1 pH 6

1.2.3.2 Buffer Salts 6

1.2.3.3 Detergents and Chaotropic Agents 6

1.2.3.4 Reducing Agents 6

1.2.3.5 Chelators or Metal Ions 6

1.2.3.6 Proteolytic Inhibitors 7

1.2.3.7 Bacteriostatics 7

1.3 An Overview of Fractionation Techniques 71.3.1 Precipitation 81.3.2 Electrophoresis 81.3.3 Chromatography 91.3.4 Expanded Bed Adsorption 101.3.5 Membrane Adsorption 10

1.4 Fractionation Strategies 101.4.1 Introductory Comments 101.4.2 Initial Fractionation 11

1.4.2.1 Clarification by Centrifugation and/orMicrofiltration 11

1.4.2.2 Ultrafiltration 11

1.4.2.3 Precipitation 11

1.4.2.4 Liquid–Liquid Phase Extraction 121.4.3 The Chromatographic Steps 12

1.4.3.1 Choice of Adsorbent 12

1.4.3.2 The Order of the Chromatographic Steps 121.4.4 The Final Step 13

1.5 Monitoring the Fractionation 141.5.1 Assay of Biological Activity 141.5.2 Determination of Protein Content 141.5.3 Analytical Gel Electrophoresis 16

1.6 The Final Product 161.6.1 Buffer Exchange 161.6.2 Concentration 161.6.3 Drying 17

1.7 Laboratory Equipment 171.7.1 General Equipment 171.7.2 Equipment for Homogenization 181.7.3 Equipment for Chromatography 19

1.7.3.1 Column Design 19

1.7.3.2 Pumps and Fraction Collectors 19

1.7.3.3 Monitoring Equipment 20

1.7.3.4 Chromatography Systems 201.7.4 Equipment for Chromatographic and

Electrophoretic Analyses 21

1.8 References 21

Protein Purification: Principles, High Resolution Methods, and Applications, Third Edition. Edited by Jan-Christer Janson# 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

3

1.1 INTRODUCTION

The development of techniques and methods for the separ-ation and purification of biological macromolecules such asproteins has been an important prerequisite for many of theadvancements made in bioscience and biotechnology overthe past five decades. Improvements inmaterials, utilization ofcomputerized instruments, and an increased use of in vivotagging have made protein separations more predictable andcontrollable, although many still consider purification of non-tagged proteins more an art than a science. However, gone arethe days when an investigator had to spend months in searchof an efficient route to purify an enzyme or hormone from acell extract. This is a consequence of the development ofnew generations of chromatographic media with increasedefficiency and selectivity as well as of newautomated chroma-tographic systems supplied with sophisticated interactivesoftware packages and data bases. New electrophoresis tech-niques and systems for fast analysis of protein compositionand purity have also contributed to increasing the efficiencyof the evaluation phase of the purification process.

In the field of chromatography, the development of newporous resin supports, new crosslinked beaded agaroses, andnew bonded porous silicas has enabled rapid growth in highresolution techniques (high performance liquid chromato-graphy, HPLC; fast protein liquid chromatography, FPLC),both on an analytical and laboratory preparative scale aswell as for industrial chromatography in columns with bedvolumes of several hundred liters. Expanded bed adsorptionenables rapid isolation of target proteins, directly from wholecell cultures or cell homogenates. Another field of increasingimportance is micropreparative chromatography, a conse-quence of modern methods for amino acid and sequenceanalysis requiring submicrogram samples. The data obtainedare efficiently exploited by recombinant DNA technology,and biological activities previously not amenable to properbiochemical study can now be ascribed to identifiable pro-teins and peptides.

Awide variety of chromatographic column packing mater-ials such as gel-filtration media, ion exchangers, reversedphase packings, hydrophobic interaction adsorbents, andaffinity chromatography adsorbents are today commerciallyavailable. These are identified as large diameter media(90–100 mm), medium diameter media (30–50 mm) andsmall diameter media (5–10 mm) in order to satisfy thedifferent requirements of efficiency, capacity, and cost.

However, not all problems in protein purification aresolved by the acquisition of sophisticated laboratory equip-ment and column packings that give high selectivity and effi-ciency. Difficulties still remain in finding optimum conditionsfor protein extraction and sample pretreatment, as well as inchoosing suitable methods for monitoring protein concen-tration and biological activity. These problems will be dis-cussed in this introductory chapter. There will also be an

overview of different protein separation techniques andtheir principles of operation. In subsequent chapters, eachindividual technique will be discussed in more detail.Finally, some basic equipment necessary for efficient proteinpurification work will be described in this chapter.

Several useful books covering protein separation and puri-fication from different points of view are available on themarket or in libraries (1–3). In “Methods of Enzymology,”for example, in older volumes 22, 34, 104, and 182 (4–7),but particularly in the most recent volume, 463 (8), anumber of very useful reviews and detailed applicationreports will be found. The booklets available from manufac-turers regarding their separation equipment and media canalso be helpful by providing detailed information regardingtheir products.

1.2 THE PROTEIN EXTRACT

1.2.1 Choice of Raw Material

In most cases, interest is focused on one particular biologicalactivity, such as that of an enzyme, and the origin of thisactivity is often of little importance. Great care should there-fore be taken in the selection of a suitable source. Amongdifferent sources there might be considerable variation withrespect to the concentration of the enzyme, the availabilityand cost of the raw material, the stability of the enzyme, thepresence of interfering activities and proteins, and difficultiesin handling a particular raw material. Very often it is com-pelling to choose a particular source because it has beendescribed previously in the literature. However, sometimesit is advantageous to consider an alternative choice.

Traditional animal or microbial sources have today, to alarge degree, been replaced by genetically engineered micro-organisms or cultured eukaryotic cells. Protein products ofeukaryotic orgin, cloned and expressed in bacteria such asEscherichia coli, may either be located in the cytoplasm orsecreted through the cell membrane. In the latter case theyare either collected inside the periplasmic space or they aretruly extracellular, secreted to the culture medium. Proteinsthat accumulate inside the periplasmic space may be selec-tively released either into the growth medium by changingthe growth conditions (9), or following cell harvesting andwashing of the resuspended cell paste. At this stage, a con-siderable degree of purification has already been achievedby choosing a secreting strain as illustrated in Figure 1.1. Inconnection with the cloning, the recombinant protein maybe equipped with an “affinity handle” such as a His-tag or afusion protein such as Protein A, glutathione-S-transferase,or maltose binding protein in order to facilitate purification.The handle is often designed such that it can be cleavedoff using highly specific proteolytic enzymes. Proteins ofeukaryotic origin, and some virus surface proteins are often

4 INTRODUCTION TO PROTEIN PURIFICATION

glycosylated why eukaryotic host cells have to be chosen fortheir production.

1.2.2 Extraction Methods

Some biological materials themselves constitute a clear ornearly clear protein solution suitable for direct applicationto chromatography columns after centrifugation or filtration.Examples include blood serum, urine, milk, snake venoms,and—perhaps most importantly—the extracellular mediumafter cultivation of microorganisms and mammalian cells,as mentioned above. It is normally an advantage to choosesuch a starting material because of the limited number ofcomponents and also because extracellular proteins are com-paratively stable. Some samples, such as urine or cell culturesupernatants, are normally concentrated before purificationbegins.

In most cases, however, it is necessary to extract theactivity from a tissue or a cell paste. This means that a con-siderable number of contaminating molecular species areset free, and proteolytic activity will make the preparationwork more difficult. The extraction of a particular proteinfrom a solid source often involves a compromise betweenrecovery and purity. Optimization of extraction conditionsshould favor the release of the desired protein and leave diffi-cult-to-remove contaminants behind. Of particular concern isto find conditions under which the already extracted protein isnot degraded or denatured while more is being released.

Various methods are available for the homogenizationof cells or tissues. For further details and discussions thereader is referred to the paper by Kula and Schütte (10).The extraction conditions are optimized by systematic vari-ation of parameters such as the composition of the extraction

medium (see below), time, temperature, and type of equip-ment used.

The proper design of an extraction method thus requirespreliminary experiments in which aliquots are taken at var-ious time intervals and analyzed for activity and protein con-tent. The number of parameters can be very large, so this partof the work has to be kept within limits by applying properjudgment. However, it is not recommended to accept asingle successful experiment. Further investigations of therequired extraction time, in particular, often pays in thelong run. The number of optimization experiments canbe reduced considerably by using chemometrics (multivariateanalysis), for which there are computer programs available(www.chemometrics.com/software).

The major problems confronted when preparing a proteinare in general denaturation, proteolysis, and contaminationwith pyrogens, nucleic acids, bacteria, and viruses. Thesecan be limited by appropriate choice of the extractionmedium, as we shall show. However, we can already statethat many of the above problems can be reduced by shortpreparation times and low temperatures. It is therefore goodbiochemical practice to carry out the first preparation stepsas fast as possible and at the lowest possible temperature.However, low temperatures are not always necessary andare sometimes inconvenient. The working temperature istherefore one of the parameters that should be optimized care-fully, especially if a preparation is to be done routinely in thelaboratory or if it is going to be scaled up to pilot or pro-duction scale.

The extract must be clarified by centrifugation and/or byfiltration before submission to column chromatography. Apreparative laboratory centrifuge is normally sufficient forthis step.

A common phenomenon when working with intracellu-larly expressed recombinant proteins is their tendency toaccumulate as insoluble aggregates known as inclusionbodies, which have to be solubilized and refolded to recovertheir native state. At first glance, the formation of insolubleaggregates in the cytoplasm might be considered a majorproblem. However, as the inclusion bodies seem to be fairlywell defined with regard to both particle size and density(11), they should provide a unique means for rapid and effi-cient enrichment of the desired protein simply by low speedfractional centrifugation andwashing of the resuspended sedi-ment. The critical step is solubilization and refolding, oftencombined with chromatographic purification under denaturat-ing conditions in the presence of high concentrations of urea orguanidine hydrochloride. This area is treated in more detail inChapter 13 and has recently been reviewed by Burgess (12).

1.2.3 Extraction Medium

To arrive at a suitable composition for the extraction mediumit is necessary initially to study the conditions at which the

Figure 1.1 Location and approximate numbers of proteins inE. coli.

1.2 THE PROTEIN EXTRACT 5

protein of interest is stable and secondly, where it is mostefficiently released from the cells or tissue. The final choiceis usually a compromise between maximum recovery andmaximum purity. The following factors have to be takeninto consideration: pH, buffer salts, detergents/chaotropicagents, reducing agents, chelators or metal ions, proteolyticinhibitors, and bacteriostatics.

1.2.3.1 pH Normally, the pH value is chosen such that theactivity of the protein is at a maximum. However, it should benoted that this is not always the pH that gives the mostefficient extraction, nor is it necessarily the pH of maximumstability. For example, trypsin has an activity optimum at pH8–9, but is much more stable at pH 3, where autolysis isavoided. The use of extreme pH values, for example, forthe extraction of yeast enzymes in 0.5 M ammonia, is some-times very efficient and is acceptable for some proteins with-out causing too much denaturation.

1.2.3.2 Buffer Salts Most proteins are maximally solubleat moderate ionic strengths, 0.05–0.1, and these values arechosen if the buffer capacity is sufficient. Suitable buffersalts are given in Table 1.1.

An acceptable buffer capacity is obtained within one pHunit from the pKa values given. The proteins as such alsoact as buffers, and the pH should be checked after additionof large amounts of proteins to a weakly buffered solution.Some extractions do not give rise to acids and bases andthus do not need a high buffer capacity. In other cases thismight be necessary, and occasional control of the pH valueof an extract is recommended.

1.2.3.3 Detergents and Chaotropic Agents In manyextractions the desired protein is bound to membranes orparticles, or is aggregated due to its hydrophobic character.In these cases one should reduce the hydrophobic interactionsby using either detergents or chaotropic agents (not both!).Some of the commonly used detergents are listed inTable 1.2. Several of them do not denature globular proteins

or interfere with their biological activity. Others, such assodium dodecyl sulfate (SDS), will do that. Quite often it isnot necessary to continue using a detergent in the bufferafter the first step(s) in the purification, so its use is restrictedto the extraction medium. In other cases it might be necessaryto use a detergent throughout the whole preparation process,leading to the final purification of a protein–detergent com-plex. More information about detergents, including theirchemical structures, can be found in Reference 13.

Detergents are amphipathic molecules. When their con-centration increases they will eventually aggregate; that is,they will form micelles at the so-called critical micelle con-centration (CMC). Because micelles often complicate purifi-cation procedures, in particular column chromatography,detergent concentrations below the CMC should be used.

Instead of using detergents to dissolve aggregates, chao-tropic agents such as urea or guanidine hydrochloride, ormoderately hydrophobic organic compounds such as ethyl-ene glycol, can be tried. Urea and guanidine hydrochloridehave proven particularly useful for the extraction and solubil-ization of inclusion bodies (12).

1.2.3.4 Reducing Agents The redox potential of the cyto-sol is lower than that of the surrounding medium whereatmospheric oxygen is present. Intracellular proteins oftenhave exposed thiol groups and these might become oxidizedin the purification process. Thiol groups can be protected byreducing agents such as 1,4-dithioerythritol (DTE), dithio-threitol (DTT) or mercaptoethanol (Table 1.3). Normally,10–25 mM concentrations are sufficient to protect thiolswithout reducing internal disulfides. In other cases a higherconcentration might be needed (14). Ascorbic acid is some-times added to polyphenol containing crude plant extractsin order to avoid oxidation and miscoloration.

1.2.3.5 Chelators or Metal Ions The presence of heavymetal ions can be detrimental to a biologically active protein,

TABLE 1.1 Buffer Salts Used in Protein Work

Buffer pK values Properties

Sodium acetate 4.75Sodium bicarbonate 6.50, 10.25Sodium citrate 3.09, 4.75, 5.41 Binds Ca2þ

Ammonium acetate 4.75, 9.25 VolatileAmmonium bicarbonate 6.50, 9.25, 10.25 VolatileTris-hydrochloride 8.21Sodium phosphate 1.5, 7.5, 12.0Tris-phosphate 7.5, 8.21

Buffer concentration refers to total concentration of buffering species. BufferpH should be as close as possible to the pKa value, and not more than one pHunit from the pKa.

TABLE 1.2 Detergents Used for Solubilization of Proteins

DetergentIonic

CharacterEffect onProtein

Critical MicelleConcentration,

% w/v

Triton X-100 Nonionic Mild, nondenaturing 0.02Nonidet P 40 Nonionic Mild, nondenaturing 0.012Lubrol PX Nonionic Mild, nondenaturing 0.006Octyl glucoside Nonionic Mild, nondenaturing 0.73Tween 80 Nonionic Mild, nondenaturing 0.002Sodium

deoxycholateAnionic Mild, denaturing 0.21

Sodium dodecylsulfate, SDS

Anionic Strongly denaturing 0.23

CHAPS Zwitter-ionic

Mild, denaturing 1.4


for two main reasons. They can enhance the oxidation ofthiols by molecular oxygen and can form complexes withspecific groups, which may cause problems. Heavy metalscan be trapped by chelating agents. The most commonlyused is ethylenediamine tetraacetic acid (EDTA) in the con-centration range 10–25 mM. An alternative is ethyleneglycol tetraacetic acid (EGTA), which is more specific for cal-cium. It should be noted that EDTA is a buffer. It is thereforebest to add EDTA before final pH adjustment. The chelatingcapacity of EDTA increases with increasing pH.

In other cases, stabilizing metal ions are needed. Manyproteins are stabilized by calcium ions. However, the divalentions calcium and magnesium are trapped by EDTA andcannot be used in combination with this chelator.

1.2.3.6 Proteolytic Inhibitors The most serious threats toprotein stability are the omnipresent proteases. The simplestsafeguard against proteolytic degradation is normally towork quickly at low temperatures. An alternative, or added,precaution is to make use of protease inhibitors (Table 1.4),especially in connection with the extraction step. Oftenthere is a need for a combination of inhibitors, for example,for both serine proteases and metalloproteases. In general,protein inhibitors are expensive, which may limit their usein large-scale applications. Proteolysis can also be reducedby rapid extraction of the fresh homogenate in an aqueouspolymer two-phase system (15) or by adsorption of theproteases to hydrophobic interaction adsorbents (16).Sometimes it is sufficient to adjust the pH to a value atwhich the proteases are inactive, but where the stability ofthe protein to be purified is maintained. A classical exampleis the purification of insulin from the pancreas.

1.2.3.7 Bacteriostatics It is wise to take precautions toavoid bacterial growth in protein solutions. The simplest

remedy here is to use sterile filtered buffer solutions as routinein the laboratory. This will also reduce the risk of bacterialgrowth in columns. A common practice for avoiding bacterialgrowth in chromatographic columns is to allow the column toflow at a reduced rate, even when it is not in operation. Somebuffers are more likely than others to support bacterialgrowth, such as phosphate, acetate, and carbonate buffers atneutral pH values. Buffers at pH 3 and below or at pH 9and above usually prevent bacterial growth, but mayoccasionally allow growth of molds.

Whenever possible it is recommended to add an antimicro-bial agent to the buffer solutions. Often used are sodium azideat 0.001M or merthiolate at 0.005%, or alcohols such as n-butanol at 1%. Sodium azide has the drawback that it is anucleophilic substance and binds metals. In cases wherethese substances may interfere with activity measurementsor the chromatography itself, it is always possible to addthe substances to solutions of the protein to be stored.

1.3 AN OVERVIEW OF FRACTIONATIONTECHNIQUES

In early work, complex protein mixtures were fraction-ated mainly by adsorption and precipitation methods. Thesemethods are still used today as preliminary steps for initialfractionation or for concentration of sample solutions.Preparative electrophoretic and chromatographic techniquesdeveloped during the l950s and 1960s made possible rationalpurification protocols and laid the foundation for the situationwe have today. The following sections give a short overviewof the various techniques normally used in preparative bio-chemical work. Chapter 2 contains an introduction to chrom-atography, including a historical review, and Chapter 15 givesan introduction to electrophoresis. Each individual chroma-tographic and electrophoretic separation technique is thentreated in detail in subsequent chapters.

TABLE 1.3 Reducing Agents

Agent Structure

Mercaptoethanol HS—CH2—CH2—OH

1,4-Dithioerythritol (DTE)

1,4-Dithiothreitol (DTT)

Ascorbic acid

TABLE 1.4 Proteolytic Inhibitors

InhibitorEnzymesInhibited

WorkingConcentration

Diisopropylfluorophosphate (DFP)

Serine proteases (avoid DFP)

Phenylmethylsulfonylfluoride (PMSF)

Serine proteases 0.5–1 mM

Ethylenediaminetetraacetate (EDTA)

Metal-activatedproteases

�5 mM

Cysteine reagents Cysteine-dependentproteases

(varying)

Pepstatin A Acid proteases 1mMLeupeptin Serin proteases 1 mM

1.3 AN OVERVIEW OF FRACTIONATION TECHNIQUES 7

1.3.1 Precipitation

Precipitation of a protein in an extract may be achieved byadding salts, organic solvents, or organic polymers, or byvarying the pH or temperature of the solution. The most com-monly used precipitation agents are listed in Table 1.5. Thestrength of a particular ion as a precipitation agent is shownby its position in the so-called Hofmeister series:

Anions: PO43– , SO4

2–, CH3COO2, Cl2, Br2, NO3

2, ClO32,

I2, SCN2

Cations: NH4þ, Kþ, Naþ, guanidine C(NH2)3

þ

The so-called antichaotropic ions to the left are the most effi-cient salting out agents. They are efficient water moleculebinders, thus increasing the hydrophobic effect in the solutionand promoting protein aggregation by facilitating the associ-ation of hydrophobic surfaces. The chaotropic salts on theright-hand side in the series decrease the hydrophobiceffect, and thus help maintain the proteins in solution.

Polar organic solvents such as ethanol promote the precipi-tation of proteins due to the decrease in water activity in thesolution as the water is replaced by organic solvent. Theyhave been widely used as precipitation agents, especially inthe fractionation of serum proteins. The following five vari-ables are usually kept under control: concentration of organicsolvent, protein concentration, pH, ionic strength, and temp-erature (17). Low temperature during the precipitation oper-ations is often necessary to avoid protein denaturation; theaddition of an organic solvent decreases the freezing pointof the solution and temperatures below 08C can be used. Inreversed phase chromatography, some proteins can be chro-matographed in solutions that contain up to �50% organicsolvent, with retention of their biological activity.

Organic polymers function in a way similar to that oforganic solvents. The most widely used polymer is polyethy-lene glycol (PEG), with an average molecular weight of either6000 or 20,000. The main advantage of PEG over organicsolvents is that it is more easily handled. It is unflammable,not poisonous, uncharged, and relatively unexpensive.Rather low concentrations are required (often less than25%) to precipitate most proteins. One disadvantage is that

high concentration solutions of PEG are viscous. PEG canalso be difficult to remove from protein solutions. However,after dilution with buffer the viscosity decreases, and becausethe substance is uncharged, the solution may be applieddirectly to an ion exchange column to further separate theproteins, simultaneously removing the polymer.

pH adjustment has been used as a simple and cheap way toprecipitate proteins. Proteins have their lowest solubility attheir isoelectric point. This is sometimes used in serumfractionation and also in the purification of insulin.

Besides pH, another parameter that influences precipi-tation of proteins in salt solutions is temperature (see below).Keeping the salt concentration constant and varying thetemperature is another way of fractionating a protein solution.

The salting out of a protein can be described by theequation

log S ¼ B� Kc

where S is the solubility of the protein in g/L of solution, B isan intercept constant,K is the salting out constant, and c is saltconcentration in mol/L.

The value of B depends on the salt used, the pH, the temp-erature, and the protein itself; K depends on the salt used andthe protein. It should be stressed that addition of a saltor another precipitating agent to a protein solution onlydecreases the solubility of the proteins. This is why a verydilute protein solution for precipitation may lead to lowrecovery, because a major part of the protein simply remainsin solution. Reproducible results can only be achieved if allthe parameters mentioned above, including the protein con-centration, are kept constant.

Centrifugation is used routinely in the protein purificationlaboratory to recover precipitates. It can also be used to sep-arate two immiscible liquid phases. Another application isdensity gradient centrifugation. Today this is used predomi-nantly for the fractionation of subcellular particles andnucleic acids. An alternative is the use of liquid–liquidphase extraction, which seems to offer several advantagesover the more classical methods.

1.3.2 Electrophoresis

Electrophoresis in free solution or in macroporous gels suchas 1–2% agarose separates proteins mainly according totheir net electric charge. Electrophoresis in gels such as poly-acrylamide separates mainly according to the molecular sizeof the proteins.

Today, analytical gel electrophoresis requiring microgramamounts of proteins is an important tool in bioscience andbiotechnology (see Chapter 15). Convenient methods forthe extraction of proteins after electrophoresis have beendeveloped, in particular protein blotting (see Chapter 18),making the technique micropreparative. There are alsomany instances where a very small amount of protein is

TABLE 1.5 Precipitation Agents

Agent Type Properties

Ammonium sulfate Salt Easily soluble, stabilizingSodium sulfate SaltEthanol Solvent Flammable, risk of

denaturationAcetone Solvent Flammable, risk of

denaturationPolyethylene glycol(PEG)

Polymer Uncharged, unflammable


sufficient for the analysis of size and composition as well asthe primary structure. Finally, there are cases where the start-ing material is extremely limited, such as protein extractsfrom small amounts of tissue (biopsies, etc.). In these cases,the protein “extract” might be just large enough for gel elec-trophoretic analysis.

Larger scale (milligrams to grams of protein) electrophor-esis was an important method for the fractionation of proteinextracts during the 1950s and early 1960s. It was carried outusing columns packed with, for example, cellulose powder asa convection depressor, as in the “Porath column” (18). Aninnate limitation of preparative column electrophoresis is thejoule heat developed during the course of the experiment.This means that the column diameter, if it is to allow sufficientcooling, should not exceed �3 cm. Several hundred milli-grams of protein can, however, be separated on such columns.Column zone electrophoresis has the advantage of allowing aprecise description of the separation parameters involved andis, besides gel filtration, the mildest separation techniqueavailable for proteins. It can be recommended for special situ-ations, but practical aspects and the excessive time requiredprecludes its routine use. Methods for large and mediumscale preparative electrophoresis have been developed, suchas the flowing curtain electrophoresis of Hannig (19) and,more recently, the “Biostream” apparatus of Thomson (20).

Isoelectric focusing, the other main electrophoretic tech-nique, separates proteins according to their isoelectric points(see Chapter 16). This technique gives very high resolution,but presents major difficulties as a preparative large ormedium scale technique. Special equipment is required toallow cooling during the focusing. Proteins often precipitateat their isoelectric point, and this precipitate can contaminatethe other bands when a vertical SephadexTM bed or columnwith sucrose gradient is used as an anticonvection mediumin the focusing experiment. Modern equipment for prep-arative isoelectric focusing (21) avoids these problems bydividing the separation chamber into smaller compartments.Another solution is to carry out the isoelectric focusing ina horizontal trough of sedimented gel particles such asSephadex (22). Here, precipitation in one zone will not dis-turb the other bands. On the other hand, the recovery of pro-teins is more tedious.

For routine preparative protein fractionation the electro-phoretic techniques have become less important than chrom-atography. Ion exchange chromatography depends onparameters similar to those for electrophoresis. Chromato-focusing fractionates proteins largely according to their iso-electric points and would therefore appear to be a moreconvenient alternative to preparative isoelectric focusing.

1.3.3 Chromatography

Separation by chromatography depends on the differentialpartition of proteins between a stationary phase (the

chromatographic medium or the adsorbent) and a mobilephase (the buffer solution). Normally, the stationaryphase is packed into a vertical column of plastic, glass, orstainless steel, and the buffer is pumped through thiscolumn. An alternative is to stir the protein solution withthe adsorbent, batchwise, and then pour the slurry onto anappropriate filter and make the washings and desorptions onthe filter.

Column chromatography has proved to be an extremelyefficient technique for the separation of proteins in biologicalextracts. Since the development of the first cellulose ionexchangers by Peterson and Sober (23) and of the first prac-tical gel filtration media by Porath and Flodin (24, 25) awide variety of adsorbents have been introduced that exploitvarious properties of the protein for the fractionation. Themore important of these properties, together with the chroma-tographic method for which they dominate the separation, areas follows:

1. Size and shape (gel filtration/size exclusion chromato-graphy, SEC).

2. Net charge and distribution of charged groups (ionexchange chromatography, IEC).

3. Isoelectric point (chromatofocusing, CF).

4. Hydrophobicity (hydrophobic interaction chromato-graphy, HIC; reversed phase chromatography, RPC).

5. Metal binding (immobilized metal ion affinity chrom-atography, IMAC).

6. Content of exposed thiol groups (covalent chromato-graphy, CC).

7. Biospecific affinities for ligands, inhibitors, receptors,antibodies, and so on (affinity chromatography, AC).

The methods often have very different requirements withregard to chromatographic conditions., including ionicstrength, pH, and various additives such as detergents, redu-cing agents, and metals. By appropriate adjustment of thebuffer composition, the conditions for adsorption and desorp-tion of the desired protein can be optimized. It should bestressed that the result of a particular chromatographic separ-ation often depends on more than one parameter. In IEC, thecharge interaction is the dominant parameter, but molecularweight and hydrophobic effects can also contribute to somedegree, depending on the experimental conditions and typeof solid phase used. In recent years the concept of multimo-dal, or mixed-mode adsorption chromatography, has receivedan increasing amount of attention, with several new productsemerging on the market (see Chapter 4 for more detailedinformation).

Highly specific methods, such as those based on bioaffi-nity (e.g., antibody–antigen interaction) or those based onthe use of in vivo fused tags such as (His)6 or glutathione-S-transferase (GST), do in some cases give a highly pure

1.3 AN OVERVIEW OF FRACTIONATION TECHNIQUES 9

protein in a single step. Normally, however, several chro-matographic methods have to be combined in orderto achieve maximal purification of a protein from a crudebiological extract. With the wide variety of chromatogra-phic media available today, in combination with a moderncomputerized chromatography system, adequate purificationcan normally be achieved within a few days to a couple ofweeks.

In recent years, columns containing a continuous, homo-geneously porous solid phase have become available. SeeChapters 2 and 9 for more information about these so-called monolithic column types. As in membrane adsorptiontechniques, the main advantage is the considerably reduceddiffusion restriction, allowing high efficiency and also highflow rates. The main disadvantage of both these techniquesis the concomitant smaller surface area per unit adsorptionmedium volume, which will restrict the nominal columnbinding capacity.

All of the chromatographic methods mentioned above aretreated in Part II of this book, which begins with a generaldescription of the concepts used in protein chromatography.

1.3.4 Expanded Bed Adsorption

The problem of removing cells and cell debris from largevolumes of whole cell cultures or cell homogenates hasencouraged the development of technologies for the directadsorption of target molecules from such feed stocks. In afluidized bed, the adsorbent particles are subject to anupward flow of liquid that keeps them separated from eachother. The resulting increased voidage allows the unhinderedpassage of cells and cell debris. In a typical fluidized bedthere is a total mixing of particles and sample in the reactor,leading to incomplete adsorption of the target moleculesunless the feed stock is recycled. Expanded bed adsorptionis a special case of fluidized bed adsorption (26), and is pri-marily applied in a pilot- or production-scale environment(26–29).

1.3.5 Membrane Adsorption

The main argument for utilizing modified membranes asmedia for protein adsorption is to solve the problem ofmass transport restriction in standard chromatography dueto the slow diffusion of proteins in the pores of the largegel particles. In membranes, most pores allow convectiveflow, and the mass transport resistance is therefore minimizedto film diffusion at the membrane matrix surface. The resultis a more efficient adsorption–desorption cycle of targetsolutes, allowing considerably higher flow rates and thusconsiderably shorter separation times. The area has beenreviewed by Thömmes and Kula (30). See Chapter 12 formore data regarding membrane separation.

1.4 FRACTIONATION STRATEGIES

1.4.1 Introductory Comments

Before attempting to design a purification protocol for aparticular protein, as much information as possible shouldbe collected about the characteristics of that protein andpreferably also about the properties of the most importantknown impurities. Useful data include approximate molecu-lar weight and pI, degree of hydrophobicity, presence ofcarbohydrate (glycoprotein) or free –SH. Some of this infor-mation might be obtained already on a DNA level, if nucleo-tide sequence data are available, but is otherwise oftencollected easily by preliminary trials using crude extracts.

Criteria with regard to the stability of the protein to bepurified should be established. Important parameters affect-ing structure are temperature, pH, organic solvents, oxygen(air), heavy metals, and mechanical shear. Special concernshould be addressed to the risk of proteolytic degradation.Finally, it is the amount of protein to be purified per batch,and the required degree of purity, that to a high degree gov-erns the techniques and methods used in the purificationprocess.

According to a study of 100 published successful proteinpurification procedures (31), the average number of steps in apurification process is four. Very seldom can a protein beobtained in pure form from a single chromatographic pro-cedure, even when this is based on a unique biospecificity.In addition to the purification steps there is often a need forconcentrations and sometimes changes of buffers by dialysisor membrane ultrafiltration.

The preparation scheme can be described as consistingof three stages:

1. The preliminary or initial fractionation stage (oftencalled the capturing stage).

2. The intermediate purification stage.

3. The final polishing stage.

The purpose of the initial stage is to obtain a stable,particle-free solution suitable for chromatography. This isachieved by clarification, coarse fractionation, and concen-tration of the protein extract. The purpose of the final stageis to remove aggregates and degradation products and toprepare the protein solution for the final formulation of thepurified protein.

Sometimes one or two of these stages coincide. An initialion exchange adsorption step can thus serve as a preliminaryfractionation applied directly to the protein extract, or a gelfiltration can give a product that is suitable as a final product,However, as the purposes of the three stages are different it isuseful to discuss them separately.

The design of the preparation scheme will differ depend-ing on the material at hand and the purpose of the purification.


If the starting material is very precious, one should favor highyield over speed and convenience. In cases where severaldifferent proteins are to be extracted from a single startingmaterial, this of course also affects the planning of thework. Finally, the final step is designed so that the productwill be suitable for its purpose, which can vary. These aspectswill be discussed below.

1.4.2 Initial Fractionation

There are many methods for the clarification of proteinsolutions. Extracts of fungal or plant origin often contain phe-nolic substances or other pigments. These can be removedby adsorption to diatomite (diatomaceous earth, Celite),either batchwise or on a short column. In order to preventoxidation and miscoloration, small amounts of ascorbicacid can be added to the crude plant extract.

Similarly, lipid material can be removed either by centrifu-gation, as the lipids will float, and one thus needs to extractthe protein solution from below, or by a chromatographicprocedure. Lipids adsorb to a number of materials. Aerosil,a fused silica, has been used for the adsorption of lipids,but agarose is sometimes a simple choice.

Contamination with nucleic acids can, in some cases,especially when preparing proteins from bacteria, constitutea problem due to their high viscosity. The classical way tosolve this problem is to precipitate the nucleic acids. Strepto-mycin sulfate and polyethylenimine have been used as preci-pitants, as have protamine sulfate and manganese salts (32).Another way to solve the problem is to add nucleases, whichcut the nucleic acids into smaller pieces, thereby reducing theviscosity. Another problem with nucleic acids or degradationproducts of nucleic acids is that, due to their low isoelectricpoints, they still are negatively charged at low pH. Anionexchangers strongly adsorb nucleic acids and are thus difficultto regenerate. The solution to this problem can in some casesbe to perform two consecutive adsorption steps. The first isexecuted at a pH below the pI of the target protein, thus pre-venting it from binding to the ion exchanger. Often, a fairlysmall amount of the ion exchanger is required in this step,which is why it is economically motivated to discard the con-taminated gel. In the second step using the same anionexchanger, the pH is increased to a value above the pI ofthe target protein, resulting in binding and subsequent elutionusing either a stepwise or continuous salt gradient.

1.4.2.1 Clarification by Centrifugation and/orMicrofiltration The clarification of any cell homogenateis usually no problem on a laboratory scale, where refri-gerated high speed centrifuges operating at speeds from20,000 rpm to 75,000 rpm, generating from �40,000g to�500,000g can be used. A useful review of centrifugationand centrifuges in preparative biochemistry is found inReference 33. As a complement to centrifugation, in recent

years, tangential or cross-flow microfiltration has receivedincreased attention, especially for large-scale applications.For a review of the advantages of cross-flow microfiltrationwe suggest Reference 34. The area is also treated in moredetail in Chapter 12.

1.4.2.2 Ultrafiltration Ultrafiltration has become awidelyused technique in preparative biochemistry. Ultrafiltrationmembranes are available with different cut-off limits for sep-aration of molecules from 1 kDa up to 300 kDa. The methodis excellent for the separation of salts and other small mol-ecules from a protein fraction with higher molecular weightand at the same time can effect a concentration of the proteins.The process is gentle, fast, and inexpensive. Ultrafiltration istreated in more detail in Chapter 12.

1.4.2.3 Precipitation Crude extracts are seldom suitablefor direct application to chromatographic columns. Prepara-tive differential centrifugation seldom results in a sufficientlyclear solution. This is one reason why it is often necessary touse other means for clarification that simultaneously concen-trate the solution and remove most of the bulk proteins. Suchan initial fractionation step should also result in the removalof proteases and membrane fragments that sometimes bindthe protein of interest in the absence of detergents. The clas-sical means is to make a fractional precipitation. Bulk proteinsin the solution are first precipitated together with residual par-ticulate matter, and then the protein of interest can be precipi-tated from the resulting supernatant solution. Sometimes theprotein of interest is allowed to remain in the mother liquorsolution for direct application to chromatographic columns,for example, hydrophobic interaction adsorption of proteinsin ammonium sulfate solutions and IEC of proteins in PEGmother liquors. The most commonly used precipitatingagents are listed in Table 1.5, together with some of their prop-erties. A typical precipitation curve is shown in Figure 1.2.

Figure 1.2 Example of a precipitation curve, showing the amountof protein precipitated with a stepwise increase in ammonium sulfateconcentration.

1.4 FRACTIONATION STRATEGIES 11

Of the various methods available for protein precipitation,the classical ammonium sulfate has some disadvantages. Theresulting protein solution often needs to be dialyzed to obtainan ionic strength that allows IEC. This problem is avoidedwhen using PEG. Organic solvents, in particular ethanoland acetone, often produce extremely fine powder-like pre-cipitates that are difficult to centrifuge and handle. Theyhave also often been shown to cause partial denaturation ofproteins, which can, for example, prevent subsequent crystal-lization. This is why organic solvents are not recommendedas first-choice precipitating agents.

1.4.2.4 Liquid–Liquid Phase Extraction A radicallydifferent way of making an initial fractionation is by partition-ing in an aqueous polymer liquid–liquid two-phase system(35). These systems often contain PEG as one phase constitu-ent and another polymer, such as dextran or even salt, as theother. Under favorable conditions it is possible to obtain theprotein of interest in the upper, normally the PEG phase. Thecontaminating bulk protein, as well as particles, will be col-lected in the lower phase and can be removed by centrifu-gation. Particles sometimes stay at the interphase and arethus also removed in the centrifugation step. Aqueouspolymer two-phase systems have been shown to be effectivetools for plasma membrane proteomics (36). By covalentattachment of affinity ligands to PEG molecules these canbe used for affinity partitioning.

1.4.3 The Chromatographic Steps

1.4.3.1 Choice of Adsorbent Preliminary separationconditions for known proteins are easily extracted usingdata bases available over the Internet. For unknown (e.g.,nonrecombinant) proteins, information regarding their chro-matographic behavior can only be obtained by preliminaryanalytical-scale experiments, for example by gel filtrationand by IEC using salt and pH gradients. Using these tech-niques, approximate values of molecular size and ionic prop-erties such as isoelectric points are obtained, information thatis fundamental to the further planning of the work. A morethorough survey of the behavior of the protein on variousadsorbents can then be done using a panel of adsorbents.This can be carried out either in a panel of parallel columnsor using tandem columns.

A classical parallel column approach was developed byScopes (37) for a panel of dye adsorbents. In this case heused up to 20 small columns containing various dye adsor-bents. The columns were equilibrated with a predeterminedapplication or starting buffer. A small volume of the proteinextract was applied to each column and the protein content(A280 absorption) and activity in the effluent measured. Apredetermined terminating buffer was then applied to eachcolumn, and the protein and enzyme activity in the effluentwere then determined. A column where the bulk of the

proteins, but not the activity, was adsorbed was chosen as a“minus column,” and an adsorbent where the reverse hap-pened was chosen as a “plus column.” These two columnsin combination effected a considerable purification of thedesired substance in the actual preparation. In an earlier butsimilar approach, a panel of parallel columns was used byShaltiel (38) for the evaluation of hydrophobic adsorbents.The technique can, however, be used for any set-up of adsor-bents such as different ion exchangers, the same ion exchan-ger under different conditions, thiol-gels, metal-chelatinggels, and so on. The elution of the columns can also be per-formed with more than two elution buffers. The purpose,however, is to get a quick idea of the behavior of a previouslyunknown protein and thus the set-up should not be enlargedbeyond what can be handled easily in the laboratory.

If the adsorbents used have well defined and continuouslyincreasing adsorption capacities for proteins, in general thepanel can also be arranged as tandem columns. This approachwas used by Porath and co-workers for the immobilized metalion (IMAC) adsorbents (39). Here, three columns (e.g., Zn,Fe, and Cu) were connected in series and a sample waspumped through all of them. After washing with startingbuffer, the three columns were disconnected and eluted sep-arately, mostly using gradients. The approach requires thatthe first column adsorbs few of the proteins present, whereasthe last adsorbs almost all of them. This technique is not asgenerally applicable as the use of parallel columns.

In Chapter 20, the use of high throughput methods in thedevelopment of industrial-scale protein purification processeswill be discussed.

1.4.3.2 The Order of the Chromatographic Steps Apriori, one would expect that the order in which the differentchromatographic steps are applied in a protein purificationprotocol is of minor importance. The total purificationfactor should be constant and the product of the factorsobtained in each individual step should be independent ofthe other steps of the protocol. In the ideal case, where eachchromatographic technique is utilized optimally with regardto the resolution and recovery, that is, within the linearregions of the adsorption isotherms (see Chapter 2), with ade-quate sample volume to column volume ratios, and with noadverse viscosity effects, this is probably true. However thereal-life situations are always far from ideal or at least suchthat adaptation to ideality becomes highly impractical. Forexample, a fractionation gel filtration step can be optimizedto give very high resolution (Chapter 3), but only at thecost of time and sample volume. To choose fractionationgel filtration as the first step, when the sample volumemight be much larger than the total volume of the column,means repetitive injections and excessive and impracticaltotal process times, which would probably also be deleteriousto the proteins in the sample solution. Likewise, to choose ACon immobilized monoclonal antibodies as the first step would


probably result in an extraordinarily high purification factor.However, the high cost of such adsorbents prohibits the use oflarge columns, which makes repeated injections of sample insmaller columns almost mandatory. This leads to long pro-cess times and the risk of product losses and/or modificationsdue to proteolytic attack. Proteolytic activity can also threatenthe stability and life length of the actual immunosorbent.Furthermore, protein-based adsorbents are difficult to main-tain to a sufficiently high degree of hygiene. There are limit-ations with regard to means for regeneration (washing) andsterilization (Chapter 9). This is why they should be savedfor the later steps of the purification protocol.

The consequence of these considerations is that there are anumber of practical rather than theoretical reasons why oneshould choose certain chromatographic techniques (31) forthe early steps and others for the final steps of a proteinpurification process. The choice is primarily governed bythe following parameters:

† the sample volume† the protein concentration and viscosity of the sample† the degree of purity of the protein product† the presence of nucleic acids, pyrogens, and proteolyticenzymes in the sample

† the ease with which different types of adsorbents can bewashed free from adsorbed contaminants and denaturedprotein.

The last parameter governs the life length of the adsorbentand, together with its purchasing price, the material cost ofthe particular purification step.

In light of what has been said above, the logical sequenceof chromatographic steps would start with more “robust”techniques that combine a concentration effect with highchemical and physical resistance and low material cost. Theobvious candidates are IEC and, to some extent, HIC. Asthe latter often requires the addition of salt for adequate pro-tein binding, it is preferably applied after salt precipitation orafter salt displacement from IEC, thereby excluding the needfor a desalting step. Thereafter, the protein fractions can pre-ferably be applied to a more “specific” and more expensiveadsorbent. The protocol is often finished with a gel filtrationstep (Fig. 1.3).

It is advisable to design the sequence of chromatographicsteps in such a way that buffer changes and concentrationsteps are avoided. The peaks eluted from an ion exchangercan, regardless of the ionic strength, be applied to a gel fil-tration column. This step also functions as a desaltingprocedure, which means that the buffer used for the gel fil-tration should be chosen so as to allow direct application ofthe eluted peaks to the next chromatographic step. The differ-ent chromatographies have, in practice, widely differentcapacities, even though it is possible to adapt several of the

methods to a larger scale. However, in the initial stages of apurification scheme it is most convenient to start with themethods that allow the application of large volumes andwhich have the highest capacities. To this category belongIEC and hydrophobic interaction, but any adsorption chroma-tographic method can be used to concentrate larger volumes,especially in batchwise operations.

1.4.4 The Final Step

The purpose of the final step is to remove possible aggregatesor degradation products and to condition the purified proteinfor its use or storage. The procedure will thus be differentdepending on the fate of the protein. Aggregates and degra-dation products are preferably removed by gel filtration. Ifthe protein is to be lyophilized, this step is also suitable fortransferring the protein to a volatile buffer (Table 1.1). Thiscan sometimes be done by IEC, but more seldom by theother forms of chromatography. If the protein solution isintended to be frozen, stored as a solution, or used immedi-ately the requirements for specific buffer salts might be lessstringent.

Several of the adsorption chromatography steps might bedesigned in such a way that they result in peaks of reasonablyhigh protein concentration. This is an advantage when gel fil-tration is chosen as a final step. Gel filtration will always result

Figure 1.3 Analysis of the methods of purification used at succes-sive steps in the purification schemes. The results are expressed as apercentage of the total number of steps at each stage. Adapted fromReference 31 by permission of the authors and publisher.

1.4 FRACTIONATION STRATEGIES 13

in dilution of the sample and is therefore often followed by aconcentration step.

If the protein is to be used for physical-chemical character-ization, especially for molecular weight studies, gel filtrationhas the advantage of giving a protein solution of defined sizeand also in perfect equilibrium with a particular buffer.Biospecific methods, by definition, give a product that ishomogeneous with respect to biological activity. This wastaken advantage of for papain, where the enzyme elutedfrom a thiol column was twice as active as any previous prep-aration. Many of the early enzyme preparations apparentlycontained molecules in which the thiol necessary for activitywas oxidized (see Chapter 8).

Proteins that, after purification and formulation, areintended for parenteral use in human beings must not containendotoxins (lipopolysaccarides, LPS) or nucleic acids. Thepurification protocols must be designed so that these com-pounds are efficiently removed, and validation studiesshould be performed to prove this. To prepare sterile proteinsolutions, aseptic filtration is used.

1.5 MONITORING THE FRACTIONATION

Proper analysis is a prerequisite for successful protein purifi-cation. Most important is the establishment of a reliable assayof the biological activity. In addition, the protein contentshould be determined in order to be able to assess the effi-ciency of the different steps. It is beyond the scope of thischapter to go into details of the particular assay methods.This is covered by the special literature dealing with theactivity in question—hormone, enzyme, receptor, and so on.

We recommend that each preparation be continuallyrecorded in a purification table (Table 1.6). In combinationwith results from gel electrophoresis, for example, this willserve as a guide for judging the reproducibility and outcomeof each preparation. In addition, each chromatography exper-iment should be accompanied by a suitable protocol such asthe one exemplified in Figure 1.4. However, the need formeasurements of biological activity and protein concen-tration—especially the latter—should not be allowed todelay the preparation, and in many cases it is sufficient tosave aliquots for analysis at one’s convenience.

1.5.1 Assay of Biological Activity

In general, biochemical activities depend on the interactionbetween molecules. This can be measured in differentways. The classical method of enzyme catalysis is only oneof these. In addition, the monitoring of the components canbe done in several ways, such as spectrophotometry, measure-ments of radioactivity, and immunological methods.Examples of these include the following:

† enzyme activity by direct spectroscopy† enzyme activity by secondary measurements onaliquots

† binding of ligand† binding of antibody.

The immunological methods require that the proteinstudied has already been purified once to allow productionof an antibody or an antiserum by immunization. Thedetection of the antigen–antibody precipitate can be doneat almost any sensitivity down to the extreme sensitivityafforded by the use of sandwich techniques and radio-actively or enzymatically labeled reagents (e.g., Chapters17 and 18).

1.5.2 Determination of Protein Content

In general, a measure of protein content is obtained uponmonitoring the effluent in chromatography by UVabsorption.However, it is not always easy to relate these measurements tothe protein content. In fact, the only certain measure ofprotein content is total amino acid analysis after hydrolysis.Strictly speaking, even this latter analysis suffers from someshortcomings, because tryptophan and cysteine normallyhave to be analyzed separately.

Large deviations from true protein values sometimesoccur in the first steps in a purification scheme. The extractitself often contains substances that interfere with the proteinanalyses. An overestimation might result, especially ifmeasurements of absorption at 280 nm are used, becausethe solutions are often turbid and absorbing substances ofnonprotein origin are present. This in turn will make thecalculated values of specific activity erroneous.

TABLE 1.6 Example of a Purification Table

MaterialVolume(mL)

Protein(mg/mL)

Total Protein(mg)

Activity(U)

TotalActivity(U)

Spec.Activity(U/mg)

Yield(%)

Purif.Factor(fold)

Extract 500 14 7000 7 3500 0.5 100 1First purif. step 50 10 500 60 3000 6 85 12

Activity (e.g., enzyme activity) is expressed as units, and specific activity as units per mg of protein (U/mg).


protein purification...protein purification principles, high resolution methods, and applications...

Documents