JPL Publication 88-8 DOE/CS-66001 -11Distribution Categories

UC-61e&UC-61f

ECUTENERGY CONVERSIONAND UTILIZATIONTECHNOLOGIES PROGRAM

Biocatalysis ProjectAnnual ReportFY 1987

iKaSA-CB-182917) ICUT JEBEEGI COBVEBSICfl 888-24110AKD UTILIZATION 3ICHNCLOGIIS) FiiCGEAH:.EICCATflLlSIS;EECJICI fijmual Eepcrt, FY 1987(JetiPropulsion Lah.) 36 p CSCL 10A Unclas

(G3/44 .01.46446

March 30,1988

Sponsored by

Energy Conversion and Utilization Technologies DivisionOffice of Energy Systems ResearchU.S. Department of EnergyThrough an agreement with

National Aeronautics andSpace Administrationby

Jet Propulsion LaboratoryCalifornia Institute of TechnologyPasadena, California

https://ntrs.nasa.gov/search.jsp?R=19880014726 2018-06-21T03:49:08+00:00Z

TECHNICAL REPORT STANDARD TITLE PAGE

1. Report No. JPL PUB 88-8

4. Title and SubtitleECUT — Energy Conversion

Technologies ProgiBlocatalysis Project Annual

2. Government Accession No.

and Utilization•am. Report Fiscal Year 1987

7. Author (j)

9. Performing Organization Name or

JET PROPULSION LAB(California Institul4800 Oak Grove DriiPasadena, Calif orn]

id Address

3RATORY:e of Technology/eLa 91109

12. Sponsoring Agency Nome and Address

NATIONAL AERONAUTICS AND SPACE ADMINISTRATIONWashington, D.C. 20546

3. Recipient's Catalog No.

5. Report DateMarch 30, 1988

6. Performing Organization Code

8. Performing Organization Report No.JPL PUB 88-8

10. Work Unit No.

11. Contract or Grant No.NAS7-918

13. Type of Report and Period Covered

JPL Publication

14. Sponsoring Agency Code

15. Supplementary Notes

16. AbstractThe Annual Report presents the fiscal year (FY) 1987 research activities and

accomplishments for the Biocatalysis Project of the U.S. Department of Energy, EnergyConversion and Utilization Technologies (ECUT) Division. The ECUT Biocatalysis Projectis managed by the Jet Propulsion Laboratory, California Institute of Technology.

The Biocatalysis Project's technical activities were organized, into three workelements:

The Molecular Modeling and Applied Genetics work element includes modeling andsimulation studies to verify a dynamic model of the enzyme carboxypeptidase; plasmidstabilization by chromosomal integration; growth and stability characteristics ofplasmid-containing cells; and determination of optional production parameters forhyper-production of polyphenol oxidase.

The Bioprocess Engineering work element supports efforts in novel bioreactorconcepts that are likely to lead to substantially higher levels of. reactor productivity,product yields, and lower separation energetics.

The Bioprocess Design and Assessment work element attempts to develop procedures(via user-friendly computer software) for assessing thegiven biocatalyst process.

economics and energetics of a

17. Key Words (Selected by Author(s)) 18. Distribution Statement

Bioengineering Unlimited/UnclassifiedConversion TechniquesMaterials (General)Chemical Engineering

19. Security Clossif. (of this report)Unclassified

20. Security Cossif. (of this page) 21. No. of Pages 22. PriceUnclassified '

JPL 0184 R 9(83

JPL Publication 88-8 DOE/CS-66001 -11Distribution Categories

UC-61e&UC-61f

ECUTENERGY CONVERSIONAND UTILIZATIONTECHNOLOGIES PROGRAM

Biocatalysis ProjectAnnual ReportFY 1987

March 30,1988

Sponsored by

Energy Conversion and Utilization Technologies DivisionOffice of Energy Systems ResearchU.S. Department of EnergyThrough an agreement with

National Aeronautics andSpace Administrationby

Jet Propulsion LaboratoryCalifornia Institute of TechnologyPasadena, California

The Biocatalysis Project is managed by the Jet Propulsion Laboratory,California Institute of Technology, for the United States Department of Energythrough an agreement with the National Aeronautics and Space Administration (NASA Task RE-152, Amendment 307; DOE Interagency AgreementDE-AIO1-81CS66001).

The Biocatalysis Project focuses on resolving the major technical barriers thatimpede the potential use of biologically-facilitated continuous chemical produc-tion processes.

This report was prepared as an account of work sponsored by an agency of theUnited States Government. Neither the United States Government nor anyagency thereof, nor any of their employees, makes any warranty, express or im-plied, or assumes any legal liability or responsibility for the accuracy, com-pleteness, or usefulness of any information, apparatus, product, or processdisclosed, or represents that its use would not infringe privately owned rights.

Reference herein to any specific commercial product, process, or service by tradename, trademark, manufacturer, or otherwise, does not necessarily constitute orimply its endorsement, recommendation, or favoring by the United StatesGovernment or any agency thereof. The views and opinions of authorsexpressed herein do not necessarily state or reflect those of the United StatesGovernment or any agency thereof.

ABSTRACT

The Annual Report presents the fiscal year (FY) 1987 research activitiesand accomplishments for the Biocatalysis Project of the U.S. Department ofEnergy, Energy Conversion and Utilization Technologies (ECUT) Division. TheECUT Biocatalysis Project is managed by the Jet Propulsion Laboratory,California Institute of Technology.

The Biocatalysis Project's technical activities were organized intothree work elements:

The Molecular Modeling and Applied Genetics work element includesmodeling and simulation studies to verify a dynamic model of the enzymecarboxypeptidase; plasmid stabilization by chromosomal integration; growth andstability characteristics of plasmid-containing cells; and determination ofoptional production parameters for hyper-production of polyphenol oxidase.

The Bioprocess Engineering work element supports efforts in novelbioreactor concepts that are likely to lead to substantially higher levels ofreactor productivity, product yields, and lower separation energetics.

The Bioprocess Design and Assessment work element attempts to developprocedures (via user-friendly computer software) for assessing the economicsand energetics of a given biocatalyst process.

111

FOREWORD

For further information about the ECUT Biocatalysis Project, contact:

Dr. James J. Eberhardt, Director Dr. Minoo N. Dastoor, ManagerEnergy Conversion and Utilization ECUT Biocatalysis Project

Technologies (ECUT), CE-12 M/S 125-112Forrestal Bldg. 5E-066 Jet Propulsion LaboratoryU.S. Department of Energy California Institute of Technology1000 Independence Avenue, S.W. 4800 Oak Grove DriveWashington, D.C. 20585 Pasadena, California 91109(202) 586-5377 (FTS) 896-5377 (818) 354-7429 (FTS) 792-7429

IV

CONTENTS

I. EXECUTIVE SUMMARY 1-1

II. PROJECT DESCRIPTION 2-1

A. THE ECUT MISSION 2-1

B. HISTORICAL BACKGROUND 2-1

C. RELEVANCE 2-4

D. GOAL AND OBJECTIVE 2-5

E. PROJECT STRUCTURE AND ORGANIZATION 2-5

F. MANAGEMENT SUPPORT FUNCTION 2-6

G. TASK DESCRIPTIONS 2-7

III. FY87 TASK ACCOMPLISHMENTS 3-1

A. MOLECULAR MODELING AND APPLIED GENETICS 3-1

B. BIOPROCESS ENGINEERING 3-3

C. PROCESS DESIGN AND ANALYSIS 3-9

IV. BIBLIOGRAPHY 4-1

FIGURES

2-1. ECUT Role in Energy Conservation Research and Development . . 2-2

2-2. ECUT Program: Projects and Lead Center 2-3

SECTION I

EXECUTIVE SUMMARY

The Biocatalysis Project sponsored by the Energy Conversion andUtilization Technologies (ECUT) Division of the U.S. Department of Energy(DOE) is an applied research effort focused on providing the enablingtechnology base for new bioprocess applications which have substantial energyimplications. The primary mission of DOE-ECUT is to provide, for the nation,a bridge between basic research and industrial application for severaltechnologies (e.g., high temperature materials, tribiology, combustion andbiocatalysis) which have the potential for making positive energycontributions. Furthermore, the ECUT mission hopes to strengthen the futureU.S. competitive position by conducting research directed towards exploitingappropriate advances in basic research in which the U.S. has demonstratedleadership.

As the national lead center, the Jet Propulsion Laboratory of theCalifornia Institute of Technology serves as the technical manager for theBiocatalysis Project. The primary objective of the Biocatalysis Project is toresolve the critical technical constraints (e.g., poor productivity, highseparation energetics, presence of aqueous medium, lack of design tools) thatimpede the utilization of biocatalysis for the production of chemicals andmaterials. To achieve this objective, the Biocatalysis Project supportsgeneric applied research, as well as development of predictive theoreticalmodels (that are experimentally verifiable) which address the technologicalbarriers to the commercial utilization of biochemical catalysis.

The ECUT Biocatalysis Project consists of three major work elements,each addressing a key technical component of the required enabling technologybase. The selection, as well as the relationship between each of these threework elements, has been defined by their scale of action. The MolecularModeling and Applied Genetics work element focuses on defining optimalmicroscale parameters for biocatalysis and pursuing practical applications ofbasic molecular biology research findings. Computer graphic models have beenadvanced which predict the dynamic behavior and intramolecular conformationalchanges of enzymes (biocatalysts), including effects of temperature andbiocatalytic inhibitions. Advances in Applied Genetics have focused on theissue of biological information (plasmid) stability. Novel high-speedmeasurement methods employing flow cytometry have been developed, which nowallow researchers to assay the fraction of plasmid-containing cells in therecombinant population in approximately three hours, more than one order ofmagnitude faster than present day plating methods. Results of suchmeasurements have been used to develop a new theory of plasmid stability whichrecognizes the limited capacity of plasmid-free cells to grow in selectivemedium. The Bioprocess Engineering work element emphasizes defining the basicengineering relationships between cellular scale events and macro-levelparameters. Laboratory tests using an advanced fluidized bed bioreactorpossessing high cell densities (5-15 x 10̂ -0 cells/ml) have demonstratedexceptionally high bioreactor productivities: 50 to 100 g/l/h (for ethanol)as compared to 5 to 10 g/l/h for current conventional processes. Similar

1-1

advances have been noted using a novel multi-membrane bioreactor design inwhich the chronic and ubiquitous problem of product feedback inhibition hasbeen significantly relieved. Substantial progress, in rates and extent ofsynthesis, has also been reported in the technology of enzyme reactions inorganic solvents (as opposed to conventional aqueous medium) for promotingappropriate chemical transformations (e.g., regiospecific oxidation ofaromatics). The Process Design and Analysis work element focuses ondeveloping user-friendly computer programs which assess the energetics andeconomics of biocatalyst chemical production processes. Modular computerprograms have been developed which provide capabilities for rapid assessmentof energy expenditures and costs for unit operations within biocatalystprocesses.

1-2

SECTION II

PROJECT DESCRIPTION

A. THE ECUT MISSION

The Energy Conversion and Utilization Technologies (ECUT) Program seeksto:

(1) Monitor advances in basic scientific research and evaluateemerging technologies for applicability to energyconservation,

(2) Conduct exploratory development and establish feasibility ofinnovative or revolutionary conservation concepts,

(3) Effect technology transfer to DOE end-use conservation programsand/or to private industry, and

(4) Identify energy conservation research needs and feed thisinformation back to basic research.

In fulfilling these aims, ECUT acts as a transmission in transferringthe power of basic research to the wheels of industry (Figure 2-1). There aremany ways to achieve these aims and since its establishment ECUT has chosen tosupport long-term, high-risk generic research and exploratory development forgeneration of more energy-efficient materials and processes. The BiocatalysisProject is one of five projects that make up the ECUT Program (Figure 2-2).

B. HISTORICAL BACKGROUND

The Biocatalysis Project was established in 1980 as the ChemicalProcesses Project with the Jet Propulsion Laboratory as the lead center. Atthat time the project had two work elements: Catalysis (chemical catalysis andbiocatalysis), and Separation. Within these work elements, several smallcontracts were completed that evaluated the energy efficiencies and potentialapplications of relatively new separation processes, such as supercriticalextraction, membrane separation, and chromatography. Research was alsostarted at the California Institute of Technology (Caltech) on chemicalcatalytic behavior models and kinetics of expression in recombinant-DNAmicroorganisms.

In 1982, the ECUT Program was reorganized and the Chemical ProcessesProject was renamed as the Biocatalysis Project. As the name suggests, theproject emphasis was also changed, now stressing biocatalysis as the primaryresearch focus. The work elements were modified to include: MolecularModeling and Applied Genetics, Bioprocess Engineering, and Process Design andAnalysis. Each of these elements addresses a key technical componentnecessary for the development of more advanced and efficient bioprocesses.These three work elements detail a systematic progression of understandingstarting with the basic understanding of biocatalytic mechanisms at themolecular level, to micro-level effects on process parameters in reactors

2-1

Iexo

VQ

•O

§

O

0)0)oiGO

(1)09co

hOV40)C

C•H

OOS

HOOw

CM

(1)

2-2

11p *Qo SCO Z -cCC X SiUJ O -Q> UJ UJ

Oz ^o§ 8 -SS 1CC N 3-W US-

Iocc<111COUJcc_J

z

icc12

CD<

h-coUJ

X

cc0

oli-es<Q.•

1

M

P

-.£2CO CO

w5>:<

CA

TA

lB

IOC

AT

CO

<ccUJ

oCO13m

I

SC

IEN

CE

ST

HE

RM

AL

<

_j

<z

zzoOcc<

m3z

§ «§ 3« Q)

§

£ -g1- «LU M

? -o<0

5 2_1 0

p! 4J,i 2Z

PLI

UJ

O ^CC 1-,

3 °o £

gCJu

m<-J CM

pu7 ^is <uZO M

si ai£ s:< _ico -"•

m3DCO

UJ_l

<

O<Q

•

2-3

(including effects of integrated separation processes), and finally to thedevelopment and assessment of new process concepts for technology transfer tothe industrial sector. With this information the production of large-volume,low-cost, energy-intensive chemicals from renewable resources may be apractical alternative in the future.

The selection as well as the relationships between each of the workelements has been defined by their scale of action. Research activities areconducted at various scales of action or specific size dimensions. Hence,Molecular Modeling and Applied Genetics, includes research activities at amolecular and cellular level, i.e., a scale of 1 urn and smaller. Tosuccessfully exploit the findings at the molecular and cellular level towardsscale-up, Bioprocess Engineering includes research in the area of engineeringkinetics and control, as well as novel concepts in reactor design. The scalein this work element is generally at the 1-meter dimension or at the reactorlevel. Finally, Process Design and Analysis focuses on activities thatoperate at the entire process level (0.4 hectares or more). A central tenetsubscribed to by the ECUT Biocatalysis Project is that in the acquisition ofan appropriate enabling technology, it is mandatory to integrate crucialresearch findings at the various levels of activity and transfer the findingsto the private sector.

C. RELEVANCE

Three main conclusions can be stated regarding the competitive statusand future potential of the United States petrochemical industry: (1) ManyUnited States industries are losing their competitive edge as the globaleconomy is expanding; the U.S. petrochemical industry is an example of justsuch an industry. (2) The United States must take advantage of itscompetitive strengths if it is to compete effectively in the future. Thisincludes the application of its world-renowned capabilities in basic science.Biotechnology is an example of scientific research whose application hastremendous potential for commercial application. (3) Although biotechnologyis being applied towards the production of specialty chemicals by the chemicalindustry, its application to the far larger commodity chemical market isconsidered too long term and risky to attract research and developmentdollars. This is sensible from the perspective of each individual firm, butit is unclear whether the specialty chemical markets can sustain all of thecompetitors who are restructuring their operations towards specialty chemicalproduction. A key government role in the future competitiveness of theindustry may be the support of generic research which can be applied to thedevelopment of economical commercial processes.

The application of biocatalytic processes to the production of large-scale chemicals would capitalize on the three major areas in which the UnitedStates currently exerts leadership and whose successful exploitation wouldcontribute to its global competitive position in the future. First, theUnited States is considered a world leader in basic research advances; ourlead in molecular biology and biochemistry is unquestioned. Commercialutilization of that lead could increase the competitive advantage of ourdomestic chemical industries. Second, we can exert our leadership further by

2-4

using the tools of macromolecular modeling and energy optimization to designentirely novel biocatalysts and to test such catalysts by design conceptsthrough computer simulation. Such a design strategy would be useless withouta means for inexpensive synthesis; however, the newly developed techniques ofsite-directed mutagenesis provide the methodology to synthesize and testexperimentally these design principles. The use of this design strategy wouldenhance our productivity in the future by significantly and intelligentlyreducing the time spent on experimental and empirical procedures in thelaboratory. Third, the United States is one of the most efficient andproductive producers of agricultural products in the world. Because biomassis often the feedstock for the biocatalyst production of organic chemicals, wehave a clear competitive advantage in biomass production within our domesticagricultural sector.

D. GOAL AND OBJECTIVE

(1) Goal. The goal of the ECUT/Biocatalysis Project is to exploit theUnited States' competitive advantage in biotechnology byfacilitating the production of chemicals efficiently viabiocatalytic processes. The Biocatalysis Project supportshigh-risk applied generic research and development aimed atadvancing an enabling technology base that will allow the rationaldevelopment and scale-up of large biocatalyst chemical productionprocesses. The Biocatalysis Project does this by conductingexploratory development on, and by establishing the feasibilityof, novel, innovative, or revolutionary basic research advances inbiotechnology.

(2) Objective. To meet this goal, the Biocatalysis Project hasfocused on developing predictive models and supporting novelbioprocessing concepts which can be utilized by commercialproducers for large-scale chemical production. The strategy ofthe Biocatalysis Project is to reduce production costs byincreasing product yields, by increasing reactor productivity, andby decreasing energy requirements for production. These problemsare generic in a very large number of biocatalyst processes.

E. PROJECT STRUCTURE AND ORGANIZATION

The activities of the Biocatalysis Project have been incorporated asthree major work elements and a Management Support function. The three majorwork elements and their supporting tasks are:

(1) The Molecular Modeling and Applied Genetics work element focuseson defining optimal microscale parameters for biocatalysts anddeveloping practical applications of basic molecular biologyresearch findings. The primary roles of this work element are toprovide a database for defining kinetic models of biocatalystreactivity, and to develop genetically engineered solutions to thegeneric technical barriers that preclude widespread applicationsof biocatalysis.

2-5

(a) Biocatalysis by Design.

(b) Chromosomal Amplification of Foreign DNA.

(c) Kinetics of Recombinant Cells.

(d) Hyperproduction and Secretion of Polyphenol Oxidase.

(2) The Bioprocess Engineering work element emphasizes defining thebasic engineering relationships between molecular scale events andmacro-level parameters required for designing scaled-upbiocatalyst chemical production processes. Additionally, thiswork element has a responsibility for establishing the technicalfeasibility of critical bioprocess monitoring and controlsubsystems.

(a) Metabolic Engineering.

(b) Immobilized Cell System for Continuous Efficient BiocatalystProcesses.

(c) Enzyme Catalysis in Nonaqueous Solutions.

(d) Applications of Molecular Hydrogen in Chemical Fermentations.

(e) Multimembrane Bioreactor for Chemical Production.

(f) Multiphase Fluidized Bed Bioreactor.

(g) Biocatalyst Hydroxylation in Organic Solvents,

(h) Separation by Reversible Chemical Association.

(i) Bioseparation of Phosphate.

\

(j) Protein Engine~el:ihg for Nonaqueous Solvents.' ~ ~ '~

(3) The Process Design and Analysis work element provides overallassessments, via energetic and economic analysis of biocatalystchemical production processes, and initiation of technologytransfer for advanced bioprocesses.

(a) Bioprocess Synthesis, Integration, and Analysis.

(b) Assessment: Biotechnology and Chemical Production.

F. MANAGEMENT SUPPORT FUNCTION

The Management Support function has three areas of responsibility: TaskManagement and Planning; The Guidance and Evaluation Panel; and IndustryTechnology Transfer.

2-6

(1) The Task Management and Planning area administers and coordinatesthe task elements of the Biocatalysis Project. Theseresponsibilities include: developing statements of work andevaluation criteria; defining the sequence of accomplishments thatwill achieve the outlined objectives; monitoring and evaluatingin-house and contract research progress; preparing andimplementing the Annual Operating Plan (AOP); technical multi-yearplans and budgets; assigning personnel; identifying new areas ofresearch and issues to be addressed; and timely publishing ofresearch results and developments and their dissemination to DOEheadquarters, industry, and other researchers.

(2) The Guidance and Evaluation Panel includes industrial and academicrepresentatives who are leading authorities in the science andtechnology of biocatalysis. The panel is responsible forreviewing ongoing activities and future project plans, as well asevaluating the Biocatalysis Project's performance.

(3) The Industry Technology Transfer area's task is to ensure strong,interactive relationships between the Biocatalysis Project andindustry that will assist the transfer of the enabling technologyrequired by the private sector for the production of chemicals viabiocatalysis.

G. TASK DESCRIPTIONS

(1) Molecular Modeling and Applied Genetics

The Molecular Modeling and Applied Genetics work element iscomprised of four research tasks: Biocatalysis by Design,Chromosomal Amplification of Foreign DNA, Kinetics of RecombinantCells, and Hyperproduction and Secretion of Polyphenol Oxidase.The focus is on defining optimal microscale parameters forbiocatalysts and developing practical applications of basicmolecular biology research findings.

(a) Biocatalysis by Design (W.A. Goddard, III, Caltech). Acritical problem in the design of new. biological systems(biopolymers, biocatalyst) is the prediction of thestructure and properties of the new system. Thus, given aspecific primary sequence of amino acids, prediction of thesecondary and tertiary structures (folding) of thefunctional protein complex is desirable. The problem isthat the conformation is determined by weak forces (van derWaals, hydrogen bonding, electrostatic), with the resultthat an enormous number of conformations give similarenergies. A further complication is that the time scale ofa system refolding from one conformation to another is slow(maybe seconds). Thus, in a molecular dynamics simulation,the system may oscillate about one conformational minimumfor 10" to 10" time steps before it begins to unfoldand refold to a better conformation. As a result, it is not

2-7

yet possible to predict the conformation (tertiarystructure) of even small proteins (say, 50 amino acids or500 atoms). However, to design new biochemical systems, itis essential that we be able to predict the structure sothat the concomitant properties can be properly predicted.

Because of difficulties in predicting structure, the focusof this task was on systems where the conformation wasknown. The various amino acids were then modified and themolecular dynamics techniques of annealing and quenchingwere used to equilibrate the system. However, in designingnew biosystems, it is critical to develop the ability topredict the globally minimum conformations so that moredrastic changes can be utilized in designing new systems.

Three overall approaches are being considered, all of whichcombine random fluctuations (to allow sampling of newregions of conformational space) with molecular dynamics(allowing the forces to guide optimization of localstructure): unconstrained optimization using Monte Carlosimulated annealing and simulated growth techniques combinedwith molecular dynamics; tube-constrained conformationsearches where the centerline of the electron density fromlow-resolution X-ray studies is used as a conformationconstraint for molecular dynamics; and pairwise-constrainedconformation searches where pairwise distances fromtwo-dimensional NMR are used.

(b) Chromosomal Amplification of Foreign DNA (G. Bertani, JPL).Chromosomal amplification (and productivity of recombinantmicroorganisms, discussed below) represent new advances inthe area of biological information stability. Informationis a dominant scientific issue in the case of geneticallyengineered organisms. Since the "new" information isexogenously added to the biological system (microorganism),it is a perturbation from the natural state and, hence, thestability characteristics of the engineered information withrespect to time are central for success. The objective ofthis research is to determine the possibility of solving theplasmid stability problem by inserting the recombinant DNAplasmid directly into the bacterial chromosome. Because theplasmid carrying the desired genetic trait would then bepart of the chromosome, it could not be easily lost as thebacterium multiplies.

(c) Kinetics of Recombinant Cells (J.E. Bailey, Caltech). Thisresearch employs experimental methods in concert withmathematical models at the molecular and population levelfor increasing the productivity of bioreactors usinggenetically engineered recombinant microorganisms. Thereare two main research areas that are necessary forsuccessful development of this method: kinetics of DNAreplication, which determines the rates of growth and

2-8

stability of desired recombinant cells; and control andmonitoring of plasmids, which are the DNA elements thatcontain the genetic instructions needed to make desiredproducts more efficiently and selectively.

Quantitative descriptions (kinetic expressions) of thegrowth and stability (plasmid retention) of recombinantcells are needed to optimize the combination of naturalcellular products and components introduced to maximize theefficiency of the overall system. Although cells can begenetically engineered to increase efficiency, specificity,and product yield, the presence of recombinant DNA can havedeleterious effects on the activity of cells with thedesired characteristics.

(d) Hyperproduction and Secretion of Polyphenol Oxidase (W.V.Dashek and A.L. Williams, Atlanta). The objectives of thisresearch are to: purify intracellular and extracellularpolyphenol oxidases from Coriolus versicolor maintained inliquid culture; and determine if intracellular polyphenoloxidase is either synthesized de novo or activated. Theapproach is to: establish the time courses for theappearance of intracellular and extracellular polyphenoloxidase and subsequently quantify oxidase specific activity;and determine optimum growth conditions of C. versicolor formaximum production of polyphenol oxidase, includingdetermination of cofactor requirements.

(2) Bioprocess Engineering

This work element consists of ten tasks: Metabolic Engineering,Immobilized Cell System for Continuous Efficient BiocatalystProcesses, Enzyme Catalysis in Nonaqueous Solutions, Applicationsof Molecular Hydrogen in Chemical Fermentations, MultimembraneBioreactor for Chemical Production, Multiphase Fluidized BedBioreactor, Biocatalyst Hydroxylation in Organic Solvents,Separation by Reversible Chemical Association, Bioseparation ofPhosphate, and Protein Engineering for Nonaqueous Solvents. Theyare primarily concerned with the definition of basic engineeringrelationships between molecular and microscale events andmacro-level parameters required to design and scale-upbioprocesses.

(a) Metabolic Engineering (J.E. Bailey, Caltech). The objectiveis to formulate different types of mathematical structuresfor analysis and synthesis of directed metabolic changesaccomplished by recombinant DNA genetic engineeringmethods. One approach is the definition of a highlystructured, single-cell model of E. coli which includes manyof the mechanistic details of nutrient assimilation, cellgrowth, and protein synthesis.

2-9

(b) Immobilized Cell System for Continuous Efficient BiocatalystProcessing (C.E. Scott, ORNL). The task objectives are toenhance productivity and operability of a fluidized bedreactor system containing immobilized microorganisms, and toinvestigate bioreactor dynamics, including the formulationand investigation of kinetic properties of biocatalystparticles. This will lead to a better understanding ofreactor behavior and control predictability. The reactorconsists of a column containing immobilized cells; thesubstrate (glucose) and nutrients are passed into the bottomof the column and the product ethanol (in water) iscontinuously withdrawn from the top. The primary advantageof this type of bioreactor is increased rates offermentation resulting from the high concentration of cellsin the reactor.

(c) Enzyme Catalysis in Nonaqueous Solutions (H.W. Blanch, U.C.Berkeley). Techniques for increasing the solubility andtransport rate of organics at the site of enzyme action mustbe found before enzymes will be widely used for productionof large-scale chemicals. The approach proposed here isbased on the use of a second organic liquid, in which thesubstrate or product is soluble to a much greater extentthan in an aqueous system. A second aspect of the use ofenzymes in nonaqueous environments is the potential to runmany reactions "backwards." In cases where Water is areaction substrate, its high activity in aqueous solutiongenerally results in a shift of equilibrium to favor thenormal products of the reaction. In a nonaqueous systemsuch reactions may be forced in the reverse direction, aswater will be present at low concentrations. Examples ofthis type of reaction are esterification and synthesis oftryptophan, an important amino acid.

(d) Applications of Molecular Hydrogen in Chemical Fermentations(P.O. Maxwell, Celgene). Glucose, derived from renewablebiomass, currently has two serious limitations as afermentation raw material: selectivity to form specific endproducts is often low and the variety of end products thatcan be made is limited. It has recently been found that iffermentation is conducted in the presence of hydrogen, theformation of oxidized coproducts is decreased so thatselectivity is increased. In addition, the types of primaryproducts formed may contain less oxygen. Two targetfermentations have been selected to demonstrate theseconcepts: fermentation of glucose to succinic acid whichwill demonstrate a desirable shift in product mixcomposition, and fermentation of glucose to 1,3-propanediolto demonstrate the ability to extend the range of endproducts.

2-10

(e) Multimembrane Bioreactor for Chemical Production (M.L.Shuler, Cornell). A four-layer, multimembrane reactorconcept is being developed: one for cell entrapment, one forsubstrate flow, one for gas flow, and one for the flow of anextractant which selectively removes the product. Taskefforts will include: proof-of-concept experiments;development of a mathematical model relating themicro-environment to reactor performance; development of aprocess simulator from the mathematical model; andexperiments to validate the model and to monitor theinterrelationship of cell physiology to immobilization.

(f) Multiphase Fluidized Bed Bioreactor (B. Allen, BattelleColumbus). This task is directed towards the development ofa novel, fluidized-bed bioreactor for the continuousproduction of low-value, high-volume chemicals. The conceptincorporates the features required to minimize the effectsof end-product inhibition, a fundamental biochemical/microbiological limitation that adversely affects theeconomics of current bioprocess technologies. End-productinhibition decreases biocatalyst activity which leads to lowproductivity, high-volume reactors, and low-productconcentration, which requires capital/energy-intensiveaqueous separation systems.

The bioreactor concept in this task combines the use of animmobilized biocatalyst with in-situ, nonaqueous sorbentseparation to minimize end-product inhibition. This isachieved by the use of multi-phase fluidized-bed (MPFB)bioreactor process. The biocatalyst is immobilized asrelatively large or dense particles (dense phase) which arefluidized in a conventional manner using the fermentationmedia. Either a solid or an insoluble, non-aqueous solvent(entrained phase) is then circulated through the dense bedto remove secreted products. After circulating through aseparation/desorption unit to recover products, theentrained phase is regenerated.

(g) Biocatalyst Hydroxylation in Organic Solvents (A. Klibanov,MIT). As discussed earlier, one of the major obstacles inutilization of enzymes in organic chemistry is the necessityto carry out enzymatic reactions in water, which is a poormedium for most industrial chemical processes. For example,the enzyme polyphenol oxidase has a great potential forselective hydroxylation of aromatic compounds, an oftendesirable but very difficult problem in organic chemistry.However, polyphenol oxidase-catalyzed oxidations cannot beconducted in water because: most phenolic substrates areonly slightly soluble in water; quinones (formed as theproduct of the enzymatic oxidation) spontaneously polymerizein water; and during that polymerization the enzyme ischemically modified and consequently inactivated.

2-11

It has recently been discovered that several differentenzymes can function in nearly anhydrous organic solvents.Under such conditions, polyphenol oxidase can quantitativelyconvert phenols to the corresponding o-quinones which thencan be chemically reduced to catechols. The overall processaffords preparative regioselective hydroxylation of aphenol. This effort includes determination of themechanism, implications and preparative use of the abovephenomenon.

(h) Separation by Reversible Chemical Association (C.J. King,U.C. Berkeley). The objective of this task is to examineand evaluate reversible chemical association, orcomplexation, using organic agents as a method forseparating polar organic substances from dilute aqueoussolutions (e.g., bioprocess product or waste streams). Thegoal is to obtain sufficient understanding of underlyingchemical, equilibrium, and transport behavior to enablerational selection of separating agents, and methods ofimplementation, as well as rational conceptual design andeconomic evaluation. In the production of carboxylic acidsand other products, processes such as fermentation yield lowconcentrations of products in water. The subsequentseparation and purification of the desired products is oftendifficult and energy intensive.

(i) Bioseparation of Phosphate (Rogers, INEL). The phosphateindustry utilizes about 0.3 quads/yr for the separation ofphosphate from apatite ore. Therefore, this task isdirected toward development of bioprocessing forsolubilization and separation of phosphate from ore. Thespecific objectives of the proposed research are to: definea microbiological system which will extract phosphate fromits ore; develop a basic understanding of the biochemicalmechanisms involved; and, through the use of modernbiotechnology, develop a bioprocessing system for transferto the phosphate industry. If a more efficient recoveryprocess can be developed, it could also be applied tophosphate mine waste because ore containing <26% phosphorusis currently considered waste and is used as mine backfill.The approach will be to gain a thorough understanding of thebiochemical interactions which cause the microbial releaseof phosphate from its ore. This will be accomplished byscreening microorganisms obtained from areas of highphosphate content (i.e., phosphate mines, process wastestreams, fertilized agricultural lands, etc.) for theirability to solubilize phosphate. Those organisms which arepositive for the desired trait will be selected for furtherbiochemical studies. When a better understanding of themechanism of solubilization has been established, the workwill be focused on development of methods to enhanceproduction. Enhancement could be through physical/chemicalstimulation, genetic manipulations, or both.

2-12

(j) Protein Engineering for Nonaqueous Solvents (F.H. Arnold,Caltech). The industrial applications of biocatalysts havebeen severely limited by constraints on the solventenvironment of proteins, which normally require an aqueousmedium for effective operation. With the advent ofconvenient methods for altering the amino acid compositionand for synthesis of entirely new proteins, it is worthwhileto consider engineering proteins that would be effective innonaqueous solvents. However, the success of a rationaldesign procedure for constructing proteins to use in organicsolvents depends on understanding relationships amongvarious factors: amino acid sequence, secondary and tertiaryprotein structure, and activity and stability in nonaqueoussolvents. The goal of this research is to begin to definethese relationships, by implementing an integral anditerative protein engineering approach based on the modelhydrophobic protein crambin.

(3) Process Design and Analysis

The Process Design and Analysis work element includes developmentof design and process technology for scaling up bioprocesses tothe pilot-plant level and overall assessments (i.e., systemsanalyses) of biocatalyst chemical production processes. This workelement is comprised of two research tasks: Bioprocess Synthesis,Integration, and Analysis; and Assessment: Biotechnology andChemical Production.

(a) Bioprocess Synthesis, Integration, and Analysis (J.D.Ingham, JPL). The purpose of this task is to assemble aseries of candidate bioprocesses and systematically conductrelevant energy-economic analyses and comparisons of theirprojected potential for commercial development within thenext decade. Bioprocesses to be assessed will includeprojected research advances (e.g., genetically engineeredmicroorganisms, bioreactor modeling and verification,membrane development, and biocatalyst immobilization) forenergy-economic comparative assessments. The fourparameters that determine bioprocess economics and totalenergy requirements are cell density, productivity, productrecovery energy and yield.

(b) Assessment; Biotechnology and Chemical Production (ChemSystems Inc.). The objective of this work is to carry outtechno-economic assessments of specific bioprocesses wheremost of the new technology has already been developed,primarily in support of technology transfer activities.

2-13

SECTION III

FY87 TASK ACCOMPLISHMENTS

A. MOLECULAR MODELING AND APPLIED GENETICS

(1) Kinetics of Recombinant Cells (J. Bailey, Caltech). This researchfocuses on control and monitoring of plasmids and kinetics of DNAexpression. Novel high-speed measurement methods employing flowcytometry are used to characterize the cell population in areactor in terms of cell size and cellular plasmid concentration.These experimental methods are used in concert with mathematicalmodels at the molecular and population level to improveunderstanding of plasmid stability and increasing productivity ofbioreactors using genetically engineered recombinantmicroorganisms.

The single-cell measurement method based upon labelling nowpermits detection of single-plasmid copies in yeast. This isaccomplished through the use of flow cytometry in which lightscattering and fluorescence measurements can be made on a singlecell in a flowing cell suspension. Using this experimentalmethod, the fraction of plasmid-containing cells in therecombinant population can be determined in approximately threehours.

Results of these measurements have been used to develop a plasmidstability model for recombinant populations propagating unstablemulticopy plasmids. The model incorporates plasmid replicationand partition functions, as well as the effect of plasmid presenceon host growth rate. It has been found that the amount of acloned enzyme produced by a genetically engineered cell dependsupon the differences between the rate of enzyme or messenger RNA(mRNA) synthesis and the rate of enzyme or mRNA degradation. Therates of mRNA and protein turnover have been found to besignificantly influenced by changes in the number of plasmids percell, the longevity of the mRNA, the physiological condition ofthe cell, and the frequency of plasmid segregation of the plasmidbetween sister cells at cellular division. For example, as thenumber of plasmids within the cell increases, the quantity ofDNA-directed mRNA increases, which in turn produces higherquantities of gene product (protein). This increase in geneproduct is not gratis: costs include the additional energy tomanufacture the gene, its mRNA, and its specific protein. Thisadditional energy cost places these cells at a competitivedisadvantage in the bioreactor when compared to non-plasmid-containing cells, leading to product production failure. Becauseplasmids are transferred between cells, there is also a chance fortheir loss in division. Normally the genetic content of theplasmid is adjusted such that the cell is dependent upon itspresence for its survival. It is then assumed that if the plasmidis lost, the necessary material (enzyme) for survival is lost and

3-1

the cell will be lost; Yet it has now been shown that theavailability of the enzyme depends on the half life of the raRNA.The longer the half life of the mRNA, the longer the cell cancontinue to grow without a plasmid. The cell that lacks a plasmidis not then penalized and can grow for some time withoutcontributing to the reactor performance. This same phenomenon ofcontinued growth without a plasmid can also occur if the quantityof mRNA is high. Thus, the overall productivity of the reactordepends on a number of variables. This model now allows workersin the field to estimate the impact of plasmid instability onbioreactor performance.

(2) Chromosomal Amplification of Foreign DNA (G. Bertani, JPL).Another approach to obviate the intrinsic instability ofplasmids, and therefore of any foreign genetic material that isbuilt into them, is to insert a copy of the plasmid, with whatevergenes are carried by it, into the chromosome of the bacterium.This approach has now been successfully demonstrated. To increasethe number of copies of the genes thus introduced on thechromosome, and make the structure equivalent to that of a typicalbacterium loaded with plasmids, an indirect method of selectionfor chromosome amplification was applied, also successfully. Suchchromosomally amplified structures can be undone by normal geneticrecombination, however. During the year, attempts at reducing theeffects of normal genetic recombination have been made, byintroducing mutations that reduce recombination in the bacterium.This work is still in progress. Another problem that is beinginvestigated is the unexpectedly low growth rate of those strainswhere plasmid amplification has been produced within thechromosome.

(3) Hyperproduction and Secretion of Polyphenol Oxidase (W.V. Dashekand A.L. Williams, Atlanta University.). Two Ph.D students haveprepared task plans that define each of their activities that willbe implemented to accomplish the objectives of this research.Experimental work has been started to determine kinetics ofproduction of extracellular oxidases from C. versicolor, establishoptimum growth conditions, and compare specific activities for theenzyme utilizing different phenolic substrates.

Experiments have been completed to unequivocally establish timecourses for the formation of intra- and extracellular polyphenoloxidases (PPO). Cycloheximide addition (to C. versicolorintracellular total protein) inhibited total protein levels by57.4% and enhanced PPO activity by a factor of 2.3, to suggestthat PPO is activated, rather than synthesized de novo. Protocolsfor the purification of extracellular PPO are being developed.Presentations of this and related work were made in St. Louis, MOand at the University of Cambridge, England.

3-2

B. BIOPROCESS ENGINEERING

(1) Protein Engineering for Nonaqueous Solvents (F.H. Arnold,Caltech). The goal of this research is to begin to define therelationship between the amino acid sequence and proteinstructure, based on the model hydrophobic protein crambin. Theunusual stability and solubility of crambin in a wide range ofnonaqueous solvents will be investigated. The structures,stabilities, and solubilities of crambin mutants will be measuredand correlated with specific alterations in the protein amino acidsequence. This research includes studies of the properties ofcrambins, design of new mutant crambins by computer/moleculargraphics, production of new mutants by site-directed mutagenesis,and detailed NMR studies of crambin stability and response tononaqueous solvents. The results of this work will be used toformulate a set of criteria and a method for rational design ofproteins to be used in the presence of nonaqueous solvents.

The synthetic gene for crambin has been expressed in E. coli. Thegene was inserted into pKK223-3, containing the tac promoter.Crambin is expressed intracellularly at reasonably high levels.The protein produced in the bacteria is reduced and unfolded.Attempts to refold the recombinant crambin in a mixture ofoxidized and reduced thiols have been successful. The resultingprotein has an NMR proton spectrum very similar to the nativematerial. The genes for five variants of crambin have beenproduced, including one with an additional disulfide bridgebetween residues 12 and 30. A preliminary set of design criteriafor engineering proteins to be stable in nonaqueous solvents hasbeen proposed and presented at the IX Enzyme EngineeringConference in. Santa Barbara.

(2) Enzyme Catalysis in Nonaqueous Solutions (H.W. Blanch, U.C.Berkeley). This research pursues conversion of cholesterol tocholestenone by using an enzyme (with and without enzyme-waterencapsulation) dispersed in an organic solvent. Kinetics ofenzymatic conversion in a two-phase system (toluene-water, withoutencapsulant) have been determined as a basis for comparison withthe encapsulated system and determination of the effects ofsurfactant concentration and stirring rate. Oxidase activity at25°C decreased after one day, probably because of denaturationat the solvent interface, followed by no further decrease beforetermination after 5 days. Although the surfactant (Span 85, >0.5vol%, used to disperse the solvent in the aqueous phase) increasedthe initial rate by a factor of about five, the effect ofsurfactant on interfacial enzymatic activity is not wellunderstood. Therefore, the effect of a biological surfactant,phosphatidylcholine (PC), on the rate was determined, and therelative effects of the two surfactants were rationalized on thebasis of differences in their chemical structures. A model wasdeveloped to account for various effects on rates of reaction in atwo-phase system, and oxidase activity assay and control of pH inthe two-phase system were investigated.

3-3

A method has been developed to avoid capsule membrane swellingwhen the capsules are transferred from the encapsulant formationsolvent (cyclohexane) to the reaction solvent (toluene orbenzene), and to control pH within the microcapsules, by usingcyclohexane-toluene (1:1) as a common solvent: the mixed solventdoes not cause swelling and allows a higher acid concentration inthe solvent. In the oxidase-catalyzed conversion of cholesterolto cholestenone within microcapsules, control of pH is required tomaximize enzymatic activity. To provide for transfer of acidacross the membrane and into the capsules, they are exposed to HC1with Aliquat 336 in the organic phase. A non-invasive techniquehas been developed to monitor acidity within the capsules by usingNMR (31p) spectrometry to determine encapsulated mono- anddihydrogen phosphate. The relative amounts of each phosphate typeare then used to calculate the pH inside the capsules. The pH formaximum oxidase activity was determined to be from 6.0 to 6.5.

Additional research concerns tryptophan synthesis and extractivecatalysis. Although it was originally planned to study ureasynthesis, attention has been redirected to tryptophan formationas a model to minimize experimental difficulties (e.g., definitionof an appropriate solvent system for urea, and development of afundamental understanding of extractive catalysis and the effectof product water on reaction equilibria). In this conversion,ammonium pyruvate or serine plus indole react to form tryptophanand water in the presence of tryptophanase. The normal reactionis hydrolysis, because of the high effective concentration ofwater in most physiological environments. A liquid membrane(cyclohexane) or micellar solution is used to isolate the enzymein an aqueous environment at pH 9 from the aqueous outer phase atpH 7, into which the product is extracted, as a result of the pHgradient across the liquid organic membrane. (Micelles consist ofmicroscopic drops of water within an organic phase, which isdispersed as an emulsion in water.) Because of limited ammoniatransport across the organic phase, it was found that serine,rather than ammonium pyruvate, must be used in a micellar 'environment.

A model has been developed and experimentally verified to describemass transfer kinetics for substrate and products of this system.Another significant accomplishment is a detailed description andclarification of the effects of solvents on the equilibria ofenzymatic reactions in two-phase systems, which will be extremelyuseful in the definition of future biocatalyst conversionprocesses based on extractive catalysis (i.e., catalyzed reactionsin two-phase systems). In this work, equilibrium equations havebeen derived in terms of partition coefficients and activities ofthe reactants; these equations can be used to predict productyields in two-phase systems. Further work on tryptophan synthesisincludes initiation of studies using EPR spectrometry and spinlabels to investigate water and enzyme interactions (e.g.,conformational mobility of enzymes within micelles).

3-4

(3) Biocatalyst Hydroxylation in Organic Solvents (A.M., Klibanov,MIT). The purpose of this research is to develop new methods forenzymatic oxidative. catalysis in organic .solvents, e.g., forselective oxidation of aromatic compounds to produce quinones andhydroxylated aromatic chemicals, for peroxidase-catalyzeddepolymerization of lignin, and for potential polymerization ofaromatic compounds.

Conditions for maximum enzymatic depolymerization of polyconiferylalcohol as a model for lignin have been determined. These includeutilization of a solvent consisting of 95% dioxane in water at apH of 5.0 with 0.5 mg/ml of the enzyme (horseradish peroxidase)and 300 ug/ml of polyconiferyl alcohol. Alternative solvents werealso investigated. While peroxidase-catalyzed depolymerization ofsynthetic lignin did not take place in aqueous solutions,significant depolymerization was accomplished in organic solventsusing either horseradish or lactoperoxidases. A mechanisticevaluation of the ability of horseradish peroxidase to catalyzethe cleavage of specific lignin interunit linkages using lowmolecular weight model compounds indicated a preference forbenzylic ether cleavage over aryl-alkyl cleavage if both werepresent. Molecular oxygen was shown to stimulatedepolymerization, and this was interpreted to be the result offormation of free radical lignin species produced by theperoxidase-initiated depolymerization reactions. The ability ofhorseradish peroxidase to depolymerize lignin in organic solventsmay enable useful low molecular weight chemicals to be producedeconomically from lignin or lignin waste products.

The depolymerization of lignin has been extended to show thatextracted lignins from milled wood or kraft pine can besuccessfully depolymerized using the same methods as forpolyconiferyl alcohol. Also, milled wheat straw has beendelignified with a modified, dioxane-soluble peroxidase. Nodepolymerization took place in aqueous solutions, to show theadvantage of enzymatic reaction in organic solvents. Studies havealso been initiated on horseradish peroxidase-catalyzed reactionsto polymerize phenols in organic solvents as an alternative tophenol-formaldehyde resins. Two inventions developed in this workhave been licensed for commercial applications: EnzymaticReactions in Liquid and Solid Paraffins: Application forTemperature Abuse Sensors; and Method and Apparatus for DetectingCompounds Utilizing Enzyme-Catalyzed Reactions in the GaseousPhase.

Separation by Reversible Chemical Association (C.J. King, LBL,U.C. Berkeley). In the production of carboxylic acids, processessuch as fermentation yield low concentrations of carboxylic acidsin an aqueous multicomponent solution. The subsequent separation,purification, and concentration of the carboxylic acids isdifficult and energy intensive.

3-5

A novel processing approach for recovery of low-volatilitycarboxylic acids has been developed. This method involves use oftwo diluents that are substantially more volatile than the amineand the acid. One (e.g., chloroform) is an "active" diluent,which solvates the complex and greatly increases the equilibriumdistribution coefficient. The other (e.g., n-heptane) is an"inert" diluent, which does not interact with the complex andgives lower distribution coefficients. The active diluent ispresent during extraction of the acid from water. It is thenremoved by distillation and replaced by the inert diluent, whichis present during back-extraction of the acid into water. Theinert diluent is then removed by distillation and replaced withthe active diluent for reuse of the solvent mixture. Equilibriummeasurements for extraction of succinic acid with amine present ina diluent containing different proportions of chloroform andheptane show that this approach is promising. Extraction oflactic acid by Alamine 336 in chloroform and m-cresol diluentsshows very high distribution coefficients, which indicates thatthis system can be applied to energy-efficient recovery of lacticacid.

Fixed-bed adsorption runs have been made with Porapak Q (WatersAssoc.) to determine the effect of particle size on adsorption.Succinic acid, acetic acid, and 1,3-butanediol have been adsorbedfrom aqueous solutions, using the bed in both wetted andnon-wetted conditions. In the latter case, separation follows theorder of volatilities. This enables a separation among solutesthat is not possible in the wetted state. The shapes ofbreakthrough curves for the non-wetted state display a number oftransfer units about a factor of 10 less than expected.Preliminary indications are that this is the result ofchannelling, caused by preferential wetting of the wall of theglass column. The possibility of using a polycarbonate column isbeing investigated. Sorption properties of a large number ofbasic resins have also been examined experimentally.

Preliminary infrared spectroscopic measurements have been made forsuccinic acid dissolved in dry amine with chloroform diluent totry to confirm the presence of specific complexes. Equilibriumdata have been obtained for extraction of ethanol by variousphenolic compounds in different diluents to attempt to identifystructural factors that lead to favorable equilibria. Manyphenolics are not amenable as alcohol extractants because they aresolids at practical temperatures that cannot be dissolved indiluents at high concentrations. Alkyl groups on the phenoldecrease solubility loss in water, but distribution coefficientsare unfavorably decreased. It was also determined that there isno significant advantage in using pentachlorophenol ordihydroxybenzenes. Sorption properties of various basic resins(e.g., polyethyleneimine, epoxy polyamines, and polyvinylpyridine)have been examined experimentally to .identify attractiveregenerable sorbents for carboxylic acids.

3-6

(5) Metabolic Engineering (J. Bailey, Caltech). The objective of thisstudy is to formulate mathematical models [i.e., the MetabolicPathway Synthesis Program (MPS)] for analysis of directedmetabolic changes accomplished by recombinant DNA methods. TheMPS program can be used to predict, in a qualitative way, theeffect of adding or deleting enzymatic activities to or from thecellular environment, to classify pathways with respect tocellular objectives, arid to extract information about metabolicregulation. MPS can be used for the identification of appropriategenotypes or genetic modifications that will redirect metabolismtowards amplified production of desirable bioproducts. Thesecapabilities have been illustrated by synthesizing the classicalconversion of glucose 6-phosphate to pyruvate (Embden-Meyerhof-Parnas, pentose phosphate, and Entner-Doudoroff pathways), andbiosynthesizing L-alanine that does not necessitate the use ofalanine aminotransferase from pyruvate.

(6) Multiphase Fluidized Bed Bioreactor (B. Allen, Battelle MemorialInstitute, Columbus). A Multiphase Fluidized Bed Reactor (MPFB)was built by Battelle Memorial Institute and its design andoperation was tested. This effort was successful at elucidatingmany operational characteristics of fluidized beds used asbioreactors for the bacterium Clostridium acetobutylicum.

The bacterium C. acetobutylicum was successfully encapsulated inalginate beads with bead sizes ranging from 0.5 mm to 3 mm (andlarger). The inclusion of aluminum oxide or iron was alsosuccessfully achieved to create beads with greater density, whichoperate more effectively as the dense phase in the fluidized bed.The bacterium was active and was viable for over 300 hours ofcontinuous operation with active production of butanol for over240 hours. Operationally, it was found that the small diameterbeads were required for successful fluidization. The specificorganism was chosen because of its acute sensitivity to productinhibition. Thus, the relief of such inhibition by the sorbentextraction in this system should be better observed.

The operational mode for the extractive or sorbent phase wasexamined. It was determined that a liquid sorbent which operatesin a co-current mode would be best to test the concept. Severalcandidate sorbents were chosen based on published distributioncoefficients for butanol in a biphasic liquid system and testedfor toxicity towards cells and for actual distribution coefficientvalues. Some sorbents were found that possessed appealingdistribution coefficient values. However, most were found to betoxic to the cells and the final choice was made to use oleylalcohol. Although not the best choice for processing, it serveswell to test the feasibility of this concept.

An actual working prototype of the MPFB system was built and usedfor cold flow model testing in order to obtain operational data.It was 3" in diameter by 6' in height with a working volume ofabout 7 to 8 liters. Bed expansion volumes, velocity

3-7

measurements, pumping and circulation requirements, and sorbentrecirculation parameters were measured. A novel measurement toolwas employed to measure bed and solvent velocities: a Fiber OpticLaser Doppler Anemometer. This instrument permitted anon-invasive measurement of fluid flows, although it had arestricted range. A major result of this work was to show thatthe fluidized bed operates as a reactor that is intermediatebetween plug flow and backmixed operation. These data will behelpful in designing larger units for scale-up.

The prototype of the MPFB was reconfigured to be sterilizable,thus permitting operation with bacterial cells. This unit willprovide a unique facility for testing authentic microbial cultureswith the MPFB concept. This unit has undergone initial testingand operation to eliminate "bugs" prior to the actual feasibilitytests. Tests to demonstrate sorbent extraction of butanol in thefluidized bed were conducted. The results indicated successfulpartitioning of butanol into the oleyl alcohol from beadscontaining butanol (no microbial encapsulation in this case).

An independent technoeconomic analysis of the MPFB concept wasconducted by the engineering firm, Bio En-Gene-Er Assoc., Inc.Their analysis was quite favorable and concluded that thisfluidized bed concept could increase the effective concentrationof the product ten-fold while maintaining actual concentrationsnear the organism below the threshold of inhibition. As currentlyconfigured, the MPFB would reduce the cost of butanol by about 34%(to yield a 30% pretax return on investment). Furtheroptimization could reduce this cost even further. Guidelines forfurther research were proposed, including a need for betterproduct stoichiometry from the microbe and an improvement in thesorbent.

(7) Multimembrane Bioreactor for Chemical Production (M. Shuler,Cornell). The Multimembrane Bioreactor with extraction was shown

- to operate and exhibit improved performance over-conventionalbatch and immobilized fermenters. In the MMB system withextraction, glucose concentration nearly approached zero, loweringproduction costs. In addition, ethanol productivity wasincreased by a factor of 5 to 7. Operational problems with thissystem were encountered which were overcome by defining nutrientrequirements. One new concept utilizing the MMB system wasattempted by operating the system with pressure cycling in the gasphase. It was anticipated that such operation would enhance masstransfer and permit increased cell densities. Experiments wereconducted to test this hypothesis.

It was subsequently demonstrated that a five-fold improvement inaverage fermentation rate for a 200 g/1 glucose feed for apressure cycle reactor could be achieved over the normaldiffusion-controlled system (8.7 g glucose consumed/hr versus1.8 g glucose consumed/hr). The pressure cycle removes ethanolmore efficiently from the cell layer (2.8 g/l/hr versus 1.1

3-8

g/l/hr). Additional characteristics of this reactor system wereexamined, such as the effect of high cell densities on operationand operation at high temperatures (above 40°C).

The effect of high cell densities on the reactor system werestudied. The analysis evaluated growth rate as a function of celldensity. The apparent growth rate at 41 g/1 was roughly half ofthat at 9 g/1 (0.16 versus 0.29/hr). The reduction appeared to bedue to cell contact blocking uptake sites at high cellconcentrations. The operation of the pressure cycle reactor athigh temperatures was also evaluated. Previous attempts tooperate such fermentations at higher temperatures in the diffusionmode showed non-performance at 40° and 45°C. In the pressurecycle mode, operation at 40°C was very effective (i.e., rapidglucose consumption), but at 45°C cell viability was lost andthe fermentation became "stuck" (or inhibited). Work with thepressure cycled MMB system is now being directed towardscontinuous operation and/or increasing the size of the system byoperating two or more units in series.

Additional areas of research involved an analysis of theseparations and aeration requirements for this extractive system,and modeling of the ethanol yeast fermentation system.Preliminary experiments to evaluate separations were carried out.In a simple two-step distillation with "spent" TBP and ca. 30 g/1of ethanol and ca. 60 g/1 of water in the TBP, a top product with80 wt% ethanol and virtually no TBP was recovered. If the spentTBP was dehydrated with MgSO^ then the top product is nearlypure ethanol. These results indicated that purification isachievable. The effects of aeration on the specific yeast strainemployed were studied. This strain operated poorly under strictlyanaerobic conditions, but required a level of oxygen at 2 to 3% of100% saturation.

An initial formulation of a mathematical model for yeast growthand metabolism was completed. After considerable work andimprovement, the model is considered to be one of the mostcomplete and powerful simulators of yeast behavior available. Itcan accurately predict transient behavior in batch culture, apredictive behavior not found in models currently available. Themodel will continue to be improved until the end of this effort tobetter fit empirical data obtained in the laboratory.

C. PROCESS DESIGN AND ANALYSIS

(1) Bioprocess Synthesis, Integration, and Analysis (J. D. Ingham,JPL). The purpose of this task is to derive or synthesize aseries of candidate bioprocesses and systematically conductrelevant energy-economic analyses and comparisons for added-valuecommodity chemicals to determine potential for process technologytransfer and commercial development. These bioprocesses will be

3-9

modified to include projected research advances (e.g., geneticallyengineered microorganisms, bioreactor modeling and verification,membrane development, and biocatalyst immobilization) forenergy-economic comparative assessments. The four parameters thatdetermine bioprocess economics and total energy requirements arecell density, productivity, product recovery energy, and yield.

Extensive utilization of bioprocesses for the production ofcommodity chemicals could significantly decrease U.S. dependenceon the supply of foreign and domestic non-renewable petroleumresources, but because of relatively high feedstock costs and somespecific limitations of biocatalyst processes, most of them arenot yet competitive with current petrochemical routes with respectto either process energy consumption or economics. Since theselimitations often result in either relatively low rates ofreaction and production, low product concentration, or both,assessments of a series of processes for ethanol (as a model of ahigh volume industrial chemical) have ,been completed to establisha more quantitative basis for estimation of the relative effectsof production rates and product concentration in the bioreactor onprocess energy and economics. The results of model bioprocesssimulations for ethanol show that energy consumption is sensitiveto product concentration, is essentially independent of rate ofproduction, and that the effect of production rate on economics isat least an order of magnitude less than the effect of productconcentration. It is also suggested that concentration is likelyto be even more important for other potential industrial chemicalbiocatalyst processes (e.g., for n-butanol or acetic acid) whereproduct recovery is more energy intensive than for ethanol. Itcan be concluded that future research advances to improve theenergy, economics, and effective synthesis of new, competitivebiocatalyst processes for industrial chemicals should be directedtoward significantly increasing product concentration, rather thanby attempting to independently maximize only rates of production.

A computer model has been-formulated-where fundamental-kineticequations and intrinsic parameters (e.g., maximum specific rates,saturation constants, and yield factors) result in fixedrelationships with respect to conversion, product yield, productconcentration, and productivity. Calculations show that there aremaximum limits for these parameters for specific bioprocesses: thehighest productivity normally will occur at a relatively low yieldand concentration. Therefore, investigations have been initiatedto establish reactor configurations and conditions that result inminimum energy requirements and production costs. Preliminarycalculations for simulated processes of rate, yield, andconcentration combinations agree with previous experimentalobservations. This bioreactor kinetic study will be completed andused to attempt to, confirm that earlier assessments are realisticand consistently applicable to potential advanced bioprocesses, toensure that assumed reactor conditions are reasonable, and thatprojected improvements could be realized.

3-10

SECTION IV

BIBLIOGRAPHY

Allison, J. M., and Goddard, W. A., "Oxidative Dehydrogenation of Methanol toFormaldehyde," J. Catal.. Vol. 92, pp. 127-135, 1985.

Allison, J. M., and Goddard, W. A., "Active Sites on Molybdate Surfaces,Mechanistic Considerations for Selective Oxidation, and Ammoxidation ofPropene," ACS Symposium Series No. 279, pp. 23-36, 1985.

Bertani, L. E., and Bertani, G., "Chromosomal Integration and Excisionof a Multicopy Plasmid Joined to Phage Determinants for Site SpecificRecombination," in Plasmids in Bacteria, D. R. Helinski, ed., PlenumPress, NY, p. 917, 1985.

Bertani, L. E., and Six, E. W., "The P2-like Phages and Their Parasite, P4,"to appear in The Viruses, editors; Wagner, R. R., and Fraenkel-Conrat, H., Plenum Publishing Corp, 1988.

Biocatalysis Research Activity Annual Report; FY 1983, Jet PropulsionLaboratory, JPL Publication 84-27, DOE/CS-66001-4, April 15, 1984.

Biocatalysis Project Annual Report; FY 1984, Jet Propulsion Laboratory,JPL Publication 85-31, DOE/CS-66001-8, April 15, 1985.

Biocatalysis Project Annual Report; FY 1985, Jet Propulsion Laboratory, JPLPublication 86-32, DOE/CS-66001-9, July 30, 1986.

Biocatalysis Project Annual Report; FY 1986, Jet Propulsion Laboratory,JPL Publication 87-11, DOE/CS-66001-10, April 30, 1987.

Celniker, S., Sweder, K., Srienc, F., Bailey, J. E., and Campbell, J. L.,"Deletion Mutations Affecting the Yeast Autonomously ReplicatingSequence, ARSI," Mol. Cell. Biology, Vol. 4, p. 2455, 1985.

Chem Systems, Inc., "Technical and Economic Assessment of Processes for theProduction of Butanol and Acetone - Phase Two: Analysis of ResearchAdvances," JPL 9950-928, DOE/CS-66001-5, Prepared for Jet PropulsionLaboratory by Chem Systems, Inc., Tarrytown, NY, August 1984.

Chem Systems Inc., "Techno-Economic Assessment of Microbial AmmoniaProduction," prepared for Jet Propulsion Laboratory by Chem SystemsInc., Tarrytown, NY, February 1986.

Chem Systems Inc., "Assessment of ORNL Immobilized Cell Ethanol FermentationProcess," prepared for Jet Propulsion Laboratory by Chem Systems Inc.,Tarrytown, NY, March 1986.

Chem Systems Inc., "Implications of ORNL Immobilization Technology on CitricAcid Production," prepared for Jet Propulsion Laboratory by Chem SystemsInc., Tarrytown, NY, June 1986.

4-1

Cho, T., and Shuler, M. L., "Multi-Membrane Bioreactor for ExtractiveFermentation," Biotech. Prog., Vol. 2, pp. 53-60, 1986.

Davison, B. H., and Scott, C. D., "Ethanol Production on an IndustrialFeedstock by Zymomonas Mobilis in a Fluidized Bed Bioreactor," EighthSymposium on Biotechnology for Fuels and Chemicals, Gatlinburg, TE,May 13-16, 1986.

Davison, B. H., and Scott, C. D., "Fermentation Kinetics of Zymomonas mobilisunder Mass Transfer Limitations," 192nd ACS National Meeting, Anaheim,CA, September 7-12, 1986.

Eggers, D. K., Lira, D. J., and Blanch, H. W., "Enzymatic Production of1-Tryptophan in Liquid Membrane Systems," submitted to Bioprocess Eng.

Eggers, D. K. , Blanch, H. W., and Prausnitz, J. M., "Extractive Catalysis:Solvent Effects on Equilibria of Enzymatic Reactions in Two-PhaseSystems," submitted to Bioprocess Eng.

England, C., "Isothermal Separation Processes Update," JPL 9950-927,DOE/CS-66001-6, prepared for Jet Propulsion Laboratory by EngineeringResearch West, Santa Monica, CA, August 1984.

Goddard III, W. A., Low, J. J., Olafson, B. D., Redondo, A., Zeiri, Y.,Steigerwald, M. L., Carter, E. A., Allison, J. N., and Chang, R., "TheRole of Oxygen and Other Chemisorbed Species on Surface Processes forMetals and Semiconductors; Approaches to Dynamical Studies of SurfaceProcesses," Proceedings of the Symposium on the Chemistry and Physics ofElectrocatalysis, Vol. 84-12, The Electrochemical Society, Ind.,Pennington, NJ, pp. 63-95, 1984.

Goddard III, W. A., "Theoretical Chemistry Comes Alive: Full Partner withExperiment," Science. Vol. 227, pp. 917-923, 1985.

Hatcher, H. J., "Biocatalysis Program Bioseparations Task," Summary Report,E. G. and G. Idaho, Inc., Idaho National Engineering Laboratory, IdahoFalls, ID, February 1986.

Ingham, J. D., Biocatalyst Processes for Production of Commodity Chemicals;Assessment of Future Research Advanced for N-Butanol Production, JetPropulsion Laboratory, JPL Publication 84-60, July 1, 1984.

Kertes, A. S., and King, C. J., "Extraction Chemistry of Fermentation ProductCarboxylic Acids," Biotechnology and Bioengineering, Vol. 28,pp. 269-282, 1986.

King, C. J., chapter on Chemical Association in Handbook of SeparationProcess Technology, R. W. Rousseau, ed., John Wiley and Sons, 1987.

King, C. J., et al., "Non-wetting Polymeric Adsorbents for Selective Uptake ofCarboxylic Acids and Glycols," presented at the Symposium, Fundamentalsof Adsorption, Meeting of the American Chemical Society, Denver,Colorado, April 1987,.and Foundation Conference on Advances inSeparation Technology, West Germany, April 1987.

4-2

Klibanov, A. M., "Polyphenol Oxidase - Catalyzed Hydroxylation of AromaticCompounds in Organic Solvents," Quarterly Report No. 1, MassachusettsInstitute of Technology, Cambridge, MA, March 1986.

Kuhn, R. H., and Linden, J. C., "Effects of Temperature and Membrane FattyAcid Composition on Butanol Tolerance of Clostridium acetobutylicum,"Eighth Symposium on Biotechnology for Fuels and Chemicals, Gatlinburg,TE, May 13, 1986.

Lee, S. B., Seressiotis, A., and Bailey, J. E., "A Kinetic Model for ProductFormation in Unstable Recombinant Populations," Biotechnology and Bio-engineering, Vol. 27, pp. 1699-1709, 1985.

Lee, S. B., and Bailey, J. E., "A Mathematical Model for v Plasmid Replication:Analysis of Wild-Type Plasmid," Plasmid. Vol. 11, No. 151, 1984.

Lee, S. B., and Bailey, J. E., "A Mathematical Model for v Plamid Replication:Analysis of Copy Number Mutants," Plasmid, Vol. 11, No. 151, 1984.

Lee, S. B., and Bailey, J. E., "Analysis of Growth Rate Effects on Productivityof Recombinant Escherichia coli Populations Using Molecular MechanismModels," Biotechnol. Bioeng.. Vol. 26, No. 66, 1984.

Lee, S. B., and Bailey, J. E., "Genetically Structured Model for Lac Promoter-Operator Function in the Chromosome and in Multicopy Plasmids: LacPromoter Function," Biotechnol. Bioeng., to be published.

Linden, J. C., and Kuhn, R. H., "Biochemistry of Alcohol Effects onClostridia," in Ethanol Toxicity at the Cellular Level, Van Uden, N.,ed., CRC Press, Cleveland, OH, in press.

Miskowski, V. M., "The Electronic Spectra of u-Peroxodicobalt (III)Complexes," Comments Inorg. Chem., in press.

Peretti, S. W., and Bailey, J. E., "A Mechanistically Detailed Model ofCellular Metabolism for Glucose-Limited Growth of Escherichia coliB/r-A," Biotechnol. Bioeng.. Vol. 28, p. 1672, 1986.

Peretti, S. W., and Bailey, J. E. , "Simulations of Host-Plasmid Interactionsin Escherichia coli; Copy Number, Promoter Strength, and RibosomeBinding Site Strength Effects on Metabolic Activity and Plasmid GeneExpression," Biotechnol. Bioeng., Vol. 29, p. 316, 1987.

Scott, C. D., "Techniques for Producing Mono-dispersed Biocatalyst Beads forUse in Columnar Bioreactors," Proceedings of the Seventh Symposium onBiotechnology for Fuels and Chemicals, Gatlinburg, TE, May 16, 1985.

Scott, C. D., "Innovative Techniques for Gel Immobilization of Microorganismsand Enzymes," International Conference on Enzyme Engineering, Helsingor,Denmark, September 1985, to be published in The Annals of the New YorkAcademy of Sciences.

4-3

Scott, C. D., and Thompson, J. E., "Mass Transfer Properties of BiocatalystBeads," Eighth Symposium on Biotechnology for Fuels and Chemicals, .Gatlinburg, TE, May 13-16, 1986.

Seo, J. H., and Bailey, J. E., "Effects of Recombinant Plasmid Content onGrowth Properties and Cloned Gene Product Formation in Escherichiacoli," Biotechnology and Bioengineering, Vol. 27, pp. 1668-1674, 1985.

Seressiotis, A., and Bailey, J. E., "Optimal Gene Expression and AmplificationStrategies for Batch and Continuous Recombinant Cultures," Biotechnol.Bioeng., Vol. 29, p. 392, 1987.

Seressiotis, A., and Bailey, J. E., "MPS: An Algorithm and Data Base forMetabolic Pathway Synthesis," Biotechnol. Left.. Vol. 8, p. 837, 1986.

Srienc, F. , Campbell, J. L., and Bailey, J. E., Biotechnol. Letters, Vol 5,p. 43, 1985. .

Srienc, F., Bailey, J. E., and Campbell, J. L., "Effect of ARSI Mutations onChromosome Stability in Saccharomyces Cerevisiae," Mol. Cell. Biology,Vol. 5, No. 7, pp. 1676-1684, 1985.

Srienc, F., Campbell, J., and Bailey, J. E., "Flow Cytometry Characterizationof Plasmid Stability in Saccharomyces cerevisiae," Third EuropeanCongress on Biotechnology, Verlag Chemie, Weinheim, Federal Republic ofGermany, Vol. Ill, pp. 111-187, 1984.

Srienc, F., Campbell, J. L., and Bailey, J. E., "Analysis of UnstableRecombinant Saccharomyces cerevisiae Population Growth in SelectiveMedium," 1984, submitted for publication.

"System for Biotechnology Assessment," Final Report, Phase I; August 1984 -July 31, 1985, Washington University, St. Louis, MO, August 1985.

Tamada, T. A., Kertes, A. S., and King, C. J., "Solvent Extraction of SuccinicAcid from Aqueous Solutions," International Conference on Solvent •Extraction, Munich, Germany, September 1986.

The ECUT Biocatalysis Project. Jet Propulsion Laboratory, ECUT ProgramBulletin, DOE, ECU-86/1, PB86-900401, Jan.-Feb., 1986.

The ECUT Biocatalysis Project, Jet Propulsion Laboratory, submitted forPublication in the ECUT Program Bulletin, May 1987.

Warren, L. F., "Electrocatalysis Research," Final Report for the PeriodAugust 25. 1982 through December 30, 1983, JPL 9950-926, DOE/CS-66001-7,prepared for Jet Propulsion Laboratory by Rockwell International,Science Center, Thousand Oaks, CA, August 1984.

Wilcox, R. E., Industry, University, and Research Institute Interest in theU.S. Department of Energy ECUT Biocatalysis Research Activity, JetPropulsion Laboratory, JPL Publication 83-90, DOE/CS-66001-3,November 1, 1983.

4-4