a I. . . .u. £15.23”: 1.1.. 1.... a. raw“? 1. n... i:«...r..«uul3 a I!!! 5' uv”.§¥‘> L. in (32...? 21.12...“ 1... H... .r $3 UBRARY ' MIC: "Ban State “i University fi This is to certify that the dissertation entitled STRATEGIES FOR IMPROVING SYNTHESIS OF QUINIC ACID AND SHIKIMIC ACID FROM D-GLUCOSE presented by JUSTAS JANCAUSKAS has been accepted towards fulfillment of the requirements for the PhD. degree in Chemistry 5% (W- Lu / II/lajor Professor’s Signature I"qu~c'I\ Z L/J 200? Date MSU is an affirmative-action. equal-opportunity employer STRATEGIES FOR IMPROVING SYNTHESIS OF QUINIC ACID AND SHIKIMIC ACID FROM D—GLUCOSE By Justas Jancauskas A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 2008 ABSTRACT STRATEGIES FOR IMPROVING SYNTHESIS OF QUINIC ACID AND SHIKIMIC ACID FROM D-GLUCOSE By Justas Jancauskas Optimization of microbial synthesis of quinic acid by recombinant E. coli strains was performed. Overexpression of E. coli shikimate/quinate dehydrogenase YdiB resulted in lower quinic acid titer and yield under fed-batch conditions than when AroE was overexpressed. Increased levels of 3-dehydroquinic acid were also observed, which indicated that YdiB was not able to reduce 3-dehydroquinic acid at a sufficiently rapid rate. The use of E. coli B as a host for quinic acid production rather than E. coli K-12 was examined and revealed that overexpression of tktA-encoded transketolase was not necessary in E. coli B, since it synthesized the same concentration of quinic acid with or without plasmid-localized WA. A decline in synthesis of quinic acid was observed as compared to the standard E. coil K-IZ QPl.I/pKD12.l38. The total synthesized hydroaromatics by the E. coli B producer harboring plasmid-localized tktA was 46 g/L while E. coli B without transketolase overexpression produced 65 g/L. Respectively, E. coli K-12 produced 72 g/L and 54 g/L of total hydroaromatics with or without transketolase overexpression. Optimum quinic acid production by E. coli K-IZ QPl .l/pKDl2.l38 under fed-batch glucose-limited culture conditions was achieved with a total cultivation time of 78 h and a reduced phosphate concentration in the medium (35 mM instead of 43 mM). It was also demonstrated that 3-dehydroquinic acid is recaptured by E. coli K-l2 under glucose-limited culture conditions and further converted to quinic acid. However, glucose-rich fermentation conditions have to be avoided, especially during early stages of quinic acid production. A new quinic acid purification method was developed. It included aromatization of 3-dehydroquinic acid by l h reflux of cell-free and protein-free culture medium, followed by filtration through activated carbon. Salt precipitation was performed with 3 volumes of 100% ethanol. Quinic acid was recrystallized from salt—free ethanol with an 85% yield. Quinic acid is a byproduct and contaminant during shikimic acid biosynthesis by E. coli SPl.l/pKD12.I38. Historically, glucose-rich conditions were used to limit quinic acid levels in the culture medium. A new variant, 112.2, of E. coli SPI .l was constructed by disrupting the genomic shikimate/quinate dehydrogenase ydiB gene. “2.2 showed no quinic acid accumulation during shikimic acid biosynthesis under fed-batch glucose- limited or glucose-rich culture conditions. E. coli 112.2/pKD12J38 produced 49 g/L of shikimic acid in 20% yield while SPl.l/pKD12.l38 synthesized 30 g/L in 13% yield under glucose-limited culture conditions. However, SPl.l/pKD12.l38 performed better under glucose-rich culture conditions and synthesized 60 g/L of shikimic acid in 26% yield, while .I.I2.2/pKD12.I38 produced SI g/L in 20% yield, respectively. Another variant of SPI.l was constructed where shikimate kinases AroK and AroL were inactivated using a scarless method rather than by transposon insertion as was done for SP] .1. Preliminary results showed that more carbon flowed into the shikimate pathway. An attempt to identify a hydroaromatics efflux system in E. coli was unsuccessful. Attempts were made to identify a shikimate dehydrogenase gene from Gluconobacter oxydans [F0 3244 and to overexpress it in E. coli. Copyright by Justas Jancauskas 2008 To my parents For their constant love and support ACKNOWLEDGMENTS It is with the great pleasure that I take the opportunity to thank all the people that I am indebted to for their help. First and foremost, I would like to thank my advisor Prof. John W. Frost for allowing me to take part in his research. His passion about scientific research, aggressiveness and strong work ethic has certainly been inspiring. I am indebted to him for his guidance and encouragements throughout my graduate studies as well as for proofreading this thesis. I would also like to thank the members of my graduate committee, Prof. Babak Borhan, Prof. Robert E. Maleczka, Jr. and Prof. David P. Weliky for their input during the preparation of thesis. I am also sincerely greatful to my undergradute advisor Prof. Romualdas Sablinskas at Kaunas University of Technology, Lithuania for his pation in biotechnology that has influenced me. I am grateful to Dr. Karen M. Draths for her friendship, patient guidance, technique guidance, invaluable suggestions and endless support in my research. I am also thankful to Carolyn Wemple for help in all the encountered bureaucratic issues during my stay at MSU. My gratitude is given to the Frost group members, past and present, including Brad Cox, Kin Sing Stephen Lee, Dr. Dongming Xie, Dr. Stephen Cheley, Dr. Jiantao Guo, Dr. Wei Niu, Dr. Wensheng Li, , Dr. Heather Stuben and Adam Bush. I am especially grateful to Dr. Man Kit Lau, Dr. Mapitso N. Molefe, Dr. Jihane Achkar, Xiaofei .Iia, Dr. Jinsong Yang for their much needed support throughout my time in the group and whose friendships that I will treasure forever. I wish to thank Dr. Ningqing Ran and Dr. Jain Yi for their discussions and guidance in experimental techniques. vi Finally, I owe the sinceriest of thanks to my dear friends Dr. Konstantinos Rampalakos, Dr. Efthalia Kalliri, Dr. N ikoleta Theodoropoulou, Dr. Chrysoula Vasilieou, Kyoungsoo Lee and his family, Diana Gaidelyte and David Riess, Algirdas Baranauskas, Arunas Gaidelis, and Fei Yam for their understanding, help and support. I am also greatfull to all of my aunts, uncles cousins and their familys for so much needed support and understanding. This thesis is dedicated to my parents, Alberta-Genovaite Jancauskiene and Zigmantas Jancauskas, my brother, Mindaugas Jancauskas and my sister-in-law Aukse Jancauskiene, my niece Agne Jancauskaite and my nephew Paulius Jancauskas. vii TABLE OF CONTENTS LIST OF TABLES ............................................................................................................. xi LIST OF FIGURES .......................................................................................................... xiii LIST OF ABBREVIATIONS .......................................................................................... xvi CHAPTER ONE ................................................................................................................. I Introduction ..................................................................................................................... I The shikimate pathway ................................................................................................ 3 3-Dehydroshikimic acid .............................................................................................. 5 Shikimic acid ............................................................................................................... 7 Quinic acid .................................................................................................................. 9 Increasing of carbon flow into the shikimate pathway ............................................. 12 References ..................................................................................................................... 18 CHAPTER TWO .............................................................................................................. 26 Optimizing microbial synthesis of (-)-quinic acid ........................................................ 26 Introduction ............................................................................................................... 26 Biocatalytic synthesis of quinic acid by fed-batch fermentation .............................. 31 A search for a better shikimate/quinate dehydrogenase ............................................ 35 E. coli B as a quinic acid producer ............................................................................ 48 Optimization of quinic acid production by E. coli QPI .l/pKD12.138 ..................... 64 Quinic acid purification ............................................................................................. 83 Discussion ................................................................................................................. 87 References ..................................................................................................................... 95 CHAPTER THREE ........................................................................................................... 99 A search for a better shikimic acid producer .................................... 99 Introduction ............................................................................................................... 99 Fermentation conditions .......................................................................................... 103 Inactivation of ydiB ................................................................................................. 104 A new SP1 .I variant ................................................................................................ 113 Shikimate dehydrogenase from Gluconobacter oxydans ........................................ 1 16 An attempt to identify hydroaromatics transport system in E. coli ......................... 119 Discussion ............................................................................................................... 129 References ................................................................................................................... 132 CHAPTER FOUR ........................................................................................................... 135 Experimental ............................................................................................................... 135 General methods .......................................................................................................... 135 Spectroscopic measurements ................................................................................... I35 Chromatography ...................................................................................................... 135 Bacteria strains and plasmids ...................................................................................... 136 Storage of bacterial strains and plasmids ................................................................ 139 Culture medium ....................................................................................................... 139 Fed-batch fermentation (general) ............................................................................ I40 Glucose-rich fermentor conditions .......................................................................... 141 viii Glucose-limited fermentor conditions ..................................................................... 142 Analysis of fermentation broths .............................................................................. 142 Genetic manipulations ................................................................................................. 144 General ........................................................................................................................ 144 Large scale purification of plasmid DNA ............................................................... 145 Small scale purification of plasmid DNA ............................................................... 145 Determination of DNA concentration ..................................................................... 146 DNA precipitation ................................................................................................... 146 Restriction enzyme digestion of DNA .................................................................... 147 Agarose gel electrophoresis .................................................................................... 147 Isolation of DNA from agarose ............................................................................... 148 Treatment of vector DNA with calf intestinal alkaline phosphatase (CIAP) .......... 148 Treatment of DNA with Klenow fragment ............................................................. 149 Ligation of DNA ..................................................................................................... I49 Preparation and transformation of competent cells ................................................. 150 Purification of genomic DNA ................................................................................. 152 PI-mediated transduction ........................................................................................ 153 Enzyme assays ......................................................................................................... 154 Shikimate dehydrogenase assay in forward direction ............................................. 154 Shikimate dehydrogenase assay in reverse direction .............................................. 155 Quinate dehydrogenase assay .................................................................................. 156 Chapter two (experimental) ......................................................................................... 156 E. coli W31 10 AaroD(new)::FRT-cat-FRT ............................................................ 156 E. coli B serAzzaroB AaroD(new)::FRT-cat-FRT ................................................... 157 E. colI' B serAzzaroB AaroD(new)::FRT ................................................................. 157 Plasmid p.l.IS.O67 ..................................................................................................... 158 Plasmid pJJS.068 ..................................................................................................... 159 Plasmid pJJS.069 ..................................................................................................... 159 Plasmid pJJS.072 ..................................................................................................... 159 Plasmid pJJS .073 ..................................................................................................... 160 Chapter three (experimental) ....................................................................................... 160 E. colI' BW251 l3 AserAzzFRT-kan-FRT ................................................................. I60 E. colI’ BW251 l3 AserA: :FRT ................................................................................. 161 E. colI' BW25113 AydI'B: :FRT- kan- FRT ................................................................. 161 E. coli .IJZkarI ........................................................................................................... 162 E. colI' JJZ ................................................................................................................ 163 E. coli BW251 13 AydI'B(HI, H2.2)::FRT—kan-FRT ............................................... 163 E. coli 112.2kan ........................................................................................................ I64 E. coli 112.2 ............................................................................................................. I64 E. colI' .I.I3cat ........................................................................................................... I65 E. coli “3 ................................................................................................................ 166 E. colI' JJ4cat ........................................................................................................... 166 E. coli .IJ4 ................................................................................................................ I67 JJScat ....................................................................................................................... I67 115 ............................................................................................................................ I68 Genomic DNA library of Gluconobacter oxydans IFO 3244 ................................. 169 ix Plasmid pJJSJSl ..................................................................................................... I69 Plasmid pJJS.l64 ..................................................................................................... I69 Plasmid pJJS.l65 ..................................................................................................... 170 References ................................................................................................................... 171 LIST OF TABLES Table I. Comparison of the impact of modification of the central metabolism in E. coli on synthesis of 3-dehydroshikimic acid and shikimic acid. ................. 13 Table 2. Comparison of the impact of modification of the central metabolism in E. can on synthesis of 3-dehydroshikimic acid and quinic acid. ..................... 27 Table 3. Shikimate dehydrogenase AroE and YdiB kinetic parameters for 3—dehydroshikimic acid. ................................................................................... 35 Table 4. Concentrations and yields of products synthesized by quinic acid producing strains with plasmid-overexpression of ydI'B or aroE. .................... 37 Table 5. YdiB specific acitivities. .................................................................................... 47 Table 6. Screening for E. coli mutant phenotype. ............................................................ 52 Table 7. Concentrations and yields of products synthesized by quinic acid producing E. colI' B strains under glucose-limited conditions. ........................ 55 Table 8. Concentrations and yields of products synthesized by quinic acid producing E. colI' QPI .l/pKD12.138 during various length of fermentation. .................................................................................................... 64 Table 9. Concentrations and yields of products synthesized by quinic acid producing E. coli K-12 strain with and without transketolase overexpression .................................................................................................. 68 Table 10. Concentrations and yields of products synthesized by quinic acid producing strain QPI .I/pKD12.138 with various phosphate concentration in the medium. ........................................................................... 69 Table l 1. Concentrations and yields of products synthesized by quinic acid producing strain QPI.1/pKD12.138 with various aromatic amino acid concentrations in the medium. ......................................................................... 73 Table 12. Concentrations and yields of products synthesized by quinic acid producing strain QPI .l/pKD12.138 under various glucose fermentation conditions. ........................................................................................................ 80 Table 13. Quinic acid purification from microbial culture medium. ............................... 86 Table 14. Quinic acid elemental analysis. ........................................................................ 87 Table 15. Shikimic acid and 3-dehydroshikimic acid molar ratios, shikimic acid yield and total hydroaromatic yield produced by recombinant E. coli under fermentor-controlled, glucose-rich conditions. .................................... 102 xi Table 16. Table 17. Table 18. Table 19. Table 20. Table 21. Concentrations and yields of products synthesized by shikimic acid producing E. colI' with ydI'B mutation. ........................................................... 106 Concentrations and yields of products synthesized by shikimic acid producing E. colI' with ydI'B mutation. ........................................................... 1 l4 Shikimate dehydrogenase Km values. ............................................................. l 17 Shikimate dehydrogenase activity .................................................................. I I9 Concentrations and yields of products synthesized by shikimic acid and quinic acid producing E. colI' with ydI'N overexpression ................................ 127 Bacterial strains and plasmids. ....................................................................... 137 xii LIST OF FIGURES Figure l. The shikimate pathway and biosynthesis of quinic acid and gallic acid. ........... 4 Figure 2. Value-added chemicals synthesized from glucose via 3- dehydroshikimic acid intermediacy. .................................................................. 6 Figure 3. Use of shikimic acid in synthetic chemistry. ...................................................... 8 Figure 4. Biologically active quinic acid derivatives. The portions from quinic acid are indicated in bold. ................................................................................ 10 Figure 5. Glycolysis and biogenesis of D-erythrose 4—phosphate and phosphoenolpyruvic acid. ................................................................................ 15 Figure 6. Plasmid maps of pKD12.112, pKD12.l38, pNR4.230 and pJJ4.l71. .............. 34 Figure 7. (A) E. coli QPI .l/pKD12.I38 and (B) QPI .l/pNR4.230 cultured under glucose-limited conditions. .............................................................................. 36 Figure 8. (A) E. coli QPl .l/pJJ4.171 cultured under glucose-limited culture conditions and (B) QP1.l/p.l.l4.l71 cultured under glucose-rich culture conditions. ........................................................................................................ 37 Figure 9. Construction of plasmid pJJS.067 ..................................................................... 39 Figure 10. Construction of plasmid pJJS.068 ................................................................... 40 Figure l 1. Construction of plasmid pJJS.O69 ................................................................... 41 Figure 12. Construction of plasmid pJJS .072 ................................................................... 42 Figure 13. Construction of plasmid pJJS.073 ................................................................... 43 Figure 14. (A) E. coli QPl.l/pJ.IS.O69 cultured under glucose-limited conditions, (B) QPl.l/p.l.15.069 cultured under glucose-rich conditions and (C) QP1.l/pJ.15.073 under glucose-limited conditions. .......................................... 45 A Figure 15. Gene deletion method in E. colI' ...................................................................... 50 Figure 16. Construction of E. coli AaroDzzFRT mutant. ................................................. 53 Figure 17. (A) E. colI' B serA::aroB AaroDzzFRT-cat—FRT/pKDl2.1 12 and (B) E. coli B serA::aroB AaroDzzFRT—cat—FRT/pKDIZ. I 38 cultured under glucose-limited conditions. .............................................................................. 55 Figure 18. Sequencing of the 5’ end of PCR product with VI primer of E. coli B serA::aroB AaroDzzFRT—cat—FRT 2.2 kb PCR product. ................................ 56 xiii Figure 19. Sequencing of the 3’ end of PCR product with V2 primer of E. colI‘ B serA::aroB AaroDzzFRT—cat—FRT 2.2 kb PCR product. ................................ 57 Figure 20. (A) E. colI' B serA::aroB AaroD(new)::FRT—cat—FRT/pKD12.1 12 and (B) E. colI' B serA::aroB AaroD(new)::FRT—cat-FRT/pKD12.I38 cultured under glucose-limited conditions. ...................................................... 60 Figure 21. (A) E. colI' B serA::aroB AaroD(new)::FRT—cat—FRT/pKD12.112 and (B) E. colI' B serA.°.°aroB AaroD(new)::FRT—cat—FRT/pKD12.138 cultured under glucose-limited conditions. ...................................................... 61 Figure 22. (A) E. coli B serAzzaroB AaroD(new)::FRT/pKD12.I 12 and (B) E. coli B serAzzaroB AaroD(new)::FRT/pKD12.138 cultured under glucose-limited conditions. .............................................................................. 63 Figure 23. E. colI' QPl .l/pKD12.l38 cultured under glucose-limited conditions: (A) 60 h and standard inoculation; (B) 60 h and non standard inoculation. ....................................................................................................... 65 Figure 24. E. coli QPI .l/pKD12.138 cultured under glucose-limited conditions: (A)l32 h; (B) 84h. ............................................................................................ 67 Figure 25. E. coli QP1.l/pKD12.l 12 cultured under glucose-limited conditions. .......... 68 Figure 26. E. colI' QP1.l/pKD12.l38 cultured under glucose-limited conditions in: (A) 35 mM KZHPO4; (B) 20 mM Kzl-IPO4 medium .................................... 70 Figure 27. Comparison of the impact of phosphate concentration on E. coli QPl .l/pKD12.138: (A) dry cell weight; (B) quinic acid synthesis. ................ 71 Figure 28. Comparison of the impact of aromatic amino acid concentrations on E. coli QPI .l/pKD12.l38 dry cell weight. ........................................................... 73 Figure 29. E. coli QPI .l/pKD12.l38 cultured under glucose-limited conditions with: (A) twofold increased aromatic amino acid; (B) 15 g/L yeast extract supplementation. .................................................................................. 76 Figure 30. Comparison of the impact of twofold increased aromatic amino acid concentration and 15 g/L yeast extract supplementation in the culture medium on dry cell weight and quinic acid synthesis by E. coli QPl .I/pKD12.l38 under glucose-limited culture conditions. ......................... 77 Figure 31. E. coli QPI .l/pKD12.l38 cultured under glucose-rich conditions for (A) 66 h and (B)36 h and subsequently switched to glucose-limited culture conditions. ............................................................................................ 81 xiv Figure 32. Quinic acid accumulation profiles obtained during cultivation of E. coli QPI .l/pKD12.138 under glucose-rich culture conditions and subsequently switched to glucose-limited culture conditions. ......................... 82 Figure 33. Conversion of 3-dehydroquinic acid to protocatechuic acid. ......................... 85 Figure 34. 'H NMR of purified ISt crop of quinic acid .................................................... 88 Figure 35. I3C NMR of purified lSt crop of quinic acid. ................................................. 89 Figure 36. (A) SP1.l/pl(D12.138 and (C) JJZ/pKD12.l38 cultured under glucose-limited conditions, and (B) SP1.1/pKD12.138 and (D) JJZ/pKD12.l38 cultured under glucose-rich conditions. ............................... 107 Figure 37. E. coli K-12 ydI'B and aroD genomic DNA locus: (A) graphic representation; (B) DNA sequence. ............................................................... l 10 Figure 38. (A).lJ2.2/pKD12.138 cultured under glucose-limited conditions and (B) JJ2.2/pKD12.138 cultured under glucose-rich conditions. ...................... l 12 Figure 39. (A) JJS/pKD12.138 cultured under glucose-limited conditions and (B) .IJ4/pKD12.l38 cultured under glucose-rich conditions. ............................... I 15 Figure 40. Modification of gene disruption method. ..................................................... 121 Figure 41. Construction of pJ15.151. ............................................................................. 124 Figure 42. Construction of pJJS.l64. ............................................................................. 125 Figure 43. Construction of pJJS.165. ............................................................................. 126 Figure 44. (A) SP1 .l/pKD12.I38 and (B) SP1 .l/JJ5.165 cultured under glucose- rich conditions, and (C) QPl .l/pKD12.138 and (D) QPl .1/pJJS.l65 cultured under glucose-limited conditions. .................................................... 127 XV Ac ADP ATP AP ApR bp CIAP Cm CmR cat DAHP DCU DHQ DHS DO DTT E4P EMP FBR FLP F RT LIST OF ABBREVIATIONS acetyl adenosine diphosphate adenosine triphosphate ampicillin ampicillin resistance gene base pair calf intestinal alkaline phosphatase chloramphenicol chloramphenicol resistance gene chloramphenicol resistance gene 3-deoxy-D-arabin0-heptulosonic acid 7-phosphate digital control unit 3-dehydroquinic acid 3-dehydroshikimic acid dissolved oxygen dithiothreitol D-erythrose 4-phosphate Embden-Meyerhof pathway feedback resistant flippase flippase recognition targen hour xvi IPTG Kan KanR kan kb M9 min mL IIL mM IIM NAD NADH NA DP NADPH NMR OD OTR isopropyl B-D-thiogalactopyranoside kanamycin kanamycin resistance gene kanamycin resistance gene kilobase pair turnover number kilogram Michaelis constant Luria broth molar minimal salts minute milliliter microliter millimolar micromolar nicotinamide adenine dinucleotide, oxidized form nicotinamide adenine dinucleotide, reduced form nicotinamide adenine dinucleotide phosphate, oxidized form nicotinamide adenine dinucleotide phosphate, reduced form nuclear magnetic resonance spectroscopy opficaldenshy oxygen transfer rate xvii ORF PEP PID PC R Phe psi PT S QA rpm SA SDS Tc tet TCA Trp TSP UV open reading frame phosphoenolpyruvic acid proportional-integral-derivative polymerase chain reaction L-phenylalanine pounds per square inch phosphotransferase system quinic acid rotations per minute shikimic acid sodium dodecyl sulfate tetracycline tetracycline resistance gene tricarboxylic acid L-tryptophan sodium 3—(trimethylsilyl)propionate-2,2,3,3-d4 L-tyrosine ultraviolet xviii QHAPT ER QN E Introduction Much of the chemical industry is based on technologies where starting materials are derived from fossil fuels. As a nonrenewable natural resource, petroleum has several environmental and geopolitical problems associated with its use. The carcinogenicity of benzene, which is derived from fossil fuels, is additionally problematic.l Interest in using renewable starting materials has grown with the increasing demand for and cost of petroleum.2 In addition to the actual cost of petroleum, the health costsl associated with fossil fuel-derived chemical building blocks such as benzene have to be considered.2'3 Carbohydrates derived from renewable feedstocks such as glucose, xylose and arabinose are abundant and can be used in biotic and abiotic processes to obtain products that are . . . .. 4 currently manufactured by the chemical Industry usmg tradItIonal technology. Fermentation processes have existed for thousands of years to meet various human needs . . . 5 - such as productIon of bread, Wine, beer, Vinegar and cheese. However, the SCIence behind these processes was not understood until French chemist Louis Pasteur discovered “germs”. Discovery of the microorganisms and invention of the Pasteurization process . . . 6 . . 7 laid the foundation for modern biotechnology. Since then the discovery of DNA, . . . . 9 . . exploration of the plasmid,8 development of sequencmg techniques, and application of polymerase chain reaction| (PCR) have revolutionized traditional microbiology and established a new global biotechnology industry. In the future, microbial synthesis alone or combined with chemical synthesis will grow and replace many currently employed chemical syntheses.ll D-Glucose being renewable and the most abundant carbohydrate monomer holds great potential as a starting material for microbial synthesis. Today, the majority of glucose in the US used in microbial synthesis is derived from corn starch processed by wet milling operations. Another major source of glucose is cellulose. However, better technology for cellulose depolymerization must be developed in order to for cellulose to find use as a glucose source.l2 To compete with chemical synthesis, microbial synthesis must produce value- added chemicals in high yield and high concentration. There are two general methods to increase the yield and concentration of desired products: Direct more carbon flow into the targeted biosynthetic pathway and eliminate byproduct formation. Chapter 2 of this dissertation will focus on increasing the yield and concentration of quinic acid synthesized by a recombinant Escherichia coli strain. Investigation of E. coli ydI'B- encoded shikimate/quinate dehydrogenase YdiB as a substitute for currently employed aroE-encoded shikimate/quinate dehydrogenase for quinic acid production will be discussed. The use of E. coli B rather then E. coli K-12 as a host will be presented as well. Optimization of the microbial synthesis of quinic acid using E. coli QP1.l/pKD12.l38 will be discussed in detail and will include the examination of such parameters as fermentation time, phosphate and aromatic amino acid levels in the production medium, together with a new method for purifying quinic acid. Chapter 3 will focus on microbial synthesis of the shikimic acid. Quinic acid is an undesirable byproduct in shikimic acid biosynthesis, which lowers the yield of shikimic acid by channeling carbon flow away from biosynthesis of shikimic acid. 2 Secondly, if quinic acid accumulates at higher levels, it complicates shikimic acid purification.l3 A single genetic modification of shikimic acid producer SP1.1 afforded the first microbial synthesis of shikimic acid where accumulation of quinic acid as byproduct has been completely eliminated. Construction and preliminary investigation of the new shikimic acid producing E. colI' .115 will be discussed as well as the use of heterologous shikimate dehydrogenase from Gluconobacter oxydans IFO3244. Hydroaromatics transport in E. colI' was also investigated as an avenue to increase yields and concentrations of microbially synthesized shikimic acid or quinic acid. The shikimate pathway The shikimate pathway exists in plants, bacteria and fungi, and is involved in the transformation of carbohydrates into aromatics including the amino acids L-tyrosine, L- phenylalanine and L-tryptophan, and vitamin precursors such as 2, 3-dihydroxybenzoate, p-hydroxybenzoate and p-aminobenzoate.l4 Mammals are incapable of de novo biosynthesis of aromatics and must acquire these metabolites from their diet. The first step of the shikimate pathway is the irreversible condensation of phosphoenolpyruvate (PEP) and D-erythrose 4-phosphate (E4P) to form 3-deoxy-D- arabino-heptulosonic acid 7-phosphate (DAHP), which is catalyzed by DAHP synthase (Figure l). E. coli has three DAHP isozymes AroF, AroG and AroH, which are feedback inhibited, respectively, by L-tyrosine, L-phenylalanine and l.—tryptophan.l5 DAHP is 6 fuIther converted to 3-dehydroquininc acid (DHQ) by aroB-encoded DHQ synthase.l A syn elimination of water from DHQ is catalyzed by 3-dehydroquinate dehydrogenase (AroD) and produces 3-dehydroshikimic acid.l7 Reduction of 3-dehydroshikimic acid by NADPH affords shikimic acid. OH OH “6:0 6H quinic aoId COzi-l H0 : ' 0H gallic acid AroE Z 9 ——-> L-phenylalanine L-tryptophan OH L-tyrosine folic acid 0 2H “9 002” N03 Q ——> H _ OH OH 3-dehydroquinic acid lAroD COZH 6H 3—dehydroshilomic acid ——-> “90 0 Lo H203PO OH 0100214 0 mid 7-phosphate 002H H0"©‘0H 0H shikimic acid 5:“ 6H chonsmic and 3-deoxy-o-arabino-heptulosonic N00 —-> AroF AroG AroH 9 ° g 0315?; '§ 3;. 5 E ‘3 I -<0 22 5’2 t In; 80"53—LQ' «133 2 g g; ’2 .2 3. Eff (9 a E E 3' 3 g 3“ ” 3»: Figure 1. The shikimate pathway and biosynthesis of quinic acid and gallic acid. Enzymes: DAHP synthase (AroF, AroG, AroH); 3-dehydroquinate synthase (AroB); 3- dehydroquinate dehydratase (AroD); shikimate dehydrogenase (AroE); shikimate kinase (AroK, AroL); S-enolpyruvoylshikimate 3-phosphate synthase (AroA); chorismate synthase (AroC); unknown(?). This reduction is catalyzed by aroE-encoded shikimate dehydrogenase,l8 which is feedback inhibited by shikimic acid.'9 A second putative shikimate/quinate dehydrogenase isozyme YdiB was recently identified in E. coli.20 The next step requires ATP and is catalyzed by two shikimate kinase isozymes, AroK and AroL.2| Phosphorylated shikimic acid is a substrate for 5-enolpyruvoylshikimate 3-phosphate synthase encoded by the aroA gene, which catalyzes enolpyruvoyl addition to the C-5 hydroxy using PEP as a substrate.22 The last enzyme of the shikimate pathway is chorismate synthase (AroC).23 lt catalyzes 1,4-trans elimination of phosphate to form chorismic acid. Chorismic acid is the branching point for biosynthesis of aromatics. All three aromatic amino acids and aromatic vitamins are synthesized from this chorismic acid. Regulation of the shikimate pathway in E. coli occurs at the protein level via feedback inhibition and at the transcriptional level. The transcription of L-tyrosine- sensitive AroF and L-phenylalanine-sensitive AroG isozymes is regulated by a . . l_ . . transcrIptIonal repressor encoded by the tyrR gene. 5 TranscrIptIon of l.-tryptophan- sensitive AroH is regulated by the trpR gene product.” The same transcription regulators control shikimate kinase aroL transcription, while aroK is constitutively . . 15 expressed In E. colI. 3-Dehydroshikimic acid 3-Dehydroshikimic acid is an intermediate in the shikimate pathway and can be viewed as a branch point from which natural products can be biosynthesized without intermediacy of chorismic acid (Figure 2). The aromatization of 3-dehydroshikimic acid to protocatechuic acid is crucial to the value of 3-dehydroshikimic acid as an intermediate for the synthesis of several commodity and fine chemicals. Protocatechuic acid is a . . . . 24 . . . . 25 . . 26 . branch pomt In the bIosyntheSIs of catechol, as, ClS-mUCOHIC aCId, vanillin, gallic acid27 and pyrogallol.27 Hydrogenation of cis-cis-muconic acid affords adipic acid.25 0 H CH3O HO OH vanillin D'glulcoseO COZH COzH H02C\/\/\ C O H H0©h OQOH adipic acid protocatechuic acid DCl)-IS l I \ COZH CooZH OH HO OH CIS, Gig-Educonic catechol OH gallic acid pyrogallol Figure 2. Value-added chemicals synthesized from glucose via 3-dehydroshikimic acid intermediacy. The Frost group developed E. colI' KL3/pJYl.2l6A as a construct capable of converting glucose into 3-dehydroshikimic acid.28 This 3-dehydroshikimic acid- synthesizing biocatalyst was constructed by disruption of the genomic aroE locus (Figure l) in E. colI', which resulted in accumulation of 3-dehydroshikimic acid in the culture medium. Integration of a second aroB copy in the genome and plasmid-localized 6 FBR expression of feedback-insensitive, aroF -encoded DAHP synthase, tktA-encoded transketolase and ppsA-encoded PEP synthase increased the flow of carbon into the shikimate pathway and produced 69 g/L of 3-dehydroshikimic acid in 35% yield. Shikimic acid It has been demonstrated that shikimic acid is a useful chiral compound. It is a six-membered carboxylic ring with three stereogenic centers, and an endocyclic olefin. Schreiber and coworkers have exploited it as the core scaffold for a large combinatorial library synthesis (Figure 3).29 Frost and coworkers reported that shikimic acid also . . 3 . . serves as an enVIronmentally-friendly precursor for phenol 0 and p-hydroxybenzmc aCId (Figure 3).3| The importance of shikimic acid increased with its use as the starting material for the manufacture of Tamiflu® (Figure 3), which is a potent inhibitor for Influenza A and Influenza B neuraminidases.32 Tamiflu was approved by the FDA and is currently marketed by Roche as an orally administered antiinfluenza drug. Use of shikimic acid in Tamiflu synthesis and other synthetic applications was previously restricted by shikimic acid price and availability. Before a microbial synthesis of shikimic acid was developed by the Frost group, the only source of shikimic acid was . . . . . . . 33 IsolatIon from the pericarps of Chinese star anIse (lllICIum sp. plants). COQH OH p-hydroxybenzoic acid 0 O O R §_N o I} COZH \/ \ H. ' " \' <— ' . —> w, o 0' = Ho“ . OH NH2 - o“' - "I- ; = H O “0 "5 0” 7 ”NY 0 shikimic acid Tamiflu I C CH phenol Figure 3. Use of shikimic acid in synthetic chemistry. Genetically engineered E. colI' SPl.l/pKD12.l38 was constructed for microbial syntheSIs of shIkImIc aCId. 4 ThIs biocatalyst was created by disrupting the shIkImate kinase-encoded aroK and aroL genomic loci in E. colI'. Also integration of a second aroB copy in the genome and overexpression of plasmid-localized feedback insensitive aroFFBR-encoded DAHP synthase and tktA-encoded transketolase permitted an increase of carbon flow into the shikimate pathway, which ultimately translated into 52 g/L concentration of shikimic acid synthesized in 18% yield from glucose.34 The SPI.1/pI(D12.I38 fermentation process has been scaled up and is licensed by Roche to provide a source of shikimic acid for the manufacture of Tamiflu. Quinic acid Quinic acid was first isolated from crude quinine (an anti-malaria drug isolated from Cine/Iona bark) in 1790.35 The structure and stereochemistry of quinic acid was aSSIgned only In 1932 by Fisher and Dangschat. 6 Qumic and is an attractive molecule for combinatorial library or natural product synthesis due to its hi ghly-functionalized six- . . . 37 . . . membered carbocyclic ring and four asymmetric centers. the a few biologically active natural products and natural product derivatives have been synthesized in whole, or in part, from quinic acid. The molecules include anti-influenza drug GS—4104~02 marketed by Roche as Tamiflu,38 (-)-sugiresinol dimethyl ether (a derivative of (-)- sugiresinol isolated from Cryptomeria japom'ca),39 the epoxycyclohexenone core of scyphostatin (a powerful inhibitor of neutral sphongomyelinase),40 (+)—eutypoxide B (a secondary metabolite of fungus Eutypa lata responsible for pathogenic vineyard die-back disease)“, the A ring of la,25-dihydroxyvitamin D3 derivatives (potential drugs for treatment of osteoporosis and psoriasis),42 the 2-iodocyclohexenone acetal portion of anticancer drug taxol,43 and the bicyclic core structure of the potent enediyne antitumoral agent esperimicin-A,“ (Figure 4). COgEt ETOT¢WNH ' H3PO4 O NHAc 9.3-4104 (-)-sugiresinol dimethyl ether OH 9“ “‘\Y\0H E '5 ' HN \ \ \ ’ ‘ O o - \ O - = O scyphostatin OH HO. (+)-eutypoxide B 9 O o HO OH O 509" 1 a,25-dihydroxy-19-norvitamin D3 HO 2-iodocyclohexenone acetal O HO HOOC OH HO / O OH O 1,5-dicaffeoquuinc acid bicyclic core of esperamicin-A1 HO HO 0 o O 0 Ho HOOC HOOC fiflm HO OH HO / 0 0” H300 / 0 OH 0 O 1-caffeoyl-5-feruoquuinc acid 1 ,5-0—diferuoquuinc acid Figure 4. Biologically active quinic acid derivatives. The portions from quinic acid are indicated in bold. - IO Recently, dicaffeoquuinic acids have received attention as potent HIV—l integrase inhibitor. Currently AIDS treatment consists of drug cocktails, which inhibit HIV-1 reverse transcriptase and protease. However, rapid mutation of HIV virus results in drug- resistant strains. Since HIV-1 is a retrovirus, it requires the integration of the proviral double-stranded DNA arising from the reverse transcription step into the host chromosome for its efficient replication, maintenance and productive infection. DNA integration is carried out by integrase, which thus constitutes another potential target in anti-retrowral drug deSIgn.4 Robinson and coworkers were the first to report that dicaffeoquuinic acids were potent inhibitors of HIV-1 integrase in tissue culture.46 More recent preclinical studies suggest, that 1,5-dicaffeoquuinic acid and its two metabolites, monomethylated l-caffeoyl-5-feruoquuinic acid and dimethylated 1, 5-0- diferuoquuinic acid (Figure 4), show the highest inhibitory activity towards HIV-l integrase.47 All biologically active natural products previously cited in this chapter had quinic acid as part of their structure. Does quinic acid by itself have any biological activity? In recent studies by Pero and coworkers it has been reported that ammonium salt of quinic acid is the active component in the Uncari'a tomentosa (Cat’s claw) extract C—Med 10069.483 The authors described that quinic acid inhibited transcriptional regulator NF—KB activity, and they postulate that this is the primary basis of the anti-inflammatory character of cat’s claw extracts. The Pero group also reports that in human studies they have observed increased tryptophan and nicotinamide concentrations in subject’s urine after ingestion of water supplemented with quinic acid.48C Additionally, no quinic acid or hippuric acid, a known metabolite derived from quinic acid in humans,49 was observed in l 1 blood serum. It was also shown that primates including humans have the greatest conversion of quinic acid to hippuric acid. This metabolism is apparently dependent on intestinal microflora, because there was no metabolism of quinic acid to hippuric acid when injected intraperitoneally.4921 As a consequence of these observations, Pero and coworkers have postulated that the biological activity of quinic acid in humans could not be assigned directly to quinic acid. Instead, Pero has hypothesized that quinic acid is converted to biologically active metabolites by microorganisms in the human gastro- intestinal tract. The Pero group suggests that biological activity of quinic acid is arising from gastro-intestinal microbes catalyzing the conversion of quinic acid via the tryptophan-quinilinate-nicotinamide-NAD branch of the shikimate pathway. This is hypothesized to account for the observed elevated levels of tryptophan and nicotinamide In urine. Nicotinamide IS a known NF-KB 02' inhibitor and an antIOXIdant.50m'n High doses of nicotinamide have therapeutic value for treatment of lipid profiles,Ode 50i.k . 50f-h . 50' . 501 diabetes, depreSSIon, J HIV, and migranes. Increasing of carbon flow into the shikimate pathway Modifications of central metabolism in E. coli to channel more carbon flow into the shikimate pathway has been the focus of considerable research activity.“ The first strategy employed was overexpression52 of feedback resistant AroF,FBR which helped to overcome feedback Inhibition of DAHP synthase caused by aromatic amino aCldS.) Insensitivity to feedback inhibition by aromatic amino acids increases the in vivo catalytic activity of DAHP synthase. Because increased carbon flow into the shikimate pathway 12 results from increased DAHP synthase activity, increased accumulation of 3- dehydroshikimic acid and DAH level was observed during E. coli ABZ834 cultivation.54 The absence of catalyticalIy-active shikimate dehydrogenase, which catalyzes conversion of 3-dehydroshikimic acid into the shikimic acid (Figure l), in E. coli ABZ834, resulted in the accumulation of 3-dehydroshikimic acid in the culture medium. Accumulation of DAH, which results from DAHP dephosphorylation, lowers 3-dehydroshikimic acid and shikimic acid concentration and yield. An approximately twofold increase in DHQ . . . . . 9 . . . . synthase speCIfic actIVIty lS reqUIredI to eliminate DAH accumulation, which can be accomplished by introducing a second copy of aroB into the genome of E. coli.55 Table 1. Comparison of the impact of modification of the central metabolism in E. coli on synthesis of 3-dehydroshikimic acid and shikimic acid. Desired goduct Entry Constructa Relevant characteristicsb [DHS]C DHS [SA]' SA yield, ,g/L yield, % g’L /° 1 KL3/pKL4.3SB aHUFFIBFt 89m 20 17 - - 2 KL3/pKL4.66i3 amFFBR’ 86m, aroFFBR 39 16 - - 3 KL3/pKD12.291A amFFBR, 5604' Pam: 41 18 - - 4 KL3/pKL5.17A amFFBR’ 56m, Pam}; MA 58 24 - - 5 KL3/pJY1.216A amFFBR’ 89m, Pam]; WA 69 35 - - ppsA 6 SP1 .1/pK012.1 12 amFFBR’ aroE’ 89m - 38 12 7 SP1.1/pKD12.138 amffBR’ 8,05, 89m, WA - - 52 18 8 SP1.1/pi<015.o71 amFFBR’ aroE’ 5904’ MA ppsA - - 66 23 9 SP1.1pts/psce.908 amFFBR aroE’ 3904' MA 71 27 a KL3: A32834 serAzzaroB; SPI.l: RB79I serAzzaroB aroL478::Tn/0 aroK/7zszR. Ptac glf g/k I)aroli‘FBR: feed-back insensitive DAHP synthase; dehydrogenase Pump: promoter locus of E. coli aroF gene; tktA: transketolase; ppsA: phosphoenolpyruvate synthase; glf. glucose facilitator from Z. mobilis; glk: glucose serA: 3-phosphoglycerate kinase. C DHS: 3-dehydroshikimic acid; SA: shikimic acid. d(mol DHS)/(mol glucose consumed). I3 Several different routes were pursued for circumventing transcriptional repression of aroF.le Use of two plasmid-localized copies of aroEFBR resulted in the synthesis of 39 g/L of 3—dehydroshikimic acid in 16% yield (Table 1., entry 2), while one plasmid— Iocalized copy of aroFfBR afforded 20 g/L of 3-dehydroshikimic acid synthesized in 17% yield (Table 1, entry I).Slf A further improvement in 3-dehydroshikimic acid synthesis was achieved once aroFFBR was overexpressed together with unmodified native promoter Pump, which presumably helped to titrate away the cellular supply of TyrR repressor protein. Synthesis of 41 g/L of 3-dehydroshikimic acid in 18% yield was observed (Table 1, entry 3).5|f These results indicated that the best configuration was to include in the plasmid one copy of ParoF and aroFFBR under its native promoter. The same study also demonstrated that higher DAHP synthase activity did not necessarily translate into higher yields or concentrations of shikimate pathway products. This indicated that the substrates for DAHP synthase were limiting factors in increasing carbon flow directed into the shikimate pathway. 14 O O oxaloacetlc acid O 3% +10%“ phosnhoenolpvrwic acid Eno I OH ATP ADP 002 P1 2‘00211 OR2 0H 0 ATP ADP P110 0 *r pyruvic? acid RIF“ I: H203P0 ngI: OH H203P0\/'\',OP03H2 O 1 ,3-diphosphoglycerate N20 NADH PI R1=H; Fizz-Poem 2-phosphoglycerata RI=POaHz; Rz=H GPhosphoolycerato OH O H203Po\/K/u‘ H oI-I D-erythrosa 4—phosphata Gpm ( OH Hzoapow H 0 glyceraldehyde 3-phosphata TktA ——-n> a”: V3" 795 ——-> ‘— ATP ADP glucose 6-phosphate 0 OH OH OH fructose 6-phosphate "RIC 0 o" OH OH fructose 1,6-diphosphata we on Hzoapo Pgil HzoaPO dihydroxyacatone phosphate Fba / o H203P0\/'\/0H Figure 5. Glycolysis and biogenesis of D-erythrose 4-phosphate and phosphoenolpyruvic acid. Enzymes: phosphoenolpyruvatezcarbohydrate phosphotransferase (PTS); glucokinase (le); phosphoglucose isomerase (Pgi); phosphofructose kinase (Pfk); fructose 1,6-diphosphate aldolase (Fba); glyceraldehyde 3- phosphate dehydrogenase (Gap); phosphoglycerate kinase (ng); enolase (Eno); pyruvate kinase (PykA, Pku); transketolase ('TktA). 15 Frost and coworkers published the first work suggesting that D—erythrose 4— phosphate (E4P) availability was an important factor limiting in vivo DAHP synthase activity.56 The reason for low E4P availability might be due to its tendency to . . . 57 . . . . . . polymerize in solution. DISSOCIation back to the monomeric form of E4P is qu1te slow. This may be the reason why E. coli cells keep low steady-state concentrations of E4P. To increase the intracellular concentration of E4P, tktA-encoded transketolase was overexpressed, which led to an increased concentration and yield of 3-dehydroshikimic acid (Table 1, entry 4)5” and shikimic acid (Table 1, entry 7) relative to (Table 1, entry 6) .34 With increased in vivo E4P availability, availability of phosphoenolpyruvate (PEP) became a limiting factor. A number of different cellular processes and enzymes compete with DAHP synthase for PEP including pyruvate kinase PykA and Pku (Figure 5), PEP carboxylase Ppc (Figure 5), and the PEP-dependent carbohydrate:phosphotransferase system (PT S) for transport of glucose and structurally related sugars into the cytoplasm. Efforts to improve intracellular PEP availability began . . . . . .' 5 . .‘9 . With genomic Inactivation of PEP carboxylase”? 8 and pyruvate kinase,5 but did not lead to significantly improved biosyntheses of aromatic amino acids. A better strategy 51d,5|c was reported by Liao and coworkers. Liao used an E. coli aroB mutant (inactive 3-dehydroquinate synthase, Figure l) with plasmid-localized aroGFBR-encoded feedback insensitive DAHP synthase, transketolase and overexpressed PEP synthase. This strategy was based on recycling pyruvic acid, which is generated by P'I‘S-mediated glucose transport, back to PEP. The construction of E. coli strain KL3/pJYl.216A in the Frost 16 group afforded the best 3-dehydroshikimic acid producer constructed so far (Table 1, entry 5).28 Plasmid pJYl.216A carried aroFFBR-encoded feedback-insensitive DAHP synthase, tktA-encoded transketolase, ppsA-encoded PEP synthase and the P aroF ' encoded promoter region of DAHP synthase. The same approach was taken to create a better shikimic acid-producing biocatalyst in SP1.l/pKD15.07lB (Table 1, entry 8).60 Another strategy was to use a non-PTS mechanism for glucose transport into the cell, thereby eliminating consumption of a mole of PEP for each mole of glucose transported into the cell. It was shown, that heterologous expression of Zymomonas mobilis plasmid-localized gif (glucose facilitator protein) and glk (glucose kinase) in a PTS-deficient E. coli strain reconstituted glucose transport and phosphorylation.“ Increase in concentration and yield of microbe-synthesized shikimic acid was observed (Table 1, entry 9) when compared to SP1 .l/pKD15.07IB results (Table 1, entry 8).60 I7 References (a) Wong, 0.; Raabe, G. K. Critical review of cancer epidemiology in petroleum industry employees, with a quantitative meta-analysis by cancer site. Am. J. Ind. Med. 1989, 15, 283—3 10. (b) Lewis, R. J. Carcinogenically active chemicals, Van Nostrand Reinhold, New York, 1991, p. 68. (c) O’Connor, S. R.; Farmer, P. B.; Lauder, I. Benzene and non-hodgkin’s lymphoma. J. Pathol. 1999, [89, 448- 453. (d) Lan Q.; Zhang, L.; Li, G.; Verrneulen, R.; Weinberg, R. S.; Dosemeci, M.; Rappaport, S. M.; Shen, M.; Alter, B. P.; Wu, Y.; Kopp, K.; Waidyanatha, S.; Rabkin, Ch.; Guo, W.; Chanock, S.; Hayes, R. B.; Linet, M.; Kim, S.; Yin, S.; Rothman, N.; Smith, M. T. Hematotoxicity in workers exposed to low levels of benzene. Science, 2004, 306, 1774-1776. (a) Tullo, A. Benzene costs hurt customers. Chem. Eng. News 2004, 82, 15-17. (b) http://www.eia.doe.gov/oiaf/aeo/index.html. Annual energy outlook 2008 with projections to 2030. (a) Yoshida, J .; Inomata, M. Trends in developments of aromatics production technologies. Aromatikkusu 2002, 54, 123-135. (b) Tullo, A. H. A new source. Chem. Eng. News 2003, 81, 16-17. (a) Frost, J. W.; Lievense, J. Prospects for biocatalytic synthesis of aromatics in the let- century. New J. Chem. 1994, 18, 341. (b) Bongaerts, J.; Kramer, M.; Muller, U.; Raeven, L.; Wubbolts, M. Metabolic engineering for microbial production of aromatic amino acids and derived compounds. Metabol. Eng. 2001, 3, 289-300. Pannuri, S.; DiSanto, R.; Kamat, S. In Kirk-Othmer encyclopedia of chemical technology online; Biocatalysis, 2003, Wiley. Junker, B. In Kirk-OIhmer encyclopedia of chemical technology online; Fermentation, 2004, Wiley. Watson, J.; Crick, F. A structure for deoxyribose nucleic acid. Nature 1953, I7 I , 737-738. Cohen, 8.; Chang, A. C. Y.; Boyer, H. W.; Helling, R. B. Construction of biologically functional bacterial plasmids in vitro. Proc. Natl. Acad. Sci. USA. 1973, 70, 3240-3244. (a) Sanger, F.; Coulson, A. R. A rapid method for determining sequences in DNA by primed synthesis with DNA polymerase. J. Mol. Biol. 1975, 94, 441—448. (b) Sanger, F.; Nicklen, S.; Coulson, A. R. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. U .S A. 1977, 74, 5463—5467. 18 10 II 12 I4 Saiki, R. K.; Scharf, 3.; Faloona, F.; Mullis, K. B.; Horn, G. T.; Erlich, H. A.; Amheim, N. Enzymatic amplification of beta-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 1985, 230, 1350-1354. http://www.eere.energv.gov/industry/chemicals/visions biocatalysishtml. Chemical industry of the future. (a) Bozell, J. J .; Landucci, R. Alternate feedstocks program technical and economic assessment; U. S. Department of Energy, Office of Industrial Technologies, 1993. (b) Lynd, L. R.; Cushman, J. H.; Nichols, R. J.; Wyman, C. E. Fuel ethanol from cellulose biomass. Science 1991, 25], 1318-1323. (c) Zaldivar, J .; Nielson, J .; Olsson, L. Fuel ethanol production from lignocellulose: A challenge for metabolic engineering and process integration. Appl. Microbiol. Biotechnol. 2001, 56, 17-34. Draths, K. M.; Knop, D. R.; Frost, K. M. Shikimic acid and quinic acid: replacing isolation from plant sources with recombinant biocatalysis. J. Am Chem. Soc. 1999,/21,1603-1604. (a) Haslam, E. In Shikimic acid: Metabolism and metabolites; Wiley: New York, 1993. (b) Bentley, R. The shikimate pathway - a metabolic tree with many branches. Crit. Rev. Biochem. Mol. Biol. 1990, 25, 307-384. (c) Herrmann, K. M. In Amino acids: Biosynthesis and genetic regulation; Herrmann, K. M., Somerville, R. L., Ed.; Addison-Wesley: Reading, 1983: p. 301. Pittard, A. J. Biosynthesis of the aromatic amino acids. In Escherichia coli and Salmonella typhimurium: cellular and molecular biology. Neidhardt, F. C. ed. ASM Press (Washington, DC: American Society for Microbiology) 1996, p. 458- 484. (a) Frost, J. W.; Bender, J. L.; Kadonaga, J. T.; Knowles, J. R. Dehydroquinate synthase from Escherichia coli: purification, cloning, and construction of overproducers of the enzyme. Biochemistry 1984, 23, 4470-4475. (b) Carpenter, E. P.; Hawkins, A. R.; Frost, J. W.; Brown, K. A. Structure of dehydroquinate synthase reveals an active site capable of multistep catalysis. Nature 1998, 394, 299-302. (a) Smith, B. W.; Turner, M. J.; Haslam, E. Shikimate pathway. 4. Stereochemistry of 3-dehydroquinate dehydratase reaction and observations on 3- dehydroquinate synthetase. J. Chem. Soc., Perkins Trans. 1 1975, I, 52-55. (b) Duncan, K.; Chaudhuri, S.; Campbell, M. S.; Coggins, J. R. The overexpression and complete amino acid sequence of Escherichia coli 3-dehydroquinase. Biochem. J. 1986, 238, 475-483. 19 18 I9 20 21 22 23 24 25 (a) Anton, I. A.; Coggins, J. R. Sequencing and overexpression of the Escherichia coli araE gene encoding shikimate dehydrogenase. Biochem. J. 1988, 249, 319- 326. (b) Chaudhuri, S.; Coggins, J. R. The purification of shikimate dehydrogenase from Escherichia coli. Biochem. J. 1985, 226, 217-223. Dell, K. A.; Frost, J. W. Identification and removal of impediments to biocatalytic synthesis of aromatics from D-glucose: rate-limiting enzymes in the common pathway of aromatic amino acid biosynthesis. J. Am. Chem. Soc. 1993, [15,11581-11589. (a) Michel, G.; Roszak, A. W.; Sauve, V.; Maclean, J.; Matte, A.; Coggins, J. R.; Cygler, M .; Lapthorn, A. J. Structures of shikimate dehydrogenase AroE and its paralog YdiB. J. Biol. Chem. 2003, 278, 19463-19472. (b) Benach, J.; Lee, I.; Edstorm, W.; Kuzin, A. P.; Chiang, Y.; Acton, T. B.; Montelione, G. T.; Hunt, J. F. The 2.3-A crystal structure of the shikimate 5-dehydrogenase orthologue Ydib from Escherichia coli suggests a novel catalytic environment for an NAD- dependent dehydrogenase. J. Biol. Chem. 2003, 278, 19176-19182. (a) DeFeyter, R. C.; Pittard, J. Genetic and molecular analysis of AroL, the gene for shikimate kinase-II in Escherichia coli K-12. J. Bacterial. 1986, I65, 226-232. (b) DeFeyter, R. C.; Pittard, J. Purification and properties of shikimate kinase-II from Escherichia coli K-12. J. Bacterial. 1986, I 65 , 331-333. (c) LObner-Olesen, A.; Marinus, M. G. Identification of the gene (aroK) encoding shikimic acid kinase-I of Escherichia coli. J. Bacteriol. 1992, l 74, 525-529. Duncan, K.; Coggins, J. R. The SerC-AroA operon of Escherichia coli: a mixed function operon encoding enzymes from 2 different amino acid biosynthetic pathways. Biochem. J. 1986, 234, 49-57. (b) Duncan, K.; Lewendon, A.; Coggins. J. R. The purification of 5-enolpyruvylshikimate 3-phosphate synthase from an overproducing strain of Escherichia coli. FEBS Lett. 1984, I65, 121-127. White, P. J.; Millar, G.; Coggins, J. The overexpression, purification and complete amino-acid sequence of chorismate synthase from Escherichia coli K 12 and its comparison with the enzyme from Neurospara crassa. Biochem. J. 1988, 25], 313-322. (a) Draths, K. M.; Frost, J. W. Environmentally compatible synthesis of catechol from D-Glucose. J. Am. Chem. Soc. 1995, 117, 2395-2400. (b) Li, W.; Xie, D.; Frost, J. W. Benzene-free synthesis of catechol: interfacing microbial and chemical catalysis. J. Am. Chem. Soc. 2005, 127, 2874-2882. (a) Draths, K. M.; Frost, J. W. environmentally compatible synthesis of adipic acid from D-Glucose. J. Am. Chem. Soc. 1994, 116,399-400. (b) Niu, W.; Draths, K. M.; Frost, J. W. Benzene-free synthesis of adipic acid. Biotechnol. Prog. 2002, 18,201-211. 20 26 27 28 29 30 31 32 33 34 35 36 37 Li, K.; Frost, J. W. Synthesis of vanillin from glucose. J. Am. Chem. Soc. 1998, [20, 10545-10546. Kambourakis, S.; Draths, K. M.; Frost, J. W. Synthesis of gallic acid and pyrogallol from glucose: replacing natural product isolation with microbial catalysis. J. Am. Chem. Soc. 2000, 122, 9042-9043. Yi, J.; Li, K.; Draths, K. M.; Frost, J. W. Modulation of phosphoenolpyruvate synthase expression increases shikimate pathway product yields in E. coli. Biotechnol. Prog. 2002, 18, 1141-1148. (a) Tan, D. S.; Foley, M. A.; Shair, M. D.; Schreiber, S. L. Stereoselective synthesis of over two million compounds having structural features both reminiscent of natural products and compatible with miniaturized cell-based assays. J. Am. Chem. Soc. 1998, 120, 8565-8566. (b) Tan, D. S.; Foley, M. A.; Stockwell, B. R.; Shair, M. D.; Schreiber, S. L. Synthesis and preliminary evaluation of a library of polycyclic small molecules for use in chemical genetic assays. J. Am. Chem. Soc. 1999, [21, 90073—90087. Gibson, J. M.; Thomas, P. S.; Thomas, J. D.; Barker, J. L.; Chandran, S. 8.; Harrup, M. K.; Draths, K. M.; Frost, J. W. Benzene—free synthesis of phenol. Angew. Chem., lntl. Ed. 2001, 40, 1945—1948. Barker, J. L.; Frost, J. W. Microbial synthesis of p-hydroxybenzoic acid from glucose. Biotechnol. Bioeng. 2001, 76, 376-390. Kim, C. U.; Lew, W.; Williams, M. A.; Liu, H.; Zhang, L.; Swaminathan, S.; Bischofberger, N.; Chen, M. S.; Mendel, D. 3.; Tai, C. Y.; Laver, W. 0.; Stevens, R. C. Influenza neuraminidase inhibitors possessing a novel hydrophobic interaction in the enzyme active site: design, synthesis, and structural analysis of carbocyclic sialic acid analogues with potent anti-influenza activity. J. Am. Chem. Soc. 1997, [[9, 681-690. Weber, W. F. Hoffmann-La Roche, Ltd., personal communication. Knop, D. R.; Draths, K. M.; Chandran, S. S.; Barker, J. L.; von Daeniken, R.; Weber, W.; Frost, J. W. Hydroaromatic equilibration during biosynthesis of shikimic acid. J. Am. Chem. Soc. 2001, 123, 10173-10182. Hofmann, F. C. Crell ’s Chemische Annalen, 1790, 2 314. Fisher, H. O., L.; Dangschat, G. Ber. Dtsch. Chem. Gas. 1932, 65, 1009. (a) Phoon, C. W.; Abell, C. Use of quinic acid as template in solid-phase combinatorial synthesis. J. Comb. Chem. 1999, I, 485-492. (b) Gonzalez C.; Carballido, M.; Castedo, L. Synthesis of polyhydroxycyclohexanes and relatives from (-)- quinic acid. J. Org. Chem. 2003, 68, 2248-2255. (c) Kaila, N.; Somers, 21 38 39 41 42 43 45 W. S.; Thomas, B. E.; Thakker, P.; Janz, K.; DeBernardo, S.; Tam, S.; Moore, W. J.; Yang, R.; Wrona, W.; Bedard, P. W.; Crommie, D.; Keith, J. C.; Jr.; Tsao, D. H. H.; Alvarez, J. C.; Ni, H.; Marchese, E.; Patton, J. T.; Magnani, J. L.; Camphausen, R. T. Quinic acid derivatives as sialyl Lewis-Mimicking selectin inhibitors: Design, synthesis, and crystal structure in complex with E-selectin. J. Med. Chem. 2005, 48, 4346-4357. ((1) Usami, Y.; Ueda, Y. Stereoselective syntheses of diastereomers of antitumor natural product pericosine A from (-)- quinic acid. Synthesis, 2007, 20, 3219-3225. Federspiel, M.; Fischer, R.; Hennig, M.; Mair, H.-J.; Oberhauser, T.; Rimmler, G.; Albiez, T.; Bruhin, J.; Estermann, H.; Gandert, C.; Goeckel, V.; Goetzoe, S.; Hoffmann, U.; Huber; G.; Janatsch, G.; Lauper, S.; Roeckel-Staebler, 0.; Trussardi, R.; Zwahlen, A. G. Industrial synthesis of the key precursor in the synthesis of the anti-influenza drug Oseltamivir phosphate (Ro 64-0796/002, GS- 4104—02): Ethyl (3R ,4S,SS)-4,5-epoxy-3-( 1 -ethyl-propoxy)- I -cyclohexene- I - carboxylate. Org. Proc. Res. Develop. 1999, 3, 266-274. Matsuo, K.; Sugimura, W.; Shimizu, Y.; Nishiwaki, K.; Kuwajima, H. Synthesis of (-)-sugiresinol dimethyl ether utilizing (-)-quinic acid. Heterocycles 2000, 53, 1505-1513. Murray, L. M.; O’Brien, P.; Taylor, R. J. K. Stereoselective reactions of a (—)- quinic acid-derived enone: application to the synthesis of the core of scyphostatin. Org. Lett. 2003, 5, 1943-1946. Barros, M. T.; Maycock, C. D.; Ventura, M. R. Enantioselective total synthesis of (+)-Eutypoxide B. J. Org. Chem. 1997, 62, 3984-3988. Ono, K.; Yoshida, A.; Saito, N.; Fujishima, T.; Honzawa, S.; Suhara, Y.; Kishimoto, S.; Sugiura, T.; Waku, K.; Takayama, H.; Kittaka, A. Efficient synthesis of 2-modified la,25-dihydroxy-l9-norvitamin D3 with Julia olefination: high potency in induction of differentiation on HL-60 cells. J. Org. Chem. 2003, 68, 7407-7415. Su, Z.; Paquette, L. A. Conversion of D-(-)-quinic acid into an enantiopure C-4 functionalized 2-iodocyclohexenone acetal. J. Org. Chem. 1995, 60, 764-766. Ulibarri, G.; Nadler, W.; Skrydstrup, T.; Audrain, H.; Chiaroni, A.; Riche, C.; Grierison, D. S. Construction of the bicyclic core structure of the enediyne antibiotic esperamicin-Al in either enantiomeric form from (-)-quinic acid. J. Org. Chem. 1995, 60, 2753-2761. Tarrago-Livak, L.; Andreola, M. L.; Fournier, M.; Nevinsky, G.A.; Parissi, V.; de Soultrait, V.R.; Litvak, S. Inhibitors of HIV-1 reverse transcriptase and integrase: Classical and emerging therapeutical approaches. Curr. Pharm. Des. 2002, 8, 595-614. 22 47 48 49 50 Robinson, E. W., Jr.; Reinecke, M. G.; Abdel-Malek, S.; Jia, Q.; Chow, S.A. Inhibitors of HIV-1 replication that inhibit HIV integrase. Proc. Natl. Acad. Sci. USA 1996, 93, 6326-6331. (a) Yang, 8.; Meng, Z.; Dong, J.; Yan, L.; Zou, L.; Tang, Z.; Dou, G. Metabolic profile of 1,5-dicaffeoquuinic acid in rats, an in vivo and in vitra study. Drug Metab. Dispos. 2005, 33, 930-936. (b) Asres, K.; Seyoum, A.; Veeresham, C.; Bucar, F.; Gibbons, S. Naturally Derived Anti-HIV Agents. Phytother. Res. 2005, 19, 557—58 1. (a) Akesson, Ch.; Lindgren, H.; Pero, R. W.; Leanderson, T.; Ivars, F. Quinic acid is a biologically active component of the Uncaria tomentosa extract C-Med 100®. Int. lmmunopharmacal. 2005, 5, 219-229. (b) Mammone, Th.; Akesson, Ch.; Gan, D.; Giampapa, V.; Pero, R. W. A water soluble extract from Uncaria tomentosa (cat’s claw) is a potent enhancer of DNA repair in primary organ cultures of human skin. Phytother. Res. 2006, 20, 178—183. (0) Personal communication, manuscript in preparation, 2007. (a) Adamson, R. H.; Bridges, J. W.; Evans, M. E.; Williams, R. T. Species differences in the aromatization of quinic acid in vivo and the role of gut bacteria. Biochem. J. 1970, [[6, 437-433. (b) Indahl, S. R.; Scheline, R. R. Quinic acid aromatization in the rat. Urinary hippuric acid and catechol excretion following the singular or repeated administration of quinic acid. Xenabiotica 1973, 3, 549- 556. (a) Pero, R.W.; Axelsson, B.; Siemann, D.; Chaplin, D.; Dougherty, G. Newly discovered anti-inflammatory properties of the benzamides and nicotinamides. Mol. Cell. Biochem. 1999, 193, 1 19-125. (b) Virag, L. Structure and function of poly (ADP-ribose) polymerase-l: role in oxidative stress-related pathologies. Curr. Vasc. Pharmacol. 2005, 3, 209-214. (c) Adams, J. D. Nicotinamdie and its pharmacological properties for clinic therapy. Drug Des. Rev. Online 2004, I, 43-52. ((1) Handfield-Jones, 8.; Jones, 5.; Peachey, R. High dose nicotinamide in the treatment of necroobiosis lipoidice. Br. J. Dermatol. 1988, I I8, 693-696. (e) Takahashi, Y.; Tanaka, A.; Nakamura, T.; Fukuwatari, T.; Shibata, K.; Shimada, N.; Ebihhara, 1.; Koide, H. Nicotinamide suppresses hyperphosphatemia in hemodiaylsis patients. Kidney Int. 2004, 65, 1099-1104. (f) Rakieten, N.; Gordon, B. S.; Beaty, A.; Cooney, D. A.; Schein, P. 8.; Dixon, R. L. Modification of renal tumorgenic effect of streptozotocin by nicotinamide spontaneous reversability of streptozotocin diabetes. Proc. Soc. Exp. Biol. Med. 1976, [5], 356-361. (g) Yamada, K.; Nonaka, K.; Hanafusa, A.; Miyazaki, A.; Toyoshima, H.; Tarui, S. Preventive and theraeuitc effects of large-dose nicotinamide injections on diabetes associated with insulitis. An observation in nonobese diabetic (NOD) mice. Diabetes 1982, 3], 749-753. (h) Wilson, B. M.; Buckingham, B. Prevention of type la diabetes melitus. Pediatr. Diabetes 2001, 2, 17-24. (j) Chouinard, G.; Young, S.N.; Annable, L.; Sourkes, T. L. Tryptophan-nicotinamide, imipramine and their combination in depression. A 23 51 52 53 54 controlled study. Acta Psychiatr. Scand. 1979, 59, 395-414. (1) Murray, M. F.; Langan, M.; MacGregor, R. R. Increased plasma tryptophan in HIV-infected patients treated with pharmacologic doses of nicotinamide. Nutrition 2001, 17, 654—556. (k) Beales, P. E.; Burr, L. A.; Webb, G. P.; Mansfield, K. J.; Pozzilli, P. Diet can influence the ability of nicotinamide to prevent diabetes in the non-obese diabetic mouse. A preliminary study. Diabetes Metab. Res. Rev. 1999, 15, 21- 28. (I) Gedye, A. Hypothesized treatment for migraines using low doses of tryptophan, niacin, calcium, caffeine, and acetylsalicyclic acid. Med. Hypothesis 2001, 56, 91-94. (m) Kamat, J. P.; Devasagayam, T. P. Nicotinamide (vitamin B3) as an effective antioxidant against oxidative damage in rat brain mitochrondria. Redox Rep. 1999, 4, 179-184. (n) Ungerstedt, J. S.; Blomback, M.; Soderstrom, T. Nicotinamide is a potent inhibitor of proinflammatory cytokines. C lin. Exp. Immunol. 2003, I3 I , 48-52. (a) Draths, K. M.; Pompliano, D. L.; Conley, D. L.; Frost, J. W.; Berry, A.; Disbrow, G. L.; Staversky, R. J.; Lievense, J. C. Biocatalytic synthesis of aromatics from D-glucose: the role of transketolase. J. Am. Chem. Soc. 1992, 114, 3956-3962. (b) Gubler, M., Jetten, M., Lee, S. H., Sinskey, A. J. cloning of the pyruvate-kinase gene (Pyk) of Corynebacterium glutamicum and site-specific inactivation of Pyk in a lysine-producing Corynebacterium lactofermentum strain. Appl. Env. Microbial. 1994, 60, 2494-2500. (c) Miller, J. E.; Backman, K.C.; O'Conner, M. J.; Hatch, R. T. Production of phenylalanine and organic-acids by phosphoenolpyruvate carboxylase-deficient mutants of Escherichia coli. J. Ind. Microbial. 1987, 2, 143-149. ((1) Patnaik, R.; Liao, J. C. Engineering of Escherichia coli central metabolism for aromatic metabolite production with near theoretical yield. Appl. Environ. Microbial. 1994, 60, 3903-3908. (e) Patnaik, R.; Spitzer, R. G.; Liao, J. C. Pathway engineering for production of aromatics in Escherichia coli: confirmation of stoichiometric analysis by independent modulation of AroG, TktA, and Pps activities. Biotechnol. Bioeng. 1995, 46, 361- 370. (f) Li, K.; Mikola, M. R.; Draths, K. M.; Worden, R. M.; Frost, J. W. Fed- batch fermentor synthesis of 3-dehydroshikimic acid using recombinant Escherichia coli. Biotechnol. Bioeng. 1999, 64, 61-73. (a) Weaver, L. M.; Herrmann, K. M. Cloning of an AroF allele encoding a tyrosine-insensitive 3-deoxy-D—arabina-heptulosonate 7-phosphate synthase. J. Bacterial. 1990, I72, 6581. (b) Mikola, M. R.; Widman, M. T.; Worden, R. M. In situ mutagenesis and chemotactic selection of microorganisms in a diffusion gradient chamber. Appl. Biochem. Biotechnol. 1998, 70- 72, 905-918. Ogino, T.; Garner, C.; Markley, J. L.; Herrmann, K. M. Biosynthesis of aromatic- compounds - l3C NMR-spectroscopy of whole Escherichia coli cells. Proc. Natl. Acad. Sci. USA 1982, 79, 5828-5832. Draths, K. M.; Frost, J. W. Genomic direction of synthesis during plasmid-based biocatalysis. J. Am. Chem. Soc. 1990, [12, 9630-9632. 24 55 56 57 58 59 60 61 Snell, K. D.; Draths, K. M.; Frost, J. W. Synthetic modification of the Escherichia coli chromosome: enhancing the biocatalytic conversion of glucose into aromatic chemicals. J. Am. Chem. Soc. 1996, 118, 5605-5614. (a) Draths, K. M.; Frost, J. W. Synthesis Using Plasmid-Based Biocatalysis: Plasmid Assembly and 3-Deoxy-D-Arabino-Heptulosonate Production. J. Am. Chem. Soc. 1990, [[2, 1657-1659. (b) Frost,J. W. US. Patent 5,168,056, 1992. Williams, J. F.; Blackmore, P. F.; Duke, C. C. MacLeod, J. K. Fact, uncertainty and speculation concerning the biochemistry of D-erythrose-4-phosphate and its metabolic roles. Int. J. Biochem. 1980, [2, 339-344. Backman, K.C. US. Patent 5,169,768, 1992. Mori, M.; Yokota, A. ; Sugitomo, S.; Kawamura, K. Patent JP 62,205,782, 1987. Chandran, S. S.; Yi, J.; Draths, K. M.; von Daeniken, R.; Weber, W.; Frost, J. W. Phosphoenolpyruvate availability and the biosynthesis of shikimic acid. Biotechnol. Prog. 2003, I 9, 808-814. (a) Parker, C.; Barnell, W. 0.; Snoep, J. L.; Ingram, L. 0.; Conway, T. Characterization of the Zymomonas mobilis glucose facilitator gene product (glf) in recombinant Escherichia coli: examination of transport mechanism, kinetics and the role of glucokinase in glucose transport. Mol. Microbiol. 1995, I5, 795- 802. (b) Snoep, J. L.; Arfman, N.; Yomano, L. P.; Fliege, R. K.; Conway, T.; Ingram, L. 0. Reconstruction of glucose uptake and phosphorylation in a glucose- negative mutant of Escherichia coli by using Zymamanas mobilis genes encoding the glucose facilitator protein and glucokinase. J. Bacterial. 1994, I 76, 2133- 2135. (c) Weisser, P.; Kramer, R.; Sahm, H.; Sprenger, G. A. Functional expression of the glucose transporter of Zymomonas mobilis leads to restoration of glucose and fructose uptake in Escherichia coli mutants and provides evidence for its facilitator action. J. Bacteriol. 1995, I 77, 3551-3554. 25 HAPTER T Optimizing microbial synthesis of (-)-quinic acid Introduction Two stereospecific quinic acid syntheses are reported in the literature.l However, commercially available quinic acid is isolated from Cinchona bark.2 The first reported microbial synthesis of quinic acid relied on heterologous expression in E. coli of the Klebsiella pneumoniae qad gene that encodes quinate dehydrogenase.3 Qad catalyzed the reduction of 3-dehydroquinic acid to quinic acid. In K. pneumaniae, Qad oxidizes quinic acid to 3-dehydroquinic acid in the presence of NAD”, which is the first step in quinic acid catabolism via the B-ketoadipate pathway.4 The reduction of 3-dehydroquinic acid by Qad in E. coli resulted in formation of quinic acid because E. coli does not have the ability to catabolize 3-dehydroquinic acid. Quinic acid biosynthesis utilizing Qad relied on maintenance in an E. coli host of two plasmids, which is problematic during high denSIty cultivation under fermentor-controlled conditions. A subsequent microbial synthesis of quinic acid developed by the Frost group utilized native E. coli shikimate dehydrogenase, which was discovered to reduce 3-dehydroquinic acid to quinic acid.5 The resulting construct, E. coli QP1.l/pKD12.112, synthesized 40 g/L of quinic acid in 16% yieid.5 26 Table 2. Comparison of the impact of modification of the central metabolism in E. coli on synthesis of 3-dehydroshikimic acid and quinic acid. Desired product [DHS], DHS [QA], QAyield, Entry Construct Relevant characteristics g/L yiled, % g/L °/o 1 KL3/pKD12.291A aroFFBR, 3804’ Pam: 41 18 - - 2 KL3/pKL5.17A amFFBR’ 8604' Pam]; MA 58 24 - - 3 KL3/pJY1.216A aroFFBR, semi Pam]; WA 69 35 - pins/1 4 JY1/pJY2.183 amFFBFi’ 86m, PamF’ MA, 60 34 - Ptac g/fg/k 5 QP1.1/pK012.112 amFFBFi, 8,05, 5904 - 4o 16 6 QP1.1/pKD12.138 ameR’ 8,05, 8%, MA 49 2o 7 QP1.1/pKD15.071 amFFBFt 6,05, 8904’ MA, ppsA 49 18 8 QP1.1/pNR4.230 aroFFBR, PtacamHORF)’ 59% 46 20 WA 0 KL3: A32834 serA::aroB; QPI .I: A82848 serA::aroB; JY I: KL3 AptsHpts/crr::KanR. b aroFFBR: feed-back insensitive DAHP synthase; serA: 3-phosphoglycerate dehydrogenase Pump: promoter locus of E. coli araF gene; tktA: transketolase; ppsA: phosphoenolpyruvate synthase; glf. glucose facilitator from Z. mobilis; glk: glucose kinase. C DHS: 3-dehydroshikimic acid; QA: quinic acid. d (mol DHS)/(mol glucose consumed). A variety of strategies have been employed for improving the yields of shikimate pathway metabolites synthesized by E. coli.6'7'8 These strategies have focused on expression of feedback-insensitive DAHP synthase as well as increasing the availability of D-erythrose 4-phosphate (E4P) and phosphoenolpyruvate (PEP). E4P and PEP are the substrates of DAHP synthase, which is the first enzyme in the shikimate pathway and catalyzes an irreversible condensation to form 3-deoxy-D-arabiI'Io-heptulosonic acid 7- phosphate (DAHP Figure I). E4P is derived from the pentose phosphate pathway and PEP is formed in the Embden-Meyerhof pathway (glycolysis) as shown in Figure 5. Initial attempts to increase carbon flow into the shikimate pathway examined 27 . . . . . 9 overexpreSSIon of feedback-Insensmve DAHP synthase Isozymes. Frost and coworkers used a 3-dehydroshikimate-synthesizing strain to demonstrate that the increases in overexpression of DAHP synthase leads to increased accumulation of 3-dehydroshikimic . . 6f . . . . aCId and byproducts In the culture supernatent. However, additional Increases In DAHP synthase overexpression failed to have a positive impact, i.e. increased in synthesized 3- dehydroshikimic acid.6f Availability of E4P or PEP might be a limiting factor, since both molecules are used as the substrates for DAHP synthase. Overexpression of plasmid- Iocalized tktA-encoding transketolase resulted in higher DAH production, which was attributed to increased E4P availability.'0 A 3-dehydroshikimic acid concentration of 41 g/L synthesized in 18% yield (Table 2, entry 1) increased to 58 g/L synthesized in 24% yield (Table 2, entry 2) once transketolase was overexpressed together with feedback- f insensitive DAHP synthase in E. coli KL3/pKL5.l7A.6 A similar trend was observed for microbial synthesis of quinic acid. E. coli QP1.l/pKD12.I 12 synthesized 40 g/L of quinic acid in 16% yield after 60 h cultivation without overexpression of transketolase (Table 2, entry 5).3 However, when transketolase was overexpressed with plasmid- Iocalized tktA, E. coli QP1.l/pKD12.I38 produced 49 g/L of quinic acid in 20% yield . . . . I after cultivation under fermentor-controlled conditions for 48h (Table 2, entry 6).l In order to eliminate PEP limitation, a couple of strategies have been 8.12‘6de employed. ' Liao and co-workers demonstrated that a twofold increase in DAH production can be achieved by overexpression of phosphoenolpyruvate synthase, an enzyme that converts pyruvate to PEP with conversion of ATP to AMP.6d The combined 28 positive effect on synthesized DAH was observed once Liao and coworkers overexpressed transketolase and phosphoenolpyruvate synthase in the presence of feedback-insensitive DAHP synthase.6e Frost and co-workers demonstrated the combined effect of transketolase and phosphoenolpyruvate synthase overexpression using E. coli KL3/pJY1.216A, which synthesized 69 g/L of 3-dehydroshikimic acid under glucose-rich culture conditions in 35% yield (Table 2, entry 3).8 The same strategy was explored for quinic acid synthesis. Overexpression of plasmid-localised ppsA-encoded phosphoenolpyruvate synthase in E. coli QP1.l/pKD15.07I resulted in synthesis of 49 g/L of quinic acid 48 h in 18% yield under fermentor-controlled conditions (Table 2, 11 entry 7). Another strategy for increasing PEP availability employed a non phosphoenolpyruvatezcarbohydrate phosphotransferase system (P'TS) to phosphorylate and transport glucose across the cytoplasmic membrane. This strategy was based on the fact that during PTS-mediated glucose transport in E. coli one molecule of PEP is used to catalyze the phosphorylation and transport of one molecule of glucose, therefore limiting PEP in viva availability. PEP used during P’TS-mediated transport of glucose is converted to pyruvate, which is ultimately channeled to the tricarboxylic acid cycle (TCA). Therefore, E. coli lacking PTS-mediated glucose transport might possess increased in viva PEP availability. Frost and coworkers showed that E. coli pts‘ host strain JYI with plasmid pJY2.I83A—localized transketolase and feedback-insensitive DAHP synthase overexpression synthesized 60 g/L of 3-dehydroshikimic acid in 34% yield (Table 2, entry 7).'2 This showed an increase in 3—dehydroshikimic acid 29 concentration and yield when compared to KL3/pKL5.l7A (Table 2, entry 2), which uses PTS for glucose transport (previously described). However, the increase was not as significant as for KL3/pJYI.216A (Table 2, entry 3), which employs phosphoenolpyruvate synthase for PEP recycling during FT S transport of glucose. Since JYI had an inactivated P'TS-mediated glucose transport system, plasmid pJY2.183A additionally carried the glf gene encoding the glucose facilitator protein from Zymomonas mobilis and the E. coli glk gene encoding the glucose kinase. Quinic acid producer host, lacking native PTS system QP1.lpts was constructed. Plasmid pKD12.l38 was modified by inserting nglfglk insert and resulted in pSC6.090 plasmid.l3 A new quinic acid producer QP1.lpts/pSC6.090 failed to grow well under fermentor-controlled . . 11 conditions. 3-Dehydroquinic acid accumulation was observed throughout the microbial synthesis of quinic acid. Therefore, another strategy was to increase shikimate dehydrogenase AroE activity. Ran reported that QP1.l/pNR4.23O had twofold higher shikimate dehydrogenase activity than QP1.l/pKD12.138.ll above However, under the same conditions it produced 46 g/L of quinic acid in 20% yield (Table 2, entry 8), QP1.l/pKD12.l38 synthesized 49 g/L of quinic acid in 20% yield (Table 2, entry 6)." Increase in shikimate dehydrogenase activity did not lead to increased quinic acid concentration and yield. 30 Biocatalytic synthesis of quinic acid by fed-batch fermentation The quinic acid-synthesizing E. coli strains were cultivated under fed-batch glucose-limited fermentor—controled conditions at 33 °C, pH 7.0 and dissolved oxygen was maintained at 10% air saturation. Plasmid maintenance relied on nutritional pressure, since all E. coli hosts had disrupted serA genes and each carried plasmid with a serA insert. Glucose addition was controlled by dissolved oxygen concentration. Under aerobic conditions, 02 is the best electron sink,” therefore a subtle change in metabolic rate can be detected immediately by dissolved oxygen concentration in the medium. When dissolved oxygen concentration (p02) decreased below a set point (10%) indicating an increased rate of metabolism due to ample concentrations of glucose, the rate of glucose addition decreased. Conversely, when pO2 increased indicating a slower rate of metabolism due to inadequate concentration of glucose in the culture medium, the rate of glucose addition was increased. The glucose addition rate was controlled by a proportional-integraI-derivative (PID) control loop. Overtime, the controller found the steady addition rate resulting in a steady state concentration of glucose of approximately 0.2 mM in the medium. A proportional gain (KC) of 0.1 was used for glucose PID control. A concentration range of 55-170 mM glucose in the fermentation medium was maintained by manually adjusting the rate of glucose addition under glucose-rich conditions. Fermentations were run in duplicate and reported results represent the average of two runs unless otherwise stated. Metabolite concentrations were determined using 'H NMR unless otherwise stated. 31 Host E. coli host QP1.1D was constructed by site-specific insertion of araB into the serA locus of E. coli ABZ848, which has inactive 3-dehydroquinate dehydratase due to a mutation in the araD gene.15 3-Dehydroquinate dehydratase (AroD) is a shikimate pathway (Figure 1) enzyme. Without AroD, E. coli QPI.I can not biosynthesize the aromatic amino acids and aromatic vitamins required for its growth and metabolism. Therefore, all QP1.I cultures had to be supplemented with aromatic amino acids L- phenylalanine, L—tyrosine and L-tryptophan, and the precursors for aromatic vitamins p- hydroxybenzoic acid, p—aminobenzoic acid, and 2,3-dihydroxybenzoic acid, when grown in minimal salt medium. A second genomic copy of the araB locus in E. coli QP1.I increased the specific activity of 3-dehydroquinate synthase to a level where accumulation of 3-deoxy-D-arabino-heptulosonic acid (dephosphorylated DAHP Figure l) was no longer observed.l6 Disruption of the serA locus in QPI.1 led to inactive 3- phosphoglycerate dehydrogenase, which is an enzyme required for L-serine biosynthesis in wild-type E. coli. A copy of the serA gene was inserted into plasmids and provided the basis for plasmid maintenance when cultures were cultivated on glucose in minimal salt medium. The construction plasmid pKD12.I 12 was previously reportedsas well as plasmid 9 pKD12.l38,” pNR4.230H and pii4.i7i.' The plasmid maps are shown in Figure 6. . . . ‘B All plasmids shared genomic elements like araFI~ R, PMCPamEaroE and serA. To overcome feedback inhibition of DAHP synthase caused by the aromatic amino acid supplements, plasmid-localized feedback-insensitive DAHP synthase encoded by aroFFBR was required. Shikimate dehydrogenase encoded by araE ensured reduction of 32 3-dehydroquinic acid to quinic acid. Plasmid pKD12.112 and pKD12.l38 had aroE expression controlled by a native PamE promoter while araE in pNR4.230 was expressed from a Pm promoter. Plasmid pJJ4.l7l was identical to pKD12.l38 except it had ydiB ORF under control of P promoter, rather than PamEaraE. Plasmid-encoded serA was lClC required to maintain the plasmid under glucose-limited conditions and to restore 3- phosphoglycerate dehydrogenase activity after araB was inserted into the serA locus of genomic DNA. Plasmids pKD12.l38 and pNR4.230 also had transektolase encoding tktA locus. 33 (Hindlll) pJJ4.171A 9.5 kb j .\ BamHl PtacydiB \ _ P lac I ‘ 1% Kpnl aroFFBR E c 0R1 EcoRl Figure 6. Plasmid maps of pKD12.112, pKD12.l38, pNR4.230 and pJJ4.171. 34 A search for a better shikimate/quinate dehydrogenase Previously in the literature, E. coli YdiB was named a putative shikimate dehydrogenase, because it showed activity towards shikimic acid,'8 and E. coli ydiB gene is in the same operon as another shikimate pathway enzyme araD-encoded 3- dehydroquinate dehydratase. Crystal structures of YdiB and AroE were found to be very similar, even though _\‘(llB share only 25% nucleotide sequence identity with aroE.'8a It was also demonstrated that YdiB catalyzes shikimic acid conversion to 3- dehydroshikimic acid and quinic acid conversion to 3-dehdyroquinic acid in the presence of NADP and NADL'gu Further evaluation of YdiB proved that it can function as a second shikimate dehydrogenase in E. coli.1 First, It was shown that plasmid-localized ycliB restored E. coli A32834 growth on glucose-minimal medium.'9 E. coli A82834 lacks shikimate dehydrogenase actIVIty due to a mutation in the aroE gene, 5 which results in an inability to grow on glucose-minimal medium lacking supplementation with shikimic acid or alternatively, supplementation with aromatic amino acids and aromatic Vitamins. I The YdiB enzyme was also characterized as a feedback-insensetive shikimate dehydrogenase with shikimic acid, although based on Km and kcat determination YdiB was not as active of a shikimate dehydrogenase relative to AroE (Table 3).” Table 3. Shikimate dehydrogenase AroE and YdiB kinetic parameters for 3-dehydroshikimic acid. Enz me -1 Y Km(mM) keat. s AroE 0.10 361 YdiB 10 83 35 An attempt was made to replace AroE with YdiB in microbial synthesis of quinic acid. E. coli QP1.l/pJJ4.l7IA did not produce any quinic acid under glucose-limited conditions in 48 h (Table 4, entry 3; Figure 8A). However, both control strains QP1.l/pKD12.I38 (Table 4, entry 1; Figure 7A) and QP1.l/pNR4.230 (Table 4, entry 2; Figure 78) produced 58 g/L in 22% yield and 57 g/L in 22% yield of quinic acid, respectively. E. coli QP1.l/pJJ4.I71 did not produce any quinic acid (Table 4, entry 4; Figure 88) when cultured glucose-rich fermentor-controlled conditions. The only hydroaromatic synthesized by QP1.l/pJJ4.I7I was 3-dehydroquinic acid, which was synthesized at 52 g/L in 21% yield (Table 4, entry 3) under glucose-limited culture conditions and at 50 g/L in 21% yield (Table 4, entry 4) when cultivated under glucose- rich conditions. Since, 3-dehydroquinic acid is a substrate for shikimate dehydrogenase AroE and YdiB (Figure l), accumulation of this metabolite instead of quinic acid suggested a lack of ydiB expression. A. B. A 70 A 70 . 5 60 twn‘: 3 50 c . E o "' " a? m ’1 I— g 50 . l— '— 3 4o . _ ..... 1 I 40 . .......... 8 § 7 g m —1 I-.. z‘ m ---------------- o o o 1f 20 20 .. 1 o 5 g 10 .. .1 - 1o ----------- ---- .- o . go gygcre 12 18 24 30 36 42 48 12 18 24 80 38 42 48 Time(h) Time(h) Figure 7. (A) E. coli QP1.l/pKD12.138 and (B) QP1.l/pNR4.230 cultured under glucose-limited conditions. Legend: (dotted bars) 3-dehydroquinic acid, (open bars) quinic acid, (black circles) dry cell weight. 36 Table 4. Concentrations and yields of products synthesized by quinic acid producing strains with plasmid-overexpression of ydiB or aroE. C IQAI. d Total Entry Construct [DHQ] QA . e 9"- yield, °/o yield, % a P1.1/ KD12.1 8 21 22 1 0 p 3 5 58 a P1.1/ NR4.230 22 22 2 O p 5 57 a P1.1/ J4.171 0 21 3 ' Q pd 52 o b P1.1/ JJ4.171 0 21 4 ' O p 50 o a P1.1/ JJ5.069 4 26 5 ' O p 62 11 b, QP1.1/pJJ5.069 3 19 48 8 a P1.1/ JJ5.073 5 12 ' O p 12 20 a . . . . b . . . * . Glucose-limited conditions. Glucose-rich conditions. Single run fermentation. CAbbreviations: DHQ — 3-dehydroquinic acid, QA — quinic acid. d(mol QA)/(mol glucose consumed). e(mol DHQ + mol QA)/(mol glucose consumed). P ------------- 888883 .s O DHQ, QA. Dry cell weight (94.) I .' '3': ’e'. t.'. .'. :t‘. O 12 ‘1 1' Time (h) 12182436424860 88883 DHQ, 0A. Dry cell Wfiight (94-) O 33 121824303642485460 'I'Ime(h) Figure 8. (A) E. coli QP1.l/pJJ4.l7l cultured under glucose-limited culture conditions and (B) QP1.l/pJJ4.171 cultured under glucose-rich culture conditions. Legend: (dotted bars) 3-dehydroquinic acid, (open bars) quinic acid, (black circles) dry cell weight. 37 Similar gene overexpression problem were previously encountered when attempts to overexpress the ORF adk-encoded of adenylate kinase from a Pm. promoter did not yield assayable enzyme activity.'9 When 30 bp upstream from E. coli genomic adk were 20. 19 included with the adk ORF, successful overexpression of Adk was achieved. The same strategy was applied for ydiB cloning and expression. Two final plasmids were constructed in order to evaluate ydiB overexpression and quinic acid production under fermentor-controlled conditions. Plasmid pJJ5.069 was designed to be identical to pKD12.l38 except P EaraE was replaced by P vdiB with 31 bp upstream genomic aro ta C. sequence included. Plasmid pJJ5.073 was constructed in the pKK223-3 cloning vector, rather than vector pSU18 as used for construction of plasmids pKD12.l38 and pJJS.069. Construction of plasmid pJJ5.067 began with PCR amplification of a 0.9 kb ydiB fragment with a 31 bp upstream sequence (ydiB+3lbp) from E. coli W3110 genomic DNA. The isolated PCR fragment was inserted into EcoRI and SmaI cloning site of the pKK223-3 vector (Figure 9). Since the target plasmid pJJS.069 had to be as close as possible to pKD12.l38, therefore plasmid pKD12.112 was chosen as the starting point for construction of the plasmid carrying the ydiB + 31 bp insert. Digestion of plasmid pKD12.l 12 with Kpnl and BamHI restriction endonucleases lead to loss of the 3.2 kb P aroE serA fragment. PCR amplified P vdiB+3lbp [(1 (7 ta ('- fragment from pJJ5.067 with Kpnl and BamHI ends was inserted into Iinerized pKD12.112 yielding plasmid pJJ5.068 (Figure 10) after ligation. Digestion of plasmid pNR8.146A with Xbal restriction endonuclease liberated an intact 3.8 kb tktA serA 38 fragment, which was cloned into pJJ5.068 previously pretreated with Xbal to afforded target plasmid pJJS.069 (Figure I I). Smal Pstl EcoRl Hindi I I Ptac PCR ydiB from E. co/iW311O genomic DNA 11) EcoFil and Smal digest Ecol-'11 Smal 1 kb ydiB 1) EcoFil digest 2) CIAP treatment T4 Ligase Pstl Hindlll Figure 9. Construction of plasmid pJJ5.067. 39 It} Xbal ng172k1b12A % BamH. PCR PtacydiB from pJJ5.067 (Smal) / EcaRl I - / serA 1) Kpnl and BamHI digest aroFFBR / EcoRl (Smal) Kpnl BamHI Kp’" 1.2 kb 1) Kpnl and BamHI digest PtacydiB 2) CIAP treatment ¢ T4 Ligase Figure 10. Construction of plasmid pJJ5.068. pNR8.146A 6.7 kB Pstl Xbal BamHl l1) Xbal digest Kpnl Xbal Xbal ECOFII ECOFII 3.8 kb :::‘::::::::::::::::::::W/////////////// 1) Xbal digeSt tktA serA 2) CIAP treatment T4 Ligase Xbal Figure 11. Construction of plasmid pJJ5.069 4| pKD12.112A 7.7 kb ¢ 1) EcaFil digest pJJ5.067 2) Kelnow treatment 5.6 kb EcoFil Smal l 1.2 kb I {3% ELM: aroFFBR 1) Smal digest 2) CIAP treatment T4 Ligase Pstl Figure 12. Construction of plasmid pJJS .072. 42 I; pNR8.146A i5 8.7 kB Pstl Xbal BamHl 2) Klenow treatment I1) Xbal digest Xbal Xbal " 3.8 kb 1)Hindllldigest L:;::::::::::::::::;::::[/////////////////, 2) Klenow treatment tktA serA 3) CIAP treatment I T4 Ligase (Hindlll) tktA , . ,.\ Ar?" II}. III pJJ5.073 I 10.6 kb I}? f/ .4” tac f w,- Figure 13. Construction of plasmid pJJ5.073. 43 Another plasmid pJJ5.073, had the same genomic elements as pJJ5.069, except that the parent vector was pKK223-3 rather than pSU18. The construction of pJJ5.073 started with digestion of pKD12.112 using EcoRI restriction endonuclease followed by agarose gel isolation of a 1.2 kb aroFFBR fragment and Klenow treatment with dNTPs. The liberated aroFFBR fragment was inserted into previously Smal digested pJJ5.067, which resulted in pJJS.072 (Figure 12). Digestion of plasmid pNR8.146A with Xbal restriction endonuclease liberated an intact 3.8 kb tktA serA fragment, which was treated with Klenow. Insertion of this Klenow—treated tktA serA fragment into Hindlll-digested and Klenow-treated pJJS.072 afforded target plasmid pJJ5.073 (Figure 13). Newly constructed plasmids were evaluated under fermentor-controlled conditions. E. coli QP1.l/pJJ5.069 produced I 1 g/L of quinic in 4% yield under glucose- limited conditions (Table 4, entry 5; Figure 14A) and 8 g/L in 3% yield under glucose- rich conditions (Table 4, entry 6; Figure 143). The major byproduct for both fermentations was 3-dehydroquinic acid, which accumulated to 62 g/L (Table 4, entry 5) and 48 g/L (Table 4, entry 6). The total yield of synthesized hydroaromatics (3- dehydroquinic acid + quinic acid) was 26% for QP1.l/pJJ5.069 cultivated under glucose- limited conditions and 19% for QP1.l/pJJ5.069 cultivated under glucose-rich conditions. These yields are close to the 22% observed for the control strains (Table 4, entry 1 and 2). This indicated that the carbon flow directed into the shikimate pathway was the same and only the last step, reduction of 3-dehydroquinic acid to quinic acid, was not as efficient for QP1.l/pJJ5.069 relative to QP1.l/pKD12.I38. > 3 . In "‘ A 5 so 3 so a E '2 5° '9' 5" 3 4° a g 3° 5 2° ._ 6 gm g E; 2:211 :55 o a £555., £222 r-i 2:511 281'] O I r T o I I fir T 12182430364248 121824303648 Time(h) Time(h) \l O DHQ. QA, Dry cell weight (glL) 8 8 8 8 8 1o :Z;l --.. 3:2; -. . . ’1'? 2-2- '2-2 I I p u a o o - '0'. ’D-" a... .I'I . '2': *2’1' 2'21 '2': t W l t I 12 18 24 30 36 48 Time(h) D Figure 14. (A) E. coli QP1.l/pJJ5.069 cultured under glucose-limited conditions, (B) QP1.l/pJJ5.069 cultured under glucose-rich conditions and (C) QP1.l/pJJ5.073 under glucose-limited conditions. Legend: (dotted bars) 3-dehydroquinic acid, (open bars) quinic acid, (black circles) dry cell weight. YdiB enzymatic activities were assayed in order to determine why 3- dehydroquinic acid reduction was deficient. Previously reported YdiB can utilize shikimic acid and quinic acid as substrates as well as NAD+ and NADP+ as cofactors. New constructs were assayed in the forward direction, i .e. reduction of 3-dehydroquinic 45 acid or 3—dehydroshikimic acid with NADH or NADPH. E. coli host QPI.I showed background shikimate/quinate dehydrogenase specific activity at around 0.1 U/mg (Table 5, entries 1—4). E. coli hosts DHSOI and RB79lserAzzaroB had similar background shikimate/quinate dehydrogenase activities relative to QPI.I. Based on the measured quinate dehydrogenase specific activity with NADPH as the cofactor (Table 5, entry 6) versus NADH as the cofactor (Table 5, entry 6), NADPH appears to be the preferred cofactor for quinate dehydrogenase. YdiB with NADPH displayed a twofold higher specific activity for reduction of 3-dehydroshikimic acid (Table 5, entry 7) relative to reduction of 3-dehydroquinic acid (Table 5, entry 5). QP1.l/pJJ5.067 had approximately tenfold higher quinate dehydrogenase activity (Table 5, entry 5) as compared to the background quinate dehydrogenase activity in host E. coli QP1.1 (Table 5, entry 1). This indicated that overexpression of ydiB was being achieved. E. coli RB79IserA::aroB/pJJ5.069 showed only background level quinate dehydrogenase activity (Table 5, entry 10). This level of quinate dehydrogenase activity is consistent with 3-dehydroquinic acid being the major product synthesized by QPI .1/pJJ5.069 under fermentor—controled conditions. The discovery that final plasmid pJJ5.069 had no or poor overexpression of plasmid-localized ydiB, served as the reason to construct a new plasmid pJJ5.073 where ydiB in the final construct was not a PCR product as was the ydiB insert in pJJS.069. Even a high fidelity DNA polymerase can result in spontaneous mutation during PCR. Table 5. YdiB specific acitivities. Entry Construct Substrate, Cofactor Specific activity,a U /mg 1 QP1.1 DHQ, NADPH 0.14 2 QP1.1 DHQ, NADH 0.09 3 QP1.1 DHS, NADPH 0.14 4 QP1.1 DHS, NADH 0.01 5 QP1.1/pJJ5.067 DHQ, NADPH 1.29 6 QP1.1/pJJ5.067 DHQ, NADH 0.21 7 QP1.1/pJJ5.067 DHS, NADPH 2.34 8 QP1.1/pJJ5.067 DHS, NADH 0.20 9 DHSct/pJJ5.068 DHQ, NADPH 0.69 10 RB791serA::aroB/pJJ5.069 DHQ, NADPH 0.13 11 FiB791serA::aroB/pJJ5.073 DHQ, NADPH 0.10 a One unit (U) of shikimate/quinate dehydrogenase corresponds to the formation of l umole of NAD(P) in the presence of 3-dehydroshikimic or 3-dehydroquinic acid per min at 25 °C. E. coli QP1.1/pJJ5.073 was evaluated under fermentor-controled glucose-limited conditions. This time it produced 14 g/L of quinic acid in 5% yield (Table 4, entry 7; Figure 14C). It was the best quinic acid producer with ydiB overexpression, but it produced fourfold less of quinic acid as compared to the control strains (Table 4, entry 1). Total hydroaromatics yield of 12% was also low as compared to the 22% produced by QP1.1/pKD12.l38. 3-Dehydroquinic acid also accumulated at leaser (19 g/L, Table 4, entry 7) concentrations as compared to the 62 g/L produced by QP1.1/pJJ5.069 (Table 4, entry 5). This could indicate that E. coli QP1.1/pJJ5.073 had less carbon flow directed to the shikimate pathway. Quinate dehydrogenase activity for RB791serA::araB/pJJ5.073 revealed 0.10 U/mg of specific activity which is at the same level as the background E. coli quinate dehydrogenase specific activity (Table 5, entry 1). Even though parent plasmid pJJ5.067 had twenty-fold higher quinate dehydrogenase activity, after inserting shared genomic elements, the final activity dropped down to the host levels, no matter how the quinate—synthesizing strains QP1.1/pJJ5.069 and QP1.1/pJJ5.073 were 47 constructed. Further pursuit of using YdiB to replace AroE as the shikimate/quinate dehydrogenase in quinate-synthesizing constructs was abandoned due to low YdiB activities in the final constructs. E. coli B as a quinic acid producer E. coli QP1.1 is 8 K-12 strain. Investigation of E. coli B as a host strain for the quinic acid production was pursued for a couple of reasons. From previous experiments in the Frost group, it was discovered that E. coli B has higher native transketolase specific activity than E. coli K-l2.2| Plasmid-localized transketolase overexpression might not be necessary for quinic acid production. Secondly, E. coli B is deficient in Lon 22 . . . . proteases, which are the major proteases catalyzmg the endoproteolytic cleavage of proteins in the cell. High and more stable DAHP synthase and transketolase activity levels might therefore be in E. coli B, especially in the stationary phase, might therefore be realized in E. coli 8.22 Strains disrupted in the [on gene produce several phenotypic changes including increased sensitivity to UV and ionizing radiation, overproduction of mucopolysaccharide, reduced Iysogenization of bacteriophages lambda and PI , and most importantly, reduced protein degradation.22d Synthesis of quinic acid requires an inactive 3-dehydroquinate dehydratase encoded by araD as well as a second copy of 3-dehydroquinate synthase encoded by 5 . . 21 . . . araB. E. coli B serA::aroB was preVIously constructed. Only aroD Inactivation was required to afford an E. coli B host suitable for synthesis of quinic acid. 48 Wanner and coworkers have developed a method for gene deletion or disruption in E. coli (Figure 15).23 The first step involves PCR of an antibiotic gene flanked by two FRT (flippase recognition target) sequences. Primers for this PCR step are designed in such a way that the first 40-50 bp sequence (H1 or H2) is homologous to the gene of interest that is going to be disrupted. The last 20 bp of the primer are homologous to the priming sequence (PI and P2) before the FRT sites in the template plasmids. The template plasmids used in this work were pKD3, which encodes for chloramphenicol resistance, or pKD4, which encodes for kanamycin resistance. In the second step, a target E. coli host is transformed with the plasmid pKD46, which expressed bacteriophage A Red recombinase from an inducible arabinose promoter. E. coli with A Red recombinase is transformed with the PCR product from the first step, and the mutants are selected on antibiotic plates. The key factor of this step is the transformation efficiency. Since, transformation is performed with linear DNA from the PCR step, endogenous nucleases start degrading this linear DNA prior to completion of the recombination step. Therefore, a high transformation efficiency of 108-109 transformants per pg of DNA is required in order to obtain several mutants. Elimination of the antibiotic resistance marker is performed by transforming the mutant E. coli with plasmid-encoded FLP flippase pCP20,24 which acts on the two FRT sites flanking the antibiotic gene. Flippase in pCP20 is transcribed from a temperature inducible promoter and the plasmid has a temperature-sensitive replicon, as well as ampicillin and chloramphenicol resistance genes. E. coli is transformed with pCP20 and ampicillin- resistant mutants are selected at 30 °C, after which a few single colonies are incubated at 43 °C in order to induce FLP synthesis and promote plasmid pCP20 loss. A mutant E. 49 coli strain with knocked-out gene is obtained with a small FRT scar (Figure 15, step 4.). Helper plasmid encoding for A Red recombinase pKD46, template plasmids pKD3 and pKD4, and FLP-encoding plasmid pCP20 were purchased from E. coli Genetic Stock Center at Yale University. PCR amplify FRT-flanked resistance gene 69 FRT FRT PI Antibiotic resistance ’ , r//////////////////7//////////,W/// ////z LW Transform strain expressing it Red recombinase 2 GeneA Ell GeneB L12 GeneC ' l l l l s 8 Select antibiotic-resistant transformants FRT .. _ . FRT Gene A ' Antibiotic reSIstance F Gene C L K//// VflflWflW/A’W/yfl '///Z I Eliminate resistance cassette using FLP expression plasmid FRT 4 Gene A Gene C ' l m 97/21 I Figure 15. Gene deletion method in E. coli. From previous experience in the Frost group, it was known that wild-type E. coli B has very low transformation efficiency. Therefore, a strategy (Figure 16) was 50 employed to first delete the araD gene in E. coli W31 10 K-12 and then use PI phage to transduce the mutation into E. coli 3.25 Construction of the E. coli W3110 araD(-) mutant started with primer HIPI and H2P2 design. The HI sequence was chosen to be first 40 bp of araD ORF (5’-ATGAAAACCGTAACTGTAAAAGATCTCGTCA- TTGGTACGG-3’) and the H2 sequence was designed to be last 40 bp of araD ORF (5’- T'TATGCCTGGTGTAAAATAGTTAATACCGTGCGCAAATCA-3’). PI (5’-GTGT- AGGCTGGAGCTGC I I C-3’) and P2 (5’—CATATGAATATCCTCCTTAG-3’) sequences were based on template pKD3.23 A 1.1 kb PCR product was electroporated into E. coli W3110/pKD46 overexpressing A Red recombinase. The mutants were selected on LB/Cm plates and the phenotype of obtained mutants screened by replication on selective plates. Since araD is gene in the shikimate pathway, E. coli bearing inactive araD will not be able to grow on glucose-minimal plates without aromatic amino acid and aromatic vitamin supplementation. Mutants also should grow on LB/Cm plates, due to FRT-cat—FRT insertion in the araD gene. Sensitivity to ampicillin should indicate a loss of pKD46 or pCP20 plasmid. Several mutants were screened for the desired phenotype. E. coli W31 10 showed sensitivity to the ampicillin and chloramphenicol but did not require aromatics or serine supplementation for growth on glucose-minimal plates (Table 6, entry 1). Mutant E. coli W3110 AaraD::FRT—cat—FRT required aromatic amino acid and aromatic vitamin supplementation while grown on glucose-minimal salts plates, and it was sensitive to ampicillin but not to chloramphenicol (Table 6, entry 4). The mutant was also confirmed by PCR. PCR primers VI and V2 (Figure 16, Step 2) were designed to have homology outside H1 and H2 region. Therefore the PCR product for wild-type E. coli W31 10 was expected to be approximately lkb and for the mutant it 51 was expected to be 1.2 kb. PCR screening revealed a 1.2 kb size product on agarose gel for the mutant and a lkb for the wild-type control. Both, phenotype screening and PCR verification suggested that obtained E. coli W31 10 AaroDzzFRT-cat—FRT was the COI'I'CCI mutant. Table 6. Screening for E. coli mutant phenotype. M9/ . “'9’ "'9’ M9] Gluc/ LB/ LB/ Entry Strain c Gluc/ Gluc/ LB Gluc Ser Aros Aros/ Ap Cm Ser 1 E. coliW311O + + + + — — + 2 E. coli B + + + + — — + 3 E. coli B serA::aroB - + — + — — + 4 E. coliW3110 - — + + _ + AaroD::FRT-cat-FRT Sa E.C0llW3110 — — + + _. + AaroD:.FFiT-cat—FRT 6 E. coli B serA::aroB — — — + _ + AaroD:.FFiT-cat-FFIT 7 E. coliW3110 AaraD::FRT- cat -FRT — — + + _ + (new) 8b E. coIiW3110 — — + + _ + AaroD:.FFiT- cat -FRT (new) 9 E. coli B serA::aroB — — — + _ + AaroDsthT- cat-FRT (new) 10 E. coli B serA::aroB - — — + — _ + AaroD:.FRT (new) a E. coli W31 10 was transduced with PI-W3l 10 AaraDzzFRT-cat-FRT. b E. coli W31 10 was transduced with PI-W3I 10 AaroD::FRT-cat-FRT (new). C Abbreviations: Gluc — glucose, Ser — L-serine, Aros - aromatic amino acids (L-phenylalanine, L-tyrosine and L- tryptophan) and aromatic vitamins (2,3-dihydroxybenzoic acid, p-aminobenzoic acid and p-hydroxybenzoic acid), Ap — ampicillin, Cm — chloramphenicol. Note: Five colonies of each strain were screened per selective plate. 52 1. PCR amplify insert with FRT-flanked resistance gene '5’; pm L \5 ”1 F5" < P2 Cm” % 2. Transform fragment into W3110 expressing A Red recombinase and select for chloramphenicol resistance ydiB fl H1 P1FRT FET L ydiF L 11 I: - l 1‘ l 4' 3’00 am” 152—H2 V2 3. P1 phage-mediated transduction to E. coli B serA::aroB 4. Eliminate chloramphenicol resistance using FLP ydiB FRT ydiF E. coliB serA::aroB AaroD:.FRT I 11 —L- J] J r ll Figure 16. Construction of E. coli AaraD::FRT mutant. Pl phage-mediated transduction was used to transfer the AaraD::FRT-—cat—FRT mutation from E. coli W3110 AaroD::FRT—cat—FRT to E. coli B serA::araB. Mutants were screened for the correct phenotype and were verified by PCR analysis. Wild-type E. coli W3110 also was used as a control strain for the PI phage-mediated transduction. E. coli W3110 mutants obtained by plate after PI transduction showed identical growth characteristics (Table 6, entry 5) as the donor strain (Table 6, entry 4). Candidate E. coli B serA::aroB AaroDzzFRT—cat—FRT mutants also showed the correct growth pattern by requiring aromatics and serine supplementation for growth on glucose-minimal plates and displaying sensitivity towards ampicillin but not to chloramphenicol. Mutants required serine supplementation due to the araB insertion into the serA locus in the parent E. coli B serA::aroB strain (Table 6, entry 3). Attempted PCR verification of the mutants 53 revealed quite unexpected results. PCR with E. coli B serA::aroB as a template afforded a 1 kb sized DNA fragment on the agarose gel as was expected. However, the PCR product from E. coli B serA::araB AaroDzzFRT—cat—FRT and an E. coli W3110 AaraD::FRT—cat—FRT (transduced) was 2.2 kb, in contrast to the 1.2 kb. This indicated that there was an additional Ikb DNA fragment inserted together with the FRT—cat—FRT cassette. Removal of the antibiotic resistance gene from the mutants (Table 6, entry 5 and 6) with plasmid pCP20-encoded flippase was unsuccessful as they still showed resistance to chloramphenicol and a PCR product size of 2.2 kb. While a new strategy was formulated and was executed, E. coli B serA::araB AaroDzzFRT—cat—FRT host was evaluated under glucose-limited fermentor-controlled conditions. E. coli B serA::araB AaroDzzFRT—cat—FRT/pKD12.I 12 produced 4 g/ L of quinic acid in 5% yield over 60 h of cultivation under fermentor-controlled conditions (Table 7, entry I). The major hydroaromatic metabolite was 3-dehydroquinic acid, which accumulated to 43 g/L. The overall hydroaromatic yield was 22%. Very similar results were obtained for the E. coli B serA::aroB AaroD::FRT—cat—-FRT /pKD12.l38, which had plasmid localized-tktA, and synthesized 4 g/L of quinic acid together with 42 g/L of 3-dehydroquinic acid with an overall 18% yield of hydroaromatics in 60 h (Table 7, entry 2). The same concentrations of hydroaromatics indicated that transketolase overexpression was not necessary in a E. coli B host, as had been anticipated. E. coli B synthesis of quinic acid without transketolase overexpression showed even higher overall yield relative to overexpression of transketolase. Interestingly, both fermentations produced less biomass at about 22 g/L (Figure 17) relative to QP1.1/pKD12.l38, which produced approximately 60 g/L of dry cell weight (Figure 7A and Figure 25A). 54 Table 7. Concentrations and yields of products synthesized by quinic acid producing E. coli B strains under glucose-limited conditions. "m a b Entry Construct Tl h e, [DHQ] [3A], QA ,TOt:| g/L yield, °/o yield, % 1 E. coli B serA::aroB AaroD::FRT—cat—FRT 60 43 4 5 22 /pKD12.112 2 E. coli B serA::aroB AaroDzzFFiT-cat—FRT 60 42 4 6 18 /pKD12.138 3* E. coli B serA::aroB AaroD (new)::FRT—cat— 60 18 40 17 24 FRT /pKD12.112 4* E. coli B serA::aroB AaroD (new)::FFtT—cat— 60 13 41 19 25 FRT/pKD12138 5 * E. coli B serA::aroB AaroD (new)::FRT—cat— 84 28 37 10 18 FRT /pKD12.112 6 * E. coli B serA::aroB AaroD (new)::FRT—cat— 84 14 32 10 15 FRT /pKD12.138 7 E. coli B serA::aroB AaroD (new)::FFIT 84 7 22 10 13 /pKD12.112 8* E. coli B serA::aroB AaroD (new)::FRT 84 11 42 17 21 /pKD12.138 * Single run fermentation. aAbbreviations: DHQ — 3-dehydroquinic acid, QA - quinic acid. b (mol QA)/(mol glucose consumed). C(mol DHQ + mol QA)/(mol glucose consumed). .> in \l O 20 D. . 0 3.1]. .- 121824303642485460 121824303642485460 Time(h) Time(h) DHQ, QA, Dry cell weight (911.) 8 8 8 8 8 DHQ, QA, Dry cell weight (g/L) 8 D Figure 17. (A) E. coli B serA::aroB AaraD::F RT-cat—FRT /pKDlZ.112 and (B) E. coli B serA::aroB AaroDzzFRT—cat—FRT/pKD 12.138 cultured under glucose-limited conditions. Legend: (dotted bars) 3-dehydroquinic acid, (open bars) quinic acid, (black circles) dry cell weight. 55 The same strategy was repeated multiple times in order to obtain an E. coli B serA::aroB AaroDzzFRT—cat—FRT mutant and every time the same 2.2kb sized DNA fragment was obtained after PCR analyzis. This clearly indicated that it was not a spontaneous mutation, and suggested a fundamental problem with the way araD was being deleted. The PCR product from a couple of different mutants was sequenced with the same VI and V2 primers in order to determine why the PCR product size was 2.2 kb rather than 1.2 kb as had been expected. Technology Support Facility at Michigan State University. 1 51 101 151 201 251 301 351 401 451 501 551 601 ctgacaggct TAAAAQAICT cctatacttt GGCGCGCCTT .ATTAAGCATT GAATCGCCAG ATGGTGAAAA AAAACTGGTG CAATAAACCC TCTTGCGAAT CCAGAGCGAT GGTGAACACT AATTCCGGAT gaccgcgtgc CGTCATTGGT ctagagaata ACGCCCCGCC CTGCCGACAT CGGCATCAGC CGGGGGCGAA AAACTCACCC TTTAGGGAAA ATATGTGTAG GAANACGTTT ATCCCATATC GAGCATTCAT agaaagggta ACGthgtag ggaacttcgg CTGCCACTCA GGAAGCCATC ACCTTGTCGC GAAGTTGTCC AGGGATTGGC TAGGCCAGGT AAACTGCCGG CAGTTTGCTC ACCAGCTCAC CAGGCGGGCA aaaaATG_AAA W aataggaact TCGCAGTACT ACAAACGGCA CTTGCGTATA ATATTGGCCA TGAGACGAAA TTTCACCGTA AAATCGTCGT ATGGAAAACG CGTCTTTCAT Sequencing was done at the Research ACCGTAACTG cttcgaagtt tcATTTAAAT GTTGTAATTC TGATGAACCT ATATTTGCCC CGTTTAAATC AACATATTCT ACACGCCACA GGTATTCACT GTGTAACAAG TGCCATACGT Figure 18. Sequencing of the 5’ end of PCR product with V1 primer of E. coli B serA::aroB AaroD::FRT—cat—FRT 2.2 kb PCR product. Legend: atgc — araD upstream sequence; ATGC — araD; atgc — P1 or P2 on pKD3; atgc — FRT; ATGC -caL Sequencing results of the 5’ end of PCR product did not show any unexpected sequence. Small letters in Figure 18 represents E. coli DNA sequence upstream of araD gene with a 100% identity. This sequence is followed by 40 bp H1 sequence of araD (ATGC) and 20 bp of PI sequence from pKD3 template (a_tgg). The entire 48 bp FRT 56 sequence (atgc) was also present. The rest of the sequence (ATGC) corresponded to cat gene sequence responsible for the chloramphenicol resistance. Sequencing of the 2.2 kb 3’ end PCR product revealed different result (Figure 19). It had the araD downstream sequence (atgc) followed by the 40 bp H2 sequence of araD (ALGC) and 20 bp of P2 sequence from the pKD3 template (a_tgg). However, only 28 bp of the FRT (atgc) sequence was found and it was followed by an E. coli [55 transposase and transactivator seguence (nmpC), which was determined using BLSAT analysis. E. coli genomic sequence revealed that araD and nmpC sequences are 1.2 x 10" bp apart. Three mutants from different cloning attempts were submitted for sequencing analysis and all of them revealed identical results. Even though araD and nmpC share only 3% sequence identity, it was postulated that deletion of the entire araD ORF was problematic. 1 ttttttagtt 51 101 151 201 251 301 351 401 451 501 551 CCTGGTGTAA cttagttcct GGGGAAATTC CCGATTTTTT TTCGCCATTT ACGTATCATT CCAGTTGGTT CCGAACTGTC GCTGGCTTTC GCTGTTTCAA ACATCCACCT cggcggggag AATAGTTAAT attccgaagt TTCTCGGCTG CTCCCGTAAA AAGGCGTTAT TGGTCCGCCC ATCGTTTTTC GCTTGATGAT ATGTATTCGA GGTTCTTACC CGGCCAGCTC ggtgttcccg ACCGTGCGCA tcctattctc ACTCAGTCAT TGCCTTGAAT CCCCAGTTTT GAAACAGGTT AGCAACCCCT GCGAAATGGG TGTTGATGGC TTGCCGGGGC CTCGCGCTGT ccgaaatatt AATCAcatat tagGGAAGGT TTCATTTCTT CAGCCTATTT TAGTGAGATC GGCCAGCGTG TGTATCTGGC TGCTCCACCC CGTTTTGTTC GCTCGGCGAT GGCGCGCCTT attgcTTATG gaatatcctc GCGAATAAGC CATGTTTGAG AGACCGTTTC TCTCCCACTG AATAACATCG TTTCACGAAG TGGCCCGGAT TTGCGTGGAT CAGCCAGTCC GGTAGCCGGC Figure 19. Sequencing of the 3’ end of PCR product with V2 primer of E. coli B serA::aroB AaraDzzFRT—cat—FRT 2.2 kb PCR product. Legend: atgc - araD upstream sequence; ATGC — araD; atg — P1 or P2 on pKD3; atgc — FRT; ATGC - ISS transposase and trans-activator, DLP12 prophage, truncated outer membrane porin. 57 A new strategy was pursued were only half of the araD ORF was to be deleted. The same HIPI primer was used in the first step (Figure 16). The H2 sequence was chosen to start at the 439 nucleotide of the araD sequence (5’-CCGGTAAATAACT- CCAGATCGATCATATCAACCAGGCCGC-3’). E. coli W3110 AaraD(new)::FRT— cat—FRT was obtained successfully, and it was verified by correct growth characteristics on selective plates (Table 6, entry 7) and PCR. The newly constructed mutant required aromatics substitution to grow on glucose-minimal salts plates and it was sensitive to ampicillin but not to chloramphenicol. PCR analysis revealed the 1.6 kb sized DNA fragment on agarose gel. A successive PI phage-mediated transduction of aroD(new):: FRT-cat-FRT from E. coli W31 10 AaroD(new):: FRT-cat-FRT to E. coli B serA::aroB afforded the desired mutant of E. coli B serA::aroB araD(new):: FRT-cat-FRT. Growth characteristics of the new E. coli B mutant revealed the correct phenotype, requiring aromatics and serine supplementation when grown on glucose-minimal salts plates and sensistivity to ampicillin but not to chloramphenicol (Table 6, entry 9). However, elimination of FRT-cat-FRT cassette using pCP20 was unsuccessful after multiple trials. A solution to the removing the insert’s drug resistance was provided by Ingram’s group successful use of FLP flippase encoded by the pFT-A plasmid in E. coli 8.26 The key difference between pFT-A and pCP20 is that in pFT-A FLP expression is controlled by a chlorotetracycline inducible promoter rather than by a temperature inducible promoter as in pCP20. Plasmid pFT-A has a temperature sensitive replicon. Therefore it can be eliminated from the host by growing at 43 °C. E. coli B serA::aroB aroD(new):: FRT- cat-FRT/pFT-A were induced with chlorotetracycline and incubated for 6 h at 30 °C. This 30 °C, 6 h culture was then used to inoculate a culture incubated at 43 °C overnight. 58 Single colonies were obtained by the streaking overnight culture on LB plates. Phenotype screening of E. coli B serA::aroB aroD(new)::FRT revealed the correct growth pattern, with mutants requiring aromatics and serine supplementation when grown on glucose-minimal salts plates and sensitivity to chloramphenicol and ampicillin. PCR analysis with VI and V2 primers revealed 0.5 kb sized DNA fragment indicating a truncated araD gene. The wild-type araD gene size is 0.8 kb. Newly constructed E. coli B mutants were evaluated under glucose-limited fermentor-controlled conditions. The results were obtained after a single fermentor run for each mutant. E. coli B serA::aroB araD(new)::FRT-cat-FRT/pKD12.1 12 and E. coli B serA::aroB araD(new)::FRT-cat-FRT/pKD12.138 showed very similar cellular growth and metabolite accumulation (Figure 20A and Figure 208). Quinic acid was the major product and it accumulated to 40 g/L in 17% yield (Table 7, entry 3) and 41 g/L in 19% yield (Table 7, entry 4) over 60 h cultivation. E. coli B serA::aroB aroD(new)::FRT-cat- FRT/pKD12.1 12 without transketolase overexpression (Table 7, entry 3) and with transketolase overexpression E. coli B serA::aroB aroD(new)::FRT-cat-FRT/pKD12.138 (Table 7, entry 4) produced the same amounts of quinic acid in the same overall total hydroaromatics yield. This indicated that transketolase overexpression in E. coli B had a negligible impact on hydroaromatics biosynthesis. The biomass accumulation remained low at approximately 20 g/L for both fermentations. 3-Dehydroquinic acid accumulated to a 13 and 18 g/L concentration, which was higher relative to E. coli K-12 synthesis of quinic acid. The molar ratio between quinic acid and 3-dehydroquinic acid decreased from 11.5 for QP1.1/pKD12.l38 to 3.1 for E. coli B serA::aroB aroD(new)::FRT-car- FRT/pKD12.I38. As shown in Figure 20A and Figure 208, quinic acid concentration 59 kept increasing until the end of fermentor run and 3-dehydroquinic acid concentration stopped increasing. Therefore, quinic acid biosynthesis by newly constructed E. coli B strains was prolonged to 84 h (Figure 21). 3" ID V O ‘1 O 88888 Di-io, oA, Dry cell weight (91L) 8 8 8 8 8 8 DHQ, GA. Dry cell weight (911.) amt"??? . . ‘3, dllI 121824303642485460 121824303642485460 'nme(h) Time(h) 0 Figure 20. (A) E. coli B serA::aroB AaraD(new)::FRT-cat—FRT/pKD12.112 and (B) E. coli B serA::aroB AaraD(new)::FRT-cat—FRT/pKD12.138 cultured under glucose-limited conditions. Legend: (dotted bars) 3-dehydroquinic acid, (open bars) quinic acid, (black circles) dry cell weight. However, this did not help to increase quinic acid concentration in the final culture supernatant. Microbial E. coli B serA::aroB aroD(new)::FRT-cat- FRT/pKDIZ.l 12 synthesis of quinic acid over an 84 h cultivation resulted in synthesized 37 g/L of quinic acid in 10% yield (Table 7, entry 5). E. coli B serA::aroB aroD(new)::FRT-cat-FRT/pKD12.138 synthesized 32 g/L of quinic acid in 10% yield (Table 7, entry 6) when cultivated under the same conditions. 3-Dehydroquinic acid accumulated at higher levels during E. coli B cultivation under fermentor-controlled conditions (Figure 20 and Figure 21) relative to E. coli K-12 (Figure 7A and Figure 25). 60 A70 560 g... E 40 8 5‘30 . . O O 3 2° - Q o I— Y19"!1'1'ILI-II't—I Iii-.f r, f ' . '.. I 12182430364248546066727884 Time(h) B. A 7O 5 60 1:” .3 5o 3 R 40 s 30 - ” 8 2° ‘ ‘1’ ‘° o 0‘. ’Tn1’i11:;iT-’i,:s.:’ri:-,"." 1‘1" 12182430364248546066727884 Time (h) Figure 21. (A) E. coli B serA::aroB AaroD(new)::FRT-cat—FRT/pKDlZJ12 and (B) E. coli B serA::aroB AaraD(new)::FRT—cat—FRT/pKD12.l38 cultured under glucose-limited conditions. Legend: (dotted bars) 3-dehydroquinic acid, (open bars) quinic acid, (black circles) dry cell weight. The observed increase in 3-dehydroquinic acid concentration might be associated with the mechanism for 3-dehydroquinic acid export/import; from/ in to the cytoplasm. 61 If E. coli B has a slower transport system for importing 3-dehydroquinic acid from the culture medium back into the cytoplasm relative to E. coli K-12, then exported 3~ dehydroquinic acid will not be recaptured and reduced to quinic acid by E. coli B at the same rate relative to E. coli K-l2. Another reason for increased 3-dehydroquinic acid levels can be explained due to loss of glucose-limited condition control throughout the fermentor run. Glucose presence in the culture medium results in inhibition of the 3- dehydroquinic acid import system in E. coli, which will be discussed later in this chapter. Therefore, loss of glucose-limited control during E. coli B cultivation resulted in elevated levels of 3-dehydroquinic acid. This indicates that a better glucose-limited control conditions needs to be elaborated in order to reduce 3-dehydroquinic acid accumulation. The mutants with successfully removed chloramphenicol resistance were evaluated under glucose-limited fermentor-controlled conditions. E. coli B serA::aroB aroD(new)::FRT/pKD12.112 synthesized 22 g/L of quinic acid in 10% yield over 84 h and the total hydroaromatics yield was l3% (Table 7, entry 7). The quinic acid and 3- dehydroquinic acid molar ratio was low again (3.1). These results are an average of two runs. Interestingly, transketolase overexpression had a profound effect this time, and E. coli B serA::aroB aroD(new)::FRT/pKD12.l38 accumulated 42 g/L of quinic acid in l7% yield over 84 h with the total synthesized hydroaromatics yield of 21% (Table 7, entry 8). Further investigation is required in order to determine why after pFT—A treatment of E. coli B serA::aroB aroD(new)::FRT-c'at—FRT, the E. coli B serA::aroB aroD(new)::FRT/pKD12.l l2 (Table 7, entry 7) synthesized less quinic acid relative to the E. coli B serA::aroB aroD(new)::FRT-cat-FRT/pKD12.l l2 (Table 7, entry 5) and the 62 E. coli B serA::aroB aroD(new)::FRT/pKD12.l38 (Table 7, entry 8) when cultivated under the same conditions. N O 38888 DHQ, QA, Dry cell weight, (911.) m: Ehfié . Ti: .5 , 12182430364248546068727884 Time(h) O N O DHQ, QA, Dry cell weight (91L) 8 8 8 8 8 O 12182430364248546066727884 Time (h) Figure 22. (A) E. coli B serA::aroB AaroD(new)::FRT/pKD12.112 and (B) E. coli B serA::aroB AaroD(new)::FRT/pKD12.l38 cultured under glucose-limited conditions. Legend: (dotted bars) 3-dehydroquinic acid, (open bars) quinic acid, (black circles) dry cell weight. 63 Optimization of quinic acid production by E. coli QP1.1/pKD12.138 NT LN Previously discussed quinic acid production by E. coli QP1.1/pKD12.l38 (Table 4, entry l; Figure 7A) was run for 48 h and it afforded 58 g/L of quinic acid in 21% yield from glucose. As Figure 7 indicates, quinic acid concentration keeps increasing until the end of the fermentor run. Therefore, maybe a longer culture time would lead to higher quinic acid concentrations. Quinic acid fermentations were run under glucose-limited conditions for 60 and 132 h. QP1.1/pKD12.l38 produced 60 g/L of quinic acid in 21% yield in 60 h (Table 8, entry 2). This result is very similar to the 48 h fermentation (Table 4, entry 1). The concentration of 3-dehydroquinic acid remained the same at 5 g/L, and the total hydroaromatics yield was 22%. The standard fermentation inoculation conditions were also probed. Conventionally, inoculum for a fermentor is prepared by inoculating a single colony of interest into 5 mL glucose—minimal salts medium. This inoculum is incubated at 37 °C for 24 h. The next day, a 5 mL culture is transferred into 95 mL of fresh glucose-minimal salts medium and incubated at 37 °C for an additional l l h. Table 8. Concentrations and yields of products synthesized by quinic acid producing E. coli QP1.1/pKDlZ.l38 during various length of fermentation. Tim a A b Total Entry Inoculation conditions h e, [DHQ] [CS/L], QA , 0 g/L yield, % weld. % 1 Standard (24h 5mL, 11h 100 mL) 48 5 58 21 22 2* Standard 60 5 60 21 22 3* 17h5mL,11h100mL 60 3 55 19 2O 4 * Standard 132 3 73 20 21 5 Standard 84 5 67 18 19 * Single run fermentation. “Abbreviations: DHQ — 3-dehydroquinic acid, QA — quinic acid. b (mol QA)/(mo| glucose consumed). C(mol DHQ + mol QA)/(mol glucose consumed). 64 €70 $50 - ' : - i 40 .5 ...... [ ..... 530 e a ------ y- ---------- 5 20 i """""" “J 1 O 3.10 232 ::= """" "'4 """" - o a... measles . r ,. ,. 121824303642485460 Time(h) B. £70 2360 . ‘ e ' ’4 8’50 ~ 3 i = g .......... .. 8 4° - T e. ., ..... .... 330 <‘ .. ........... .. O 20 t 4 g 10 --------------------- ~— 1 o D‘Afi “jig? I” ' r” r If: IE 121824303642485460 Time (it) Figure 23. E. coli QP1.1/pKD12.l38 cultured under glucose-limited conditions: (A) 60 h and standard inoculation; (B) 60 h and non standard inoculation. Legend: (dotted bars) 3-dehydroquinic acid, (open bars) quinic acid, (black circles) dry cell weight. A new inoculation protocol was tested where a 5 mL culture was incubated for l7 h rather than 24 h, hoping that a fresher culture will lead to synthesis of a higher concentration of quinic acid. However, quinic acid accumulated at 55 g/L concentration in 19% yield after 60 h cultivation (Table 8, entry 3). Accumulation of 3-dehydroquinic acid remained very similar at 3 g/L and the total yield of synthesized hydroaromatics was 65 20%. Synthesis of hydroaromatics using the standard inoculation conditions and using the newly tested inoculation conditions can be compared in Figure 23A and Figure 23B. interestingly, initiation of the fermentor run with a fresh inoculant produced more biomass throughout entire fermentation process as compared to the standard inoculation conditions. The highest level of biomass was 59 g/L of dry cell weight using the fresh inoculant (Figure 238) compared to SI g/L of dry cell weight using the standard inoculant (Figure 23A). E. coli QP1.1/pKD12. I38 cultivation time was prolonged to 132 h. This time 73 g/L of quinic acid was synthesized in 20% yield (Table 8, entry 4). While more quinic acid was synthesized, the yield remained almost the same. Accumulation of 3- dehydroquinic acid also remained the same. The highest quinic acid production was achieved at approximately 102 h, when the quinic acid concentration reached 76 g/L. The highest dry cell weight was observed at 36 and 42 h at 53 g/L. Cultivation of E. coli QP1.1/pKD12.l38 for 84 h was also examined. It produced 67 g/L of quinic acid in 18% yield (Table 8, entry 5). Measured quinic acid concentrations varied between 60 and 84 h with and averag quinic acid concentration of 68 g/L. 66 3? \l O DHQ, QA, Dry cell weight (911.) 8 8 8 8 8 o .1—‘5 EhtB—l' I 12 24 B. A 70 _ _ ' _ 5% w '1 ....... «.4 .. ............ E .9 50 . lunf ----------- ....... in... ----------- = 40 _ ----- .4 ----- i ----- i ------------ .... ...... 8 g j z. 30 ............ . “4 ............................ 0 <5 20 .. . -. ........... , ............................. C . j. g" 10 . Efgir’w -- :3: ...... .. .---.. ..... o o [11 .-- F 1 12182430364248546066727884 Time (it) Figure 24. E. coli QP1.1/pKD12.l38 cultured under glucose-limited conditions: (A)l32 h; (B) 84h. Legend: (dotted bars) 3-dehydroquinic acid, (open bars) quinic acid, (black circles) dry cell weight. EEEEQI QE TRANSKE I QLASE QVEREXPRESfiIQE Transketolase overexpression in quinate synthesizing E. coli K-12 constructs was also investigated. Fermentation time was extended beyond 60 in order to capture the highest concentrations of quinic acid synthesized by the constructs. E. coli 67 QP1.1/pKD12.112 did not have plasmid-localized tktA and it synthesized 52 g/L of quinic acid over 84 h (Table 9, entry 2; Figure 25). This result is an average of two runs. it clearly indicates that transketolase overexpression is required for E. coli K-12 strains, because QP1.1/pKD12.l38 synthesized 67 g/L of quinic acid (Table 9, entry 1). Quinic acid yield also increased from 15%, without transketolase overexpression in QP1.1.pKD12.1 12 to 19% with transketolase overexpression in QP1.1/pKD12.l38. Table 9. Concentrations and yields of products synthesized by quinic acid producing E. coli K-12 strain with and without transketolase overexpression. Time, a [QA], b Total Entry inoculation conditions h [DHQ] 9 QA . 0 g/L yield, % yield, % 1 QP1.1/pKD12.138 84 5 67 18 19 2 QP1.1/pKD12.112 84 2 52 15 15 “Abbreviations: DHQ — 3-dehydroquinic acid, QA — quinic acid. b (mol QA)/(mol glucose consumed). C(moi DHQ + mol QA)/(mol glucose consumed). €70 ‘160 43.. -. $40 0 ‘ . T1 ‘— f . $ - t ........................... E - 1 030 520 . "1+«~ ...................... “i ............ . ..... $10 . ................. . ......................................... D o _r‘[1r' T. I. ‘. I. r r-. t '3' IE?‘ 12182430364248546066727884 Time(h) Figure 25. E. coli QP1.1/pKD12.112 cultured under glucose-limited conditions. Legend: (dotted bars) 3-dehydroquinic acid, (open bars) quinic acid, (black circles) dry cell weight. 68 E EN N PE The detailed description of the composition of fermentation medium is described in Chapter 4. The key ingredients are potassium phosphate, required to maintain buffer conditions and also as a phosphorous source for the bacteria. Aromatic amino acids are required as a supplement due to an inactivated shikimate pathway. Therefore, the concentration of aromatic amino acids added in the medium can control the biomass and ultimately influence quinic acid synthesis. Three precursors for aromatic vitamin synthesis p-hydroxybenzoic acid, p-aminobenzoic acid, and 2,3-dihydroxybenzoic acid are also required as supplements. Ammonium iron citrate is required as the iron source and citric acid is used as metal ion chelator, because E. coli does not have cataboiic pathway for citric acid degradation. Table 10. Concentrations and yields of products synthesized by quinic acid producing strain QP1.1/pKD12.l38 with various phosphate concentration in the medium. Time, QA , b Total Entry [KgHPO4] h [DHQ] [ 1 QA , c g/L yield, % yield, % 1 43 mM (standard) 60 5 60 21 22 2* 35 mM 60 5 62 22 23 3 * 20 mM 60 3 51 21 22 * Single run fermentation. “Abbreviations: DHQ — 3-dehydroquinic acid, QA — quinic acid. b (mol QA)/(mol glucose consumed). C(mol DHQ + mol QA)/(mo| glucose consumed). Separation of inorganic salts from quinic acid fermentation broth is an important step in purification of quinic acid, which will be described in detail later in this chapter. Reduction of unconsumed potassium phosphate could lead to an easier process for isolation and purification of quinic acid. Various potassium phosphate concentrations were investigated for cultivation of QP1.1/pKD12.l38 under fermentor-controlled, glucose-limited conditions. 69 P A 70 @650 - - T ----- H .c: g 50 . l ----- 4. ...... =3 40 a ..... 5 m1. 8 1 g 30 e """ <' 20 °. . 2 1° *' ------ ”-1 0 o.,....._.:%? .- .. r1 12 18 24 30 36 42 48 54 60 Time(h) B A 70 5 60 ‘5 -. .g 50 _ r E 40 . Q ..... 7 e E 30 . ............ 5' 2° ------ 4 $.10“? .. 0 on. .F ['1 'f" .1" 121824303642485460 Time (h) Figure 26. E. coli QP1.1/pKD12.l38 cultured under glucose-limited conditions in: (A) 35 mM KZHPO4; (B) 20 mM K2HPO4 medium. Legend: (dotted bars) 3- dehydroquinic acid, (open bars) quinic acid, (black circles) dry cell weight. A standard potassium phosphate concentration was 7.5 g/L (43 mM) of KZHPO4. Quinic acid was synthesized at 62 g/L concentration in 22% yield in 35 mM KZHPO4 medium (Table 10, entry 2) and was approximately the same relative to the standard cultivation conditions (Table 10, entry 1). Therefore, a 60 h quinic acid synthesis by E. coli QP1.1/pKD12.l38 could be efficiently achieved in 35 mM phosphate medium. Quinic acid production declined from 60 g/L for the QP1.1/pKD12.l38 cultivated in the standard 70 43 mM KZHPO4 medium (Table 10, entry 1) to 51 g/L cultivated in 20 mM KZHPO4 cultivation medium (Table 10, entry 3). Reduced phosphate concentrations in the medium resulted in reduced biomass (Figure 27 A). interestingly, fermentations reached stationary phase at approximately the same time (30 h), but the cell growth rate during exponential phase (0 — 30 h) was higher under reduced phosphate concentrations than compared to the QP1.1/pKDlZ.l38 cultivation under standard conditions (Figure 27). Dry cell weight (g/L) Quinic acid (g/L) A. so 50‘ 40. aoi 20? 10E 0 o . 0 CD 0130 01:10 0 CD 0 OD CD 0 12 18 24 30 36 42 Y r Time (h) 4854 70 60* 50. 40* 301 201 10‘ O as? 0 K3 C] O GOD COD 1:] G O I T r Y 121824303642485460 Time(h) Figure 27 . Comparison of the impact of phosphate concentration on E. coli QP1.1/pKD12.l38: (A) dry cell weight; (B) quinic acid synthesis. Legend: (diamonds) 43 mM KZHPO4, (squares) 35 mM KZHPO4 and (circles) 20 mM KZHPO4. 71 Aromatic amino acid supplementation was another investigated culture medium component. The concentration of aromatic amino acids was reduced by 25%, 50% and 75% relative to the typically employed concentration. The results are summarized in Table 11 and shown as a single entry and not as an average value, because the synthesized metabolite concentrations were widely dispersed even for the same concentration of aromatic amino acids. The main reason for high result dispersion was difficulty in controlling glucose—limited conditions. Reduced concentration of aromatic amino acids translated into reduced biomass formation (Figure 28 A). Since at lower concentrations of aromatic amino acids there was less biomass, use of p02 control to maintain glucose-limited conditions was more difficult. The max stir speed varies depending on fermentor set up in order to maintain the same oxygen transfer rate (OTR). Fermentors were previously calibrated for OTR and they revealed that under the same conditions (volume of culture and airflow) the stir rate has to be adjusted differently in order to obtain the same OTR. The small differences in fermentor vessel shape, relative location of sparger and impellers have major effect on OTR. Due to this reason, the maximum stir rate to maintain the same OTR for each experiment in Table 1 1 is different and listed in parentheses. Cultures were purged with l vvm (volume of airflow per volume of culture medium per minute) of airflow under standard cultivation conditions. 72 Table 11. Concentrations and yields of products synthesized by quinic acid producing strain QP1.1/pKD12.l38 with various aromatic amino acid concentrations in the medium. - Entry Aromatic amino Al‘ll'stnW, Stir (gar: stir), Tirae, [DH 01:8 [32:] Q Ab .Totgl acud conc. g/L yield, % yield, % 1 100% 1.0 1100 (1100) 84 5 67 18 19 2* 25% reduced 1.0 1000 (1000) 60 23 34 13 23 3* 25% reduced 1.0 940 (940) 84 13 50 14 17 4* 50% reduced 0.5 1000 (1000) 60 31 31 13 26 5* 50% reduced 1 .0 740 (940) 84 5 58 17 18 6* 50% reduced 0.5 940 (940) 84 16 40 14 20 7* 75% reduced 0.5 750 (1100) 84 8 43 18 22 8* 100% increased 1 .O 1 100 (1 100) 84 4 64 18 19 9* 15 g/L yeast 1.0 1100 (1100) 84 7 68 18 20 extract * Single run fermentation. aAbbreviations: DHQ — 3-dehydroquinic acid, QA — quinic acid. b (mol QA)/(mol glucose consumed). C(mol DHQ + mol QA)/(mol glucose consumed). Note: aromatic amino acids 100% are: 0.7 g/L of L-tyrosine, 0.35 g/L of l.- tryptophane, 0.7 g/ L L-phenylalanine. 60 00. ,t 50 J ‘0 0’ °’ 0» .. 52 0 9.800.? 3540‘ ofli' '3 c .Q’ I -—"'UU $30 [:1 .. ca E 8I O AAAA 520 DXAA AAAA 10+2 I 03' 182430364248546066727884 Time(h) d '0 Figure 28. Comparison of the impact of aromatic amino acid concentrations on E. coli QP1.1/pKD12.138 dry cell weight. Legend: (diamonds) 100%, (open circle) 25% reduced, (solid circle) 25% reduced, (solid square) 50% reduced, (open square) 50% reduced, (dashes) 50% reduced, (triangular) 75% reduced. 73 E. coli QP1.1/pKD12.l38 produced 34 g/L (Table 11, entry 2) in 60 h and 50 g/L (Table 11, entry 3) of quinic acid 84 h when aromatic amino acid concentration were reduced by 25% in culture medium. Accumulation of 3-dehydroquinic acid was observed at higher levels of 23 g/L (Table l 1, entry 2) and 13 g/L (Table 11, entry 3). The relatively high 3-dehydroquinic acid concentration can be explained by imperfect maintenance of glucose—limited culture conditions, because glucose concentration rose to 4 g/L towards the end of the fermentation. By contrast, a typical glucose concentration for glucose-limited culture conditions is bellow 0.03 g/L. Both fermentations were run under standard conditions (i vvm and OTR-optimized stir rate), but it was already obvious that p02 response is slower than during the standard fermentation conditions. Cultures with a 50% reduced in the concentration of aromatic amino acids had to be run with either 0.5 vvm airflow (Table l 1, entry 4 and entry 6) or reduced stir rate (Table 1 1, entry 5) in order to maintain glucose-limited conditions. Due to better maintenance of glucose-limited culture conditions, QP1.1/pKD12.l38 synthesized higher quinic acid concentration at 58 g/L (Table l 1, entry 5) and 40 g/L (Table 11, entry 6) relative to 31 g/L (Table 1 l , entry 4). Accumulation of the major byproduct 3-dehydroquinic acid at 16 g/L (Table 11, entry 6) was related to the glucose concentration in the medium. At the same time, quinic acid was produced at lower concentration relative to the control experiment (Table l 1, entry 1). Reduction by 25% of aromatic amino acid concentration, yielded even lower biomass accumulation. However, quinic acid concentrations remained approximately the same at 43 g/L (Table 11, entry 7). Reduced biomass ultimately resulted in reduced airflow and lower stir rates in order to maintain glucose- limited culture conditions. Interestingly, reduction in aromatic amino acids by 25% and 74 50% resulted in approximately the same biomass accumulation, while a 75% reduction in aromatic amino acids had a major impact on the biomass formation (Figure 28A). The final concentration of synthesized quinic acid with different levels of aromatic amino acid supplementation varied from 40—50 g/L. The general trend observed was that reduction in aromatic amino acid supplementation reduced biomass formation and cultures’ oxygen requirement. Therefore, in order to establish reliable and reproducible fermentation conditions, more fermentor runs will be required to screen for optimal airflow and impeller speed. The effect of a twofold increase in aromatic amino acid concentration in the culture medium was also examined. initially standard fermentation conditions were used with the normal concentration of aromatic amino acids in the culture medium. Once cells reached the end of the exponential growth phase at approximately 18 h after the beginning of fermentation an additional 50% of the normal concentration of aromatic amino acids were added twice in 6 h intervals, which led to the final twofold increase in the concentration of aromatic amino acids in the culture medium. This incremental addition was done in order to avoid the feedback inhibition of plasmid-localized aroFFBR by L-tyrosine.9 Quinic acid accumulated to 64 g/L in 18% yield over 84 h (Table 11, entry 8). The highest quinic acid concentration of 66 g/L was reached at 78 h (Figure 29 A). Accumulation of 3-dehydroquinic acid remained at a low level (4 g/L) and the total yield of synthesized hydroaromatics was the same at 19% (Table 11, entry 8) as for the control experiment (Table 11, entry 1). The increase in aromatic amino acids did not have any effect on total yield on synthesized hydroaromatics. 75 A 70 5; 60 - ...... r .......... F E _ F .g’ 50 ------ - --------------- 3 g . m 0 i L § 40 , ----- . -. r a. 30 i ----------- .. ...... .. a ..... ..-. L. D e i <1“ 20 ----- + r ------------ --- ~--« 1 O O" 10 . ..... .... ..... , ........... . ..... .... g o._Ln.'n—l . .' .' T. r 'u ‘F 'E ‘1" 12182430364248546066727884 Time (h) B. A 70 _ _ — g r : 60 mrm -- .C a: J"... i ..... - ......... .. 5o . ‘3’ o 0 . " i ii (i ii = 40 . ..... p. .. ... 4 1.. -i --«-i - ..... a , 7 J g 30 ... ............... .. ..... - ..- - 5" 20 ...... a E i .3 - e g“ 10 D o . 12182430364248546066727884 Time(h) Figure 29. E. coli QP1.1/pKD 12.138 cultured under glucose-limited conditions with: (A) twofold increased aromatic amino acid; (B) 15 g/L ' yeast extract supplementation. Legend: (dotted bars) 3-dehydroquinic acid, (open bars) quinic acid, (black circles) dry cell weight. A positive effect of shikimic acid biosynthesis during glucose-rich, fermentor-controlled . . . . . 3 conditions With yeast extract supplementation was prewously reported.l Chandran et. a1. 76 reported that shikimic acid titter and yield increased from 62 g/L and 26% to 84 g/L and 33%.13 Yeast extract supplementation was also investigated for quinic acid biosynthesis, where 15 g/L of yeast extract was added in the beginning of the fermentation. The final synthesized quinic acid concentration was 68 g/L in 84 h (Table 11, entry 9). It was synthesized in 18% yield with 20% total yield of synthesized hydroaromatics. Accumulation of 3-dehydroquinic acid was observed at 7 g/L at the end of the fermentation. Although glucose concentration was not detectable during the fermentor run, 3-dehydroquinic acid concentration kept increasing during the first half of the fermentor run to 21 g/L (Figure 29 B). Apparently, yeast extract had the same effect as glucose and caused 3-dehydroquinic acid accumulation. ‘1 O i»- H. -i {>0}. DOD .9013 .1 [>03 D C? CD 1111 m 0 I DUO >1. <95 :30 0H5-.. .. 12182430364248546066727884 Time (h) Figure 30. Comparison of the impact of twofold increased aromatic amino acid concentration and 15 g/L yeast extract supplementation in the culture medium on dry cell weight and quinic acid synthesis by E. coli QP1.1/pKD 12.138 under glucose- limited culture conditions. Legend: (open squares) dry cell weight standard conditions, (open triangular) dry cell weight twofold increased aromatic amino acids, (open circles) dry cell weight 15 g/L yeast extract supplementation, (solid squares) quinic acid standard conditions, (solid triangular) quinic acid twofold increased aromatic amino acids, (solid circles) quinic acid 15 g/L yeast extract supplementation. QA, Dry cell weight (g/L) 8 B 8 8 8 8 77 interestingly, the quinic acid accumulation profile for the standard fermentor run (solid squares, Figure 30) was identical with the yeast extract supplemented fermentor run (solid circles, Figure 30). Twofold increase in aromatic amino acid supplementation (solid triangular, Figure 30) probably caused DAHP synthase inhibition early in the fermentor run. Therefore, quinic acid synthesis in the first 30 h was slower, but seemed to resume to a normal rate after 36 h. However, the total amount of synthesized quinic acid did not catch up with the standard fermentation. Yeast extract supplementation resulted in increased cell biomass (open circles, Figure 30) early in the fermentor run. However, after 48 h it reached the same level as the standard fermentation (open squares, Figure 30) and it remained the same throughout fermentor run. Counter intuitively, a lower biomass profile was observed for the twofold aromatic amino acid (open triangular, Figure 30) supplementation. " {E0 {111W - D Nil 0 _-, U ._,l 0 l O, y. Under glucose-limited, fermentor-controlled conditions, time-dependent the change in concentration of 3-dehydroquinic acid during microbial synthesis of quinic acid suggested that 3-dehydroquinic acid could be exported outside the cell prior to the reduction step and imported back into the cytoplasm (recaptured) with subsequent reduction to afford quinic acid (Figure 7, Figure 24A, Figure 248, and Figure 26). To test this hypothesis, Ran constructed E. coli QP1.1/pNR4.276 incapable of de novo quinic acid synthesis from glucose.ll The construct had inactivated plasmid-localized DAHP FB . . . . . synthase aroF R together With other shared genomic elements for qumic aCld production. Genomic expression of all three DAHP synthase isozymes (AroF, AroG and 78 AroH) was inhibited by adding aromatic amino acids to the culture medium. This ensured no carbon flow into the shikimate pathway (Figure l). The construct was cultured under glucose-limited conditions and no formation of quinic acid or 3- dehydroquinic acid was observed in the culture medium. Addition of 5 g/L of 3- dehydroquinic acid resulted in formation of 2.5 g/L quinic acid with 2.1 g/L of 3- dehydroquinic acid remaining after 30h under glucose-limited, fermentor-controlled conditions. Another way to test whether initially formed 3-dehydroquinic acid is recaptured over the course of microbial synthesis of quinic acid is to initially run the microbial synthesis under glucose-rich culture conditions and then switch to glucose- limited culture conditions. Metabolite recapture would be implicated if 3-dehydroquinic acid synthesized under glucose-rich culture conditions decreases in concentration and is replaced by quinic acid when the culture is switched to glucose-limited culture conditions. interestingly, for the most times when 3—dehydroquinic acid formation was observed at higher concentrations, the total yield of synthesized hydroaraomtics increased for E. coli K-12 QP1.1/pKD12.138 (Table 11, entry 2 and entry 4) and for E. coli B serA::aroB aroD(new):: FRT-('at-FRT/pKDIZJIZ (Table 7, entry 3), E. coli B serA::aroB aroD(new):: FRT-('at-FRT/pKD12J38 (Table 7, entry 4) relative to the standard 19-22% yield obtained by QP1.1/pKD12.l38 (Table 8, entry 5 and entryl). Therefore, maybe fermentor controlled culture conditions where 3-dehydroquinic acid is first formed under glucose-rich conditions with subsequent conversion to quinic acid under glucose-limited conditions will lead to higher quinic acid concentration and yields. 79 Table 12. Concentrations and yields of products synthesized by quinic acid producing strain QP1.1/pKD12.l38 under various glucose fermentation conditions. Time, a QA , b Total Entry Fermentation type h [DHQL [ 1 QA . 0 g/L yield, °/o yield. % 1* Glucose-limited 132 3 73 18 19 2* Glucose-rich 66h" 132 8 48 13 15 3* Glucose-rich 36h" 84 16 4O 10 14 * Single run fermentation. H Fermentor was run under glucose-rich culture conditions for 66 h or 36 h and then switched to glucose-limited culture conditions. “Abbreviations: DHQ - 3-dehydroquinic acid, QA — quinic acid. b (mol QA)/(moi glucose consumed). C(mol DHQ + mol QA)/(moi glucose consumed). E. coli QP1.1/pKD12.l38 was cultivated under standard glucose-rich culture conditions for 66 h and then switched to glucose-limited culture conditions. The total synthesized quinic acid concentration after 132 h was 48 g/L in 13% yield, while 3- dehydroquinic acid accumulated at 8 g/L (Table 12, entry 2). The total yield of synthesized hydroaromatics was 15%, which is lower than the 19% observed during the control experiment (Table 12, entry 1). The concentration of 3-dehydroquinic acid increased until 36 h and then started a gradual decline simultaneous with an increase in quinic acid concentrations (Figure 31A). Even though the culture medium had approximately 10 g/L of glucose, the rate of 3-dehydroquinic acid accumulation in the culture medium began to decrease after 36 h. Once the fermentor-controlled culture was switched to glucose-limited culture conditions (after 66 h) a steady decline in the concentration of 3-dehydroquinic acid was observed, but this did not correlate with a comparable rate in the increase in concentration of quinic acid. This experiment was revisited with another fermentor run where glucose-rich conditions were changed to glucose-limited conditions after 36 h. This time, 40 g/L of quinic acid was synthesized in 10% yield over 84 h (Table 12, entry 3). The total yield of synthesized hydroaromatics 8O was 14%. As with the previous experiment, where the concentration of 3-dehydroquinic acid rose faster than that of quinic acid under glucose-rich conditions (until 42 h, Figure 31 B). After the culture was switched to glucose-limited conditions, the concentration of 3-dehydroquinic acid declined faster than the concentration of quinic acid increased. A. _ 70 g: so 3... 355 5° . . a 40 , M.,, 0% o it 56 3° 'i’ O" 20. ........ . "‘*'"*H I ii 0 10 . ; Mr 0.;-1,1311}, Em 12 24 132 DHQ, QA, Dry cell weight, Glucose (gIL) 8 8 8 8 12182430364248546066727884 Time (h) Figure 31. E. coli QP1.1/pKD12.l38 cultured under glucose-rich conditions for (A) 66 h and (8)36 h and subsequently switched to glucose-limited culture conditions. Legend: (dotted bars) 3-dehydroquinic acid, (open bars) quinic acid, (black bars) glucose, (black circles) dry cell weight 81 80 [3 El [31:1 70 D DECIDE at: D :1 50 E] €50 D A AAA/31315.94) :40 AAA-00C 030 [:1 A 00 A o 20 ADD C10 10 éé oué..............e... 12 24 36 48 60 72 84 96108120132 Time(h) Figure 32. Quinic acid accumulation profiles obtained during cultivation of E. coli QP1.1/pKD12.l38 under glucose-rich culture conditions and subsequently switched to glucose-limited culture conditions. Legend: (squares) glucose-limited culture conditions, (triangular) glucose-rich culture conditions until 66 h, (circles) glucose-rich culture conditions until 36 h. These experiments indicate that it is very important to maintain strictly glucose—limited culture conditions during the first 60 h of microbial quinic acid synthesis, because the rate of quinic acid formation is the highest during glucose-limited conditions (squares, Figure 32) and the rate of quinic acid synthesis does not recover to the same rate after exposure to glucose-rich culture conditions (triangular and circle, Figure 32). A loss of glucose-limited culture condition control towards the end of the fermentation process does not have a drastic impact since the rate of quinic acid accumulation is restored to the normal rate once the residual glucose is consumed from the culture medium. Even though the total concentration and yield of microbe-synthesized quinic acid was lower than for the control experiment, it was demonstrated that de novo synthesized and exported 3—dehydroquinic acid can be later transported back into the cytoplasm and reduced to quinic acid. Even under glucose-rich culture conditions, quinic acid 82 concentration kept increasing (Figure 32), which indicates that a portion of 3- dehydroquinic acid gets reduced to quinic acid prior to export. However, recapture of 3- dehydroquinic acid under glucose-limited conditions is also important in quinic acid biosynthesis. The hydroaromatics transport system in E. coli was not yet been identified. An attempt to identify such a transport system will be described in Chapter 3. Hydroaromatics transport in E. coli may be an evolutionary remnant from which a progenitor microbe to E. coli exploited 3-dehydroquinic acid and quinic acid as a sole source of carbon for growth and metabolism. Klebsiella pi-zeumoniae, which is evolutionary closely related to E. coli, is capable of exploiting quinic acid as sole carbon . 4 . . . source for growth and metabolism. The glucose concentration in the culture medium can control transport of hydroaromatics, since it would be expected that glucose would inhibit other carbon source uptake pathways. Therefore, under glucose-rich culture conditions, the recapture of 3-dehydroquinic acid is slows with the highest recapture rates apparently realized under glucose-limited culture conditions. For quinic acid-producing E. coli, a remnant hydroaromatic transport pathway is beneficiary, since it is the combination of 3-dehydroquinic acid reduction prior to export and after recapture that affords the highest concentrations and yields of microbe-synthesized quinic acid. Quinic acid purification Purification of quinic acid from fermentation broth faces a couple of challenges. First, there is the need to separate away inorganic salts. Secondly, there are biosynthetic byproducts that must be removed. Cell-free culture medium was obtained by centrifugation of crude culture medium at 10,000 g for 15 min. Protein was removed by acidification of cell-free culture medium 83 to pH 3 with concentrated H2504 with stirring at low temperature (0 or 4 °C) for at least two hours. Precipitated proteins were filtered through Whatman No.1 filter paper. No filtering agent, such as Celite, was required, because proteins coagulated in large pieces. Alternatively, refluxing of the acidified cell-free culture medium was used to separate protein and aromatize 3-dehydroquinic acid simultaneously. All proteins precipitated after refluxing. However some adhered tightly to the sides of the flask. Refluxing neutral cell-free culture medium did not result in separation of protein from the cell-free culture medium. industrially, ultrafiltration of cell-free culture medium is likely a preferred route for protein removal. Accordingly, ultrafiltration of cell-free culture medium was primarily used elaboration of a purification scheme for microbe-synthesized quinic acid. Ultrafiltration employed Millipore 10 kDa membrane to separate proteins from the cell- free culture medium. Previously by the Frost group it was demonstrated that protocatechuic acid could be obtained in a 100% conversion yield from 3-dehydroquinic acid by refluxing culture medium containing 3-dehydroquinic acid (Figure 33).27 Therefore, refluxing of cell-free, protein-free culture medium was used to aromatize 3-dehydroquinic acid, which was the major byproduct in quinic acid microbial synthesis. Neutral or cell-free, protein-free culture medium acidified to pH 3 was refluxed for i h to aromatize 3-dehydroquinic acid. Aromatized byproduct was then removed by adsorption on activated carbon. Treatment with activated carbon was used to decoiorize the dark brown cell—free, protein-free culture medium after aromatization of 3-dehydroquinic acid. Activated carbon was added to cell-free, protein-free culture medium at room temperature and the mixture was 84 stirred for 1.5 h. After filtration through a Ceiite pad, the resulting solution was pale yellow. Ho, COZH 002” ' culture medium 0 _ OH reflux HO OH OH 3-dehydroquinic protocatechuic aCId acid Figure 33. Conversion of 3-dehydroquinic acid to protocatechuic acid. A pH effect was observed during treatment with activated carbon. Cell-free, protein free culture medium at pH 3 required 2% (w/v) of activated carbon (20 g of charcoal for l L of broth), while at pH 7 7-8% of activated carbon was required to obtain the same pale yellow cell-free, protein-free culture medium. After treatment with activated carbon at pH3 broth, a small portion of this solution was neutralized and the pale yellow color remained. Therefore, decolorization was not dependent on maintenace of an acidic pH. Grades of activated carbon employed included: activated carbon Norit® 3A3 100 mesh and Darco® KB 100 mesh. Decolorization required 2% (w/v) of Norit charcoal, while 4% (w/v) of Darco charcoal was required to reach the same decolorization level. For elaboration of all subsequent steps in the purification of microbe-synthesized quinic acid, Norit® activated carbon was used. Separation of quinic acid from inorganic salts in the culture medium was the most challenging step. Organic solvent extraction of quinic acid could not be used since quinic acid has very low solubility in organic solvents. A negative adsorption method was examined where a strong anion exchange resin (AG1-X8) and strong cation exchange 85 resin (Dowex 50 WX4) were used to remove inorganic salts from the cell-free, protein- free, decolorized culture medium. The resulting salt-free solution was evaporated to dryness and quinic acid was recrystalized from ethanol, yielding 18% final purification yield based on the quinic acid originally present in the culture medium. in order to remove inorganic salts from 1 L of cell-free, protein-free, decolorized culture medium, 1 L of cation exchange resin and 1 L of anion exchange resin were required. This was unlikely to be an industrially viable process. Table 13. Quinic acid purification from microbial culture medium. Volume [GA] GA Yield Step Description (L) (g/L) (g) (°/o) 1 Ultrafiltration of cell-free broth through 10 kDa 1.20 33 40 membrane 2 Reflux 1 h, acidify cell-free, protein-free culture 0.96 40 38 95 medium to pH 3, 4% (w/v) charcoal treatment 3 Concentrate fivefold, add three volumes of EtOH, 0.82 46 38 95 remove inorganic salt precipitate 4 Concentrate to dryness, dissolve in boiling EtOH, 1.18 30 35 88 filter insoluble precipitates 5 Collect QA that precipitated after 30 min at room 22.1 55 st temperature, (1 crop). chill at 4 °C overnight nd Collect QA precipitate (2 crop). Concentrate fourfold, chill at 4 °C overnight rd 7 Collect QA precipitates (3 crop) 5.6 14 8 Total QA recovered 34.1 85 After screening of numerous methods, the final quinic acid purification was obtained (Table 13). After filtration of ethanol precipitated inorganic salts, the resulting filtrate was concentrated to dryness and the residue was redissolved in 1 L of boiling ethanol. Solids remaining in solution were removed by filtration. Partial concentration of quinate-containing filtrate led to formation of the first crop of quinic acid precipitate and was filtered. Two more rounds of concentration, chilling and precipitate collection were performed. Overall, 85% of quinic acid in the starting culture medium was isolated. 86 The quality of quinic acid was determined by elemental analysis (Table 14). The third crop of quinic acid precipitates was not submitted for elemental analysis, due to its pale yellow color. Table 14. Quinic acid elemental analysis. expt. (cald.) entry sample (%) C H 1 St 43.79 (43.75) 6.35 (6.29) 1 crop 2 nd 43.50 (43.75) 6.53 (6.29) 2 crop 3 Aldrich 43.41 (43.75) 6.61 (6.29) First crop of purified quinic acid was also evaluated by 1H and ”C NMR analysis with D20 as a solvent and TSP as an internal standard. Proton NMR (Figure 34) revealed correct quinic acid peaks as they were compared to literature reported values and standard NMR, which was obtained from quinic acid sample purchased from Aldrich. Proton NMR also revealed that purified quinic acid had some residual EtOH, which methionine triplet can be seen at 1.15 ppm and methilyne quadruplet can be seen at 3.61 ppm (Figure 34). Also, a triplet at 1.25 ppm and a quadruplet at 4.21 ppm corresponds to methionine and methilyne of ethyl group in quinic acid ethyl ester (Figure 34), which was obtained during purification of quinic acid in step 4 (Table 13). Carbon NMR (Figure 35) analysis revealed quinic acid peaks which were identical to literature reported values and to standard quinic acid NMR. 87 Figure 34. 1H NMR of purified 1St crop of quinic acid. 88 El ._'_ ._..._.._ i _ .,.__-._....__. ., ._ . i . l Figure 35. 13C NMR of purified lSt crop of quinic acid. 89 iT TITT‘I'TiilITT—T ir|iliiiTT‘lllT—TTT‘WFY—TT—TT—TWFT-I l TITTi—WTTITlTIIiTTITTFTTI TT'TliijiiTiiTiT it TIT; 180 .itiiriitii. 91'1”.”“11‘. .11 ' irrTi‘v 1 1‘0 120 100 80 60 40 20 160 220 200 240 Discussion The first reported microbial synthesis of quinic acid relied on heterologous expression in E. coli of the Klebsiella pneumoniae qad gene, which encodes quinate 3 . . . . . dehydrogenase. in a later variant a native E. coli aroE-encoded shikimate dehydrogenase in an E. coli aroD mutant lacking 3-dehydroquinate dehydratase enzyme activity, resulted in an E. coli construct, which synthesized quinic acid from glucose.D Homologous overexpression of all genetic elements avoids the differences between K. pneumoniae and E. coli” in codon usage, promoter strength and protein folding. More recently, another E. coli shikimate/quinate dehydrogenase encoded by ydiB locus was . l8 . . . . . . discovered. Disruption of the genomic ydiB sequence in E. coli construct that synthesized shikimic acid resulted in complete elimination of byproduct quinic acid formation during synthesis of shikimic acid. This clearly indicated that YdiB is involved in quinic acid biosynthesis. Therefore, overexpression of plasmid-encoded ydiB rather than aroE as a shikimate/quinate dehydrogenase was evaluated. However, E. coli QP1.1/pJJ5.069 and QP1.1/pJJS.073 failed to produce quinic acid at the same concentration and yield as the control strain relaying on aroE overexpression (Table 4). The primary reason was low YdiB specific activity (Table 5). Deletion of the entire ORF of the araD gene was unsuccessful in E. coli B and only half of aroD was deleted in the final construct, which yielded E. coli B capable of synthesizing quinic acid. However, the E. coli B quinic acid producers (Table 7) did not synthesize quinic acid in as high a concentration and yield as quinic acid-synthesizing constructs based on E. coli K-12 (Table 4, entry 1). Plasmid-localized transketolase 90 overexpression had no effect on synthesized hydroaromatics concentrations by E. coli B serA::aroB AaroD(new)::FRT—cat—FRT/pKD12.112 (Figure 20A) relative to E. coli B serA::aroB AaroD(new)::FRT—cat—FRT/pKD12.138 (Figure 208). On the other hand, transketolase overexpression had an impact on biomass accumulation in the beginning of cultivation, where from 0 — 24 h of cultivation of E. coli B serA::aroB AaroD(new)::FRT—cat—FRT/pKD12.138 cells grew slower with transketolase overexpression (Figure 20B) as compared to the E. coli B serA::aroB Aar0D(new)::FRT— cat—FRT/pKD12.1 12 without transketolase overexpression (Figure 20A). Overexpression of transketolase did not have an impact on biomass production E. coli K- 12 QP1.1/pKD12.l38 (Figure 248) relative to E. coli K-12 QP1.1/pKD12.112 (Figure 25) as they accumulated the same biomass throughout the cultivation. interestingly, the total of synthesized the hydroaromatics, quinic acid and 3-dehydroquinic acid, by E. coli B serA::aroB aroD(new)::FRT-cat-FRT/pKD12.112 was 65 g/L in 18% yie1d(T able 7, entry 5), while E. coli K-12 QP1.1/pKD12.112 synthesized 54 g/L in 15% yield(Table 9, entry 2) and QP1.1/pKD12.138 synthesized 72 g/L in 19% yield (Table 9, entry 1). This indicates that E. coli B serA::aroB aroD(new)::FRT-cat—FRT/pKDl2.112 is a better hydroaromatics (3—dehydroquinic acid + quinic acid)-synthesizing construct relative to E. coli K-12 QP1.1/pKD12.112 when transketolase is not overexpressed. However, with transketolase overexpression E. coli K-12 QP1.1/pKD12.l38 is a better host strain for synthesizing hydroaromatics relative to than E. coli B serA::aroB aroD(new)::FRT-cat- FRT/pKD12.112 without transketolase overexpression or even E. coli B serA::aroB aroD(new)::FRT-cat-FRT/pKD12.138 with transketolase overexpression, which synthesized 46 g/L in 15% yield of total hydroaromatics (Table 7, entry 6). 91 Accumulation of higher levels than usual of 3-dehydroquinic acid by E. coli B serA::aroB aroD(new)::FRT-cat-FRT/pKD12.112 and E. coli B serA::aroB aroD(iiew)::FRT-car-FRT/pKD12.138 may be explained by E. coli B recapture mechanism for 3-dehydroquinic acid transported into the culture medium is not as active/efficient as the recapture mechanism in E. coli K-12. The optimal quinic acid production time for E. coli K-12 QP1.1/pKD12.138 was determined to be 78 h (Figure 24 B) with a standard inoculum preparation. When a fresher inoculum was used, quinic acid accumulation was slower (Figure 23 B) than for the experiment with the standard inoculum preparation (Figure 23 A), even though a fresher inoculm resulted in higher biomass accumulation. The transketolase overexpression in quinate-producing E. coli K-12 strain had a more profound effect and it increased quinic acid concentration in the medium from 52 g/L synthesized by QP1.1/pKD12.112 in 15% yield (Table 9, entry 2) to 67 g/L synthesized by QP1.1/pKD12.l38 in 18% yield (Table 9,entry 1). An interesting effect was observed for QPl .1/pKD12.138 cultivated with reduced potassium phosphate or aromatic amino acid concentrations. During both cultivation conditions, reduction in accumulated biomass was observed (Figure 27 A and Figure 28 A) although only in the presence of aromatic amino acids was a reduction in dissolved oxygen requirement observed. E. coli QP1.1/pKD12.l38 did not display a reduced oxygen requirement even when the dry cell weight decreased from 55 g/L to 40 g/L upon lowering the concentration of potassium phosphate in the culture medium. However, when biomass decreased to 40 g/L with lowered concentration of aromatic amino acid, E. coli QP1.1/pKD12.l38 required less dissolved oxygen in the medium and therefore 92 airflow and/or impeller rate had to be changed. Supplementation of fermentation medium with 15 g/L of yeast extract did not afford higher concentrations and yields of quinic acid (Table 11, entry 9). Another experiment could be done in the future where addition of yeast extract is done not in the beginning of a fermentor run, but towards the end, when E. coli might lack some essential nutrients. Further optimization of QP1.1/pKD12.l38 culture conditions is required in order to determine the impact on quinic acid biosynthesis with reduced airflow. Large volume industrial fermenors can not support a 1 vvm airflow rate. Reduction of airflow will reduce dissolved oxygen concentration in the culture medium, which ultimately can translate to reduction in quinic acid concentrations and yield. Therefore, optimization of impeller speeds, impeller shape and size, and introduction of baffles inside the fermentor is required in order to detremine the best aeration conditions compatible with large-volume fermentor runs. To summarize, the best quinic acid microbial synthesis by E. coli QP1.1/pKD12.138 over 78 h was determined with standard inoculation and culture medium and afforded 70 g/L of quinic acid in 18% yield and a low 5 g/L level of byproduct 3-dehydroquinic acid. Recapturing on 3-dehydroquinic acid was shown to be an important step for quinic acid producing E. coli K-12. it was also shown that maintaining glucose-limited conditions during the first 60 h of fermentation is very important since during this time quinic acid biosynthesis is at its highest rate. 3-Dehydroquinic acid recapturing was slow in E. coli QP1.1/pKD12.l38 when fermentor-controlled culture conditions were switched from glucose-rich to glucose-limited conditions. Loss of glucose-limited control, which leads to glucose-rich conditions, towards the end of a fermentor run does not have a substantial impact on quinic acid synthesis. 93 A purification procedure was developed for microbe—synthesized quinic acid that allowed for isolation of 85% quinic acid originally in the culture medium in pure form. it involved aromatization of 3-dehydroquinic acid by refluxing cell-free, protein—free culture medium, treatment with activated carbon to remove aromatized 3-dehydroquinic acid product and decolorize culture medium, and the use of EtOH to selectively precipitate inorganic salts from quinic acid. EtOH was also used to purify quinic acid by crystallization. A further optimization of quinic acid purification process required where spray drying could be used of cell-free, protein-free culture medium. This could reduce the number of unit operations in the purification process and would amiable to large volume fermentor runs. Quinic acid and inorganic salt powder mixture obtained after spray drying could be redissolved in hot ethanol and insoluble slats filtered from solution. After filtration and partial concentration, quinic acid could be isolated in pure form. 94 References (a) Bestmann, H. J.; Held, H. A. Stereospecific synthesis of optically pure quinic acid and shikimic acid from D-arabinose. Angew. Chem., Int. Ed. Engl,. 1971, [0, 336-337. (b) Hiroya, K.; Ogasawara, K. A concise enantio- and diastereo- controlled synthesis of (-)-quinic acid and (-)-shikimic acid. Chem. Commun. 1998, 2033-2034. Haslam, E. in Shikimic Acid: Metabolism and Metabolites; Wiley & Sons: New York, 1993, p.56. Draths, K. M.; Ward, T. L.; Frost, J. W. Biocatalysis and nineteenth century organic chemistry: Conversion of D-glucose into quinoid organics. J. Am. Chem. Soc. 1992, [[4, 9725-9726. (a) Mitsuhashi, 8.; Davis, B. D. Aromatic biosynthesis X111. Conversion of quinic acid to 5-dehydroquinic acid by quinic dehydrogenase. Biochim. Biophys. Acta 1954, I5, 268-280. (b) Davis, B. D.; Gilvarg, C.; Mitsuhashi, S. Enzymes of aromatic biosynthesis: Quinic dehydrogenase from Aerabacter aeragenes. Methods Enzymal. 1955, 2, 307 -31 1. Draths, K. M.; Knop, D. R.; Frost, K. M. Shikimic acid and quinic acid: replacing isolation from plant sources with recombinant biocatalysis. J. Am Chem. Soc. 1999,/21,1603-1604. (a) Draths, K. M.; Pompliano, D. L.; Conley, D. L.; Frost, J. W.; Berry, A.; Disbrow, G. L.; Staversky, R. J.; Lievense, J. C. Biocatalytic Synthesis of Aromatics from D-Glucose: The Role of Transketolase. J. Am. Chem. Soc. 1992, [14,3956-3962. (b) Gubler, M.; Jetten, M.; Lee, S. H.; Sinskey, A. J. Cloning of the Pyruvate-Kinase Gene (Pyk) of Corynebacterium glutamicum and Site- Specific inactivation of Pyk in a Lysine- Producing Corynebacterium lactofermentum Strain. Appl. Env. Microbiol. 1994, 60, 2494-2500. (c) Miller, J. E.; Backman, K. C.; O’Conner, M. J.; Hatch, R. T. Production of Phenylalanine and Organic-Acids by Phosphoenolpyruvate Carboxylase-Deficient Mutants of Escherichia coli. J. Ind. Microbial. 1987, 2, 143-149. ((1) Patnaik, R.; Liao, J. C. Engineering of Escherichia coli Central Metabolism for Aromatic Metabolite Production with Near Theoretical Yield. Appl. Environ. Microbiol. 1994, 60, 3903-3908. (e) Patnaik, R.; Spitzer, R. G.; Liao, J. C. Pathway Engineering for Production of Aromatics in Escherichia coli: Confirmation of Stoichiometric Analysis by independent Modulation of AroG, TktA, and PpsA Activities. Biotechnol. Bioeng. 1995, 46, 361-370. (f) Li, K.; Mikola, M. R.; Draths, K. M.; Worden, R. M.; Frost, J. W. Fed-Batch Fermentor Synthesis of 3- Dehydroshikimic Acid Using Recombinant Escherichia coli. Biotechnol. Bioeng. 1999,64,61-73. 95 i3 14 (a) Flores, N.; Xiao, J.; Berry, A.; Bolivar, F.; Valle, F. Pathway Engineering for the Production of Aromatic Compounds in Escherichia coli. Nat. Biotechnol. 1996, 14, 620-623. (b) Chen, R.; Yap, W. M. G. J.; Postma, P. W.; Bailey, J. E. Comparative Studies of Escherichia coli Strains Using Different Glucose Uptake Systems: Metabolism and Energetics. Biotechnol. Bioeng. 1997, 56, 583-590. (c) Chen, R.; Hatzimanikatis, V.; Yap, W. M. G. J.; Postma, P. W.; Bailey, J. E. Metabolic Consequences of Phosphotransferase (PT S) Mutation in a Phenylalanine-Producing Recombinant Escherichia coli. Biotechnol. Prag. 1997, 13, 768-775. ((1) Baez, J. L.; Bolivar, F.; Gosset, G. Determination of 3-Deoxy-D- Arabina-Heptulosonate 7-Phosphate Productivity and Yield from Glucose in Escherichia coli Devoid of the Glucose Phosphotransferase Transport System. Biotechnol. Bioeng. 2001, 73, 530—535. (e) Flores, S.; Gosset, G.; Flores, N.; De Graaf, A. A; Bolivar, F. Analysis of Carbon Metabolism in Escherichia coli Strains with an Inactive Phosphotransferase System by 13C Labeling and NMR Spectroscopy. Metabol. Eng. 2002, 4, 124-137. Yi, J.; Li, K.; Draths, K. M.; Frost, J. W. Modulation of Phosphoenolpyruvate Synthase Expression Increases Shikimate Pathway Product Yields in E. coli. Biotechnol. Prag. 2002, I8, 1 141-1 148. (a) Ogino, T.; Garner, C.; Markley, J. L.; Herrmann, K. M. Biosynthesis of Aromatic Compounds: 13C NMR Spectroscopy of Whole Escherichia coli Cells. Proc. Natl. Acad. Sci. USA 1982, 79, 5828-5832. (b) Weaver, L. M.; Herrmann, K. M. Cloning of an AroF Allele Encoding a Tyrosine-Insensitive 3-Deoxy-D- Arabina-Heptulosonate 7-Phosphate Synthase. J. Bacteriol.1990, I 72, 6581-6584. Draths, K. M.; Pompliano, D. L.; Conley, D. L.; Frost, J. W.; Berry, A.; Disbrow, G. L.; Staversky, R. J .; Lievense, J. Biocatalytic Synthesis of Aromatics from D- Glucose: the Role of Transketolase. J. Am. Chem. Soc. 1992, I [4, 3956—3962. Ran, N. Synthesis of aromatics and hydroaromatics from D-giucose via a native and a variant of the shikimate pathway. Ph.D. dissertation. Michigan State University, 2004. Yi, J.; Draths, K. M.; Li, K.; Frost, J. W. Altered Glucose Transport and Shikimate Pathway Product Yields in E. coli. Biotechnol. Prog. 2003, 19, 1450- 1459. Chandran, S. C.; Yi, J.; Draths, K. M.; von Daeniken, R.; Weber, W.; Frost, J. W. Phosphoenolpyruvate availability and the biosynthesis of shikimic acid. Biotechnol. Prog. 2003, I 9, 808-814. Erickson, L. E.; Fung, D. Y-Ch. in: Handbook of anaerobic fermentations; Dekker: New York, 1988. 96 15 16 18 19 20 21 22 23 Pittard, J.; Wallace, B. J. Distribution and function of genes concerned with aromatic biosynthesis in Escherichia coli. J. Bacterial. 1966, 9] , 1494-1508. Snell, K. D.; Draths, K. M.; Frost, J. W. Synthetic modification of the Escherichia coli chromosome: enhancing the biocatalytic conversion of glucose into aromatic chemicals. J. Am. Chem. Soc. 1996, 118, 5605-5614. Knop, D. R.; Draths, K. M.; Chandran, S. S.; Barker, J. L.; von Daeniken, R.; Weber, W.; Frost, J. W. Hydroaromatic equilibration during biosynthesis of shikimic acid. J. Am. Chem. Soc. 2001, 123, 10173-10182. (a) Michel, G.; Roszak, A. W.; Sauve, V.; Maclean, J.; Matte, A.; Coggins, J. R.; Cygler, M.; Lapthorn, A. J. Structures of Shikimate Dehydrogenase AroE and its Paralog YdiB. J. Biol. Chem. 2003, 278, 19463-19472. (b) Benach, J.; Lee, 1.; Edstorm, W.; Kuzin, A. P.; Chiang, Y.; Acton, T. B.; Montelione, G. T.; Hunt, .1. F. The 2.3-A Crystal Structure of the Shikimate 5-Dehydrogenase Orthologue YdiB from Escherichia coli Suggests a Novel Catalytic Enviroment for an NAD- dependent Dehydrogenase. J. Biol. Chem. 2003,278, 19176-19182. Jancauskas, J. Strategies for improving synthesis of 3-dehydroshikimic acid and shikimic acid from D-glucose. M.S. thesis. Michigan State University, 2006. Brune, M.; Schumann, R.; Wittinghofer, F. Cloning and sequencing of the adenylate kinase gene (adk) of Escherichia coli. Nuc. Acid Res. 1985, I3, 7139- 7151. Chandran, S.S. Manipulation of genes and enzymes of the shikimate pathway. Ph.D. Dissertation, Michigan State University, 2000. (a) Studier, F. W.; Rosenberg, A. H.; Dunn, J.; Dubendorff, J. W. Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol. 1990, [85, 60-89 (b) Studier, F. W.; Moffatt, B. A. Use of bacteriophage T7 RNA polymerase to direct selective high—level expression of cloned genes J. Mol. Biol. 1986, [89, 113-130. (c) Grodberg, J.; Dunn, J. J. ampT encodes the Escherichia coli outer membrane protease that cleaves T7 RNA polymerase during purification. J. Bacterial. 1988, I70, 1245-1253. ((1) Phillips,T. A.; van Bogelen, R. A.; Neidhardt, F. C. [on gene product of Escherichia coli is a heat-shock protein. J. Bacterial. 1984, [59, 283-287. Datsenko, K.A.; Wanner, B.L. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA. 2000, 97, 6640-6645. 97 24 25 26 27 Cherepanov, P. P.; Wackernagel, W. Gene disruption in Escherichia coli: TcR and KmR cassettes with the option of Pip-catalyzed excision of the antibiotic- resistance determinant. Gene 1995, 158, 9-14. Miller, J. H. A Short Course in Bacterial Genetics; Cold Spring Harbor Laboratory: Plainview, NY, 1992. (a) Posfai, G., Koob, M.D., Kirkpatrick, H.A., Biattner, F.R. Versatile insertion plasmids for targeted genome manipulations in bacteria: isolation, deletion, and rescue of the pathogenicity island LEE of the Escherichia coli 0157:H7 genome. J. Bacteriol. 1997, I79, 4426-4428. (b) Martinez-Morales, F., Borges, S., Martinez, A., Shanmugam, K.T., Ingram, L.O. Chromosomal integration of heterologous DNA in Escherichia coli with precise removal of markers and replicons used during construction J. Bacterial. 1999, 18], 7143-7148. Li, W.; Xie, D.; Frost, J. W.. Benzene-free synthesis of catechol: interfacing microbial and chemical catalysis. J. Am. Chem. Soc. 2005, 127, 2874-2882. 98 CHAPTER THREE A search for a better shikimic acid producer Introduction Previously reported construction of a shikimic acid producing E. coli host strain SP1.1. began with the homologous recombination of the araB gene into the serA locus of E. coli RB791 resulting in RB791 serA::aroB.4 An additional copy of araB in the final shikimic acid producing strain, increased the specific activity of 3-dehydroquinate synthase to a level where the 3-dehydroquinic acid synthesis rate from 3-deoxy-D- arabino-heptulosonic acid 7-phosphate (DAHP) (Figure 1) was no longer an impediment to carbon flow through the shikimate pathway.I Disruption of serA locus in RB791 serA::aroB led to inactive 3-phosphoglycerate dehydrogenase, which is an enzyme required for L-serine biosynthesis in wild type E. coli. Therefore, a copy of the serA gene was inserted in plasmids and that provided the basis for plasmid maintenance when cultures were cultivated under minimal salt conditions. In order to accumulate shikimic acid in fermentation media, shikimate kinases isozymes AroK and AroL had to be inactivated (Figure I). This was done by two successive P1 phage-mediated transductions to transfer araL478::Tn/0 and araKzszR loci from ALO8072 into RB791 serA::aroB, which afforded the final shikimic acid producing E. coli host SP1.1. Increases in the intracellular concentration of phosphoenolpyruvate resulted in the direction of more carbon flow into the shikimate pathway when feedback resistant DAHP 3 . . . . . synthase and transketolase were expressed. The benchmark strain for shikimic aCId 99 production, SP1.1/pKD12.138, had overexpressed aroE-encoded shikimate dehydrogenase, tktA-encoded transketolase and araFFBR-encoded DAHP synthase and synthesized 52 g/L of shikimic acid from glucose in 18% yield with a 24% total combined yield of shikimic acid, 3-dehydroshikimic acid and quinic acid.4 Overexperession of transketolase ensured higher E4P availability inside the cell and therefore the direction of increased carbon flow into the shikimate pathway was observed (Figure 1). E. coli SP1.1/pKD12.112, which lacked plasmid-localized transketolase overexpression produced 38 g/L of shikimic acid from glucose in 12% yield and a 15% total of shikimic acid, 3-dehydroshikimic acid and quinic acid yield under glucose-rich conditions.4 With increased E4P availability, PEP availability became a limiting factor for shikimic acid biosynthesis. Two strategies were employed to increase PEP availability. E. coli SP1.1/pKD15.07iB with expression of plasmid-localized ppsA- encoded phosphoenolpyruvate synthase in addition to AroFFBR, TktA and AroE, synthesized 66 g/L of shikimic acid in 23% yield and a combined 29% yield of shikimic acid and hydroaromatic byproducts.3a This strategy relied on the fact that pyruvate, which is generated from PEP during PFS glucose transport, was recycled back to PEP by phosphoenolpyruvate synthase. The second strategy employed Glf-mediated glucose transport by facilitated diffusion in a PTS-inactive E. coli strain SP1.1pts/pSC6.09OB. Shikimic acid was synthesized at 71 g/L in 27% yield with combined of shikimic acid and hydroaromatic byproducts 34% yield. The strategies examined so far for improving shikimic acid biosynthesis were based on increasing carbon flow into the shikimate pathway. As described earlier, they 100 employed increasing specific activities of DAHP synthase and transketolase, and increasing intracellular phosphoenolpyruvate levels. With increased carbon flow into the shikimate pathway, synthesized shikimic acid yield was increased from 12% (Table 15, entry 1) to 23% (Table 15, entry 4) and total hydroaromatic yield was increased from 15% (Table 15, entry 1) to 29% (Table 15, entry 4).”4 However, the ratio between shikimic acid and 3-dehydroshikimic acid declined from 5.9 to 4.1 (Table 15). Previous work determined that accumulation of 3-dehydroshikimic acid and quinic acid is caused by hydroaromatic equilibration where shikimic acid is transported back inside the cell and converted to quinic acid through intermediacy of 3-dehydroshikimic acid and 3— dehydroquinic acid.4 It was also shown that quinic acid accumulation during shikimic acid biosynthesis was successfully reduced from 19 g/L, under glucose-limited culture conditions to 4 g/L, under glucose-rich culture conditions.4 Glucose-rich culture conditions also yielded a higher concentration and yield of shikimic acid. Previous work has also determined that shikimate dehydrogenase AroE exhibits linear mixed-type inhibition with 3-dehydroshikimic acid and an inhibition constant (K,) of 0.16 mM associated with shikimic acid.5 Therefore, accumulation of 3-dehydroshikimic acid (decrease in SA/DHS) during hydroaromatic equilibration can be caused by feedback inhibition of aroE-encoded shikimate dehydrogenase by shikimic acid. 101 Table 15. Shikimic acid and 3-dehydroshikimic acid molar ratios, shikimic acid yield and total hydroaromatic yield produced by recombinant E. coli under fermentor-controlled, glucose-rich conditions. E S R I h SA Total ntr train e evant c aracteristics - b c y Ram Yield Yield BR 0 o 1 SP1.1/pK012.112 serA, aroFF ’ PtacaroE 5.9 12 /o 15 /o 2 SP1.1/pKD12.138 serA, aroFFBR’ PtacaroE’ ”(M 3.2 180/0 240/0 - BR 3 SP1 .1pts/pSCG.O90A P13 /59'A. afoFF . PlacaroE. tktA. Ptac 9” 4.7 27% 34% glk BB 4 SP1H1/pKD15071B serA. aroFF ’ PtacaroE’ tktA, ppSA 4.1 230/0 290/0 “(mol produced shikimic acid)/(mol produced 3-dehydroshikimic acid). b(mol shikimic acid)/(mol glucose consumed). C(mol shikimic acid + mol 3-dehydroshikimic acid + mol quinic acid)/(mol glucose consumed). Removing feedback inhibition from shikimate dehydrogenase by shikimic acid might decrease 3-dehydroshikimic acid accumulation and therefore increase synthesized shikimic acid concentration and yield. Two strategies were developed to obtain a feedback insensitive shikimate dehydrogenase. One strategy called for identification of a non-E. coli shikimate dehydrogenase that was insensitive to shikimic acid inhibition followed by heterologous expression of the enzyme in an E. coli shikimate-producing strain. Quinate dehydrogenase Qad from Klebsiella pneumoniae and shikimate dehydrogenase AroD from Bacillus subtilis were investigated as alternatives to E. coli shikimate dehydrogenase AroE.6 However, B. subtilis shikimate dehydrogenase was more sensitive to shikimic acid inhibition relative to E. coli AroE and was not evaluated under fermentor-controlled conditions.6 Overexpression of plasmid-localized qad afforded very low shikimate dehydrogenase activity and Qad was not evaluated under . . 6 fermentor-controlled conditions. 102 The second strategy for obtaining a feedback insensitive shikimate dehydrogenase was directed evolution of E. coli wild-type shikimate dehydrogenase.6 It was previously shown that use of feedback resistant DAHP synthase led to increased carbon 'flow into the . . 7 . . . . shikimate pathway. A high-throughput screening of mutant shikimate dehydrogenase library was performed. However, no improvement was obtained and mutant AroE was inhibited by shikimic acid at the same level as wild-type AroE. Fermentation conditions The impact of E. coli genetic modifications on the yields and concentrations of synthesized shikimic acid and shikimate pathway byproducts was evaluated under fed- batch controlled conditions. Fermentations were run under glucose-rich and glucose- limited conditions in a 2.0 L working volume fermentor. A concentration range of 55- 170 mM glucose in the fermentation medium was maintained by manually adjusting the rate of glucose addition under glucose-rich conditions while a steady state concentration of approximately 0.2 mM glucose was maintained under glucose-limited conditions. During cultivation under glucose-limited conditions, glucose addition was controlled automatically through PID control loop by maintaining a steady concentration of dissolved oxygen. Glucose-rich conditions can lead to excessive generation of acetic acid, which is toxic to E. coli and many other microbes. 8 Glucose-limited conditions minimize generation of acetic acid but can lead to excessive CO2 generation resulting in lower product yields.8 However, it was previously discovered that shikimic acid producing E. coli had to be cultivated under glucose-rich conditions in order to obtain higher concentration and yield and more important to minimize quinic acid 103 accumulationf' Quinic acid accumulation in the shikimic acid broth at 10% or higher concentration relative to the concentration of shikimic acid complicates shikimic acid purification by crystallization.9 Thus, mostly glucose-rich conditions were used to evaluate synthesis of shikimic acid and shikimate pathway byproducts. A temperature of 36 °C and pH 7 .0 were maintained. Dissolved oxygen concentration was maintained at 20% of air saturation under both glucose-rich and glucose-limited conditions. All fermentations were run in duplicate and reported results represent the average of two runs unless otherwise stated. Metabolite concentrations were determined using lH NMR. Fermentations were terminated once metabolite concentrations stopped increasing. Inactivation of ydiB E. coli native second shikimate/quinate dehydrogenase YdiB was previously discovered and was shown to have shikimate dehydrogenase activity in vitraI0 and in vivo by restoring E. coli A82834 growth on glucose-minimal plates.6 Chapter 2 revealed, that overexpression of plasmid-localized ydiB in quinic acid producing strain resulted in quinic acid accumulation in the fermentation broth. However, YdiB was not as efficient in quinic acid production (Chapter 2) or shikimic acid production6 as compared to AroE. The overexpression of plasmid-localized ydiB in shikimic acid producer SP1.1/pJJ4.171A led to equilibration of shikimic, 3—dehydroshikimic and quinic acids to almost 1:1:1 molar ratio under glucose-rich culture conditions.6 This supported a hypothesis, that YdiB may be responsible for quinic acid accumulation during microbial synthesis of shikimic acid. To evaluate this hypothesis, a shikimic acid producing E. coli 104 host with deleted genomic ydiB sequence was constructed. As mentioned earlier, production of shikimic acid is performed under glucose-rich conditions in order to minimize quinic acid accumulation. Chapter 2 revealed that recapturing of 3- dehydroquinic acid from culture medium is reduced or totally inhibited under glucose- rich culture conditions. Therefore, both glucose-rich and glucose-limited conditions were evaluated for shikimic acid synthesis using E. coli ydiB mutant host. The construction of ydiB mutant began with deletion of the entire ydiB ORF in E. coli BW25113 using the Wanner methodologyll described in Chapter 2 and afforded BW25113 Ati‘diBzzFRT-kan-FRT. E. coli BW25113 was chosen due to high transformation efficiency and it was previously used by the Wanner group to generate . l . . . various gene knock—outs.I A succeSIful Pl phage-mediated transduction of the ydiBzzFRT-kan—FRT mutation from BW25113 ydiB::FRT-kan-FRT to E. coli SP1.1 resulted in SP1.1 AydinFRT-kan—FRT, which was designated as E. coli JJ2/(an. The mutation was confirmed by PCR analysis. Additionally, E. coli JJZkan showed no sensitivity to kanamycin due to the FRT-kan-FRT insertion in genomic DNA and it was insensitive to chloramphenicol and tetracycline due to mutations in araK and araL. respectively. E. coli JJ2/(an was treated with pCP20 plasmid—encoded FLP, which led to E. coli JJ2 (SP1.1 AycliBzzFRT). Deletion of the ydiB gene in JJ2 was confirmed by PCR analysis and growth characteristics on selective plates. E. coli JJZ showed sensitivity towards kanamycin, which indicated loss of the FRT-kan-FRT fragment, sensitivity to ampicillin, which indicated loss of the pCP20 plasmid, and resistance to chloramphenicol and tetracycline, like the parent E. coli host SP1 .1. 105 Table 16. Concentrations and yields of products synthesized by shikimic acid producing E. coli with ydiB mutation. 0 SA Total Entry Strain [SA] . yieldd’ [DH/fit [DH/E]. 10:1 yields: g/L % 9 9 g % 1a SP1.1/pKD12.138 30 13 9 o 12 21 2b SP1.1/pKD12.138 6O 26 11 o 7 33 3a JJ2/pKD12.138 23 10 13 27 o 26 4b 312/me 2.133 17 3 3 41 o 34 5&5 JJ2/pK012.112 18 9 7 12 o 19 Ga- JJ2/pKDi 2.152A 31 15 7 o o 18 7a JJ2.2/pKD12.138 49 20 13 4 o 27 8b JJ2.2/pKD12.138 51 20 12 3 o 26 “Glucose-limited conditions. bGlucose-rich conditions. * Single run fermentation. (Abbreviations: shikimic acid (SA), 3-dehydroshikimic acid (DHS), quinic acid (QA). d(mol SA)/(mol glucose consumed). e(mol SA + mol DHS + mol QA)/(mol glucose consumed). E. coli SP1.1/pKD12.138 was used as a control strain and it synthesized 30 g/L of shikimic acid in 13% yield for glucose over 60 h under glucose-limited culture conditions (Table 16, Entry 1.). Quinic acid accumulated at 12 g/L and 3—dehydroshikimic acid at 9 g/L with the total hydroaromatics yield of 21%. Cultivation of SP1.1/pKD12.138 under glucose-rich culture conditions resulted in a twofold increase in shikimic acid concentration (60 g/L) and twofold increase in yield (Table 16, entry 2) relative to culturing under glucose-limited culture conditions (Table 16, entry 1). Formation of quinic acid was reduced to 7 g/L, when E. coli QP1.1/pKD12.l38 was cultivated under glucose-rich culture conditions (Table 16, entry 2) relative to 12 g/L obtained under glucose-limited culture conditions. However 3-dehydroshikimic acid accumulated to concentrations in excess 11 g/L (Table 16, entry 2). The total yield of synthesized hydroaromatics increased to 33% (Table 16, entry 2) when E. coli SP1.1 was cultivated 106 under glucose-rich culture conditions relative to totally yield of 21% obtained under glucose-limited culture conditions. Metabolite accumulation reveals that quinic acid started to accumulate in the early stage of the fermentor-controlled cultivation and kept increasing until the end of the cultivation under glucose-limited conditions (Figure 36A). Under glucose-rich culture conditions quinic acid started to accumulate during the second half of the cultivation (Figure 368). A. B. 60 60 .A50 “A50 H l as. 53 . i 6:40 GE“ "-4 ................. I3 32.? 0'030 Q 30 "1 ----------------- m—g (:93 lo . g§ 20 0'5 20 -- ------ “-1 ------ $510 $510 ------ ”fill. -- o. o . L. 121824303642485460 121824303642485460 Time(h) Time(h) C. D. 60 60 "50 A50 .. 5.5 53 l . gs 4° d~4° '- ii ; 01174,30 Eé'ao ii i '3 ’ 0 e -; ’ i; ii 2 (i): 1 $3 1!: 1 k 08” . : 08” 1 ii? mg m}- ,3. O1 Ffi-i'fi 0 Er y l i . t 1‘ - 121824303842485460 121824303642485460 Time(h) T110901) Figure 36. (A) SP1.1/pKD12.138 and (C) JJ2/pKD12.138 cultured under glucose- Iimited conditions, and (B) SP1.1/pKDlZ.138 and (D) JJ2/pKD12.l38 cultured under glucose-rich conditions. Legend: shikimic acid (open bars), 3-dehydroshikimic acid (grey bars), 3-dehydroquinic acid (dotted bars), quinic acid (black bars), dry cell weight (circles). 107 E. coli JJZ/pKD12J38 synthesized 23 g/L of shikimic acid under glucose-limited conditions in 10% yield together with 27 g/L 3-dehydroquinic acid and 13 g/L 3- dehydroshikimic acid (Table [6, entry 3). The total yield of synthesized hydroaromatics was 26%, which was higher than 2l% compared to E. coli SPl.l/pKD12.l38 cultured under the same conditions (Table I6, entry I). The major product synthesized by E. coli JJZ/pKD12.l38 was 3-dehydroquinic acid rather than shikimic acid and the highest concentration of 3-dehydroquinic acid was 30 g/L, which was reached at 42 h (Figure 36 C). The concentration of 3-dehydroquinic acid synthesized by E. coli JJ2/pKD12.l38 actually matched the concentration of shikimic acid (30 g/L) synthesized by SPl.l/pKDl2.|38 at 60 h when cultured under identical conditions (Figure 36 A; Table 16, entry 1). Cultivation of E. coli JJZ/pKD12.l38 under glucose-rich conditions afforded only l7 g/L of shikimic acid in 8% yield over 60 h fermentor run (Table 16, entry 4). The major metabolite synthesized by E. coli JJZ/pKD12.l38 was 3- dehydroquinic acid and it accumulated at 40 g/L concentration by the end of the fermentor run with the highest concentration of 50 g/L obtained at 42 h (Figure 36 D). This time, the total yield of synthesized hydroaromatics was 34% (Table I6, entry 4), which is almost the same as 33% synthesized by E. coli SPI .l/pKD12.l38 when cultured under the same conditions. Dry cell weight for E. coli JJ2/pKD12.l38 dropped approximately 10 g/L as compared to E. coli SPl.l/pKD12.l38 under the same conditions (Figure 36). Interestingly, no quinic acid was detected in JJZ/pKD12.l38 fermentations. This indicated that YdiB was essential for quinic acid accumulation in SP1.1/pKD12.l38 when cultured under glucose-limited and glucose-rich culture conditions. Accumulation of 3—dehydroquinic acid as a major metabolite indicated that 108 3-dehydroquinic acid dehydratase AroD was not as active as it was in SP1.1/pKD12.138. To further probe this possibility, synthesis of shikimic acid using E. coli JJ2 as a host strain was probed where plasmid-encoded aroD was used in order to reconstitute the lost of genomic AroD activity. Plasmid pKD12.152A was constructed by introducing an aroDi insert to pKD12.112. E. coli JJZ/pKD12.112 synthesized 18 g/L of shikimic acid in 9% yield together with 12 g/L of 3—dehydroquinic acid and 7 g/L of 3—dehydroshikimic acid over 60 h glucose—limited cultivation (Table 16, entry 5). No quinic acid accumulation was observed at any time during the fermentor run. E. coli JJ2/pKD12.152A synthesized 31 g/L of shikimic acid in 15% yield along with 7 g/L of 3- dehydroshikimic acid under glucose-limited conditions (Table 16, entry 6). Most importantly, E. coli JJ2/pKD12.152A did not accumulate 3-dehydroquinic acid or quinic acid. Total yield of synthesized hydroaromatics for JJZ/pKD12.152 was 18% (Table 16, entry 6), which was comparable to the 19% total yield of hydroaromatics synthesized by E. coli JJZ/pKD12.112 (Table 16, entry 5). The lower yields and concentrations were obtained by E. coli JJ2/pKD12.112 (Table 16, entry 5) and JJZ/pKD12.152 (Table 16, entry 6) relative to E. coli JJZ/pKDl2.l38 (Table 16, entry 3) because plasmid pKD12.112 and pKD12.152 did not cary transketolase tktA insert as did plasmid pKD12.l38. However it was clearly shown, that plasmid overexpression of araD removed 3—dehydroquinic acid accumulation in the culture. This indicated that deletion of the entire ydiB gene from the E. coli SP1.1 genomic DNA resulted in reduced AroD activity in vivo. A closer examination of the araD locus revealed that the distance between ydiB and aroD gene is only 30 bp (ATGC in Figure 37). 109 Additionally Coggins and co-workers predicted that aroD has its own promoter, which is part of the ydiB sequence shown as boxed letters in Figure 37.12 Therefore, a new strategy was developed where deletion of ydiB gene was carried in such a way that the predicted promoter sequence was left intact. F) ydiN ydiB 1% aroD F L1 >L 8.1 51 101 151 201 251 301 351 401 451 501 551 601 651 701 751 801 851 901 951 1001 1051 1101 1151 1201 1251 1301 1351 1401 1451 1501 1551 1601 atggatgtta ccgccacagt gattgccatt ggagcaattg gatgccgaac ctgccaaact ctgcgtggct gagcggtttt gtgcctcaac attaaactct cgcgcagcgg tcgccgatca accaatggca taatgatatc ccggaaaata ttatcgcccg tacctatatg aaggattaaa aaacaactgg ggtgggggcc ataacaccga gatatcaaag ggcaattggc ttaaccgtcg gttaatgaaa gcaagccttt caaaagtggg agtctgttac ataacocgca tatgaggaag cgaattgatt aaatgcagaa gccttcgaag agccctcaaa cgtgtgaata atcaacacca cggcacgggc gcaaaacgat gcgcaggggg ggatgagttc acaccgattg gctgaagccc tatgaaaccc gggttgatgg taaagcctta tggataacga atgcgcggaa tgttgatgaa tcgttaatga catattcgcg ggtgctgtta caattgaagg ttcgataaag tgtcgtcacg tggcttccgc cttgagaatg cctatcctat gaaaaagcgg tagctttcct ctggtgtatc ttaacaccag tgatggctat ccattaaaga ggggccggtg tttaaaagaa ccctcgcctt gtcaccgatc cgacatttta aatcattggt atccgggact tCtggthCt gaatgcgtgt ttattgcagc aggcgcaaca 39G 99 9 aaaacgattg cacattatgg tQQQQtthq AAAACCGTAA CATCGTCTCG TCGCCTATCG TATGCCGACC CCGTGAGACC AAGAAGGCGG CGTGCAGCCA TACCGGTGAT ATGTGAAAGT GAAGAAATCA TCCTAAGATT TTGCCGCGAC ACGATGTCGA ATTTGGCTCG GGCAAATCTC atggatacgg actggcaaag tqcctgaCAG CTGTAAAAGA CTGATGGCGA TGAAGCGGAC TCTCCAATGT ATGCCAGAAA CGAGCAGGCG TCGACAGCGG GATCAGGTTA AGTCATGTCC TTGCCCGTCT GCGCTGATGC CCTGGAGATG TGGCAAAAAC GCGGCAACTT GGTAAATGAT catgttgttg atttccctct GCTGACCGCG TCTCGTCATT AAGATATCGC TTTGATATTC GGAGTCTGTC AACCGCTGCT ATTTCCACCG CCTGGTTGAT AAGAAACCGT AACCATGACT GCGCAAAATG CGCAAAGTAC CAGGAGCAGT TGGCGTAATT TTGGTGCGGT TTGCGCACGG tggcaagggg ggaatatgtt TGCAGAAAGG GGTACGGGCG CAGCGTGAAA TGGAATGGCG ATGGCGGCAG GTTTACCTTC AGGCTTATAT ATGATCGATC CGCCTACGCC TCCATAAAAC CAATCCTTCG CAGCGATGTG ATGCCGATCG TCTCGTCTGG AAAAAAAGCG TATTAACTAT ctgaacagt aaacag to GT TG CACCTAAAAT TCCGAAGCTC TGTGGACCAC CAAAAATTCT CGCAGTGCCA TGCACTCAAT TGGAGTTATT CACGCGCATG GCCGGAAGCC ACGCCGATAT CTGACGTTGC TCCAATTATC CTGGTGAAGT TCTGCGCCAG TTTACACCAG Figure 37. E. coli K-12 ydiB and araD genomic DNA locus: (A) graphic representation; (B) DNA sequence. Legend: (atgc) ydiB sequence, (ATGC) aroD sequence, ( boxed atgc) predicted aroD promoter sequence, (a_tgg) primer sequences for generating “22 mutant. 110 Construction of the new shikimic acid producer started from construction of E. coli BW25113 AydiB(Hl, H2.2)::FRT-kan-FRT using Wanner gene disruption methodology, where H1 (single underlined) and H22 (double underlined) sequences are shown in Figure 37. A successful Pl phage-mediumted transduction of a new ydiB mutation into SP1.1 afforded SPl.l AydiB(H1, H2.2)::FRT-kan-FRT, which was named as ”2.2/can. This mutant was treated with pCP20 plasmid-encoded FLP and resulting E. coli SP1.1 AycliB(Hl, H2.2)::FRT was named 112.2. Mutants were successfully verified by PCR analysis. E. coli 1.12.2kcm was insensitive to kanamycin, tetracycline and chloramphenicol, while “2.2 and SPl.l showed sensitivity only to kanamycin. The growth pattern on glucose-minimal salt plates was identical with 1.12.2kan, 1.12.2 and SP1 .I requiring L-serine, aromatic amino acid and aromatic vitamin supplementation. E. coli JJ2.2/pKD12.l38 was evaluated under glucose—limited conditions and it synthesized 49 g/L of shikimic acid in 20% yield together with 4 g/L of 3-dehydroquinic acid and 12 g/L of 3-dehydroshikimic acid (Table 16, entry 7). Accumulation of 3- dehydroquinic acid was reduced as compared to the 27 g/L synthesized by JJZ/pKD12.138 (Table 16, entry 3) under the same conditions, however it was not totally eliminated. For comparison, the control strain SP1.1/pKD12.l38 did not produce any 3- dehydroquinic acid. No accumulation of quinic acid was observed at any point during the cultivation of ”2.2/pKD12.l38 under fermentor-controlled glucose-limited culture conditions (Figure 38A). A continuous increase in shikimic acid concentration was observed (Figure 38A). The total synthesized shikimic acid concentration of 49 g/L (Table 16, entry 7) for ”2.2/pKD12.l38 was approximately the same as the total shikimic acid (23 g/L) plus 3-dehydroquinic acid (27 g/L) for JJZ/pKD12J38 (Table 16, 111 entry 3) and higher that the total shikimic acid (30 g/L) plus quinic acid (12 g/L) synthesized by SP1.1/pKD12.l38 (Table 16, entry 1). This indicates that YdiB is essential for quinic acid equilibration with shikimic acid when shikimate-synthesizing E. coli constructs are cultivated under glucose-limited culture conditions. Cultivation of “2.2/pKD12.l38 under glucose-rich culture conditions yielded 51 g/L of shikimic acid (Table 16, entry 8), which was lower as compared to the 60 g/L of shikimic acid produced by SP1.1/pKD12.l38 under the same glucose-rich culture conditions. However, ”2.2/pKD12.l38 did not produce quinic acid at any point during the fermenor run (Figure 38B) although of 3 g/L of 3-dehydroquinic acid was synthesized (Figure 38 B). The total yield of hydroaromatics synthesized by E. coli JJ2.2/pKD12.138 was 26 — 27% (Table 16, entry 7 and Table 16, entry 8), which is higher than the 21% total yield of hydroaromatics synthesized by SP1.1/pKD12.l38 under glucose-limited culture conditions (Table 16, entry 1), but lower than the 33% total yield of hydroaromatics synthesized by SP1.1/pKD12.l38 under glucose-rich culture conditions (Table 16, entry2). 60 60 5%: 5° . as: -" 4o .......... , .- 215. $5. 0.3 3° """"" i; """" US$30 'g? 20 i ---------- t-~+b4 Jo:§20 $3 10 ----- --E--E $8109»- 0 ‘ ' . . i. .' 0‘ ' " 121824303642485460 121824303642485460 Time(h) Time(h) Figure 38. (A)JJ2.2/pKD12.l38 cultured under glucose-limited conditions and (B) JJ2.2/pKD12.138 cultured under glucose-rich conditions. Legend: shikimic acid (open bars), 3-dehydroshikimic acid (grey bars), 3~dehydroquinic acid (dotted bars), quinic acid (black bars), dry cell weight (circles). 112 A new SP1.1 variant As mentioned earlier, shikimic acid producing host E. coli SP1.l was constructed by introducing shikimate kinase aroL478::Tn/0 and aroKzszR mutations from AL08072 into RB791 serA::aroB, which resulted in SP1 .I resistance to tetracycline and chloramphenicol. Resistance genes can introduce a polar effect on the downstream genes. In order to eliminate the posibility of the polar effects a new shikimic acid producing host was constructed using Wanner gene deletion methodology. It was also suggested in the literature that E. coli shikimate kinase encoded by the aroL gene is a principal shikimate kinase responsible for the majority of shikimic acid phosphorylation inside the cell.'3 Therefore, it was postulated that an E. coli mutant with only aroL mutation might be able to grow on glucose-minimal salts conditions without aromatic amino acid and aromatic vitamin supplementation. The remaining minor shikimate kinase activity encoded by aroK might channel enough shikimic acid down the shikimate pathway to ensure de novo biosynthesis of adequate level of aromatic amino acid and aromatic vitamins to sustain growth. Deletion of the principal shikimate kinase encoded by aroK also needed to be evaluated for whether accumulation of shikimic acid in the culture medium would still take place. To answer these questions, E. coli JJS (RB791 serA::aroB AaroKzzFRT Aar0L22FRT) was constructed, which carried a double shikimate kinase knock-out and 1.14 (RB791 serA::aroB AaroLzzFRT) was constructed, which carried only aroL knock-out. Construction of both mutant strains was performed using Wanner gene disruptions methodology and both strains were cured of antibiotic 113 resistance. Gene deletions were performed directly in E. coli RB791 serA::aroB rather than BW251 13, as during 112 and 1122 construction. Table 17. Concentrations and yields of products synthesized by shikimic acid producing E. coli with ydiB mutation. [SA]C. SA yieldd, [DHS], [DHQ], [QA], Total entry Strain Mutation g/L % 9 IL 9 IL g/L yielde, % 1a SP1.1/pKD12.138 a’OK'va’OL' 30 13 9 o 12 21 2b SP1.1/pKD12.138 aFOKva’OL' 60 26 11 o 7 33 31" JJ4/pKD12.138 3'01“, AarOL' 18 3 16 o 23 10 4a. JJ5/pKD12.138 Aa'OK'v AarOL' 29 12 12 o 28 27 “Glucose-limited conditions. bGlucose—rich conditions. * Single run fermentation. cAbbreviations: shikimic acid (SA), 3-dehydroshikimic acid (DHS), quinic acid (QA). d(mol SA)/(mol glucose consumed). e(mol SA + mol DHS + mol QA)/(mol glucose consumed) E. coli 114/pKD12.l38 was evaluated under glucose-rich, fermentor-controlled culture conditions. Culture medium was not supplemented with aromatic amino acids and aromatic vitamins. E. coli 114/pKD12.l38 grew faster that SP1.1/pKD12.l38 and accumulated of 70 g/L of biomass by 36 h (Figure 39 B), which is double that of 35 g/L of biomass produced by SP1.1/pKD12.l38 under the same conditions (Figure 36 B) by 30 h. This clearly indicated that there was enough shikimate kinase AroK activity to channel shikimic acid downstream into shikimate pathway to biosynthesize adequate concentrations of aromatic amino acids and aromatic vitamins needed for growth. However, 114/pKD12.l38 produced only 18 g/L of shikimic acid in 3% yield over 60 h (Table 17, entry 3). Synthesis by 114/pKD12.l38 of 3-dehydroshikimic acid and quinic acid was of 16 g/L and 23 g/L, respectively (Table 17, entry 3), consisted higher concentration of these hydroaromatic byproducts relative to SP1.1/pKD12.l38 (T able 17, entry 2). Presumably, the remaining AroK activity in 114/pKD12.l38 channeled more 114 carbon flow down the shikimate pathway than had been initially anticipated. Accumulation of higher biomass led to reduced product yields and the total yield of synthesized hydroaromatics was 10% (Table 17, entry 3), which was threefold lower than the 33% total yield of hydroaromatics synthesized by SP1.1/pKD12.138 under glucose- rich culture conditions. A double shikimate kinase aroK and aroL knockout was evaluated under glucose- limited conditions. E. coli 1.15/pKD12.l38 produced 29 g/L of shikimic acid in 12% yield over 60 h fermentor run (Table 17, entry 4) and it was close in performance to the 30 g/L of shikimic acid synthesized in 13% yield by SP1.1/pKD12.138 under glucose- limited conditions (Table 17, entry 1). However, it also accumulated higher levels of 3- dehydroshikimic acid (12 g/L) and quinic acid (28 g/L, Table 17, entry 4). The total yield of synthesized hydroaromatics was 27%, which is significantly higher relative to 21% total yield of synthesized hydroaromatics synthesized by SP1.1/pKDl2.l38 (Table l7,entry1). A. B. 60 604 ------------------ 0m. . 9 70 A o 0 .A50 550 G 'mg 5% < 9:40 0.40 E .1. g. as? . I: U) 0820 30:20 3310- $101 0‘ O-' 121824303642485460 121824303642485460 Time(h) 11ml") Figure 39. (A) JJS/pKDlZ.l38 cultured under glucose-limited conditions and (B) JJ4/pKD12.l38 cultured under glucose-rich conditions. Legend: shikimic acid (open bars), 3-dehydroshikimic acid (grey bars), 3-dehydroquinic acid (dotted bars), quinic acid (black bars), dry cell weight (circles). 115 Shikimate dehydrogenase from Gluconobacter oxydans As mentioned earlier, several attempts to identify a better shikimate dehydrogenase than E. coli AroE were made including overexpression of quinate dehydrogenase Qad from Klebsiella pneumoniae and shikimate dehydrogenase AroD . .. 6 . . from Bczczllus submits. A new shikimate dehydrogenase was reported from Glucmmbac‘rer oxydans lFO 3244 and it showed in vitro specific activity with shikimic acid and 3-dehydroshikimic acid as substartes.l4 Additionally this shikimate dehydrogenase showed no in vitro activity with quinic acid, 3-dehydroquinic acid and protocatechuic acid. More interestingly, G. orvdans 1FO 3244 shikimate dehydrogenase has almost eightfold higher Km value for shikimic acid as compared to E. coli shikimate dehydrogenase AroE (Table 18). During shikimic acid synthesis, 3-dehydroshikimic acid accumulation accounted for 14% of the total hydroaromatic and was by far the major byproduct (Table 16, entry 2). It was previously postulated that accumulation of 3- dehydroshikimic acid and quinic acid during shikimic acid biosynthesis is due to hydroaromatic equilibration, where shikimic acid is recaptured by the cells and converted back to 3-dehydroshikimic acid, 3-dehydroquinic acid and quinic acid in the reverse direction of the shikimate pathway (Figure l).4 The conversion of shikimic acid to 3- dehydroshikimic acid and of 3-dehydroquinic acid to quinic acid is performed by the . . . 4 . . . same E. colt shikimate dehydrogenase AroE. Therefore, overexpressron of shikimate dehydrogenase from G. oxyclans lFO 3244 with higher K value for shikimic acid should "1 slow down shikimic acid/3-dehydroshikimic acid equilibration. Additionally, if quinic 116 acid is not a substrate for G. oxydans [FD 3244 shikimate dehydrogenase, then 3- dehydroquinic acid also might not be a substrate thereby precluding the reduction of 3- dehydroquinic acid to unwanted quinic acid. Successful heterologous overexpression of . . I. Glzwmmbacter genes In E. ('()[l has been demonstrated by several groups. 5 Therefore, an attempt to overexpress shikimate dehydrogenase from G. arvdans 1FO 3244 in E. coli shikimic acid producer was made. Table 18. Shikimate dehydrogenase Km values. Organism Km' mM shikimic acid 3-dehydroshikimic acid 6. oxydans IFO 3244 0.5 (ref. 14) 0.2 (ref. 14) E. co/iK-12 0,055 (ref. 108) 0.1 (ref. 6) The gene sequence for G. nxydans [FD 3244 shikimate dehydrogenase was not available. However, the entire genome sequence for G. oxydans H621 was available from the NCBI'6 and the ERGOI7 database. Interestingly, the NCBI and ERGO gene sequences for shikimate dehydrogenase from G. mydanx H621 did not match. ERGO annotated only one shikimate dehydrogenase, which uses PQQ (pyrroloquinoline quinone) as a cofactor, rather than NAD(P). Although, many bacteria can biosynthesize the cofactor PQQ required for several dehydrogenases, wild-type E. coli can not produce 18 PQQ. Therefore, use of this enzyme in an E. coli shikimic acid producer was not pursued. The shikimate dehydrogenase gene from the NCBI database had 19% identity with E. coli shikimate dehydrogenase AroE at the protein level. No data was available as to whether this dehydrogenase was PQQ or NAD(P) depended. The putative shikimate dehydrogenase gene sequence in ERGO was annotated as fructose S-dehydrogenase. 117 The assumption was made that there was no important sequence difference between G. oxydans [FD 3244 and G. oxydans H621. PCR primers for shikimate dehydrogenase ORF from G. mydcms H621 were designed based on the NCBI sequence. G. 0.1-Mans lFO 3244 was ordered from the National Institute of Technology and Evaluation in Japan and genomic DNA was purified using a Qiagen genomic DNA isolation kit. Purified genomic DNA from G. mydans [PO 3244 was used as a template for PCR. The 0.8 kb PCR product of was isolated using agarose gel and cloned under a P tac promoter between the EcoRI and Smal sites of pKK223-3 vector. The resulting plasmid was transformed into E. coli AB2834 host and the transformation mixture was . . . l9 . . . . plated on glucose-minimal salts plates. E. coli A82834 has Inactive shikimate dehydrogenase AroE, ad as consequence, this mutant is not able to grow on minimal salt medium without shikimic acid or aromatic amino acid and aromatic vitamins supplementation. 1f ABZ834 is transformed with a plasmid encoding active shikimate . . . . 6 dehydrogenase, transformants wrll able to grow on glucose-minimal salts medium. However, no growth after 7 days was observed in this case, which indicated that either the shikimate dehydrogenase gene sequence from G. arwlans H621 was PQQ depended or this sequence was annotated incorrectly by NCBl. Genomic plasmid library approach was then taken to isolate shikimate dehydrogenase from G. arrdans [PO 3244. Purified genomic DNA was partially digested with BamHl restriction endonuclease followed by ligation of 1-10 kb genomic DNA pieces into the BamHI site of pBluescript SK (—) vector. The resulting plasmid library was electroporated into E. coli A82834 and the transformants plated on glucose- minimal salts medium. After 48 h of incubation at 37 °C, two colonies were identified 118 and plasmids from these colonies were isolated. Multiple enzyme digestion of these plasmids yielded identical restriction enzyme digestion patterns, which indicated that genomic DNA pieces were identical or very similar. Overexpression of G. oxydans IFO 3244 shikimate dehydrogenase in AB2834 was investigated. Cell lysates were assayed in the forward direction using 3-dehydroshikimic acid as a substrate and NADPI4 as cofactor and in the reverse direction using shikimic acid and NADPH. Overexpression levels were only twofold higher (Table 19, entry 2 and entry 3) relative to the ABZ834 background specific activity (Table 19, entry 1). As a consequence, further efforts directed towards heterologous expression of shikimate dehydrogenase from G. oxydans IFO 3244 were abbandoned. Table 19. Shikimate dehydrogenase activity. . . . . a entry Source Specuflc actuvuty , U/mg 3-dehydroshikimic acid shikimic acid 1 A82834 0.0005 0.0002 2 A82834/pBlue IF03244 #1 0.001 0.001 3 A82834/pBlue IFO3244 #2 0.001 0.001 “ One unit (U) of shikimate dehydrogenase corresponds to the formation of l umole of NADP in the presence of 3-dehydroshikimic or consumption of l umole of NADPH in the presence of shikimic acid per min at 25 °C. An attempt to identify hydroaromatics transport system in E. coli To identify the hydroaromatics transport system in E. coli is a crucial step in optimizing the production of shikimic acid and quinic acid. Increasing the rate of such transport system or rate might result in increased production of hydroaromatics. Several active efflux pumps have been identified in E. coli in response to treatment with chemicals or antiobiotics. Increased expression of the efflux pumps leads to decreased intracellular concentration of the externally added compounds, resulting in E. coli with a 119 higher tolerance to these compounds. The efflux systems range from broad to very narrow substrate specificity. For example, E. coli efflux system AcrAB-ToIC is upregulated in response to exogenous compounds such as salicylic acidzo, methylviologenZI or bile salts22 and a wide range of compounds are the substrates for this pump.23 On the other hand CusCFBA complex exports only copper and silver ions from E. coli cells.24 Another specific efflux pump was identified for export of p- hydroxybenzoic acid in E. 6011.23 It was shown that this pump is triggered by internal and external p-hydroxybenzoic acid, a native metabolite of E. coli. Interestingly, only a few aromatic carboxylic acids were identified as a substrate for this pump. Pittard and co-workers have characterized a system encoded by the shiA locus, which is apparently responsible for shikimic acid transport in E. coli.26 Later, Frost and co-workers showed that shiA knockout E. coli was still transporting the shikimic acid. Therefore, there must be another system involved in shikimic acid or hydroaromatics efflux.2 The E. coli ydiN gene was found in the same operon as 3-dehydroquinate dehydratase encoded by araD and the second shikimate/quinate dehydrogenase ydiB (Figure 37A). The function of YdiN is unknown, however it was annotated as an amino acid/amine MFS transporter27 or multidrug resistance gene.'7 Additionally, upregulation of ydiN was observed at the transcriptome level during shikimic acid biosynthesis.28 Might YdiN be involved in hydroaromatics transport across the cellular membrane? To gain insights into the possible function of YdiN, SP1.1 AydiszFRT mutant was constructed. If E. coli YdiN is involved in hydroaromatics transport alone or together in complex with some other 120 proteins, deletion of this gene will result in an inactive hydroaromatics efflux pump and therefore synthesized shikimic acid concentration should decrease. Construction of SPI.1 AydiszFRT started from generating E. coli BW25113 AydiN::FRT-kcm-FRT mutant using previously described gene inactivation methodology. Multiple trials were performed to generate the desired BW25113 AydiNxFRT-kan-FRT mutant, but only one colony was obtained. Once again P1 phage—mediumte transduction was used to transfer yc/z‘N mutation to an E. coli SP1.1 host. However, all trials were unsuccessful and SPI.1 A_vdiN::FRT-kan-FRT could not be obtained. If YdiN is responsible for multidrug resistance, as it was annotated, deletion of this gene probably resulted in hypersensitivity to antibiotics and therefore no colonies were observed on selective LB/kan plates. A new selection method was designed. Instead of inserting antibiotic resistance gene in a desired locus of genomic DNA, a serA insertion will be used. Therefore, a host lacking 3-phosphoglycerate dehydrogenase SerA activity has to be used for this method and mutant selection will be based on growth on glucose-minimal medium in the absense of any added antibiotics. This method should overcome issues related to antibiotic hypersensitivity. rpiA p1 p‘2 serA 1’ng (a) E:>:>I:>::> <:1 BW25113 0.7 kB 1.2 kB 0.5 kB mil: P1 FRT yng (b) ‘-—? LL-I C: BW25113 AserA2::FF?T 0.7 kB 0.5 kB FRT FRT " P2 serA (ORF) 1.4 kB ‘ (C) m 'fl//////////////A/////17////////// m PNR9-280 Figure 40. Modification of gene disruption methbd. 121 Since all hosts for microbial production in the Frost research group are serA (-), these hosts can be used for gene deletion using this method and the serA insertion can be eliminated using fiippase in the final target host. E. coli serA transcription is controlled by two promoter sequences P1 and P2, where P2 is the principal promoter (Figure 40 A).29 A new host was constructed where entire P2 and a part of P1 promoter sequence together with ORF of serA were deleted using standard gene deletion method described earlier and resulted in BW25113 AserAzzFRT-kan-FRT. This host had 169 bp deletion upstream from serA ORF and a 100 bp downstream of serA. Kanamycin resistance gene was eliminated using pCP20 plasmid-encoded flippase and resulted in the E. coli BW25113 AserAzzFRT ready to be used in the newly designed deletion method (Figure 40 B). The Frost group member constructed plamsid pNR9.280 where chloramphenicol resistance gene in pKD3 was replaced with the P2 serA sequence, flanked between two FRT sites (Figure 40C). The newly constructed system (Figure 40) was used to knock-out the ydz‘N gene in E. coli SP1 .1. Plasmid pNR9.280 was used as template for the initial PCR step where H1 and H2 40 bp primers where designed to be homologous to the 5’ and 3’ ends of the ydiN gene. Successful PCR yielded 1.4 kbp DNA size band on agarose gel. The PCR product was electroporated in BW251 l3 AserAzzFRT/pKD46 expressing A. Red recombinase and mutants were selected on glucose-minimal plates. Approximately 200 colonies appeared on the selective plates, but after second round of replication on glucose-minimal plates only 5 of them continued to grow, indicating that the majority of the colonies were false positives. Genomic DNA was isolated from all 5 candidates for the PCR verification test. Verification primers were designed to have homology outside the ydiN region and PCR 122 reaction with 5 candidate genomic DNA resulted in a 1.2 kbp sized DNA band rather than 1.4 kbp sized as had been anticipated. Control PCR with E. coli K-12 genomic DNA also resulted in 1.2 kbp sized DNA, which indicated that none of the 5 candidates had a ydiN deletion. A second PCR test was performed with primers designed juSt outside serA and it resulted in approximately 300 — 400 bp sized DNA, which indicated that serA deletion was still present in the host. Construction of ydiN mutant was unsecsesful and was stopped. Instead of deleting ydiN gene and expecting a lower concentration of synthesized shikimic acid, overexpression of the ydiN gene in a shikimic acid or quinic acid producer might lead to increased hydroaromatics accumulation in the medium, if ydz'N is alone responsible for hydroaromatics transport. Construction of a ydiN overexpressing plasmid began with PCR amplification of a 1.2 kb ydiN ORF fragment from E. coli W3110 genomic DNA. Isolated PCR fragment was inserted into EcoRI and Pstl cloning site of the pKK223-3 vector under the Pm promoter and yielded plasmid pJJS.151 (Figure 41). Plasmid pJJS.151 served as a template for PCR of the 1.4 kbp size P .ydiN fragment, far which was eventually treated with Kelnow. Plasmid pKD12.l 12 was Iinerized with Sall \‘diN restriction endonuclease and treated with Klenow, subsequent ligation with PM; insert fragment afforded pJJS.I64 (Figure 42). Transketolase encoded by the 2.2 kbp tktA insert was excised from pNR8.146 with BamHI restriction endonuclease and was treated with Klenow. Plasmid p.115.l64 was Iinerized with Xbal restriction nuclease and treated with Klenow. DNA insert tktA was ligated to Iinerized pJJS.I64 and afforded the target plasmid pJJS.l65 (Figure 43). Newly constructed plasmid, pJJS.165 was evaluated for 123 shikimic acid production under glucose-rich culture conditions and for quinic acid production under glucose-limited culture condition. Smal Pstl EcoRl Hindi I 1 PCR ydiN from E. coIiW3110 genomic DNA l1) EcoRI and Pstl digest {if 11' - iii p 51113213: 3 l7 ECOR' PS" 1‘ ' 1.2 kb XX:\ /1 A! ydiN x “Pry-.77" . I), , ‘Mwu E 1) EcoRI and Pstl digest E 2) CIAP treatment T4 Ligase Pstl Hindl II 5.8 kb Figure 41. Construction of pJJ5.151. 124 PCR PmydiN from pJJ5.151 l1) Klenow treatment 1.6 kb 1) Sail digest PtacydiN 2) Klenow treatment 3) CIAP treatment 1 T4 Ligase 9 PS" (Sail) \90 ,3 +9 (fa .; . a/w‘” \ \//V \6 /. \ 9.3 kb Figure 42. Construction of pJJ5.l64. 125 Pstl ‘5 (Sail) \ 153'» \<\\ N“ 4522127,» k {9,90 6.7 kB 1W serA a; pJJ5.164 : 1 9.3 kb § Pstl Xbal BamHl 1) BamHl digest 2) Klenow treatment BamHl BamHl 2.2 kb 1)Xbald1gest .............. 2) Klenow treatment MA 3) CIAP treatment j T4 Ligase Pstl (Sail) (Sail) p tacky/7V .-. :1 pJJ5.165 EcoRl Kpnl Figure 43. Construction of pJJ5.l65. 126 Table 20. Concentrations and yields of products synthesized by shikimic acid and quinic acid producing E. coli with ydiN overexpression. c . d [DHS]. [DHQ], [QA], . a Total Entry Construct [SA] , SA yield , g/L g/L 9 IL QA yield , . g/L °/0 °/o yield , % 1b SP1.1/pKD12.138 60 26 11 0 7 - 33 2b SP1.1/pJJ5.165 51 20 10 0 5 - 26 3a QP1.1/pKD12.138 - - 5 56 21 21 a, OP1.1/JJ5.165 - - - 6 44 14 16 “Glucose-limited conditions. bGIucose-rich conditions. * Single run fermentation. (Abbreviations: shikimic acid (SA), 3-dehydroshikimic acid (DHS), quinic acid (QA). d(mol SA)/(mol glucose consumed). e(mol QA)/(mol glucose consumed). f(mol SA + mol DHS + mol QA)/(mol glucose consumed). A. B. 60 60 g g 50 1 E--. ...... 5 g 50 "V - ..... , .. ........... " 40 F "- r“ 934° gs 3g 30 ;~ ------ ~~ ---------- .230 8 § 20 . - ..... 13.--. P.“ F--. 1 §§ 20 $510 ”infill-11‘ -- $510 0 . L o . 121824303842485460 121824303642485460 c. Time(h) 0. Time(h) $70 $70 3 3 E 60 3360 u ' § 50 '3 50 8 4° 3 40 3 30 530 5 20 3'20 ' 10 , ' 101 g 0 g o- g g E 12182430364248 12182430304248 Time(h) Time(h) Figure 44. (A) SP1.1/pKD12.138 and (B) SP1.1/JJ5.165 cultured under glucose-rich conditions, and (C) QP1.1/pKD12.l38 and (D) QP1.1/pJJ5.165 cultured under glucose-limited conditions. Legend: shikimic acid (open bars), 3-dehydroshikimic acid (grey bars), 3-dehydroquinic acid (dotted bars), quinic acid (black bars), dry cell weight. 127 SPl.l/pJJS.l65 synthesized 51 g/L of shikimic acid in 20% yield over 60 h fermentations (Table 20, entry 2), which was lower than the control strain SPl .l/pKD12.l38 production of 60 g/L of shikimic acid in 26% yield (Table 20, entry 1). Accumulation of byproducts, 3-dehydroshikimic acid and quinic acid was observed (Table 6, entry 2) at approximately the same concentration of 10 g/L and 5 g/L, respectively. The total yield of hydroaromatics synthesized by SPl.l/pJJ5.l65 was 26%. Dry cell weight and byproduct accumulation profile for the control strain SPl.l/pKD12.l38 and the new construct SPl.l/pJJS.l65 looked very similar throughout the respective fermentation runs (Figure 44 A and B). An larger decrease in synthesized products was observed for quinic acid production under glucose-limited conditions. E. coli QPI .l/pJJS.l65 synthesized 44 g/L of quinic acid in 14% yield over 48 h fermentation (Table 20, entry 4) while QP1.1/pKD12.l38 produced 58 g/L of quinic acid in 21% yield under the same conditions (Table 20, entry 3). Accumulation of 3-dehydroquinic acid was about the same level at 6 g/L and the total yield of hydroaromatics declined to 16% (Table 6, entry 4). The highest observed biomass level was 69 g/L at 36 h, while the control strain accumulated 61 g/L of biomass at 60 h (Figure 44C and Figure 44D). These experiments clearly indicated that overexpression of ydiN did not have a positive effect on shikimic acid or quinic acid biosynthesis. Either YdiN is not responsible for hydroaromatics transport or it is working in the complex with other proteins. 128 Discussion Inactivation of YdiB activity in a shikimic acid producer led to elimination of quinic acid accumulation in the fermentation broth under glucose—limited and glucose- rich conditions. However, the shikimic acid concentration under glucose—limited conditions was reduced from 60 g/L synthesized by SP1.1/pKD12.l38 to 51 g/L synthesized by JJZ.2/pKDl2.l38. Interestingly, 3-dehydroquinic acid accumulation was still observed for ”2.2/pKD12.138, while SP1 .l/pKD12.l38 showed no 3-dehydroquinic acid accumulation in the medium. This suggests that E. coli ”2.2 had lower 3- dehydroquinate dehydratese AroD activity as compared to SPI .I. Since it was predicted that the promoter sequence for aroD is within the ~vdiB ORF'Z, E. coli ”2.2 was constructed by deleting only a part of ydiB, which left the predicted promoter sequence for aroD intact. This helped to reduce the concentration of 3-dehydroquinic acid synthesized by 112.2/pKDIZJ38 as compared to JJZ/pKD12.l38. However, synthesis of 3-dehydroquinic acid was not completely eliminated (Table [6). It was demonstrated for the first time that YdiB is involved in the synthesis of quinic acid as a byproduct during the synthesis of shikimic acid under both glucose-rich and glucose-limited conditions. However, the full role of YidB in E. coli is not fully understood. The total concentration of hydroaromatics synthesized by SPl.l/pKD12.l38 was 78 g/L, while JJZ/pKD12J38 and “2.2/pKD12.l38 synthesized 66 g/L (Table 16, entry 2, entry 4 and entry 8). Successful deletion of ydiB channeled more carbon flow towards the shikimic acid and hydroaromatic byproducts under glucose-limited conditions. E. coli SPl.l/pKD12.l38 produced Sl g/L of total hydroaromatics under glucose-limited conditions, while JJZ/pKDlZ.l38 synthesized 63 g/L and “2.2/pKD12.l38 synthesized 66 g/L (Table 16, 129 entry 1, entry 3 and entry 7). There was no observed quinic acid accumulation as well. Interestingly, total synthesized hydroaromatics by JJZ.2/pKDlZ.l38 under glucose- limited and glucose-rich conditions was the same at 66 g/L. Elimination of quinic acid accumulation in shikimic acid production should simplify the purification of product shikimic acid. This eliminates problems encountered with in co-crystallization of quinic acid with shikimic acid during purification of the desired shikimic acid.9 The new shikimic acid construct JJS/pKDl2.l38, which possessed deleted aroK and aroL, did not show any improvements in shikimic acid production over SPl.l/pKD12.l38. Shikimic acid production using a single shikimate kinase aroL knockout E. coli JJ4/pKDl2.l38 did not produce as much shikimic acid as the double shikimate kinase knockout SPl.l/pKD12.l38 (Table 17, entry 3 and 2). However, JJ4/pKD12.l38 was able to grow in culture medium lacking supplementation with aromatic amino acids and aromatic vitamins channeled more carbon downstream the shikimate pathway. The downside of this accomplishment was an overly abundant supply of aromatics that translated into two-fold increase in the biomass (Figure 39 and Figure 36 B) and seven-fold decline in the synthesized hydroaromatics. This is consistant with the need to control biomass formation in order to achieve high concentration and yields of hydroaromatics. Identification of a shikimate dehydrogenase that is highly selective for reduction of 3-dehydroshikimic acid over 3-dehydroquinic acid along with reduced sensitivity to feedback inhibition by shikimic acid remains an attractive goal. Along these lines, the selectivity reported for G. oxydcms shikimate dehydrogenase calls further investigation. [30 However, a new strategy for isolating NADPH-dependent shikimate dehydrogenase from G. 0.x:vdans will be needed. In route to delineate the system exploited by E. coli to export hydroaromatics. E. coli lacking YdiN activity could not be constructed using two different selection methods. With the alternative strategy of YdiN overexpression with plasmid-localized ydiN, shikimic and quinic acid production declined as compared to the control strains (Table 20). Alternative candidates of gene encoding functions essential to hydroaromatic export will need to be explored in the future. l3l References Snell, K. D.; Draths, K. M.; Frost, J. W. Synthetic modification of the Escherichia coli chromosome: enhancing the biocatalytic conversion of glucose into aromatic chemicals. J. Am. Chem. Soc. 1996, 118, 5605-5614. Laner-Olesen, A.; Marinus, M. G. Identification of the gene (aroK) encoding shikimic acid kinase—I of Escherichia coli. J. Bacteriol. 1992. [74, 525-529. (a) Chandran, S. S.; Yi, J.; Draths, K. M.; von Daeniken, R.; Weber, W.; Frost, J. W. Phosphoenolpyruvate availability and the biosynthesis of shikimic acid. Biotechnol. Prog. 2003, I9, 808-8 l4. (b) Yi, J.; Li, K.; Draths, K. M.; Frost, J. W. Modulation of Phosphoenolpyruvate Synthase Expression Increases Shikimate Pathway Product Yields in E. coli. Biotechnol. Prog. 2002, I8, I I4l-l I48. Knop, D. R.; Draths, K. M.; Chandran, S. S.; Barker, J. L.; von Daeniken, R.; Weber, W.; Frost, J.W. Hydroaromatic equilibration during biosynthesis of shikimic acid. J. Am. Chem. Soc. 2001, [23, [0173-10182. Dell, K. A.; Frost, J. W. Identification and removal of impediments to biocatalytic synthesis of aromatics from D-glucose: Rate-limiting enzymes in the common pathway of aromatic amino acid biosynthesis. J. Am. Chem. Soc. 1993, [15, Il58l-I1589. Jancauskas, J. Strategies for improving synthesis of 3-dehydroshikimic acid and shikimic acid from D-glucose. M.S. thesis. Michigan State University, 2006. (a) Weaver, L. M.; Herrmann, K. M. Cloning of an aroF allele encoding a tyrosine-insensitive 3-deoxy-D-arabino-heptulosonate 7—phosphate synthase. J. Bacteriol.l990, I72, 6581-6584. (b) Li, K.; Mikola, M. R.; Draths, K. M.; Worden, R. M.; Frost, J. W. Fed-batch fermentor synthesis of 3-dehydroshikimic acid using recombinant Escherichia coli. Biotechnol. Bioeng. 1999, 64, 6l-73. (a) Konstantinov, K. B.; Nishio, N.; Yoshida, T. Glucose Feeding Strategy Accounting for the Decreasing Oxidative Capacity of Recombinant Escherichia coli in Fed-Batch Cultivation for Phenylalanine Production. J. Ferment. Bioeng. 1990, 70, 253-260. (b) Konstantinov, K. B.; Nishio, N.; Seki, T.; Yoshida, T. Physiologically Motivated Strategies for Control of the Fed-Batch Cultivation of Recombinant Escherichia coli for Phenylalanine Production. J. Ferment. Bioeng. 1991, 7], 350-355. (c) Kleman, G. L.; Strohl, W. R. Acetate Metabolism by Escherichia coli in High-Cell-Density Fermentation. Appl. Environ. Microbiol. 1994, 60, 3952-3958. I32 l0 l3 l4 l6 l7 l8 Draths, K. M.; Knop, D. R.; Frost, J.W. Shikimic acid and quinic acid: replacing isolation from plant sources with recombinant biocatalysis. J. Am Chem. Soc. 1999,/21,1603-1604. (a) Michel, G.; Roszak, A. W.; Sauve, V.; Maclean, J.; Matte, A.; Coggins, J. R.; Cygler, M.; Lapthorn, A. J. Structures of shikimate dehydrogenase AroE and its paralog YdiB. J. Biol. Chem. 2003, 278, l9463-l9472. (b) Benach, J.; Lee, I.; Edstorm, W.; Kuzin, A. P.; Chiang, Y.; Acton, T. B.; Montelione, G. T.; Hunt, J. F. The 2.3-A crystal structure of the shikimate 5-dehydrogenase orthologue YdiB from Escherichia coli suggests a novel catalytic enviroment for an NAD- dependent dehydrogenase. J. Biol. Chem. 2003,278, l9l76-l9l82. Datsenko, K.A.; Wanner, B.L. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA. 2000, 97, 6640-6645. Duncun, K.; Chaudhuri, S.; Campbell, M. S.; Coggins, J. R. The overexpression and complete amino acid sequence of Escherichia coli 3—dehydroquinase. Biochem J. 1986, 238, 475. (a) DeFeyter, R. C., J. Pittard. Genetic and molecular analysis of aroL, the gene for shikimate kinase II in Escherichia coli K-IZ. J. Bacterial. 1986, [65, 226- 232. (b) DeFeyter, R. C., J. Pittard. Purification and properties of shikimate kinase II from Escherichia coli K-l2. J. Bacterial. 1986, I65, 33 l-333. Adachi, 0.; Ano, Y.; Toyama, H.; Matsushita, K. Purification and properties of NADP-dependent shikimate dehydrogenase from Glucanabacter axydans [FD 3244 and its application to enzymatic shikimate production. Biasci. Biotechnol. Biochem. 2006, 70, 2786-2789. (a) Park, Y.-C.; Kim, S.-J.; Choi, J.-H.; Lee, W.-H.; Park, K.-M.; Kawamukai, M.; Ryu, Y.-W.; Seo, J.-H. Batch and fed-batch production of coenzyme Q10 in recombinant Escherichia coli containing the decaprenyl diphosphate synthase gene from Glucanabacter subaxydans. Appl. Microbial. Biotechnol. 2005, 67, l92-l96. (b) Cheng, H.; Jiang, N.; Shen, A.; Feng, Y. Molecular cloning and functional expression of D—arabitol dehydrogenase gene from Glucanabacter axyclans in Escherichia coli. FEMS Microbiol. Lett. 2005, 252, 35-42. httg//www.ncbi.nlm.nih.gov/sites/entrez?db=genome http://ergo.integ ratedgenomics.com/ERGO/ Matsushita, K.; Arents, J. C.; Bader, R.; Yamada, M.; Adachi, 0.; Postma, P. W. Escherichia coli is unable to produce pyrroloquinoline quinone (PQQ). Microbiol. 1997, 143, 3l49-3156. I33 19 20 2| 22 23 24 25 26 27 28 29 Pittard, J.; Wallace, B. J. Distribution and function of genes concerned with aromatic biosynthesis in Escherichia coli. J. Bacterial. 1966, 91, 1494-1508. Martin, R. G.; Rosner, J. L. Binding of purified multiple antibioticresistance repressor protein (MarR) to mar operator sequences. Proc. Natl. Acad. Sci. USA 1995, 92, 5456—5460. Rosner, J. L.; Storz, G. Regulation of bacterial responses to oxidative stress. Curr. Top. Cell. Regal. 1997, 35, l63—I77. Saier, M. J.; Paulsen, I.; Sliwinski, M.; Pao, S.; Skurray, R.; Nikaido, H. Evolutionary origins of multidrug and drug-specific efflux pumps in bacteria. FASEB J. 1998, 12, 265-274. Nikaido, H.; Zgurskaya, H. I. AcrAB and related multidrug efflux pumps of Escherichia coli. J. Mol. Microbiol. Biotechnol. 2001, 3, 2l5—218. Franke, 5.; Grass. G.; Rensing, C.; Nies, D. H. Molecular analysis of the copper- transporting efflux system CusCFBA of Escherichia coli. J. Bacteriol. 2003, 185, 3804—3812. Van Dyk, T. K.; Templeton, L. J.; Cantera, K. A.; Sharpe, P. L.; Sariaslani, F. S. Characterization of the Escherichia coli AaeAB Effiux Pump: A metabolic relief valve? J. Bacteriol. 2004, [86, 7196-7204. (a) Pittard, J.; Wallace, B. J. Gene controlling the uptake of shikimic acid by Escherichia coli. J. Bacteriol. 1966, 92, 1070-1075. (b) Brown, K. D.; Doy, C. H. Transport and utilization of biosynthetic intermediate shikimic acid in Escherichia coli. Biachim. Biophys. Acta 1976, 428, 550-562. (c) Whipp, M. J .; Camakaris, H.; Pittard, A. J. Cloning and analysis of the shiA gene, which encodes the shikimate transport system of Escherichia coli K-12 Gene 1998, 209, l85-l92. http://BioCyc.org/ECOLl/substring—search?type=NIL&object=ydiN Johansson, L. Metabolic analysis of shikimic acid producing Escherichia coli. Ph. D. Disertation, Lund University, Sweeden, 2006. Yang, L.; Lin, R.T.; Newman, E.B. Structure of the Lrp-regulated serA promoter of Escherichia coli K-12. Mal. Microbial. 2002,43, 323-333. I34 CHAPTER FQQR Experimental General methods Spectroscopic measurements IH NMR spectra were recorded on a Varian 300 MHz VX-300 FT-NMR spectrometer. Chemical shifts were reported in parts per million (ppm) downfield from internal sodium 3-(trimethylsilyl)propionate-2.2.3.3—d4 (TSP, a = 0.00) with D20 as solvent. TSP was purchased from Lancaster. UV and visible measurements were recorded on a Hewlett-Packard 8452A Diode Array Spectrophotometer equipped with HP 89532A UV-Visible Operating Software or on a Agilent 8453 UV/Vis equipped with Agilent ChemStation A.l0.0 ISI | or on a Beckman DU 530 UV/Vis spectrophotometer. Chromatography Gas chromatography was performed on an Agilent 6890N equipped with an HP-S capillary column (30 m x 0.25 mm x 0.25 micron). Temperature programming began with an initial temperature of 120 °C for 3 min. The temperature was increased to 210 °C at a rate of 15 °C/min, and held at the final temperature for l min. The split injector was maintained at a temperature of 300 °C and the FID detector was kept at 350 °C. Samples analyzed by gas chromatography were derivatized using bis(trimethylsilyl)trifiuoro- acetamide and quantified relative to an internal standard of dodecane against a calibration CUI'VC. I35 Dowex 50Wx8-200 (H*) and Dowex lx8-400 (Cl') were purchased from Sigma- Aldrich. Previously used Dowex 50 (H“‘) was cleaned by treatment with bromine. An aqueous suspension of resin was adjusted to pH 14 by addition of solid KOH. Bromine was added to the solution until the suspension turned a golden yellow color. Additional bromine was added (l-2 mL) to obtain a saturated solution. The mixture stood at room temperature overnight, and the Dowex 50 resin was collected by filtration and washed exhaustively with water followed by 6 N HCl. Dowex 50 (H*) was stored at 4 °C. AG- lX8 (acetate form and chloride form) and hydroxyapatite Bio-Gel HTP gel were purchased from Bio-Rad. Bacteria strains and plasmids All the strains and plasmids used are shown in Table 21. E. coli K-12 strain RB79l was obtained from the American Type Culture Collection (ATCC strain 53622). E. coli A82834,l A82848l were obtained from the E. coli Genetic Stock Center at Yale University. E. coli KL3,2 QPl.l,3 SPl.l,4 were constructed in the lab previously. Plasmid constructions were carried out in E. coli DHSa, which is available from lnvitrogen. Plasmid pKK223-35 is available from GE Healthcare. Plasmid pSU186 were obtained previously by this lab. Plasmids pKD12.l l2,4 pKD12.l38,“t pNR8.l46,7 were constructed in the lab previously. I36 Table 21. Bacterial strains and plasmids. Strain/ Plasmid Relevant Characteristics Source Strain DH50. P ¢801acZAM15 A(lacZYA—argF) U169 recAI ena'Al Invitrogen hst I 7( rk_, ’"k+ ) phaA A: supE44 thi-I gyrA96 relAI E. coli W3l IO wild-type K-12 ATCC E' coli BW251 '3 lacfI rrnBTl4 AlacZWJl6 hst5 l4 AaraBADAH33 CGCS ArhaBADLD78 Glucanobacter wild-type NBRC axydans IFO3244 RB79' W3l IO lacL8lq ATCC RB79IserA::aroB RB79| serA::aroB Lab4 A82834 tsx-352 gin V42 h’ araE353 malT352 CGSC AB2848 tsx-356 gan42 aroD352 LAM- CGSC AL0807 F leuB6 thi-l lach ton A2 I Alach hst supE44 rfrbDI Lab8 araK::CmR araL478::Tn/0 KL3 A82834 serA::aroB Lab2 JYI KL3 AptsHpts/crr::KanR Lab9 SPI '1 RB791 serA::aroB araL478::Tnl () araK I 7::CmR Lab4 SPI JP” SPLI AptsHpts/crrzzKanR Labl0 QPI .l A82848 serA::aroB Lab3 QPI '11”! QPl.l AptsHpts/crr::KanR Lab7 E. coli B serA::aroB Lab” E. coli W3l IO AaraD(new)::FRT-cat-FRT Chapter 2 E. coli B serA::aroB AaraD(new)::FRT-cat-FRT Chapter 2 E. coli B serA::aroB AaraD(new)::FRT Chapter 2 BW251 l3 AserAzzFRT-kan-FRT Chapter 3 BW25l l AserA::FRT Chapter 3 BW251 l3 A_vcliB::FRT-kan-F RT Chapter 3 BW25l l3 AydiB(H l , H2.2)::FRT-kan-FRT Chapter 3 JJ2kan SP1 .I A__\-'diB::FRT-kan-FRT Chapter 3 JJ2 SPI .l AycIiB::FRT Chapter 3 JJ2.2kan SP1 .I AydiB::FRT-kan-FRT Chapter 3 JJ2.2 SPI .l AydiB(H l , H2.2)::FRT Chapter 3 JJ3cat RB791 serA::aroB AaroK ::FRT-cat-F RT Chapter 3 JJ3 RB791 serA::aroB AaraK::FRT Chapter 3 JJ4cat RB791 serA::aroB AaroLzzFRT-cat-FRT Chapter 3 JJ4 RB79l serA::aroB AaroLzzFRT Chapter 3 JJ5cat RB791 serA::aroB AaroK::FRT AaraLzzFRT-cat-F RT Chapter 3 JJ5 RB791 serA::aroB AaraKzzFRT AaraL::FRT Chapter 3 I37 Table 2 I (continued). Strain/ Plasmid Relevant Characteristics Source Plasmid pKK223-3 ApR, Prac GE Healthcare pBluescrip SK (—) ApR [,,, Z Stratagene pKD3 ApR, FRT flanked CmR cosc'2 pKD4 ApR, FRT flanked KmR CGSCl2 pKD46 ApR, araC, P,,,aBy,/3, exa, ts-pAlOl replicon CGSCl2 pCP20 ApR CmR ,Flp, x c1857+ CGSCI3 PFT'A ApR, TcR ,Flp Labl4 pKL5-17A Cm R tktA in pKDIRI 29lA Labz pSC6.090 P,,,,.glfglk, araFFBR ,,tktA P,a,araE, serA Labll PJY I 'ZI6A ApR, s,erA araFFBR P,,,OF, tktA, P,,,CppsA Lab” PJYZ-183 Cm: ,serA, aroFFBR P,,,OF, Fungi/glk, tktA Lab9 PNR4-230 Ap:, araFFBR ,P,a,araE, serA, tktA Lab7 PN R4276 ApR , ppsA, P,,,CaraE, serA, tktA Lab7 pNR8.I46 serA, tktA in p346 Lab7 PNR9-280 ApR, FRT- flanked serA Lab PKDIZ'I '2 ApR, araFF:R ,P.,,,,,aroE serA In pSU I8 Lab“ pKD12.l38 ApR, aroFFB R,,t/