1 METABOLIC ENGINEERING OF Escherichia coli TO EFFICIENTLY PRODUCE SUCCINATE IN MINERAL SALTS MEDIA By KAEMWICH JANTAMA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008
2 2008 Kaemwich Jantama
3 To my parents, Prateep and Nongluck Jantama. Their love, laughter, support, and sacrifice have been constant. Because of this, in all aspects of my life, I am the luckiest person in the world. I am truly grateful.
4 ACKNOWLEDGEMENTS First and forem ost, I would especially like to express my sincere gratitude and deep appreciation to to my microbiology advisor, Di stinguished Professor Lonnie O. Ingram, for his extensive scientific support, for generously providing research f acilities and also for his helpful advice, constructive comments and guidance in many other areas throughout my dissertation work at Microbiology and Cell Science Departme nt, University of Florida. His extensive knowledge as a researcher and e xperience in academia taught me invaluable lessons that I will carry with me throughout my career. I extend my gratitude to my chemical engine ering advisor, Professor Spyros A Svoronos, for his warm guidance, immeasurab le attention, constant support, and continuous encouragement to persevere through the obstacl es and frustrations of my graduate study at Chemical Engineering Department, University of Florid a. He was an incredibly gifted mentor. I also thank my supervisory committees, Prof essor Keelnathan T Sh anmugam, Professor Ben Koopman, and Associated Professor Yiider Tseng, for their valuable discussions, suggestions, and insights. My generous appreciation goes out to all I ngrams lab members, especially Dr. Xueli Zhang and Jonathan C Moore for their warm frie ndship, support, kindness, helpful suggestions, encouragement, and generous contribution during my research work. Thanks always go to Sean York and Lorraine Yomano for always help keeping chemicals and equipments in proper working order. I feel lucky to have worked with such talented researchers, whose passion for research and friendly nature made every day enjoyable.
5 My heartly thanks must be extended to Shirley Kelly at Chemical Engineering Department, and Janet Lyles at Microbiology and Cell Science Departments for providing excellent and superb help through all my paper work. I am obliged to Ministry of Science, Technology and Environment, Thailand for providing the Royal Thai Government Scholarship throughout six years of my abroad study in the United States of America. I am greatly indebt ed to all Thai people who have devoted me to the most valuable investment. I promise all of my philosophy and knowledge I have learnt will be returned to our nation as a whole. Finally, I wish to express my infinite thanks and appreciation to my lovely parents, Mr. Prateep and Mrs. Nongluck Jantam a, and my aunt, Miss Nongkran Jaidej, for all their grateful love and care, precious spiritua l support, patronage and sincer e encouragement. Least but not last, I want to thank a half of my soul, Dr. Sirima Suwannakut, who has been my greatest supporter and companion. Her patience and love has helped me achieve my goals throughout graduate school. I also give th anks to the Maynard family for being my home away from home while I lived in Florida. Without their inexhaustible patience, this work would not have been accomplished. My success belongs to you.
6 TABLE OF CONTENTS page ACKNOWLEDGEMENTS.............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES.......................................................................................................................10 LIST OF ABBREVIATIONS........................................................................................................ 11 ABSTRACT...................................................................................................................................17 CHAP TER 1 INTRODUCTION..................................................................................................................19 Succinate and its Importance.................................................................................................. 20 Literature Review.............................................................................................................. .....23 Brief Summary of Enzymes Invol ving in Anaerobic Ferm entation................................ 23 Succinate-Producing Microorganisms.............................................................................27 Succinate Producing Pathways in Microorganisms........................................................28 Basis for Increased Succinate Pr oduction in Previous Developed E. coli Strains .......... 30 Previously Developed Succini c Acid Producing Strains in E. coli .................................32 Objective.................................................................................................................................36 Genetic Manipulation of E. coli ......................................................................................37 Metabolic Evolution in Metabolic E ngineered E. coli ....................................................37 Metabolic Flux Analysis of Metabolic Engineered E. coli .............................................38 2 GENERAL PROCEDURES................................................................................................... 39 Medium Preparation............................................................................................................. ..39 Growth of Bacterial Cultures.................................................................................................. 39 Plasmid Preparation by Alkaline Lysis Method ..................................................................... 40 DNA Amplification by Polymerase Chain Reaction (PCR)................................................... 40 Agarose Gel Electrophoresis of DNA.................................................................................... 41 Restriction Endonuclease Digestion of DNA ......................................................................... 41 DNA Ligation.........................................................................................................................42 Preparation of E. coli Competent Cells ..................................................................................42 Preparation of E. coli Competent Cells by CaCl2 Method.............................................. 42 Preparation of E. coli Competent Cells by Electr o-transform ation Method for E. coli Carrying Temperature Sensitive Plasmid....................................................................... 42 Transformation of Competent Cells....................................................................................... 43 Transformation of E. coli By Heat Shock Method .......................................................... 43 Transformation of E. coli by electroporation ..................................................................43 Anaerobic Ferm entation ......................................................................................................... 44 Analytical Methods............................................................................................................. ....44
7 3 COMBINING METABOLIC ENGINEERIN G AND MET ABOLIC EVOLUTION TO DEVELOP NONRECOMBINANT STRAINS OF Escherichia coli ATCC8739 THAT PRODUCE SUCCINA TE AND MALATE........................................................................... 46 Introduction................................................................................................................... ..........46 Materials and Methods...........................................................................................................46 Strains, Media and Growth Conditions........................................................................... 46 Genetic Methods..............................................................................................................48 Deletion of mgsA and poxB Genes ..................................................................................48 Fermentations.................................................................................................................. 51 Analyses..........................................................................................................................52 Results and Discussion......................................................................................................... ..52 Construction of KJ012 for Succi nate Production by Deletion of ldhA adhE and ackA ..............................................................................................................................52 Improvement of KJ012 by Metabolic Evolution............................................................. 55 Construction of KJ032 and KJ060.................................................................................. 56 Construction of KJ070 and KJ071 by Deletion of Methylglyoxal Synthase ( mgsA ) ......61 Construction of KJ072 and KJ073 by Deletion of Pyruvate oxidase ( poxB ) ..................62 Fermentation of KJ060 and KJ073 in AM1 Medium Containing 10%(w/v) Glucose.... 62 Conversion of Other Substrates to Succinate .................................................................. 63 Production of Malate in NBS Medium C ontaining 1 m M Betaine and 10%(w/v) Glucose........................................................................................................................64 Conclusions.............................................................................................................................64 4 BATCH CHARACTERIZATION AND M ETABOLIC DIS TRIBUTION OF EVOLVED STRAINS OF Escherichia coli ATCC8739 TO PRODUCE SUCCINATE..... 67 Introduction................................................................................................................... ..........67 Materials and Methods...........................................................................................................67 Strains, Media and Growth Conditions........................................................................... 67 Fermentations.................................................................................................................. 67 Analyses..........................................................................................................................68 Calculation Specific Production Ra tes for Excreted Metabolites ................................... 68 Metabolic Flux Analysis.................................................................................................. 69 ATP and Cell Yield Analysis..........................................................................................70 Fermentation Pathways................................................................................................... 71 Results and Discussion......................................................................................................... ..75 Batch Characterization of Evolved E. coli Strains that Pr oduce Succinate.....................75 Effect of Glucose Concentra tions to Succinate Production ..................................... 75 Effect of Initial Inocula and Acetat e on Succinate Production in KJ060.................82 Effect of Low Salt Media (AM1) on Succinate Production in KJ060 ..................... 86 Effect of low Salt Media (AM1) on Succinate Production in KJ073 ....................... 86 Metabolic Flux Analysis in E. coli S trains To Produce Succinate.................................. 87 Metabolic Flux Distributions in E. coli W ild Type.................................................. 87 Metabolic Flux Distributions in Mutant Strains .......................................................91 Effect of Gene Deletions on ATP and Cell Yields.......................................................... 93 Conclusions.............................................................................................................................95
8 5 ELIMINATING SIDE PRODUCTS AND IN CR EASING SUCCINATE YIELDS IN ENGINEERED STRAINS OF Escherichia coli ATCC8739................................................ 96 Introduction................................................................................................................... ..........96 Materials and Methods...........................................................................................................97 Strains, Media and Growth Conditions........................................................................... 97 Deletion of FRT Markers in the adhE ldhA and focA-pflB Regions .............................97 Construction of Gene Deletions in tdcDE, and aspC ....................................................104 Removal of FRT Site in ackA Region And Construction of citF sfcA and pta-ackA Gene Deletions ...........................................................................................................105 Fermentations................................................................................................................ 107 Analyses........................................................................................................................107 Results and Discussion......................................................................................................... 108 Elimination of FRT Sites in KJ073 to Produce KJ091..................................................108 Deletion of tdcD and tdc E Decreased Acetate Production............................................108 Citrate lyase ( citF ) Deletion Had no Effect on Ac etate or Succinate Production.........112 Deleting aspC had no Effect on Succinate Yield.......................................................... 113 Deleting sfcA Had no Effect on Succinate Yield ...........................................................113 Deleting the Combination of aspC and sfcA Im proved Succinate Yield...................... 115 Reduction in Pyruvate and Acetate by Deletion of pta .................................................116 Conclusions...........................................................................................................................117 6 GENERAL CONCLUSIONS AND FUTURE DIRECTIONS............................................ 120 General Accomplishments.................................................................................................... 120 Future Works........................................................................................................................122 Future use of Succinate as Bioplastics.................................................................................. 123 LIST OF REFERENCES.............................................................................................................125 BIOGRAPHICAL SKETCH.......................................................................................................139
9 LIST OF TABLES Table page 1-1 Comparison of succinate production by natural producers............................................... 28 1-2 Comparison of succinate production by E. coli ................................................................33 2-1 PCR parameters for the am plification of specific genes. ..................................................41 3-1 Composition of mineral salts m edia (excluding carbon source)........................................ 47 3-2 Escherich ia coli strains, plasmids, and primers used in this study....................................49 3-3 Fermentation of glucose in minera l salts m edium by mutant strains of E. coli .................57 4-1 Escherich ia coli strains used in this study......................................................................... 68 4-2 Reactions used in Metabolic Flux An alysis (MF A) for succinate production under anaerobic condition............................................................................................................ 73 4-3 Stochiometric relationship between the me tabolic inte rmediates and metabolites and the network reactions (matrix K represented) for an anaerobic succinate production in glucose minimal medium in E. coli ...............................................................................75 4-4 Fermentation profile for succinate production in 5% a nd 10% glucose NBS m edium..... 77 4-5 Comparison of KJ060 on metabolite pr oduction using 10%(w/v) glucose (~556 m M) as substrate in NBS salt medium with diffe rent initial acetate concentrations and initial cell density........................................................................................................... ....83 4-6 Comparison of KJ060 and KJ073 on meta bolite production using 10%(w/v) glucose (~556 m M) as substrate in AM1 salt me dium with initial cell density............................. 84 4-7 Specific production rate of extracellular m etabolites........................................................ 88 4-8 Metabolic fluxes distribution of an anaerobic succinate production in 5%(w/v) and 10%(w/v) glucose NBS of various E. coli strains..............................................................90 5-1 Escherich ia coli strains, plasmids, and primers................................................................. 98 5-2 Fermentation of 10%(w/v) glucose in mi neral salts AM1 m edium (1mM betaine) by mutant strains of E. coli ...................................................................................................114
10 LIST OF FIGURES Figure page 1-1 Possible production routes to succinate ba sed products as commodity and specialty chem icals...................................................................................................................... ......21 1-2 Central metabolic pathway of E. co li .................................................................................24 3-1 Fermentation of glucose to succinate................................................................................. 54 3-2 Coupling of ATP production and growth to succinate and m alate production in engineered strains of E. coli. Solid arrows connect NADH pools..................................... 55 3-3 Steps in the genetic engineer ing and m etabolic evolution of E. coli ATCC 8739 as a biocatalyst for succinate and malate production................................................................ 59 3-4 Growth during metabolic evolution of KJ012 to produce KJ017, KJ032, and KJ060. ..... 59 3-5 Fermentation products during the metabolic evolution of strains for succinate and m alate production.............................................................................................................. .60 3-6 Production of succinate and malate in m ineral salts media (10% glucose) by derivatives of E. coli ATCC 8739...................................................................................... 63 4-1 Ferm entation Pathway of E. coli under anaerobic condition.............................................72 4-2 Proposed succinate producti on pathway from glucose......................................................85 5-1 Strain constructions....................................................................................................... ...109 5-2 Standard pathway for the anaerobic metabolism of glucose in E. coli. ...........................110 5-3 Expanded portion of metabolism illustrating the pathways of additional genes that have been deleted (solid crosses) in the context of standard central metabolism. ......... 111
11 LIST OF ABBREVIATIONS Ace acetate AceF lipoate acetyltransferase AceE E1 subunit of pyruvate dehydrogenase complex ACKA acetate kinase ADHE alcohol dehydrogenase ADP adenosine diphosphate AM1 Alfredo Mertinez medium version 1 AMP adenosine monophosphate ATP adenosine triphosphate ATCC american type culture collection atm atmosphere Av average Bio biomass bp base pair BSA bovine serum albumin C degree Celsius C2 2-carbon compound cAMP cyclic adenosine monophosphate CDW cell dried weight CoA coenzyme A cm centimeter C/N carbon-nitrogen ratio
12 CSL corn steep liquor DAACS superfamily of transporter dATP deoxyadenosine triphosphates DctA dicarboxylate DAACS transporter dCTP deoxycytidine triphosphate dGTP deoxyguanosine triphosphate DHAP dihydroxyacetone-phosphate DNA deoxyribonucleic acid dNTP deoxynucleo tide triphosphate dTTP deoxythymidine triphosphate E1 subunit of PDH complex; AceE dimer E2 subunit of PDH comlex; AceF core E3 subunit of PDH complex; LpdA dimer EC enzyme commission on nomenclature number EDTA ethylene diamine tetraacetic acid EM extracellular or excreted malate EP extracellular or excreted pyruvate FBA flux balance analysis Fe-S ferrous-sulfur compound FLP recombinase enzyme FNR fumarate-nitrate reductase regulatory protein For formate FRT FLP recombinase recognition site
13 g gram Gluc or G glucose GAPDH glyceraldehyde 3-phosphate dehydrogenase h hour HPLC high performance liquid chromatography HPr histidyl phosphorylated protein IPTG isopropyl-D-thiogalactoside kg kilogram kPa kilopascal Km michalis constant kb kilo base pair L or l liter lac lactate LB Luria-Bertani LpdA lipoamide dehydrogenase LDHA lactate dehydrogenase M molarity mal malate MFA metabolic flux analysis mg milli gram MH2 menaquinol (reduced form) MH4 cultured medium (yeast extract based medium) min minute
14 ml milli liter mM milli molar MMH3 cultured medium (yeast extract based medium) MOPS 3-[N-morpholino] propanesulfonic acid MQ menaquinone (oxidized form) ms milli second MW molecular weight N normality NADH nicotinamide adenine dinucleotide (reduced form) NAD or NAD+ nicotinamide adenine dinucleotide (oxidized form) NBS New Brunswick synthetic medium NaOAc sodium acetate vi specific volumetric flux of intermediates and compounds OAA oxaloacetate OD optical density PCK phosphoenolpyruvate carboxykinase PCR polymerase chain reaction PDH pyruvate dehydrogenase complex PdhR pyruvate dehydrogenase complex regulator protein PEP phosphoenolpyruvate PFLB pyruvate fomate-lyase Pi inorganic phosphate PO4 phosphate group
15 POXB pyruvate oxidase PPC phosphoenolpyruvate carboxylase PPS phosphoenolpyruvate synthase Prod productivity p.s.i. pounds per square inch PTS phosphotransferase system PYC pyruvate carboxylase PYK pyruvate kinase pyr pyruvate Red Red recombinase RnaseA ribonuclease A rpm revolution per minute v or vol volume V volts sec second SDS sodium dodecyl sulphate sp. species STP standard temperature and pressure suc succinate TAE tris-acetate-EDTA TBE tris-borate-EDTA TCA tricarboxylic acid TDCD propionate kinase
16 TE tris-EDTA buffer TF transferring U unit UV ultraviolet W weight YE yeast extract YATP cell yield per mole of ATP consumed YATP MAX maximum theoretical cell yield
17 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy METABOLIC ENGINEERING OF Escherichia coli TO EFFICIENTLY PRODUCE SUCCINATE IN MINERAL SALTS MEDIA By Kaemwich Jantama August 2008 Chair: Spyros A. Svoronos Cochair: Lonnie O. Ingram Major: Chemical Engineering Succinic acid, which is one of the building bl ock chemicals, is currently produced by the hydrogenation of petroleum-derived maleic anhydride. The increase in price of oil and petroleum derivatives has made the microbi al production of succinate an ec onomically attractive option for succinate as a renewable commodity chemi cal. In this dissertation, I accomplished the construction of Escherichia coli ATCC 8739 derivatives to produce succinate in mineral salts medium under anaerobic condi tions. This work was done by a combination of metabolic engineering and metabolic evolution. Both stra tegies allowed us to engineer the central metabolism of E. coli to direct the carbon flow to succinate and to select the strains that produce succinate with high titer, producti vity, and yield. Strain KJ073 ( ldhA adhE ackA ( focA pflB ) mgsA poxB ) produced succinate with molar yield of 1.2 per mole of glucose used. This strain was further engineered for improvements in succinate production by eliminating acetate, malate, and pyruvate as significan t side products. Strain KJ122 ( ldhA adhE ackA ( focA pflB ) mgsA poxB tdcDE citF aspC sfcA ) significantly increase d succinate yield (1.46 mol/mol glucose), succinate titer (680-700 mM), and average volumetric pr oductivity (0.9 g/l-h). Strain KJ134 ( ldhA adhE ( focA pflB ) mgsA poxB tdcDE citF aspC sfcA ( pta-
18 ackA )) produced less pyruvate a nd acetate with the succinate yield, titer, and average productivity of 1.53 mol/mol glucose used, 600 mM, and 0.8 g/l-h, respectively. Strains KJ122 and KJ134 produced near theoretical yields of succinate during simple, anaerobic, batch fermentations using mineral salts medium without the addition of plasmid or foreign gene, and any complex nutrient. Both may be useful as biocatalysts for the commercial production of succinate. Batch experiments were conducted with a wild type of E. coli and four succinateproducing strains in order to establish how metabolic fluxes changed as a result of gene deletions and metabolic evolution. The flux through pyr uvate dehydrogenase (P DH) complex was added to the classical anaerobic fermentation pathways and mutants lacking pyruvate formate-lyase (PFLB) increased the flux through this pa thway to produce NADH required for succinate production. Also, the mutants utilized phosphoenolpyr uvate synthase (PPS) to convert pyruvate produced during glucose phosphorylation back to phosphoenolpyruvate (PEP). This provided additional PEP utilized for producing succinate. Three mutants had pflB deleted, and these exhibited considerably higher fl ux to oxaloacetate (OAA). Increas ed glucose concentration did not affect the fluxes significantly, other than decreasing lactate pr oduction for the wild type from low to lower levels and decreasing the flux to biomass for the mutant strains. ATP generation was also studied. The wild type and a pflB+ mutant had ATP yield close to the maximum theoretical values. However, the pflBmutants had very low ATP yield. Correspondingly, the cells yields (YATP) were somewhat below the maximum literature value for the wild type and pflB+ mutant, and were considerably above that maximum for the pflBmutants. It is hypothesized that the strains may activate some unknown ATP generating pathways.
19 CHAPTER 1 INTRODUCTION The upward trend in the price of oil is expect ed to further in crease demand for alternative bio-based chemicals (Nordhoff et al., 2007). The use of petroleum and its derivatives for industrial chemical production could be reduced in the future by a combination of environmental concerns over toxic by-products, resistance to biodegradation, and problems associated with recycling (Hatti-Kaul et al., 2007; Sauer et al., 2008). The co ncept of green chemistry was introduced in the early 1990s to encourage the use of technologies that reduce the generation of toxic substances and offer environmentally friendly alternatives to petroleum-based products (Anastas and Warner, 1998). This is now shifting to white biotechnology, the production of renewable chemicals by microbial and enzymatic routes (Lorenz and Zinke, 2005). Microbially produced organic acids represen t potential building block molecules for the chemical industry (Lorenz and Zinke, 2005). The versatility, substrate select ivity, regioselectivity, chemoselectivity, enantioselectivity and catalysis at ambient temperatures and pressures make the production of chemicals using bi ological systems more attractive. Although biocatalysts have more commonly been used for the production of hi gh-value products such as fine chemicals and pharmaceuticals (Breuer et al., 2004; Schmid et al ., 2001; Thomas et al., 2002), the market share of white biotechnology products is predicted to ri se from 5% to 20% by 2010 (Hatti-Kaul et al., 2007; Sauer et al., 2008). However, the success of increased market share of white biotechnology products is challenged by the cost-competitivenes s of existing chemical processes with capital assets already in place for commodity chemical production. Succinic acid has been identified by the U.S. Department of Energy as one of the top 12 building block chemicals that could be produced from renewable feedstocks (Werpy and Petersen, 2004). Current succinic acid producti on by the hydrogenation of petroleum-derived
20 maleic anhydride is too expensive for widespr ead use as a platform chemical. Inexpensive microbial processes could provide succinic acid as a renewable building block molecule for conversion into chemical intermediates, specialty chemicals, food ingred ients, green solvents, pharmaceutical products, and biodegradable plas tics (Lee et al., 2004; Wendisch et al., 2006; Willke andVorlop, 2004; Zeikus et al., 1999). Potent ially high volume products that can be made from succinic acid include tetrahydrofuran, 1,4-butanediol, succindiam ide, succinonitrile, dimethylsuccinate, N-methyl-pyrrolidone 2-pyrrolidone, 1,4-diaminobutane, and butyrolactone (Sauer et al., 2008) The microbial production of succinic acid from carbohydrates offers the opportunity to be both greener and more cost effective than petroleum-based alternative products. In addition, microbial succinate production incorporates CO2, a primary greenhouse gas, providing further incentiv e for production by white biotechnology. Succinate and its Importance Succinate and its derivatives have been of comm er cial interest for many applications. It is alternatively used to produce new specialty ch emicals and materials for which the demand is growing rapidly (Sado and Tajima, 1980). Examples for applications in many industries are shown in Figure 1-1. First, succinate is used as an anti-microbial agent, as a flavoring agent, and as an acidulant/pH modifier. In the food market, newly introduced flavor enhancers are sodium succinate and dilysine succinate, used in low sodium food, which can replace monosodium glutamate (Jain et al., 1989). Sec ond, succinate has been used in the production of health-related agents including pharmaceuticals, antibiotics, am ino acids, and vitamins. Third, because of its structure as a linear saturated dicarboxylic acid, succinic acid can be used as an intermediate chemical and would be converted to many chem icals applied in chemical industry such as butanediol, tetrahydrofuran, butyrolactone, and other four carbon chemicals (Dake et al., 1987).
21 It also has been used in the chemical industry as a surfactant, detergen t, extender, and foaming agent and as an ion chelator in which it is used in electroplating to pr event corrosion and pitting of metals. Diethyl succinate is a useful solvent for cleaning metal surfaces or for paint stripping. Ethylene diaminedisuccinate has also been used to replace ethylene diamine tetraacetic acid (EDTA) (Bergen and Bates, 1984). A polymerizat ion product of succinic acid and 1,4 butanediol has the potential to become a biodegradable plastic (Zwicker et al., 1997). Succinylation of lysine residues improves the physical and functio nal attributes of soy proteins in foods, and succinylation of cellulose is applied in impr oving water absorbability (Wollenberg and Frank, 1988). It is evident that it is important to deve lop technologies for the pr oduction of succinate to supply the industrial needs. Figure 1-1 Possible production routes to succinate based products as commodity and specialty chemicals (Zeikus et al., 1999).
22 The total market size for uses of succi nic acid is more than $400,000,000 per year. Currently, more than 15,000 tonnes of industr ial succinic acid is sold, and is produced petrochemically from butane or oxidation of be nzene through maleic an hydride. The price of succinic acid varies between $5.90-8.80/kg depe nding on its purity (Zeikus et al., 1999). However, succinate produced fermentatively is about 5,000 tonnes per year and sold at about $2.20/kg to the food market. Fermentation-based pr oduction utilizes cheap agricultural products such as corn, starch, molasses, or cheap sugars such as glucose or sucrose as carbon substrates. The production cost of succinate by fermentation is lower than that by petrochemistry thus making fermentation-derived succinate sold at a lower proposed price at about $0.55/kg if the production size will be above 75,000 tonnes (Zeikus et al., 1999). Becaus e of its economics promise, fermentation-derived succinate has the potential to become a large volume commodity chemical that is forming the basis for suppl ying many important interm ediate and specialty chemicals. Moreover as a small molecule ch emical, succinate would replace many chemicals derived from benzene and intermediate petrochemi cals, resulting in a la rge reduction in pollution from the manufacture and consumption of ove r 250 benzene-derived chemicals (Ahmed and Morris, 1994). Succinate production by fermentation has distin ct advantages over productions of other organic acids because carbon dioxide gas is consumed during succinate production. The process would decrease pollution caused by the greenhous e gas. However, the production of succinate and most of its derivativ es is currently at the advanced research and developmental stage. Since the demand of succinate in many applications is high and increasing every year, interest in anaerobic fermentation has intensified especially as how it relates to the utilization of cheap
23 sugar sources such as glucose, and other agri cultural carbohydrates to produce higher-value fermentation derived succinate. Literature Review Brief Summary of Enzymes Involving in Anaerobic Fermentation In m any microorganisms including Escherichia coli the process of oxidative phosphorylation cannot occur under anaerobic conditi ons and the cell produces energy from the process of degrading the origin al substrate known as substrate level phosphorylation. The cell is trying to maximize the formation of energy. Th e reducing power, NADH, which is produced during degradation of substrate, ha s to be reoxidized for the proce ss to continue, as its supply is not limitless. It is important to note that pyruva te produced by the glycolytic pathway will enter the tricarboxylic acid cycle (TCA), at least in part, to provide essential precursors for biosynthesis. The NADH, produced in the cycle ca nnot be converted to ATP as the cells have no supply of oxygen to drive oxidative phosphoryla tion. However, the cell can recycle NADH by means of reduction of some carbon intermediates that are accumul ated under these conditions. In other words, the reducing equivalents react w ith the accumulated carbon intermediates to reduce them in their turn (Figure 1-2). In the central anaerobic metabolic pathwa y, pyruvate is assimilated to re-oxidize NADH via lactate dehydrogenase and alcohol dehydrogenase activities resulting in lactate and alcohol productions, respectively. In the simplest method of hydrogen disposal, pyruvate is reduced to lactate at the expense of NADH. The reac tion is catalyzed by a cytoplasmic lactate dehydrogenase encoded by ldhA The enzyme is jointly induced by acid pH and anaerobiosis. Lactate can be produced from dihydroxyacetonephosphate (DHAP). DHAP is converted to methylglyoxal by product of mgsA (methyglyoxal synthase) and is subsequently converted to lactate by glyoxalase activities encoded by gloAB (Wood, 1961).
24 Figure 1-2. Central metabolic pathway of E. coli Solid arrows represent central fermentative pathways. Dotted arrow represen ts microanaerobic pathway ( poxB ). Dash arrow represents minor lactate producing pathway (mgsA, gloAB ). Genes: pykAF : pyruvate kinase, ldhA : lactate dehydrogenase, pflB : pyruvate formate-lyase, pta : phosphate acetyltransferase, ackA : acetate kinase, adhE : alcohol dehydrogenase, ppc : PEP carboxylase, aceEF/lpdA : acetyltransferase/dihydrolipoamide acetyltransferase component of the pyruvate dehydrogenase complex, mdh : malate dehydrogenase, fumABC : fumarase, frdABCD : fumarate reductase, fdh : formate dehydrogenase, mgsA : methyglyoxal synthase, gloAB : glyoxylase, and poxB : pyruvate oxidase (adapted from Clark, 1989). Pyruvate formate-lyase encoded by pflB which is responsible fo r anaerobic conversion of pyruvate to acetyl~CoA and formate, is posttra nslationally interconverted between active and inactive forms. The enzyme synthesis is increas ed by anaerobiosis and can be raised further by pyruvate (Knappe and Sawers, 1990). Acetyl~CoA produced from pyruvate can be used to generate ATP from ADP by conversion to acetate, or to dispose off extra reducing equivalents by
25 conversion to ethanol. The first process de pends on the consecutive action of phosphate acetyltransferase probably encoded by pta and acetate kinase encoded by ackA Synthesis of these enzymes is not significantly changed by the respiratory condition of the cell. Consequently, most of the acetyl~CoA is excreted as acetate by cells growing on glucose under aerobic condition. In the absence of glucose, external ac etate is mostly utilized by reversal of the pathway catalyzed by acetyl~CoA synthethase, encoded by acs (Kumari et al., 1995). Acetyl~CoA is also converted to etha nol under anaerobic fermentation. The pathway involves a consecutive reduction of the acet yl group of acetyl~CoA to acetaldehyde, and acetaldehyde to ethanol at the expense of NA DH. The reactions are catalyzed by a single polypeptide, which is alcohol dehydrogenase, encoded by adhE The propinquity of the two sites of reduction might minimize escape of the acetal dehyde, which is chemically reactive. ADHE protein has dual enzyme activities, which ar e alcohol dehydrogenase and coenzyme~A-linked acetaldehyde dehydrogenase. However, alcohol dehydr ogenase is more sensitive to inactivation by the aerobic metabolism (Clark, 1989). The assimilation of PEP also o ccurs via carboxylation in whic h it generates succinic acid. For PEP carboxylation, fumarate reductase is acti vated and re-oxidizes NADH using fumarate as an electron acceptor. Endogenous or exogenous carbon dioxide is combined with PEP by phosphoenolpyruvate carboxylase encoded by ppc The oxaloacetate formed is reduced to malate by the activity of malate dehydrogenase encoded by mdh Malate is dehydrated to fumarate by fumarase enzymes encoded by fumABC whose anaerobic induction depends on FNR regulation. Fumarate is finally reduced to succinate by fumara te reductase. The net result is disposal of four reducing equivalents (4H+ + 4e-). Fumarate reductase encoded by frdABCD can accept electrons from various primary donor enzymes through me naquinone. Fumarate reductase is induced
26 anaerobically by fumarate but is repressed by oxygen or anaerobica lly by nitrate (Cecchini et al., 2002). Pyruvate dehydrogenase multi-enzyme co mplex is composed of products of aceEF and lpdA genes. The reaction is the gateway to the TCA cycle, producing acetyl~CoA for the first reaction. The enzyme complex is composed of multip le copies of three enzymes: E1, E2 and E3, in stoichiometry of 24:24:12, respectively. The E1 dimers (encoded by aceE ) catalyze acetylation of the lipoate moieties that are attached to the E2 subunits. The E2 subunits (encoded by aceF ) are the core of pyruvate dehydrogenase complex and exhibit transacetylation. The E3 component is shared with 2-oxoglutarate dehydr ogenase and glycine cleavage multi-enzyme complexes. Pyruvate is channeled through the cat alytic reactions by attachment in thioester linkage to lipoyl groups carrying acetyl group to su ccessive active sites. This enzyme complex is active under aerobic condition (CaJacob et al., 1985). Pyruvate can be converted to CO2, Acetyl~CoA, and NADH via the enzyme complex. Under micro-aerobic condition, pyruvate oxidase encoded by poxB is responsible for generating C2 compounds from pyruvate during the tr ansition between aerobic and strict anaerobic growth condition. This enzyme couples the electron from pyruvate to ubiquinone and decarboxylates pyruvate to generate carbon dioxide and acetate (Abdel-Hamid et al., 2001). Under both aerobic and anaerobic respirations, the versatilit y of the electron transport system for generating proton motive force is made possible by employing ubiquinone or menaquinone in the plasma membrane as a diffusi ble electron carrier or adaptor to connect a donor modular unit functionally to an acceptor modular unit. The types of electron carrier and donor modulars used for electron transport depend on the pattern of gene expression in response to the growth conditions. In anaerobic conditi ons, the electron donor modular units are primary
27 dehydrogenases of the flavoprotein kind. The acc eptor modular units consist of terminal reductases requiring various components, such as Fe-S. In general, when the terminal acceptor has a relatively high redox potential such as oxygen, ubiquinone is used as the redox adaptor i.e. pyruvate oxidase case. When the terminal acceptor has a relatively low redox potential such as fumarate, menaquinone is used instead for exampl e reduction of fumarate to succinate (Cecchini et al., 2002). Succinate-Producing Microorganisms Succinate is an intermediate produced in th e metabolic pathway of several anaerobic and facultative microorganisms. Many propi onate-producing bacteria such as Propionibacterium species, typical gastrointestinal bacteria such as E. coli Pectinatus sp., Bacteroides sp., rumen bacteria such as Ruminococcus flavefaciens Actinobacillus succinogens Anaerobiospirillum succiniciproducens, Bacteroides amylophilus, Prevotella ruminicola Succinimonas amylolytica Succinivibrio dextrinisolvens Wolineela succinogenes and Cytophaga succinicans produce succinic acid from sugars and amino acids (Bryan t and Small, 1956; Bryant et al., 1958; Davis et al., 1976; Guettler et al., 1996a, b; Scheifi nger and Wolin, 1973; Van der Werf et al., 1997). Most of the succinate-producing bacteria have been isolated and cultured from the rumen, because succinate is required for a precursor of propinate that is subsequent to oxidation for providing energy and biosynthe tic precursors in such an animal (Weimer, 1993). Many bacteria have been described with th e natural ability to produce succinate as a major fermentation product (Guettler et al., 1998; Table 1-1). Some of these such as Actinobacillus succinogenes (Guettler et al., 1996 a,b; Meynial-Salles et al., 2007), Anaerobiospirillum succiniciproducens (Glassner and Datta, 1992), and Mannheimia succinoproducens (Lee et al., 2006; Song et al., 2007) can produce at high rates (up to 4 g/l-h) with impressive titers of succinate (300-900 mM) and high yi elds (>1.1 mol succinate/mol
28 glucose). In a recent study with a native succinate producer, A. succiniciproducens, electrodialysis, sparging with CO2, cell recycle, and batch feeding were combined (MeynialSalles et al., 2007). However, these natural produ cers require complex media ingredients, which add cost associated with production, purification, and waste disposal. Succinate Producing Pathways in Microorganisms Succinic acid producing bacteria produce varying am ounts of succinic acid as well as other products, including ethanol lactic acid, and formic aci d during mixed acid fermentation. The rumen bacteria produce succinic acid in very high concentrations, along with acetate, pyruvate, formate, and ethanol. E. coli produces succinate as a mi nor fermentation product, typically 12 mol/100 mol glucose (Wood, 1961). Unlike E. coli the rumen bacteria such as A. succiniciproducens forms succinate up to 120 mol/100 mol glucose (Nghiem et al., 1997; Samuelov et al., 1991). Several different pathways can produce succinic acid (Lee et al., 2004). One pathway involves phosphoenolpyruvate (PEP) carboxylation that is catalyzed by PEP carboxylase or PEP carboxykinase. The other pathway involves pyruvate carboxylation. Two different enzymes, malic en zyme and pyruvate carboxylase in metabolic pathways are responsible for pyruvate carboxylat ion. Malic enzyme catalyzes th e conversion of pyruvate into malic acid while pyruvate carboxylase catalyzes the conversion of pyruva te into oxaloacetate (Lee et al., 2004). Table 1-1. Comparison of succinate production by natural producers a Organism Medium/Conditiona Succinate Titer (mM)b Succinate Yield (mol/mol) Reference Actinobacillus succinogenes FZ53 130 g/l glucose supplemented with 15 g/l CSL and 5 g/l YE, 80 g/l MgCO3, anaerobic batch fermentation, 78 h incubation 898 [1.36] 1.25 Guettler et al., 1996a Anaerobiospirillum succiniciproducens ATCC 53488 120 g/l glucose in peptone/YE based medium, integrated membrane-bioreactor-electrodialysis with CO2 sparging, 150 h incubation 703 [0.55] 1.35 MeynialSalles et al., 2007
29 Table 1-1. ( Continued ) Organism Medium/Conditiona Succinate Titer (mM)b Succinate Yield (mol/mol) Reference Actinobacillus succinogenes 130Z 100 g/l glucose supplemented with 15 g/l CSL and YE, 80 g/l MgCO3, anaerobic batch fermentation, CO2 sparging, 39 h incubation 678 [2.05] 1.37 Guettler et al., 1996b Mannheimia succiniciproducens ( ldhA pflB pta-ackA ) 63 g/L glucose in MMH3 (yeast extract based medium), fed batch fermentation, 0.25 vol/vol/min CO2 sparging, 30 h incubation 444 [1.75] 1.16 Lee et al., 2006 Actinobacillus succinogenes ATCC 55618 70 g/l glucose with flour hydrolysate and 5 g/l YE, anaerobic batch fermentation with 4% inoculum, 65 h incubation 302 [0.55] 1.18 Du et al., 2007 Anaerobiospirillum succiniciproducens ATCC 53488 50 g/l glucose, 2% CSL, and 25 ppm tryptophan, neutralized with 5.5 M NaCO3, saturated medium of 0.3 atm partial pressure of CO2, 29.5 h incubation 289 [1.16] 1.04 Guettler et al., 1998 Succinivibrio dextrinosolvens ATCC 19716 15 g/l of each CSL and YE, 10 0 g/l glucose, and 80 g/l MgCO3, batch fermentation, 36 h. 226 [0.74] NR Guettler et al., 1998 Corynebacterium glutanicum R 40 g/l glucose (121 g total glucose) in Defined mineral salt medium with 400 mM NaHCO3 fed batch fermentation, 6 h incubation 195 [3.83] 0.29 Okino et al., 2005 Prevotella ruminocola ATCC 19188 15 g/l of each CSL and YE, 10 0 g/l glucose, and 80 g/l MgCO3, batch fermentation, 36 h incubation 160 [0.52] NR Guettler et al., 1998 Mannheimia succiniciproducens MBEL55E KCTC 0769BP 18 g/L glucose in MH4 (YE based medium) supplemented with 119 mM NaHCO3 a continuouscell-recycle membrane reactor with the CO2 partial pressure of 101.3 kPa gas (100% CO2), 6 h incubation 144 [2.83] 1.44 Song et al., 2007 a Abbreviations: CSL, corn steep liquor ; YE, yeast extract; NR, not reported. b Average volumetric productivity is shown in brackets [g/l-h] ben eath succinate titer. The molar yield was calculated based on the pr oduction of succinate fr om metabolized sugar during both aerobic and anaerobic conditions. Biomass was generated predominantly during aerobic growth. Succinate was produced pr imarily during anaerobi c incubation with CO2, H2, or a mixture of both. Normally under anaerobic conditions, the PE P carboxylation pathway is a major pathway to produce succinic acid. A. succiniciproducens and A. succinogenes have been demonstrated to be the most effeicient succinate producing st rains. Both strains pr oduce succinic acid through four reactions catalyzed by PEP carboxykinase, ma late dehydrogenase, fumarase, and fumarate dehydrogenase. In contrast, E. coli utilizes multiple pathways to form succinic acid (Van der
30 Werf et al., 1997). Unde r anaerobic conditions, E. coli utilizes glucose to primarily produce of acetate, formate, and ethanol, as well as smaller amounts of lactate and succinate (Figure 1-2). Basis for Increased Succinate Production in Previous Developed E. coli Strains It is accep ted that the enzyme generally regarded as the dominant carboxylating activity for succinate production is PEP carboxylase duri ng growth (Gokarn et al., 2000; Karp et al., 2007; Keseler et al., 2005; Mill ard et al., 1996; Unden and Kleefeld, 2004; Vemuri et al., 2002b; Wang et al., 2006). Previous studies show ed that the overexpression of a native ppc gene in E. coli resulted in higher specific succ inate production (Milla rd et al., 1996), higher specific growth rate, and lower specific acetate pr oduction due to more carboxylation of PEP to replenish TCA cycle intermediates (Farmer and Liao, 1997). Howe ver, since PEP is required for the glucose transport system, overexpressing ppc also decreases the glucose uptake rate by 15-40% without significantly increasing succinate yi eld (per glucose) as compared to an isogenic control (Chao and Liao, 1994; Gokarn et al., 2000). It has also been report ed that overexpression of the E. coli phosphoenolpyruvate carboxylase ( ppc ) is not helpful for succinate production in the absence of a mutation in pps (Chao and Liao, 1994; Gokarn et al., 2000; Kim et al., 2004; Millard et al., 1996). This failure of the native PPC to increase succinate yields diverted most research attention to a new metabolic design, overexpression of the PYC (pyruvate carboxylase) from Lactobacillus lactis or Rhizobium etli as the carboxylating step (Vemuri et al., 2002a, b; Gokarn et al., 2000; Lin et al., 2005a, b, c) rath er than pursuing furthe r work with the native repertoire of E. coli genes. Succinate produced by E. coli using the pathway generally regarded as the native fermentation pathway (phosphoe nolpyruvate carboxylase; ppc ) waste the energy of phosphoenolpyruvate by producing inor ganic phosphate. One ATP is lost per succinate produced by this pathway (Figure 1-2). Conserving this energy as ATP by using alternative enzyme
31 systems represents an opportunity to increase ce ll growth and co-select for increased succinate production. Based on known genes in E. coli, three other enzyme routes ( sfcA, maeB and pckA ) for succinate production were envisioned that would conserve ATP and could thereby increase growth. However, all carboxylation steps in these alternative routes are thought to function in the reverse direction (decar boxylation) primarily for gluconeogenes is during growth on substrates such as organic acids (Keseler et al., 2005; Oh et al., 2002; Stols a nd Donnelly, 1997; Samuelov et al., 1991; Sanwal, 1970). Rumen bacteria such as A. succinogenes produce succinate as a primary product during glucose fermentation using the energy c onserving phosphoenolpyruvate carboxykinase for carboxylation (Kim et al., 2004; McKinlay et al ., 2007). Many researchers have studied the overexpression of PCK to increase carbon flow to succinate. Previous in vestigators have noted that the kinetic parameters of phosphoenolpyr uvate carboxylase (PPC) and phosphoenolpyruvate carboxykinase (PCK) may have important eff ects on carboxylation and succinate production (Millard et al., 1996; Kim et al., 2004). The Km towards bicarbonate for E.coli phosphoenolpyruvate carboxylase (PPC) is 0.15 mM (M orikawa et al., 1980), 9-fold lower (13 mM) than E. coli phosphoenolpyruvate carboxykinase (P CK) (Krebs and Bridger, 1980). Although overexpressing PCK from E. coli in multi-copy plasmid increased PCK activity by 50fold, it was reported to have no effect on su ccinate production (Mil lard et al., 1996). In E. coli K12, activities for both phosphoe nolpyruvate carboxylase and phosphoenolpyruvate carboxykinase we re reported to be equal in vitro (Van der Werf et al., 1997) with the former serving as the primary r oute to succinate. Succi nate production was also not increased when PCK from A. succiniciproducens was overexpressed in E. coli K12 (Kim et al., 2004). This enzyme also has a high Km for bicarbonate (30 mM; La ivenieks et al., 1997).
32 However, when A. succiniciproducens PCK was overexpressed in a ppc mutant of E. coli K12, succinate production was increase d 6.5-fold (Kim et al., 2004). Previously Developed Succinic Acid Producing Strains in E. coli Since E. coli has an ability to grow fast without a requirem ent of complex nutrients, and is easy to manipulate its metabolic pathways by genetic engineering, it ha s a potential to become a target microorganism for strain improvement and process design for succinate production. However, the feasibility of in creasing succinate production yield in this microorganism through metabolic engineering has not yet been fully developed. In the past decade, many research groups have been studying extensively to obtain high production yield of succinic acid by metabolic engineering of E. coli strains (Table 1-2). During glycolysis, NADH is generated and re-oxidized through the reduc tion of organic intermediates derived from glucose. Unfortuna tely, no more than 0.2 mol of su ccinate is produced per mol of glucose consumed by E. coli during fermentation (Lee et al ., 2004). Analysis of metabolism in silico has been used to design gene knockouts to create a pathway in E. coli that is analogous to the native succinate pathway in M. succiniciproducens (Lee et al., 2005, 2006). The resulting strain, however, produced low levels of succinate Andersson et al. (2007) reported the highest levels of succinate produ ction by an engineered E. coli (339 mM) containing only native genes. Pyruvate carboxylation in recombinant E. coli is the major target pathway for redirecting pyruvate to succinic acid. Strain NZN111 wa s engineered by inact ivating two genes (pflB encoding pyruvate formate-lyase and ldhA encoding lactate dehydrogena se), and over-expressing two E. coli genes, malate dehydrogenase (mdh ) and phosphoenolpyruvate carboxylase ( ppc ), from multicopy plasmids (Stols et al., 1997). Fermen tation of this strain revealed that pyruvate was accumulated in the growing medium.
33 Table 1-2. Comparison of succinate production by E. coli a Organism Medium/Conditiona Succinate Titer (mM)b Succinate Yield (mol/mol) Reference E. coli AFP111 ( pflAB, ldhA, ptsG ) Rhizobium etli pyc overexpressed 40 g/l glucose (90 g total glucose) in medium supplemented with 20 g/l tryptone, 10 g/l YE and 40 g/l MgCO3 dual phase-fed batch fermentation, 76 h incubation 841 [1.31] 1.68 Vemuri et al., 2002,b E. coli HL27659k/pKK313 ( iclR sdhAB ackA-pta poxB, pstG ) Sorghum vulgare pepc overexpressed 106 g/l glucose in medium supplemented with 20 g/l tryptone, 32 g/l YE and 2 g/l NaHCO3 fed batch fermentation under complete aerobic condition, 59 h incubation 499 [1.00] 0.85 Lin et al., 2005d Bacterial Isolate 130Z ATCC 55618 50 g/l glucose supplemented with 1% CSL, 0.6% YE, and 2 g/l MgCO3 neutralized with 10 N NaOH, 0.3 atm of CO2, 29.5 h incubation 388 [1.55] 1.40 Guettler et al., 1998 E. coli SBS550MG ( ldhA adhE iclR ackA-pta ), L. lactis pyc Bacillus subtilis citZ 20 g/l glucose (100 g total glucose) LB supplemented with 1 g/l NaHCO3, 200 mg/l ampicillin, and 1mM IPTG. 100% CO2 at 1L/min STP headspace, repeated fed-batch fermentation, 95 h incubation 339 [0.42] 1.61c Sanchez et al., 2005a; Cox et al., 2006 E. coli AFP184 ( pflB ldhA pts ) 102 g/l glucose supplemented with 15 g/l CSL, dual phase aerobic growth and anaerobic production, sparging with air followed by CO2, 32 h incubation 339 [1.27] 0.72 c Andersson et al., 2007 E. coli SBS550MG ( ldhA adhE iclR ackA-pta ), Overexpression of L. lactis pyc Bacillus subtilis citZ 20 g/l glucose LB supplemented with 1 g/l NaHCO3, 200 mg/l ampicillin, and 1mM IPTG. 100% CO2 at 1L/min STP headspace, batch fermentation, 24 h. incubation 162.6 [0.80] 1.61c Sanchez et al., 2005a; Cox et al., 2006 E. coli SBS110MG ( ldhA adhE ), Lactococcus lactis pyc 20 g/l glucose LB supplemented with1.5 g/l NaHCO3 and 0.5g MgCO3, 200 mg/l ampicillin, and 1mM IPTG. Dual pahse with 100% CO2 at 1L/min STP headspace, 168 h incubation 130 [0.09] 1.24 c Sanchez et al., 2005a; Sanchez et al., 2006 E. coli NZN111 (W1485 pflB ladhA ), E. coli mdh overexpressed 20 g/l glucose LB supplemented with 0.5 g MgCO3, 1.5 g/l NaOAc, 0.1 g/l ampicillin, and 10 M IPTG, 44 h incubation, sealed serum tube. 108 [0.22] 0.98 c Stols et al., 1997 E. coli JCL1208, E. coli ppc overexpressed 11 g/l glucose LB supplemented with 0.15 g MgCO3, 0.1 g/l carbenicillin, and 0.1 mM IPTG, 44 h incubation, anoxic CO2 charging at 1 atm headspace, 18 h incubation 91 [0.60] 0.44 c Millard et al., 1996 E. coli GJT Sorghum pepC 40 g/l glucose LB supplemented with 27.78 g/l MgCO3, simple batch fermentation in sealed airtight flask 80 [no data] 0.42 c Lin et al., 2005c E. coli HL51276k ( iclR icd sdhAB ackA-pta poxB, pstG ), Sorghum sp. pepC S8D mutation 10.8 g/l glucose LB supplemented with 2g/l NaHCO3 50 mg/l kanamycin, 1 mM IPTG, aerobic batch reactor, 50 h incubation 68 [0.16] 1.09 c Lin et al., 2005b
34 Table 1-2. ( Continued ) Organism Medium/Conditiona Succinate Titer (mM)b Succinate Yield (mol/mol) Reference E. coli SBS880MG ( ldhA adhE fdhF ), L. lactis pyc 20 g/l glucose LB supplemented with1.5 g/l NaHCO3 and 0.5g MgCO3, 200 mg/l ampicillin, and 1mM IPTG. Dual phase with 100% CO2 headspace, 168 h incubation 60 [0.04] 0.94 c Sanchez et al., 2005b a Abbreviations: CSL, corn steep liq uor; YE, yeast extract; NR, not reported. b Average volumetric productivity is shown in brackets [g l-h] ben eath succinate titer. c The molar yield was calculated based on the production of succinate from metabolized sugar during both aerobic and anaerobic conditions. Biomass was generated predominantly during aerobic growth. Succinate was produced pr imarily during anaerobi c incubation with CO2, H2, or a mixture of both. Malate dehydrogenase (mdh ) was expressed in this strain to dissipate accumulated pyruvate to succinic acid. This strain produced 108 mM succinate with a molar yield of 0.98 mol succinate per mol of me tabolized glucose (Donnelly et al., 19 98; Millard et al ., 1996; Stols and Donnelly, 1997; Chatterjee et al., 2001). Other researchers have pursued alternative approaches that express heterologous genes from plasmids in recombinant E. coli (Table 1-2) The pyruvate carboxylase ( pyc ) from Rhizobium etli was over-expressed from a multicopy plasmid to direct carbon flow into succinate production route and succinate was produced from the strain e xpressing heterologous pyc about 841 mM (Gokarn et al., 2000; Vemu ri et al., 2002a, b). Strain SBS550MG was constructed by inactivating the isocitrate lyase repressor ( iclR ), adhE ldhA and ackA, and over-expressing Bacillus subtilis citZ (citrate synthase) and R. etli pyc from a multi-copy plasmid (Sanchez et al., 2005a). This strain has inactivated the iclR gene, which encodes a tr anscriptional repressor protein of glyoxylate bypass, resulting in constitutive activation of glyoxylate bypass via flux through isocitrate. This strain achieved a very hi gh yield, 1.6 mol succinate per mol glucose used with an average anaerobic productivity rate of 10 mM/h (Sanchez et al., 2005a).
35 Another example of carboxylation of pyruvate into oxaloacetate by pyruvate carboxylase ( pyc) was developed. AFP111, a spontaneous mutation strain in the ptsG gene encoding the glucose specific permease of phosphotransferase system, produced succinate as a major product when the R. etli PYC was overexpressed, resu lting in the yield of 0.96-mol/mol glucose used. The further development for an efficient succi nate production was achieved by employing this strain in a dual phase fermentation in which an initial aerobic gr owth phase was allowed prior to an anaerobic growth phase. The final succinic acid concentration obtained was reaching the productivity of 1.3 g/l-h (Clar k, 1989; Vemuri et al., 2002a, b; Gokarn et al., 2000). Not limited to anaerobic fermentation, production of succinic acid under aerobic conditions was also characterized. The most efficient producing strain HL27659k was able to achieve a succinate yield of 0.91 mol/mol glucos e used at a dilution rate of 0.1/h. Strain HL27659k was engineered by mutating succinate dehydrogenase (sdhAB ), phosphate acetyltransferase ( pta ), acetate kinase ( ackA ), pyruvate oxidase (poxB ) glucose transporter ( ptsG ), and the isocitrate lyase repressor ( iclR ). This strain produced less than 100 mM succinate and required oxygen-limited fermentation conditi ons (Cox et al., 2006; Lin et al., 2005a, b, c; Yun et al., 2005). IclR mutation in the strain exhibits high citrate synthase and malate dehydrogenase activities resulting in highly efficient succinic acid production as a major product under aerobic conditions without pyruvate accumulation (Lin et al., 2005d). Complex, multi-stage fermentation processes ha ve also been investigated to improve succinate production using recombinant E. coli (Table 1-2). In many of these, aerobic growth phase is followed by an anaerobic fermentation phase and includes sparging with CO2, H2, or both (Andersson et al., 2007; Nghiem et al., 1999; Sanchez et al., 2005a, b; Sanchez et al., 2006; Vemuri et al., 2002a, b).
36 Objective Many microorganisms that produce succi nate naturally unde r obligate anaerobic condition such as A. succinogenes, A. succiniciproducens, and M. succiniciproducens have been studied extensively for succinate producti on under anaerobic condi tion. Even though high succinate production rate have been achieved fr om these organisms, the byproducts such as acetate, lactate, formate, and ethanol have been also in high level. Unfortunately, no genetic technique is available to engi neer these microorganisms to produce homo-succinic acid under anaerobic conditions via gene de letions or disruptions. These mi croorganisms require a complex source of nutrients such as whey based medium and yeast extract for optimal growth. Moreover, to obtain high succinate producti on rate, these microorganisms also require external supplies of carbon dioxide, hydrogen, or the mixture of gases, which raise the production cost. A bacterium that exhibits fast growth, no requirements of sp ecial and expensive sour ces of nutrients during growth, and has available techniques for ge netic manipulations, would be an ideal microorganism for succini c acid production. Since E. coli exhibits fast growth, is able to grow in the minimal medium, and many genetic techniques can be applied to it, it would be a good target microorganism to be developed and studied for pract ical succinate production in the industry. All the methods for producing succinate from E. coli published have involved rich media such as Lurie-Bertani (LB) broth, which contai n sources of amino acids, proteins, and other chemicals from yeast extract and peptone. Cont aminating proteins a nd cell byproducts would have to be removed from the final product. Thus the separation process requires removal of the impurities including cells, proteins, organic ac ids, and other impurities. Antibiotics and isopropyl-D-thiogalactoside (IPTG) used for ma intaining plasmid and inducing gene overexpression increase the cost of succinate production. Carbon dioxide and hydrogen gases
37 supplied to the reactor during the fermentation pro cess also raise the produc tion cost. The cost of inefficiencies related with media compos itions, downstream processing including product recovery, concentration, and purif ication has been very high. To become more attractive, the production of succinate in E. coli could be performed in a minimal medium. However the succinate production by E. coli in a minimal medium has not yet been studied. Hence, this project will study succinate produ ction in Mineral Salt Medium. To attain efficient succinate production, the specific goals will be as described below. Genetic Manipulation of E. coli E. coli ATCC 8739 will be used as a host to alter the m etabolic pathway producing succinic acid as a major fermented product. All the possible native central metabolic genes that are responsible for producing primarily anaerobic fermentation products in E. coli will be inactivated by genetic e ngineering techniques. The native gene s of central anaerobic metabolism including adhE (alcohol dehydrogenase E), ldhA (lactate dehydrogenase A), ackA (acetate kinase A), and pflB (pyruvate formate-lyase B) can be eliminated from chromosomal DNA of the parental host strain. To obtain high succini c acid production yield, the carbon flux through the phosphoenolpyruvate carboxylation route should be active rather than that through pyruvate during anaerobic fermentation (Figure 1-2). Othe r genes involved in producing organic acids other than succinate should be also inactivated so that the mutant E. coli strain channels the phosphoenolpyruvate to succinic acid. Moreover the foreign genes using during genetic manipulations can be eliminated from the E. coli genome. Metabolic Evolution in Metabolic Engineered E. coli Genetica lly modified E. coli strains obtained from gene deletions will be subsequently selected for the best representative clone via me tabolic adaptation or evol ution. The cultures will be repeatedly transferred into fresh minimal me dium for a period of time to achieve a clone in
38 which the spontaneous mutations that occurred during selection result ed in phenotypes that exhibits fast cell growth, rapi d consumption of different car bon sources, and high production yield and productivity of succinic acid, but low production of other organic acids. Metabolic Flux Analysis of Metabolic Engineered E. coli Flux distributions of metabolic intermediate s and metabolic excretes will be determined by using Flux Balance Analysis (FBA) (Jin et al., 1997; Nielsen and Villadsen, 1994; Villadsen et al., 1997; Vallino and Stephanopoulos, 1993; Stephanopoulos and Vallino, 1991). The deletions of central anaerobic metabolic gene s will affect the part itioning of carbon fluxes around the phosphoenolpyruvate and pyruvate nodes. To better unde rstand the redistribution of metabolic intermediates, the experimental data from fermentation profiles of the mutant strains constructed in this study will be examined to provide further insight into the dynamics of the succinic acid production pathway. Thus, the analyzed FBA data will be useful in comparing the production of succinic acid from the E. coli mutant strains in which we would know which combinatorial enzymes would suitably be deleted from the central anaerobic metabolic pathway to achieve the highest succinic acid production yield.
39 CHAPTER 2 GENERAL PROCEDURES Medium Preparation For strain construction, E. coli wild type and mutant strain s were grown in Luria-Bertani (LB) medium containing 1%(w/v) tryptone, 0.5%(w/v) yeast extract and 0.5%(w/v) NaCl. Antibiotics, kanamycin: 25 g/ml, chloramphenicol: 40 g/ml, tetracycline: 12.5 g/ml, and ampicillin: 50 g/ml, were used for selection of the E. coli transformants ha rboring antibioticresistant genes. The resulting succinic acid-producing strains were maintained on mineral NBS or AM1 salt medium with trace metal. Anaer obic fermentation cultures were composed of minimal media containing various concentrations of glucose as carbon source depending on the experiment, and supplemented with 100 mM pota ssium bicarbonate and 2 mM betaine HCl. The base used in the anaerobic fermentation experiments is a mixture of 6N potassium hydroxide and 3M potassium carbonate. Potassium carbonate is not only used for ne utralization; it also provides carbon dioxide indirectly for the PEP carboxylation pathway. Me dia were sterilized by autoclaving at 15 p.s.i for 20-30 min. For agar plates, 15%(w/v) bacter iology agar was added before autoclaving. Growth of Bacterial Cultures E. coli wild type and m utant stra ins were grown at either 30 C or 37 C, depending on the experiment, in LB broth with shaking at 200 rpm or overnight on LB agar plates. For selection of antibiotic resistant colonies, the E. coli cells were grown on LB agar plates containing antibiotics at different concentrations depending on the e xperiment. For metabolic evolution and batch experiment, NBS and AM1 salts medium supplemen ted with glucose were used instead of LB medium for cultivating the strains.
40 Plasmid Preparation by Alkaline Lysis Meth od A single colony of E. coli was inoculated in 3.0 ml of LB broth, containing antibiotics and incubated with shaking at 37 C for 16-18 hours. The E. coli culture was transferred to 1.5 ml microcentrifuge tube and centrifuged at 12,000 rpm for 1 min. The pellet was resuspended in 250 l of Alkaline lysis I solution [50 mM glucos e, 10 mM EDTA and 25 mM Tris pH 8.0 with RNaseA]. Two hundreds and fifty microliters of Alkaline lysis II solution [0.2N NaOH, 1% (w/v) SDS] were added to the cell mixture. Th e cell suspension was mixed gently to prevent shearing of chromosomal DNA. Three hundreds a nd fifty microliters of Alkaline lysis III solution [60 ml of 5M potassium acetate, 11.5 ml of acetic acid, 28.5 ml distilled water] were added to the cell mixture. The mixture was im mediately centrifuged at 14,000 rpm for 15 min at room temperature. A supernatant was removed from the white cell pellet (chromosomal DNA and cell debris) and 0.1 volume of sodium acetate, pH 5.2, was added to supernatant and mixed well by inversion. Two volumes of iso-propanol or absolute ethanol was added allowing incubation at C for 15 min. The mixture was centrif uged at 14,000 rpm for 10 min at room temperature. The DNA pellet was rinsed in 70% ethanol, air dried and th en resuspended in 20 l TE buffer (10 mM Tris-HCl pH 8.0, 1 mM EDTA) or st erile distilled water. DNA Amplification by Polymera se Chain Rea ction (PCR) The standard PCR reaction was performed us ing 10X PCR Master Mix solution (Qiagen, Valencia, CA) in a PCR reaction of 50 l. Twenty five microliters of master mix containing 10 mM of each dNTP (dATP, dGTP, dCTP and dTTP ), PCR reaction buffer (20 mM Tris-HCl pH 8.8, 10 mM KCl, 10 mM (NH4)2SO4, 2 mM MgSO4, 1%(v/v) Triton X-100, 1 mg/ml nucleasefree BSA, and Taq polymerase enzyme), 40 pmole of each primer (forward and reverse strand primers), and 50 ng of either plasmid or chromosomal DNA template and distilled water, were
41 added to the mixture. The reaction wa s performed in automated Mastercycler gradient PCR machine (Table 2-1). After the amplification reaction was finished, an aliquot of the PCR reaction mixture was examined on 0.8% (w/v) agarose ge l electrophoresis. Table 2-1. PCR parameters for the amplificatio n of specific genes. The extension time is depending on the length of the genes (1 kb/min). PCR profile to amplify genes Step Period Temperature ( C) Time (min) Number of cycles 1 Pre-denaturing 95 5 min 1 2 Denaturing Annealing Extension 95 55 72 30 sec 30 sec Vary 30 3 Extra-extension 72 10 min 1 Agarose Gel Electrophoresis of DNA To analyze the size of DNA frag ments and restriction patterns, the PCR product and DNA fragments were subjected to agarose gel electrophoresis. The appropriate amount of agarose powder was dissolved in 1X TBE buffe r [89 mM Tris-HCl, 89 mM boric acid, 25 mM EDTA pH 8.0] or 1X TAE buffer [40 mM Tris -HCl, 40 mM acetic aci d, 25 mM EDTA pH 8.0] under boiling temperature to ensure the homogeneity of the gel solution. Five microliters of loading dye [0.1%(w/v) bromopheno l blue, 40%(w/v) Ficoll and 5 mM EDTA)] was added and mixed well to the DNA samples before loading into the wells of th e solidified gel. The electrophoresis was performed at a constant vo ltage, 80 V, for 1 hour. After completion of electrophoresis, the gel was stained with 2 g/ml ethidium bromide for 5-10 minutes and destained in distilled water for 10 min. The DNA bands were visualized under UV light and photographed by a gel documentation system. Restriction Endonuclease Digestion of DNA The 20 l reaction m ixture, 0.2-1 g of plasmid DNA was com posed, 1X restriction endonuclease buffer, restriction endonuclease enzy me, approximately 5 U, and sterile distilled
42 water. The restriction endonuclease buffer, the am ount of restriction endo nuclease used and the optimum condition for digestion were selected acc ording to the manufacturers instructions. DNA Ligation DNA ligation reaction was carried out using T4 DNA ligase. The linearized vector was com bined with insert DNA at the molar ratio of 1:5. The ligation mixture contained 1X ligation buffer [50 mM Tris-HCl pH 7.6, 10 mM MgCl2, 1 mM ATP, 1 mM dithiothreitol and 5%(w/v) polyethylene glycol-8000], T4 D NA ligase and sterile distilled wa ter in a final volume of 10-20 l. The ligation mixture was incubated at 14-16 C for overnight. The amount of T4 DNA ligase (1-10 units) added to the r eaction depends on the amount of total DNA in the reaction. Preparation of E. coli Competent Cells Preparation of E. coli Competent Cells by CaCl2 Method A single colony (diameter of about 2-3 mm) of E. coli was inoculated into 3 ml of LB broth and incubated at 37 C for overnight. Cells were dilu ted to 1:100 in LB medium and incubated at 37 C with shaking until the OD600 was 0.3-0.5. The culture was centrifuged at 3,000 rpm, 4 C for 10 min. The pellet was re-suspended and washed in 5 ml of ice-cold CaCl2 for 2 times. After washing the cell, the white cell pelle t was re-suspended in 2 ml of ice-cold CaCl2 and placed on ice for 1 hour. Glycerol was a dded into the cell suspension at 15%(v/v) final concentration then 200 l aliquots were stored at C. Preparation of E. coli Competent Cells by Ele ctro-transformation Method for E. coli Carrying Temperature Sensitive Plasmid A single colony (diameter of about 2-3 mm) of E. coli was inoculated into 3 ml of LB broth and incubated at 30 C for overnight. Cells were dilute d to 1:100 in LB medium with ampicillin and L-arabinose, and incubated at 30 C with viscous shaking until the OD600 reached 0.3-0.5. The culture was centrifuged at 3,000 rpm, 4 C for 10 min. The pellet was re-suspended
43 and washed in 5 ml of sterile ice-cold nano-pur e water for 4 times. After washing the cell, the white cell pellet will be re-suspended in 1 ml of sterile ice-cold nano-pure water. Eighty microliters of aliquot were dispensed into electroporation cuvette. Transformation of Competent Cells Transformation of E. c oli by Heat Shock Method Plasmid DNA, 1-3 g, was mixed gently with 200 l of E. coli competent cells and placed on ice for 30 min. The cells were heat-shocked at 42 C for 90 seconds and incubated on ice for additional 5 min. The transformed cells were mixed with 800 l of LB broth and incubated at 37 C for 1 hour. Two hundreds microliters of transformed cells were then plated on LB agar plates containing suitable antibiotics de pending on antibiotic resist ant genes harbored in the plasmid, and incubated overnight at 37 C for non-temperature sensitive plasmids or 30 C for temperature sensitive plasmids. Transformation of E. c oli by electroporation Linearized DNA, 100 ng-10 g (in 5-10 l of sterile water), was mixed with electroporated competent cells, a nd the mixture was transferred to an ice-cold 0.2 cm Gene Pulser cuvette. The cuvette was inc ubated on ice for 5 minutes. The cells were pulsed by using Bio-Rad gene pulser under the conditions used with E. coli (2,500 V, pulse length 5 ms). Then 1 ml of 1 M ice-cold LB broth was added to the cuvette immediately and the solution was transferred to a sterile 15 ml t ube. The tube was incubated at 30 C with 150 rpm shaking for 1 hour. Transformed cells, 200 l, was spread on LB agar plates containing suitable antibiotics depending on antibiotic resistant genes harbored in the DNA fragments, and incubated overnight at 37C.
44 Anaerobic Fermentation Seed cultures and fermentations were grown at 37 C, 100 rpm, in mineral salts NBS (Causey et al., 2004) or AM1 me dium (4.2 g/l total salts; Ma rtinez et al., 2007) containing 10%(w/v) glucose supplemented with 100 mM pota ssium bicarbonate and 2 mM betaine HCl. Fermentations were maintained at pH 7.0 by au tomatically adding a mixture of 3M potassium carbonate and 6N potassium hydroxide. Fermentations were carried out in a container with a 350 ml working volume out of 500 ml total volume Temperature was controlled by means of submersion of containers in a thermo-regulated water bath. A magnetic s tirrer beneath the bath mixed the cultures continuously. Samples were re moved from the containers during fermentation aseptically by syringe connected to the vessels. To perform metabolic evolution in the mutant strains, seed cultures were inoculated to the vessels at the final OD550 of 0.1. Twenty-four hour cultures were daily withdrawn from the vessels, and transferred into the new fresh medium by diluting the cultures to final OD550 of 0.1. The cultures were transferred daily until the phenotypes of cultures improve in terms of growth rate, fast substrate consumption, and high succinic acid production yield. Analytical Methods Fermentation samples were removed during fermentation for the measurement of cell mass, organic acids, and sugars. Cell mass was estimated from the optical density at 550 nm (0.33 mg of cell dry wei ght/ml) with a Bausch Lomb Spectronic 70 spectrophotometer. Organic acids and sugars were determined by using high performance liquid chromatography, HPLC, (Hewlett Packard 1090 series II) equipped w ith UV and refractive index detectors with a Bio-Rad Aminex HPX-87H ion exclusion column. Th e mobile phase used in the HPLC system is 4 mM sulfuric acid. Cultures co llected from the fermentor were centrifuged to separate cells and
45 supernatant. The supernatant was further filtrated passing through a 0.2 m filter prior to injecting to the HPLC. The 10 linjection volumes were automatically analyzed. Organic acids were separated in the column according to their molecular weight and structure.
46 CHAPTER 3 COMBINING METABOLIC ENGINEERIN G AND MET ABOLIC EVOLUTION TO DEVELOP NONRECOMBINANT STRAINS OF Escherichia coli ATCC 8739 THAT PRODUCE SUCCINATE AND MALATE Introduction The ferm entative production of succinate fr om renewable feedstocks will become increasingly competitive as petroleum prices increase. Succinate can serve as a substrate for transformation into plastics, solv ents, and other chemicals curren tly made from petroleum (Lee et al., 2004; Lee et al., 2005; Mc Kinlay et al., 2007; Wendisch et al., 2006; Zeikus et al., 1999). A variety of genetic approaches have been used to engineer strains of E. coli for succinate production with varying degrees of success (Table 1-2). Again complex ingredients have been used in the media with thes e recombinants. Many succinate-producing strains have been developed by deleting competing pathways a nd overexpressing native genes using plasmids. In this chapter, we describe novel strains of E. coli ATCC 8739 that produce succinate at high titers and yields in mine ral salt media during simple, pH-c ontrolled, batch fermentations without the addition of heterol ogous genes or plasmids. During development, an intermediate strain was characterized that produced malate as the dominant product. Materials and Methods Strains, Media and Growth Conditions NBS m ineral salts medium (Causey et al., 2004) supplemented with 100 mM KHCO3, 1 mM betaine HCl, and sugar (2%(w/v) to 10%(w/v)) was used as a fermentation broth in most studies and for maintenance of strains (Table 3-1). A new low salt medium, AM1 (4.2 g/l total salts; Martinez et al ., 2007), was developed during the latter st ages of this investigation and used in fermentations with KJ060 and KJ073 (Table 3-1). This medium was supplemented with 100 mM KHCO3 and sugar as indicated and includes 1 mM betaine when initial sugar concentrations
47 were 5%(w/v) or higher. No gene encoding antibi otic resistance, plasmid, or other foreign gene is present in strains devel oped for succinate production ex cept in intermediates during construction. Strains used in this study are summ arized in Table 3-2. Derivatives of E. coli ATCC 8739 were developed for succinate production by a unique combination of gene deletions and selections for increased producti vity. Cultures were grown at 37oC in modified LB broth (per liter: 10 g Difco tryptone, 5 g Difco yeast extr act, 5 g sodium chloride) (Miller, 1992) only during strain construction. Antibiotics were included as appropriate. Table 3-1. Composition of mineral sa lts media (excluding carbon source) Component Concentration (mmol/l) aNBS + 1 mM betaine AM1 + 1 mM betaine KH2PO4 25.72 0 K2HPO4 28.71 0 (NH4)2HPO4 26.50 19.92 NH4H2PO4 0 7.56 Total PO4 80.93 27.48 Total N 53.01 47.39 bTotal K 84.13 1.00 MgSO4 7H2O 1.00 1.50 CaCl2 2H2O 0.10 0 Thiamine HCl 0.015 0 Betaine-KCl 1.00 1.00 (mol/l) c FeCl3 6H2O 5.92 8.88 CoCl2 6H2O 0.84 1.26 CuCl2 2H2O 0.59 0.88 ZnCl2 1.47 2.20 Na2MoO4 2H2O 0.83 1.24 H3BO3 0.81 1.21 MnCl2 4H2O2 0 2.50 Total Salts 12.5 g/l 4.1 g/l a NBS + 1 mM betaine: NBS medium amended with betaine (1 mM). b Calculation includes KOH used to neutralize betaine-HCl stock. c Trace metal stock (1000X) was prepared in 120 mM HCl.
48 Genetic Methods Plasm ids and primers used in this study are summarized in Table 3-2. Methods for chromosomal deletions, integration, and removal of antibiotic resistance genes have been previously described (Datsenko and Wanner, 2000; Grabar et al ., 2006; Posfai et al., 1997; Zhou et al., 2006a). Sense primers contain sequences corre sponding to the N-terminus of each targeted gene (boldface type) followed by 20 bp (underlined) corres ponding to the FRT-kanFRT cassette. Anti-sense primers contain sequences corresponding to the C-terminal region of each targeted gene (boldface type) followed by 20 bp (underlined) corresponding to the cassette. Amplified DNA fragments were electroporated into E. coli strains harboring Red recombinase (pKD46). In resulting recombinants, the FRT -kanFRT cassette replaced the deleted region of the target gene by homologous recombination ( double-crossover event). The resistance gene (FRT -kanFRT) was subsequently excised from the chromosome with FLP recombinase using plasmid pFT-A, leaving a scar region contai ning one FRT site. Chromosomal deletions and integrations were verified by testing for antibiotic markers, PCR analysis, and analysis of fermentation products. Generalized P1 phage trans duction (Miller, 1992) was us ed to transfer the ( focA-pflB ) ::FRT -kanFRT mutation from strain SZ204 into strain KJ017 to produce KJ032. Deletion of mgsA and poxB Genes A m odified method was developed to delete E. coli chromosomal genes using a two-step homologous recombination process (Thomason et al., 2005). With this method, no antibiotic resistance genes or scar sequen ces remain on the chromosome af ter gene deletion. In the first recombination, part of the target gene was replaced by a DNA cassette containing a chloramphenicol resistance gene ( cat ) and a levansucrase gene ( sacB ). In the second recombination, the cat-sacB cassette was replaced with native sequences omitting the region of
49 deletion. Cells containing the sacB gene accumulate levan during incubation with sucrose and are killed. Surviving recombinants are highly enriched for loss of the cat-sacB cassette. A cassette was constructed to f acilitate gene deletions. The cat-sacB region was amplified from pEL04 (Lee et al., 2001; Thomason et al., 2005) by PCR using the JM catsacB primer set (Table 3-3), digested with Nhe I, and ligated into the corresponding site of pLOI3421 to produce pLOI4151. The cat-sacB cassette was amplified by P CR using pLOI4151 (template) and the cat -up primer set ( EcoR V site included in each primer), digested with EcoR V, and used in subsequent ligations (Jantama et al., 2008a). Table 3-2. Escherichia coli strains, plasmids, and primers used in this study Relevant Characteristics Sources E. coli Strains Wild type ATCC 8739 ATCC KJ012 E. coli ATCC 8739 Wild type, ldhA:: FRT adhE:: FRT ackA:: FRT This study KJ017 KJ012, improved strain selected from 10% glucose, NBS This study KJ032 KJ017, ldhA:: FRT adhE:: FRT ackA:: FRT ( focA pflB ) :: FRT This study KJ060 KJ032, improved strain selected from 10% glucose without initial acetate, NBS This study KJ070 KJ060, mgsA This study KJ071 KJ070, improved strain selected from 10% glucose, NBS This study KJ072 KJ071, poxB This study KJ073 KJ072, improved strain selected from 10% glucose, AM1 This study SZ204 ( focA pflB ) :: FRTkanFRT Zhou, 2003 Plasmids pKD4 bla FRTkan -FRT Datsenko, 2000 pKD46 bla exo (Red recombinase), temp erature-conditional replicon Datsenko, 2000 pFT-A bla flp temperature-conditional replicon and FLP recombinase Posfai, 1997 pEL04 cat-sacB targeting cassette Lee, 2001 Thomason, 2005 pLOI3421 1.8 kbp SmaI fragment containing aac Wood, 2005 pLOI4151 bla cat; cat-sacB cassette This study pCR2.1-TOPO bla kan; TOPO TA cloning vector Invitrogen pLOI4228 bla kan; yccT-mgsA-helD (PCR) from E.coli B cloned into pCR2.1-TOPO vector This study pLOI4229 cat-sacB cassette PCR amplified from pLOI4151 ( EcoR V digested) cloned into mgsA in pLOI4228 This study pLOI4230 PCR fragment amplified from pLOI4228 (using mgsA -1/2 primers), kinase treated, then selfligation This study pLOI4274 bla kan; poxB (PCR) from E.coli C cloned into pCR2.1-TOPO vector This study pLOI4275 cat-sacB cassette PCR amplified from pLOI4151 ( EcoR V digested) cloned into poxB of pLOI4274 This study
50 Table 3-2. (Continued ) Relevant Characteristics Sources pLOI4276 PCR fragment amplified from pLOI4274 (using poxB -1/2 primers), kinase treated, then selfligation This study Primer sets ldhA 5 ATGAACTCGCCGTTTTATAGCACAAAACAGTACGACAAGAAGTAC GTGTAGGCTGGAGCTGCTTC 3 5 TTAAACCAGTTCGTTCGGGCAGGTTTCGCCTTTTTCCAGATTGCT CATATGAATATCCTCCTTAG 3 This study adhE 5 ATGGCTGTTACTAATGTCGCTGAACTTAACGCACTCGTAGAGCGT GTGTAGGCTGGAGCTGCTTC 3 5 TTAAGCGGATTTTTTCGCTTTTTTCTCAGCTTTAGCCGGAGCAGC CATATGAATATCCTCCTTAG 3 Zhou, 2003 ackA 5 ATGTCGAGTAAGTTAGTACTGGTTCTGAACTGCGGTAGTTCTTCA GTGTAGGCTGGAGCTGCTTC 3 5 TCAGGCAGTCAGGCGGCTCGCG TCTTGCGCGATAACCAGTTCTTC CATATGAATATCCTCCTTAG 3 Zhou, 2003 focA-pflB 5 TTACTCCGTATTTGCATAAAAACCATGCGAGTTACGGGCCTATAA GTGTAGGCTGGAGCTGCTTC 3 5 ATAGATTGAGTGAAGGTACGAGTAATAACGTCCTGCTGCTGTTCT CATATGAATATCCTCCTTAG 3 This study JMcatsacB 5 TTAGCTAGCATGTGACGGAAGATCACTTCG 3 5 CCGCTAGCATCAAAGGGAAAACTGTCCATAT 3 This study cat -up 5 AGAGAGGATATCTGTGACGGAAGATCACTTCG 3 5 AGAGAGGATATCGAATTGATCCGGTGGATGAC 3 This study mgsAup/down 5 CAGCTCATCAACCAGGTCAA 3 5 AAAAGCCGTCACGTTATTGG 3 This study mgsA 1/2 5 AGCGTTATCTCGCGGACCGT3 5 AAGTGCGAGTCGTCAGTTCC 3 This study poxBup/down 5 AAGCAATAACGTTCCGGTTG 3 5 CCACTTTATCCAGCGGTAGC 3 This study poxB1/2 5 GACGCGGTGATGAAGTGAT 3 5 TTTGGCGATATAAGCTGCAA 3 This study The mgsA gene and neighboring 500 bp regions ( yccT-mgsA-helD 1435 bp) were amplified using primer set mgsA -up/down and cloned into the p CR2.1-TOPO vector (Invitrogen) to produce plasmid pLOI4228. A 1000-fold diluted prep aration of this plasmid DNA served as a template for inside-out amplification using the mgsA -1/2 primer set (both primers within the mgsA gene and facing outward). The resulting 4958 bp fragment containing the replicon was ligated to the amplified, EcoR V-digested cat-sacB cassette from pLOI4151 to produce pLOI4229. This 4958 bp fragment was also us ed to construct a second plasmid, pLOI4230 (phosphorylation and self-ligation). In pLOI4230, the central region of mgsA is absent ( yccTmgsA-mgsAhelD ). After digestion of pLOI4229 and pLOI4230 with Xmn I (within the vector), each served as a template for amplification using the mgsA -up/down primer set to produce the linear DNA fragments for integration step I (yccT-mgsAcat-sacBmgsAhelD ) and step II ( yccT-mgsA- mgsAhelD), respectively. After electroporation of the step I fragment into KJ060 containing
51 pKD46 (Red recombinase) and 2 h of incubation at 30oC to allow expression and segregation, recombinants were selected for chloramphenico l (40 mg/l) and ampicillin (20 mg/l) resistance on plates (30oC, 18 h). Three clones were chosen, grown in LB with ampicillin and 5% (w/v) arabinose, and prepared for elect roporation. After electroporation with the step II fragment, cells were incubated at 37oC for 4 h and transferred into a 250-ml flask containing 100 ml of modified LB (100 mM MOPS buffer added and NaCl omitt ed) containing 10% (w/v) sucrose. After overnight incubation (37oC), clones were selected on modified LB plates (no NaCl; 100 mM MOPS added) containing 6% (w/v) sucrose (39oC, 16 h). Resulting clones were tested for loss of ampicillin and chloramphenicol resistance. Cons truction was further confirmed by PCR analysis. A clone lacking the mgsA gene was selected and designated KJ070. The poxB gene was deleted from KJ071 in a manner analogous to that used to delete the mgsA gene. Additional primer sets ( poxB -up/down and poxB -1/2) used to construct the poxB deletion are included in Table 3-2 together with the corresponding plasmids (pLOI4274, pLOI4275, and pLOI4276). The resulti ng strain was designated KJ072. Fermentations Seed cultures and ferm entations were grown at 37 C, 100 rpm in NBS or AM1 mineral salts medium containing glucose, 100 mM KHCO3 and 1 mM betaine HCl. These were maintained at pH 7.0 by the automatic addi tion of KOH during initial experiments. Subsequently, pH was maintained by adding a 1:1 mixture of 3M K2CO3 and 6N KOH. Fermentations were carried out in small fermen tation vessels with a working volume of 350 ml. Fermentations were inoculated at either an initial OD550 of 0.01 (3.3 mg CDW/l) or 0.1 (33.3 mg CDW/l) as indicated. No antibiotic resistance gene was present in the strains that were tested. Fermentation vessels were sealed except for a 16 -gauge needle, which served as a vent for
52 sample removal. Anaerobiosis was rapidly ach ieved during growth with added bicarbonate serving to ensure an atmosphere of CO2. Analyses Sa mples were removed during fermentation for the measurement of cell mass, organic acids, and sugars. Cell mass were estimated from the optical density at 550 nm (0.33 mg of cell dry weight/ml) with a Bausch Lomb Spectronic 70 spectrophotometer. Organic acids and sugars were determined by using high perf ormance liquid chromatography, HPLC, (Hewlett Packard 1090 series II) equipped with UV and refr active index detectors with a Bio-Rad Aminex HPX-87H ion exclusion column. The mobile phase used in the HPLC system is 4 mM sulfuric acid. Cultures collected from the fermentor were previously centrifuged to separate cells and supernatant. The supernatant was further filtrated passing through a 0.2 m filter prior to injecting to HPLC. Ten microliters of injection volume were automatically analyzed. Organic acids were separated in the column depending on their retention times according to their molecular weight and structure. Results and Discussion Construction of KJ012 for Succinate Production by Deletion of ldhA adhE and ackA E. coli prod uces a mixture of lactate, acetate, ethanol and succinate during glucose fermentation (Figure 3-1). Major pathways leading to lactate, acet ate and ethanol were eliminated by deleting genes encoding fermentative D-lactate dehydrogenase, acetate kinase, and alcohol dehydrogenase to construct KJ012 ( ldhA::FRT adhE::FRT ackA::FRT ). This strain retained only the succinate pa thway with malate dehydrogenase and fumarate reductase as primary routes for NADH oxidation (Figure 3-2). Although strain KJ012 grew well in complex media (not shown), poor growth was observed in mineral-based media such as NBS and acetate
53 was produced as the most abundant product from sugar metabolism (Table 3-3). In this NBS mineral salts medium, succinate t iter and cell yield for KJ012 were 8-fold to 7-fold lower than the unmodified parent, E. coli ATCC 8739. Poor growth and glucose fermentation could re sult from insufficient capacity to oxidize NADH using only the succinate pathway (Figure 3-1), from a deficiency of metabolites and precursors for biosynthesis (error in metabolic part itioning) due to shifts in pool sizes resulting from gene deletions, or a combination of both. Gr owth was increased 5-fold (Table 3-3) in the same medium (plus 100 mM MOPS for pH control) by providing mild aeration (100 rpm, 100 ml NBS broth, 250-ml flask) indicating that the NADH oxidation capacity limited growth. Growth was also increased (5-fold) by replacing NBS mi neral salts medium with LB indicating a deficiency in metabolic partitioning. With the co mplex nutrients of LB, succinate titers were increased 18-fold over NBS medium indicating that the succinate pathway served as the primary route for NADH oxidation in KJ012. From these resu lts, we concluded that the poor performance of KJ012 under anaerobic conditions is a comple x problem resulting primarily from suboptimal metabolic partitioning of carbon between growth and fermentation products. It is likely that previous investigators have made similar observations with E. coli engineered for succinate producti on and that this is the basis for their reliance on complex medium and two-phase process (aerobic growth and anaerobic producti on) (Andersson et al., 2007; Millard et al., 1996; Sanchez et al., 2005a; Stols et al., 1997; Vemu ri et al., 2002a, b). Absent a clear approach to decrease the co mplexity of this problem in engineered E. coli, we elected to pursue an adaptive genetic strategy termed metabolic evolution. Rather than incremental studies to identify and solve spec ific problems in metabo lic partitioning, natural selection was used to solve the aggregate pr oblem. The production of ATP for growth by KJ012
54 and all subsequent derivatives in this paper is obligately c oupled to malate and succinate production during the oxidation of NADH (Figure 32). Selection for improved growth and ATP production during serial cultivati on provides a route to co-select improved growth and improved production of these dicarboxylic acids. Analogous approaches have been used to develop E. coli strains for the production of Lalanine (Zhang et al., 2007) and l actate (Zhou et al., 2003; Zhou et al., 2006a, b). Figure 3-1. Fermentation of glucose to succinate Central metabolism indicating genes deleted in constructs engineered for succinate production. Solid arrows represent central fermentative pathways. Dashed arrow represen ts microaerophilic pathway ( poxB ). Dotted arrows show pathways that normally function during aer obic metabolism, pyruvate dehydrogenase ( pdh) and the glyoxylate bypass (aceAB). Crosses represent the gene deletions performed in this study to obtain KJ012 ( ldhA, adhE, ackA ), KJ032 ( ldhA, adhE, ackA, focA-pflB ), and KJ070 (ldhA, adhE, ackA, focA-pflB, mgsA ), and KJ072 ( ldhA, adhE, ackA, focA-pflB, mgsA, poxB ). Genes and enzymes: ldhA lactate dehydrogenase; focA formate transporter; pflB pyruvate formate-lyase; pta phosphate acetyltransferase; ackA, acetate kinase; adhE, alcohol dehydrogenase; ppc, phosphoenolpyruvate carboxylase; pdh, pyruvate dehydrogenase complex; gltA, citrate synthase; mdh, malate dehydrogenase; fumABC, fumarase isozymes; frdABCD, fumarate reductase; fdh, formate dehydrogenase; mgsA methylglyoxal synthase; gloAB glyoxylase I and II; poxB pyruvate oxidase; aceAB, isocitrate lyase; acnAB aconitase; acs, acetyl-CoA synthetase, aldA and aldB, aldehyde dehydrogenase isozymes; and icd isocitrate dehydrogenase.
55 Figure 3-2. Coupling of ATP pr oduction and growth to succinate and malate production in engineered strains of E. coli. Solid arrows connect NAD H pools. Dotted arrows connect NAD+ pools. During glycolysis under an aerobic conditions, the production of ATP for cell growth is obligately coupled to the oxidation of NADH. Improvement of KJ012 by Metabolic Evolution KJ012 ( ldhA::FRT adhE::FRT ackA::FRT) grew poorly in com parison to the parent E. coli ATCC 8739, exhibited lower rates of succina te production, and provide no better molar yields (Table 3-3). Despite these re sults, serial transfer of this st rain was tried as a method to coselect improved growth and su ccinate production based on the following rationale. The primary pathway for the glucose fermentation into succinat e (Figure 3-1 and 3-2) is generally thought to use phosphoenolpyruvate carboxylase ( ppc ) for the carboxylation step (Unden and Kleefeld, 2004; Fraenkel, 1996; Keseler et al., 2005; Millard et al., 1996; Gottschalk, 1985; Karp et al., 2007). This carboxylating enzyme does not conserve the high-energy phosphate in phosphoenolpyruvate and reduces the net ATP available for growth. Serial transfers of KJ012 with selection for improved growth offered an opp ortunity to select for mutational activation of
56 alternative routes for succinat e production (Figure 3-2) that maintained redox balance and increased ATP yields. Metabolic evolution was carried out by seque ntially subculturing under various regimens using small, pH-controlled fermentors (Figure 3-3). Selection was begun using 5%(w/v) glucose with serial transfers at 120-h in tervals (Figure 3-4A and 3-5A). Addition of betaine, a protective osmolyte (Purvis et al., 2005; Underwood et al., 2004; Zhou et al., 2006b), increased cell growth and cell yield and allowed more frequent transf ers. Beginning at transfer 27, a bicarbonate solution was used to maintain pH and provide additional CO2 for succinate production. The rapid subsequent improvement confirmed that CO2 had become limiting (Figure 3-5A). With further selection in 5%(w/v) glucose, molar yields of succinate improved to 0.73 per mole of glucose metabolized (Table 3-3). The glucose concentration was doubled and transfers continued (Figure 3-4B and 3-5B) with modest improvement in growth (initial 24 h) and a small decline in succinate yield. With 10%(w/v) glucose, unwanted co-products (acetate, formate, and lactate) were abundant despite the absence of the primary lactate dehydrogenase ( ldhA ) and acetate kinase ( ackA ) (Table 3-3). A clone was isolated from the last transfer and designated KJ017 ( ldhA::FRT adhE::FRT ackA::FRT). Construction of KJ032 and KJ060 The gene en coding pyruvate formate-lyase ( pflB ) was deleted from KJ017 to eliminate the loss of reductant as formate and excessive pr oduction of acetyl~CoA, a potential source of acetate. The upstream formate transporter ( focA ) in this operon was also deleted. As expected, this deleted strain (KJ032) did not grow without acetate (Figure 3-4C). Deletion of pflB is well known to cause acetate auxot rophy under anaerobic conditio ns (Sawers & Bock, 1988).
57 Table 3-3. Fermentation of glucose in mineral salts medium by mutant strains of E. coli Succinate Yieldc Fermentation Products (mM) e Straina Culture Conditions Media, Gluc (w/v) Cell Yieldb (g/L) mol/mol g/g Av. Vol. Prod d (g/L-h) Suc Mal Pyr Ace Lac For E. coli wild type f 0.1 OD550, 0.1 mM betaine 5%, NBS 2.00.2 0.190.02 0.12 0.120.01 493 33 0 152 0 98 4 262 9 KJ012 f 0.1 OD550, 0.1 mM betaine 5% NBS 0.30.1 0.200.01 0.13 0.040.01 60.4 26 1 KJ012 0.1 OD550, 0.1 mM betaine Shaken flask f 5% NBS + MOPS 1.5 0.10 0.06 0.02 10 226 16 KJ012 0.1 OD550 Luria Broth 5% LB 1.5 0.70 0.50 0.09 108 61 <2 14 1st TF: No betaine, 0.1 OD550, 120 h transfers 5%, NBS 0.3 0.13 0.09 0.072 6 26 <2 3rd TF: 2 mM betaine, 0.1 OD550, 96 h transfers 5%, NBS 0.7 0.28 0.18 0.128 26 71 <2 40th TF: 1 mM betaine, 0.1 OD550, 24 h transfers, 3M K2CO3+6N KOH 5%, NBS 2.3 0.73 0.48 0.251 204 179 <2 151 KJ012 ( ldhA, ackA, adhE ) (KJ017) 40th TF: 1 mM betaine, 0.1 OD550, 24 h transfers, 3M K2CO3+6N KOH 10%, NBS 1.7 0.74 0.49 0.354 288 181 38 199 2nd TF: 1 mM betaine, 0.1 OD550, 48 h transfers, 20 mM NaOAc, 3M K2CO3+6N KOH 5%, NBS 1.0 1.47 0.97 0.260 212 44 15th TF: 1 mM betaine, 0.01 OD550, 24h transfers, 5 mM NaOAc, 3M K2CO3+6N KOH 10%, NBS 1.4 1.07 0.71 0.736 596 331 9 170 <2 KJ032 ( ldhA, ackA, adhE, focA, pflB ) (KJ060) 5th TF: 1 mM betaine, 0.01 OD550, 24 h transfers, No NaOAc, 3M K2CO3+6N KOH 10%, NBS 1.4 1.04 0.69 0.711 579 318 9 161 <2
58 Table 3-3. ( Continued ) a Clones were isolated from the fermentation broth at various point s and assigned strain numbers, indicated by numbers in parenth esis b Cell yield estimated from optical density (3 OD550nm = 1 g l-1 CDW). c Succinate yields were calculate d based on glucose metabolized. d Average volumetric productivity was calculated for tota l incubation time. e Abbreviations: suc, succinate; mal, malate; pyr, pyruvate; ace, acetate; lac, lacate; for, formate. f Average of 3 or more fermentations with standard devia tions. Aerobic shaken flask (100 rpm; 100 ml NBS, 250-ml flask). Succinate Yieldc Fermentation Products (mM) e Straina Culture Conditions Media, Gluc (w/v) Cell Yieldb (g/L) mol/mol g/g Av. Vol. Prod d (g/L/h) Suc Mal Pyr Ace Lac For 1st TF: 1 mM betaine, 0.01 OD550, 24 h TF, 3M K2CO3+6N KOH, 5%, NBS 1.0 1.06 0.70 0.361 294 219 25 102 KJ070 ( ldhA, ackA, adhE, focA, pflB, mgsA) (KJ071) 50th TF: 1 mM betaine, 0.01 OD550, 24 h transfers, 3M K2CO3+6N KOH 10%, NBS 1.1 0.71 0.47 0.419 341 626 <2 76 2nd TF: 1 mM betaine, 0.01 OD550, 24 h transfers, 3M K2CO3+6N KOH 10%, NBS 1.3 0.97 0.64 0.663 539 186 <2 95 6th TF: 1 mM betaine, 0.01 OD550, 24 h transfers, 3M K2CO3+6N KOH 10%, AM1 1.2 1.34 0.88 0.733 596 38 4 112 KJ072 ( ldhA, ackA, adhE, focA pflB, mgsA, poxB ) (KJ073) 45th TF: 1mM betaine, 0.01 OD550, 24 h transfers, 3M K2CO3+6N KOH 10%, AM1 1.5 1.26 0.83 0.858 699 313 103 172 KJ073 f 1mM betaine, 3M K2CO3+6N KOH 0.01 OD550 inoculum 10%, AM1 2.3 .1 1.20 .09 0.77 .03 0.82 .01 668 118 3 55 2 183 7 KJ060 f 1mM betaine, 3M K2CO3+6N KOH 0.01 OD550 inoculum 10%, AM1 2.2 .1 1.41 .07 0.92 .05 0.90 .04 733 9 39 7 250 6 21 KJ060 f 1mM betaine, 3M K2CO3+6N KOH 0.60 OD550 inoculum 10%, AM1 2.2 .1 1.61 .12 1.05 .09 0.77 .04 622 17 1.5 180 3 21 KJ071 f 1mM betaine, 3M K2CO3+6N KOH 0.01 OD550 inoculum 10%, NBS 1.5 .0 0.78 .02 0.53 .01 0.33 .04 280 516 4 58 5 64
59 Figure 3-3 Steps in the genetic engineer ing and metabolic evolution of E. coli ATCC 8739 as a biocatalyst for succinate and malate producti on. This process represents 261 serial transfers providing over 2000 generations of growth-based selection. Clones were isolated from the final culture of each regimen and assigned strain designations, shown in parenthesis in Table 3-3. Figure 3-4 Growth during metabolic evolution of KJ012 to produce KJ017, KJ032, and KJ060. Strain KJ012 was sequentially transferre d in NBS medium c ontaining 5%(w/v) (A) and 10%(w/v) (B) glucose, respectively to produce KJ017. After deletion of focA and pflB the resulting strain (KJ032) was initiall y subcultured in medium supplemented with acetate (C). Acetate levels were d ecreased and subsequen tly eliminated during further transfers to produce KJ060. Broken line represents fermentation by KJ017 without acetate, added for comparis on. Symbols: optical density at OD550nm,
60 Figure 3-5. Fermentation products during the metabolic evolution of strains for succinate and malate production. Cultures were supplemen ted with sodium acetate as indicated. Black arrows represent the transition betw een fermentation conditi ons as indicated by text. No formate and only small amounts of lactate were detected during metabolic evolution of KJ032. No formate and lactate were detected during metabolic evolution of KJ070 and KJ072. A (5% w/v glucose) and B (10% w/v glucose), KJ012 to KJ017; C (5% w/v glucose) and D (10% w/v glucose), KJ032 to KJ060; E. 10% glucose, KJ070 to KJ071; F. 10% glucose, KJ072 to KJ073. Symbols for all: succinate; formate; acetate; malate; ,lactate; and pyruvate. Growth and succinate production by KJ032 were restored by the addition of 20 mM acetate (Figure 3-4C and 3-5C). Production of formate a nd acetate were substantially reduced as a result of pflB (and focA ) deletion. Although this strain required ace tate for growth, additional acetate was also produced during fermentation. The same phenomenon was previously reported for pflB deleted strains during the construction of E. coli K-12 biocatalysts for pyruvate production (Causey et al., 2004). Lactate leve ls were also reduced in KJ 032 (Table 3-3; Figure 3-5C). Subsequent transfers were accompanied by improvements in growth and succinate production. Added acetate was reduced, inocula size was re duced, and glucose concentration was doubled (10%(w/v)) during subsequent tran sfers (Figure 3-5D). After reduc ing added acetate to 5 mM, an unstable population emerged that produced elevated levels of malate at the expense of succinate.
61 After further transfers, acetate was omitted and a strain was developed that was no longer auxotrophic. However, succinate yields declined upon elimination of added acetate while malate and acetate levels increased. A small amount of pyruvate was also produced. A clone was isolated from the last tr ansfer and designated, KJ060 ( ldhA:: FRT adhE::FRT ackA::FRT ( focB-pflB ) ::FRT). Construction of KJ070 and KJ071 by De letion of Methylglyoxal Synthase ( mgsA ) The sm all amount of lactate present in the fermentation broths of various strains is presumed to originate from the methylglyoxal pathway (Figure 3-1; Grabar et al., 2006). Although this represents a small loss of yield, lact ate production by this pa thway is indicative of methylglyoxal accumulation, an inhibitor of bot h growth and glycolysis (Egyud and SzentGyorgyi, 1966; Grabar et al., 2006; Hopper and Cooper, 1971). Production of methylglyoxal and lactate were eliminated by deleting the mgsA gene (methylglyoxal synthase) in KJ060 to produce KJ070 ( ldhA::FRT adhE::FRT ackA::FRT ( focA-pflB ) :: FRT mgsA). Strain KJ070 was initially subcultured in 5%(w/v) glucose (Figure 3-5E). Deletion of mgsA is presumed to have increased glycolytic flux as ev idenced by the accumulation of pyr uvate in the medium (Table 33). This increase in glycolytic flux may also be responsible for the further decline in the succinate/malate ratio due to increased production of oxaloacetate, an al losteric inhibitor of fumarate reductase (Iverson et al., 2002; Sanwal, 1970). At transfer 21, glucose was doubled to 10 %(w/v) and transfers c ontinued. This higher level of glucose and subsequent transfers resu lted in further increas es in malate production, exceeding succinate in latter tran sfers (Figure 3-5E). Increased production of malate versus succinate in 10%(w/v) glucose is also consistent with increased glycolytic flux and inhibition of furmarate reductase by oxaloacetate. At transf er 50, 1.3 moles of ma late and 0.71 moles of
62 succinate were produced per mole of glucose metabolized (Table 3-3). Significant amounts of acetate were also produced. A new strain was isolated from the final subculture and designated KJ071 ( ldhA::FRT adhE::FRT ackA::FRT ( focA-pflB ) :: FRT mgsA). This strain may be useful for malate production. Construction of KJ072 and KJ073 by Deletion of Pyruvate oxidase ( poxB ) Although conversion of glucose to acetate is redox neutral, partitioning of carbon to acetate decreases the yield of succinate and mala te. Pyruvate oxidase represents a potential source of acetate and CO2 during incubation under microaer ophilic conditions (Causey et al., 2004) and was targeted for gene deletion. However, deletion of poxB to produce KJ072 ( ldhA::FRT adhE::FRT ackA::FRT ( focA-pflB ) ::FRT mgsA poxB ) did not reduce acetate production indicating that alternative pathways are involved. Eliminating poxB resulted in unexpected changes in fermentation products, an increase in succinate and decrease in malate (Table 3-3; Figure 3-5F). The mechanism for this improvement in succinate production is unknown but may be related to other activities of pyruvate oxidase such as acetoin production, decarboxylation, and carboligation (Ajl and Werkman, 1948; Chang and Cronan, 2000). Strain KJ072 was subjected to 40 further r ounds of metabolic evol ution in AM1 medium, a lower salts medium, with 10 %(w/v) glucose (Table 3-3; Fi gure 3-5F). Improvements in growth, cell yield and succinate production were observed during these transfers. Malate, pyruvate and acetate levels also increased. A clon e was isolated from the final transfer and designated KJ073 ( ldhA ::FRT adhE ::FRT ackA ::FRT ( focA pflB )::FRT mgsA poxB ). Fermentation of KJ060 and KJ073 in AM 1 Medium Containing 10%( w/v) Glucose Figure 3-6 shows batch fermentations with KJ060 and KJ073, the two best biocatalysts for succinate production. Although growth was co mpleted within the initi al 48 h of incubation,
63 succinate production continued fo r 96 h. One-third of succinate production occurred in the absence of cell growth. These strains produced succinate titers of 668-733 mM, with a molar yield of 1.2-1.6 based on glucose metabolized. Acetate, malate, and pyruvate accumulated as undesirable co-products and detracted from the pot ential yield of succinate (Table 3-3). The maximum theoretical yield of succinate from glucose and CO2 (excess) is 1.71 mol per mole glucose based on the following Equation 3-1: 7 C 6 H 12 O 6 + 6 CO 2 12 C 4 H 6 O 4 + 6 H 2 O (3-1) Figure 3-6. Production of succina te and malate in mineral salts media (10% glucose) by derivatives of E. coli ATCC 8739. A. Succinate production by KJ060 in AM1 medium. B. Succinate production by KJ 073 in AM1 medium. C. Production of malate by KJ071 in NBS medium. Fermentations were inoculated at a level of 33 mg DCW/l. Symbols for all: glucose; succinate; malate; cell mass. Conversion of Other Substrates to Succinate Although this study primarily focused on the conv ersion of glucose to succinate is well known that E. coli has the nativ e ability to metabolize a ll hexose and pentose sugars that are constituents of plant cell walls (Asghari et al., 1996; Underwood et al., 2004). Some strains of E. coli can also metabolize sucrose (Moniruzzaman et al., 1997). Strain KJ073 was tested for utilization of hexoses, pentoses, and using tube cultures containing 2% sugar. In all cases, these sugars were converted primarily to succinate. Strain KJ073 al so metabolized glycerol to
64 succinate. During incubation with 2% glycerol 143 mM glycerol was metabolized to produce 127 mM succinate with a molar yield of 0.89, 89% of the theoretical maximum. Production of Malate in NBS Medium Contai ning 1 mM Betaine and 10% (w/v) Glucose During growth-based selections, cultures were observed to vary in their production of malate (Table 3-3), a potentia lly useful alternative product. Malate was the most abundant product from KJ071 with 10%(w/v) glucose, almost double that of succinate (Table 3-3; Figure 3-5E). This strain produced 516 mM malate with a molar yield of 1.44 based on metabolized glucose. Conclusions The f ermentative metabolism of E. coli is remarkably adaptable. Derivatives can be readily engineered and evolved to circumvent numerous deletions of genes concerned with native fementation pathways and increase fluxe s through remaining enzymes to maintain redox balance, ATP production, and growth. Though arguabl y more challenging, cells can make such adaptive changes in mineral salts media while balancing carbon partitioning for biosynthetic needs. After eliminating the primary routes for NADH oxidation (lactat e dehydrogenase, alcohol dehydrogenase) and acetate production (acetate kinase), growth and ATP production remain linked to NADH oxidation and the pr oduction of malate or succinate (Figure 3-2). Growth-based selections for NADH oxidation and ATP production cannot readily distinguish between malate and succinate as end products, since the precursors of both serve as electron acceptors. During these investigations, KJ071 was de veloped that produces more malate than succinate. This strain and further derivatives may be useful for malate production. Other strains such as KJ073 and KJ060 may be useful for succinate production. Deletion of pflB the primary source of acetyl~CoA during anaeerobic growth, resulted in an auxotrophic requirement for acetate (S awers and Bock, 1988). This requirement was
65 eliminated through metabolic evolution, presumably due to increased prod uction of acetyl~CoA by other routes such as pyruvate dehydrogenase (de Graef et al., 1999). Ma ny shifts in metabolic products were unanticipated. Th e increase in malate during selections after deletion of mgsA is unexplained. Methylglyoxal is a me tabolic inhibitor that is produced in response to an imbalance in metabolism (Grabar et al., 2006). Elimination of methylglyoxal produc tion may have provided a growth-related advantage such as increased grow th rate, a shorter lag af ter inoculation, etc. The reduction in malate and shift to higher succinate production after a poxB deletion was also surprising. Little change in the acetate level was observe d indicating that eith er this enzyme was a minor source of acetate or that other route for acetate production functionally replaced it. After deletion of poxB succinate was again produced as the dominant dicarboxylic acid. With the best strains for succinate production, KJ060 and KJ073, malate and acetate remained as abundant coproducts (Table 3-3; Figure 3-5D and 3-5F). Elimination of thes e represents a further opportunity to increase yields. All previously engineered E. coli developed for succinate production have used complex media and plasmids with antibiotics for mainte nance. Most have achieved only low titers of succinate in simple batch fermentations, requiri ng more complex processes to achieve higher titers (Table 1-1 and Table 1-2). A variety of genetic approaches ha ve been reported that increase succinate production from glucose by recombinant E. coli in complex medium. In our initial construct, growth and sugar metabolism were very poor in mineral salts medium but very robust in complex, LB medium. It is likely that similar observations led previous investigators to their reliance on complex nutrients (T able 1-1 and Table 1-2). Comp lex media containing vitamins, amino acids, and other macromolecular precursors may mask potential regulatory problems in metabolism and biosynthesis that were created by metabolic engineering.
66 Many investigators have also used heterol ogous genes and complicated processes that include sparging with gas (CO2, H2, O2 or air) and dual aerobic and anaerobic process steps. This complexity of process and nutrients would be expected to increase the cost of construction, materials, purification, and waste disposal. In contrast, strains KJ060 and KJ073 produced high titers of succinate (600700 mM) in simple batch fermentatio ns (10%(w/v) sugar) using mineral salts medium without any comple x nutrients or foreign genes.
67 CHAPTER 4 BATCH CHARACTERIZATION AND METABO LIC DIS TRIBUTION OF EVOLVED STRAINS OF Escherichia coli ATCC 8739 TO PRODUCE SUCCINATE Introduction In chapter 3, I described the developm ent of E. coli strains that produc e succinate under anaerobic conditions in mineral salts medium (NBS or AM1) without the requirement for heterologous gene expression, rich nutrients, or antibiotics by combining metabolic engineering and metabolic evolution methods. In this chapter, I describe the effect of glucose concentration, inoculum size, growth supplements, and types of media on succinate produc tion. In this chapter, metabolic flux analysis (MFA) was performed to assess the distribution of intracellular metabolites when glucose concentration was varied. Materials and Methods Strains, Media and Growth Conditions Strains u sed in this study are summarized in Table 1 and indicated by an asterisk. No gene encoding antibiotic resistance (or other fore ign genes) is present in these strains. The precursors of the mutants used are also listed. NBS (Causey et al., 2004) mineral salts medium supplemented with 100 mM KHCO3, 1 mM betaine HCl, and glucose (5%(w/v) or 10%(w/v)) was used as a fermentation medium in most studies and for maintenance of strains. Fermentations Seed cultures and ferm entations were grown at 37 C, in a rotary shaker at 100 rpm in NBS or AM1 mineral salts media containing glucose, 100 mM KHCO3 and 1 mM betaine HCl. These were maintained at pH 7.0 by the automatic addition of KOH during initial experiments. Subsequently, pH was maintained by adding 1:1 mixtures of 3M K2CO3 and 6N KOH. Fermentations were carried out in small fermen tation vessels with a working volume of 350 ml.
68 Seed cultures were inoculated at an initial OD550 of 0.1 (33.3 mg CDW/l). Each experiment was carried out in triplicate. Analyses Cell m ass was estimated from the optical dens ity at 550 nm (OD 1.0 = 333 mg of cell dry weight/l) by using a Bausch Lomb Spectronic 70 spectrophotom eter. Organic acids and sugars were determined by using high performance li quid chromatography (Grabar et al., 2006). Ethanol was analyzed using gas chromatography (Ohta et al., 2001) Table 4-1 Escherichia coli strains used in this study Strain Relevant Characteristics Source ATCC 8739* Wild type ATCC KJ012 strain C, ldhA:: FRT adhE:: FRT ackA:: FRT Jantama et al., 2008a KJ017* improved strain obtained from metabolic evolution of KJ012 with 10% glucose, NBS Jantama et al., 2008a KJ032 KJ017, ( focA pflB ) :: FRT Jantama et al., 2008a KJ060* improved strain obtained from metabolic evolution of KJ032 with 10% glucose, NBS Jantama et al., 2008a KJ070 KJ060, mgsA Jantama et al., 2008a KJ071* improved strain obtained from metabolic evolution of KJ070 with 10% glucose, NBS Jantama et al., 2008a KJ072 KJ071, poxB Jantama et al., 2008a KJ073* improved strain obtained from metabolic evolution of KJ072 with 10% glucose, NBS Jantama et al., 2008a Indicates st rain used in this study. Calculation Specific Production Rates for Excreted Meta bolites The specific production rates of several excr eted metabolites during exponential growth were calculated using two measurements 24 or 48 hours apart and based on the log mean average cell concentration during that time interval (Aristidou, 1995): tX )t(c)tt(c ri i i (4-1)
69 where ci is the concentration of metabo lite i in the reactor (mmol/L) and X the log mean concentration of biomass (gCDW/L) )t(X )tt(X ln )t(X)tt(X X (4-2) Metabolic Flux Analysis Flux distributions of metabolic intermediate s and fermentation products were determined by standard methods (Jin et al., 1997; Nielse n and Villadsen, 1994; Vallino and Stephanopoulos, 1993; Stephanopoulos and Vallino, 1991). The intracellular fluxes were calculated using a stoichiometric model and applying mass balan ces around intracellular metabolites (Aristidou, 1994; Aristidou et al., 1999; Stephanopoulos et al., 1998). Figure 1 shows the fermentation pathways of E. coli. The reactions involved in the fermentation network are presented in Table 4-2. The material balances result in a set of linear algebraic equations that can be expressed in matrix notation as: r = K (4-3) where r is the vector for the net specific formation rates of 16 metabolites, v is the vector of up to 15 metabolic fluxes (mmol/g CDW-h) shown in Figure 1 (T = [1, 2, ,13, EM, EP]), and K is the matrix of stoichiometric coefficients presen ted in Table 4-3. The concentrations of eight extracellular products (glucose, succ inate, acetate, formate, lactate, ethanol, excreted malate, and excreted pyruvate) represented in the boxes of Figure 4-1 were measured, and their specific production rates were calculated using Equation 4-1. The net accumulation rates of intracellular metabolites were assumed to be zero. This is justified because pseudo-steady state can be assumed due to the high turnover of the metabolit e pool or, alternatively, balanced growth, and because the dilution term due to growth can be neglected due to low intracellular levels
70 (Stephanopoulos et al., 1998). For the NADH balance, it is assumed that the NADH production and consumption must be equal and this is also included as a constraint for the calculations. The solution of Equation 4-3 is obtained using the pseudo-invers e matrix since the system is overdetermined. This corr esponds to the least squares estimate. = (KT K) KT r (4-4) However, the least squares solution gave in some cases a small negative number for fluxes of two extracellular products, the formate flux, v7, and the excreted pyruvate flux, vEP. As these metabolites are not supplied in the medium, a negative flux is not possible. In such cases, the calculated negative flux was set to 0 and the calculation of v was repeated after removing from K the column for that flux. The net NADH specific production and utiliz ation rates per mole of glucose are NADHP/G = 1 134)( v vv (4-5) NADHU/G = 1 121196) 2( v vvvv (4-6) ATP and Cell Yield Analysis Reaction 5 is reversible, and the direction a ffects ATP production or consumption. If the reaction is forward (v5 > 0) it is catalyzed with PYK produc ing one mole of ATP per mole of PEP consumed, yielding the following net ATP s ynthesis rate per mole of glucose used ATP/G = 1 3854) ( v vvvv (provided v5 > 0) (4-7) If reaction 5 is reverse, one mole of ATP is consumed but this is converted to AMP (Table 2), which is equivalent to the loss of two moles of ATP to ADP. ATP/G = 1 3854) 2( v vvvv (provided v5 < 0) (4-8)
71 The cell yield (YATP) defines the amount of energy (A TP) required for the formation of microbial cells (g CDW/mole ATP). Th is can be calculated from the fluxes by YATP = ATP/G)(consumed) glucose of (moles produced biomass ofgCDW (4-9) Fermentation Pathways In this study we assume that succinate production occurs via the fermentative TCA route (Figure 1). In addition, E. coli can perform the glyoxylate bypa ss, which is an alternative pathway producing succinate. The glyoxylate bypass is normally active u nder aerobic conditions when it grows on acetate or fatty acids as carbon s ources (Cortay et al., 1989) but it is repressed by glucose (Nimmo and Nimmo, 1984). Under anaerobic conditions, IclR (isocitrate lyase transcriptional repressor) represses the transcription of the aceBAK operon that encodes isocitrate lyase, malate synthase, and isocitrate dehydr ogenase/kinase/phosphoprylase by binding to its promoter proximal site. Some E. coli strains, e.g. BL21, operate the glyoxylat e bypass during anaerobic growth due to cons titutive transcription of aceA and aceB and no transcription of iclR (Phue and Shiloach, 2004) We have confirmed that the succinate-producing st rains constructed in this study did not produce de tectable enzymatic activity of is ocitrate lyase (unpublished data). Therefore, the glyoxylate bypass was not cons idered in this flux analysis study. For many years, the PDH complex for pyr uvate conversion was considered inactive under anaerobic conditions (Guest et al., 1989; Unden et al., 2002). However, de Graef et al. (1999) showed low but signi ficant PDH activity when E. coli grows under anaerobic conditions. Kim et al. (2007) further showed that the PDH activity in the anaerobic cell could be significantly elevated by a single mutation in lpdA gene. The PDH complex provides an alternative route to acetyl~CoA, NADH, and CO2 from pyruvate. An extra NADH gained from the PDH activity can supply the additional requi rement of reducing power to produce succinate.
72 The enzymatic activity of this complex was very high in the succinate-p roducing strains even when they grew under anaerobic conditions (un published data). Thus the PDH pathway was included in Figure 1 and th e metabolic flux analysis. Figure 4-1. Fermentation Pathway of E. coli under anaerobic condition. vi, represents the fluxes used in the calculation for metabolic flux analysis. The carbon flow through citrate, isocitrate, and biosynthesis is assumed to be negligible to simplify the metabolic flux calculation. The boxes represent the measured metabolites. Solid arrows represent central fermentative pathways. Enzymes listed are pdh, pyruvate dehydrogenase; mdh, malate dehydrogenase; fumA and fumB, fumarase isozymes; frd, fumarate reductase; mgsA methylglyoxal synthase; poxB pyruvate oxidase; ppc phosphoenol pyruvate carboxylase; pdh, pyruvate dehydrogenase complex; ldhA, lactate dehydrogenase; pflB, pyruvate-formate lyase; adhE, alcohol dehydrogenase, ackA; acetate kinase; tdcD, propionate kinase with acetate kinase activity; tdcE, ketobutyrate formate-lyase ; acnAB aconitase; gltA, citrate synthase; gloAB glyoxalase.
73 Table 4-2. Reactions used in Metabolic Flux Analysis (MFA) for succi nate production under an aerobic condition Reaction Enzyme (s) EC number 1 Glucose + PEP Glucose-6P + pyruvate Glucose: PTS enzyme system; I, HPr, IIGlc, IIIGlc 2 Glucose-6P Biomass 3 Glucose-6P + ATP 2Glyceraldehyde 3-P + ADP a. Glucose-6P Fructose 6-P b. Fructose-6P+ATP Fructose 1,6-diP+ADP c. Fructose 1,6-diP 2Glyceraldehyde 3-P Glucose-6P isomerase Phosphofructokinase Fructose-diP aldolase 22.214.171.124 126.96.36.199 188.8.131.52 4 Glyceraldehyde-3P + NAD+ + Pi + ADP PEP + ATP + NADH + H++ H2O a. Glyceraldehyde-3P + NAD+ + Pi Glycerate-1,3-diP +NADH+H+ b. Glycerate-1,3-diP + ADP 3-Phosphoglycerate + ATP c. 3-Phosphoglycerate 2-Phosphoglycerate d. 2-Phosphoglycerate PEP + H2O 3P-Glyceraldehyde dehydrogenase 3PGlycerate kinase Phosphoglycerate mutase Enolase 184.108.40.206 220.127.116.11 18.104.22.168 22.214.171.124 5 PEP + ADP Pyruvate + ATP Pyruvate + ATP + H2O PEP + AMP + Pi Pyruvate kinase (PYK) Phosphoenolpyruvate synthase (PPS) 126.96.36.199 188.8.131.52 6 Pyruvate + NADH Lactate + NAD+ Lactate dehydrogenase 184.108.40.206 7 Pyruvate + HSCoA Formate + Acetyl~CoA Pyruvate formate-lyase 220.127.116.11 8 Acetyl~CoA +Pi + ADP Acetate + HSCoA + ATP a. Acetyl~CoA+Pi Acetyl~P + HSCoA b. Acetyl~P + ADP Acetate + ATP Acetate phosphotransferase Acetate kinase 18.104.22.168 22.214.171.124 9 Acetyl~CoA + 2NADH + 2H+ Ethanol + HSCoA + 2NAD+ a. Acetyl~CoA + NADH + H+ Acetaldehyde + HSCoA + NAD+ b. Acetaldehyde + NADH + H+ Ethanol + NAD+ Aldehyde dehydrogenase Alcohol dehydrogenase 126.96.36.199 188.8.131.52 10 PEP + CO2 OAA + Pi PEP carboxylase 184.108.40.206 11 OAA + NADH Malate + NAD+ Malate dehydrogenase 220.127.116.11 12 Malate + NADH + H+ Succinate + H2O + NAD+ Malate Fumarate + H2O Fumarate + NADH + H+ Succinate + NAD+ Fumarase ABC Fumarate reductase 18.104.22.168 1.3.5.
74 Table 4-2. (Continued ) Reaction Enzyme (s) EC number 13 Pyruvate + HSCoA + NAD+ Acetyl~CoA + CO2 + NADH + H+ Pyruvate Dehydrogenase (PDH) Complex: AceF: lipoate acetyltransferase AceE: E1p subunit of PDH LpdA: lipoamide dehydrogenase 22.214.171.124 126.96.36.199 188.8.131.52 EM Malate transport Dicarboxylate DAAC S transporter, DctA No EC # EP Pyruvate transport No EC #
75 Table 4-3. Stochiometric relationship between th e metabolic intermediates and metabolites and the network reactions (matrix K represented) for an anaer obic succinate production in glucose minimal medium in E. coli Flux To 1 2 3 4 5 6 7 8 9 10 11 12 13 EMEP Glucose used 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 succinate 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 Acetate 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 Formate 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 Lactate 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 Ethanol 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 Excreted Malate 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 Excreted Pyruvate 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 Pyruvate 1 0 0 0 1 -1 -1 0 0 0 0 0 -1 0 -1 Glucose-6P 1 -1 -1 0 0 0 0 0 0 0 0 0 0 0 0 Glyceraldehyde-3P 0 0 2 -1 0 0 0 0 0 0 0 0 0 0 0 Acetyl-CoA 0 0 0 0 0 0 1 -1 -1 0 0 0 1 0 0 PEP -1 0 0 1 -1 0 0 0 0 -1 0 0 0 0 0 OAA 0 0 0 0 0 0 0 0 0 1 -1 0 0 0 0 Malate 0 0 0 0 0 0 0 0 0 0 1 -1 0 -1 0 NADH 0 0 0 1 0 -1 0 0 -2 0 -1 -1 1 0 0 Results and Discussion Batch Characterization of Evolved E. coli Stra ins that Produce Succinate Effect of Glucose Concentratio ns to Succinate Production KJ012 ( ldhA, adhE, ackA ) was constructed to direct carbon flow to oxaloacetate (OAA) synthesis and to prev ent NADH oxidation from lactate and alcohol production. KJ012 strain growth and succinate production were im proved by sequentially sub-culturing for several generations to enrich for desira ble performance via metabolic evol ution (Jantama et al., 2008a). The strains are presumed to contain multiple spontaneous mutations among their genomes, which led to higher efficiency in carbon partitioning and energy to accommodate the gene modifications. KJ017, evolved from KJ012, exhibited in creased glycolytic flux and reduced fermentation time versus that of KJ012. These resu lts were consistent w ith 4 to 5-fold higher yields of succinate for both glucose concentr ations. In 5%(w/v) glucose fermentation, the average volumetric productivity of succinate and biomass yield of KJ017 were 7 and 10-fold
76 higher, respectively than those of KJ012 (Table 4-4). This in dicated that the spontaneous mutations, which occurred during metabolic evol ution, caused the ac tivation of succinate production routes and the increas e in biomass yield. It show ed that the anapleurotic (phosphoenolpyruvate carboxylatio n) pathway replenishes and conserves carbon sources that would otherwise be assimilated via pyruvate form ate-lyase activity and be further lost to CO2 and H2 via formate hydrogen-lyase (encoded by fdh and hyc ) activity. This pathway also probably plays an essential role in promoting gr owth in the microorganism when levels of highenergy intermediates such as OAA or acetyl~ CoA are high (Kornberg, 1966). After completing fermentation of 5% (w/v) glucose NBS, the amount of succinate produced was ~229 mM (Table 4-4). The molar succinate yield of KJ017 was 0. 92 mol/mol glucose used. KJ017 completed the fermentation of 5% (w/v) glucose within 96 h but the remaining glucose was still about 50 mM when performing the fermentation in 10% (w/v ) glucose even though the fermentation time was prolonged. The reducing equivalent s need to be recycled to NAD+ by the action of malate dehydrogenase ( mdh ) and fumarate reductase ( frd ), thus enabling glycolysis and enhancing succinate production (Velick and Furf ine, 1963). However, a higher NADH/NAD+ ratio resulting from high glycolytic flux (i.e 10%(w/v) gl ucose) may lower the enzyme activity of glyceraldehyde 3-phosphate dehydrogenase (G APDH). This is due to lack of NAD+ binding that affects a conformational change of the enzyme activity state (Hillman, 1979). The production yield of succinate from 10%(w/v) glucose NBS wa s slightly increased to 1.08 mol/mol glucose utilized. The yield was increased about 17%, co mpared to the yield produced from 5%(w/v) glucose fermentation. The increased yield can be explained by the high glycolytic flux since succinate titer and average productivity were increased about 115% and 111%, respectively.
77 Table 4-4: Fermentation profile for succinate production in 5% and 10% glucose NBS medium a Time required to complete fermentation. b Total succinate produced calculated after fermentation was completed. c Production yield calculated on a basis of mole succinate produced per mole glucose metabolized at the end of exponential growth. d Average volumetric productivity calculated on the basis of the maximum succinate level produced per total incubation time or fermentatio n finished depending on the culture condition. e Maximum biomass generated during fermentation calculated basing on bact erial cell molecular weight (CH2N0.25O0.5), 25.5 g/mol (Abbot and Claman, 1973) KJ017 produced high levels of formate and acet ate (Table 4-4) because the intracellular pyruvate pool was likely utilized by pyruvate formate-lyase activity resulting in carbon wasting and lower production yield of succinate. Highe r succinate production yields of KJ060 ( ldhA, adhE, ackA, pflB ) were obtained in both glucose concentr ations compared to KJ017. This is an effect of gene deletions re sulting in more direct carbon fl ux through an anapleurotic pathway to reoxidize NADH accumulated during glycolysis rather than through pyruvate dissimilation routes for maintaining redox balance. The os motic stress resulted from high glucose concentration did not appear to affect the succinate production in this stra in since the average productivity was still increased when hi gher glucose concentration was used. Suc. Production Strain Timea (h) Titerb (mM) Yieldc (mol/mol) Avg. Prod. (g/l-h)d Ace (mM) For (mM) Lac (mM) Eth (mM) Mal (mM) Pyr (mM) Bioe (mM) 5% GLUCOSE NBS WT 48 49 3 0.19 0.02 152 30 262 19 98 24 153 39 33 10 98 2 KJ012 144 6 1 0.20 0.01 26 1 11 1 KJ017 96 229 15 0.92 0.03 226 33 207 35 18 10 101 6 KJ060 96 332 3 1.17 0.08 106 6 19 2 15 3 93 6 KJ071 96 258 10 0.93 0.04 65 28 90 4 53 4 54 11 KJ073 72 339 3 1.24 0.13 0.12 0.01 0.04 0.00 0.28 0.02 0.41 0.00 0.31 0.01 0.54 0.03 115 5 47 3 6 2 91 2 10% GLUCOSE NBS WT 120 101 17 0.18 0.02 190 14 120 6 102 6 70 10 26 2 102 1 KJ012 144 13 2 0.28 0.08 45 1 14 1 12 3 KJ017 144 493 8 1.08 0.02 271 20 235 12 22 3 92 8 KJ060 144 660 5 1.21 0.03 117 7 137 4 9 1 70 5 KJ071 144 401 8 0.80 0.01 75 2 616 13 63 3 60 2 KJ073 96 664 5 1.12 0.03 0.10 0.01 0.01 0.00 0.59 0.00 0.65 0.02 0.33 0.01 0.80 0.01 121 12 122 4 98 2 81 2
78 The succinate production in E. coli by PEP carboxylation requires an external source of carbon via bicarbonate. CO2 gas generated during acid-base ne utralization by supplementation of a mixture of K2CO3 and KHCO3 contributed to higher carbon fixation by PPC activity. KJ060 exhibited a faster rate of su ccinate production (PEP carboxylation) in 10%(w/v) glucose medium since the average productivity was increased al most 59% compared to that from 5%(w/v) glucose medium. This implies that the carboxylati on of PEP at the PEP node is preferred. The carboxylation during succinate produ ction from KJ060 would be a benefit to an environment since the strain efficiently perf orms carbon dioxide fixation. CO2 gas generated from industrial processes can directly be used to produce succinate from ine xpensive carbon sources such as glucose, corn, and/or CSL, resulting in reduced levels of greenhouse gases. Production of acetate was still high even though KJ060 cannot generate acetyl~CoA due to lack of the pyruvate route (pyruvate formatelyase and acetate kinase). However, an overflow metabolism would be one of the reasons fo r high acetate level (Andersen and von Meyenburg 1980; Doelle and Hollywood, 1976; Meyer et al., 1984; Vemuri et al., 2006). This phenomenon is induced when the rate of glycolysis exceeds a critical value in which it is the consequence of an imbalance between glucose uptake and the demand for energy and biosynthesis resulting in by-product formation from pyruvate (Aristidou et al., 1995; Farmer and Liao, 1997). Moreover, during the absence of oxygen, oxidative phosphorylati on is inhibited leading to inefficiency in NADH oxidation. KJ060 was adapted to maintain the redox balance via accumulation of succinate. With high glucose concentrations, the rate of NADH ge nerated during glycolysis is greater than the rate of NADH oxidation (Vemuri et al., 2006). The carbon flow diversion to acetate could be a means to limit NADH in the cell since the flux th rough acetyl~CoA and acetate does not generate any reducing e quivalents (el-Mansi et al., 1989).
79 Methylglyoxal is a toxic interm ediate generated during glycolys is when glycolytic flux is increased (Grabar et al., 2006). The cell avoids the adverse effect of the toxic substance by converting the methylglyoxal to lactate, via gl yoxylase activity, leading the accumulation of lactate. The mgsA gene encoded for methylglyoxal syntha se was deleted from KJ060 to improve growth and to eliminate residual lactate. The resulting strain, KJ071 shows no lactate accumulation but exhibited lower succinate produ ction yield (0.80-090 mol/mol glucose used) than KJ060 (Table 4-4). However the growth rates of both strains were not significantly different (data not shown). Surprisingly, KJ071 accumulated pyruvate (~63 mM) as well as a very high level of malate (~616 mM) in 10%(w/v) glucose medium. The average succi nate productivity of KJ071 (0.33 g/l-h) was significantly lower than that of KJ060 (0.65 g/l-h). The reason for unexpected level of malate produc ed during fermentation of KJ071 is still unclear. However, we speculate that this st rain had increased flux through OAA vi a PPC activity, rather than via partitioning of carbon flow through pyr uvate catabolism, resulting in low level of acetate. It is generally known that in the presence of oxygen, the PDH comp lex oxidatively-decarboxylates pyruvate to acetyl~CoA with the conservation of reductant as NADH. Si nce KJ071 is unable to produce acetyl~CoA due to pflB mutation, the PDH activity ac tivated to some extent, to compensate the lack of pflB activity, even in the absence of oxygen. The malate accumulation was a consequence of the lower flux through PDH activity in which the rate of NADH generation was not fast e nough to efficiently convert fumarate to succinate. It is likely that the lack of NADH also influenced th e conversion rate of malate to fumarate, which then became the rate-limiting step in this pathway. Moreover, the feedback inhibition of anaerobic metabolism would account for malate accumulation. The increase in glycolytic flux may cause the decline in the succinate/malate ratio due to increased production of oxaloacetat e, an allosteric inhibitor of
80 fumarate reductase (Iverson et al., 2002; Sanwal, 1970). In addition, high levels of glucose cause the catabolite repression of mdh expression due to the catabolite control of cyclic AMP (cAMP) regulatory protein (Park et al., 1995). Also high level of ma late can alloster ically inhibit phosphoenolpyruvate carboxylase (Wang et al., 2006). To reduce the product feedback inhibition of both enzymes, the strains might excrete malate out of the cell to maintain the intracellular level that keeps the carboxylation pr ocess functioning. However, how the mgsA deletion affected the metabolic shift is still unclear. Decreasing acetate production during succina te fermentation would improve downstream processing by simplifying the purification. Pyruva te oxidase (POXB) is thought to be responsible for cell survival during stationary phase by oxidation of pyruvate resulting in acetate accumulation. Pyruvate oxidase is still functiona l at low growth rates (stationary phase) and under microaerobic conditions (Abdel-Hamid et al., 2001). An elevated intracellular pyruvate level accumulated from high glycolytic flux activ ates pyruvate oxidase, which directs pyruvate to acetate via a single step decarboxylation with a concomitant reduction of flavoprotein (Hager, 1957). The effect of poxB deletion showed that KJ073 ( ldhA, adhE, ackA, pflB, mgsA, poxB ) could completely utilize 10%(w/v) glucos e within 96 h and had the highest average succinate productivity in both 5%(w/v) and 10%(w/v) glucose concentrations (0.54 and 0.80 g/lh, respectively) (Table 4-4). The lower succinate yield observed from KJ073 compared to KJ060 might be due to the effect of the mutation that caused the strain to accumulate malate since KJ073 was derived from KJ071. However, KJ073 ad apted itself to produce less malate (about 80%) compared to KJ071 and produced ~664 mM succinate from 10%(w/v) glucose NBS. Unexpectedly, KJ073 still produced a cetate at the same level as that of KJ060. This indicates that deletion of poxB did not lower the acetat e level but promoted the growth and increased
81 glycolytic flux instead (Jantama et al., 2008a). Ve muri et al., (2005) revealed that the removal of POXB generally decreased the expression of PT S genes while increasing the expression of glucokinase, an alternative glucose uptake route in E. coli (Curtis and Epstein, 1975). Moreover, they also showed that the elimination of POXB activity led to increa sed expression of a key Entner-Doudoroff gene, edd (encoding phosphogluconate dehydrat ase), suggesting increased utilization of this pathway. POXB allows cells to avoid generating ATP in converting pyruvate to acetate. They propose that E. coli might respond to poxB deletion by metabolizing more glucose via the Entner-Doudoroff pathway that generates less ATP than the Embden MeyerhofParnas pathway (Vemuri et al., 2005). E. coli prefers limiting its ATP synthesis when glucose uptake rate is increased (Cha o and Liao, 1994; ; Causey et al., 2003; Koebmann et al., 2002; Patnaik et al., 1992). This result suggests the pr esence of other acetate producing pathways. The cells might compensate for the poxB deletion by increasing the rela tive expression of other genes involved in acetate formation. Fi rst, under anaerobic conditions, OA A is reduced to malate, and citrate can be converted into OAA and acetate via acetyl~CoA by citrate lyase (encoded by citDEF ) activity to recycle the in tracellular OAA pool for other me tabolic functions (Nilekani and Sirvaraman, 1983). Second, the study of the tdcD gene has revealed that it encodes a protein with acetate/propionate kinase activity (Hesslinger et al ., 1998). Based on the sequence similarity, the tdcD product is highly similar to the ackA encoded acetate kinase (Reed et al., 2003). Furthermore, the tdcE gene located downstream of tdcD in the same operon has a pyruvate/ -ketobutyrate formate-lyas e activity, which is a pfl like protein (Hesslinger et al., 1998). Both genes are involved in the metabo lism of L-threonine under anaerobic growth conditions. It is likely that the low formate a nd high acetate levels from fermentation of 5%(w/v) glucose in NBS implies that both gene products in KJ060 and its derivatives are functioning.
82 The increased levels of pyruvate in KJ071 and KJ073 also resulted in a major reduction in the succinate yield and productivity compared to KJ060. The high levels of pyruvate might result from high malic enzyme activity due to increased activation of phosphoenolpyruvate carboxylase and malate dehydrogenase. The activitie s of these enzymes were found to be higher when acetate or even glucose was present in the cultures (Siddiquee et al., 2004). Effect of Initial Inocula and Acetat e on Succinate Production in KJ060 Many attempts to produce succinate with E. coli derivatives at high cell densities have been reported to increase the glucose conversion rate to succinate (San chez et al., 2005a; b; Sanchez et al., 2006). KJ060 was used to study the e ffect of initial inocula and acetate levels on succinate production. The results showed that th e higher initial cell dens ities did not promote higher glucose consumption over lower cell densit y cultures (Table 4-5 and 4-6). However, 50100 mM glucose still remained in the medium after fermentation was prolonged upto 144 h (data not shown). The requirement of biomass generation in lowe r inocula culture might have increased the rate of glucose consumption as glucose was utiliz ed and the intermediate products were used in biosynthesis without accumulation of intracellular metabolites. In contrast, the glucose utilization in high inocula culture might lead in faster accu mulation of intracellular me tabolites than that of low inocula cultures. The metabolites generated dur ing glycolysis and succinate production were accumulated to such a critical level that they c ould inhibit some enzymes involved in glycolysis and anapleurotic pathways. This may result in fe edback inhibition of glyc olysis and anapleurotic pathways, thus decreasing the rate of succinate production.
83 Table 4-5. Comparison of KJ060 on metabolite production using 10%(w/v) glucose (~556 mM) as substrate in NBS salt medium with diffe rent initial acetate concentrations and initial cell density. Suc. Production Suc. Productivityd Initial OD550 Initial Acetate (mM) Timea (h) Titerb (mM) Yieldc (mol/mol) Avg. Prod. (g/l-h) Max. Prod. (g/l-h) Ace (mM) Mal (mM) Pyr (mM) Bioe (mM) 0.01 10 96 692 16 1.24 0.03 0.85 0.02 1.57 0.13 153 21 86 1 32 5 74 4 0.2 10 >144 633 30 1.33 0.06 0.71 0.03 0.90 0.04 135 12 85 5 34 2 66 4 0.01 5 >144 626 6 1.20 0.01 0.70 0.01 1.40 0.03 150 11 91 1 39 1 73 4 0.2 5 >144 596 12 1.35 0.03 0.58 0.03 1.20 0.08 151 22 75 2 20 2 66 2 0.01 0 >144 664 16 1.19 0.01 0.65 0.02 1.47 0.20 150 2 113 19 16 1 73 4 0.2 0 >144 648 2 1.24 0.03 0.61 0.01 1.30 0.08 146 10 81 1 28 2 71 4 0.01f 0 120 696 9 1.27 0.04 0.68 0.01 1.33 0.10 140 12 83 6 43 9 83 3 0.2f 0 120 690 30 1.32 0.06 0.68 0.03 1.23 0.05 171 5 39 2 10 1 77 3 a Time required to complete fermentation. b Total succinate produced calculated after fermentation was completed. c Production yield calculated on a basis of mole succinate produced per mole glucose metabolized at the end of exponential growth. d Maximum volumetric producti vity calculated on the basis of th e most productive 24-h period. Average volumetric productivity calculated on the basis of the maximum succinate level produced per total incubation time or fermentation finished depending on the culture condition. e Maximum biomass generated during fermentation cal culated basing on bacter ial cell molecular weight (CH2N0.25O0.5), 25.5 g/mol (Abbot and Claman, 1973) f The medium was supplemented with 0.1% each of peptone and yeast extract. From our studies, the succinate production yi eld was in the range of 1.1-1.6 mol/mol glucose used. The variation in the producti on yield might be dependent on the NADH availability and the functionality of glyoxylate bypass. The glyoxylate bypass might be operative in E. coli ATCC 8739 derivatives in which isocitrate lyase genes ( aceAB ) are constitutively transcribed while transcription of the isocitrate lyase repressor ( iclR ) is repressed even under anaerobic conditions (P hue and Shiloach, 2004). The glyoxylate bypass converts 2 moles of acetyl ~CoA and 1 mole of OAA to 1 mol of succinate and 1 mole of malate without the requirement of N ADH (Kornberg, 1966). However, 1 mole of malate from glyoxylate bypass requires an extra mole of NADH to be further converted to succinate. The only way to gain one extra mole of NADH is through the assimilation of
84 pyruvate via the pyruvate dehydrogenase (PDH) complex, which produces 1 mole of acetyl~CoA, 1 mole of CO2 and 1 mole of NADH from one mole of pyruvate. The NADH produced from PDH activity provided more reducing equivalents to reduce fumarate to succinate resulting in a greater molar succ inate yield than 1 mol/mol glucos e used. In figure 4-2, I propose a new succinate production pathwa y that leads the maximum yield of 1.71 mol/mol glucose used. The pathway includes the activat ion of PDH activity and of th e glyoxylate bypass. The succinate production yield of 1.1-1.2 mol/mol glucose used was observed in the cultures that started with low inocula. It might imply that only the P DH activity is activated but not glyoxylate bypass, resulting in lower succinate produ ction yield than we propose. In contrast, the higher succinate production yield close to that which we propose was obtained fr om the cultures inoculated with high cell densities. It also implies that P DH is activated and glyoxylate bypass might be operative under these conditions. Table 4-6. Comparison of KJ060 and KJ073 on metabolite production using 10%(w/v) glucose (~556 mM) as substrate in AM1 salt medium with initial cell density. Suc. Production Suc. Productivityd Strain + Initial OD550 Timea (h) Titerb (mM) Yieldc (mol/mol) Avg. Prod. (g/l-h) Max. Prod. (g/l-h) Ace (mM) Mal (mM) Pyr (mM) Bioe (mM) KJ060+0.01 120 733 40 1.41 0.07 0.90 0.04 2.21 0.24 250 37 39 17 87 2 KJ060+0.20 >144 636 17 1.48 0.09 0.78 0.02 1.49 0.04 211 8 18 11 79 1 KJ060+0.40 >144 668 41 1.49 0.04 0.82 0.05 1.45 0.40 192 6 27 19 1 0 86 5 KJ060+0.60 >144 622 31 1.61 0.12 0.77 0.04 1.07 0.13 180 13 17 5 6 1 78 3 KJ073+0.01 96 668 9 1.20 0.09 0.82 0.01 1.95 0.19 183 27 118 13 55 2 97 1 a Time required to complete fermentation. b Total succinate produced calculated after fermentation was completed. c Production yield calculated on a basis of mole succinate produced per mole glucose metabolized at the end of exponential growth. d Maximum volumetric producti vity calculated on the basis of the most productive 24-h period. Average volumetric productivity calculated on the basis of the maximum succinate level produced per total incubation time or fermentation finished depending on the culture condition. e Maximum biomass generated during fermentation cal culated basing on bacter ial cell molecular weight (CH2N0.25O0.5), 25.5 g/mol (Abbot and Claman, 1973)
85 Figure 4-2. Proposed succinate producti on pathway from glucose. The pathway shows maximum theoretical yield (1.71 mol/mol glucose us ed) of succinate produced during anaerobic fermentation. The blue arrow represen ts the flux through pyruvate dehydrogenase (PDH) complex that provides the extra source of NADH. The red arrow represents the carbon flow through glyoxylate bypass that yields malate and succinate from the condensation of glyoxylate a nd acetyl~CoA. +; gained, ; consumed. CIT; citrate, ACO; aconitate, IS CIT; isocitrate. The succinate production yield from high initial cell density cultures was also higher when acetate was supplied. Unlike the yield, th e maximum volumetric productivity from the low inocula (with or without initial acetate) cultures was higher than that in cultures with higher inocula (Table 4-5). In addition, the combinati on of low initial cell density growth with acetate or yeast extract/peptone supplementation pr omoted growth and increased the glucose consumption rate. The fermentation time was reduced to 96 h when acetate was added. The low glucose consumption observed in the la te log phase was possibl y due to a lack of growth supplements that are essential for gluc ose transport. To prove this, cultures were supplemented with yeast extract and peptone. The results showed that the supplemented cultures
86 at both high and low initial cell densities cons umed glucose more rapidly, completing the fermentation by 120 h compared to 144 h fo r non-supplemented cultures (Table 4-5). Effect of Low Salt Media (AM1) on Succinate Production in KJ060 The succinate production yield obtained from KJ060 in AM1 was about 1.41 mol/mol glucose utilized. The production yield was dram atically increased (upto 1.61 mol/mol glucose used) with a large inoculum; however, the ma ximum volumetric productivity was decreased (Table 4-6). In addition, glucose consumption wa s reduced when the initial cell density was high (data not shown). The highe st titer of succinate was achieved w ith the lowest inoculum (Table 46). This indicates that, in the low initial cell concen tration case, the produc tion of succinate is associated with biosynthesis. In other word s, part of the glucose was consumed. The intermediates generated from catabolism were us ed for biosynthesis during succinate production. The use of these intermediates for biosynthesi s increased glycolytic flux and accompanied the succinate production without metabolite accumulation, resulting in in a shorter time to complete fermentation. In contrast, due to the lower bios ynthetic demands in the cultures with higher cell densities, metabolic intermediates were accumulate d to a level that caused feedback inhibition of glucose catabolism. Surprisingly, the malate pr oduction of KJ060 in AM1 medium was lower than that in NBS, and the pyruvate concentration was very lo w. Unfortunately, the level of acetate in AM1 was 60% higher than that from NBS medium. We hypothesize that the high level of nitrogen salts in AM1 media may activate acetate production by unknown pathways. Effect of low Salt Media (AM1) on Succinate Production in KJ073 KJ073 was used to study the effect of AM1 medium on succinate production. The results showed that the average succina te productivity was approximately the same but the succinate production yield was increased a bout 7% higher than cultures grown in NBS based medium
87 (Table 4-4 and Table 4-6). However, when compared to KJ060 in AM1 medium, KJ073 exhibited lower succinate yields and productivities by 14% and 12%, respectively (Table 4-6). This suggests that the lower yield and rate of succinate produ ction resulted from a greater accumulation of malate and pyruvate in KJ073 than those in KJ060. The carbon losses to these intermediates affected the yield and rate. Metabolic Flux Analysis in E. coli Strains to Produce Succinate For each of the five strains studied, batch ferm entations were carried out in triplicate with 5%(w/v) and 10%(w/v) glucose. Table 4-7 presents the measured glucose consumption rates and specific production rates of the excreted meta bolites during the growth phase. For each metabolite, the values listed are in the order of the experiments. For example, the second run with wild type in 10%(w/v) glucose gave gl ucose consumption rate 20.77 mmol/gCDW-h, succinate production rate 2.59 mmol/gCDW-h, a cetate production rate 18.00 mmol/gCDW-h, etc. Using Equation 3, the fluxes were calculat ed for each run. The resulting means and standard deviations are presented in Table 48. Below we discuss how the metabolic fluxes change from the wild type to the mutants, and how they are affected by increased glucose concentration. Metabolic Flux Distributions in E. coli Wild Type The wild type has 5 branch points as can be seen in Figure 4-1. The first is at Glucose-6P, where v1 splits to v2 for biosynthesis and v3 for catabolism. The catabolic flux is much larger, and the split ratio is not significantly affect ed by increasing glucose concentration. The second branch point is at PEP, which is where v4 splits into v5 plus v1 to pyruvate and v10 to OAA. For both glucose concentrations, ov er 90% of PEP flux flows to pyruvate through v1+v5 (Table 4-8). That the majority of the flux is to pyruvate is expected since the apparent Km
88 (Michaelis-Menten constant) for PEP of PYK (pyruvate kinase) (0.2 mM, Kornberg and Malcovati, 1973) is considerably lower than that of PPC (phos phoenolpyruvate carboxylase) (0.8 mM, Smith et al., 1980). Table 4-7. Specific production ra te of extracellular metabolites Strains Wild Strain KJ017 KJ060 KJ071 KJ073 a Specific production rate (mmol/gCDW/h) 5%b 10%b 5%b 10%c 5%b 10%c 5%b 10%b 5%c 10%d Glucose used 22.40 21.51 21.00 16.86 20.77 19.79 12.75 11.51 11.46 14.13 14.39 14.66 15.76 16.93 16.83 19.65 15.93 15.84 17.11 17.60 17.06 9.94 10.96 9.82 15.28 16.14 16.06 25.96 21.23 23.76 Succinate 2.70 2.36 2.74 2.24 2.59 2.57 10.31 11.84 11.35 15.06 15.35 15.76 19.86 21.50 21.93 23.96 19.01 18.53 17.74 17.44 16.39 12.65 14.06 12.82 16.57 18.02 17.62 30.62 25.87 28.65 Acetate 19.20 18.44 17.00 13.92 18.00 17.80 10.14 10.75 10.64 8.36 8.42 8.76 7.58 7.72 6.30 6.69 5.71 5.11 6.83 6.51 6.20 5.91 7.00 6.19 4.78 5.33 4.48 6.72 6.74 6.77 Formate 35.95 35.06 34.73 25.45 33.29 32.22 9.59 9.75 8.96 8.49 8.43 8.80 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Lactate 1.31 0.92 1.17 0.39 0.50 0.54 0.32 0.69 0.59 0.80 0.93 1.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Ethanol 19.83 19.00 17.17 15.25 18.08 17.24 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Excreted Mal ate 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2.39 1.73 2.45 6.26 7.27 6.87 0.00 0.00 0.00 2.93 2.91 2.40 3.68 2.31 2.49 Excreted Pyruvate 0.00 0.00 0.00 0.07 0.08 1.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.83 0.85 0.98 0.00 0.00 0.00 0.00 0.00 0.00 2.46 2.96 2.86 2.56 2.04 2.10 a Specific production rates of extracellular metabo lites calculated from the data during growth phase b calculated from data at 0 and 24 h c calculated from data at 0 and 48 h d calculated from data at 24 and 48 h The next node is at pyruvate where v1 plus v5 splits into v6 to lactate, v7 to formate and acetyl~CoA, v13 to acetyl~CoA, and vEP to extracellular pyruvate. Under anaerobic conditions, PFLB has the highest affinity for pyruvate, with a lower Km for pyruvate than that of LDHA (PFLB = 2.0 mM; Knappe et al., 1974, LDHA = 7.2 mM; Tarmy and Kaplan, 1968a, b), thus explaining the higher flux through formate and acetyl~CoA ( v7) than through lactate (v6) (Table
89 4-8). Two-fold lower flux of LDHA activity wa s observed in 10%(w/v) glucose compared to 5%(w/v) glucose. In higher glycolytic flux, E. coli requires more ATP to produce osmoprotectants, thus favoring the ATP producing PFLB path. This could explain some downregulation of the activity of LDHA. Low or no flux through the LDHA pathway has been previously observed in glucose-grown cells under excess glucose conditions (de Graef et al., 1999). There is also flux through the PDH path way. This flux might serve to limit pyruvate accumulation and to balance the NAD/NADH ratio, since flux through PFLB is an overall redoxneutral process. It has been previously reported that the PDH complex activity is not dependent on the presence of oxygen but is mediated by th e internal redox state as reflected in the NAD/NADH ratio (de Graef et al., 1999). The flux through PDH exhibited in E. coli wild type might be affected by the NAD H requirement for reduction of acetyl~CoA and acetaldehyde produced via PFLB activity. In addition, NADH is consumed in the routes towards succinate production ( v11 and v12). Increase in glucose to 10%(w/ v) increased the flux through PDH in proportion to the increase in glycolytic flux. Finally, there was little or no flux to extracellular pyruvate. The fourth branch point is at Acetyl~CoA, where v13 plus v7 split into v8 to acetate and v9 to ethanol. The flux through ADHE ( v9), along with minor contributions from the fluxes to lactate (v6) and succinate ( v11 and v12), fulfill the requirement for NADH regeneration (NADHP/G NADHU/G), as required for the redox balance. Another observation for the flux distribution in wild type is the partition of acetyl~CoA into equimolar amounts of ethanol and acetate ( v8 v9). Stephanopoulos and Vallino (1991) and Aristidou et al. (1999) al so mention this observation. The last potential branch point is at malate, wher e the split can lead to succinate or extracellular malate. No appreciable amount of excreted malate was observed.
90 Table 4-8. Metabolic fluxes distribution of an anaerobic succinate production in 5%(w /v) and 10%(w/v) gluc ose NBS of various E. coli strains Strains ECC KJ017 KJ060 KJ071 KJ073 Flux Toa (mmol/gCDW/h) 5% 10% 5% 10% 5% 10% 5% 10% 5% 10% Glucose, v1 21.60.7 19.12.0 11.90.7 14.40.3 16.50.6 17.12.2 17.30.3 10.20.6 15.80.5 23.72.4 Biomass, v2 1.00.2 1.00.2 0.91.4 0.50.1 0.80.2 0.40.2 1.00.3 0.30.1 0.70.2 0.40.2 Gly-3P, v3 20.60.9 18.12.0 11.00.7 13.90.4 15.70.8 16.72.3 16.20.5 10.00.5 15.10.6 23.32.2 PEP, v4 41.21.8 36.33.9 22.01.4 27.80.7 31.41.6 33.44.6 32.41.0 19.91.0 30.21.2 46.54.4 Pyruvate, v5 16.91.2 14.71.7 -1.01.2 -2.50.1 -6.70.3 -7.00.8 -9.10.4 -3.70.2 -6.10.1 -9.21.2 Lactate, v6 1.10.2 0.50.1 0.50.2 0.90.1 0.00.0 0.00.0 0.00.0 0.00.0 0.00.0 0.00.0 Formate, v7 35.10.7 30.34.3 9.50.4 8.10.2 0.00.0 0.00.0 0.00.0 0.00.0 0.00.0 0.00.0 Acetyl-P, v8 18.40.9 16.52.3 10.50.3 9.50.2 8.20.5 7.01.0 7.20.3 6.40.5 5.60.3 8.60.3 Ethanol, v9 18.71.4 16.91.4 0.00.0 0.00.0 0.00.0 0.00.0 0.00.0 0.00.0 0.00.0 0.00.0 OAA, v10 2.60.2 2.50.2 11.20.8 15.40.4 21.11.1 22.73.2 24.00.7 13.30.7 20.20.7 31.23.1 Malate, v11 2.50.2 2.50.2 11.20.8 14.90.3 20.60.9 22.13.1 23.70.7 13.20.7 19.80.7 30.32.9 Succinate, v12 2.50.2 2.50.2 11.20.8 14.90.3 20.60.9 19.92.9 16.90.7 13.20.7 17.10.7 27.52.2 PDH, v13 2.11.3 3.00.3 0.90.5 2.40.1 9.30.3 8.11.2 7.80.3 6.50.4 6.30.0 10.30.6 EM, vEM 0.00.0 0.00.0 0.00.0 0.00.0 0.00.0 2.20.4 6.80.5 0.00.0 2.70.3 2.80.7 EP, vEP 0.00.0 0.10.0 0.00.0 0.00.0 0.00.0 1.50.1 0.00.0 0.00.0 3.10.3 3.10.4 ATP/G 2.60.0 2.60.1 1.60.4 1.30.0 0.60.0 0.60.0 0.30.0 0.90.0 0.50.1 0.60.1 YATP 4.90.3 5.70.7 9.61.9 7.30.3 25.91.2 13.92.4 50.714.427.02.0 16.22.0 15.10.6 NADHP/G 2.00.1 2.10.0 1.90.3 2.10.0 2.50.0 2.40.0 2.30.1 2.60.0 2.30.0 2.40.0 NADHU/G 2.00.1 2.10.0 1.90.3 2.10.0 2.50.0 2.40.0 2.30.1 2.60.0 2.30.0 2.40.0 a Specific volumetric flux during growth phase
91Metabolic Flux Distributio ns in Mutant Strains For the mutant strains (KJ017, KJ060, KJ071, and KJ073) the glucose-6-P branch point again shows the majority of the car bon flux going to glyceraldehyde-3-P ( v3). For all mutant strains, increasing glucose concentration decreased the flux v2 to biomass, indicating that the mutants are less capable to dealing with high osmolality than the wild type. The split of carbon flux at the PEP node is critical since it determines the flux to the succinate production route (Figure 4-1). For strain KJ017 lacking LDHA, ACKA and ADHE, the flux towards succinate productio n is dramatically increased (T able 4-8) since this pathway solely functions to recycle NAD+ for glycolysis, resulting in th e net specific carbon flux from PEP to OAA ( v10) more than four times higher than that of the wild type. It is also noteworthy, that the need to regenerate NAD+ caused flux v5 to reverse sign, indicating that some of the pyruvate generated during glucose transport is converted back to PEP through PPS. In the strains without PFLB (KJ060, KJ071, and KJ073), the specific flux v10 from PEP to OAA increased even further to as high as 67% of v4. This indicates that th e absence of PFLB caused the strains to shift the flux from pyruvate breakdown to PEP utilization. Th e relative splits at this node were not appreciably affect ed by glucose concentration. The splits at the pyruvate node were signifi cantly affected by the ge ne deletions in the mutants. In KJ017, as in the wild type, the ma jority of the flux to pyruvate goes to formate ( v7) with some flux through PDH ( v13) and, surprisingly, a small amount of flux to lactate ( v6). It is hypothesized that an alternativ e parthway for producing lactat e may have become active, a candidate being the methylglyoxal synthase. No extracellular pyruvate was observed in KJ017. In strains KJ060, KJ071 and KJ073, no lactat e was observed. Strain KJ071 did not excrete pyruvate, but strain KJ060 did produce so me extracellular pyruvate at 10%(w/v) glucose and strain KJ073 produced some at both glucose leve ls. For all strains th e major flux out of the
92 pyruvate node was the flux through PDH (v13) leading to the production of acetate. This flux dramatically increased, up to 8 fold, in the pflBstrains with respect to KJ017. The increase in the PDH flux is in agreement with the rate of succinate production (Table 4-8). This suggests that NADH produced from the PDH activity can incr ease the flux to succinate and balance the NADH production and consumption. From Table 4-8, the redox balance is accomplished by having v4 + v13 v11 + v12 in the strains lacking the ability to produce ethanol and lactate. Strain KJ032, the parent strain of KJ060 and the first strain with a pflB deletion, required acetate as a source of acetyl~C oA to grow anaerobically. Afte r the metabolic evolution KJ060 grew under anaerobic condition without acetate. The only pathway permitting the production of acetyl~CoA as source of C2 for biosynthesis from pyruvate w ould be the PDH complex. Besides the effect of internal redox state, PDH complex activity observed during anaerobic fermentation in the strains lacking PFLB could be explaine d by the regulation of FNR (fumarate-nitrate regulatory protein) and intracellular pyruvate. Quail and Gu est (1995) explaine d that FNR and PdhR repressor proteins control pdh transcription. In the absence of pyruvate, PdhR binds to the promoter of the pdh operon and inhibits transcription. Increased intracellular pyruvate levels because of the cells inability to dissimilate pyruvate might result in binding of pyruvate with PdhR thus releasing PdhR from the promoter region of the pdh operon. Control by FNR alone might not be enough to repress pdh transcription, reflecting a gr eatly increased flux through the PDH complex in the strain lacking PFLB activity. The benefit of activation of the PDH complex to succinate production is that the flux via the PDH complex pr ovides an extra mole of NADH that can be used to reduce malate to succinate. The last node for the mutants is that of v11 to malate. KJ017 produced no extracellular malate, converting all intracellular malate to succinate. KJ060 produced small amount of
93 extracellular malate at high glucose levels. The decrease in NADH production via the PDH pathway, may have limited its availability fo r succinate production. Strain KJ071 produced significant amount of extracellular malate at 5%(w/v) glucose during exp onential growth. At 10%(w/v) glucose, no malate was observed durin g exponential growth, but a very high amount of malate (titer of 520 mM, Jant ama et al., 2008a) was produced in the ensuing stationary phase. Strain KJ073 produced the highest amount of succinate, particularly at 10%(w/v) glucose, splitting the flux at the node to 85% succinate at 5%(w/v) glucos e and to almost 90% succinate at 10%(w/v) glucose. The redox balance proba bly controls this split, with higher NADH production through PDH favoring succinate. It is interesting that mutant strains KJ060, KJ071, and KJ073 have high flux to acetate ( v8), even though ackA was deleted from their genomes. These strains might activate other enzymes to compensate for ACKA activity. One of the enzymes involved in the degradation of threonine is a possibility since it exhibits propionate kinase ( tdcD ) activity (Figure 4-1) and is an acetate kinase homologue (H esslinger et al., 1998). Effect of Gene Deletions on ATP and Cell Yields Neglecting carbon loss through OAA, the wild ty pe strain converts acetyl~CoA into approximately equimolar amounts of ethanol and acetate that yields an additional mole of ATP per mole of glucose used. Adding this to the tw o net moles of ATP gained from glycolysis (50% of PEP is from the PTS system that does not produce ATP) yields a maximum net ATP/G of 3 (Tempest and Neijssel, 1987). However, during fe rmentation wild type strain also produces succinate. The carbon flux to the succinate produ ction pathway does not produce ATP, resulting in production of ATP/G less than 3 (2.6) for both glucose co ncentrations. The amount of ATP produced depends on the split flux ratio from PEP to pyruvate ( v5) via PYK and to OAA ( v10) via
94 PPC. In wild type strain most of the carbon flux fr om PEP flows to pyruvate, v5. This strain produces the highest amo unt of ATP/G among the five strains studied. For the strains lacking ACKA, the net maximum ATP/G is 2 because they can produce ATP only from glycolysis under anaerobic condi tions (Tempest and Neijssel, 1987). Strain KJ017 produces 1.6 ATP/G at 5%(w/v) glucose a nd 1.3 ATP/G at 10%(w/v) glucose (Table 48), somewhat lower than the theoretical maximum. This can be explained by the increase in carbon flux to the succinate production route ( v10) and the loss of energy to produce PEP from pyruvate via PPS ( v5 becomes negative). In the stra ins without PFLB and ACKA (KJ060, KJ071, and KJ073), ATP production is considerably lowe r, less than 1 ATP/G. This is attributed to substantial increase in the specific flux v10 from PEP to OAA and a substantial increase in the flux through PPS ( v5<0), which consumes 2 moles of ATP. As seen, the production of ATP in all mutant strains was low compared to the wild type. How these strains resolve the energy shortage to maintain a certain intracellular ATP level used for promoting cell growth and metabolic activity is interesting. The strains probably gain the benefit of generating ATP using other unknown pathways. This should be investigated further. It is known that the growth yield of an or ganism is proportional to the amount of ATP produced by its catabolic processes. Cell yield (YATP) is defined as the gram of dry weight of microbial cells produced per mo le of ATP formed (Bauchop a nd Elsden, 1960). The metabolic flux analysis of the data allowed us to calculate the anaerobic cell yield (YATP) neglecting maintenance and the results are shown in Table 4-5. A value for YATP MAX of 10.3 g CDW per mole ATP (Hempfling and Mainzer, 1975) has been generally accepted as the maximum YATP for the anaerobic growth of E. coli in minimal glucose media. The values estimated for the wild type and KJ017 are somewhat below YATP MAX, as would be expected since maintenance energy
95 is not considered. However, the values for the strains without PFLB and ACKA (KJ060, KJ071, and KJ073) are considerably higher than YATP MAX, particularly for KJ071 at 5% w/v glucose. A probable explanation is that mo re ATP was generated than our flux analysis indicates. Conclusions The metabolism of E. coli is remarkably adaptive. Some genetic manipulations are obviously needed to engineer metabolism for the production of chemicals such as succinate with high titer, productivity, and yield close to the maximum theoretical from glucose. The combination of gene deletions and metabolic ev olution can result in significant changes in metabolite fluxes. Some interesting changes id entified in this study are increased flux through PDH for the pflB mutants and reversal of the PEP to pyruvate reaction, from the PYK-catalyzed forward reaction for the wild type to th e PPS-catalyzed reverse reaction for all ackA mutants. Analysis of ATP generation and cell yield suggests that the pflB mutants may contain an alternative route for ATP production. This should be investigated further. Metabolic flux analysis of mu tant strains that we deve loped for succinate production provided us with insights and information on how the intracellular metabolite fluxes change at branch points as a result of the genetic manipulations. This information may guide us in future genetic modifications with the goal of direct ing more carbon flux to succinate production and reducing the flux to byproducts su ch as acetate and malate.
96 CHAPTER 5 ELIMINATING SIDE PRODUCTS AND INCREASING SUCCINATE YIELDS IN ENGI NEERED STRAINS OF Escherichia coli ATCC 8739 Introduction Many approaches have been investigated for the microbial production of succinic acid. E. coli has been engineered with heterologous genes to increase succina te production using both aerobic, anaerobic, and two-step combination processes (Chatterj ee et al., 2001; Jantama et al., 2008a; Sanchez et al., 2005a, b; Vemuri et al., 20 02a, b; Wu et al., 2007). Rumen bacteria have been identified and investigated that have the natural ability to produce large amounts of succinic acid anaerobically (Jantama et al., 2008a; Zeikus et al., 1999). However, most rumen bacteria and recombinant E. coli require complex nutrients (Janta ma et al., 2008a; Song and Lee, 2006; Song et al., 2007) and accumulate side products dur ing fermentation. Reducing both the levels of side products and the use of complex nutrients offer opportunities to im prove product yields and reduce costs associated with downstream pro cessing, purification, and waste disposal. In Chapter 3, I described that derivatives of E. coli ATCC 8739 were recently constructed that produce succinic acid in minimal salts medi a using simple, batch fermentations without the use of foreign genes. The best biocatalyst, KJ073 ( ldhA::FRT adhE::FRT ( focA pflB ) ::FRT ackA ::FRT mgsA poxB ), produced 1.2 mole of succinic acid per mole of metabolized glucose. However, these strains also accumulate d large amounts of acetate, malate, and pyruvate as side products. By identifying the source of th ese side products, I developed new strains that approach the maximum theoretical yield (1.71 mo l succinate per mol glucose) in simple batch fermenations.
97Materials and Methods Strains, Media and Growth Conditions New derivatives of E. coli C (ATCC 8739) were devel oped for succinate production using a unique combination of gene deletions co upled with growth-based selection. Strains, plasmids, and primers used in this study are summarized in Table 5-1. During strain construction, culture s were grown at 37oC in modified LB (per lite r: 10 g Difco tryptone, 5 g Difco yeast extract, 5 g sodium chloride) (Mi ller, 1992) and supplemente d with antibiotics as appropriate (Jantama et al., 2008a ; Zhang et al., 2007). No gene encoding antibiotic resistance, plasmid, or foreign gene is pres ent in the final strains developed for succinate production. After construction, strains were grown and maintained in AM1 medium (Martinez et al., 2007). This medium was supplemented with 100 mM KHCO3 and glucose (as indicated). Betaine (1 mM) was also added when the initial glucose concentration was 5%(w/v) or higher. Deletion of FRT Markers in the adhE, ldhA and focA-pflB Regions The strategy used to make sequential gene deletions and remove the FRT markers from the adhE, ldhA and focA-pflB loci has been described previo usly (Datsenko and Wanner, 2000; Grabar et al., 2006; Jantama et al., 2008a; Zha ng et al., 2007). Plasmid pLOI4151 was used as a source of a cat-sacB cassette and Red recombinase (pKD 46) was used to facilitate doublecrossover, homologous recombination events. Chlora mphenicol resistance was used to select for integration. Growth with sucrose was used to select for loss of sacB With this approach, successive deletions were constructed to produce derivatives of KJ079 that eliminated all FRT sites. Primers and plasmids are listed in Table 5-1.
98Table 5-1. Escherichia coli strains, plasmids, and primers Relevant Characteristics Sources E. coli strains KJ073 ldhA::FRT adhE::FRT ( focA pflB ) ::FRT ackA ::FRT mgsA poxB Jantama et al., 2008a KJ076 KJ073, ackA::cat-sacB translational stop sequence This study KJ079 KJ076, ackA::translational stop sequence This study TG200 KJ079, adhE::cat-sacB This study TG201 TG200, adhE This study TG202 TG201, ldhA::cat-sacB This study TG203 TG202, ldhA This study TG204 TG203, ( focA pflB ) ::cat-sacB This study KJ091 TG204, ( focA pflB ) This study KJ098 KJ091, tdcDE This study KJ104 KJ098, citF:: translational stop sequence This study KJ110 KJ104, aspC This study KJ119 KJ104, sfcA::translational stop sequence This study KJ122 KJ110, sfcA::translational stop sequence This study KJ134 KJ122, pta-ackA ::translational stop sequence This study Plasmids pKD46 bla exo (red recombinase), temper ature-conditional replicon Datsenko, 2000 pEL04 cat-sacB cassette Lee, 2001 Thomason, 2005 pCR2.1TOPO bla kan; TOPO TA cloning vector Invitrogen pLOI2228 cat ; FRT -cat-FRT cassette MartinezMorales et al., 1999
99Table 5-1. ( Continued ) Relevant Characteristics Sources pLOI2511 bla kan ; FRT -kanFRT cassette Underwood et al., 2002 pLOI4162 bla cat; ligation of cat-sacB cassette ( Pac I digested) from pLOI4146 and Pac I digested pLOI4161 Jantama et al., 2008b pLOI4158 bla kan; ackA (PCR) from E.coli C (using JM ackA F1/R1 primers) cloned into pCR2.1-TOPO vector Jantama et al., 2008b pLOI4159 Sma I /Sfo I digested cat-sacB cassette from pLOI4162 cloned into the PCR amplified inside-out product from pLOI4158 (using JM ackA up1/down1) Jantama et al., 2008b pLOI4160 Pac I digestion of pLOI4159, then self-ligated Jantama et al., 2008b pLOI4280 bla kan; aspC (PCR) from E.coli C (using aspC up/down primers) cloned into pCR2.1-TOPO vector Jantama et al., 2008b pLOI4281 Sma I /Sfo I digested cat-sacB cassette from pLOI4162 cloned into the PCR amplified inside-out product from pLOI4280 (using aspC 1/2 primers) Jantama et al., 2008b pLOI4282 PCR fragment amplified insi de-out product from pLOI4280 (using aspC 1/2 primers), kinase treated, then self-ligated Jantama et al., 2008b pLOI4283 bla kan; sfcA (PCR) from E.coli C (using sfcA up/down primers) cloned into p CR2.1-TOPO vector Jantama et al., 2008b pLOI4284 Sma I /Sfo I digested cat-sacB cassette from pLOI4162 cloned into the PCR amplified inside-out product from pLOI4283 (using sfcA 1/2 primers) Jantama et al., 2008b pLOI4285 Pac I digestion of pLOI4284, then self-ligated Jantama et al., 2008b pLOI4413 bla kan; ychE-adhE-ychG (PCR) fr om E.coli C (using up/down adhE primers) cloned into pCR2.1-TOPO vector Jantama et al., 2008b pLOI4419 PCR fragment amplified in side-out product from pLOI4413 (IOadhE up/down using primers), kinase treated, then self-ligated Jantama et al., 2008b pLOI4415 bla kan; ycaO-focA-pflB-pflA (PCR) from E.coli C (using up focA /Mid pflA primers) cloned into pCR2.1-TOPO vector Jantama et al., 2008b pLOI4421 PCR fragment amplified inside -out product from pLOI4415 (using IOycaO up/IO-mid pflB down primers), kinase treated then self-ligated Jantama et al., 2008b
100Table 5-1. ( Continued ) Relevant Characteristics Sources pLOI4430 bla kan; hslJ-ldhA-ydbH (PCR) from E.coli C (using ldhA A/C primers) cloned into pCR2.1-TOPO vector Jantama et al., 2008b pLOI4432 PCR fragment amplified inside -out product from pLOI4424 (using IOldhA up/down primers), kinase treated, then self-ligated Jantama et al., 2008b pLOI4515 bla kan; tdcG-tdcFED-tdcC (PCR) from E.coli C (using tdcDE up/down primers) cloned into pCR2.1-TOPO vector This study pLOI4516 Sma I /Sfo I digested cat-sacB cassette from pLOI4162 cloned into the PCR amplified inside-out product from pLOI4515 (using tdcDE F7/R7 primers) This study pLOI4517 PCR fragment amplified insi de-out product from pLOI415 (using tdcDE F7/R7 primers), kinase treated, then self-ligated This study pLOI4629 bla kan; citF (PCR) from E.coli C (using citF up2/down2 primers) cloned into pC R2.1-TOPO vector Jantama et al., 2008b pLOI4630 Sma I /Sfo I digested cat-sacB cassette from pLOI4162 cloned into the PCR amplified inside-out product from pLOI4629 (using citF 2/3 primers) Jantama et al., 2008b pLOI4631 Pac I digestion of pLOI4630, then self-ligated Jantama et al., 2008b pLOI4710 bla kan; pta-ackA (PCR) from E.coli C (using ackA up/ pta down primers) cloned into pCR2.1-TOPO vector Jantama et al., 2008b pLOI4711 Sma I /Sfo I digested cat-sacB cassette from pLOI4162 cloned into the PCR amplified inside-out product from pLOI4710 (using ackA2/pta2 primers) Jantama et al., 2008b pLOI4712 Pac I digestion of pLOI4711, then self-ligated Jantama et al., 2008b Primer sets JM ackA-F1/R1 5GCCTGAAGGCCTAAGTAGTA3 5GCACGATAGTCGTAGTCTGA3 Jantama et al., 2008b JMackAup1/ down1 5GTTGAGCGCTTCGCTGTGAG3 5GCCGCAATGGTTCGTGAACT3 Jantama et al., 2008b JMcatsacBup3/ down3 5CTCACCTCGAGTGTGACGGAAGATCACTTCG3 5GTGCAGGATCCATCAAAGGGAAAACTGTCCATAT3 Jantama et al., 2008b WM adhE A/C 5ATGGCTGTTACTAATGTCGCTGAACTTAAC GCACTCGTAGAGCGTCGGCACGTAAGAGGTTCCAA 3 5TTAAGCGGATTTTTTCGCTTTTTTCTCAGCTTTAG CCGGAGCAGCACACTGCTTCCGGTAGTCAA 3 This study
101Table 5-1. ( Continued ) Relevant Characteristics Sources WM ldhA A/C 5ATGAAACTCGCCGTTTATAGCACAAAACAGTACGACAAGAAGTACGGCACGTAAGAGGTTCCAA 3 5TTAAACCAGTTCGTTCGGGCAGGTTTCGCC TTTTTCCAGATTGCTACACTGCTTCCGGTAGTCAA 3 This study WM pflB A/C 5TTACTCCGTATTTGCATAAAAACCATGCGAGTTACGGGCCTATAACGGCACGTAAGAGGTTCCAA 3 5TTACATAGATTGAGTGAAGGTACGAGTAATAACGTCCTGCTGCTGTTCTACACTGCTTCCGGTAGTCAA 3 This study tdcDE up/down 5CGCCGACAGAGTAATAGGTT3 5TGATGAGCTACCTGGTATGG3 This study tdcDE F7/R7 5CGATGCGGTGGCCAATTAAG3 5GACGACGTGCTGGATTACGA3 This study citF up2/down2 5GGGTATTCAGGCGTTCGATA3 5GCCCGAGAGGATGACTATGT3 Jantama et al., 2008b citF 2/3 5GGTGATCGATGTTGTGCATC3 5CCCGTTCTTGTCGTTGAGAT3 Jantama et al., 2008b Up/downadhE 5CCGCTGTCTGATAACTGGTC3 5GCATAAGCGGATGGTCACTG3 This study IOadhE up/down 5GCTGCTCCGGCTAAAGCTGA3 5ACGCTCTACGAGTGCGTTAA3 This study upfocA /Mid pflA 5AGATCGCCAGCCGCTGCAAT3 5AACCGTTGGTGTCCAGACAG3 This study IO-ycaO up/IOmid pflB down 5GCCTACATTGCGTAGGCTAT3 5GCAGCAGGACGTTATTACTC3 This study ldhA A/C 5ATGAAACTCGCCGTTTATAG3 5TTAAACCAGTTCGTTGCCC3 This study IO-ldhA up/down 5CGTTCGATCCGTATCCAAGT3 5AGGCTGGAACTCGGACTACT3 This study aspC up/down 5 TCCATCGCTTACACCAAATC3 5TGGGGGATGACGTGATATTT3 Jantama et al., 2008b aspC 1/2 5 AGATAACATGGCTCCGCTGT3 5AGGAGCGGCGGTAATGTTC3 Jantama et al., 2008b sfcAup/down 5CTATGCTTGATCGGCAACCT3 5ACGATCGCCTGGTTTTAATG3 Jantama et al., 2008b sfcA1/2 5TACCGCCGTACCTCCATCTA3 5CGTAAGGGATATAAAGCGAACG3 Jantama et al., 2008b ackA up/ pta down 5CGGGACAACGTTCAAAACAT3 5ATTGCCCATCTTCTTGTTGG3 Jantama et al., 2008b
102Table 5-1. ( Continued ) Relevant Characteristics Sources ackA 2/ pta 2 5AACTACCGCAGTTCAGAACCA3 5TCTGAACACCGGTAACACCA3 Jantama et al., 2008b
103 To remove the FRT site in the adhE region, hybrid primers (WM adhE A/C) for adhE::FRT target region were designed to contain approximately 50 bp of homology to the 5 and 3 regions of adhE::FRT site and 20 bp corresponding to cat-sacB gene from pLOI4151. These primers were used for PCR amplification of the cat-sacB cassette using pLOI4151 as a template. The resulting PCR product was used to replace the FRT site in adhE region with a cat-sacB cassette by a double-crossover, homologous recombination event with selection for resistance to chlorampheni col, to produce TG200. The adhE gene and surrounding sequen ce were amplified from E. coli ATCC 8739 using up/down adhE primers. The PCR product containing ych Eadh E- ych G (3.44 kb) was cloned into pCR2.1-TOPO, yielding pLOI4413. A second set of primers (IOadhE up/down) was used to amplify the inside-out product with pLOI4413 as a template and Pfu polymerase to yield a bluntended product in which a 2.6 kbp internal segment of adhE sequence was deleted. This insideout PCR product was kinase-treated and self-l igated, resulting in pLOI4419. The PCR product amplified from pLOI4419 (up/down adhE primers) was used to replace the cat-sacB cassette in TG200 with the desired chromosomal sequence by another double, homologous recombination event, with sucrose selection for loss of sacB The resulting strain was designated TG201 (KJ079 with the FRT removed from adhE region). The FRT sites in the ldhA and ( focA-pflB ) regions were removed in a manner analogous to that used to delete the adhE ::FRT site. Additional primer sets ( ldhA A/C and IOldhA up/down) used to remove the FRT site in ldhA are included in Table 5-1 together with the corresponding plasmids (pLOI4430 and pLOI4432) Strain TG202 was produced by replacing this region in TG201 with the PCR product from pLOI4151 (WMldhA A/C primers). The cat-
104 sacB cassette in TG202 was replaced with the PCR product from pLOI4432 ( ldhA A/C primers) with sucrose selection for loss of sacB to produce TG203. Primer sets (upfocA /Mid pflA and IOycaO up/IO-midpflA down) and corresponding plasmids (pLOI4415 and pLOI4421) used to remove the FRT site in ( focA-pflB ) are included in Table 5-1. Strain TG204 was produced by replac ing this region in TG203 with the PCR product from pLOI4151 (WM pflB A/C primers). The cat-sacB cassette in TG204 was replaced with the PCR product from pLOI4421 (up focA/ Mid pflA primers) with sucrose selection for loss of sacB to produce KJ091. KJ091 is a derivative of KJ073 in which all FRT sites have been removed from the adhE ldhA and ( focA-pflB ) regions of the chromosome. The transformation efficiency of the cat-sacB fragment into E. coli was approximately 100 colonies per g DNA. More than 90% of recombinants selected for chloramphenicol resistance were sensitive to sucrose. After the second recombination and selection with sucrose, greater than 90% of colonies were of the de sired genotype (chloramphenmicol sensitive and insensitive to sucrose). Construction of Gene Deletions in tdcDE and aspC The tdcDE gene and neighboring 1000 bp regions ( tdcG-tdcFED-tdcC 5325 bp) were amplified using tdcDE up/down primers and cloned into the pCR2.1-TOPO vector to produce plasmid pLOI4515. To facilitate the sequential deletion of chromosomal DNA, plasmid pLOI4162 was constructed with a removable cat-sacB cassette and the option to include an 18-bp segment (5GCCTAATTAATTAATCCC3) of synthetic DNA with stop codons in all reading frames. This plasmid is composed of synthetic seque nces and parts of plasmids pLOI2228 (MartinezMorales et al., 1999), pLOI2511 (Underwood et al., 2002), and pEL04 (Lee et al., 2001;
105 Thomason et al., 2005). The construction detail s of pLOI4162 were pr eviously described (Jantama et al., 2008b). The sequence of pLOI 4162 is accessible from Genbank with an accession number of EU531506. A 1000-fold diluted preparation of pLOI4515 D NA served as a template for inside-out amplification using the tdcDE F7/R7 primers (both primers within the tdcDE gene and facing outward). The resulting 6861 bp fragment containi ng the replicon was ligated to the amplified, Sma I /Sfo I-digested cat-sacB cassette from pLOI4162 (JM catsacB up3/down3 primers) to produce pLOI4516. This 6861 bp fragment was also used to construct a second plasmid, pLOI4517 (kinase treated, self -ligation) containing a deletion of tcdD and tdcE The PCR fragments amplified from pLOI4516 and pLOI4517 ( tdcDE up/down primers) were used to replace tdcDE region in KJ091. The resulting clones were tested for loss of ampicillin and chloramphenicol resistance and designated KJ098. The aspC gene was deleted from KJ104 in a manner analogous to that used to delete the tdcDE gene. Additional primer sets ( aspC up/down and aspC 1/2) used to construct the aspC deletion are included in Table 5-1 together with the corresponding plasmids (pLOI4280, pLOI4281, and pLOI4282). The resulting strain was designated KJ110. Neither KJ098, nor KJ110 contain any intervening sequence within the respective deleted regions ( tdcDE and aspC ). Removal of FRT Site in ackA Region and Construction of citF sfcA and pta-ackA Gene Deletions To eliminate the FRT site in the ackA region of KJ073, plasmids containing sequences of the desired mutation were constructed prev iously (Jantama et al., 2008b). Briefly, E. coli ATCC 8739 genomic DNA was used as the template for PCR amplification of ackA with the JM ackA F1/R1 primers that bind approximately 200 bp upstream and downstream of the ackA gene. The linear product was cloned into pCR2 .1-TOPO (Invitrogen, Carlsbad, CA) to produce
106 pLOI4158. Plasmid pLOI4158 was then used as a template for inside-out PCR with JM ackA up1/down1 primers and Pfu polymerase to yield a blunt-e nded product that lacks an 808bp internal segment of ackA The Pac I-flanked cat sacB cassette ( Sma I/ Sfo I fragment from pLOI4162) was then ligated into the blunt PCR product to produce pLOI4159. Plasmid pLOI4159 served as a template for PCR amplification (JM ackA F1/R1 primers). This PCR product was used to repla ce the FRT site in the ackA region of KJ073 by double-crossover homologous recombination, with selection for ch roramphenicol resistance. The resulting clone was designated KJ076. Plasmid pLOI4159 was also digested with Pac I to remove the cat-sacB cassette and selfligated to produce pLOI4160, retaining the 18-bp translational stop sequence. Plasmid pLOI4160 served as a PCR template (JM ackA F1/R1 primers). This amplified fragment was used to replace the cat-sacB cassette in KJ076 by double-crossover hom ologous recombination with selection for loss of sacB After removal of pKD46 by growth at el evated temperature, the resulting strain was designated KJ079. In this strain, the deleted region has been replaced by the 18-bp translational stop sequence. The strategy used above to remove the FRT site from the ackA region was employed to make sequential deletions of citF sfcA and pta-ackA and to replace the deleted regions with the 18-bp translational stop sequence. Additional primer sets ( citF up2/down2 and citF 2/3) used to construct the citF deletion are included in Table 5-1 t ogether with the co rresponding plasmids (pLOI4629, pLOI4630, and pLOI4631). The re sulting strain wa s designated KJ104. The sfcA gene was deleted from strains KJ104 and KJ110, resulting in strains designated KJ119 and KJ122, respectively. Additional primer sets ( sfcA up/down and sfcA 1/2) used to
107 construct the sfcA deletions are included in Table 5-1 together with the corresponding plasmids (pLOI4283, pLOI4284, and pLOI4285). The pta-ackA operon (including the synt hetic translational stop sequence) was deleted from KJ122 to produce strain KJ134. Additional primer sets ( ackA up/ pta down and ackA 2/ pta 2) used to construct this deletion are included in Table 5-1 together with the corresponding plasmids (pLOI4710, pLOI4711, and pLOI4712). St rain KJ134 does not c ontain any FRT sites or foreign genes. Fermentations Seed cultures and fermentati ons were incubated at 37 C (100 rpm) in AM1 mineral salts medium (Martinez et al., 2007) containing 10%(w/v) glucose (555 mM), 100 mM KHCO3, and 1 mM betaine HCl. A mixture of 3M K2CO3 and 6N KOH was added to maintain pH and supply CO2. Differences in base composition (mixtures of 1:1, 4:1, 6:1) had little effect on fermentation. Fermentations were carried out in ve ssels with a working volume of 350 ml. Fermentations were inoculated at an initial OD550 of 0.01 (3.3 mg CDW/l) unless indicated otherwise. Fermentation vessels we re sealed except for a 16-gauge needle that served as a vent and a port for sample removal. Anaerobiosis was rapidly achieved during growth. Added bicarbonate served to ensure an atmosphere of CO2. Analyses Cell mass was estimated from the optical density at 550 nm (OD 1.0 = 333 mg of CDW/l) by using a Bausch Lomb Spectronic 70 spectrophotom eter. Organic acids and sugars were determined by using high performance li quid chromatography (Grabar et al., 2006).
108Results and Discussion Elimination of FRT Sites in KJ073 to Produce KJ091 The central anaerobic fermentation genes in E. coli ATCC 8739 wild type were sequentially deleted (Jantama et al., 2008a) us ing the method of Datsenko and Wanner (2000) with PCR products and removable antibiotic markers (FRT recognition sites and FLP recombinase). These constructions, in combination w ith metabolic evolution, were used to select mutant strains in which both gr owth and succinate production we re increased (Jantama et al., 2008a). The resulting strain KJ073 (Figure 5-1) produced 1.2 mole of succinate per mole of metabolized glucose (Jantama et al., 2008a). However, the deletion methods used for these constructions also left a single 82 to 85 bp genetic scar (FRT site ) in the region of each deleted gene ( ackA, ldhA, adhE, focA-pflB ). These FRT sites served as recognition sites for FLP recombinase (Storici et al., 1999) during removal of the antibiotic gene. Accumulation of these FRT sites in the chromosome significantly complicated further gene deletions. Prior to further strain improvements, all FR T sites in KJ073 were replaced with native DNA containing the desired deleti on (Grabar et al., 2006; Zhang et al., 2007; Jantama et al., 2008a). The resulting strain, KJ091, contains specific deletions in ackA, ldhA, adhE, focA-pflB, mgsA, and poxB and lacks all FRT sites. Production of succinate and co-products by strain KJ091 were equivalent to those of KJ073 (Table 5-2). Strain KJ091 was used as the parent for further improvements in succinate production (Figure 5-1). Deletion of tdcD and tdcE Decreased Acetate Production During the anaerobic fermentati on of glucose by unmodified E. coli (Figure 5-2), pyruvate formate-lyase ( pflB ) serves as the primary source of acetyl~CoA, precursor of acetyl~P (Karp et al., 2007; Kessler and Knappe, 1996). Acetyl~P is conve rted to acetate primarily by acetate kinase (ackA ). The abundance of acetate as a fe rmentation product in strains KJ073 and
109 KJ091 was surprising since these strains contain deletions in both ackA and pflB (Figure 5-2). Acetate production increased durin g growth-based selection (Jan tama et al., 2008a) consistent with the mutational activation of alternative pathways th at increase the production of acetyl~CoA, acetyl~P, and the yield of ATP (Fig ure 5-2 and 5-3). Elimination of acetate production represented a further oppor tunity to redirect metabolism and improve succinate yield. Figure 5-1. Strain constructions. A propionate kinase with acetate ki nase activity is encoded by the tdcD gene (Figure 53), but is typically produced only for the degradation of threonine (Hesslinger et al., 1998; Reed et al., 2003). During anaerobic growth w ith 10%(w/v) glucose, expression of tdcD could functionally replace ackA increasing the production of AT P and acetate from acetyl~P and providing a competitive growth advantage. The adjacent tdcE gene in the same operon is similar to pflB and encodes a -ketobutyrate formate-lyase (Hessl inger et al., 1998) with pyruvate formate-lyase activity. It is possible that increased expression of tdcE during anaerobic growth with 10% (w/v) glucose could in crease the production of acetyl~C oA (immediate precursor of acetyl~P) and waste potential reductant as form ate (Figure 5-3). To test this hypothesis, tdcD and tdcE (adjacent genes) were simultaneously deleted from KJ091 to produce KJ098. Deletion of
110 these two genes reduced acetate production by 42 % and increased succinate yield by 10% in KJ098 in comparison to KJ091, establishing the importance of this unexpected pathway in diverting carbon flow away from succinate (Table 5-2). Surprisingly, th e production of malate by KJ091 was also eliminated as a result of the ne w deletion in KJ098. The level of pyruvate in the broth (KJ098) also declined by 40%, an intermed iate that would be pr edicted to increase upon elimination of the alternative rout e for pyruvate metabolism provided by tdcE and tdcD Figure 5-2. Standard pathway for the anaerobic metabolism of glucose in E. coli. Solid arrows indicate reactions expected to be functional. Solid cro sses indicate dele ted genes. Boxed crosses represent key deletions used to construct initial strain for succinate production, KJ012 ( ldhA, adhE, ackA ). The dashed line represents oxidation of pyruvate to acetate by PoxB, a process th at is typically functional only under microaerophilic conditions. The dotted lines indicate reactions that are primarily associated with aerobic metabolism. Genes and enzymes: ppc phosphoenolpyruvate carboxylase; ldhA lactate dehydrogenase; pflB pyruvate formate-lyase; focA formate transporter; pta phosphate acetyltransferase; ackA, acetate kinase; adhE, alcohol dehydrogenase; pdh, pyruvate dehydroge nase complex; gltA, citrate synthase; mdh, malate dehydrogenase; fumA, fumB, and fumC, fumarase isozymes; frdABCD, fumarate reductase; fdh, formate dehydrogenase H; hyc, hydrogenase 3; icd, isocitrate dehydrogenase; acs acetyl~CoA synthetase; mgsA methylglyoxal synthase; poxB pyruvate oxidase; aldA aldehyde dehydrogenase; and aldB aldehyde dehydrogenase. The tdcE gene ( -ketobutyrate formate-lyase, homologous to pflB ) and tcdD gene (propionate ki nase, homologous to ackA ) are shown in parenthesis and are typically expressed duri ng threonine degradation.
111 The mechanisms responsible for reductions in malate (a problem contaminant during succinate purification) and pyruvate resulti ng from deletions of tdcD and tdcE are unknown. As expected, a reduction in acetate was accompanied by a small increase in succinate yield. However, the decreased side product formation was accompanied by a decrease in the volumetric productivity of succinate. Figure 5-3. Expanded portion of metabolism illustra ting the pathways of additional genes that have been deleted (solid crosses) in the context of standard central metabolism. Succinate and acetate are principal produc ts (boxed) from fermentations by KJ073 and derivatives. Genes and enzymes: citDEF citrate lyase; gltA citrate synthase; aspC asparte aminotransferase; ppc phosphoenolpyruvate carboxylase; sfcA NAD+linked malic enzyme; fumA & fumB fumarase; frdABCD fumarate reductase; pykA & pykF pyruvate kinase; tdcE -ketobutyrate formate-lyase (homologue of pflB ); pta phosphate acetyltransferase; tcdD propionate kinase (homologue of ackA ). Pyruvate dehydrogenase is thought to func tion primarily during aerobic metabolism. NAD+ -linked malic enzyme (sfcA ) is thought to function primarily for decarboxylation during gluconeogenesis. Ge nes encoding pyruvate formate-lyase ( pflB ) and acetate kinase ( ackA ) have been deleted in all strains described in this study and are shown in parenthesis. Acetyl ~P is hypothesized to be converted to acetate by unspecified reactions (indicated by ?) such as the phosphorylation and desphosphorylation of proteins, sponta neous hydrolysis, and other reactions.
112Citrate lyase ( citF ) Deletion Had no Effect on A cetate or Succinate Production Despite the elimination of two pathways fo r acetate production (Figure 5-2), significant amount of acetate was produced by strain KJ 098 (Table 5-2). Under anaerobic conditions, oxaloacetate is partitioned between a reduced product, malate, and a more oxidized intermediate, citrate (Figure 5-2). Citrate represents a potential source of this acetate. Citrate can be converted to oxaloacetate and acet ate by citrate lyase ( citDEF ), a mechanism proposed to recycle the intracellular oxaloacetate pool fo r other metabolic functions (Nilekani and Sivaraman 1983). Together with citrate synthase ( gltA ), these two enzymes could form a futile cycle that wastes the energy conserved in acetyl~CoA and accumulates acet ate. Citrate lyase expr ession is associated with growth on citrate as a co -substrate (Lutgens and Gottschal k, 1980; Kulla and Gottschalk, 1977) and this gene would not be expected du ring fermentations with excess glucose unless activated by mutations. Citrate lyase is a multi-enzyme complex com posed of three different polypeptide chains ( citDEF ) that are each essential for activ ity (Quentmeier et al., 1987). The citF gene was deleted from KJ098 to test the possibility that a citrate lyase futile cycl e is a primary source of acetate during glucose fermentation. Howeve r, acetate levels were essentially unchanged in the resulting strain KJ104, eliminating citrate lyas e as a primary source. Deletion of citF was accompanied by a small increase in pyruvate, but had no effect on succcinate yield (Table 5-2). Deletion of citF (KJ104) adversely affected the growth and re duced cell yield by 22% as compared to KJ098. Kulla (1983) reported that citrate lyase activity could not be completely eliminated in E. coli since it appears to serve a protec tive rather than a cat abolic purpose. High internal citrate pools resulting from citrate lyase de letion were proposed to interfere with v ital cell functions by chelating divalent cations. In addition, the accumulation of citrat e could result in an imbalance between glutamate and aspartate pool sizes during biosynthesis that may also adversely affect
113 growth. Despite a reduced cell yield, succinate yield, titer, average productivity, and acetate levels with KJ104 were comparable to those with KJ098 (Table 5-2). Deleting aspC had no Effect on Succinate Yield Aspartate aminotransferase ( aspC ) is a multifunctional enzyme that catalyzes the synthesis of aspartate, phenylalanine and other compounds by transamination. In this reaction, Laspartate is synthesized from oxaloacetate (an intermediate from PEP carboxylation) by transamination with L-glutamate. Aspartate is a constituent of proteins and participates in many other biosynthetic pathways. About 27 percent of the cellular nitrog en has been estimated to flow through aspartate (Reitzer, 2004). Aspartate bi osynthesis and succinate production share a common intracellular pool of oxaloacetate (Figure 5-3). Deletion of aspC could increase succinate production by reducing carbon flow into aspartate, but ma y also create an auxotrophic requirement that would prevent anaerobic growth in minimal salts medium such as AM1. However, deletion of aspC from KJ104 to produce KJ110 had no effect on succinate yield, cell yield, or acetate (Table 5-2). Thus aspartase ( aspC ) did not appear to divert significant levels of oxaloacetate away from succinate in this strain. In KJ110, aspartate may be produced by an alternative route such as aspartate ammonia-lyase ( aspA ) using ammonia and fumarate as substrates. Deleting sfcA Had no Effect on Succinate Yield Significant amounts of pyruvate are present at the end of fermentation with KJ104 and other strains of E. coli engineered for succinate production (Tab le 5-2). This pyruvate represents an unwanted product and a further opportunity to increase succinate yield. The high level of pyruvate could result from the decarboxylation of malate to pyruvate by malic enzyme ( sfc A) as illustrated in Figure 5-1.
114Table 5-2. Fermentation of 10% (w/v) glucose in mineral salts AM1 medium (1mM beta ine) by mutant strains of E. coli Succinate Yieldb Fermentation Products (mM) d,e,f Strain Culture Conditions (pH control, inoculum) Cell Yielda (g/L) mol/mol g/g Av. Vol. Prod c (g/L-h) Glucose Used (mM) Suc Mal Pyr Ace For KJ073 3M K2CO3+6N KOH (1:1) 0.01 OD550 inoculum 2.3 .1 1.20 .09 0.77 .03 0.82 .01 564 2 668 120 3 60 2 180 7 trace KJ091 3M K2CO3+6N KOH (1:1) 0.01 OD550 inoculum 2.2 .1 1.19 .02 0.78 .01 0.84 .01 576 687 109 72 155 trace KJ098 3M K2CO3+6N KOH (1:1) 0.01 OD550 inoculum 2.3 .1 1.30 .04 0.85 .02 0.79 .01 496 1 644 <1 42 88 KJ104 3M K2CO3+6N KOH (4:1) 0.01 OD550 inoculum 1.8 .1 1.31 .01 0.86 .01 0.78 .03 485 5 630 5 5 78 90 0 KJ104 3M K2CO3+6N KOH (6:1) 0.01 OD550 inoculum 1.9 .1 1.30 .01 0.85 .01 0.77 .01 482 620 3 94 81 KJ110 3M K2CO3+6N KOH (4:1) 0.01 OD550 inoculum 2.0 .1 1.28 .02 0.84 .01 0.79 0.01 500 640 0 4 76 110 1 KJ119 3M K2CO3+6N KOH (4:1) 0.01 OD550 inoculum 2.0 .1 1.29 .01 0.85 .01 0.81 .01 506 655 0 4 60 8 100 4 KJ122 3M K2CO3+6N KOH (4:1) 0.01 OD550 inoculum 2.3 .1 1.36 .02 0.89 .01 0.83 .01 500 680 0 <1 120 1 90 3 KJ122 3M K2CO3+6N KOH (6:1) 0.01 OD550 inoculum 2.0 .2 1.40 .02 0.92 .01 0.88 .04 510 8 700 0 6 59 110 KJ122 3M K2CO3+6N KOH (6:1) 0.15 OD550 inoculum 2.1 .1 1.46 .08 0.96 .05 0.84 .06 470 685 0 <1 120 3 122 KJ134 3M K2CO3+6N KOH (6:1) 0.01 OD550 inoculum 2.3 .1 1.53 .03 1.00 .02 0.75 .02 397 5 606 5 13 22 37 a Cell yield estimated from optical density (3 OD550nm = 1 g CDW/l). b Small amounts of glucose remained in some fermentations. Succi nate yields were calculated based on glucose metabolized. c Average volumetric productivity was calcu lated for total incubation time (96 h). d Abbreviations: Suc, succinate; Mal, malate; Pyr, pyruvate; Ace, acetate; Lac, lacate; For, formate. e Lactate was absent from all broths. f All data represent an average of 3 or mo re fermentations with standard deviations.
115 This enzyme is thought to function primar ily during gluconeogenesis (Oh et al., 2002; Stols and Donnelly, 1997; Unden and Kleefeld, 2004) rather than during the anaerobic catabolism of 10% w/v glucose. Although reductive carboxylation of pyruvate to form malate is thermodynamically favored, the kinetic parameters of this enzyme favor the dehydrogenation and decarboxylation under physiological conditi ons (Stols and Donnelly, 1997). Over-expression of this enzyme to carboxylate py ruvate has been previously us ed to construct strains of E. coli for succinate production (Sto ls and Donnelly, 1997). If malic enzyme ( sfcA ) functions to carboxylate pyruvate in KJ104 (and related strains) and contributes to succinate production, deleti on of this gene would be expected to reduce succinate yield and increase the le vel of side products such as pyruvate. Alternatively, if malic enzyme functions to decarboxylate malate in KJ104, deletion of sfcA would be expected to increase succinate yield and decrease the leve l of pyruvate. To our surprise deletion of the sfcA gene from KJ104 to produce KJ119 had no measur able effect on succinate production, growth, or side product levels (Table 5-2) in comparison to KJ104. Thes e results clearly demonstrated that malic enzyme is unimportant for succinate production in KJ104 and related strains, in sharp contrast to the succinate-pr oducing strains developed by Stol s and Donnelly (1997) in which over expression of malic enzyme was used as the primary route for succinate production. Deleting the Combination of aspC and sfcA Improved Succinate Yield Although no significant benefits were obs erved from individual deletions of sfcA or aspC from KJ104, we proceeded to test the effect of deleting both genes in combination. The combined deletion of both sfcA and aspC (strain KJ122) resulted in an unexpected increase in succinate yield and titer with a small reduction in acetate (Table 4-2), in comparison to the parent strain KJ110 and related strains (KJ1 04 and KJ119). The combined deletion ( aspC and sfcA ) in KJ122 increased succinate yield, succinate tite r, and average productiv ity by 8%, 13%, and 14%,
116 respectively as compared to KJ104. Although th e mechanism is unknown, it is possible that single deletions were ineffective because they were compensated in part by increased flow through the remaining enzyme activ ity, malic enzyme or aspartase (Figure 5-3), dampening any potential benefit. The increase in succinate yield and titer are presumed to result from increases in the availability of oxaloacetate and malate that allow a larger fraction of carbon to proceed to succinate. Strain KJ122 (Table 5-2) produced 1.4-1.5 mo l succinate per mol of glucose, 85% of the maximum theoretical yield (1.71 mol per mol gluc ose). To produce this high level of succinate and fully reduce malate, additional reductant was required. Although the source of this additional reductant is unknown, additional NAD(P)H could be provided by an increase in pyruvate flow through pyruvate dehydrogenase ( pdh ) or increased glucose me tabolism through the pentose phosphate pathway, together with the glyoxyl ate bypass. Although PDH co mplex is thought to function primarily during oxidative metabolism (Guest et al., 1989), it remains active during fermentation (de Graef et al., 1999) subject to me tabolite regulation (Kim et al., 2007). Further evidence of a functional PDH under anaerobic co nditions has been reported by Zhou et al. (2008). Strains containing deleti ons of pyruvate formate-lyase ( pflB ), lactate dehydrogenase ( ldhA ), and fumarate reductase ( frdBC ) are able to grow anaerobi cally and produce ethanol, but unable to grow after deletion of pdhR the transcriptional activator for pyruvate dehydrogenase. Increased function of pyruvate dehydrogenase would also offer an alternative route for acetyl~CoA, a potential precursor of acetate (Figure 5-2 and 5-3). Reduction in Pyruvate and Acetate by Deletion of pta Although KJ122 produced excellent succinate yields (1.4 mol/m ol glucose), acetate and pyruvate accumulated in the broth remained as a further opportunity to in crease yield. Pyruvate is presumed to accumulate from a metabolic ove rflow from glycolysis since pyruvate excretion
117 was reduced during growth in lower concentrati ons of sugars (data not shown). This excess pryuvate may be partially responsible fo r promoting acetyl~CoA production by pyruvate dehydrogenase (Figure 5-3). Excess acetyl~CoA could then be converted to acetyl~P by phosphotransacetylase ( pta ). Acetyl~P could be converted to acetate by a variety of routes including phosphorylation and de phosphorylation of proteins (ene rgy depleting futile cycle), non-enzymatic hydrolysis, etc. Deletion of phosphotransacetylase ( pta ) in KJ122 to produce KJ134 reduced acetate and pyruvate accumulation with a small further increa se in succinate production (Table 5-2). These results clearly indicate that acetyl~P from phosph ate acetyltransferase is the primary source of acetate in KJ122 and is consistent with pyr uvate dehydrogenase or the pentose phosphate pathway serving as the source of acetyl~CoA. The modest level of acetate is presumed to represent an aggregate of many low flux metabolic reactions and nonenzymatic hydrolysis. Biomass remained high. Volumetric productivity declined by 17% in comparison to KJ122 but remained equivalent to that of the star ting strain for this investigation, KJ073. With this strain, large amounts of glucose remained unmetabolized after the 4-day fermentation. Conclusions The metabolism of E. coli is remarkably adaptive. While some genetic changes are obviously needed to reengineer metabolism for a product such as succin ate through examination of central metabolic pathways, additional mutations were clearly needed to achieve theoretical yields of succinate from glucose. Engineered strains initially grew very poorly (Jantama et al., 2008a). Growth appears to have been limited in part by the low ATP yield (predicted by pathways to be approximately 1 ATP per glucose), although this is quite adequate for other rapidly growing fermentative organisms such as Zymomomas mobilis Growth-based selection for improved strains such as KJ012 through KJ091 co-selected for increased efficiency of ATP
118 production by increasing glycolytic flux (Jantama et al., 2008a), increasing the yield of ATP per glucose (directing more carbon flow into ac etyl~CoA, acetyl~P and acetate by recruiting pdh, pta, tdcD, tdcE ), and increasing succinate production to close the redox balance. These results clearly establish an initial ATP limitation as th e basis for growth based selection to increase succinate production. Further deletions (pta, tdcD, tdcE ) were required to eliminate alternative routes for acetate-linked ATP production. Some of the other observed improvement s in strains are not readily explained. These include the unexpected improvement observed after deletion of both aspC and sfcA when neither alone had a measurable effect. The beneficial effects of a poxB deletion during anaerobic fermentation (expected to be nonfunctional during anaerobic fermentation) and an mgsA deletion (very minor pathway) were also unanticipated (J antama et al., 2008a). With the deletion of pflB, tdcD, tdcE, and adhE obvious routes to acetate have been removed (Figure 5-2 and 5-3). Yet significant levels of acetate persis ted in strain KJ122. Deletion of pta encoding phosphate acetyltransferase was required to reduce the pr oduction of this undesirable co-product. The remaining pathway for acetate production fr om acetyl~P is unknown but could involve phosphorylation and dephosphorylat ion, non-enzymatic hydrolysis, pentose phosphate pathway, or others. The E. coli chromosome contains a large reservoi r of genes with diverse capabilities. Although activities have been described for most of these gene s, these vastly under estimate the catalytic potential of existing enzymes and the expanded range of activities that could be developed by relatively small mutational changes in structure or regulation. It is likely that some of the changes observed in this study such as the functional replacement of acetate kinase by the tdcD product (propionate kinase) an d pyruvate formate-lyase by the tdcE product ( -ketobutyrate
119 formate-lyase) have resulted from a regulatory mutation since both are within the same operon. Increased function of pyruvate dehydrogenase under anaerobic c onditions has been previously described as a structural mutati on that reduces the extent of ac tivity inhibition by NADH (Kim et al., 2007). This enzyme is a likely source of the additional NADH required for redox balance when succinate yields exceed 1.0 mol/mol glucose. A similar mutation may have occurred during the development of KJ073 and derivatives. Alternatively, highe r pyruvate dehydrogenase activity would be expected if NADH pools were reduced by high flux to succinate. The development of KJ134 for the production of succinat e at near theoretical yields during simple batch fermentations without complex nutrients represents an importa nt step toward the development of bio-based succinic acid as a platform chemical for renewable products.
120 CHAPTER 6 GENERAL CONCLUSIONS AND FUTURE DIRECTIONS General Accomplishments Derivatives of E. coli ATCC 8739 were engineered to prod uce either succinate or malate as the major product in mineral salts media usin g simple fermentations (anaerobic stirred batch with pH control) without the addition of pl asmids or foreign genes. This was done by a combination of gene deletions (genetic engine ering) and metabolic e volution with over 2000 generations of growth-based selection. After deletion of other fermentation genes ( ldhA adhE ackA ), the pathway for malate and succinate pro duction remained as the primary route for the regeneration of NAD+. Under anaerobic conditions, ATP pr oduction for growth was obligately coupled to malate dehydrogenase and fumarate reductase activities by the requirement for NADH oxidation. Selecting strains for improved gr owth co-selected for in creased production of succinate or malate. Additional deletions were introduced as further improvements ( focA, pflB, poxB, mgsA ). The best succinate biocatalysts, strains KJ060 ( ldhA adhE ackA focA, pflB ) and KJ073 ( ldhA adhE ackA focA, pflB, mgsA, poxB ), produce 622-733 mmol of succinate with molar yields of 1.2 to 1.6 per mole of metabolized glucose. The best malate biocatalyst, strain KJ071 ( ldhA adhE ackA focA, pflB, mgsA ), produced 516 mM malate with molar yields of 1.4 per mole of glucose metabolized. We attempted to reduce the fermentation time for succinate production by using high inocula. Although, high inocula increased succ inate yield per mole of glucose, average fermentation rates were lower than with lower inocula. Supplementation with acetate increased the yield of succinate and succinate productivy. Supplementation wi th yeast extract and peptone also increased the rate of sugar metabolism by increasing cell yield.
121 Metabolic flux analysis (MFA) shows that the specific succinate production rate was highest during exponential growth This is in agreement w ith the increase in the NADH production rate from the activity of the pyruvate dehydrogenase (PDH) complex. This NADH is mused for the reduction of fumarate to succinate, and redox balance is acheived. The cell yields (YATP) calculated were lower than the maximum theoretical cell yield (YATP MAX) in wild type and KJ017 due to energy uncoupling in which AT P produced from the fermentation process was used for cellular maintenance rather than for bios ynthesis. However, the values for the strains without PFLB and ACKA (KJ060, KJ071, a nd KJ073) are considerably higher than YATP MAX. A probable explanation is that mo re ATP was generated than our flux analysis indicates. A probable explanation is that mo re ATP was generated than our flux analysis indicates. KJ073 has been further improved by removi ng residual FLP recombinase sites (FRT sites) from the chromosomal regions of gene dele tions to create a strain devoid of foreign DNA, strain KJ091 ( ldhA adhE ackA ( focA pflB ) mgsA poxB ). KJ091 was further engineered for improvements in succinate production by el iminating acetate, malate, and pyruvate as significant co-products. Deletion of the propionate kinase ( tdcD ; acetate kinase homologue) and -ketobutyrate formate-lyase ( tdcE ; pyruvate formate-lyase ho mologue) genes reduced the acetate level by 50% and increased succinate yiel d (1.3 mol/mol glucose) by almost 10%, as compared to KJ091 and KJ073. Deletion of two genes encoding enzymes involved in oxaloacetate metabolism, aspartate aminotransferase (aspC ) and the NAD+-linked malic enzyme ( sfcA ) (KJ122) significantly in creased succinate yield (1.5 mol/m ol glucose), succinate titer (700 mM), and average volumetric productivity (0.9 g/l/h). Residual pyruvate and acetate levels were substantially decreased by deletion of pta encoding phosphotransacetylase, to produce KJ134 ( ldhA adhE ( focA pflB ) mgsA poxB tdcDE citF aspC sfcA ( pta-ackA )). Strains
122 KJ122 and KJ134 produced near theoretical yields of succinate during simple, anaerobic, batch fermentations using mineral salts medium. Both may be useful as biocatalys ts for the commercial production of succinate. Future Works A primary technology for use of succinic acid is selective reduction to produce the wellknown butanediol (BDO), tetrahydrofuran (THF) a nd gamma-butyrolactone (GBL) families. The hydrogenation/reduction chemistry for the conversi on of succinic acid to this family of compounds is well established and is similar to the conversion of maleic anhydride to the same families of compounds (Werpy et al., 2004). For th is application, product purity must be taken into consideration during the development of strains because impurities from the fermentation might otherwise affect or inhibit the subsequent hydroge nation/reduction process. Significant improvement in fermentative succ inate production, including further strain development and process engineering is required for it to be competitive with the petrochemical production routes. Productivity improvements are required to reduce the capital and operation costs of the fermentation. A minimum productivity of 2.5 g/l-h has been proposed for the process needs (Werpy et al., 2004). A final titer is also important when considering overall process costs in which a high final titer will reduce overall separation and concentrating costs. In an ideal process, the fermentation would be performed at low pH, without requiring any neutralization. The cost of neutralizatio n is not necessarily cost prohibitive, but the conversion of the salt to the free acid does add significant costs. Metabolic e volution offers an excellent way to select for improved strains that produce su ccinate with higher titers, yi elds, and the productivities, and tolerate acidic conditions. The exact pathway for gl ucose transport in which the strains are used to import glucose into the cells is yet to be elucidated. A better understanding of what mechanism the strains use for glucose transport and of how the strains metabolize glucose would
123 allow us to engineer the strain s for higher rates of glucose cata bolism, resulting in increases in succinate productivities. The inhibition or toxicity from products accumulated during fermentation has to be removed. The high levels of accumulated succinat e and acetate are likely to inhibit the glucose consumption and harm the cells. Product removal from the culture media is challenging. The use of base neutralization by simultaneously ge nerating carbon dioxide a nd forming precipitated succinate salts would be one me thod to reduce the inhibitory effects of accumulated products. However, the type and concentration of base, and time of neutralization would need to be extensively studied in order to improve the process development. Future use of Succinate as Bioplastics The synthesis of polymers based on monome rs from renewable feedstock is rapidly growing since the global environmen t issue and the rapid depletion of the oil reserves have been extensively concerned. Renewable monomers with a large functional diversity are available from a wide range of resources (Noordover et al., 2006 ). An active research using succinate as a renewable monomers substrate is also in progress for the syntheses, structures, physical properties, and degradation of biodegradable aliphatic polyeste rs (Kuwabara et al., 2002). Poly(butylene succinate) (PBS) is one of the most promising polyesters, and its good biodegradability in the natural environment has been reported (Tokiwa and Suzuki, 1977). PBS and related copolymers have been used as enviro nmentally biodegradable thermoplastics, as well as bioabsorable/biocompatible medical materials since they have useful mechanical properties due to their high molecular weight (Azim et al., 2006). Moreover, random copolymers such as poly(butylene succinateco -butylene adipate), poly(butylene succinateco -ethylene succinate), and poly(butylene succinateco -hexamethylene succinate) have been also developed to control the physical properties and the speed of biodegradati on (Gan et al., 2001).
124 Poly(propylene succinate) (PPS ) is a relatively new biodegrad able polymer that is also produced using monomers from renewable resources (Hartlep et al., 2002; Kim et al., 2004). PPS has gained increasing interest, since it exhibits a highest biodegradation rate compared to that of PBS, poly(ethylene succinate), poly(propylene se bacate), and other biodegradable polymers due to its low crystallinity (Bik iaris et al., 2006; Umare et al ., 2007). However, PPS has low mechanical properties that limit its applications. The applications of biodegradable polymers including PBS, PPS, and their copolymers, have been also found in biomedical fields such as for constructing absorbable bone plates, other surgical fixation devices, surgical sutures, and controlled release drug carriers (Song and Sung, 1995). The biocompatibility, toxicity, and immunogeni city, desirable mechanical properties, and predictable biodegradation rate of polymers are currently studied and developed. As seen, succinate based polymers have a wide variety of applications. The demands for succinate are rapidly increasing. Production of succinate fe rmentatively, including strain developments, process design, and downstream purification procedures would be further extensively developed in order to supply a good source of succinate to its growing market.
125 LIST OF REFERENCES Abbott BJ, Clam an, A. 1973. The relationship of substrate, growth rate, and maintenace coefficient to single cell protein pr oduction. Biotechnol Bioeng 15:117-127. Abdel-Hamid AM, Attwood MM, Guest JR. 2001. Pyruvate oxidase cont ributes to the aerobic growth efficiency of Escherichia coli Microbiol 147:1483-1498. Ahmed I, Morris D. 1994. Replacing petrochemical with biochemicals. Institute for Local Reliance, Washington, DC. Ajl SJ, Werkman CH. 1948. Enzymatic fixation of carbon dioxide in -ketoglutaric acid. Proc Natl Acad Sci USA 34:491-498. Andersen KB, von Meyerburg K. 1980. Are growth rates of Escherichia coli in batch cultures limited by respiration? J Bacteriol 144:114-123. Andersson C, Hodge D, Berglund KA, Rova U. 2007. Effect of different carbon sources on the production of succinic acid usi ng metabolically engineered Escherichia coli Biotechnol Prog 23(2):381-388. Anastas P, Warner J. 1998. Green Chemistry: Theory and Practice. London, UK: Oxford University Press. Aristidou AA. 1994. Application of Metabolic and Biochemical Engineering Techniques for the Enhancement of Recombinan t Protein Production in Escherichia coli. Ph.D. Thesis, Department of Chemical Engineerin g, Rice University, Houston, TX. Aristidou AA, San KY, Bennett GN. 1995. Metabolic engineering of Escherichia coli to enhance recombinant protein production through acetat e reduction. Biotechnol Prog 11: 475-478. Aristidou AA, San KY, Bennett GN. 1999. Metabolic flux analysis of Escherichia coli expressing the Bacillus subtilis acetolactate synthase in ba tch and continuous cultures. Biotechnol Bioe ng 63(6):737-749. Asghari A, Bothast RJ, Doran JB, I ngram LO. 1996. Ethanol production from hemicellulose hydrolysates of agricultural residues us ing genetically engineered Escherichia coli strain KO11. J Industrial Microbiol 16:42-47. Azim H, Dekhterman A, Jiang Z, Gross RA. 2006. Candida antarctica lipase B-catalyzed synthesis of Poly(butylenes succinate): shorter chain building blocks also work. Biomacromolecules 7:3093-3097. Bauchop T. and Elsden SR. 1960. The growth of microorganisms in relation to their energy supply. J Gen Microbiol 23:457-469.
126 Bergen WG, Bates DB. 1984. Ionorphores: their effect on producti on efficiency and mode of action. J Anim Sci 58:1456-1483. Bikiaris DN, Papageorgiou GZ, Achilias DS. 2006. Synthesis and comparative biodegradability studies of three Poly(alkylene succin ate)s. Polym Degrad Stab 91:31-43. Breuer M, Ditrich K, Habicher T, Hauer B, Kesse ler M, Stuermer R, Zelinski T. 2004. Industrial methods for the production of optically active intermediate s. Angew Chem Int Ed 43:788. Bryant MP, Small N. 1956. Characteristics of two ne w genera of anaerobic curved rods isolated from the rumen of cattle. J Bacteriol 72:22-26. Bryant MP, Bouma C, Chu H. 1958. Bacteroides ruminicola n. sp. And the new species Succinomanas amylolytica Species of succinic acid produci ng anaerobic bacteria by the bovine rumen. J Bacteriol 76:15-23. CaJacob C, Gavino GR, Frey PA. 1985. Pyruvate dehydrogenase complex of Escherichia coli : Thiamin pyrophosphate and NADH-dependent h ydrolysis of acetyl~CoA. J Biol Chem 260(27):14610-14615. Causey TB, Zhou S, Shanmugam KT, Ingram LO. 2003. Engineering the metabolism of Escherichia coli W3110 for the conversion of sugar to redox-neutral and oxidized products. Homoacetate production. Pr oc Natl Acad Sci USA 81:4348-4352. Causey TB, Shanmugam KT, Yomano LP, Ingram LO. 2004. Engineering Escherichia coli for efficient conversion of glucose to pyr uvate. Proc Natl Acad Sci USA 101:2235-2240. Cecchini G, Schroder I, Gunsalus RP, Maklas hima E. 2002. Succinate dehydrogenase and fumarate reductase from Escherichia coli Biochimica et Biophysi ca Acta 1553:140-157. Chang YY, Cronan JE Jr. 2000. Conversion of Escherichia coli pyruvate decarboxylase to an alpha-ketobutyrate oxidase. Biochem J 352:717-724. Chao YP, Liao JC. 1994. Metabolic respons es to substrate futile cycling in Escherichia coli J Biol Chem 269:5122-5126. Chatterjee R, Cynthia SM, Kathleen C, Da vid PC, Donnelly MI. 2001. Mutation of the ptsG gene results in increased production of succinate in fermentation of glucose by Escherichia coli Appl Environ Microbiol 67:148-154. Clark, DP. 1989. The fermentation pathways of Escherichia coli FEMS Microbiol Rev 63:223234. Cortay JC, Bleicher F, Duclos B, Cenatiempo Y, Gautier GC, Prato JL, and Cozzone AJ. 1984. Utilization of acetate in Escherichia coli : structural organization and differential expression of the ace operon. Biochemie 71:1043-1049.
127 Cox SJ, Shalel LS, Sanchez AM, Lin H, Peercy B, Bennett GN, San KY. 2006. Development of a metabolic network design and optimiza tion framework incorporating implement constraints: A succina te production case study. Metab Eng 8:46-57. Curtis SJ, Epstein W. 1975. Phosphorylation of Dglucose in Escherichia coli mutants defective in glucosephosphotransferase, mannosephosphotransferase, and glucokinase. J Bacteriol 122:1189-1199. de Graef MR, Alexeeva S, Snoep JL, de Mattos MJ. 1999. The steady-state internal redox state (NADH/NAD) reflects the external redox state a nd is correlate d with catabolic adaptation in Escherichia coli J Bacteriol 181:2351-2357. Dake S, Gholap RV, Chaudhuri RV. 1987. Carbonylation of 1,4butanediol diacetate using rhodium complex catalyst: a kinetic study. Ind Eng Chem Res 26:1513-1518. Datsenko KA, Wanner BL. 2000. One-step in activation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci USA 97:6640-6645. Davis CP, Cleven D, Brown J, Balish E. 1976. Anaerobiosprillium a new genus of spiral-shaped bacteria. Int J Syst Bacteriol 26:498-504. Doelle HW, Hollywood N. 1976. Effect of specifi c growth rate and gluc ose concentration on growth and glucose metabolism of Escherichia coli K-12. Microbiol 17: 23-33. Donnelly MI, Millard CS, Clark DP, Chen MJ, Ra thke JW. 1998. A novel fermentation pathway in an Escherichia coli mutant producing succinic acid, acetic acid, and ethanol. Appl Biochem Biot echnol 70:187-198. Du C, Lin SK, Koutinas A, Wang R, Webb C. 2007. Succinic acid produc tion from wheat using a biorefining strategy. Appl Microbiol Biot echnol 76(6):1263-1270. el-Mansi EM, Holms WH. 1989. Control of carbon flux to acetate excretion during growth of Escherichia coli in batch and continuous cultures. J Gen Microbiol 135:2875-2883. Egyud LG, Szent-Gyorgyi A. 1966. On the regula tion of cell division. Proc Natl Acad Sci USA 56:203-207. Farmer WR, Liao JC. 1997. Reduction of aerobic acetate production by Escherichia coli Appl Environ Microbiol 63:3205-3210. Fraenkel DG. 1996. Chapter 14, Glycolysis. In A. Bck, R. Curtiss III, J. B. Kaper, F. C. Neidhardt, T. Nystrm, K. E. Rudd, and C. L. Squires (ed.), EcoSal Escherichia coli and Salmonella : cellular and molecular biology. [On line.] http://www.ecosal.org. ASM Press, Washington, D.C.
128 Gan Z, Abe H, Kurokawa H, Doi Y. 2001. Solid-S tate microstructures, thermal properties, and crystallization of biodegradable Poly(butyl ene succinate) (PBS) and its copolyesters. Biomacromolecules 2:605-613. Glassner DA, Datta R. 1992. Process for the pr oduction and purification of succinic acid. US Patent 5,143,834. Gokarn RR, Eiteman MA, Altman E. 2000. Metabolic analysis of Escherichia coli in the presence and absence of carboxylating enzymes phosphoenolpyruvate carboxylase and pyruvate carboxylase. Appl E nviron Microbiol 66:844-1850. Gottschalk G. 1985. Bacterial metabolis m. 2nd ed. Springer-Verlag, New York. Grabar TB, Zhou S, Shanmugam KT, Yomano LP, Ingram LO. 2006. Methylglyoxal bypass identified as source of chiral contamination in L(+) and D(-) lactate fermentations by recombinant Escherichia coli Biotechnol Lett 28:1527-1535. Guest JR, Angier SJ, Russell GC. 1989. Structure, expression, an d protein engin eering in the pyruvate dehydrogenase complex of Escherichia coli Ann NY Acad Sci 573:76-99. Guettler MV, Jain MK, Rumler D. 1996a. Method for making succinic acid, bacterial variants for use in the process, and methods of obtaining variants. US patent 5,573,931. Guettler MV, Jain MK, Soni BK. 1996b. Process for making succinic acid, microorganisms for use in the process and methods of obtaini ng the microorganisms. US patent 5,505,004 Guettler MV, Jain MK, Soni BK. 1998. Proces s for making succinic acid, microorganisms for use in the process and methods of obtaining the microorganisms. US Patent 5,723,322. Hager LP. 1957. Trypsin activation of a ferricyanide-linked pyruvic acid oxi dation. J Biol Chem 229(1):251-263. Hartlep H, Hussmann W, Prayit no N, Meynial-Salles I, Zeng AP. 2002. Study of two-stage processes for the microbial production of 1,3-propanediol fro m glucose. Appl Microbiol Biotechnol 60:60-66. Hatti-Kaul R, Tornvall U, Gustafsson L, Borj esson P. 2007. Industrial biotechnology for the production of bio-based chemical s-a cradle-to-grave perspective. Trends in Biotechnol 25(3):119-124. Hempfling WP, Mainzer SE. 1975. Effects of vary ing the carbon source limiting growth on yield and maintenance characteristics of Escherichia coli in continuous culture. J Bacteriol 123:1076-1087.
129 Hesslinger C, Fairhurst SA, Sa wers G. 1998. Novel keto acid formate-lyase and propionate kinase enzymes are components of an anaerobic pathway in Escherichia coli that degrades L threonine to propiona te. Mol Microbiol 27(2):477-492. Hillman JD. 1979. Mutant analysis of glyceraldehyde 3-phosphate dehydrogenase in Escherichia coli Biochem J 179(1):99-107 Hopper DJ, Cooper RA. 1971. The regulation of Escherichia coli methylglyoxal synthase: a new control site in glycolysis. FEBS Lett 13:213-216. Iverson TM, Luna-Chavez C, Croal LR, Cecch ini G, Rees DC. 2002. Crystallographic studies of the Escherichia coli quinol-fumarate reductase wi th inhibitors bound to the quinol-binding site. J Biol Chem 277:16124-16130. Jain MK, Datta R, Zeikus JG. 1989. High value organic acids fermentation-emerging processes and products. In: Ghosh TK (ed) Bioprocess engineering: the first generation. Ellis Harwood, Chichester, pp 366-398. Jantama K, Haupt MJ, Svoronos SA, Zhang X, Moore JC, Shanmugam KT, Ingram LO. 2008a. Combining metabolic engineering and meta bolic evolution to develop nonrecombinant strains of Escherichia coli C that produce succinate a nd malate. Biotechnol Bioeng 99(5):1140-1153. Jantama K, Zhang X, Moore JC, Svoronos SA, Shanmugam KT, Ingram LO. 2008b. Eliminating side products and increasing succ inate yields in engineered strains of Escherichia coli C. Biotechnol Bioeng. DOI10.1002/bit.22005. Jin S, Ye K, Shimizu K. 1997. Metabo lic flux distributions in recombinant Saccharomyces cerevisiae during foreign protein pr oduction. J Biotech 54:161-174. Karp PD, Keseler IM, Shearer A, Latendresse M, Krummenacker M, Paley SM, Paulsen IT, Collado-Vides J, Gamma-Castro S, Peralta-Gi l M, Santos-Zavaleta A, Penaloza-Spinola M, Bonavides-Martinez C, Ingraham J. 2007. Multidimensional annotation of Escherichia coli K-12 genome. Nucl Acids Res 35(22):7577-7590. Kessler D, Knappe J 1996. Anaerobic dissimilation of pyruvate In Neidhardt FC, Curtiss III R, Ingraham JL, Lin ECC, Low KB, Magasani k B, Reznikoff WS, Riley M, Schaechter M, Umbarger HE, editors. Escherichia coli and Salmonella : cellular and molecular biology. ASM Press, Washington, D.C. pp 199-205. Keseler IM, Collado-Vides J, Gamma-Castro S, Ingraham J, Paley S, Paulsen IT, Peralta-Gil M, Karp PD. 2005. Ecocyc: A comprehensive database resource for Escherichia coli Nucl Acids Res 33:D334-D337.
130 Kim DY, Yim SC, Lee PC Lee WG, Lee SY, Chang HN. 2004. Batch and continuous fermentation of succinic aci d from wood hydrolysate by Mannheimia succiniciproducens MBEL55E. Enzyme Microb Biotechnol 35:648-653. Kim Y, Ingram LO, and Shanmuga m KT. 2007. Construction of an Escherichia coli K-12 mutant for homoethanologenic fermentation of glucose or xylose without foreign genes. Appl Environ Micr obiol 73: 1766-1771. Kornberg HL and Malcovati M. 1973. Control of in situ of the pyruvate kinase activity of Escherichia coli FEBS Letters 32(2):257-259 Knappe JH, Balschkowski P, Grobner P, Schmitt T. 1974. Pyruvate formate-lyase of Escherichia coli: The acetyl-enzyme intermediate. Eur J Biochem 50:253. Knappe J, Sawers G. 1990. A radical-chemical route to acetyl~CoA: th e anaerobiclly induced pyruvate formate-lyase system of Escherichia coli. FEMS Microbiol Rev 75:383-398. Koebmann BJ, Westerhoff HV, Snoe p JL, Nilsson D, Jensen PR. 2002. The glycolytic flux in Escherichia coli is controlled by the demand fo r ATP. J Bacteriol 184:3909-3916. Kornberg H. 1966. The role and control of the glyoxylate cycle in Escherichia coli Biochem J 99: 111. Krebs A, Bridger WA.1980. "The kinetic prope rties of phosphoenolpyruvat e carboxykinase of Escherichia coli Can J Biochem 58(4):309-318. Kulla H, and Gottschalk G. 1977. Energy-depe ndent inactivation of citrate lyase in Enterobacter aerogenes J Bacteriol 132(3): 764-770. Kulla HG. 1983. Regulatory citrate lyase mutants of Salmonella typhimurium J Bacteriol 153: 546-549. Kumari S, Tishel R, Eisenbach M, Wolfe AJ. 1995. Cloning, characteriza tion, and functional expression of acs, the gene which encodes acetyl coenzyme A synthetase in Escherichia coli J Bacteriol 177(10):2878-2886. Kuwabara K, Gan Z, Nakamura T, Abe H, Doi Y. 2002. Molecular mobility and phase structure of biodegradable Poly(butylene succi nate) and Poly(butylene succinateco -butylene adipate). Biomacromolecules 3:1095-1100. Laivenieks M, Vieille C, Zeikus JG. 1997. Cloning, sequencing, and overexpression of the Anaerobiospirillum succiniciproducens phosphoenolpyruvate carboxykinase ( pckA ) gene. Appl Environ Microb iol 63(6): 2273-2280.
131 Lee EC, Yu D, Martinez de Velasco J, Tessa rollo L, Swing DA, Court DL, Jenkins NA, Copeland NG. 2001. A highly efficient Escherichia coli -based chromosome engineering system adapted for recombinogenic targeti ng and subcloning of BAC DNA. Genomics 73:56-65. Lee SY, Hong SH, Lee SH, Park SJ. 2004. Fermentative production of chemicals that can be used for polymer synthesis. Macromol Biosci 4:157-164. Lee SJ, Lee DY, Kim TY, Kim BH, Lee J, Lee SY. 2005. Metabolic engineering of Escherichia coli for enhanced production of succinic acid, based on genome comparison and in silico gene knockout simulation. Appl Environ Microbiol 71:7880-7887. Lee SJ, Song H, Lee SY. 2006. Genome -based metabolic engineering of Mannheimia succiniciproducens for succic acid production. A ppl Environ Microbiol 72(3):1939 1948. Lin H, Bennett GN, San KY. 2005a. Ch emostat culture characterization of Esherichia coli mutant strains metabolically engineered fo r aerobic succinate production: A study of the modified metabolic network based on meta bolite profile, enzyme activity, and gene expression profile. Metab Eng 7:337-352. Lin H, Bennett GN, San KY. 2005b. Metabolic engineering of aerobic succinate production systems in Escherichia coli to improve process productivi ty and achieve the maximum theoretical succinate yield. Metab Eng 7:116-127. Lin H, Bennett GN, San KY. 2005c. Effect of carbon sources differing in oxidation state and transport route on succinate production in metabolically engineered Escherichia coli. J Ind Microbiol Biotechnol 32:87-93. Lin H, Bennett GN, San KY. 2 005d. Fed-batch culture of a metabolically engineered Escherichia coli strain designed for high-level su ccinate production and yield under aerobic conditions. Biotechnol Bioeng 90:775-779. Lorenz P, Zinke H. 2005. White biotechnology: differences in US and EU approaches? Trends in Biotechnol 23(12):570-574. Lutgens M, and Gottschalk G. 1980. Why a co-substrate is required for the anaerobic growth of Escherichia coli on citrate. J Gen Microbiol 119:63-70. Martinez-Morales F, Borges AC, Martinez A, Shanmugam KT, Ingram LO. 1999. Chromosomal integration of heterologous DNA in Escherichia coli with precise removal of markers and replicons used during constr uction. J Bacteriol 181:71437148. Martinez A, Grabar TB, Shanmugam KT, Yoma no LP, York SW, Ingram LO. 2007. Low salt medium for lactate and etha nol production by recombinant Escherichia coli B. Biotechnol Lett 29:397-404.
132 McKinlay JB, Vieille C, Zeikus JG. 2007. Prosp ects for a bio-based succinate industry. Appl Microbiol Biotechnol 76:727-740. Meyer HP, Leist C, Fiechter A. 1984. A cetate formation in continuous culture of Escherichia coli K12 D1 on defined and complex media. J Biotechnol 1: 355-358. Meynial-Salles I, Dorotyn S, Soucaille P. 2007 A new process for the continuous production of succinic acid from glucose at high yield, titer and productivity. Biotechnol Bioeng 99(1):129-135. Millard CS, Chao YP, Liao JC, Donnelly MI. 1996. Enhanced production of succinic acid by overexpression of phosphoenol pyruvate carboxylase in Escherichia coli Appl Environ Microbiol 62:1808-1810. Miller JH. 1992. A short course in bacter ial genetics. A laboratory manual and handbook for Escherichia coli and related bacteria. Plainview: Cold Spring Harbor Laboratory Press. p 456. Moniruzzaman M, Lai X, York SW, Ingram LO 1997. Extracellular me libiose and fructose are intermediates in raffinose catabolism dur ing fermentation to ethanol by engineered enteric bacteria. J Bacteriol 179:1880-1886. Morikawa, M., K. Izui, et al. 1980. Regulation of Escherichia coli phosphoenolpyruvate carboxylase by multiple effectors in vivo. Estima tion of the activities in the cells grown on various compounds. J Biochem (Tokyo) 87(2):441-449. Nghiem NP, Davison BH, Suttle BE, Richards on GR. 1997. Production of succinic acid by Anaerobiospirillum succinoproducens Appl Biochem Biot echnol 63-65: 565-576. Nghiem PP, Donnelly MI, Mill ard CS, Stols L. 1999. Method for the production of dicarboxylic acids. US Patent 5,869,301. Nielsen J, Villadsen J. 1994. Bioreaction engineering principles. New York: Plenum Press. Nilekani S, Sivaraman C.1983. Purificati on and properties of citrate lyase from Escherichia coli. Biochemistry 22(20):4657-4563. Nimmo GA and Nimmo HG.1984. The regulatory properties of isocit rate dehydrogenase kinase and isocitrate dehydrogenase phosphatase from Escherichia coli ML308 and the roles of these activities in the control of isocitrate dehydrogenase. Eur J Biochem 141:409-414. Noordover BA, van Staalduinen VG, Duchatea u R, Koning CE, van Benthem R A.T.M Mak M, Heise A, Frissen AE, van Haveren J. 2006. Coand terpolyesters based on isosorbide and succinic acid fo r coating applications: synthe sis and characterization. Biomacromolecules 7:3406-3416.
133 Nordhoff S, Hocker H, Gebhardt H. 2007. Rene wable resources in the chemical industryBreaking away from oil? Biotechnol J 2:1505-1513. Oh MK, Rohlin L, Kao KC, Liao JC. 2002. Gl obal expression profiling of acetate-grown Escherichia coli J Biol Chem 277:13175-13183. Ohta K, Beall DS, Shanmugam KT, Ingr am LO. 1991. Genetic improvement of Escherichia coli for ethanol production: chro mosomal integration of Zymomonas mobilis genes encoding pyruvate decarboxylase and alc ohol dehydrogenase II. Appl Eviron Microbiol 57:893. Okino S, Inui M, Yukawa H. 2005. Production of organic acids by Corynebacterium glutanicum under oxygen deprivation. Appl Microbiol Biotec hnol 68:475-480. Park SJ, Cotter PA, Gunsalus RP. 1995. Regulation of malate dehydrogenase ( mdh ) gene expression in Escherichia coli in response to oxygen, carbon, and heme availability. J Bacteriol 177(2):6652-6656. Patnaik R, Roof WD, Young RF, Liao JC. 1992. Stimulation of glucose metabolism in Escherichia coli by a potential futile cycl e. J Bacteriol 174:7527-7535. Phue JN, Shiloach J. 2004. Transcri ption levels of key metabolic genes are the cause for different glucose utilization pathways in E. coli B (BL21) and E. coli K (JM109). J Biotechnol 109: 21-30. Posfai G, Koob MD, Kirkpatrick HA, Blattner FR. 1997. Versatile insertion plasmids for targeted genome manipulations in bacteria: Isolation, de letion, and rescue of the pathogenicity island LEE of the Escherichia coli O157:H7 genome. J Bacteriol 179:4426-4428. Purvis JE, Yomano LP, Ingram LO. 2005. Enha nced trehalose production improves growth of Escherichia coli under osmotic stress. Appl Environ Microbiol 71:3761-3769. Quail MA, Guest JR. 1995. Purification, characte rization and mode of action of PdhR, the transcriptional repressor of the pdhR-aceEF-lpd operon of Escherichia coli. Mol Microbiol 15(3):519-29. Quentmeier A, Holzenburg A, Mayer F, Antranikia n G. 1987. Reevaluation of citrate lyase from Escherichia coli Biochim Biophys Acta 913(1):60-5. Reed JL, Vo TD, Schilling CH, Palsson BO. 2003. An expanded genome-scale model of Escherichia coli K-12 (iJR904 GSM/GPR). Ge nome Biol 4(9):R54. Reitzer, L. 2004. Chapter 184.108.40.206, Biosynthesis of Glutamate, Aspartate, Asparagine, L-Alanine, and D-Alanine. In A. Bck, R. Curtiss III, J. B. Kaper, F. C. Neidhardt, T. Nystrm, K. E. Rudd, and C. L. Squires (ed.), EcoSal Escherichia coli and Salmonella : cellular and molecular biology. [Online.] http://www.ecosal.org. Washington, D.C., ASM Press
134 Sado T, Tajima M, Noguchi Research Foundatio n. 1980. Carbonylation of 1,4-butanediol by homogeneous catalyst. Japa n Kakai Tokkyo Koho 8,051,037. Samuelov NS, Lamed R, Lowe S, Zeikus JG. 1991. Influence of CO2-HCO3 levels and pH on growth, succinate production, and enzyme activities of Anaerobiospirillum succinoproducens Appl Environ Micr obiol 57:3013-3019. Sanchez AM, Bennett GN, San KY. 2005a. Nove l pathway engineering design of the anaerobic central metabolic pathway in Escherichia coli to increase succinate yield and productivity. Metab Eng 7:229-239. Sanchez AM, Bennett GN, San KY. 2005b. Efficient succinic acid production from glucose through overexpression of pyruvate carboxylase in an Escherichia coli alcohol dehydrogenase and lactate dehydrogenas e mutant. Biotechnol Prog 21:358-365. Sanchez AM, Bennett GN, San KY. 2006. Batch cu lture characterization and metabolic flux analysis of succinate-producing Escherichia coli strains. Metab Eng 8:209-226. Sanwal BD. 1970. Regulatory mechanisms invol ving nicotinamide adenine nucleotide as allosteric effectors: A control of glucos e 6-phosphate dehydrogenase. J Biol Chem 245(7):1625-1631. Sawers G, Bock A. 1988. Anaerobic regul ation of pyruvate formate-lyase from Escherichia coli K-12. J Bacter iol 170:5330-5336. Sauer M, Porro D, Mattanovich D, Branduardi P. 2008. Microbial production of organic acids: expanding the markets. Trends in Biotechnol 26(2):100-108. Scheifinger CC, Wolin MJ. 1973. Propionate forma tion from cellulose and soluble sugars by combined cultures of Bacteroides succinogenes and Selenomonas ruminantium Appl Microbiol 26:789-795. Schmid A, Dordick JS, Hauer B, Kiener A, Wubbolts M, Witholt B. 2001. Industrial biocatalysis today and tomorrow. Nature 409:258. Siddiquee KA, Arauzo-Bravo MJ, Shimizu K. 2004. Effect of a pyruvate kinase ( pykF -gene) knockout mutation on the control of gene expression and metabolic fluxes in Escherichia coli FEMS Microbiol Lett 235(1):25-33. Smith TE, Balasubramanian KA, and Beezley A. 1980. Escherichia coli Phosphoenolpyruvate carboxylase: Studies on the mechanism of syne rgistic activation by nucleotides. J Biol Chemistry 255(4):1635-1642. Song DK, Sung YK. 1995. Synthesis and characteri zation of biodegradable Poly(1,4-butanediol succinate). J Appl Polym Sci 56(11):1381-1395.
135 Song H, Lee SY. 2006. Production of succinic acid by bacterial fermentation. Enzyme Microb Technol 39:352-361. Song H, Lee JW, Choi S, You JK, Hong WH Lee SY. 2007. Effects of dissolved CO2 levels on the growth of Mannheimia succiniciproducens and succinic acid production. Biotechnol Bioeng 98(6):1296-1304. Stephanopoulos G, Vallino JJ. 1991. Network rigidity and metabolic engineering in metabolite overproduction. Science 252:1675-1681. Stephanopoulos G, Aristidou AA, Nielsen J. 1998. Metabolic e ngineering: Prin ciples and Methodologies. Acadamic Press. Sandiego, CA. Stols L, Donnelly MI. 1997. Production of su ccinic acid through overexpression of NAD dependent malic enzyme in an Escherichia coli mutant. Appl Environ Microbiol 63: 2695-2701. Storici F, Coglievina M, Bruschi CV. 1999. A 2m DNA-based maker recycling system for multiple gene disruption in the yeast Saccharomyces cerevisiae Yeast 15:271-283. Tarmy EM, Kaplan NO. 1968a. Chemical characte rization of D-lactate dehydrogenase from Escherichia coli B. J Biol Chem 243:2579. Tarmy EM, Kaplan NO. 1968b. Kinetics of Escherichia coli B D-lactate dehydrogenase and evidence for pyruvate controlled change in conformation. J Biol Chem 243:2587. Tempest DW, Neijssel OM. 1987. Growth yi eld and energy distribution. In: Neidhardt FC, editor. Escherichia coli and Salmonella typhimurium. Washington, DC: American Society for Microbiology pp. 797 Thomas SM, DiCosimo R, Nagarajan V. 2002. Bio catalysis: applications and potentials for the chemical industry. Trends in Biotechnol 20:238. Thomason L, Court DL, Bubunenko M, Constantino N, Wilson H, Datta S, Oppenheim A. 2005. Current Protocols in Molecula r Biology In Ausubel FM, Brent R, Klingston RE, Moore DD, Deidman JD, Smith JA, and Struhl K, editors. Re combineering: Genetic engineering in bacteria using homologous recomb ination. New York: John Wiley & Sons. p 1.16.1-1.16.21. Tokiwa Y, Suzuki T. 1977. Hydrolysis of polyesters by lipases. Nature 270:76-78. Umare SS, Chandure AS, Padney RA. 2007. Synthe sis, characterization and biodegradable studies of 1,3-propanediol based poly esters. Polym Degrad Stab 92:464-479.
136 Unden G, Achebach S, Holighaus G, Tran HG, Wackwitz B, Zeuner Y. 2002. Control of FNR function of Escherichia coli by O2 and reducing conditions. J Mol Microbiol Biotechnol. 4(3):263-268. Unden G, Kleefeld A. 2004. Chapter 3.4.5, C4-Dicarboxylate Degradation in Aerobic and Anaerobic Growth. In A. Bck, R. Curtiss III, J. B. Kaper, F. C. Neidhardt, T. Nystrm, K. E. Rudd, and C. L. Squires (ed.), EcoSalEscherichia coli and Salmonella : cellular and molecular biology. [Online.] http://www.ecosal. org. ASM Press, Washington, D.C. Underwood SA, Buszko ML, Shanmugam KT, Ingram LO. 2002. Genetic changes to optimize carbon partitioning between ethanol and biosynthesis in ethanologenic Escherichia coli. Appl Environ Microb iol 68(12):6263-6272. Underwood SA, Buszko ML, Shanmugam KT, I ngram LO. 2004. Lack of protective osmolytes limits final cell density and volumetric productivity of ethanologenic Escherichia coli KO11 during xylose fermentation. Appl Environ Microbiol 70(5): 2734-2740. Velick, SF, Furfine C. 1963. Glyceraldehyde 3-phosphate dehydrogenase; in Lardy H, Myrbach K. (eds): The Enzymes. New York Academic Press, vol. 7, pp. 243-273. Vallino JJ, Stephanopoulos G. 1993. Metabolic flux distributions in Corynebacterium glutamicum during growth and lysine overpr oduction. Biotechnol Bioeng 41:633-646. Van der Werf MJ, Guettler MV, Jain MK, Zei kus JG. 1997. Environmental and Physiological factors affecting the succinate product ratio during carbohydrate fermentation by Actinobacillus sp. 130Z. Arch Microbiol 167:332-342. Vemuri GN, Eiteman MA, Altman E. 2002a. Effects of growth mode and pyruvate carboxylase on succinic acid production by metabolically engineered strains of Escherichia coli Appl Environ Microbiol 68:1715-1727. Vemuri GN, Eiteman MA, Altman, E. 2002b. Succinate production in dual-phase Escherichia coli fermentations depends on the time of transition from aerobic to anaerobic conditions. J Ind Microbiol Biotechnol 28:325-332. Vemuri GN, Minning TA, Altman E, Eiteman MA. 2005. Physiological responses of central metabolism in Escherichia coli to deletion of pyruvate ox idase and introduction of heterologous pyruvate carboxylase Biotechnol Bioe ng 90(1):64-76. Vemuri GN, Altman E, Sangurdekar DP, Khodursky AB, Eiteman MA. 2006. Overflow metabolism in Escherichia coli during steady-state growth: transcriptional regulation and effect of the redox. Appl E nviron Microbiol 72(5):3653-3661.
137 Wang J, Tan H, Zhaou ZK. 2006. Over-expression, purification, and characterization of Recombinant NAD-malic enzyme from Escherichia coli K12. Protein Expr Purif 53(1):97-103. Weimer PJ. 1993. Effects of dilution rate a nd pH on the ruminal cellulolytic bacterium Fibrobacter succinogenes S85 in cellulose fed continuous culture. Arch Microbiol 160: 288-294. Wendisch VF, Bott M, and Bernhard JE. 2006. Metabolic engineering of Escherichia coli and Corynebacterium glutamicum for biotechnological production of organic acids and amino acids. Current Opinion in Microbiol 9:1-7. Werpy T, Petersen G, editors. 2004. Top Value Added Chemicals From Biomass. US Department of Energy. http:// www1.eere.energy.gov/biomass/pdfs/35523.pdf Willke T, Vorlop KD. 2004. Industrial bioconversion of renewable resources as an alternative to conventional chemistry. Appl Microbiol Biotechnol 66:131-142. Wood BE, Yomano LP, York SW, Ingram LO. 2005. Development of industrial mediumrequired elimination of the 2,3-butanediol fe rmentation pathway to maintain ethanol yield in an ethanologenic strain of Klebsiella oxytoca Biotechnol Prog 21:1366-1372. Wood WA. 1961. Fermentation of carbohydrates and related co mpounds. In: Gunsalus IC, Stainer RY (eds). The bacteria, vol 2. Academic Press, New York London, pp 59-149. Wollenberg RH, Frank P. 1988. Modified succinimides in fuel composition. US patent 4,767,850. Wu H, Li ZM, Zhou L, Ye Q. 2007. Improved su ccinic acid production in an aerobic culture of a pflB ldhA double mutant of Escherichia coli by enhancing the anaplerotic activities in preceding aerobic culture. Appl E nviron Microbiol 73(24):7837-7843. Yun NR, San KY, Bennett GN. 2005. Enhancemen t of lactate and succinate formation in adhE or pta-ackA mutants of NADH dehydrogenase-deficient Escherichia coli J Appl Microbiol 99:1404-1412. Zeikus JG, Jain MK, Elankovan P. 1999. Bi otechnology of succinic acid production and markets for derived industrial products Appl Microbiol Bi otechnol 51:545-552. Zhang X, Jantama K, Moore JC, Shanmugam KT, Ingram LO. 2007. Production of L-alanine by metabolically engineered Escherichia coli Appl Microbiol Bi otechnol 77: 355-366. Zhou S, Shanmugam KT, Ingram LO. 2003. Functional replacement of the Escherichia coli D-()-lactate dehydrogenase gene ( ldhA ) with the L-(+)-lactate dehydrogenase gene ( ldhL ) from Pediococcus acidilactici. Appl Environ Micr obiol 69:2237-2244.
138 Zhou S, Grabar TB, Shanmugam KT, Ingram LO. 2006a. Betaine tripled the volumetric productivity of D-(-)-lactate by Escherichia coli strain SZ132 in mineral salts medium. Biotechnol Lett 28:671-676. Zhou S, Shanmugam KT, Yomano LP, Grabar TB, Ingram LO. 2006b. Fermentation of 12% (w/v) glucose to 1.2 M lactate by Escherichia coli strain SZ194 using mineral salts medium. Biotechnol Lett 28:663-670. Zhou S., Iverson AG, Grayburn WS. 2008. Engin eering a native homoethanol pathway in Escherichia coli B for ethanol production. Biotechnol Lett 30:335-342. Zwicker N, Theobald U, Zahner H, Fiedler HP. 1997. Optimization of fermentation conditions for the production of ethylene-diamine-disuccinic acid by Amycolatopsis orientalis J Ind Microbiol Biotechnol 19:280-285.
140 BIOGRAPHICAL SKETCH Kae mwich Jantama was born in Chiang Mai, Thailand in August 1978. He attended Montfort College, Chiang Mai, Thailand until he finished his high school. He matriculated at Chiang Mai University, Thailand for his underg raduate study in food science and technology major. He finished Bachelor of Science with fi rst class honor and received the best study award throughout his undergraduate program in March 1999. Since May 1999, he began his graduate study at Institute of Molecular Biology and Genetics, Mahidol University, Salaya Campus, Thailand. During his study, he was awarded the graduate fellowship from the National Scienc e and Technology Developm ent Agency (NSTDA), Ministry of Science, Technology and Environment, Thailand. He finished his Master of Science (Molecular Genetics and Genetic Engineering) in January 2002. In the year 2002, he was aw arded the Royal Thai Governme nt Scholarship to pursue his Ph.D. study abroad. He started his graduate study at University of Florida since August 2002 and finished Master of Engineering (Chemical E ngineering) in August 2004. After which, he continued his Doctor of Philosophy program in the Chemical Engineering Department under supervision of Professor Spyros A Svoronos. He joined his research work with Distinguished Professor Lonnie O Ingram at Microbiology a nd Cell Science Department since October 2004. He graduated his Doctor of Ph ilosophy degree in August 2008. Once he graduates from University of Florida, he will serve as a faculty member at Biotechnology Department, Instit ute of Agricultural Technology, Suranaree University of Technology, Nakorn Ratchasima, Thailand.