UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Covalent catalyst in the UDP-glucose dehydrogenase Ge, Xue 2000

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2000-486389.pdf [ 6.62MB ]
Metadata
JSON: 831-1.0061493.json
JSON-LD: 831-1.0061493-ld.json
RDF/XML (Pretty): 831-1.0061493-rdf.xml
RDF/JSON: 831-1.0061493-rdf.json
Turtle: 831-1.0061493-turtle.txt
N-Triples: 831-1.0061493-rdf-ntriples.txt
Original Record: 831-1.0061493-source.json
Full Text
831-1.0061493-fulltext.txt
Citation
831-1.0061493.ris

Full Text

C o v a l e n t C a t a l y s i s i n t h e U D P - G l u c o s e D e h y d r o g e n a s e R e a c t i o n by X U E G E B.Sc, Beijing University, 1991 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF CHEMISTRY We accept this thesis as conforming to the required standard: THE UNIVERSITY OF BRITISH COLUMBIA February, 2000 © Xue Ge, 2000 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of C^6-j2iw'r/vy The University of British Columbia Vancouver, Canada Date AAcvrcL\ [ / DE-6 (2/88) 11 Abstract UDP-glucose dehydrogenase from Streptococcus pyogenes (EC 1.1.1.22) is a NAD +-dependent enzyme that catalyzes the 4-electron oxidation of UDP-glucose to UDP-glucuronic acid without the release of an aldehyde intermediate. This enzyme is interesting because its single active site carries out two sequential oxidations, converting an alcohol to a carboxylic acid, whereas most NAD+-dependent dehydrogenases catalyze only a single step of oxidation. The recombinant dehydrogenase was purified to homogeneity and determined to have a kcat of 1.8 ± 0.1 s"1 and an apparent Km of 20 ± 4 pM for UDP-glucose (in 50 mM Trien-HC1 buffer, pH 8.7, 30 °C, with 0.5 mM NAD + and 2 mM DTT). The studies on the enzymatic reaction in H2 i 8 0 showed that only a single solvent-derived oxygen atom is incorporated into the product UDP-glucuronic acid, suggesting that an aldehyde intermediate is involved in the reaction mechanism as opposed to an imine intermediate linked via a lysine residue. The role of the conserved cysteine residue was explored using the two mutant enzymes, Cys260Ser and Cys260Ala. The dramatic loss of activity with the mutant enzymes indicates that Cys260 is catalytically important. The direct observation of a covalent adduct generated from the incubation of the Cys260Ser mutant with NAD + and either UDP-glucose or UDP-gluco-hexodialdose (UDP-Glc-6-CHO) is the best evidence to date for covalent catalysis in the mechanism of UDP-glucose dehydrogenase. Ser260 was identified as the covalently labeled amino acid residue by peptic digestion of the enzyme-adduct and analysis of the peptide mixture using neutral Il l loss mass spectrometry coupled to HPLC. The formation of an ester intermediate from the attachment of a UDP-sugar to the serine residue of the Cys260Ser mutant convincingly supports the involvement of a thioester intermediate in the mechanism employed by the wild type enzyme. UDP-(6,6-di-H)glucose was synthesized to investigate the two hydride transfer steps in the reaction catalyzed by the wild type enzyme. No primary kinetic isotope effect was observed indicating that neither of the two hydride transfer steps is rate-limiting. The best candidate for the rate-limiting step is the hydrolysis of the thioester intermediate. The fact that the Cys260Ala mutant readily oxidizes UDP-Glc-6-CHO indicates that in the absence of an active site nucleophile the mutated dehydrogenase is capable of catalyzing the second oxidation step without the involvement of covalent catalysis. The oxidation is very likely to occur through the hydrated form of the aldehyde that resembles the thiohemiacetal intermediate. The observation that the Cys260Ala mutant does not appreciably catalyze the oxidation of UDP-glucose could be explained if the aldehyde intermediate is tightly bound in the mutant active site and there is no mechanism by which it can be hydrated and proceed forward in the second oxidation step. N A D H formed from the first oxidation step is proposed to be released after the formation of the (thio)hemiacetal intermediate. IV Table of Contents Abstract ii Table of Contents iv List of Figures ix Abbreviations and Symbols .xii Acknowledgements xvi Dedication xvii Chapter One NAD+-Dependent Dehydrogenases 1 1.1 Introduction 2 1.2 Dehydrogenases 3 1.2.1 General Features of the Structure of Dehydrogenases 4 1.2.2 Stereospecificity of Dehydrogenases 5 1.3 Dehydrogenation of Alcohols 7 1.3.1 Alcohol Dehydrogenase 8 1.3.2 Lactate Dehydrogenase 12 1.4 Dehydrogenation of Aldehydes 14 1.4.1 D-Glyceraldehyde-3-phosphate Dehydrogenase 15 1.4.2 Aldehyde Dehydrogenases 17 1.4.3 Oxidation of Aldehydes by Alcohol Dehydrogenases 19 1.5 Dehydrogenases that Catalyze Sequential Oxidations 20 1.5.1 UDP-glucose Dehydrogenase 21 1.5.1.1 Biological Roles of UDP-glucuronic Acid 21 1.5.1.2 The Discovery of the Enzyme 24 1.5.1.3 Previous Mechanistic Studies on Bovine Liver UDP-Glucose Dehydrogenase 24 1.5.1.4 Previous Studies on Bacterial UDP-glucose Dehydrogenases 32 1.5.2 Histidinol Dehydrogenase 32 1.6 Aim of the Thesis 36 Chapter Two Mechanistic Studies on UDP-Glucose Dehydrogenase 38 2.1 Introduction 39 2.2 Overexpression of Recombinant UDP-Glucose Dehydrogenase in E. coli 40 2.3 Purification of the Recombinant UDP-glucose Dehydrogenase 41 2.4 Kinetic Characterization of UDP-Glucose Dehydrogenase 43 2.5 Investigation of Solvent Oxygen Incorporation into UDP-Glucuronic Acid 45 2.6 Synthesis of UDP-(6,6-di-2H)Glucose and Primary Kinetic Isotope Effect Studies.... 55 2.6.1 Synthesis of UDP-(6,6-di-2H)Glucose 56 2.6.2 Studies of the Primary Kinetic Isotope Effect Using UDP-(6,6-di- H)Glucose.... 61 2.7 Summary 61 2.8 Experimental Methods 62 2.8.1 General 62 2.8.2 Protein Determination 62 2.8.3 Strains, Plasmid and Media 63 2.8.4 Purification of Recombinant UDP-Glucose Dehydrogenase 63 2.8.5 Purity Assessment of UDP-glucose Dehydrogenase 65 2.8.6 Molecular Weight Determination of UDP-glucose Dehydrogenase 65 2.8.7 Kinetic Assay of UDP-glucose Dehydrogenase 66 2.8.7.1 Direct Assay Method 66 2.8.8 Solvent-Derived Oxygen Incorporation 68 2.8.9 Synthesis of UDP-(6,6-di-2H)Glucose 69 (a) 1,2-O-Isopropylidene-D-glucuronolactone 69 (b) D-6,6-Di-2H-glucose 70 (c) UDP-(6,6-di-2H)Glucose 71 2.8.10 Primary Isotope Effect Experiment 72 Chapter Three Mechanistic Studies on the Cys260Ser Mutant 73 3.1 Introduction 74 3.2 Purification of the UDP-glucose Dehydrogenase Cys260Ser Mutant 75 3.3 Mutant Enzyme Activity Analysis 77 3.4 Evidence for the Involvement of the Serine Residue in the Mechanism 78 3.4.1 Discovery of the Adduct 78 3.4.2 Identification of the Peptide Containing the Adduct 80 3.4.3 Identification of the Catalytic Nucleophile 90 3.4.4 Stability of the Adduct 93 3.5 Kinetic Studies on the Cys260Ser Mutant 96 3.5.1 Initial Attempts 96 3.5.2 Identification of the Oxidation State of the Adduct 97 3.5.3 Studies with UDP-[6-3H]glucose 98 3.5.4 Re-examination of the Burst 100 3.6 Kinetic Implications and Conclusion 102 3.7 Experimental Methods 109 3.7.1 General 109 3.7.2 Purification of the Cys260Ser Mutant 110 3.7.3 Analysis of the Mutant Enzyme Activity 112 3.7.4 Protein Adduct Formation 112 3.7.5 Mass Spectra 113 3.7.6 Proteolysis 113 3.7.7 MS Analysis of the Proteolytic Digest 114 3.7.8 Chemical Sequencing 115 3.7.9 Stability of Peptide Adduct 115 3.7.10 Investigation of the Adduct by HPLC 116 3.7.11 Stability of UDP-glucuronic Acid 116 3.7.12 Kinetic Studies for the Burst Formation 117 3.7.13 FT-ICR Mass Spectra of the Labeled Peptides : 117 3.7.14 Experiments with UDP-[6-3H]glucose 118 Chapter F o u r Mechanist ic Studies on the Cys260Ala Mutan t 119 4.1 Purification of the UDP-Glucose Dehydrogenase Cys260Ala Mutant 120 4.2 Kinetic Analysis 120 4.3 Studies on the Oxidation of UDP-Glc-6-CHO by the Cys260Ala Mutant 122 4.4 Investigation of the First Oxidation of UDP-glucose by the Cys260Ala Mutant 125 vii 4.4.1 Studies on the Reduction of UDP-Glc-6-CHO by the Cys260Ala Mutant 125 4.4.2 Effect of UDP-glucose Preincubation with the Cys260Ala Mutant on the Oxidation of UDP-Glc-6-CHO 127 4.4.3 Deuterium "Washout" Experiment 128 4.5 Further Investigation of the Reduction and Oxidation of UDP-Glc-6-CHO 131 4.5.1 Comparison of the Incubation of UDP-Glc-6-CHO and NADH with the Wild Type Enzyme and the Cys260Ala Mutant 131 4.5.2 HPLC Analysis of the Reduction Products of UDP-Glc-6-CHO by the Cys260Ala Mutant 133 4.5.3 HPLC Analysis of the Oxidation Products of UDP-Glc-6-CHO by the Cys260Ala Mutant 134 4.6 Mechanistic Implications and Conclusion 136 4.7 Experimental Methods 138 4.7.1 General 138 4.7.2 Purification of the Cys260Ala Mutant 139 4.7.3 Analysis of the Mutant Enzyme Activity 139 4.7.4 Kinetic Constants for the Oxidation of UDP-Glc-6-CHO by the Cys260Ala Mutant 140 4.7.5 Studies on the Reduction of UDP-Glc-6-CHO 140 4.7.5.1 UV Analysis 140 4.7.5.2 Kinetic Analysis of the Reduction of UDP-Glc-6-CHO by the Cys260Ala Mutant 141 4.7.5.3 Effect of Preincubation with UDP-glucose on the Oxidation of UDP-Glc-6-CHO 141 4.7.5.4 Deuterium "Washout" Experiment 142 4.7.6 HPLC Analysis of the Reduction Products 143 4.7.7 HPLC Analysis of UDP-Glc-6-CHO Oxidation Products 143 C h a p t e r F i v e S u m m a r y o f C o n c l u s i o n s 1 4 4 R e f e r e n c e s 1 5 0 Appendix 160 List of Figures Figure 1.1 The anti/pro-R and syn/pro-S correlation of the hydride transfer to NAD + with the substrate binds from the top 7 Figure 1.2 Enzyme-substrate ternary complex of horse liver alcohol dehydrogenase 10 Figure 1.3 Proton shuffling from the hydroxyl of alcohol substrate 11 Figure 1.4 Active site structure of the lactate dehydrogenase ternary complex 13 Figure 1.5 General mechanism for an aldehyde dehydrogenase 14 Figure 1.6 Mechanism of glyceraldehyde-3-phosphate dehydrogenase 16 Figure 1.7 Kirkwood's mechanism 29 Figure 1.8 Kinetic mechanism followed by UDP-glucose dehydrogenase 30 Figure 1.9 Alternatives to thioester formation in the dehydrogenase reaction 35 Figure 2.1 UDP-glucose dehydrogenase reaction coupled with the reduction of INT by diaphorase 44 18 Figure 2.2 Possible products generated from hydrolysis in H2 O by two different mechanisms 46 Figure 2.3 Structures of flavin mononucleotide, FMN (oxidized form) and FMNH 2 (reduced form) 49 Figure 2.4 1 3 C spectra of the carboxyl groups of UDP-glucuronic acid (A) generated in 50 % H 2 1 8 0 and (B) generated in H 2 0 51 Figure 2.5 MALDI TOF MS spectra of UDP-glucuronic acid (A) generated in 50% H218O/50% H 2 1 6 0 and (B) generated in H 2 1 6 0 54 Figure 2.6 Snetkova's synthetic route for D-[6,6-di-3H] glucose 56 Figure 2.7 Moss's synthetic route for D-[6,6-di- H] glucose 57 Figure 2.8 Enzymatic coupling of UDP to glucose 57 Figure 2.9 Synthetic route for UDP-(6,6-di-2H)glucose 58 Figure 2.10 'H NMR spectra of (A) deuterated and (B) undeuterated UDP-glucose 60 Figure 3.1 Electrospray mass spectra of (A) the Cys260Ser mutant, (B) the Cys260Ser mutant after incubation with UDP-glucose, and (C) the Cys260Ser mutant after incubation with UDP-Glc-6-CHO 79 X Figure 3.2 Proposed structures of adduct (A)ester intermediate, (B)hemiacetal intermediate 80 Figure 3.3 Inactivation of glycosidases by 2-deoxy-2-fluoro glycosides. The sugar moiety can vary for different glycosidases 81 Figure 3.4 Electrospray mass spectrometry experiments on Cys260Ser mutant peptic digests (UDP-glucose was employed): (A) labeled sample, total ion chromatogram in normal MS mode; (B) labeled sample, TIC in neutral loss mode; (C) mass spectrum of peptide A in panel B; (D) mass spectrum of peptide B in panel B and (E) unlabeled sample, in neutral loss mode 85 Figure 3.5 Electrospray mass spectrometry experiments on Cys260Ser mutant peptic digests (UDP-glucose was employed): (A) labeled sample, total ion chromatogram in normal MS mode; (B) unlabeled sample, in neutral loss mode; (C) labeled sample, TIC in neutral loss mode; (D) mass spectrum of peptide A in panel C; (E) mass spectrum of peptide B in panel C 86 Figure 3.6 The adduct structure of glycosidase and cleavage site 87 Figure 3.7 The cleavage site of the adduct of Cys260Ser mutant of UDP-glucose dehydrogenase 88 Figure 3.8 Mass spectra of labeled peptide A and peptide B (doubly charged) 89 Figure 3.9 Possible structure of the second labeled peptide 90 Figure 3.10 Tandem MS/MS daughter ion spectrum of unlabeled peptide (m/z 820.7 in doubly charged state) 92 Figure 3.11 HPLC analysis of the hydrolysis products of labeled mutant enzyme. (A)hydrolysis sample, (B)hydrolysis sample spiked with authentic UDP-glucuronic acid 95 Figure 3.12 UV profile of the incubation of Cys260Ser mutant with UDP-glucose (solid line) and UDP-Glc-6-CHO (dashed line). Enzyme concentration is 0.026 mM 100 Figure 3.13 Structure of UDC and alkylation of a cysteine residue 102 Figure 3.14 Proposed mechanism for Cys260Ser mutant 104 Figure 3.15 Proposed mechanism for UDP-glucose dehydrogenase 106 Figure 3.16 Hydrated form of UDP-Glc-6-CHO 109 XI Figure 4.1 Incubation of the Cys260Ala mutant with UDP-Glc-6-CHO and NAD + . [enzyme] = 0.1 mg/mL, [aldehyde] = 1 mM, [NAD+] - 3 mM 121 Figure 4.2 The oxidation of UDP-Glc-6-CHO catalyzed by Cys260Ala mutant 123 Figure 4.3 Reaction catalyzed by the Cysl49Ala mutant of glyceraldehyde-3-phosphate dehydrogenase 124 Figure 4.4 Reaction catalyzed by the wild type glyceraldehyde-3-phosphate dehydrogenase (GAPDH) 124 Figure 4.5 Incubation of UDP-Glc-6-CHO with the Cys260Ala mutant and NADH. [enzyme] = 0.1 mg/mL, [UDP-Glc-6-CHO] = 0.5 mM, [NADH] = 140 mM 126 Figure 4.6 'H NMR spectra at the region 3.7 - 4.0 ppm for (A)recovered UDP-glucose from incubation of deuterated UDP-glucose with Cys260Ala, NADH and NAD + , (B)authentic deuterated UDP-glucose, and (C)undeuterated UDP-glucose 130 Figure 4.7 Kinetic trace of UDP-Glc-6-CHO incubated with NADH and (A) wild type UDP-glucose dehydrogenase (solid line), (B) Cys260Ala mutant (dashed line) 131 Figure 4.8 Possible reactions occur when UDP-Glc-6-CHO was incubated with the wild type UDP-glucose dehydrogenase and NADH 132 Figure 4.9 HPLC analysis of incubation products of 0.21 mM UDP-Glc-6-CHO with 0.125 mg/mL Cys260Ala and 0.188 mM NADH 134 Figure 4.10 HPLC analysis of incubation of 0.21 mM UDP-Glc-6-CHO with 0.125 mg/mL Cys260Ala mutant and 0.6 mM NAD + 135 Figure 5.1 Mechanisms for the wild type enzyme and the Cys260Ser and Cys260Ala mutants 146 Figure A . l Multiple alignment of UDP-glucose dehydrogenase from Streptococcus pyogenes (SA-UDPGDH) with UDP-glucose dehydrogenase from bovine liver (Bov-UDPGDH), GDP-mannose dehydrogenase from Pseudononas (GDPMDH), UDP-N-acetylmanno-saminuronic acid dehydrogenase from E. coli (UDPNAMDH) and a hypothetical sequence from Salmonella (St-Hypo). Strictly conserved residues are given in reverse-background and identities in 4 of the 5 sequences are hashed 161 Figure A . 2 Direct plots of kinetic data for the wild type dehydrogenase reaction 162 Figure A . 3 Direct plots of kinetic data for the dehydrogenase reactions catalyzed by the Cys260Ala mutant 163 Abbreviations and Symbols 8 chemical shift s extinction coefficient A340 absorbance at 340 nm ATP adenosine triphosphate ca. circa Ci Curie Cys260Ala UDP-glucose dehydrogenase with a Cys—»Ala mutation at residue 260 Cys260Ser UDP-glucose dehydrogenase with a Cys—»Ser mutation at residue 260 d doublet D deuterium Da dalton(s) DNA deoxyribonucleic acid DTT dithiothreitol E. coli Escherichia coli Enz enzyme HPLC high pressure/performance liquid chromatography ESI MS electrospray ionization mass spectrometry Hz hertz I N T Q X oxidized /?-iodonitrotetrazolium violet INTred reduced />-iodonitrotetrazolium violet IPTG isopropyl-l-thio-P-D-galactopyranoside J coupling constant (in NMR) kb kilobase kDa kilodaltons koat catalytic rate constant (turnover number) Kj dissociation constant for an enzyme-inhibitor complex K m apparent Michaelis constant LB Luria Bertani medium LC/MS liquid chromatography/mass spectrometry LSIMS liquid second ionization mass spectrometry m multiplet MALDI MS matrix assisted laser desorption ion mass spectrometry min minute NAD + nicotinamide adenine dinucleotide, oxidized form NADH nicotinamide adenine dinucleotide, reduced form NADP + nicotinamide adenine dinucleotide phosphate, oxidized form NADPH nicotinamide adenine dinucleotide phosphate, reduced form NMR nuclear magnetic resonance PAGE polyacrylamide gel electrophoresis Pi orthophosphate ppm parts per million psi pounds per square inch RNA ribonucleic acid rpm revolutions per minute s singlet SDS sodium dodecyl sulfate t triplet TBAHS tetrabutylammonium hydrogen sulphate TIC total ion chromatogram TLC thin layer chromatography Trien triethanolamine UDP uridine diphosphate UDPG uridine 5'-diphosphate glucose UDP-Glc-6-CHO UDP-a-D-g/wco-hexodialdose UMP uridine 5' -monopho sphate UV ultraviolet Vis visible V m a x maximal reaction rate v/v volume to volume ratio Standard Abbreviations for Amino Acids: A Ala alanine C Cys cysteine D Asp aspartic acid E Glu glutamic acid F Phe phenylalanine G Gly glycine H His histidine I Ile isoleucine K Lys lysine L Leu leucine M Met methionine N Asn asparagine P Pro proline Q Gin glutamine R Arg arginine S Ser serine T Thr threonine V Val valine W Trp tryptophan Y Tyr tyrosine Acknowldgements xvi It is a great pleasure to give my sincere thanks to those who helped with the various stages of this thesis over the past few years. They also contribute to make my stay here a very enjoyable one. First of all, I would like to thank my supervisor, Dr. Martin Tanner, for all of his help, advice, encouragement and patience. Without him, none of this would have been possible. I have learned a great deal from him. I also offer many thanks to both past and present members of the Tanner's group, for their friendship, help and suggestions, both scientific and otherwise. I am also grateful to Dr. Stephen Withers, the people in Withers' group and my friends in Vancouver. In addition, the help from the staff of the NMR Laboratory and Mass Spectrometry Laboratory was greatly appreciated. Finally, I would like to thank my family, my parents and sister, for their love, inspiration, support, and encouragement. xvii For my family Chapter One NAD+-Dependent Dehydrogenases 2 1.1 Introduction Most of the biochemical reactions that comprise life are mediated by a series of remarkable biological catalysts known as enzymes. Enzymes differ from ordinary chemical catalysts in several aspects, such as higher reaction rate, milder reaction conditions, greater reaction specificity and the capacity for regulation. Enzymes catalyze a wide variety of chemical reactions using different mechanisms such as acid-base catalysis, covalent catalysis, metal ion catalysis, electrostatic catalysis, proximity and orientation effects and preferential binding of the transition state complex. Acid-base catalysis involves the donation or abstraction of a proton to or from a reactant so as to stabilize the reaction's transition state. Enzymes often use ionizable amino acid side chains as general acids/bases at the catalytic site. Covalent catalysis occurs through nucleophilic or electrophilic attack of the catalyst on the substrate with transient formation of a covalent bond to give a reactive intermediate. The active site can contain various nucleophiles such as the thiol of cysteine, the hydroxyl of serine, the s-amino group of lysine, the imidazole moiety of histidine and the carboxylate of glutamate and aspartate. In addition, electrophilic coenzymes like thiamine pyrophosphate and pyridoxal phosphate can be used. Metal ions in the metalloenzymes or metal-activated enzymes can function by binding to substrates to properly orient them for reaction, or electrostatically stabilizing or shielding negative charges. Metal ion-bound water molecules are potent sources of hydroxide ions at neutral pH that could act as nucleophiles. The charged groups in the enzyme active site can be arranged to provide a low dielectric constant where the electrostatic interactions are much stronger than they are in aqueous solutions. Electrostatic catalysis therefore may also serve to stabilize the transition states of the enzymatic reactions. Thoroughly understanding the mechanisms of enzymes can help us to design and develop potent inhibitors that may serve as new drugs, or engineer enzymes to produce novel compounds, or develop new catalysts with valuable industrial applications. 1.2 Dehydrogenases Dehydrogenases are enzymes that catalyze oxidation-reduction reactions. The substrates of the reactions can be alcohols or aldehydes. Primary and secondary alcohols can be oxidized to the corresponding aldehydes and ketones, respectively. Aldehydes can be either reduced to alcohols or oxidized to carboxylic acids. The dehydrogenases discussed in this thesis utilize either nicotinamide adenine dinucleotide (NAD+) or nicotinamide adenine dinucleotide phosphate (NADP+) as coenzyme. Most of them are specific for only one of the coenzymes, a few use both. During the reaction catalyzed by a dehydrogenase, NAD + or NADP + is reduced to NADH and NADPH while the substrate is oxidized. In most cases, the reaction is reversible, so that carbonyl compounds may be reduced by NADH or NADPH. The discussion here will focus on NAD+-dependent dehydrogenases. NAD+ N A D H 4 N A D P + N A D P H NAD+-dependent dehydrogenases have been studied in detail, especially alcohol dehydrogenases. Some reviews on the mechanism of the reactions catalyzed by NAD +-dependent dehydrogenases and properties of NAD + can be found in the works of Fersht1 and Oppenheimer . 1.2.1 General Features of the Structure of Dehydrogenases The dehydrogenases usually contain two domains: a catalytic domain and a coenzyme binding domain. The catalytic domains are different in structure for each dehydrogenase. However, the coenzyme binding domains of all the dehydrogenases whose structures have been solved have the same tertiary structure. This general feature is known as the Rossmann fold and commonly consists of six strands of parallel P sheets with five intervening ot-helices.3 A pattern of GXGXXG in the amino acid sequence has been observed for the Rossmann fold in many dehydrogenases.4 5 1.2.2 Stereospecificity of Dehydrogenases NAD + functions as a coenzyme in oxidation-reduction reactions by reversibly accepting a hydride at the 4-position of the nicotinamide ring to yield NADH. It is generally thought that a direct transfer of a hydride occurs, but a radical intermediate cannot be totally ruled out.5 The 4-position in the dihydronicotinamide ring is prochiral. The C-4 proton may be labeled according to the RS convention. H o H e N H 2 I R N A D H It was known in the early 1950s that the hydride transfer is stereospecific between the substrate and NAD + . 6 ' 7 The evidence is based on studies with yeast alcohol dehydrogenase. The enzyme transfers 1 mole of deuterium from CH 3CD 2OH to NAD + . When the NAD 2H that is isolated from this reaction mixture is incubated with unlabeled acetaldehyde and the enzyme, all the deuterium is lost from the NAD H and is incorporated into the alcohol that is generated. Therefore, the deuterium or hydrogen is transferred stereospecifically to one face of NAD + and then transferred back from the same face. The reaction is also stereospecific with respect to substrate. When acetaldehyde is reduced by (i?)-NAD H in the presence of yeast alcohol dehydrogenase, only a single enantiomer, (i?)-monodeuteroethanol, is produced. On the other hand, incubating C H 3 C D O with ordinary NADH will yield the (^ -enantiomer. Unlabeled acetaldehyde and deuterated coenzyme are generated by the oxidation of the (R)-enantiomer of C H 3 C H D O H with NAD + . The stereospecificity of the hydride transfer has since been supported by numerous studies on different dehydrogenases. Through the use of deuterated substrates, the side of NAD + that the dehydrogenases transfer the hydride to has been determined. The enzymes have been classified as "A" or "B" on this basis. Hydride is transferred to the pro-R position of NADH in Class A dehydrogenases and to the pro-S position of NADH in Class B dehydrogenases. For example, alcohol and lactate dehydrogenases belong to Class A, glyceraldehyde-3-phosphate and glutamate dehydrogenases belong to Class B. The stereospecificities of more than 150 NAD-linked enzymes were summarized in You's paper. The origin of this specificity remains to be determined. It has been observed that Class A enzymes catalyze the reduction of the more reactive carbonyls, while Class B is associated with the less reactive ones. There exists an anhlpro-R and syn/pro-S correlation and this can be rationalized if the substrate binding domain is always on top of the coenzyme (Figure 1.1). The NAD + is bound to Class A enzymes (pro-R), such as lactate dehydrogenase, with the nicotinamide ring in the anti conformation about the glycosidic bond. When the syn rotamer of the nicotinamide ring is bound to Class B enzymes, such as glyceraldehyde-3-phosphate dehydrogenase, the hydride of the substrate is transferred to the pro-S position of the nicotinamide ring. This has been observed for all dehydrogenases that oxidize alcohols. The opposite correlation applies to enzymes that bind the substrate below the coenzyme. Glutathione dehydrogenase is one example for this correlation. 7 Substrate Substrate NH 2 OH OH OH OH Anti/pro-R Syn/pro-S Figure 1.1 The ant\/pro-R and syn/pro-S correlation of the hydride transfer to NAD + with the substrate binding from the top. 1.3 Dehydrogenation of Alcohols The mechanism of alcohol oxidation generally proceeds in two steps, the formation of an alkoxide anion followed by the transfer of hydride.2'9'10 The hydride transfer occurs with neutralization of the charges on the anionic alkoxide and the cationic pyridinium. The stereospecificity of the hydride transfer is highly conserved in related dehydrogenases. Traditionally, alcohol dehydrogenases have been grouped into the short-chain and the long-chain dehydrogenase families according to their size. Short-chain dehydrogenases have molecular weights up to 25 kDa, with about 250 residues per subunit. This family of dehydrogenases shows pro-S specificity (Class B dehydrogenase) and has no requirement for metal ions. Insect alcohol dehydrogenase (Drosophild), bacterial ribitol dehydrogenase (Bacillus megaterium), prokaryotic glucose dehydrogenase (Klebsiella aerogenes) and hydroxysteroid dehydrogenase are some examples. The coenzyme binding domain is assigned to the N-terminal half of this group." A review on the reaction mechanism and 8 three-dimensional structure of this type of dehydrogenase can be found in Jornvall's paper.12 Long-chain dehydrogenases have about 350-375 residues with molecular weights of 35 000-40 000 Da, show pro-R specificity (Class A dehydrogenase) and require zinc. The coenzyme-binding domain is at the C-terminal half of the sequence, opposite to that of the short-chain family. Liver and yeast alcohol dehydrogenases are common representatives of this group because of their known structures and mechanisms. The properties of these two enzymes will be discussed in the next section. 1.3.1 Alcohol Dehydrogenase The NAD+-dependent alcohol dehydrogenases catalyze the interconversion of various alcohols and the corresponding aldehydes and ketones. Most of the alcohol dehydrogenases are zinc-metalloenzymes with broad substrate specificity. The substrates can be aliphatic or aromatic alcohols. Among numerous dehydrogenases, the enzymes from horse liver and yeast have been studied in the greatest detail.2'9'10'13 The horse liver enzyme (LADH) is a dimer with molecular weight of 80 kDa. Each subunit binds one NAD + and two zinc ions, one of which is directly involved in catalysis while the other plays a structural role. The yeast enzyme is a tetramer with molecular weight of 145 kDa. Each subunit binds one NAD and one zinc ion. Both enzymes use the same overall reaction mechanism, but show different rate-limiting steps and pH dependence. Extensive structural studies have been done on the horse liver alcohol dehydrogenase. The enzyme has two subunit forms, E and S, which differ by six amino acids in their primary sequences.14 The enzyme can exist as EE, ES or SS. Each subunit is divided into two domains, the coenzyme binding domain and the catalytic domain.15 These two domains are separated by a deep active-site cleft. The coenzyme binding domain has the same folding structure as many other NAD -dependent dehydrogenases. Two Zn ions bind to the catalytic domain, and the catalytic zinc ion is located at the bottom of a hydrophobic pocket formed at the junction of two domains.16 The structure of the reactive ternary complex containing the enzyme, NAD + , and 4-bromobenzyl alcohol has been solved at 2.9-A resolution.16 This was possible because there is a favorable equilibrium between this complex and the enzyme-bound ternary complex of the reaction products, NADH and 4-bromobenzaldehyde. Figure 1.2 illustrates the active site structure of the enzyme-substrate ternary complex that was obtained from the crystallographic studies.16'17 The catalytic zinc ion is coordinated to the sulfur atoms of Cysl74 and Cys46, the imidazole nitrogen NE2 of His67, and the alkoxide of the alcohol substrate. In the free enzyme, a water molecule provides the fourth metal ligand. The alcohol substrate is thought to bind as the alkoxide and is supported by the pH dependence study of the binding of substituted alcohols.18 Studies on the kinetic isotope effects are also consistent with the hydride transfer taking place from an alkoxide.19 10 H (His-67) 4 1 (Cys-174) \ > ^ , S - ( C y s ^ 6 ) NADT s . ! / binding N . H 2 \ ~*Zn 2 + domain j \ \ 0 R R' Substrate binding pocket Figure 1.2 Enzyme-substrate ternary complex of horse liver alcohol dehydrogenase. The mechanism of the oxidation of an alcohol by liver alcohol dehydrogenase proceeds first by the binding of NAD + followed by a conformational isomerization of the catalytic domain.20'21 This adjustment brings the enzyme to the right ionization state for the substrate binding. An alcohol substrate then binds to the enzyme-NAD+ binary complex and directly coordinates to the catalytic zinc, replacing the water. The removal of the hydroxyl proton from the bound alcohol has been proposed to take place via a proton-relay system involving His51, a ribose hydroxyl of the bound NAD + , and Ser48 based on the X-ray 22 structure and chemical modification studies (Figure 1.3). 11 (Ser-48) ^ / T \ H a ^ H - ' w H , n H ' ^ ( H i s - 5 1 ° O OH N H 2 (Ser-48) + H B ! JL / ^ M ~ * 0 ' ^ O ' O H x (H i s -51 ) N H 2 Figure 1.3 Proton shuffling from the hydroxyl of alcohol substrate. The next step after the formation of the ternary complex and deprotonation is the hydride transfer to NAD + , followed by the release of the resulting aldehyde. The dissociation of the aldehyde is faster than the rate of hydride transfer, so that the two ternary complexes, ERCH 2 OHNAD + and ERCHONADH, do not have time to equilibrate. The final step is the dissociation of NADH, which is usually the rate-determining step in the oxidation reaction catalyzed by liver alcohol dehydrogenase. NADH is known to bind more tightly than NAD + , so even though the equilibrium constant for the oxidation reaction in solution greatly favors NAD + and alcohol, the equilibrium position of the enzyme-bound species is close to unity. In the reduction direction, the rate-limiting step can be different depending on the aldehyde 12 substrate. For aromatic aldehydes, the dissociation of the enzyme-alcohol complex is rate-limiting, whereas for the reduction of acetaldehyde the hydride transfer is the rate-limiting step. For the yeast alcohol dehydrogenase, the enzme-product complexes dissociate rapidly so that the hydride transfer step is rate-limiting.1 1.3.2 Lactate Dehydrogenase NAD+-dependent L-lactate dehydrogenase (LDH) catalyzes the reversible oxidation of L-lactate to pyruvate: OH O I II CH3-CH-COOH + NAD + C H 3 - C - C O O H + NADH + H + Lactate dehydrogenases from various species are usually tetramers of molecular weight 140 kDa and are pro-R specific (Class A dehydrogenases). They do not require metal ions for catalysis, and the enzyme from bacterial sources is usually allosterically activated by fructose-l,6-diphosphate. The animal enzyme has three type of isoenzymes, M (muscle type), H (heart muscle type) and X (testis type). These isoenzymes have different amino acid compositions and kinetic properties. The reaction catalyzed by the lactate dehydrogenase proceeds via an ordered mechanism that is very similar to the alcohol dehydrogenase with coenzyme binding first, followed by the substrate binding. The enzyme therefore binds lactate or pyruvate only in the presence of coenzyme. The postulated structure of the active ternary complex of the dogfish enzyme is illustrated in Figure 1.4.9 The carboxylate group of the substrate forms a salt 13 bridge with the guanidinium of Argl71. The hydroxyl of the lactate forms a hydrogen bond with the unprotonated imidazole ring of His 195. This residue not only helps to orient the substrate close to the C-4 position of the nicotinamide ring, but also acts as an acid-base catalyst, removing the proton from the lactate hydroxyl during the oxidation. NH 2 R-N+ x> - H -(His-195) < - : N ^ N H H 3C" I OY - > 0 i • H 2 N o ^ N H 2 C I NH / (Arg-171) Figure 1.4 Active site structure of the lactate dehydrogenase ternary complex. In the reaction catalyzed by lactate dehydrogenase, hydride transfer is very fast and the dissociation of pyruvate is slow, so that the two ternary complexes equilibrate before the establishment of the steady state. The bound equilibrium position is pH dependent and favors lactate and NAD + at neutral pH. The free equilibrium position however favors the products pyruvate and NADH. In the steady state of lactate oxidation, the dissociation of NADH is the rate-limiting step that is analogous to the reaction catalyzed by alcohol dehydrogenase.1 14 1.4 Dehydrogenation of Aldehydes The carbonyl carbon of an aldehyde is electron poor because of the rc-bond to oxygen. To remove a hydride from this positively, polarized carbon is therefore difficult. A general solution used by the NAD+-dependent aldehyde dehydrogenases is to first add a nucleophile to give a tetrahedral intermediate (eg. hemiacetal), and then to remove the hydride yielding an acyl-nucleophile product. A final hydrolysis of this acyl adduct generates the free acid. In the case of formaldehyde dehydrogenase, the covalent intermediate is a thiohemiacetal formed between the substrate and glutathione.23 For glyceraldehyde-3-phosphate dehydrogenase, the adduct is a thiohemiacetal formed between the substrate and a cysteine residue. The mechanism of glyceraldehyde-3-phosphate dehydrogenase has been well elucidated and generally serves as a paradigm for the NAD+-dependent aldehyde oxidation. In this mechanism (Figure 1.5), an active site cysteine adds to the carbonyl to form a thiohemiacetal intermediate that anchors the substrate and activates the C-l carbon to promote hydride transfer. O B : - — H O S Enz H Figure 1.5 General mechanism for an aldehyde dehydrogenase. The thiohemiacetal is then oxidized by NAD + to yield a thioester and the hydrolysis of the thioester gives the free acid. In the case of glyceraldehyde-3-phosphate dehydrogenase, 15 the thioester is phosphorylated by orthophosphate. The details of the mechanism of this enzyme will be discussed in 1.4.1. Not all aldehyde oxidations require covalent catalysis. Alcohol dehydrogenases and histidinol dehydrogenase have been reported to catalyze aldehyde oxidations without any active site thiol groups. Details will be discussed in 1.4.3 and 1.5.2. 1.4.1 D-Glyceraldehyde-3-phosphate Dehydrogenase D-Glyceraldehyde-3 -phosphate dehydrogenase catalyzes the reversible oxidative phosphorylation of glyceraldehyde-3-phosphate to 1,3-diphosphoglycerate, using NAD + as a coenzyme. a . / H OPCV H — C — O H + N A D + + HP042- H — C — O H + N A D H + H + CH 2OP0 3 2- CH 2OP0 3 2-The enzyme is tetrameric with a molecular weight of about 150 kDa. Allostery in coenzyme binding has been well documented. Muscle and bacterial enzymes exhibit negative cooperativity24, whereas the yeast enzyme shows positive cooperativity25. On the other hand, glyceraldehyde-3-phosphate binds independently to all four subunits. The coenzyme binding domain and the catalytic domain are separated by an active site cleft with the essential cysteine residue in its center.26'27 The coenzyme binding domain is very similar to that of other dehydrogenases; NAD + is bound in the same way as in liver alcohol dehydrogenase and lactate dehydrogenase except for a 180° rotation around the ribose-nicotinamide glycoside bond, making it the pro-S specific enzyme (Class B dehydrogenase). 16 Glyceraldehyde-3-phosphate dehydrogenase catalyzes the reaction in a multistep manner (Figure 1.6) and can be separated into an oxidoreduction step and a phosphorolysis step. The reaction begins with the formation of a thiohemiacetal between glyceraldehyde-3-phosphate and a cysteine residue of the enzyme. The next step is a rapid oxidation of the thiohemiacetal by NAD + to form a thioester intermediate. Then the NADH that was produced dissociates, and a second NAD + associates. Finally, the thioester is attacked by orthophosphate to give the acylphosphate, 1,3-diphosphoglycerate, in the presence of the newly bound NAD + . The acyl transfer is very slow in the absence of bound NAD + . The rate-limiting step is the release of NADH at high pH (pH 8.4) and phosphorolysis of the acyl-enzyme at low pH (pH 5.4).28 E - S H O H O N A D + E — S — C — R N A D + E - S - C 'i \ R O E - S - C N A D H - E — S - C R N A D H Figure 1.6 Mechanism of glyceraldehyde-3-phosphate dehydrogenase. 17 The structure of the enzyme-substrate complex has been deduced from various crystal structure and model-building studies on different species. Cysl49 has been reported to provide the essential thiol for the bacterial enzyme from Bacillus stearothermophilus21 His 176 is located near Cysl49 and has been postulated to act as a chemical activator by enhancing the reactivity of the Cysl49 thiol, possibly through the formation of an ion pair with the imidazole ring. Further thiol activation might be introduced by the formation of a charge-transfer complex with the nicotinamide of NAD + . Hisl76 also functions as a general acid to protonate the carbonyl oxygen of glyceraldehyde-3-phosphate and facilitate the formation of the tetrahedral intermediate. A similar activation could occur during the phosphorolysis step as well, by forming a hydrogen bond between the imidazole ring of His 176 and the oxygen of the carbonyl group of the thioester. His 176 has also been suggested to be the general base catalyst facilitating the hydride transfer from the thiohemiacetal. Both Cysl49 and His 176 are conserved in at least five sequenced enzymes from bacteria, yeast and 30 muscle tissue who share more than 30% sequence identity. 1.4.2 Aldehyde Dehydrogenases Most of the aldehyde dehydrogenases are NAD + specific with a few NADP + utilizing examples and they oxidize a variety of aldehyde substrates to the corresponding carboxylic acids. This superfamily is comprised of enzymes of narrow and broad specificities. Broad specificity enzymes include class 1 (cytosolic), class 2 (mitochondrial), class 3 (tumor associated), and microsomal enzymes. The narrow specificity enzyme group contains betaine aldehyde dehydrogenases and succinate, glutamate, methylmalonate or 2-hydroxymuconate semialdehyde dehydrogenases. 18 X-ray crystal structures of a few enzymes from class 1, 2 and 3 aldehyde dehydrogenases were solved recently, indicating that they all share a similar overall structure. ' ' Each subunit of the aldehyde dehydrogenase contains a coenzyme-binding domain and a catalytic domain. The coenzyme-binding domain comprises residues at the N-terminal half of the sequence and has the Rossmann fold structure. Mammalian liver class 1 and class 2 aldehyde dehydrogenases have been studied extensively. They are tetrameric enzymes with identical subunits of molecular weight about 55 kDa. The active form of the enzyme has been found to be a dimer of dimers. It has been postulated that the mechanism of an aldehyde dehydrogenase is similar to that of glyceraldehyde-3-phosphate dehydrogenase. A cysteine residue acts as a nucleophile, attacking the substrate carbonyl group to form a thiohemiacetal. This covalent intermediate is then oxidized to a thioester by hydride transfer to NAD + . Hydrolysis of the thioester gives the acid product. When sequences of NAD + and/or NADP+-dependent aldehyde dehydrogenases are aligned, including representative examples of broad and narrow specificity enzymes, a single invariant cysteine residue is found.34'35 Through chemical modification36'37 and site-directed mutagenesis studies38 on sheep liver cytoplasmic and rat liver mitochondrial aldehyde dehydrogenases, Cys302 has been found to be located at the active site and is thought to provide the nucleophilic thiol. Cysteine 302 is the only strictly conserved cysteine residue among all the available sequences of mammalian class 1 and class 2 aldehyde dehydrogenases. X-ray crystal structural studies on the bovine mitochondrial32 and the sheep liver cytosolic aldehyde dehydrogenases have shown that Cys302 is indeed located at the active site. 19 A pre-steady state burst of NADH has been observed with the rat liver mitochondrial enzyme, indicating that the rate-limiting step occurs after the formation of NADH. 3 9 Deacylation has been proposed to be the rate-limiting step for this enzyme, the horse liver mitochondrial enzyme40 and human liver cytosolic enzyme41. The most important evidence is that the value of kcat is greater for aldehydes with electron-withdrawing groups than for those without.39'42 The highly conserved Glu268 has been proposed as a general base in a role similar to His 176 in glyceraldehyde-3-phosphate dehydrogenase, activating Cys302 for formation of the thiohemiacetal.37'43 The fatty aldehyde dehydrogenase from Vibrio harveyi whose sequence diverges significantly from other aldehyde dehydrogenases and shows preference for NADP + also has a conserved cysteine residue 4 4 The enzyme is a homodimer with a molecular weight about 110 kDa and it catalyzes the oxidation of long-chain fatty aldehydes. Cys289 in this enzyme corresponds to Cys302 of mammalian class 1 and class 2 enzymes and has been shown to be the essential cysteine residue for catalysis using site-directed mutagenesis 4 4 1.4.3 Oxidation of Aldehydes by Alcohol Dehydrogenases Alcohol dehydrogenases from different sources have also been reported to catalyze the dismutation of an aldehyde into a mixture of a carboxylic acid and the corresponding alcohol.2'45'46'47'48 It has been postulated from studies on horse liver alcohol dehydrogenase that the hydrated gem-diol is the substrate for oxidation.49 Aldehydes are present in solution as an equilibrium mixture of non-hydrated and hydrated forms. Formaldehyde is completely hydrated and is one of the best substrates for the dismutation. Acetaldehyde is about 50% hydrated, whereas benzaldehyde is about 10% hydrated. A question has been raised as to 20 whether the dehydrogenase directly binds the hydrated form of an aldehyde or catalyzes the hydration of the bound aldehyde substrate. It has been proposed that it is possible for horse liver alcohol dehydrogenase to catalyze sequential oxidation of alcohol to carboxylic acid by an ordered bi-bi mechanism at low substrate concentration.50 Henehan and Oppenheimer have observed that the dismutation (a lag phase) precedes the linear steady state net production of NADH in the horse liver alcohol dehydrogenase catalyzed oxidation of aldehydes.51 A possible mechanism for the formation of hydrate at this enzyme's active site has been postulated by Olson's group through their molecular dynamics studies.52 The hydrate formation could be catalyzed by delivering the zinc-bound hydroxide to the bound aldehyde substrate. However, a complete understanding of the origin of the enzyme-bound hydrate form of an aldehyde requires further experimental examination. 1.5 Dehydrogenases that Catalyze Sequential Oxidations This is a small class of enzymes that catalyze the four-electron (two-fold) oxidation of an alcohol to the corresponding carboxylic acid. It includes UDP-glucose dehydrogenase, histidinol dehydrogenase, GDP-mannose dehydrogenase, UDP-N-acetyl-glucosamine dehydrogenase and UDP-N-acetyl-mannosamine dehydrogenase. UDP-glucose dehydrogenase and histidinol dehydrogenase have been studied extensively and little is known about the other enzymes. 21 1.5.1 UDP-glucose Dehydrogenase UDP-glucose dehydrogenase (UDPGDH, EC 1.1.1.22) catalyzes the NAD+-dependent oxidation of UDP-glucose (UDPG) to UDP-glucuronic acid (UDPGA) without the release of an aldehyde intermediate and is the subject of this thesis. UDP-glucose UDP-glucuronic acid 1.5.1.1 Biological Roles of UDP-glucuronic Acid UDP-glucose dehydrogenase is crucial for both eucaryotes and procaryotes, since it provides the only pathway for formation of UDP-glucuronic acid in all organisms except plants. UDP-glucuronic acid plays a wide variety of roles in different organisms. In many strains of pathogenic bacteria, such as group A and C streptococci54, Streptococcus pneumoniae type 35 5 and Escherichia coli K5 5 6 , UDP-glucuronic acid is used in the biosynthesis of a polysaccharide antiphagocytic capsule. The formation of this capsule is known to be essential for bacterial virulence ' and serves to protect the bacteria from the host's immune system.59'60 It has been reported that the acapsular, mutant strains are nonpathogenic and very sensitive to phagocytosis.61 In Streptococcus pneumoniae type 3, the basic component of its capsular polysaccharides is cellobiuronic acid, a disaccharide consisting of glucuronic acid p (1—»4) 22 linked to glucose. E. coli K5 capsular polysaccharide is constructed from the glucuronic acid P (l->4) linked to N-acetylglucosamine. Group A streptococci (Streptococcus pyogenes) are human pathogens that colonize the skin and mucous membranes of their host. The hyaluronic acid capsule on the surface of Streptococcus pyogenes has been shown to be a major virulence determinant for group A streptococci in the human host. Hyaluronic acid is a high molecular weight glycosaminoglycan that is composed of 250 - 25 000 (3 (l->4) linked disaccharide units that is synthesized by the alternate addition of UDP-glucuronic acid and UDP-N-acetylglucosamine with a (3 (1—>3) linkage by the membrane-associated enzyme hyaluronate synthase. In addition to its role as a virulence factor in pathogenic streptococci, hyaluronate is a ubiquitous component of vertebrate and invertebrate connective tissues. The connective tissues such as cartilage, tendon, skin, and blood vessel walls, consist of collagen and elastin fibers embedded in a gel-like matrix known as ground substance. Ground substance is composed largely of glycosaminoglycans, unbranched polysaccharides of alternating uronic acid and hexosamine residues. Solutions of glycosaminoglycans are slimy and mucus like due to their high viscosity and elasticity. Hyaluronic acid is an important glycosaminoglycan component of ground substance, synovial fluid (the fluid that lubricates the joints), and the vitreous humor of the eye. Depressed levels of UDP-glucose dehydrogenase have been observed in cells lining the synovium of rheumatoid arthritic joints. Hyaluronate has high molecular mass and numerous mutually repelling anionic groups (glucuronic acid residues) that make it a rigid and highly hydrated molecule. Hyaluronate solutions have a viscosity that is shear dependent (an object under shear stress has equal and opposite forces applied across 23 its opposite faces). This viscoelastic behavior makes it serve as an excellent biological shock absorber and lubricant and it is used as a viscoelastic material in some surgical procedures.64 UDP-glucuronic acid also plays a role as a detoxifying agent in the secretion of waste metabolites from the livers in mammals. It helps to solubilize various nonpolar poisons and degradation products by forming glycosidic links to them. The formation of water soluble glucuronides allow these compounds to be eliminated from the body in the aqueous media of urine or bile.65 UDP-glucuronic acid also provides D-glucuronic acid to other glycosaminoglycans, chondroitin sulfate, heparan sulfate, heparin, and through C5 epimerization, dermatan sulfate. Chondroitin-4-sulfate has N-acetyl-D-galactosamine-4-sulfate residues in place of hyaluronate's N-acetyl-D-glucosamine residues. It is a major component of cartilage and other connective tissues. It has been found that mutation of the UDP-glucose dehydrogenase gene of Drosophila melanogaster (designated sugarless) leads to an absence of heparan sulfate side chains on proteoglycan core proteins and is identical in phenotype to the classical wingless mutation.66 Heparin consists predominantly of alternating a (1—>4)-linked residues of D-glucuronate-2-sulfate and N-sulfo-D-glucosamine-6-sulfate. It occurs in the intracellular granules of the mast cells that line arterial walls, especially in the liver, lungs and skin. It inhibits the clotting of blood by binding to antithrombin and is widely used clinically for postsurgical patients.67 It is also the required substrate for glucuronidation (glucuronosylation) of many substances, including xenobiotics, opioids, androgens and heme proteins through the action of UDP-glucuronosyl transferases in the liver. In eukaryotes, the UDP-glucuronic acid is also used as the precursor of UDP-D-xylose. 24 1.5.1.2 The Discovery of the Enzyme UDP-glucose dehydrogenase was first detected in bovine liver in 1954 by 68 Strominger's group . They partially purified an enzyme from a water extract of calf liver acetone powder and found that this enzyme can oxidize UDP-glucose to UDP-glucuronic acid. This discovery outlined for the first time the biosynthetic pathway of UDP-glucuronic acid required in the synthesis of glucuronides. The enzyme has subsequently been shown to be present in other organisms including plants, yeast and bacteria. The enzyme was purified to homogeneity fifteen years later by Zalitis and Feingold from fresh bovine liver homogenate after ammonium sulfate treatment and chromatography.69 The enzyme is a homohexamer with a subunit of 52 kDa. Since then the dehydrogenase has also been partially purified from a number of eukaryotic and prokaryotic organisms. The enzyme from E. coli is a dimer with a subunit of 47 kDa. UDP-glucose dehydrogenase from Streptococcus pyogenes was the first to be cloned and overexpressed.70 A year later, the primary structure of the enzyme from bovine liver was 71 79 determined. Molecular cloning of the human and mouse enzymes were achieved in 1998. 1.5.1.3 Previous Mechanistic Studies on Bovine Liver UDP-Glucose Dehydrogenase UDP-glucose dehydrogenase is unusual and of interest because its single active site carries out two sequential oxidations involving two N A D + molecules, whereas most N A D + -dependent dehydrogenases catalyze only a single step of oxidation. It functionally combines activities similar to alcohol and aldehyde dehydrogenases. 25 The mechanism of this enzyme has been studied by several groups in the 1970's and much of the work has focused on the bovine liver enzyme. Initial attempts to detect an intermediate in the reaction were unsuccessful. This suggested that all intermediates must remain bound to the enzyme during the reaction. The overall reaction is found to be irreversible. When UDP-glucuronic acid and NADH are incubated with the enzyme, no production of UDP-glucose can be detected nor is there any enzyme-dependent oxidation of NADH. The logical candidate for the intermediate would be the aldehyde produced from the first 2-electron oxidation of the primary alcohol group. Nelsestuen and Kirkwood enzymatically synthesized UDP-a-D-g/wco-hexodialdose (UDP-Glc-6-CHO), the aldehyde corresponding to UDP-glucose, from UDP-galactose through the use of galactose oxidase and UDP-glucose 4-epimerase.74 They have shown that UDP-Glc-6-CHO can be converted to UDP-glucuronic acid by UDP-glucose dehydrogenase and NAD + . It can also be reduced to UDP-glucose by incubating with the enzyme and NADH. The fact that UDP-glucose is an inhibitor of the reduction of UDP-Glc-6-CHO suggests that the two nucleotide sugars bind to the same site on the enzyme. This enzyme bound aldehyde intermediate cannot be trapped by reaction with carbonyl trapping agents, such as hydroxylamine, thiosemicarbazide or cyanide. HO UDP-Glc-6-CHO 26 These observations indicate that the enzymatic oxidation of UDP-glucose to UDP-glucuronic acid proceeds by two steps with UDP-Glc-6-CHO as the enzyme-bound intermediate. The first oxidation step appears to be the reversible oxidation of UDP-glucose to UDP-Glc-6-CHO with the concomitant reduction of the first equivalent of NAD + . This step is highly specific for the configuration at carbon 4 of the glucose ring because UDP-galactose cannot undergo the oxidation reaction with this enzyme and NAD + . The second step, formation of UDP-glucuronic acid from UDP-Glc-6-CHO coupled with the reduction of the second equivalent of NAD + , accounts for the irreversibility of the overall reaction. It is well known that the oxidation of free aldehydes to free acids is, in general, a thermodynamically irreversible process. The second oxidation step is less specific toward substrate structure because both UDP-Glc-6-CHO and UDP-Gal-6-CHO can be oxidized by UDP-glucose dehydrogenase. Ridley and Kirkwood have shown that UDP-glucose dehydrogenase from bovine liver abstracts the pro-R hydrogen stereospecifically at carbon 6 of the glucose moiety of the UDP-glucose since enzymatic incubation of UDP-Glc-6-CHO with tritium-labeled NADH results in the stereospecific labeling of the pro-R hydrogen at carbon-6.75 However, they also found that a prolonged incubation of the bovine liver enzyme with UDP-glucose, NAD + and tritium-labeled NADH results in the equivalent labeling of both hydrogens at carbon-6. In this experiment, a large ratio of NADH to NAD + was used to favor the exchange reaction rather than the oxidation of UDP-glucose. The observation of the random labeling suggests that the pro-S hydrogen at carbon-6 is also abstracted by a reversible process which must involve a derivative of the carboxyl group of UDP-glucuronic acid that is capable of 27 reversible hydrogenation-dehydrogenation. It is then the hydrolysis of this derivative that accounts for the irreversibility of the overall reaction. It is well established that oxidation of thiohemiacetals to thiol esters is a reversible reaction and it appears to be a general feature in the mechanism employed by the dehydrogenases that oxidize aldehydes. Bovine liver UDP-glucose dehydrogenase has 12 cysteine residues per subunit. Titration of the enzyme with 5,5'-dithiobis-(2-nitrobenzoate) (DTNB) has shown that six thiol groups are necessary for activity.77 Since the enzyme is a hexamer, it seems that one critical thiol group per subunit is essential for catalytic activity. DTNB treated enzyme completely loses the activity and lacks the ability to either oxidize or reduce UDP-Glc-6-CHO due to the modification of an essential cysteine residue. The 5-thio-2-nitrobenzoate group of DTNB is bulky and negatively charged and could prevent the access of the substrate to the active site. Replacement of the 5-thio-2-nitrobenzoate group with a less bulky cyanide fully restores the enzyme's ability to reduce UDP-Glc-6-CHO but not to oxidize it.76 This indicates that the essential thiol of a cysteine residue is involved in the second reversible dehydrogenation step. This cysteine residue is protected from modification by the addition of UDP-glucose or UDP-xylose indicating that it is located at the active site. The requirement of an active site cysteine residue for catalysis has also been indicated in other modification studies using iodoacetate78, iodoacetamide,78 5-[[(iodoacetamido)ethyl] 79 80 81 amino]naphthalene-l-sulfonic acid , iodoacetamidofluorescein , and 6,6'-dithionicotinate . It has been indicated that Cys275 in each subunit provides the reactive thiol by proteolytic 71 80 digestion and sequencing studies on the labeled enzyme. ' 28 In 1975, Kirkwood proposed that the initial oxidation of UDP-glucose generated UDP-Glc-6-CHO in the active site and the essential thiol added to its carbonyl to form a 76 thiohemiacetal. This proposal also explains the tight binding of the aldehyde intermediate. However, Kirkwood observed that when the essential thiol group was derivatized with cyanide to prevent the formation of the thiohemiacetal, the enzyme still did not release aldehyde into the incubation medium. When the cyanide modified enzyme was incubated with UDP-glucose and NAD + , and then reduced with NaBH4, a covalent enzyme-substrate 82 adduct was identified. This adduct appeared to be the reduction product of a Schiff s base formed between the aldehyde intermediate and an active site lysine. The enzyme can be labeled in the same manner by treatment with UDP-Glc-6-CHO alone, followed by NaBH4 reduction. Derivatization (by treatment with formaldehyde, followed by NaBH4 reduction) of 6 of the approximately 168 lysine residues per enzyme molecule results in destruction of 47% of the enzyme activity and UDP-glucose offers considerable protection against inactivation and decreases the extent of derivatization. These results suggest that an essential lysine is in the vicinity of the active site and is involved in the mechanism. Kirkwood therefore proposed that the oxidation of UDP-glucose generates an imine intermediate covalently bound with a lysine residue and that the aldehyde is never actually formed in the normal reaction. The overall mechanism used by bovine liver dehydrogenase was proposed by Kirkwood in 1977 (Figure 1.7). It proceeds with an unusual oxidation of the C-6 hydroxyl of UDP-glucose that involves a chemically unprecedented displacement of hydride to form the imine intermediate directly without aldehyde formation. This intermediate is then transferred to an active site cysteine residue to form a thiohemiacetal intermediate, again this would involve an unusual displacement of an amine by water. A hydride transfer from the 29 thiohemiacetal intermediate then takes place to produce an enzyme-bound thioester intermediate and another molecule of N A D H . Finally, irreversible hydrolysis of the thioester produces the product UDP-glucuronic acid. O U D P Figure 1.7 Kirkwood's mechanism. However, the suggestion of imine formation is somewhat at odds with earlier 1 8 0 labeling studies which showed that only one solvent oxygen atom is incorporated into the Ql product acid. If an imine intermediate is formed, the two oxygen atoms on the C-6 of UDP-glucuronic acid would be expected to be introduced from the solvent. A sequestered water molecule derived from the original alcohol would be needed to hydrolyze the thioester in order to explain the 1 8 0 labeling studies. The initial velocity and product inhibition studies on UDP-glucose dehydrogenase have shown that UDP-glucuronic acid is a competitive inhibitor of UDP-glucose and a non-competitive inhibitor of N A D + . This enzyme has been proposed to follow a bi-uni-uni-bi 30 OA ping-pong mechanism (Figure 1.8) , in which UDP-glucose binds to the enzyme first followed by the binding of NAD + , which is reduced and released. The second mole of NAD + is then bound, reduced and released. The irreversible step in the reaction occurs after the release of the second mole of NADH but before the release of UDP-glucuronic acid. UDPG NAD+ NADH NAD + NADH UDPG A B P B P R _1_J t I t t E EA EAB-EA'P EA' EA'B-ER'P ER E Figure 1.8 Kinetic mechanism followed by UDP-glucose dehydrogenase. The proposed kinetic mechanism is consistent with Kirkwood's chemical mechanism. Both redox steps are reversible. The hydrolysis of the thioester is not only the irreversible step but also been postulated to be the rate-limiting step by examination of rate equations. The ordered binding of substrates is well established for dehydrogenases, however most of the 2-electron dehydrogenases bind the coenzyme NAD + first. The stereospecificity with regards to NAD + is pro-S (Class B dehydrogenase) for both steps, and with regards to UDP-glucose, the pro-S and then the pro-R hydrogens are sequentially removed.83 Half-of-the-sites reactivity in bovine liver UDP-glucose dehydrogenase has been observed during binding studies with the substrates. The enzyme was found to bind only three molecules of UDP-glucose at saturation per hexamer. Iodoacetate and iodoacetamide have been used to investigate the active site thiol groups and half-of-the-sites reactivity has 87 also been observed. It has been shown that iodoacetate reacted exclusively with the active 31 site thiol groups in a biphasic manner. Three of the six subunits are rapidly inactivated by carboxymethylation, and the other three react at a rate that is an order of magnitude lower. The loss of enzyme activity is greater than the degree of incorporation of iodoacetate, indicating that the carboxymethylation of one subunit influences the catalytic activity of its neighbor. The same study has also shown that the active site thiols are protected by UDP-glucose and UDP-xylose but not by NAD + or NADH. Studies on the reaction with iodoacetamide also demonstrated the similar biphasic inactivation of the enzyme. The six subunits have been found to be identical. Therefore, the half-of-the-sites reactivity may be due to the induction of conformational asymmetry following the derivatization of one half of the subunits. These results suggest that the fundamental unit in the hexamer is a dimer, and the bovine liver UDP-glucose dehydrogenase functions as a trimer of dimers. Half-of-the-sites reactivity is not unusual for aldehyde dehydrogenases. Precedents have been established by studies on glyceraldehyde-3-phosphate dehydrogenase and several other aldehyde dehydrogenases.88'89 Studies on partially denatured samples of the enyzme led to the speculation that two active sites, one on each subunit, might be involved in the overall reaction. Eccleston and Kirkwood suggested that the aldehyde intermediate generated from the first oxidation step, which is covalently bound to a lysine residue via an imine linkage, is transferred to a cysteine thiol in the active site of an adjacent subunit where the second oxidation takes place.90 However, this proposal is at odds with the results obtained from resonance energy transfer experiments indicating that the six active sites of the hexamer were spatially remote.79 32 1.5.1.4 Previous Studies on Bacterial UDP-glucose Dehydrogenases The only bacterial UDP-glucose dehydrogenase that has been purified and characterized in the early studies is that from a strain of E. coli, MC 153 in the 1970s.91'92 It is a dimer with subunit molecular weight of 47 kDa and has 2 cysteine residues per subunit. Only one paper, published in 1973, has documented the results of mechanistic studies on this enzyme 9 1 The enzyme is reported to be active as a dimer with a pH optimum of 9.0. The Km value is 1.0 mM for UDP-glucose and 0.05 mM for NAD + , respectively. UDP-glucuronic acid is found to be a competitive inhibitor with UDP-glucose at high UDP-glucose concentrations, while it is an inhibitor that displays positive cooperativity at low UDP-glucose concentrations. In the case of the bovine liver enzyme, no cooperativity has been observed. NADH is a strictly competitive inhibitor with NAD + and uncompetitive with UDP-glucose, exhibiting the same inhibition pattern as the bovine liver enzyme. Inhibition of the bacterial enzyme by UDP-xylose is competitive with respect to UDP-glucose and noncompetitive with NAD + . The observation that UDP-xylose does not function as a cooperative inhibitor is in contrast to the enzymes from bovine liver and other eukaryotic species that use UDP-xylose as a donor of the D-xylosyl moiety , and suggests that D-xylose is not present in the complex saccharides of the E. coli strain MCI 53. 1.5.2 Histidinol Dehydrogenase The enzyme histidinol dehydrogenase catalyzes the NAD+-dependent two-fold oxidation of L-histidinol to histidine without the release of an aldehyde intermediate. This is the final step in the biosynthesis of L-histidine in bacteria, fungi, and plants. 33 F=N = N HN x> Histidinol H N V s> x / Dehydrogenase H-CH 2 f y < CH 2 -NH 2 2NAD+ 2 NADH H NH2 CH2OH COO" Histidinol Histidine Histidinol dehydrogenase is a dimer with a molecular weight of 83 kDa. Similar to UDP-glucose dehydrogenase, no aldehyde intermediates were observed or trapped and an active site thiol appeared to be important for histidinol dehydrogenase activity. Most of the mechanistic studies have been carried on the enzyme from Salmonella typhimurium. This enzyme was generally thought to employ a mechanism similar to that of UDP-glucose dehydrogenase. The observation that the synthetic compound L-histidinaldehyde can be used as a substrate for either reduction or oxidation by histidinol dehydrogenase suggests that histidialdehyde is an intermediate in catalysis.94 When histidinol dehydrogenase was inactivated by 7-chloro-4-nitro-2,l,3-benzoxadiazole, a modification reagent that resembles the substrate histidinol, two conserved cysteine residues, Cysll6 and Cys377, were covalently labeled.95 Since the presence of L-histidinol protected Cysll6 from modification, it is assumed to be in the active site and supply the essential thiol that participates in the second oxidation step. The reaction catalyzed by histidinol dehydrogenase also follows a bi-uni-uni-bi ping-pong mechanism.96 Histidinol binds first to the enzyme followed by NAD + and histidine is the last product to be released. The purified enzyme does not contain either coenzyme or 34 substrates. It has been observed that the two hydrides are transferred to the pro-R (Class A dehydrogenase) position of NADH in both steps, and that first the pro-S and then the pro-R hydrogens of substrate histidinol are removed.97'98 The stereochemistry of NADH is opposite from that of UDP-glucose dehydrogenase. Unlike UDP-glucose dehydrogenase, no half-of-the-sites reactivity has been found in histidinol dehydrogenase. Zn has been found to bind to histidinol dehydrogenase and is essential for enzyme activity99, whereas no metal ion is required by UDP-glucose dehydrogenase. Each subunit of histidinol dehydrogenase binds one zinc atom. In the case of the zinc containing alcohol dehydrogenases, horse liver alcohol dehydrogenase binds two zinc atoms per subunit, whereas yeast alcohol dehydrogenase binds only one zinc atom per subunit. In horse liver alcohol dehydrogenase, one zinc is involved in catalysis and the other one plays a structural role. The catalytic zinc is ligated to two cysteine sulfur atoms and a nitrogen atom from a histidine imidazole ring. Conserved histidine residues have also been proposed to be the candidates for zinc ion ligands in histidinol dehydrogenase.100'101 Chemical modification studies have shown that cysteine is essential for the second oxidation.95 However, subsequent site-directed mutagenesis experiments indicated that the conserved cysteine residues of histidinol dehydrogenase are not essential for catalysis.102'103 The Grubmeyer group reported that Cysll6Ser, Cysl53Ser, Cysll6Ala, Cysl53Ala and the double mutant Cysll6,153Ala of Salmonella typhimurium histidinol dehydrogenase showed 102 normal kcat and Km values. The mutagenesis studies by Ohta's group have also shown that neither of the two conserved cysteine residues in cabbage histidinol dehydrogenase, Cysll2 and Cysl49 is involved in the reaction.103 They concluded that this enzyme does not follow the glyceraldehyde-3-phosphate dehydrogenase paradigm which involves a thiohemiacetal in 35 the NAD+-dependent aldehyde oxidation. Therefore site-directed mutagenesis studies may lead to conclusions different from those reached by chemical modification studies, especially when a bulky thiol-derivatizing reagent is used. An active site nucleophile must still be involved in the oxidation of the aldehyde intermediate. Alternative mechanisms have been proposed.102 The possible pathways include the direct oxidation of an aldehyde hydrate, and the oxidation of aldehyde addition products formed with non-cysteine amino acid side chains such as serine or histidine followed by hydrolysis (Figure 1.9). O O" R- - H X O R - ^ X X = OH X = O-Ser X = O-Glu X = N-Lys X = N-His Hydrate Hemiacetal Hemiacetal Hemiaminal Hemiaminal Acid Ester Anhydride Amide Acylimidazole Figure 1.9 Alternatives to thioester formation in the dehydrogenase reaction. The most likely alternative is the addition of water to histidinaldehyde to form the aldehyde hydrate (gem-diol). A gem-diol entity was first postulated to serve as the form of substrate oxidized by lactate dehydrogenase during glyoxylate oxidation.104 Many alcohol dehydrogenases, such as that from Drosophila melanogaster45 and horse liver alcohol dehydrogenase49, have been reported to be able to further oxidize their aldehyde products via hydrated species. The detailed studies on horse liver alcohol dehydrogenase have been 36 discussed in section 1.4.3. Mutation of the essential nucleophilic cysteine residue to an alanine residue in the glyceraldehyde-3-phosphate dehydrogenase from E. coli. generated a mutant enzyme that has substantial nonphosphorylating activity and used a different mechanism from that of the parent enzyme.105 The mutant enzyme converts glyceraldehyde-3-phosphate into 3-phosphoglycerate instead of 1,3-diphosphoglucerate. Catalysis studies indicated that no acylenzyme intermediate is formed during the catalytic event and a gem-diol species serves as a reacting substrate. 1.6 Aim of the Thesis This thesis describes experiments designed to investigate the mechanism utilized by UDP-glucose dehydrogenase from Streptococcus pyogenes. Studies on this enzyme would provide insight into the mechanism employed by the small class of enzymes that catalyze the NAD+-dependent two-fold oxidation of an alcohol to the corresponding carboxylic acid. Kirkwood's mechanism for the bovine liver enzyme proceeds with an unusual displacement of hydride that leads directly to the formation of an imine intermediate. 18 However, this is at odds with earlier O labeling studies which showed that the product acid has only one oxygen atom incorporated from the solvent. We suspect that the aldehyde is the reaction intermediate (product of the first oxidation step) in the dehydrogenase mechanism. Histidinol dehydrogenase was originally thought to employ a mechanism similar to UDP-glucose dehydrogenase because a cysteine residue was found to be essential for catalysis by chemical modification studies. However, site-directed mutagenesis experiments indicated that the conserved cysteine residues of histidinol dehydrogenase are not essential for enzyme activity. Thus chemical modification studies can be misleading. 37 We used site-directed mutagenesis to study the UDP-glucose dehydrogenase mechanism. Site-directed mutagenesis is a better way to make subtle and controlled changes in essential amino acid residues than chemical modification. The hasB gene for the UDP-glucose dehydrogenase from Streptococcus pyogenes was the first and only one that had been cloned prior to this thesis work. In addition, very little is known about the bacterial enzyme. Therefore this enzyme was chosen for study. Since UDP-glucuronic acid is used in the biosynthesis of a polysaccharide antiphagocytic capsule in many strains of pathogenic bacteria, and the formation of capsule is known to be essential for virulence, elucidation of the mechanistic mode of the action of a bacterial enzyme may shed light on the development of novel antibiotic drugs. In this thesis, Chapter Two describes the overexpression and purification of the recombinant Streptococcus pyogenes dehydrogenase enzyme, and its subsequent characterization. Chapters Three and Four are devoted to the mechanistic studies on the two mutants, Cys260Ser and Cys260Ala, respectively. Based on the results of these studies, the reaction mechanism utilized by UDP-glucose dehydrogenase is proposed. Chapter Five is the summary of the conclusions that are drawn from the studies conducted in this thesis. 38 Chapter Two Mechanistic Studies on UDP-Glucose Dehydrogenase 39 2.1 Introduction As discussed in Chapter One, the most detailed studies on UDP-glucose dehydrogenase have focused on the properties and mechanism of the bovine liver enzyme, and very little is known about the bacterial enzyme. The enzyme from E. coli strain MC 153 has been shown to be smaller in size and contains less cysteine residues per subunit than the bovine liver enzyme. The enzyme from E. coli consists of two identical subunits with a subunit molecular weight of 47 kDa and contains 2 cysteine residues per subunit.92 On the other hand, the bovine liver enzyme is composed of six subunits with a molecular weight of 52 kDa and 12 cysteine residues per subunit. The lower molecular weight and possibly less structural complexity (a dimer versus a hexamer) of the bacterial enzyme may reflect its less complex control requirements. In 1993, Dougherty and van de Rijn cloned and expressed the first UDP-glucose dehydrogenase gene, hasB from group A streptococci (Streptococcus pyogenes)™ It was not until several years later that another bacterial UDP-glucose dehydrogenase gene, the gene KfaC of E. coli K5 was cloned and expressed.106 The sequence of UDP-glucose dehydrogenase from Streptococcus pyogenes shows 31.1% identity to the bovine liver enzyme71 and 53.8% identity to the E. coli K5 enzyme106. It also shares 74% sequence similarity and 57% sequence identity with the S. pneumoniae enzyme. ' There are two cysteine residues among the 402 amino acids of the dehydrogenase from Streptococcus pyogenes. Cys260 is the only one that is conserved.71 Lys204 and Lys320 are also conserved.71 A sequence alignment of the Streptococcus pyogenes UDP-glucose dehydrogenase with the bovine liver UDP-glucose dehydrogenase, GDP-mannose dehydrogenase from Pseudomonas, UDP-N-acetylmannosaminuronic acid 40 dehydrogenase from E. coli and a hypothetical sequence from Salmonella71 is shown in Appendix A. In this work, the recombinant /zasB-encoded UDP-glucose dehydrogenase was purified and characterized. The enzymatic reaction was carried out in 1 80 labeled water in order to examine the extent of solvent isotope incorporation. In addition, di-deuterated UDP-glucose (UDP-(6,6-di- H)glucose) was synthesized and used to investigate the possibility of primary kinetic isotope effects on the reaction rate. 2.2 Overexpression of Recombinant UDP-Glucose Dehydrogenase in E. coli Preparation of homogeneous proteins from natural sources usually requires a multistep purification procedure and the low expression level of the protein makes it difficult to obtain large quantities of material. The molecular biological techniques of cloning and overexpression of the gene that codes for a protein allow one to produce relatively large amounts of proteins required for extensive mechanistic studies. Molecular cloning or genetic engineering involves the insertion of a foreign DNA into a vector. A vector is a double-stranded DNA molecule that acts as a vehicle for the foreign DNA segment and contains the necessary sequences required for replication. Plasmids are the most commonly used vectors and are circular DNA duplexes of 1 - 200 kb. The vector with its added DNA segment, or insert, can be introduced into a suitable host organism, such as E. coli or yeast in which the insert sequence is expressed. Usually the cloned gene is engineered to have properly positioned regulatory sequences for RNA and protein synthesis, so that the host may produce large quantities of the RNA and protein specified by that gene. This is 41 called overexpression since the expression level of the gene is significantly higher than that of a gene coding for a typical endogenous protein. The hasB gene from group A streptococci was found to encode UDP-glucose dehydrogenase by Dougherty and van de Rijn through gene comparison studies on encapsulated strains and acapsular mutants.70 They cloned the hasB gene into the inducible pl5AT7 expression vector under the control of the T7 promoter to give the recombinant plasmid pGAC147. The T7 promoter is a sequence of DNA that can be recognized by RNA polymerase to initiate transcription. The plasmid pGAC147, which was kindly provided by Dr. Ivo van de Rijn of Wake Forest University Medical Center, North Carolina, was transformed into E. coli JM109 (DE3), grown in the presence of chloramphenicol, and induced with isopropyl-l-thio-13-D-galactopyranoside (IPTG). This resulted in a high level of expression of a protein with a molecular weight of approximately 47 kDa (by sodium dodecyl sulfate - polyacrylamide gel electrophoresis) and that demonstrated high levels of UDP-glucose dehydrogenase activity. 2.3 Purification of the Recombinant UDP-glucose Dehydrogenase The previous partial purification of UDP-glucose dehydrogenase from E. coli required 2-mercaptoethanol in all purification buffers for protein stability. The procedure included several ammonium sulfate precipitation steps at different concentration levels and three types of liquid chromatography with ammonium sulfate precipitation after each column. The purification procedure for the overexpressed protein developed in this work is far less complicated. Dithiothreitol (DTT) was used instead of 2-mercaptoethanol; all protein manipulations were performed at 4 °C; and 10% glycerol was added to all purification 42 buffers. The crude cell lysate was prepared by subjecting the cells to two passes through a chilled French pressure cell followed by removal of the insoluble cell debris by ultracentrifugation. The crude cell lysate was then partially purified by loading it onto a diethylaminoethylcellulose (DE52) weak anion-exchange column and eluded with a step-wise gradient of NaCl in pH 8.7 triethanolamine-HCl (Trien-HCl) buffer. This step removed the strong-binding negatively charged impurities from the crude lysate. The eluent fractions were assayed for UDP-glucose dehydrogenase activity using the direct continuous assay method described in the next section. The most active fractions were pooled, concentrated and desalted by dialyzing against salt-free buffer overnight. The resulting protein was further purified by anion-exchange HPLC, eluting with a linear gradient of 0 - 0.2 M NaCl in pH 8.7 Trien-HCl buffer. Fractions were assayed again for activity and the purity of the dehydrogenase fractions was analyzed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE). The purified UDP-glucose dehydrogenase appeared to be homogeneous as judged by SDS-PAGE. The purification gave about 30-35 mg of protein per 500 mL culture, which accounts for 60% - 7 0 % of the activity in the crude extract. The 12 % SDS-PAGE gel that stained with Coomassie Brilliant Blue showed a single protein band with a molecular weight about 46 kDa, using (3-lactoglobulin (19 kDa), carbonic anhydrase (29 kDa) and BSA (66 kDa) as molecular weight standards. The protein band with this molecular weight was also the biggest band in the gel of crude cell lysate. The molecular weight of UDP-glucose dehydrogenase was determined by electrospray ionization mass spectrometry (ESI MS). This technique was developed in the 1980s and proved to be a breakthrough for the direct characterization of macromolecules, particularly 43 proteins and nucleic acids.109'110 Only molecules that ionize in the solvent can be generated as gas phase ions by this "soft" ionization method. In this technique, a high velocity flow of nebulizer gas shears microdroplets from the liquid sample stream in the ionspray inlet. A high voltage is applied to the sprayer and each droplet exists in a highly charged state containing both solvent and the charged species of interest. While it is believed that most of the net charge resides near the surface of the droplet, details on the mechanism of the electrospray ionization process are not fully understood. It has been proposed that the microdroplets rapidly reduced to nanodroplets (small solvated macro-ions), which then shrink by evaporation to yield the molecular ions ultimately detected by mass spectrometry.111 The molecular weight of UDP-glucose dehydrogenase was determined to be 45 489 ± 6 Da by ESI MS, consistent with the number predicted from the gene sequence, 45 484 Da. 2.4 Kinetic Characterization of UDP-Glucose Dehydrogenase UDP-glucose dehydrogenase is an NAD+-dependent enzyme and one of the products, NADH, shows a characteristic absorption at 340 nm. In addition, the overall reaction is effectively irreversible. This enables one to follow the reaction progress continuously by monitoring the absorbance increase at 340 nm. Using this method, the enzyme purified as described in the previous section showed a specific activity of 2.4 ±0 .1 units/mg. (A unit is defined as the amount of dehydrogenase that produces 1 pmole UDP-glucuronic acid per minute at 30 °C). Activity was measured by mixing the enzyme with UDP-glucose and NAD + and monitoring the rate of NADH production using a UV spectrometer. The kinetic constants were Km = 20 ± 4 pM for UDP-glucose, Km = 65 ± 6 pM for NAD + , £ c a t = 1.8 ± 0.1 44 s"1 (in 50 mM Trien-HCl buffer, pH 8.7, 2 mM DTT, 30 °C). The Km of one substrate was measured under fixed, saturating concentration of the other substrate. A coupled assay method was also used to measure the catalytic constants. The reaction catalyzed by the dehydrogenase is coupled to a second enzymatic reaction in which NADH is used to reduce j?-iodonitrotetrazolium violet (INT, 2-[4-iodophenyl]-3-[4-nitro-phenyl]-5-phenyltetrazolium chloride). This is catalyzed by diaphorase and serves to regenerate NAD + (Figure 2.1). The reduced form of INT has a characteristic absorption at 500 nm and a larger extinction coefficient of 12 800 M"1 cm"1 as compared to 6 220 M"1 cm"1 for NADH. Using this method, the Km values of UDP-glucose and NAD + have been found to be 9.4 ± 0.3 pM and 61 ± 8 pM respectively, and the kcat was determined to be 1.7 ± 0.02 s"1 (in 50 mM Trien-HCl buffer, pH 8.7, 30 °C). The value of kcat is consistent with the value obtained by the direct assay method. The reason for a lower Km value for UDP-glucose is unclear. Figure 2.1 UDP-glucose dehydrogenase reaction coupled with the reduction of INT by diaphorase. 45 Further characterization of UDP-glucose dehydrogenase from S. pyogenes was done by Robert Campbell in our laboratory at the same time the work described in this thesis was carried out and his results are summarized in this paragraph.112 The optimum pH for enzyme activity is 8.5 - 9.0. Neither NADH nor NAD + is found to be tightly bound to the dehydrogenase since the UV spectrum of the enzyme did not show any significant absorption above 300 nm and addition of excess UDP-glucose did not cause any changes. The enzyme activity is dependent on the presence of reduced thiols and is not affected by treatment with ethylene diamine tetraacetate (EDTA, disodium salt) or carbonyl trapping reagents. The enzyme appears to be active as a monomer as judged by size exclusion chromatography. This is in contrast to the half-of-the-sites reactivity observed from the bovine liver enzyme. This observation also does not support the possibility of more than one subunit involved in the reaction as mentioned in Chapter One. Very recent X-ray crystallographic work, however, suggests the enzyme is a dimer (unpublished results from our laboratory). The initial velocity and product inhibition kinetic studies indicate the enzyme follows a bi-uni-uni-bi ping-pong mechanism with UDP-glucose bound first and UDP-glucuronic acid released last, which is consistent with that of the bovine liver UDP-glucose dehydrogenase and histidinol dehydrogenase. UDP-glucuronic acid and UDP-xylose were found to be competitive inhibitors with respect to UDP-glucose and showed K\ values of 200 pM and 2.7 pM, respectively. 2.5 Investigation of Solvent Oxygen Incorporation into UDP-Glucuronic Acid As described in Chapter One, Ordman and Kirkwood observed that a lysine residue is essential for the dehydrogenase activity of each subunit in the bovine liver enzyme. They 46 proposed that the aldehyde is not an actual intermediate in the reaction and that UDP-glucose is converted directly to an enzyme-linked imine involving an active site lysine residue. The formation of the imine intermediate would expel the original oxygen atom that was present at the C-6 position of UDP-glucose (Figure 2.2). Therefore the two oxygen atoms of the product would likely come from the solvent. Figure 2.2 Possible products generated from hydrolysis in H 2 1 8 0 by two different mechanisms. However, this appears to be at odds with the 1 80 labeling study of the beef liver enzyme in which only one solvent oxygen atom was found to be incorporated into the Q1 product. In that study, UDP-glucuronic acid was obtained by the enzymatic oxidation of UDP-glucose in 20 percent excess H 2 1 8 0. Free D-glucuronic acid was obtained by hydrolysis, converted to the O-trimethylsilyl derivative, and analyzed by mass spectrometry. The data indicated that the derivative contained approximately 18 percent excess l 8 0, suggesting that each molecule of D-glucuronic acid contained one atom of oxygen from the solvent. Further 47 experiments showed that there is no non-enzymatic exchange reaction between the oxygen atoms of D-glucuronic acid or UDP-glucuronic acid and the medium. When the labeled UDP-glucuronic acid was decarboxylated with UDP-glucuronate carboxy-lyase, the resulting UDP-xylose contained no excess 1 80. This proved that the 1 80 was located in the carboxyl group of UDP-glucuronic acid. Therefore, only one oxygen atom of the carboxyl group of UDP-glucuronic acid derives from water and an imine intermediate is unlikely to occur in the case of the bovine liver enzyme. It is possible however, that a tightly bound water derived from the C-6 hydroxyl of UDP-glucose could be re-introduced during the hydrolysis step. Histidinol dehydrogenase was originally thought to employ a mechanism similar to UDP-glucose dehydrogenase and an essential lysine residue was thought to participate in the reversible oxidation-reduction interconversion of histidinal and histidinol.113 A Schiff base generated from a lysine residue and histidinal was proposed as a covalently bound intermediate and was thought to protect the unstable histidinal against decomposition at alkaline pH. The Schiff base intermediate is analogous to the imine intermediate in the reaction catalyzed by UDP-glucose dehydrogenase. In 1990, Grubmeyer and Insinga showed that only a single oxygen was derived from solvent in the histidinol dehydrogenase reaction.114 They were able to prove the location of the oxygen using isotope-induced NMR chemical shift experiments. The chemical shift of any NMR-active nucleus depends on the induced magnetic fields of the circulating electrons in its immediate environment. The substitution of different isotopes, such as 1 8 0 for 1 60, will perturb the electronic environment of the carbon nuclear that directly bond to it and induce a measurable change in the chemical shift. Batiz-Hernandez and Bernheim proposed that the isotope induced shift arose from the changes in 4 8 vibrational levels of bonds between the observed nucleus and the isotope.115 The electrical fields generating from the oscillations of the bonded atoms in the applied magnetic field diminish the absolute shielding at the observed nucleus. When a lighter isotope is substituted by a heavier one, its average distance to the observed nucleus is slightly shorter with a smaller vibrational amplitude, producing smaller electric fields which, when felt at the probe nucleus, do not deshield it as strongly as would those from the lighter isotope. This accounts for the almost universally observed trend of shift to higher field (lower ppm numbers) on substitution of a heavier for a lighter isotope at the probe nucleus. Risley and Van Etten have observed an upfield isotopic shift caused by the substitution of l s O for 1 6 0 on the 1 3 C N M R chemical shift of directly bonded carbon atoms116, an effect that is additive in multiply substituted carbon atoms117. This technique can be used to study reaction pathways. 1 1 8 ' 1 1 9 The magnitude of the isotope-induced shift is determined principally by the structure of the carbon-oxygen functional group. 1 2 0 ' 1 2 1 An 1 8 0 isotope-induced shift in 1 3 C N M R spectroscopy was used by Grubmeyer and Insinga in the investigation of the origin of the two carboxyl oxygen atoms of histidine. They incubated the Salmonella typhimurium histidinol dehydrogenase with highly enriched L-(hydroxymethyl-13C)histidinol and N A D + in 50% H 2 1 8 0 . The 1 3 C N M R spectrum of the carboxyl carbon of the product L-(carboxy-13C)histidine showed two peaks (50:50) separated by 0.02 ppm, indicating incorporation of a single solvent oxygen. If both carboxyl oxygen atoms come from solvent, three peaks (25:50:25) would be expected as shown in a non-enzymatic solvent equilibration experiment of the carboxyl group of (l- l3C)acetate. No non-enzymatic equilibration of the carboxyl group oxygen atoms with solvent was observed with 49 histidine. The retention of one original histidinol oxygen provided strong evidence against the existence of a Schiff base intermediate in the reaction catalyzed by histidinol dehydrogenase. The combination of carrying out the reaction in 1 8 0 labeled water and examining the 18 13 O isotope-induced shift in C N M R spectroscopy has been demonstrated as a direct and convenient method to study the solvent isotope incorporation, better than the traditional indirect mass spectrometric method in which sample isolation and handling are inconvenient and time consuming. We have used this method.to investigate the origin of the carboxyl oxygens of UDP-glucuronic acid with an enzyme from bacterial sources. In our studies, the 1 3 C was present at natural abundance levels as opposed to a 13C-enriched substrate. In order to simplify the experiment and remove the need to purify the product from excess N A D + / N A D H , an N A D + regenerating system was employed. Flavin mononucleotide (riboflavin phosphate)(FMN) (Figure 2.3) and oxygen continuously oxidize the N A D H produced during the reaction and maintains the initial concentration of N A D + . C H 2 O P 0 3 2 - C H 2 O P 0 3 2 -( C H O H ) 3 ( C H O H ) 3 0 H O F M N F M N H 2 Figure 2.3 Structures of flavin mononucleotide, F M N (oxidized form) and F M N H 2 (reduced form). 50 This NAD + regenerating system was first reported by Jones and Taylor122 and then used in Grubmeyer and Insinga's O labeling studies on the histidinol dehydrogenase. FMN is a natural hydride acceptor for many in vivo enzyme-catalyzed oxidations of NADH. Jones and Taylor have demonstrated it to be an effective and convenient reagent for in situ recycling of NAD + by oxidizing the NADH generated from NAD+-dependent enzymatic reactions. FMN is highly soluble in aqueous solution and the reduced FMN (FMNH2) can be oxidized rapidly in air so that the FMN concentration remains effectively constant. Instead of adding the 2 equivalents of NAD + with respect to UDP-glucose required in the normal reaction, only one fifth equivalent of NAD + was used in the presence of the same amount of FMN. The reaction was carried out in 50 mM pH 8.0 phosphate buffer with 50% H 2 1 8 0 and 2 mM DTT. When the reaction was initiated by the addition of enzyme, the color of the reaction solution changed from yellow to green. After 20 hours incubation at 37 °C, the color changed back to yellow and remained that color. The reaction appeared to go to approximately 90% completion as monitored by thin layer chromatography (TLC). A 1 3 C spectrum of the carboxyl group of UDP-glucuronic acid produced via the UDP-glucose dehydrogenase reaction in 50% l 8 0 water is shown in Figure 2.4. Two peaks (176.673 ppm and 176.647 ppm) of approximately the same intensity were observed indicating that a single oxygen atom has been incorporated from solvent. The two peaks are separated by 0.026 ppm, which is the same as the O induced upfield shift of the (1-C)acetate and close to the 0.02 ppm shift of L-(carboxy- C)histidine produced in the histidinol dehydrogenase reaction. In a control experiment, the reaction was carried out in l 6 0 1 ^ water and the C spectrum showed only one peak for the carboxyl carbon of the product. 51 •—|—i— i— i— i— i— i -177.0 176.5 ppm 177.0 ppm I i — i — ' 176.5 (A) (B) Figure 2.4 1 3 C spectra of the carboxyl groups of UDP-glucuronic acid (A) generated in 50 % H 2 1 8 0 and (B) generated in H 2 0 . The mass spectrometric method was attempted to confirm the 1 3 C N M R results. The product UDP-glucuronic acid generated from the enzymatic reaction in H 2 1 8 0 was partially purified from the reaction mixture by applying the diluted reaction solution to a Bio-gel P2 column and eluting with water. The purified UDP-glucuronic acid was first analyzed by liquid secondary ion mass spectrometry without success. Then the recently developed matrix assisted laser-desorption-ionization time-of-flight mass spectrometer (MALDI TOF MS) was used. This mass spectrometer consists of an ion source and a collector situated at opposite 52 ends of an evacuated tube. The technique of MALDI was introduced in 1988 as a method of transferring large, labile molecules into the gas phase as intact ions.123 In this technique the analyte of interest is mixed with a large molar excess of a matrix compound that is usually a weak organic acid and strongly absorbs laser light at a wavelength greater than 300 nm. The analytes can be peptides, proteins, carbohydrates, nucleic acids, synthetic polymers or other natural products with molecular weights of up to several hundred thousand daltons. The mixture solution is put on a probe and evaporated to dryness so that the analyte is incorporated into the crystals of matrix. After the probe is inserted into the mass spectrometer, ionization of analyte into the gas phase is achieved by applying a laser pulse to the sample mixture. The matrix molecules absorb the laser light, the thermal relaxation of excited matrix molecules leads to evaporation of the matrix and transfers the non-volatile peptides into the gas phase. Ions are accelerated and their flight times down the evacuated tube measured. This technique provides high ion yields of the intact analyte and tolerates sample impurities commonly encountered in samples of biological origin, such as salts and buffers. It was clear from our studies that MALDI TOF mass spectrometry is more sensitive than the liquid secondary ion mass spectrometry and can provide more accurate quantitative information in analyzing the compounds, especially phosphorylated sugars bearing isotopes. The results from the MALDI TOF mass spectrometric analysis also supported the notion that a single solvent oxygen atom was incorporated. In the MALDI TOF spectrum, a 18 50:50 mixture of unlabeled and singly O labeled UDP-glucuronic acid were observed (Figure 2.5). In Figure 2.5A, two peaks with m/z values of 648.3 ± 0.3 and 650.3 ± 0.3 and 18 having approximately the same intensity, correspond to the unlabeled and singly O labeled UDP-glucuronic acid with three sodium atoms bound (M-2H++3Na+). Another two peaks 53 with m/z values of 670.3 and 672.3 correspond to the unlabeled and singly 1 80 labeled UDP-glucuronic acid with four sodium atoms bound (M-3H++4Na+). The peak with a m/z of 656.4 corresponds to the starting material, UDP-glucose with four sodium atoms bound. The observation of singly labeled peaks, but not doubly labeled peaks, in the mass spectrum 18 indicates that only a single O oxygen was incorporated from the solvent into the product UDP-glucuronic acid purified from the enzymatic reaction in H2 1 8 0. Both the 1 3 C NMR and MALDI MS spectra demonstrated that only one solvent oxygen atom is incorporated into the product acid, suggesting that an imine intermediate is unlikely to participate in the reaction pathway. It is not proof, however, because a tightly bound water could be reintroduced during the hydrolysis step. The conserved lysine residue might be the general base responsible for deprotonation of the substrate alcohol or could participate in the first oxidation through electrostatic stabilization of the alkoxide form of the substrate alcohol. We propose that the mechanism of the UDP-glucose dehydrogenase reaction involves a bound aldehyde intermediate that is not released from the active site to any appreciable extent and is inaccessible to external aldehyde trapping reagents. The putative aldehyde intermediate has been independently synthesized by Robert Campbell in our laboratory and was demonstrated to be a kinetically competent substrate for the dehydrogenase reaction (kcat = \ .Q s'1, Km= 14 pM). 1 2 4 This fact supports the notion that aldehyde is a true intermediate in the reaction and is tightly bound or effectively sequestered in the active site of the enzyme. 54 (A) 6 4 8-3 6 30.3 6 5 6.4 6 7 0.3 6 7 2-3 u — 1 — 6 5 0 6 7 0 m/z (B) 6 4 8 . 5 i 6 7 0 . 5 i 6 5 6 . 4 —r 1 1 • i — 650 670 m/z Figure 2.5 MALDI TOF MS spectra of UDP-glucuronic acid (A) generated in 50% H218O/50% H 2 1 6 0 and (B) generated in H 2 1 6 0. 55 2.6 Synthesis of UDP-(6,6-di-2H)Glucose and Primary Kinetic Isotope Effect Studies As discussed in Chapter One, the rate-limiting step can be different for different dehydrogenases. The dissociation of NADH is the rate-limiting step for horse liver alcohol dehydrogenase and lactate dehydrogenase, whereas the hydride transfer is the rate-limiting step for yeast alcohol dehydrogenase.1 With glyceraldehyde-3-phosphate dehydrogenase, the rate-limiting step is the release of NADH at high pH and phosphorolysis of the acyl-enzyme at low pH.1 The deacylation step has been proposed to be rate-limiting for class 1 and class 2 aldehyde dehydrogenases.40'41 In the case of the bovine liver UDP-glucose dehydrogenase, the hydrolysis of the thioester intermediate has been postulated as the rate-limiting step.84 When a C-H bond is broken or formed during the rate-determining step, isotopic substitution of the proton for a deuterium usually results in an overall two to fifteen fold 125 126 slowing in the reaction rate. ' A simplified explanation for this effect is that the vibrational zero-point energy for a bond to deuterium is lower than that of a bond to hydrogen. Therefore a greater amount of energy is required for a C-D bond to reach the transition state in a bond cleaving reaction, as compared to a C-H bond, resulting in differences in the reaction rates. This effect is called a primary kinetic isotope effect and is often used to probe the rate-limiting step of a reaction. In this work, UDP-(6,6-di- H)glucose was synthesized and used to examine if either of the two hydride transfer steps is rate-limiting. If a hydride transfer step is rate-determining, the substitution of deuterium atoms for the protons at C-6 of UDP-glucose would result in a primary kinetic isotope effect on the reaction rate. 56 2.6.1 Synthesis of UDP-(6,6-di-2H)Glucose The synthesis of UDP-[6,6-di-3H]glucose was reported by Snetkova and coworkers. It consisted of the chemical synthesis of D-[6,6-di- HJglucose followed by the enzymatic coupling of UDP to the labeled glucose.127 In their synthesis D-[6,6-di-3H]glucose was obtained by the reduction of the methyl ester of P-methyl glucuronate with gaseous tritium in the presence of nickel boride catalyst followed by removal of the methyl protecting group using H2SO4 (Figure 2.6). O H O H O H Figure 2.6 Snetkova's synthetic route for D-[6,6-di-3H] glucose. Moss reported a simpler synthetic route to D-[6,6-di-3H]-glucose128 that could be easily adopted for the synthesis of D-(6,6-di-2H)glucose. Moss synthesized D-[6,6-di-HJglucose by the reduction of l,2-0-isopropylidene-D-glucofuranurono-6,3-lactone (monoacetone-D-glucuronolactone) with NaB 3H4 followed by the hydrolysis of the isopropylidene group (Figure 2.7). The starting material is prepared from commercially available D-glucuronolactone, 5 7 T Figure 2.7 Moss's synthetic route for D-[6,6-di-3H] glucose. The enzymatic coupling of UDP by Snetkova and coworkers was carried out by the successive action of the enzymes (Figure 2.8), hexokinase, phosphoglucomutase and UDP-glucose pyrophosphorylase without losing the tritium label at the 6 position. OP032- OUDP Figure 2.8 Enzymatic coupling of UDP to glucose. First the C-6 of the glucose is phosphorylated by the hexokinase and ATP. Then phosphoglucomutase transfers the phosphate group from C-6 to C - l of the glucose and generates the a-anomer. Glucose- 1,6-diphosphate is added to keep the enzyme 58 phosphoglucomutase phosphorylated (active). UMP is coupled to the resultant glucose-1-phosphate to generate the product UDP-glucose by treatment with UDP-glucose pyrophosphorylase and a 10-fold excess of UTP. There is also a chemical method to synthesize the nucleoside diphosphates using phosphoromorpholidates.129 However, the enzymatic procedure is much simpler. We adapted the Moss synthetic route for D-(6,6-di-2H)glucose and Snetkova's enzymatic UDP coupling. The overall synthetic route of deuterated UDP-glucose is illustrated in Figure 2.9. 5 4 * Enzymes involved: hexokinase, phosphoglucomutase UDP-glucose pyrophosphorylase inorganic pyrophosphatase Figure 2.9 Synthetic route for UDP-(6,6-di-2H)glucose. In this synthesis, the two free hydroxyl groups of D-glucuronolactone 1 were protected by treating 1 with acetone and H 2 S0 4 to give compound 2. The protected lactone 2 was 59 reduced by sodium borodeuteride (NaB2H4) to generate 3. Deprotection of 3 by trifluoroacetic acid (TFA) yielded the deuterated glucose 4. The enzymatic coupling of UDP was carried out in a single incubation analogous to Snetkova's paper. Inorganic pyrophosphatase hydrolyzes the liberated pyrophosphate (P2) and drives the reaction in the direction of UDP-glucose. Certain modifications in the amounts of enzymes and chemicals were made. Less glucose-1,6-diphosphate was added (1/200 the amount of deuterated glucose) to initiate the reaction catalyzed by phosphoglucomutase. This is to ensure that only a small amount of undeuterated UDP-glucose would be generated. The purification procedure has been revised to a simpler one as well. Deuterated UDP-glucose was purified by applying it to a diethylaminoethylcellulose (DE52) ion exchange column and eluting with a linear I gradient of 0-400 mM triethylamine bicarbonate. Fractions containing the deuterated UDP-glucose were further purified by passage through a column of Amberlite IR-120 (plus) resin, followed by desalting on a Bio-gel P-2 column with elution by water. Both 'H NMR spectroscopic (Figure 2.10) and Liquid Secondary Ion mass spectrometric (LSIMS) analyses indicated that the substrate was greater than 95% enriched with deuterium at C-6 position of glucose. 60 Figure 2.10 'H NMR spectra of (A) deuterated and (B) undeuterated UDP-glucose. In spectrum B, signals due to H6" overlap with those of H3" (triplet, 3.77 ppm) and H5" (doublet, 3.89 ppm). (Arrows point to the difference between the two spectra) 61 2.6.2 Studies of the Primary Kinetic Isotope Effect Using UDP-(6,6-di-2H)Glucose The rates of the UDP-glucose dehydrogenase reaction for both the labeled and unlabeled substrates were measured under saturating conditions. Substrate concentrations were calculated from the absorption at 260 nm using an extinction coefficient (e) of 8700 M"'cm"'.124 Under these conditions, the ku/kD was determined to be 1.1 ±0 .1 . The error is the standard deviation of the data from five sets of experiments. It appears that there is no primary kinetic isotope effect and neither of the two hydride transfer steps is rate-limiting. This observation is consistent with the previous speculation based on the comparison of rate equations. We propose that a good candidate for the rate-determining step is the hydrolysis of the thioester intermediate (studies in Chapter Three strongly support the existence of this intermediate). 2.7 Summary The recombinant Streptococcus pyogenes UDP-glucose dehydrogenase was over-expressed in E. coli and purified to apparent homogeneity by a two-step procedure. The kinetic analysis of this enzyme gave an apparent Km of 20 + 4 pM for UDP-glucose and a kcat of 1.8 ± 0.1 s"1 (in 50 mM Trien-HCl buffer, pH 8.7, 2 mM DTT, 0.5 mM NAD + , 30 °C). Both the direct and coupled continuous assay methods are valid for the enzyme kinetic studies. The solvent oxygen incorporation studies showed that only one solvent-derived oxygen atom is incorporated into the product UDP-glucuronic acid generated from the enzymatic reaction in H2180. This observation indicates that the imine is unlikely to be an 62 intermediate in the dehydrogenase mechanism. We propose that a bound aldehyde intermediate is involved. We found that the isotope-induced shift caused by 1 80 on the 1 3 C NMR chemical shift of directly bonded carbon atoms is a direct and very convenient method to investigate the solvent isotope incorporation. In addition, MALDI TOF mass spectroscopy is a very sensitive technique to analyze enzymatic reaction products without need for extensive purification. By comparing the dehydrogenase reactions with protonated and deuterated UDP-glucose, kulkv was determined to be 1.1 ± 0.1 indicating that there is no primary kinetic isotope effect in either of the two hydride transfer steps. Therefore, the hydride transfers steps are not rate-limiting. A likely candidate would be the hydrolysis of the thioester intermediate or possibly binding/release events. 2.8 Experimental Methods 2.8.1 General UDP-glucose and NAD + were obtained from Sigma. All other buffers and chemicals were obtained from Sigma, Boehringer Mannheim Biochemicals, Aldrich, or Fisher Scientific, unless otherwise stated. All protein manipulations were performed at 4 °C, unless otherwise noted. 2.8.2 Protein Determination 1 3 0 Protein concentration was determined by the method of Bradford , using bovine serum albumin (BSA) as the standard. Both the protein assay solution and the standard protein were purchased from Bio-Rad. 63 2.8.3 Strains, Plasmid and Media The pGAC147 plasmid that contains the cloned UDP-glucose dehydrogenase gene from Streptococcus pyogenes (hasB gene) was kindly provided by Dr. Ivo van de Rijn of the Wake Forest University Medical Center, Winston-Salem, North Carolina, USA. E. coli JM109 (DE3) was used as the host organism for the overexpression of the recombinant UDP-glucose dehydrogenase. Cells were grown in TYPG media that contains 8 g of tryptone, 8 g of yeast extract, 2.5 g of NaCl, 1.25 g of K2HPO4, and 2.5 g of glucose per 500 mL of distilled water. Tryptone and yeast extract were from Difco. Short Protocols in Molecular Biology by Ausubel et al. was used for the general microbiology techniques. 1 3 1 Plasmid DNA was prepared using the Promega WIZARD Mini Plasmid Preps kit. 2.8.4 Purification of Recombinant UDP-Glucose Dehydrogenase The pGAC147 plasmid was transformed into CaC -^competent E. coli JM109 (DE3) cells and inoculated onto LB-agar plates containing 25 pg/mL chloramphenicol. After overnight incubation at 37 °C, a single colony was used to inoculate 500 mL of TYPG media containing 25 pg/mL chloramphenicol. The cell culture was grown at 37 °C with vigorous shaking (280 rpm) until the culture reached an optical density (measured at 600 nm) of 0.7-1.0. Isopropyl-l-thio-p-D-galactopyranoside (IPTG) (48 mg) was added at this point to a final concentration of 0.4 mM to induce the overexpression of UDP-glucose dehydrogenase. After three hours of further growth, the cells were harvested by centrifugation for 15 min at 5000 rpm (Sorvall GSA rotor) and the cell pellet (about 2 g - 2.4 g) was stored at -78 °C. 64 The cell pellet was later thawed rapidly with warm water and resuspended in 5 mL cold 50 mM triethanolamine-HCl (Trien-HCl) buffer, pH 8.7, containing 2 mM DTT, 10% (v/v) glycerol, 1.5 mM phenylmethanesulfonyl fluoride (PMSF), 1 mg/L pepstatin, and 1 mg/L aprotinin (buffer A). The stock solutions of PMSF, pepstatin (5 mg/mL) and aprotinin (5 mg/mL) were prepared in ethanol. The resuspended cells were lysed by two passes through a chilled French pressure mini cell (SLM Aminco) at 10,000 psi. Following ultracentrifugation at 30,000 rpm for 45 minutes to pellet the cell debris, the resulting supernatant (lysate) was diluted with an equal amount of buffer A. The diluted cell lysate was loaded onto a 15 mL column of diethylamino-ethylcellulose (DE52, Whatman) that had been preequilibrated with buffer A (in the cold room, 5 °C). The column was washed with buffer A (50 mL) and then eluted successively with buffer A containing 0.1 M NaCI (20 mL), 0.15 M NaCI (20 mL) and 0.2 M NaCI (20 mL). The fractions were collected when there was absorbance at 280 nm monitored by an UV detector (Spectra/Chrom Flow Thru UV monitor controller, Spectrum) connected to the column. The eluent fractions were assayed for dehydrogenase activity. The active dehydrogenase fractions were pooled and concentrated using a Millipore Ultrafree centrifugal concentrator (10 kDa nominal molecular mass cut-off, Millipore) and dialyzed against 1 liter of buffer B (50 mM Trien-HCl buffer, pH 8.7, 2 mM DTT, 10% glycerol) for 12 hours. The resulting solution was passed through a 0.22 pm filter (Sterile Acrodisc Syringe Filters), frozen in liquid nitrogen, and stored at -78 °C. The DE-52 purified protein solution was quickly thawed in warm water and applied (about 25 mg of total protein in each injection) to a Waters AP-1 Protein-Pak Q column (10 x 100 mm) that had been preequilibrated with filtered and degassed buffer B (at 22 °C). The 65 HPLC system used consisted of a Waters 625 LC system, monitored at 280 nm using a Waters 486 tunable absorbance detector. The column was washed with buffer B (10 mL) and eluted with a linear gradient of NaCI (0 - 0.2 M, over 40 minutes at 1 mL/min) in buffer B. The eluent fractions were assayed and the active fractions were pooled, frozen in liquid nitrogen, and stored at -78 °C. Unless noted otherwise, all manipulations were performed at 4°C. 2.8.5 Purity Assessment of UDP-glucose Dehydrogenase The purity of the dehydrogenase was judged by 12% SDS-PAGE, using a Bio-Rad Mini PROTEAN II electrophoresis system. Protein bands were visualized by staining with Coomassie Brilliant Blue. The dehydrogenase band was compared to (3-lactoglobulin (19 kDa), carbonic anhydrase (29 kDa) and BSA (66 kDa) as molecular weight standards. 2.8.6 Molecular Weight Determination of UDP-glucose Dehydrogenase The molecular weight of UDP-glucose dehydrogenase was determined on an electrospray mass spectrometer, performed by Shouming He in the lab of Dr. Stephen Withers, Department of Chemistry, UBC. Mass spectra were recorded on a Perkin-Elmer Sciex API 300 triple quadrupole mass spectrometer equipped with an electrospray ion source 132 • and interfaced with a Michrom HPLC. Protein samples were injected onto a microbore PLRP reversed phase column (1 x 50 mm) on the Michrom HPLC and eluted with a linear gradient of 0 - 80% acetonitrile in water (containing 0.05% trifluoroacetic acid) over 6 minutes with a flow rate of 50 pL/min. The eluent was introduced into the MS spectrometer that was operated in a single quadrupole mode. The quadrupole mass analyser was scanned 66 over a m/z range of 400 to 3000 Da with a step size of 0.5 Da and a dwell time of 1 ms per step. Protein molecular weights were determined from the mass spectroscopic data using deconvolution software Multiview 1.1 supplied by SCIEX. 2.8.7 Kinetic Assay of UDP-glucose Dehydrogenase 2.8.7.1 Direct Assay Method UDP-glucose dehydrogenase activity was assayed by measuring the rate of NADH production at 340 nm using a Varian Cary 3E UV-visible spectrophotometer equipped with a circulating water bath. The extinction coefficient of NADH was taken as 6220 M"1 cm"1. All assays were performed at 30 °C in 50 mM Trien-HCl buffer, pH 8.7, containing 2 mM DTT, 0.5 mM UDP-glucose and 0.5 mM NAD + (1 mL total volume). Steady-state kinetic constants were obtained by addition of UDP-glucose dehydrogenase (0.01 mg/mL) to a solution of various concentrations of UDP-glucose and NAD + . The reaction was typically followed for approximately 5 min or until 10 % of the starting material was consumed. Rates were determined by following the increase in absorbance at 340 nm during the first 40 seconds after the initiation with the enzyme, at 6 substrate concentrations ranging from 2.5 pM to 150 pM UDP-glucose with 500 pM NAD + or 10 pM to 500 pM NAD + with 163 pM UDP-glucose. The slopes were calculated using a least squares analysis with Cary 3 software version 3.0. The resulting calculated initial velocities were plotted as a function of substrate concentration, and the kinetic parameters were determined by a direct fit of the data to an enzyme kinetic expresion using the computer program Grafit version 3.0 (Erithacus Software, 67 1994). The reported error values are based on the deviation of the data from the calculated curve-of-best-fit. 2.8.7.2 Coupled Assay Method In the continuous coupled assay, diaphorase was employed as the auxiliary enzyme to couple the formation of NADH to the reduction of ;?-iodonitrotetrazolium violet (INT). The initial rates were measured at 30 °C by following the increase in absorbance at 500 nm using a Varian Cary 3E UV-visible spectrophotometer. The extinction coefficient of the reduced INT was taken as 12 800 M ' W 1 . The assay mixtures in 50 mM Trien-HCl buffer, pH 8.7, contained 0.65 mM INT, 0.043 mg (1.6 units) of diaphorase, 0.01 mg/mL UDP-glucose dehydrogenase, 0.5 mM of NAD + with various concentrations of UDP-glucose (5.1 pM to 122.9 pM) or 0.2 mM of UDP-glucose with variable concentrations of NAD + (7.8 pM to 313.5 pM). The total volume of the assay solution was 800 pL. The reaction was initiated by the addition of UDP-glucose dehydrogenase. The slopes were calculated using a least squares analysis with Cary 3 software version 3.0. The kinetic constants were determined using an identical method to that used in the direct assay. The error reported with the data is indicative of the deviation of the data from the curve-of-best-fit. The extinction coefficient of the reduced INT (1.28 ± 0.01 x 104 M^cm"1) was obtained by an enzymatic assay that involved incubation of 0.65 mM INT and 1.6 units diaphorase with various concentrations of NADH (from 7 pM to 57 pM) in 50 mM Trien-HCl buffer, pH 8.7. The reaction was followed to completion by monitoring at 500 nm. The concentrations of NADH and the corresponding final absorbance at 500 nm after the reaction was completed were used to calculate the extinction coefficient of reduced INT. 68 2.8.8 Solvent-Derived Oxygen Incorporation The UDP-glucose dehydrogenase reaction was performed in a capped 1.5 mL microcentrifuge tube with 50 mM phosphate buffer, pH 8.0, 25 pmol UDP-glucose, 5 pmol NAD , 5 pmol FMN, 2 mM DTT, 0.35 mg of UDP-glucose dehydrogenase, in a final volume 18 of 500 pL containing 50% H2 O (Cambridge Isotope Laboratories). Soon after the reaction was initiated with the enzyme, the color of the solution changed from yellow to green. The mixture was incubated at 37 °C for 20 hours with the lid of the tube being opened periodically. The final color was orange. The reaction was monitored by thin-layer chromatography (TLC). PEI cellulose F TLC plates were used with a 0.3 M LiCl aqueous solution as the developing reagent. After the incubation, 8 pmol of EDTA was added, and the 13 reaction mixture was lyophilized, redissolved in 0.5 mL D2O, and C NMR spectra were recorded. The control experiment was carried out in 100%H216O. Further purification of the product UDP-glucuronic acid was carried out by applying the diluted reaction solution onto a 40 mL Bio-gel P2 column and eluting with water. The purification was monitored by UV detector (Spectra/Chrom Flow Thru UV monitor controller, Spectrum) with a 254 nm filter. The resulting eluent was lyophilized and submitted for MS analysis. 1 3 C NMR spectra were collected on a Varian XL-300 at 75 MHz with proton decoupling. MALDI TOF MS spectra were obtained from a Bruker Biflex MALDI time-of flight mass spectrometer operated in the reflex mode. 69 2.8.9 Synthesis of UDP-(6,6-di-2H)Glucose The compound UDP-(6,6-di-2H)glucose was synthesized according to the procedure of Moss and Snetkova , with some modifications to reaction times, temperatures, and methods of purification as described below. Thin layer chromatography was performed on aluminum-backed sheets of silica gel 60F254 (Merck) of thickness 0.2 mm. Compounds were visualized by UV or spraying with a solution containing H2SO4 (31 mL), ammonium molybdate (21 g), and Ce(S04)2 (1 g) in water (500 mL) with 3 minutes heating at 110 °C. Flash column chromatography was performed using Silica Gel 60 (230 - 400 mesh, E. Merck, Darmastadt). 'H NMR spectra were recorded on Bruker instruments at the indicated field strengths: an AC-200 at 200 MHz, a WH-400 at 400 MHz. Liquid secondary ion mass spectra (LSIMS) were obtained on a Kratos Concept II HQ Mass Spectrometer. (a) 1,2-O-Isopropylidene-D-glucuronolactone D-Glucuronolactone (5 g) was dissolved in 125 mL acetone containing 4 mL of concentrated H2SC>4. The resulting yellow solution was stirred at room temperature for five hours until no more starting material was present as judged by TLC (TLC, ethyl acetate/methanol, 9.5:0.5). Na2C03 (15 g) was added and the solution was refluxed for four hours. The white precipitate was filtered off and the solvent was removed by rotary evaporation. The resulting syrup was redissolved in 70 mL ethyl acetate, washed with saturated NaHC0 3 and saturated NaCl (three times), and the organic phase was removed. The aqueous phase was extracted with ethyl acetate three times and the organic portions were 70 combined and washed with saturated NaCI. All the organic fractions were pooled, dried using MgSCU and rotary evaporated to dryness to give a yellow solid (1.4 g, 23% yield). (b) D-6,6-Di-2H-glucose 1,2-O-Isopropylidene-D-glucuronolactone (1.14 g, 5.32 mmol) and NaB 2H 4 (229 mg, 5.47 mmol, 1.03 eq.) were added to 120 mL ethanol. The reaction mixture was stirred at room temperature for 5 hours. After 2 hours, TLC (ethyl acetate/methanol, 9:1) showed a new product and no trace of starting material. The reaction was quenched by addition of NH4CI and the solution was stirred until the evolution of gas ceased. Then the solution was filtered through Celite, eluted with ethanol, and the solvent was removed using rotary evaporation. The residue was purified by flash column chromatography (ethyl acetate/methanol 9:1) to afford compound 3(1.151 g, 5.18mmol, 97% yield). Compound 3 (0.235g, 1.06 mmol) was dissolved in 25 mL water and trifluoroacetic acid (25 mL) was added to the solution. The reaction mixture was stirred at room temperature until TLC (chloroform/methanol 8.5:1.5) showed a new product and no starting material. The solvent was removed by rotary evaporation to give a brown syrup. Water (10 mL) was added, and evaporated and this was repeated four times until the color of the syrup was light. The resulting syrup was pumped under vacuum overnight to give an off-white solid, D-(6,6-di- H) glucose (0.186 g, 1.02 mmol, 96% yield). The 'H NMR spectrum was identical to that for the unlabeled glucose (mixture of anomers) with the exception of peaks in the area of 3.6-3.9.133 The LR LSIMS spectrum showed a peak with a m/z of 182 (m/z for unlabeled glucose = 180 Da). 71 (c) UDP-(6,6-di-2H)Glucose The enzymatic coupling of UDP was carried out in an incubation mixture of the following composition: D-(6,6-di- H) glucose (8 mg, 44 pmol, l . lmM) , ATP (72.7 mg, 132 pmol), UTP (64 mg, 132 pmol), glucose-1,6-diphosphate (0.165 pmol), MgS04 (36 mg, 146 pmol), 70 m M Tris-HCl, pH 7.8, (40 mL), hexokinase (66 units), phosphoglucomutase (109 units), UDP-glucose pyrophosphorylase (12.5 units), and inorganic pyrophosphatase (33 units). The incubation was carried out at 30 °C for 20 hours. The reaction mixture was loaded onto a 65 mL DE52 column and eluted with a linear gradient of 0 - 400 m M triethylamine bicarbonate buffer (800 mL total). The triethylamine bicarbonate buffer was prepared by bubbling CO2 into 400 m M triethylamine solution over 10 hours. The separation was monitored by an U V detector (Spectra/Chrom Flow Thru U V monitor controller, Spectrum) with a 254 nm filter. The eluent fractions were assayed for UDP-glucose by incubating with N A D + and wild type UDP-glucose dehydrogenase. The fractions containing deuterated UDP-glucose were pooled and evaporated to dryness under reduced pressure. The product was redissolved in water, lyophilized, resuspended in water, and lyophilized again. Then the product was passed through a column of Amberlite IR-120 (plus) resin (10 mL, N a + form, eluted with water), followed by desalting on a Bio-gel P-2 column (2.5 x 45 cm, eluted with water). The product is a white solid (2.65 pmol, yield 6%). ! H N M R (400 MHz, D 2 0) : 5 = 3.46 (t, J = 10.0 Hz, IH, H4"), 3.53 (m, IH, H2"), 3.77 (t, J= 9.44 Hz, IH, H3"), 3.89 (d, J = 10.4 Hz, IH, H5"), 4.20 - 4.28 (m, 4H, H3' , H4', H5'), 4.37 (d, J = 3.4 Hz, IH, H2'), 5.6 (dd, J(H,H) = 3.5 Hz, J(H,P) = 7.2 Hz, IH, H I " ) , 5.98 (m, 2H, H I ' , H5), 7.94 (d, J= 8.1 Hz, IH, H6). A l l chemical shifts were reported using the 8 scale in ppm. This spectrum was identical 72 to that of unlabeled UDP-glucose with the exception of the peaks at 3.6 - 3.95 ppm. LR-LSI MS: m/z = 613 (M-H++2Na+). NMR spectroscopic and LSIMS mass spectroscopic analyses indicated that the extent of deuterium incorporation was > 95%. 2.8.10 Primary Isotope Effect Experiment The UDP-glucose dehydrogenase reaction was carried out under saturating conditions. The reaction was monitored by the absorbance at 340 nm. The rates of oxidation were determined five times for the deuterated and the nondeuterated substrates. The assay mixtures (500 pi) contained 50 mM Trien-HCl buffer, pH 8.7, 2 mM DTT, 1.25 mg UDP-glucose dehydrogenase, 0.5 mM or 0.25 mM UDP-glucose (either deuterated or nondeuterated sample), and either 1.2 mM or 0.6 mM NAD + . All of these components, except UDP-glucose, were mixed and divided into two cuvettes. Reactions were initiated by the addition of either deuterated or nondeuterated UDP-glucose. The initial velocities were calculated from the observed rate of NADH formation as described in Section 2.8.7. The error reported for the kinetic isotope effect is the standard deviation of the data points from the average determined rates. Chapter Three Mechanistic Studies on the Cys260Ser Mutant 74 3.1 Introduction Cysteine residues are important for the activity of bovine liver UDP-glucose dehydrogenase, as shown by chemical modification studies that employ reagents such as 5,5'-dithiobis-(2-nitrobenzoate)/KCN76 and iodoacetate78 among others. The disadvantages of using the method of chemical modification to study catalytically essential amino acid residues are that modifying reagents may not be specific for the residues of interest and the reagent may be bulky or bear charges that interfere with the substrate binding (accessibility to the active site). Therefore chemical modification studies alone can sometimes provide misleading information about the involvement of the amino acid residues in the enzyme catalysis. This was the case with histidinol dehydrogenase (refer to Section 1.5.2), an enzyme that was thought to employ a similar mechanism to that of UDP-glucose dehydrogenase. The technique of site directed mutagenesis, in which a specific amino acid residue in a protein sequence can be changed to another amino acid, makes more controlled and subtler changes to enzyme structure than chemical modification. In the case of histidinol dehydrogenase, site directed mutagenesis experiments102'103 gave results differing from those obtained in chemical modification studies95. Mutation of the conserved cysteine residues showed no effect on the dehydrogenase activity, therefore cysteine residues are not catalytically involved in the mechanism of the dehydrogenase reaction. This is in contrast to the results from chemical modification studies which led to the assumption that an active site cysteine residue was important. Chemical modification studies also led to the wrong assignment of the catalytically important cysteine residues in aldehyde dehydrogenase.134'135 When the assigned cysteine residues were replaced by alanine, the mutant enzymes were fully active. It was proposed that 75 the introduction of a large and relatively hydrophobic N-ethylsuccinimidyl group of N-ethylmaleimide (the reagent used to modify thiol groups) and not the loss of the thiol group caused the observed inactivation. The substitution of these cysteines by tryptophan rather 136 than alanine residues produces inactive enzymes. There are two cysteine residues, Cys260 and Cysl62, in the UDP-glucose dehydrogenase from Streptococcus pyogenes. Amino acid sequence comparison of Streptococcus pyogenes UDP-glucose dehydrogenase to similar sequences from other species revealed that Cys260 is conserved in all the sequences known to date. Thus, Cys260 is an excellent candidate for the active site nucleophile in the dehydrogenase. By mutating Cys260 to serine or alanine and studying the mutant dehydrogenase activity, one might gain important evidence for the catalytic role of this residue. The serine and alanine residues are smaller in size than the cysteine residue and might not interfere with the substrate binding. The mutants should retain the active site structure, and the overall folding of the protein should be relatively unaffected. The serine residue could serve as a nucleophile, whereas an alanine residue has no nucleophilic functional group. The hydroxyl of the serine residue might be able to replace the thiol group of cysteine to attack the carbonyl carbon of the aldehyde intermediate. Studies on the serine mutant of UDP-glucose dehydrogenase will be described in this chapter. The properties of the alanine mutant will be discussed in the next chapter. 3.2 Purification of the UDP-glucose Dehydrogenase Cys260Ser Mutant The plasmid for the Cys260Ser mutant was kindly provided by Dr. Ivo van de Rijn of the Wake Forest University Medical Center, North Carolina, USA. The plasmid was produced using an adaptation of enzymatic PCR 1 3 7 using plasmid pGAC147. The mutation 76 was confirmed by sequencing the entire gene. The plasmid encoding the Cys260Ser mutant was transformed and overexpressed in E. coli JM 109. The mutant protein was purified using an analogous procedure to that employed for the wild type enzyme, which was described in Chapter Two. The fractions that contained the mutant protein were identified by SDS-PAGE analysis instead of the activity assay used in the purification of the wild type enzyme. A slightly lower pH buffer (pH 8.5 - 8.6 instead of 8.7) was needed for HPLC purification in order to separate the mutant enzyme from a lower molecular weight (ca. 44 kDa) protein. Initial problems were encountered in the purification that seemed to indicate that the Cys260Ser mutant was less stable than the wild type enzyme. To keep the mutant protein active, the anion-exchange HPLC column and the solvents were kept in ice during the purification or the whole HPLC system was moved to the cold room. The gradient for HPLC was also modified from 0 - 0.25 M NaCl in triethanolamine-HCl buffer over 30 minutes to 0 - 0.05 M NaCl over 5 minutes, 0.05 - 0.1 M NaCl over 30 minutes and 0.1 - 0.15 M NaCl over 10 minutes. The protein samples were not frozen from the point of cell lysis until just prior to HPLC injection. Only the cell pellets, partially purified proteins ready to be injected to HPLC, and pure protein were frozen in liquid nitrogen and stored at -78 °C. The purity of the mutant protein was judged by SDS-PAGE analysis. After the gel was stained by Coomassie Brilliant Blue, only one protein band with a molecular weight about 46 kDa was observed. The purification gave about 30-38 mg of protein per 500 mL culture. The molecular weight of the mutant enzyme was determined to be 45 473 + 6 Da by electrospray mass spectrometry, consistent with the expected 45 468 Da from the gene sequence with the initiator Met. 77 3.3 Mutant Enzyme Activity Analysis The purified Cys260Ser mutant enzyme showed less than 0.1% of the wild type enzyme activity when assayed under saturating conditions with UDP-glucose and NAD + . The reaction catalyzed by the Cys260Ser mutant was about 3300 fold slower than that catalyzed by the wild type enzyme as judged by maximum velocity. This is consistent with the notion that cysteine 260 is important for catalysis. In a later investigation, two-phase kinetics were observed as will be discussed in further detail in section 3.5. The importance of dithiothreitol (DTT) was investigated by carrying out the reaction in an assay buffer without DTT (50 mM triethanolamine-HCl buffer, pH 8.7 containing 3 mM NAD + and 3 mM UDP-glucose). The Cys260Ser mutant enzyme was dialyzed overnight at 5 °C against 50 mM triethanolamine-HCl buffer (pH 8.7) without DTT before adding to the assay buffer. The amount of NADH produced in this reaction was similar to that produced in a buffer with DTT. This is in contrast to the wild type enzyme in which DTT is crucial for the enzyme activity. However, this is not unexpected since there is no cysteine residue in the active site and no thiol reducing reagent is necessary for the activity. UDP-g/wco-hexodialdose (the putative aldehyde intermediate, UDP-Glc-6-CHO) could also be oxidized by Cys260Ser mutant under saturating conditions (3 mM NAD + , ImM UDP-Glc-6-CHO), and the maximal rate was similarly reduced to be less than 0.1 % of rate for the wild type enzyme. 78 3.4 Evidence for the Involvement of the Serine Residue in the Mechan ism 3.4.1 Discovery of the Adduc t The Cys260Ser mutant and wild type UDP-glucose dehydrogenase were separately incubated at 30 °C for 1 hour in the presence of 3 mM NAD + and 1 mM of either UDP-glucose or UDP-Glc-6-CHO in 50 mM triethanolamine-HCl buffer, pH 8.7. The resulting proteins were analyzed directly using an electrospray ionization mass spectrometer (ESI MS) interfaced with a reversed-phase HPLC system under denaturing conditions (0.05 % trifluoroacetic acid in 25 % acetonitrile in water as eluent). In the case of the wild type enzyme, only the expected mass of the protein was observed (45 489 ± 6 Da). In the case of the Cys260Ser mutant however, 90% of the protein had a mass of 46 034 ± 6 (Figure 3.1). Thus the incubation of the mutant enzyme with the substrate and coenzyme resulted in the formation of a covalent adduct with a mass greater by 562 Da than that of the free enzyme. The adduct survived 45 minutes of denaturation treatment with 6 M urea (as analyzed by ESI MS), further indicating that the adduct was covalently attached to the mutant protein. The mass of UDP-glucuronic acid is 580, therefore the mass of the resulting protein was increased by an amount corresponding to the covalent attachment of one molecule of UDP-glucuronic acid, less one molecule of water. Dialysis of the mutant-adduct for 24 hours against pH 8.7, substrate-free, triethanolamine-HCl buffer at 5 °C, resulted in a loss of less than 10% of the adduct (as analyzed by ESI MS), indicating that the attachment was kinetically stable in the active site of the folded protein at this temperature. 79 (A) 1.185 -1.0e5 - | 9.064 -8.004 -7.0e4 -6.0e4 -5.0e4 -4.0e4 -3.0e4 -2.064 -1.094 -41177.0 ^ 428B6.0 g | j 47137.0 48000 49227.0 (B) (C) 90000 -aoooo -70000 -40000 -30000 -20000 -10000 -1.0«5 -9.0«4 -8.004 -7.0a4 -6.0e4 -5,0e4 -4.0»4 -3.084 -2.0e4 -1.084 -43058 45474 44075.0 I o j, 47843.0 489G7.0 P -• ^ ' 43093.0 " U , ' J U 1 ( |l. .m.-.,., vL &„ i , M Figure 3.1 Electrospray mass spectra of (A) the Cys260Ser mutant, (B) the Cys260Ser mutant after incubation with UDP-glucose, and (C) the Cys260Ser mutant after incubation with UDP-Glc-6-CHO. 80 We propose that the adduct was formed by the addition of UDP-glucuronic acid to the serine residue of Cys260Ser mutant via an ester linkage (Figure 3.2A). It is also possible that the adduct was a hemiacetal (Figure 3.2B) formed by the addition of the aldehyde intermediate to the serine residue. However, it is very unlikely that this hemiacetal would survive the denaturing conditions employed in the ESI MS analysis. Enzyme OH I O OUDP (A) Enzyme O H O ^ ^ ^ ^ ° v H O ^ ^ ^ A OH I OUDP ( B ) Figure 3.2 Proposed structures of adduct: (A) ester intermediate, (B) hemiacetal intermediate. 3.4.2 Identification of the Peptide Containing the Adduct The traditional strategy employed for the identification of the labeled residue includes the synthesis and use of a radiolabeled inactivator to generate radiolabeled enzyme, followed by proteolytic digestion of the radiolabeled protein. Subsequent HPLC separation of the resulting peptide mixture is guided by the presence of radioisotope and peptide sequencing of the purified labeled peptide completes the identification. This approach is time consuming and technically demanding. In addition, complex syntheses are frequently required for the introduction of a radioisotope into the inactivator. 81 To circumvent the need for radiolabels in the identification and isolation of the labeled peptides, an electrospray tandem mass spectroscopy (MS/MS) based technique has previously been developed for the identification of active site residues serving as 138 139 nucleophiles in glycosidases. ' Nonradioactive mechanism-based inhibitors, 2',4'-dinitrophenyl 2-deoxy-2-fluoro glycosides, were used to covalently modify active site residues in the glycosidases. The glycosidase was covalently labeled via the formation of a glycosyl-enzyme intermediate (Figure 3.3).140,141 After the enzyme was proteolytically digested, the resulting peptide mixture was subjected to reversed-phase HPLC that was directly interfaced to an electrospray mass spectrometer (LC/MS) 1 4 2. The mass spectrometer could then be used as a detector, either to measure the total ion count or to measure at a specific mass as a function of retention time. N 0 2 O, ,0 Y Glycosidase Glycosidase Figure 3.3 Inactivation of glycosidases by 2-deoxy-2-fluoro glycosides. The sugar moiety can vary for different glycosidases. In LC/MS there are three methods used to select the labeled peptide in the complex peptide digestion mixture. These methods can be used by themselves or in combination. First, 82 if the labeled peptide elutes at a different retention time on the reversed-phase HPLC column as compared to its unlabeled counterpart, this could be used to screen the MS data for the disappearance of a specific peptide ion upon the inhibition. Second, the labeled peptide has a characteristic mass shift due to the addition of the mass of the label: This could be used to screen the MS data for a new peptide ion species. Third, if the covalent linkage between the inhibitor and the peptide is more labile than other bonds such as the amide bond, the labeled peptide can be identified by using the tandem mass spectrometer set up in the neutral loss mode. In this technique the peptide ions are subjected to limited fragmentation by collision with an inert gas (argon or nitrogen) in a collision cell. In a triple quadrupole mass spectrometer, the collision cell is the second quadrupole that is located between the two mass analyzers, the first and third quadrupoles. If the most labile linkage is cleaved homolytically in the collision cell and a neutral species of a pre-determined mass is lost, the peptide will retain its original charge. The first and third quadrupoles are scanned coordinately, but offset by the mass of the lost neutral species. The linked first and third quadrupoles are used as a detector, and only peptide ions that lose a predetermined mass upon collision can pass through both quadrupoles and be detected. This is a rapid, sensitive, nonisotopic, and conclusive alternative to the standard radioactive method. To investigate whether the Ser260 is the attachment site of the adduct, the labeled mutant (using UDP-glucose as the substrate) was subjected to pepsin digestion, and the resulting peptides were separated by reversed-phase HPLC using the ESI/MS as a detector. Pepsin is known to cut proteins preferentially on the C terminal side of phenylalanine, leucine, tryptophan and tyrosine. Other residues may be cleaved, but with variable rates. Trypsin also cleaves the protein at specific positions (arginine and lysine) and generates 83 predictable fragments; it was not used because the tryptic digestion requires basic conditions in which the adduct may not be stable (the ester bond could be hydrolyzed) and the peptides are often too big to be sequenced by tandem mass spectroscopy. When using LC/MS to separate peptide mixtures, a flow-splitting device is placed between the column and the electrospray ionization ion source so that part of the flow is directed into the ion source of the mass spectrometer and part of the flow is directed to a fraction collector.143 The ratio of flows was set up to allow 90% of the sample to be collected for further analysis and 10% of the sample was directed into the mass spectrometer. When the mass spectrometer was scanned in the normal LC/MS mode, the total ion chromatogram (TIC) displayed a large number of peaks, arising from every peptide in the digestion (Figure 3.4A). The peptide bearing the adduct was identified in the second experiment on the same sample using the tandem mass spectrometer in neutral loss mode. The spectrometer was first set to scan for the mass loss of m/z 563, which corresponds to the mass of the UDP-sugar fragment lost by the cleavage of the bond between the carbonyl carbon (C-6 of the UDP-sugar) and the serine oxygen from the peptide adduct. However, no signal was detected. When the instrument was scanned for the mass loss m/z 281.5, corresponding to the loss of an adduct from a doubly charged peptide, two peaks with HPLC retention times of 18.28 min and 25.59 min were observed in the total ion chromatogram, labeled peptide A and peptide B (Figure 3.4B). By analyzing these peaks in normal mode, two doubly charged peptides m/z A = 987.5 ± 1 (Figure 3.4C) and B = 1101.5 ± 1 (Figure 3.4D) were identified. No peaks were detected in the neutral loss spectrum of the control sample in which the unlabeled Cys260Ser mutant was digested in an analogous manner (Figure 3.4E). Both 84 peptides could be purified and obtained in microgram quantities from the collected fractions split off prior to entry into the mass spectrometer and subjected to further analysis. Identical results were obtained when UDP-Glc-6-CHO was employed (Figure 3.5). 85 1000 1100 1200 1300 m/z, amu 1400 1800 Peptide B 11CJ1.5 IU U l 800 1000 1100 1200 1300 m/z, amu 36 40 Figure 3.4 Electrospray mass spectrometry experiments on Cys260Ser mutant peptic digests (UDP-glucose was employed): (A) labeled sample, total ion chromatogram in normal MS mode; (B) labeled sample, TIC in neutral loss mode; (C) mass spectrum of peptide A in panel B; (D) mass spectrum of peptide B in panel B and (E) unlabeled sample, in neutral loss mode. 86 A B C D 15000-12000' S000' 6000 3000 1200 1300 m/z, amu Peptide B 11( 1.5 — J J I I J I lH J -800 900 1000 1 100 1200 1300 1400 m/z, amu 13.75 16.04 21 Figure 3.5 Electrospray mass spectrometry experiments on Cys260Ser mutant peptic digests (UDP-Glc-6-CHO was employed): (A) labeled sample, total ion chromatogram in normal MS mode; (B) unlabeled sample, in neutral loss mode; (C) labeled sample, TIC in neutral loss mode; (D) mass spectrum of peptide A in panel C; (E) mass spectrum of peptide B in panel C. 87 . The technique of neutral loss tandem mass spectroscopy relies on the loss of a neutral species, indicating that the pyrophosphate linkage was in a protonated state during the cleavage. This is reasonable since the HPLC eluent contained 0.05 % trifluoroacetic acid (TFA) and 25 % acetonitrile. It should also be noted that in the labeled glycosidase, the labile bond is between a glutamate side chain and an anomeric hydroxyl (Figure 3.6), and a distinctly different cleavage pattern was observed. In the case of the glycosidase, an acetal bond between the anomeric carbon of the carbohydrate and the oxygen of the glutamate 138 carboxyl group was broken. In our study, the bond that was broken was an ester bond between the C-6 carbonyl carbon of the UDP-sugar and the oxygen of a serine side chain (Figure 3.7). Apparently in the glycosidase studies, the acetal bond was more readily cleaved than the ester bond. In our studies, the ester bond was the most labile bond because the linkage was not through an acetal. This is an interesting finding since it may generalize the use of neutral loss mass spectroscopy to any species linked to an enzyme via a normal ester linkage. Glycosidase Figure 3.6 The adduct structure of glycosidase and cleavage site. 88 Cys260Ser Figure 3.7 The cleavage site of the adduct of Cys260Ser mutant of UDP-glucose dehydrogenase. The masses of the two labeled singly charged peptides were 1974 + 1 and 2202 ± 1, as calculated from the doubly charged peptide masses of 987.5 and 1101.5 respectively. Identical results were obtained regardless of whether UDP-glucose or UDP-Glc-6-CHO was employed. Interestingly two more doubly charged peptides with m/z of 787.0 and 900.0 could sometimes be detected in the mass spectrum of either peptide A (m/z 989.0) or peptide B (m/z 1102.5) (Figure 3.8) when scanned in normal mode. These two peptides have a mass increase of 161 from the unlabeled peptide instead of 563. The mass difference between 161 and 563 is 402, corresponding to the loss of a UDP fragment (Figure 3.9), which is analogous to the loss of the label in the glycosidase studies. o 18000 1 5 0 0 0 4 12000 £ 9000 6000H 3000H Peptide A 989.0 787.0 376.5 1324.0 600 900 Peptide B 1200 m/z, amu 1 5 5 2 . 0 •la...|.|*!J !,!•!, I' r,M.|„ 1 5 0 0 48000H 40000H g- 32000 "I 24000 CD 1 6 0 0 0 H 8000 900.0 1102.5 399.5 i ^ i 11 iii ii - L i .lL„. Ji iliyiJ.,!.,,:,!,,..^.!..!. 6 0 0 900 1 1200 m/z, amu 1 4 6 9 . 0 ••"..•'L-fj--'. 1 5 0 0 Figure 3.8 Mass spectra of labeled peptide A and peptide B (doubly charged). 90 Figure 3.9 Possible structure of the second labeled peptide. 3.4.3 Identification of the Catalytic Nucleophile Since the masses of the singly charged peptides bearing adducts are 1974 and 2202, the corresponding masses of the unlabeled peptides were calculated to be 1412 and 1640 by subtracting the mass of 563 and adding one for the hydrogen of the serine side chain. The candidate peptides were identified by computer analysis of the amino acid sequence of the Cys260Ser mutant. The masses 1412 ± 1 and 1640 ± 1 were used to search all possible peptide sequences. Only one peptide fragment was identified to correlate to each of the mass 1412 and 1640, GYGGYSLPKDTKQ and GYGGYSLPKDTKQLL respectively. Both peptides contain the same 13 amino acids including the site of mutation, Ser260 residue, with one peptide sequence two amino acids longer than the other one. Ser260 is very likely the catalytic nucleophile that formed an ester bond with the UDP-sugar. Therefore the sequences for the two labeled peptides containing adduct at Ser260 are peptide A = 91 GYGGYXLPKDTKQ and peptide B = GYGGYXLPKDTKQLL where X is the acylated Ser260. Sequences of peptides can be obtained by ah electrospray MS/MS sequencing experiment in which a daughter ion scan mode is used. In this experiment, the parent ion was selected in the first quadrupole (Ql) and subject to collision-induced fragmentation in a collision cell in the second quadrupole (Q2), then the masses of the daughter ions produced were detected in the third quadrupole (Q3). Unfortunately, sequence information on the labeled peptides could not be obtained by this experiment because the labeled peptide cannot be further fragmented. Since the unlabeled peptide mixture of the control sample had a similar HPLC retention time to the labeled sample, each peptide fraction whose retention time was around that of the labeled peptide was collected and analyzed by ESI MS for the unlabeled peptides that correspond to the labeled peptides A and B. The sequence of unlabeled peptide 1640 was confirmed to be the same as predicted by additional fragmentation in the daughter ion scan mode by ESI MS/MS. The ion peaks represent peptides loss fragments from both N-terminus and C-terminus were observed (Figure 3.10). To confirm the identity of the labeled peptide B, N-terminal sequencing by the Edman degradation method was performed, and the first five residues GYGGY were conclusively identified. The phenylthiohydantoin (PTH) derivative for the sixth residue (acylated Ser260) could not be identified by the quality of the sequencing data. It is unclear whether a serine bearing a UDP-sugar adduct could be converted to a phenylthiohydantoin derivatized amino acid necessary for visualization. In any event, it is unclear whether the ester linkage would survive six rounds of Edman degradation analysis. 92 G--Y-b2: 220.7-b3: 277.9 b4: 334.7-b5: 499.2-bn CH-1 CO NH-CH 1 Rn Rn+1 •K4 b9: 923.5. b10: 1038.8-*-b112+: 569.6 •T 4 b12: 1267.1 K- • Q - -L -L b13: 1395.8 1.7e5-1.6e5-1.5e5-1.485-1.3e5-1.2e5-1.185-u 1.0eS-£ • 10 9.0e4-c o 1: 8.0e4-7.0e4-6.0e4-5.0a4-4.0e4-3.084-2.084-1.084-942.5 blO 1038.8 bl2 bl3 1 ( ) 5 5 6 1142.6 1267.1 1385.8 n i l ' I I . , h « 7 . ^ ° 6 1508 1200 1400 Figure 3.10 Tandem MS/MS daughter ion spectrum of unlabeled peptide (m/z 820.7 in doubly charged state). 93 3.4.4 Stability of the Adduct The labeled peptides were quite stable under mildly acidic or neutral conditions. The pH for peptic digestion buffer was 2, and the enzyme kinetic assay buffer had a pH of 8.7. The adduct survived in both conditions. However, a 15 minute incubation of the peptide-adduct A at pH 10.2 and 40 °C resulted in loss of the adduct and produced an unmodified peptide as analyzed by ESI MS. If the pH of a solution of the peptide-adduct was adjusted to 10.2 and then immediately neutralized, negligible loss of the adduct was observed. This reactivity toward base-catalyzed hydrolysis is consistent with an ester linkage. An ion-pair reversed-phase HPLC method was employed to investigate the stability of the adducts. The nucleotides and nucleotide sugars involved in the UDP-glucose dehydrogenase catalyzed reaction can be separated using the ion-pair reversed-phase HPLC method developed by Meynial et al.144 In this method, hydrophobic tetrabutylammonium hydrogen sulfate (TBAHS) is added to the elution buffer and serves as a counterion that retains charged species on the reversed-phase column. UDP-glucose, UDP-Glc-6-CHO, NAD + and NADH can be easily separated from UDP-glucuronic acid by eluting with a linear gradient of acetonitrile in water. The separation can be monitored at 262 nm, the wavelength of maximal absorbance for the uridine chromophore. A sample of the enzyme-adduct was obtained by incubation of the Cys260Ser mutant with UDP-glucose and NAD + at 30 °C for 40 minutes, followed by dialysis against either 50 mM phosphate buffer (pH 8.7) or 50 mM phosphate buffer (pH 8.7) with 8 M urea. After the dialysis, the pH was adjusted to 12 by addition of 1 M NaOH and the sample was incubated at 45 °C for 22 minutes and then neutralized with 0.1 M phosphate buffer (pH 7). A peak with the same retention time as UDP-glucuronic acid was observed in both cases and this 94 peak co-eluted with authentic UDP-glucuronic acid (Figure 3.11). A control experiment was carried out on a sample of unlabeled mutant enzyme under identical conditions. HPLC analysis showed that no peak was observed at the retention time of UDP-glucuronic acid. Peaks with the retention times of UMP and UDP were also observed in the HPLC analysis of enzyme sample (Figure 3.11). The reason for their appearance is that UDP-glucuronic acid is unstable in very basic solutions. If a solution of UDP-glucuronic acid is adjusted to pH 12 and incubated at 50 °C for 20 minutes, 75 % of the UDP-glucuronic acid will decompose with UMP and UDP being observed in the HPLC trace. A similar incubation at a pH above 13 will result in the complete decomposition of UDP-glucuronic acid. The decomposition of UDP-glucuronic acid to release UMP might be due to the attack of the deprotonated C-2 hydroxyl at the a- phosphorus to generate D-glucuronic acid 1,2-cyclic phosphate.145'146 The observation of UDP cleaved from UDP-glucuronic acid could be due to the base-catalyzed hydrolysis of the glycosyl ester.146 Attempts to identify the hydrolysis product using the labeled peptide were unsuccessful due to the limited amount of labeled peptide that could be purified from the peptic digestion mixture and the low stability of UDP-glucuronic acid under the basic conditions. O O D-glucuronic acid 1,2-cyclic phosphate 95 (A) A262 0.'00 5.'00 lbldo li.'do 2o!6o 2s'.6o 3o'.6o Time (min) (B) 0.10- 1 0.0B-2 0.06- 1 1 A262 0.04-0.02-J 0.00-J I A 0.00 5.00 10.00 15.00 20.00 25.00 30.00 Time (min) Figure 3.11 HPLC analysis of the hydrolysis products of labeled mutant enzyme. (A)hydrolysis sample, (B)hydrolysis sample spiked with authentic UDP-glucuronic acid. (1 = UMP, 2 = UDP, 3 = UDP-glucuronic acid) 96 3.5 Kinetic Studies on the Cys260Ser Mutant 3.5.1 Initial Attempts Prolonged incubation with either UDP-glucose or UDP-Glc-6-CHO resulted in the formation of the corresponding ester intermediate, suggesting that the Cys260Ser mutant can slowly catalyze both oxidation steps, but hydrolysis of the unnatural ester linkage is extremely slow, and therefore the adduct accumulates. Since the hydrolysis of the ester intermediate is very slow, a stoichiometric burst of NADH produced during the formation of the adduct would be expected. Incubating the Cys260Ser mutant (1 mg/mL) with UDP-Glc-6-CHO and NAD + resulted in a burst over two minutes. However, a very slow burst over forty minutes was observed for the incubation of the mutant enzyme with UDP-glucose and NAD + . Both bursts were followed by an extremely slow, yet similar, rate of production of NADH that is presumably due to turnover of the adduct via hydrolysis. Theoretically the burst for UDP-glucose should be twice that for UDP-Glc-6-CHO. However, by following the increase in absorbance at 340 nm due to the production of NADH, initial attempts showed a burst of approximately 0.7 equivalent of NADH per enzyme subunit for both cases. One possible explanation for this observation could be that the equilibrium constant for adduct formation was unfavorable. Since both oxidation steps should be reversible and the aldehyde is higher in energy than the alcohol, more of the enzyme could be driven to the adduct form in the presence of the aldehyde. To test this the coupled assay method using diaphorase and j^ -iodonitrotetrazolium violet (INT) (section 2.2) was employed to pull the reaction toward product by consuming NADH as it is produced. A burst of similar size was observed. In addition, an excess amount of NAD + had no obvious effect on the size of the burst. These results raised the question of whether the two adducts were at the same 97 oxidation state and the experiments in sections 3.5.2 and 3.5.3 were performed. Subsequent studies (section 3.5.4) however brought into question the reproducibility of the burst kinetics data. 3.5.2 Identification of the Oxidation State of the Adduct The mass of peptide B (m/z 1101.5, doubly charged state) was determined at high resolution using a 9.4 Tesla Fourier transform ion cyclotron resonance mass spectrometer (FT-ICR MS) 147'148>149 coupled with a microelectrospray ion source . The mass value for peptide B (neutral species) obtained from treatment of the Cys260Ser mutant with UDP-glucose and NAD + is 2200.91 ± 0.06. The value for peptide B obtained from the incubation with UDP-Glc-6-CHO and NAD + is 2200.88 ± 0.06. Mass measurements were externally calibrated using mellitin. The data are consistent with the molecular formula C90H138N20O40P2 (neutral peptide) with a calculated mass of 2200.89. The FT-ICR MS results indicate that peptide B formed from either source is at the same net oxidation state. Since the adduct generated from the incubation of Cys260Ser with UDP-glucose and NAD + appears to be in the same oxidation state as that with UDP-Glc-6-CHO, the results of initial kinetic burst studies raised the question of whether a "hidden" hydride acceptor is responsible for the first oxidation and is only regenerated by NAD + following hydrolysis of the ester intermediate. For example, a tightly bound cofactor such as lipoic acid could accept the first hydride (converting a disulfide to two thiols) and NAD + could accept the second one. S Lipoic acid 98 If the re-oxidation of this cofactor by NAD + only occurred after ester hydrolysis, then only one equivalent of NADH would be observed in the burst with UDP-glucose. This was investigated by using UDP-[6- H] glucose as substrate to track the fate of the hydrides. 3.5.3 Studies with UDP-[6-3H]glucose First we investigated whether the "hidden" hydride acceptor is associated with the mutant enzyme. It is conceivable that some electrophilic center such as dehydroalanine or aldehydic serine could be present in the active site and could accept a hydride. This would not be detected in an electrospray mass spectrum of the enzyme since the mass change would be within error of the measurement. - N H - C - C O - - N H - C H - C O -II X C H 2 0 * ^ H Dehydroalanine Aldehydic serine The Cys260Ser mutant (1.05 mg/mL) was incubated with 3 mM NAD + , 1 mM cold UDP-glucose and 3.67 pM UDP-[6-3H]glucose (5 pCi) in 50 mM triethanolamine-HCl buffer, pH 8.7, at 30 °C for 1 hour (in a total volume of 300 pL). Then the reaction mixture was loaded onto a Sephadex G-25 size exclusion column and the mutant protein was isolated in the first fraction to be eluted. The fractions containing UDP-glucose and NAD + were eluted after the protein and interestingly the UDP-glucose eluted before the NAD + . No radioactivity was detected in the protein fraction. Duplicate experiments showed the same 99 result. More than 60% of the protein fraction had the adduct covalently attached as analyzed by ESI MS. Next we examined the possibility of a hydride accepted by a species that is associated with the mutant enzyme but is able to exchange the accepted hydrogen with solvent (as would be the case with a disulfide / dithiol species). The reaction mixture in a total volume of 200 pL contained 0.5 mg/mL Cys260Ser mutant, 3.1 mM NAD + , 1 mM cold UDP-glucose and 5.5 pM UDP-[6-3H]glucose (5 pCi) in 50 mM triethanolamine-HCl buffer, pH 8.7, and was incubated at 29 °C for about 1.5 hours. The reaction solution was then lyophilized and evaporated H 2 O was collected in a liquid nitrogen cooled trap. The total radioactivity of the collected water was measured to be about 62% of the radioactivity in the original reaction mixture. However, the total radioactivity in the H 2 0 collected from a control experiment that lacked enzyme gave 78% of the radioactivity that was added. This indicates that the solvent of the tritiated UDP-glucose stock solution is contaminated by 3 H isotope and only about 20% of total radioactivity in the stock sample was actually in the UDP-glucose. Since the percentage of tritiated UDP-glucose that actually gets converted to product is approximately 1.1 %, the background contamination of tritiated solvent overshadowed any tritium released by the enzymatic reaction. Therefore no conclusion could be drawn from this experiment. Tritium in UDP-[6-3H]glucose stock solvent could be eliminated by repeatedly lyophilizing the sample, redissolving it in water and lyophilizing again, if further study is necessary. The above results provide no evidence for the notion that a "hidden" hydride acceptor is associated with the mutant enzyme and subsequent X-ray crystallographic studies have recently confirmed this (Robert Campbell in our lab, personal communication). 100 3.5.4 Re-examination of the Burst The size of the burst for reactions carried out with the Cys260Ser mutant was re-examined using the enzyme freshly eluted from the anion-exchange HPLC column. The purity of the enzyme was assessed by SDS-PAGE and only samples without visible impurity from the Coomassie Brilliant Blue stain were used. The reactions were initiated by addition of the mutant enzyme and UV spectra were recorded at the same time for UDP-glucose and UDP-Glc-6-CHO. The ratio of the bursts resulting from the incubation with UDP-glucose (3 mM) to that with UDP-Glc-6-CHO (0.8 mM) is 1.6 (Figure 3.12). The reproducibility of this data was not good however and the error in this ratio was as high as 30 %. 0 2 0 4 0 6 0 8 0 1 0 0 1 2 0 1 4 0 1 6 0 Time (min) Figure 3.12 UV profile of the incubation of Cys260Ser mutant with UDP-glucose (solid line) and UDP-Glc-6-CHO (dashed line). [UDPG] = 3 mM, [UDP-Glc-6-CHO] = 0.8 mM. [Cys260Ser] = 0.02 mM, [NAD+] = 3 mM, [DTT] = 1 mM, 50 mM Trien-HCl buffer, pH 8.7, 30 °C. 101 In Figure 3.12, the burst for UDP-Glc-6-CHO was reached two minutes after the reaction was initiated. In the case of UDP-glucose, a very slow burst over forty minutes was observed. This observation suggests that the first oxidation of UDP-glucose to UDP-Glc-6-CHO is slower than the second oxidation of UDP-Glc-6-CHO to UDP-sugar-enzyme ester intermediate. The burst for UDP-Glc-6-CHO was less than 1 equivalent of the amount of enzyme added (ca. 0.6 equivalent). This could be due to the presence of inactive/denatured enzyme. Both bursts were followed by an extremely slow rate of production of NADH that is presumably due to turnover of the common adduct via hydrolysis. After about 85 minutes, the NADH production rates for both reactions remained constant and the rate for UDP-glucose was 1.6 times that of UDP-Glc-6-CHO, which is the same as the ratio of their bursts. This suggests that the rates of hydrolysis of the ester intermediate in the two reactions are the same as expected. It is still unclear why the burst was not reproducible and why the burst was not twice as large with UDP-glucose as compared to UDP-Glc-6-CHO. Different batches of enzyme gave different sizes of bursts. Impurities in the enzyme sample might play a role and the factors affecting the size of the burst need to be investigated further in detail. It is also conceivable that the mutant enzyme has different stability in the presence of UDP-glucose or UDP-Glc-6-CHO since the ESI MS spectra of the incubation mixture with UDP-glucose almost always showed the presence of some free mutant enzyme, whereas no free mutant enzyme was present in the spectra of the incubation with UDP-Glc-6-CHO (Figure 3.1 serves as an example). 102 3.6 Kinetic Implications and Conclusion Conservative mutations at Cys260 of UDP-glucose dehydrogenase from Streptococcus pyogenes cause a dramatic loss of activity indicating that this residue is involved in catalysis. The importance of an active site cysteine residue has been well documented in the case of UDP-glucose dehydrogenase from bovine liver. Chemical modification studies using relatively nonspecific thiol labeling reagents such as iodoacetate78, iodoacetamide78, 5,5'-dithiobis-(2-nitrobenzoate)/KCN76'77 and 6,6'-dithionicotinate81 all indicate that a cysteine is required for catalysis. The observation that the cysteine is protected from modification by the addition of UDP-glucose or UDP-xylose suggests that the catalytically important cysteine is located in the active site.77 The importance of an active site thiol has also been supported by the specific irreversible inactivation of wild type enzyme from Streptococcus pyogenes by uridine 5'-diphosphate-chloroacetol (UDC) (Figure 3.13).112 O OH OH UDC O II C I - C H 2 - C - C H 2 - O U D P + Enzyme Enzyme S - C H 2 - C - C H 2 - O U D P II O SH Figure 3.13 Structure of UDC and alkylation of a cysteine residue. 103 Inactivation by UDC is slowed by the addition of known competitive inhibitors, UDP-glucuronic acid and UDP-xylose, indicating that the process is active site-directed. Incubation with the truncated compound, chloroacetol phosphate, resulted in no detectable inactivation when tested under comparable conditions. This suggests that UDC is bound in the place of UDP-glucose and acts as an affinity label.112 If the active enzyme is denatured and then treated with excess iodoacetate, two acetate units were found to be attached to the two cysteines as analyzed by ESI MS. However, a similar incubation using an enzyme previously inactivated by UDC showed that only one acetate unit was attached to the enzyme. This observation indicates that the inactivation was due to the alkylation of an active site cysteine thiol. UDP-glucose dehydrogenase from Streptococcus pyogenes has two cysteine residues, Cys260 and Cysl62. Cys260 is strictly conserved when compared to other known UDP-glucose dehydrogenases from different sources, whereas Cysl62 is not. The dramatic loss in activity of the Cys260Ser mutant and the specific inactivation of the wild type enzyme with UDC support the notion that Cys260 is the catalytically important thiol. When the Cys260Ser mutant was treated with UDC, no enzyme-UDC adduct was formed as analyzed by ESI MS. The dramatically reduced activity of the Cys260Ser mutant indicates that Cys260 is important in catalysis, but does not directly probe its role. All of the previous studies suggest that the thiol is involved in covalent catalysis, however, the best evidence to date is the direct observation of a covalent adduct involving the mutated residue. Prolonged incubation with either UDP-glucose or UDP-Glc-6-CHO resulted in the formation of the corresponding ester intermediate, indicating that Cys260Ser mutant can slowly catalyze both oxidation steps, but hydrolysis of the unnatural ester linkage is extremely slow, and therefore the adduct 104 accumulates. This is reminiscent of the enzyme, hydroxy-3-methylglutaryl-CoA reductase, which is capable of converting (7?)-mevalonate and coenzyme A to the thioester, (5)-hydroxy-3-methylglutaryl-CoA, via a four-electron oxidation without hydrolysis.151 We propose that the first oxidation of UDP-glucose gives an enzyme bound aldehyde intermediate. The second oxidation proceeds via a hydride transfer to NAD + from a hemiacetal intermediate that was formed from the addition of the active site serine hydroxyl to the aldehyde intermediate, generating NADH and a covalently bound ester intermediate. Slow hydrolysis of the ester gives the final product UDP-glucuronic acid. The proposed mechanism is shown in Figure 3.14. C260S C260S OH O HO NAD+ NADH Cys260Ser HO' OUDP UDP-glucose Aldehyde intermediate Hemiacetal intermediate UDP-glucuronic acid Ester adduct Accumulates Figure 3.14 Proposed mechanism for Cys260Ser mutant. 105 The ability of the Cys260Ser mutant to slowly oxidize the aldehyde via formation of a hemiacetal intermediate was not totally unexpected. Many proteases such as serine proteases152 and esterases have serine residues at their active sites instead of cysteine and use serine to form tetrahedral intermediates with their substrates.153'154 In the case of rat liver mitochondrial aldehyde dehydrogenase, replacement of the catalytic cysteine with serine resulted in a mutant enzyme with low, but measurable activity. Other examples, such as studies on thiolase155 and aspartate-p-semialdehyde dehydrogenase156, also show that replacing the active site cysteine with serine resulted in a mutant enzyme that showed some activity, but was a poorer catalyst compared to the native enzyme. This is in accord with the chemical trend that RS" is a better nucleophile than RO".1 5 7 The accumulation of the ester intermediate in the reaction catalyzed by the Cys260Ser mutant indicates that hydrolysis is the rate-limiting step. This is similar to the enzyme aldehyde dehydrogenase, in which a pre-steady state burst of NADH was observed showing that the rate-limiting step occurred after NADH was formed. Chloroacetaldehyde was oxidized more rapidly than acetaldehyde indicating that aldehydes with electron-withdrawing groups were oxidized more rapidly than those without. Therefore, deacylation has been proposed to be the rate-limiting step for aldehyde dehydrogenases from different sources.40'41 It is interesting to notice that no pre-steady state burst of NADH formation was found T O with the Cys302Ser mutant of rat liver mitochondrial aldehyde dehydrogenase. This observation suggested that the rate-limiting step occurred prior to substrate oxidation and NADH formation. In addition, chloroacetaldehyde was oxidized more rapidly than propionaldehyde. Therefore, the rate-limiting step was proposed to be the formation of the 106 serine hemiacetal intermediate with the Cys302Ser mutant when a less nucleophilic residue was at the active site. The formation of an ester intermediate at the serine residue of the Cys260Ser mutant convincingly supports the involvement of a thioester intermediate in the mechanism employed by the wild type UDP-glucose dehydrogenase. It is very likely that the Cys260Ser mutant catalyzes the oxidation of UDP-glucose via a mechanism similar to that of the wild type enzyme. Therefore, we propose that the mechanism used by wild type UDP-glucose dehydrogenase from Streptococcus pyogenes is the one shown in Figure 3.15. UDPGDH UDPGDH UDP-glucose Aldehyde intermediate Thiohemiacetal intermediate UDP-glucuronic acid Thioester intermediate Figure 3.15 Proposed mechanism for UDP-glucose dehydrogenase. 107 In this mechanism, the C-6 hydroxyl of UDP-glucose is first oxidized to form an aldehyde intermediate and generates the first molecule of NADH. The aldehyde is tightly bound to the enzyme and is inaccessible to carbonyl trapping reagents. The observation that a single oxygen atom derived from solvent is incorporated into the acid product is consistent with this mechanism, and argues against the involvement of an imine intermediate as the product of the first oxidation. The second oxidation is initiated by the addition of an active site cysteine thiol (Cys260) to the aldehyde to form a thiohemiacetal intermediate. A hydride transfer from the thiohemiacetal intermediate to NAD + produces a covalently bound thioester intermediate and a second molecule of NADH. The above two steps are reversible. Hydrolysis of the thioester to give the product UDP-glucuronic acid is the final irreversible step. In the case of the Cys260Ser mutant, hydrolysis of the ester intermediate is rate-limiting. It is also likely that in the case of the wild type enzyme, hydrolysis of the thioester intermediate is the rate-limiting step. This proposal is consistent with the observation that there was no primary kinetic isotope effect when using C-6 deuterated UDP-glucose, which indicates that the two hydride transfer steps are not rate-limiting steps. The kcal value for the oxidation of synthetic UDP-Glc-6-CHO by wild type enzyme is 1.0 s"1.124 This is very similar to the value obtained for UDP-glucose measured under identical conditions (&cat = 1.2 s"1).124 In the burst study of the Cys260Ser mutant however, the formation of the ester intermediate from the oxidation of UDP-glucose is slower than from the oxidation of UDP-Glc-6-CHO. The finding of almost identical kcal values for oxidation of UDP-glucose and UDP-Glc-6-CHO by wild type enzyme therefore supports the notion that hydrolysis of thioester intermediate is a very good candidate for the rate-limiting 108 step. We speculate that the formation of the thioester from the oxidation of UDP-glucose is slower than from the oxidation of UDP-Glc-6-CHO in the wild type enzyme as well. The formation of a thioester intermediate is common to the mechanism of dehydrogenases that oxidize aldehydes, with glyceraldehyde-3-phosphate dehydrogenase serving as the best example. The mechanism of glyceraldehyde-3-phosphate dehydrogenase is well elucidated in which an active site cysteine adds to form a thiohemiacetal intermediate, and this thiohemiacetal is oxidized to thioester intermediate by a hydride transfer to NAD + . The hydrolysis of the ester intermediate with the Cys260Ser mutant is found to be much slower than the hydrolysis of the thioester intermediate with the wild type enzyme. Different rates of ester and thioester hydrolyses have also been observed with P-ketoacyl thiolase from Zoogloea ramigera. This enzyme uses covalent catalysis and forms an acetyl-S-enzyme intermediate155 with the active site cysteine nucleophile. When this residue was replaced by a serine, the mutant enzyme showed a large decrease in activity and the acetyl-O-enzyme intermediate was detected. The acetyl-O-Ser enzyme intermediate was found to be more thermodynamically stable than the acetyl-S-Cys enzyme intermediate and a lower hydrolysis rate was observed accordingly. Therefore, the greater stability of the ester (oxoester) intermediate than the thioester intermediate may contribute to the lower rate of hydrolysis in addition to the subtle geometry changes in the case of the Cys260Ser mutant of UDP-glucose dehydrogenase. The 'H NMR spectrum of UDP-Glc-6-CHO in D 2 0 indicates that more than 95 % of the aldehyde is present as a hydrate (Figure 3.16).124 There are two possibilities for the oxidation process of UDP-Glc-6-CHO. The enzyme, either the wild type or the Cys260Ser mutant, may catalyze the oxidation of only the minor unhydrated form of UDP-Glc-6-CHO 109 present in solution and the hydrated form of aldehyde could bind nonproductively in a competitive manner. Alternatively, the enzyme may bind the hydrated form of the aldehyde and catalyze its dehydration prior to oxidation. Since the adduct was observed, it seems unlikely that the oxidation proceeds directly through the hydrated form of the aldehyde. The UDP-glucose dehydrogenase from Streptococcus pyogenes and the bovine liver enzyme share 31 % identity.71 The sequence similarities between the two enzymes suggest that they share a common ancestry and likely use a common mechanism. The enzyme from E. coli shares 53 % sequence identity with streptococcal enzyme106 and the same mechanism should apply to it. 3.7 Experimental Methods 3.7.1 General The plasmid for Cys260Ser was provided by Dr. Ivo van de Rijn from the Wake Forest University Medical Center, North Carolina, USA. UDP-g7wco-hexodialdose (UDP-Glc-6-CHO) and UDP-chloroacetol (UDC) were generously provided by Robert Campbell and Rafael Sala respectively from our laboratory. UDP-[6-3H]glucose was obtained from H OUDP Figure 3.16 Hydrated form of UDP-Glc-6-CHO. 110 Sigma Radiochemicals. UDP-glucose dehydrogenase wild type enzyme was prepared as described in Chapter Two. A l l other chemicals were obtained as described in Chapter Two, unless otherwise noted. 3.7.2 Purif icat ion of the Cys260Ser Mutan t The plasmid encoding the Cys260Ser mutant was transformed and overexpressed in E. coli J M 109 cells. The overexpression and purification procedure is analogous to that of the wild type enzyme (Section 2.8.4) with some modifications. The plasmid was transformed into CaCi2-competent E. coli JM109 (DE3) cells and inoculated onto LB-agar plates containing 25 pg/mL chloramphenicol. After overnight incubation at 37 °C, a single colony was used to inoculate 500 mL of T Y P G media with 25 pg/mL chloramphenicol. The cell culture was grown at 37 °C with vigorous shaking (280 rpm) until the culture reached an optical density (measured at 600 nm) of 0.7 - 1.0. Isopropyl-l-thio-/?-D-galactopyranoside (IPTG) (48 mg) was added at this point to a final concentration of 0.4 m M to induce the overexpression of the mutant enzyme. After three hours of further growth, the cells were harvested by centrifugation for 15 min at 5000 rpm (Sorvall GSA rotor) and the cell pellet (about 2 g - 2.4 g) was frozen in liquid nitrogen and stored at -78 °C. The cell pellet was later thawed rapidly with warm water and resuspended in 5mL cold 50mM triethanolamine-HCl buffer, pH 8.7, containing 2 m M DTT, 10% (v/v) glycerol, 1.5 m M phenylmethanesulfonyl fluoride (PMSF), 1 mg/L pepstatin, and 1 mg/L aprotinin (buffer A). The stock solutions of PMSF, pepstatin (5 mg/mL) and aprotinin (5 mg/mL) were prepared in ethanol. The resuspended cells were lysed by two passes through a chilled French I l l pressure mini cell (SLM Aminco) at 10,000 psi. Following ultracentrifugation at 30,000 rpm for 45 minutes to pellet the cell debris, the resulting supernatant (lysate) was diluted with an equal amount of buffer A . The diluted cell lysate was loaded onto a 15 mL column of diethylamino-ethylcellulose (DE52, Whatman) that had been preequilibrated with buffer A (in the cold room 5 °C). The column was washed with buffer A (50 mL) and then eluted successively with buffer A containing 0.1 M NaCI (20 mL), 0.15 M NaCI (20 mL) and 0.2 M NaCI (20 mL). The fractions were collected when there was absorbance at 280 nm monitored by an U V detector (Spectra/Chrom Flow Thru U V monitor controller, Spectrum) connected to the column. The fractions containing the mutant enzyme were determined by SDS-PAGE analysis, using a Bio-Rad Mini PROTEAN II electrophoresis system. Then the fractions containing the mutant enzyme were pooled and concentrated to about 4-5 mL, the resulting protein was dialysed against 1 L of Buffer B (50 m M Trien-HCl buffer with 2 m M DTT and 10% glycerol, pH 8.5 - 8.6) for 12 hours. In order to get the best result, after the cells were lysed, protein samples were not frozen until this point. The frozen partially purified protein aliquots were quickly thawed before injection (about 25 mg of total protein in each injection) onto a Waters AP-1 Protein-Pak Q column (10 x 100 mm) that had been preequilibrated with filtered and degassed buffer B. The HPLC system consisted of a Waters 625 L C system, monitored at 280 nm using a Waters 486 tunable absorbance detector. The HPLC purification was carried out by either keeping the HPLC column and buffers in ice or moving the whole HPLC system to the cold room. The column was washed with buffer B and eluted with a step-wise gradient of NaCI in buffer B, 112 0 - 0.05 M NaCI in 5 min, 0.05 - 0.1 M NaCI in 30 min and 0.1 - 0.15 M NaCI in 30 min. Fractions containing the mutant enzyme were determined by SDS-PAGE analysis. The purity of the mutant protein was judged by 12 % SDS-PAGE. Protein bands were visualized by staining with Coomassie Brilliant Blue. The molecular weight of Cys260Ser mutant was determined by ESI MS (Perkin-Elmer Sciex API 300), under identical conditions to those described for the wild type enzyme in Chapter two. The electrospray mass spectrometry was performed by Shouming He in the lab of Dr. Stephen Withers, Department of Chemistry, University of British Columbia. 3.7.3 Analysis of the Mutant Enzyme Activity Protein concentration was determined by the method of Bradford, using bovine serum albumin (BSA) as the standard. Both the protein assay solution and the standard enzyme were purchased from Bio-Rad. The mutant enzyme activity was measured under saturating conditions. A l l assays were performed at 30 °C in 50 m M Trien-HCl buffer, pH 8.7, containing 1 m M DTT, 3 m M UDP-glucose and 3 m M N A D + (0.5 mL total volume). Generation of N A D H was monitored at 340 nm (s = 6220 M " 1 cm"1) using a Varian Cary 3E UV-visible spectrophotometer equipped with a circulating water bath. 3.7.4 Protein Adduct Formation Protein adduct formation was performed by incubating the Cys260Ser mutant (1 mg/mL) with 3 m M N A D + and 1 m M of either UDP-glucose or UDP-Glc-6-CHO in 50 m M 113 Trien-HCl buffer, pH 8.7, at 30 °C for 1 hour or more. The resulting proteins were analyzed by ESI MS or used immediately for proteolysis or stored at -78 °C until needed. 3.7.5 M a s s S p e c t r a All mass spectra (intact LC/MS, tandem MS (MS/MS) neutral loss, and MS/MS daughter-ion spectra) were recorded with a Perkin-Elmer Sciex API 300 triple quadrupole mass spectrometer equipped with an electrospray ion source and interfaced with a HPLC system (Michrom BioResources). The mass spectrometry was performed by Shouming He. Analyses of the proteins and protein adduct were carried out by intact LC/MS as described in Section 2.8.6. The protein sample was introduced into the mass spectrometer through a microbore PLRP column (1 x 50 mm) and eluted with a gradient of 20 - 80 % acetonitrile in water with 0.05 % trifluoroacetic acid at a flow rate of 50 pL/min over 6 min. The mass spectrometer was scanned over a range of 400-3000 Da with a step size of 0.5 Da and dwell time of 1 ms per step. 3.7.6 P r o t e o l y s i s The protein adduct sample (14 pL in total with 0.096 mg protein) was mixed with 62 pL of 50 mM phosphate buffer, pH 2, followed by immediate addition of freshly made pepsin solution in 50 mM phosphate buffer, pH 2, (24 pL, stock concentration 0.4 mg/mL) to a final concentration of 0.096 mg/mL pepsin. The total volume was 100 pL with a final concentration of protein adduct or mutant enzyme of 0.96 mg/mL. The amount of pepsin added was 1/10 of the amount of protein to be proteolysed. The digest was incubated at 37 °C 114 for 6 hours and quenched by freezing in liquid nitrogen, and then submitted to MS analysis. Identical conditions were used for the proteolysis of the control sample. 3.7.7 MS Analysis of the Proteolytic Digest The peptide mixture generated by peptic digestion was separated by reversed-phase HPLC on an Ultra-fast Microprotein Analyser (Michrom BioResources). In each of the MS experiments, the peptic digest was loaded on a C18 column (Reliasil, 1 mm x 150 mm) and eluted with a gradient of 0 - 40 % solvent B over 40 min at a flow rate of 50 pL/min (Solvent A : 0.05 % (v/v) trifluoroacetic acid and 2 % (v/v) acetonitrile in water. Solvent B: 0.045 % (v/v) trifluoroacetic acid and 80 % (v/v) acetonitrile in water). A post column splitter was set up between the HPLC system and mass spectrometer, sending 90 % of the sample into a fraction collector and 10 % into the mass spectrometer. HPLC total ion chromatogram (TIC) of the whole digestion mixture was recorded first in single-quadrupole scan mode. The TIC of the digestion mixture for the second HPLC run was then collected in the triple quadrupole neutral loss mode, scanning for the loss of m/z 281.5 Da which corresponds to the loss of the UDP-sugar adduct from a peptide in a doubly charged state. The collision gas was argon/nitrogen (9:1). HPLC fractions corresponding to the labeled peptide were collected and saved for other analysis. The m/z mass range of 400 to 1800 Da was scanned with a step size of 0.5 Da and a dwell time of 1.5 ms per step. The MS/MS daughter ion spectrum was obtained in the triple quadrupole daughter scan mode by selectively introducing the m/z 820 unlabeled doubly charged peptide from Ql into the collision cell Q2 and observing the daughter ions in Q3. Ql was locked on m/z 820 115 and Q3 was scanned in a range of 400 to 1800 with a step size of 0.5 and a dwell time of 1.5 ms per step. 3.7.8 Chemical Sequencing The peptic digestion sample was separated by C18 reversed phase HPLC column coupled with mass spectrometer to check for the purity. The fractions containing the labeled peptide was collected after the post column splitter for sequencing. Peptide B (2201 Da, in 25 % acetonitrile and 0.05 % TFA in water) was coupled to arylamine-functionalized polyvinylidene fluoride membranes (Sequelon AA, Milligen/Bioresearch) using N-ethyl-N'-(3-dimethylaminopropyl) carbodiimide before the solid phase sequence analysis. Peptide sequences were determined by solid phase Edman degradation on a Milligen/Bioresearch model 6600 protein sequenator. Standard protocols with on-line HPLC analysis of the resulting phenylthiohydantoins was employed. Edman degradation analysis was performed by Suzanne Perry of the Nucleic Acid Protein Service Unit, Biotechnology Lab, University of British Columbia. 3.7.9 Stability of Peptide Adduct Purified peptide B solution was adjusted to pH 10.2 by addition of 0.1 M NaOH and incubated at 40 °C for 15 min. The mass of the resulting peptide was measured by ESI MS (Perkin-Elmer Sciex API 300). 116 3.7.10 Investigation of the Adduct by HPLC The protein adduct was obtained by incubating the Cys260Ser mutant (1 mg/mL) with NAD + (3 mM), UDP-glucose (3 mM) and DTT (1 mM) in 50 mM Trien-HCl buffer, pH 8.7, at 30 °C for 40 minutes to 1 hour. The solution was then dialyzed against either 1 L 50 mM phosphate buffer, pH 6, or 50 mM phosphate buffer, pH 6, with 8 M urea. Two rounds of dialysis were performed with each round lasting 20 hours. The resulting solution was adjusted to pH 12 with 1 M NaOH and incubated at 45 °C for 22 minutes, followed by neutralization using 50 mM phosphate buffer, pH 6. The resulting sample was lyophilized and redissolved in 40 pL H2O. An aliquot of the sample (20 pL) was injected onto a Waters Radial-Pak 8NVC18 reversed-phase HPLC column pre-equilibrated with solvent C (Solvent C: 50 mM phosphate buffer, pH 7, with 2.5 mM TBAHS) and eluted with a gradient of 0 - 2 % solvent D in 18 minutes followed by 2 % - 30 % solvent D over 40 minutes (Solvent D: 50 mM phosphate buffer, pH 7, with 50 % (v/v) acetonitrile and 2.5 mM TBAHS). The eluent was monitored at 262 nm. Another aliquot (20 pL) was mixed with authentic UDP-glucuronic acid (15 pmole, 3 times the amount that eluted from the previous injection), injected onto the HPLC column, and eluted with the same gradient. 3.7.11 Stability of UDP-glucuronic Acid The samples of UDP-glucuronic acid solution were adjusted to pH 12 and pH 13 by addition of 1 M NaOH and incubated at 50 °C for 20 minutes. An aliquot of the sample (20 pL) was injected onto a Waters Radial-Pak 8NVC18 reversed-phase HPLC column pre-equilibrated with solvent C and eluted with a gradient of 0 - 20 % solvent D over 30 minutes. The eluent was monitored at 262 nm. More aliquots (20 pL) were mixed with either authentic 117 UMP or UDP and analyzed by the same ion-pair reversed-phase HPLC column eluting with the same gradient as above. 3.7.12 Kinetic Studies for the Burst Formation . The burst experiment was carried out on a Varian Cary 3E UV-visible spectrophotometer equipped with a circulating water bath. The reaction mixtures generally contain 1 mg/mL Cys260Ser mutant, 1 mM DTT, 3 mM NAD + , 3 mM UDP-glucose or 0.8 -1 mM UDP-Glc-6-CHO in 50 mM Trien-HCl buffer, pH 8.7, with a volume of 500 pL in total. The reactions were initiated by the addition of the Cys260Ser mutant. The reaction solution was mixed by being pipetted up and down with pipettes having long thin tips. In this way the cuvettes were not moved after the UV spectrometer started recording. The reaction was followed by measuring the absorbance increase at 340 nm. 3.7.13 FT-ICR Mass Spectra of the Labeled Peptides FT-ICR mass spectra were kindly obtained by Stone D.-H. Shi at the National High Magnetic Field laboratory, Tallahassee, Florida using a house built 9.4 Tesla FT-ICR mass spectrometer coupled with a microelectrospray ion source. The sample was diluted with a solution of water : methanol (50 : 50) containing 0.25 % acetic acid. The solution microdroplets generated by the electrospray process is passed through a heated metal capillary (heating current 3.0-4.5 A). Ions were accumulated in the first octupole of the instrument for 5 to 20 seconds and then transferred into the ICR cell through the second octupole. 512K data was required. Mass measurements were externally calibrated using mellitin. 118 3.7.14 Experiments with UDP-[6-3H]glucose (A) The reaction was carried out in 50 mM Trien-HCl buffer, pH 8.7, containing 3.67 pM UDP-[6-3H]glucose (5 pCi), 1 mM UDP-glucose (cold), 3 mM NAD + and the Cys260Ser mutant enzyme (1.05 mg/mL) with a total volume of 300 pL. The solution was incubated at 30 °C for 1 hour and loaded onto a Sephadex G-25 size exclusion column (80 mL, purchased from Pharmacia Biotech Company) preequilibrated with 100 mM Trien-HCl buffer, pH 8.0, containing 1 mM DTT. The column was eluted with 100 mM Trien-HCl buffer, pH 8.0, containing 1 mM DTT. The separation was monitored by an UV detector with a 280 nm filter (Spectrum). Three fractions corresponding to the mutant enzyme, UDP-glucose and NAD + were collected. 1 mL sample was removed from each fraction, mixed with 15 mL of Scinti Verse E (general purpose scintillation cocktail, Fisher Scientific Company) and counted for 3 H isotope on a Packard Tr 1900 TRI-CARB liquid scintillation counter. Before each run of counting, 3 H and l 4 C standards were run to assess the machine performance. (B) A sample containing the Cys260Ser mutant enzyme (0.5 mg/mL), 5.5 pM UDP-[6-3H]glucose (5 pCi), 1 mM cold UDP-glucose and 3.1 mM NAD + in 200 pL 50 mM Trien-HCl buffer, pH 8.7, was incubated at 29 °C for 1.5 hours. The sample solution was then degassed by cycles of freezing in liquid nitrogen, applying vacuum, and then thawing. The frozen sample was put on a vaccum system and the evaporated H 2 0 was collected using a liquid nitrogen trap. The H 2 0 sample was mixed with 15 mL Scinti Verse E solution and counted on a Packard Tr 1900 TRI-CARB liquid scintillation counter. A control experiment that lacked the enzyme was performed under identical conditions Chapter Four Mechanistic Studies on the Cys260Ala Mutant 120 UDP-glucose dehydrogenase from Streptococcus pyogenes has two cysteine residues, Cys260 and Cysl62. Cys260 is conserved in all known sequences of UDP-glucose dehydrogenase from different sources. This residue was mutated to serine and alanine in order to gain insight into its role in the dehydrogenase mechanism. The preceding chapter described the properties of the Cys260Ser mutant. This chapter will discuss the results from studies on the Cys260Ala mutant. 4.1 Purification of the UDP-Glucose Dehydrogenase Cys260Ala Mutant The plasmid pGAC400 encoding for the Cys260Ala mutant was provided by Dr. Ivo van de Rijn from the Wake Forest University Medical Center. The overexpression and purification procedures are analogous to those used for the Cys260Ser mutant except that ampicillin was used instead of chloramphenicol. The purity of the enzyme was assessed by SDS-PAGE analysis. A protein band with a molecular weight of about 46 kDa was the only band visible on the gel after staining with Coomassie Briliant Blue. The purification gave about 22 - 25 mg of protein per 500 mL culture. The molecular weight of the Cys260Ala mutant enzyme was determined to be 45 459 ± 6 Da by electrospray mass spectrometry (ESI MS), consistent with the expected 45 452 Da from its gene sequence. 4.2 Kinetic Analysis The Cys260Ala mutant showed very little dehydrogenase activity when incubated with UDP-glucose and NAD + under saturating conditions (less than 0.01% of the wild type activity). This is in contrast to what was observed with histidinol dehydrogenase which was 121 thought to employ a mechanism similar to that of UDP-glucose dehydrogenase. Replacement of the conserved cysteine residues of histidinol dehydrogenase with serine or alanine resulted in mutants with normal catalytic constants. ' Therefore, it is very likely that histidinol dehydrogenase catalyzes the oxidation of histidinol to histidine via a different mechanism. It is also important to note that sequence alignment analysis shows no similarities between the two enzymes. A surprising observation was that UDP-g/wco-hexodialdose (UDP-Glc-6-CHO) was readily oxidized by the Cys260Ala mutant in the presence of NAD + (Figure 4.1). No protein adduct was observed when the Cys260Ala mutant was incubated with either UDP-glucose or UDP-Glc-6-CHO in the presence of NAD + as analyzed by ESI MS. 0 2 4 6 8 1012 1 4 1 6 1 8 2 0 2 2 2 4 2 6 28 30 Time (min) Figure 4.1 Incubation of the Cys260Ala mutant with UDP-Glc-6-CHO and NAD + . [Cys260Ala] = 2.2 pM, [UDP-Glc-6-CHO] = 1.0 mM, [NAD+] = 3.0 mM. 122 4.3 Studies on the Oxidation of UDP-Glc-6-CHO by the Cys260AIa Mutant The kinetic constants for the oxidation of UDP-Glc-6-CHO by the Cys260Ala mutant were determined to be kcat = 0.19 + 0.01 s"1 and Km = 0.26 + 0.04 mM. The kcat value was within an order of magnitude of that observed with the wild type enzyme (kcat = 1.2 s"1). The Km value for the Cys260Ala mutant was 18 fold higher than that of the wild type enzyme (Km = 0.014 mM), suggesting that the substrate binding is weaker. The mutation prevents UDP-Glc-6-CHO from forming a covalent bond with the mutant enzyme as would occur in the case of the wild type enzyme (thiohemiacetal intermediate), and in addition some noncovalent bonds, such as hydrogen bonds and hydrophobic interactions, may be weakened. It appears that the Cys260Ala mutant is able to catalyze the second oxidation (oxidation of the aldehyde intermediate) somewhat efficiently, even without a catalytic nucleophile. Previous studies on a sample of the bovine liver UDP-glucose dehydrogenase in which the active site thiol was cyanide-derivatized also showed that the modified protein retained a limited ability to catalyze the second oxidation step.76 It is unlikely that other amino acid residues near the active site could be a surrogate for Cys260. Therefore the Cys260Ala mutant may catalyze the oxidation of UDP-Glc-6-CHO via a different mechanism from that of the wild type enzyme and the Cys260Ser mutant. The oxidation of UDP-Glc-6-CHO by the Cys260Ala mutant is likely to occur through the hydrated form of the aldehyde (UDP-Glc-6-CH(OH)2) that resembles the thiohemiacetal intermediate and is able to bind in an analogous fashion. UDP-Glc-6-CHO was found to exist primarily in the hydrated form when free in solution.124 The alternative mechanism is shown in Figure 4.2. 123 C260A H C H 3 HO NAD+ N A D H V S'» HO' OUDP OUDP Figure 4.2 The oxidation of UDP-Glc-6-CHO catalyzed by the Cys260Ala mutant. Some alcohol dehydrogenases, such as the Drosophila melanogaster and the horse liver alcohol dehydrogenases, are able to catalyze the dismutation of an aldehyde to a mixture of a carboxylic acid and the corresponding alcohol. It has been postulated from studies on horse liver alcohol dehydrogenase that the enzyme further oxidizes its aldehyde products via hydrated gem-diol species.49 Mutation of the essential nucleophilic cysteine residue (Cysl49) to an alanine in the glyceraldehyde-3-phosphate dehydrogenase from E. coli generated a new enzyme that can catalyze the oxidation of substrates but lacks phosphorylating function.105 The mutant enzyme converts glyceraldehyde-3-phosphate into 3-phosphoglycerate instead of 1,3-diphosphoglycerate with the wild type enzyme. Mechanistic studies on this mutant (Cysl49Ala) indicated that no acylenzyme intermediate is formed during the catalytic event and a different mechanism (Figure 4.3) might be employed from the native enzyme (Figure 4.4), in which a gem-diol species was proposed to serve as the substrate. 124 Cys149Ala Cys149Ala I 1 C H 3 ^ u CH 3 HO. P |_| O, 3 -OH NAD+ NADH H — C — O H V 7 , H — C — O H I I CH 2 OP0 3 2- CH 2 OP0 3 2" Figure 4.3 Reaction catalyzed by the Cysl49Ala mutant of glyceraldehyde-3-phosphate dehydrogenase. GAPDH GAPDH GAPDH GAPDH 0 \ / H u n , S C K ^ s n OP0 3 2-r i u ^ / L I M i n t MAnw ^<s. / U^. / SH T 1 SH / , H NAD+ NADH I V A 1 H P ° 4 - 2 " H — C — O H H _ C _ 0 H H — C — O H - H _ C _ 0 H CH 2 OP0 3 2- CH 2 OP0 3 2- CH 2 OP0 3 2- CH 2 OP0 3 2-Figure 4.4 Reaction catalyzed by the wild type glyceraldehyde-3-phosphate dehydrogenase (GAPDH). It is interesting to note that the replacement of the catalytically important cysteine residue (Cys302) in rat liver mitochondrial aldehyde dehydrogenase resulted in a mutant T O lacking dehydrogenase activity. This Cys302Ala mutant did not catalyze the oxidation of the aldehyde substrate, propionaldehyde. Even though propionaldehyde has a hydrated form in aqueous solution, the mechanism shown in Figure 4.2 apparently does not apply to the alanine mutant of rat liver mitochondrial aldehyde dehydrogenase. 125 4.4 Investigation of the First Oxidation of UDP-glucose by the Cys260Ala Mutant Two possibilities can explain why the Cys260Ala mutant does not readily oxidize UDP-glucose to UDP-glucuronic acid. One is that the first oxidation step is extremely slow. If this is the case, it would invoke a dual role for the Cys260 thiol(ate); it could assist in the deprotonation of the C-6 hydroxyl in the first oxidation step, and in the second oxidation step it could be employed in thioester formation. The other possibility is that the UDP-Glc-6-CHO produced from the first oxidation of UDP-glucose is tightly bound in the mutant active site, and there is no mechanism by which it can be hydrated and proceed forward in the second oxidation step. To investigate these two possibilities, the reduction of UDP-Glc-6-CHO was studied and a deuterium "washout" experiment was performed. The results of these studies will be discussed in the following sections. 4.4.1 Studies on the Reduction of UDP-Glc-6-CHO by the Cys260Ala Mutant In order to test whether the mutation affected the first oxidation step, we decided to investigate the reverse reaction, namely the reduction of UDP-Glc-6-CHO with NADH. UDP-Glc-6-CHO was incubated with the Cys260Ala mutant and NADH, and the reaction was followed by the decrease of absorbance at 340 nm. The UV trace (Figure 4.5) indicates that NADH was consumed and the reduction was proceeding. The kinetic constants of the reduction reaction were then measured with smaller amounts of added enzyme under initial velocity conditions. The values of kcat and Km were determined to be 1.89 + 0.05 s"1 and 58 ± 7 pM respectively. 126 E c § 0.8 co 03 OJ O c ro -Q O w Si < 0.6 0.4 0 20 40 60 80 100 120 140 160 Time (min) Figure 4.5 Incubation of UDP-Glc-6-CH0 with the Cys260Ala mutant and NADH. [Cys260Ala] = 2.2 pM, [UDP-Glc-6-CHO] - 0.092 mM, [NADH] = 0.15 mM. In order to be reduced, UDP-Glc-6-CHO has to be bound in the active site of the Cys260Ala mutant in the non-hydrated form and in the presence of NADH. Since UDP-Glc-6-CHO is readily reduced by the Cys260Ala mutant, the first oxidation of UDP-glucose must occur according to microscopic reversibility. The enzyme-bound aldehyde form of UDP-Glc-6-CHO and NADH should be the products of the first oxidation reaction. It is possible that the Cys260 thiol(ate) is not directly involved in the deprotonation of the C-6 hydroxyl in the first oxidation step. 127 4.4.2 Effect of UDP-glucose Preincubation with the Cys260Ala Mutant on the Oxidation of UDP-Glc-6-CHO To further investigate the oxidation of UDP-glucose to UDP-Glc-6-CHO by the Cys260Ala mutant, an experiment was devised to test whether a preincubation with UDP-glucose and NAD + would affect the oxidation of UDP-Glc-6-CHO to UDP-glucuronic acid. If the Cys260Ala mutant can oxidize UDP-glucose to UDP-Glc-6-CHO and this intermediate is tightly bound in a non-hydrated form that cannot be further oxidized, the enzyme would effectively be inhibited. The reaction rate would therefore be expected to decrease when compared to a reaction without preincubation. In this experiment, the Cys260Ala mutant was preincubated with UDP-glucose and NAD + for ten minutes before an equivalent amount of UDP-Glc-6-CHO was added. The resulting reaction rate was compared to that obtained without preincubation. Since the first oxidation is reversible, a coupling enzyme, diaphorase, was added in order to drive the reaction direction toward oxidation by the consumption of NADH as it is produced. Diaphorase catalyzes the reduction of />-iodonitrotetrazolium violet (INT) in concert with oxidizing NADH to NAD + . The results showed that the preincubated mutant catalyzed the oxidation of UDP-Glc-6-CHO at a rate that was close to that of the control, which does not suggest a tightly bound species is inhibiting the enzyme. It is possible that the coupling reaction did not affect the first oxidation because both NADH and UDP-Glc-6-CHO are tightly bound in the active site of the Cys260Ala mutant and neither could be released. It may be the case that NADH dissociates only after a nucleophile adds to the aldehydic carbonyl. Since there was no 128 nucleophile present, this did not occur. To verify the notion that NADH is not released, a deuterium "washout" experiment was conducted. 4.4.3 Deuterium "Washout" Experiment Given that the first oxidation step occurs, the incubation of C-6 doubly deuterated UDP-glucose (UDP-(6,6-di-2H)glucose) with the Cys260Ala mutant and NAD + would generate UDP-Glc-6-CHO containing one deuterium atom at the carbonyl carbon and NADH 2 2 containing one deuterium atom at C-4 (NAD H). If NAD H can be released and undeuterated NADH is present in the reaction medium, the Cys260Ala mutant would re-bind either NADH or NAD 2 H and reduce UDP-Glc-6-CHO. If undeuterated NADH is bound to the mutant enzyme, a mono-deuterated UDP-glucose could be expected to be generated. Analyzing the number of deuterium atoms in the recovered UDP-glucose could shed light into the first oxidation step catalyzed by the Cys260Ala mutant. UDP-(6,6-di-2H)glucose (1.2 mM) was incubated with 20 mM NADH, 4 mM NAD + and the Cys260Ala mutant at 30 °C for 23 hours. Separation of the UDP-glucose from other components in the reaction mixture was achieved using a diethylamino-ethylcellulose (DE52) column and eluting with a linear gradient of ammonium bicarbonate buffer. A size exclusion (Bio-gel P-2) column and ion-exchange resin were used to further purify the resulting UDP-glucose fraction and prepare it for NMR and MS analysis. The 'H NMR spectrum (Figure 4.6A) (in D2O) shows only peaks for the protons at C-3 and C-5 of the glucose ring in the region of 3.7 - 3.95 ppm. These peaks are the same as those for the authentic di-deuterated UDP-glucose sample (Figure 4.6B). The NMR spectrum of undeuterated UDP-glucose at this region shows a different peak profile (Figure 4.6C) 129 because the chemical shifts for C-6 protons overlap with those of C-3 and C-5 protons. Liquid secondary ion mass spectrometry analysis also confirmed that the recovered UDP-glucose contains two deuterium atoms. These results confirm that NADH is not released after the first oxidation catalyzed by the Cys260Ala mutant. UDP-Glc-6-CHO and NAD 2 H generated from the first oxidation are tightly bound in the active site of the Cys260Ala mutant. Since the hydride transfer is stereospecific, the deuterium at C-4 of the dihydronicotinamide ring of NAD H must have been transferred back to the C-6 of the glucose of UDP-Glc-6-CHO in the reduction reaction regenerating the doubly deuterated UDP-glucose. 130 Figure 4.6 'H NMR spectra at the region 3.7 - 4.0 ppm for (A) recovered UDP-glucose from incubation of deuterated UDP-glucose with Cys260Ala, NADH and NAD + , (B) authentic deuterated UDP-glucose, and (C) undeuterated UDP-glucose. 131 4.5 Fur ther Investigation of the Reduction and Oxidat ion of U D P - G l c - 6 - C H O 4.5.1 Compar ison of the Incubation of U D P - G I c - 6 - C H O and N A D H wi th the W i l d Type Enzyme and the Cys260Ala Mutan t When UDP-Glc-6-CHO was incubated with NADH and the wild type UDP-glucose dehydrogenase, a different kinetic trace was observed from that obtained with the Cys260Ala mutant (Figure 4.7). In the case of the wild type enzyme, NADH was first consumed, then regenerated. There was no net change of NADH after incubation for over 2 hours. However, with the Cys260Ala mutant, NADH was consumed. o 0.6 w < 0.4 1 • 1 1 1 i i i i i r -t • -•\ — - \ -> i i i i i > i i i i i i i i 0 20 40 60 80 100 120 140 160 Time (min) Figure 4.7 Kinetic trace of UDP-Glc-6-CHO incubated with NADH and (A) the wild type UDP-glucose dehydrogenase (solid line), (B) the Cys260Ala mutant (dashed line). [UDP-Glc-6-CHO] = 0.075 mM, [NADH] = 0.15 mM, [DTT] = 2 mM, 50 mM Trien-HCl buffer, pH 8.7, 30 °C. [wild type] = 0.22 pM, [Cys260Ala] = 0.22 pM. 132 With the wild type enzyme, UDP-Glc-6-CHO was first reduced to UDP-glucose in the presence of NADH and NAD + was generated. Once NAD + was present, either UDP-glucose or UDP-Glc-6-CHO could be oxidized by the wild type enzyme, with NADH regenerated concurrently. Eventually, a net dismutation would produce equal amounts of UDP-glucose and UDP-glucuronic acid, and the amount of NADH would return to the starting level (Figure 4.8). In this process, NAD + is truly acting as a cofactor, as opposed to a reagent, since there is no net consumption. The driving force for the dismutation is the greater stability of the acid and the alcohol as compared to the aldehyde. NADH NAD+ "O. O H. O OUDP OUDP Figure 4.8 Possible reactions when UDP-Glc-6-CHO is incubated with the wild type UDP-glucose dehydrogenase and NADH. 133 In the case of the Cys260Ala mutant, NADH was consumed. A close look at the amount of NADH reduced showed that it equals ca. 78 % of the UDP-Glc-6-CHO added. However, the Cys260Ala mutant was found to be able to oxidize UDP-Glc-6-CHO efficiently in the presence of NAD + , with a kcat only 6 times lower than the wild type enzyme. Did some oxidation of UDP-Glc-6-CHO occur during the incubation? To answer this question, the reaction products were analyzed by HPLC. 4.5.2 HPLC Analysis of the Reduction Products of UDP-Glc-6-CHO by the Cys260Ala Mutant The technique of ion-pair reversed phase HPLC (Section 3.4.4) was used to identify the products of the incubation of 0.21 mM UDP-Glc-6-CHO with the Cys260Ala mutant and 0.188 mM NADH. The reaction mixture was incubated at 35 °C for 30 minutes. Aliquots of the reaction mixture were injected onto the reversed-phase HPLC column without further purification. The HPLC buffers contained tetrabutylammonium hydrogen sulfate (TBAHS) as a counterion to retain charged analytes on the reversed-phase column. The column was eluted with a step-wise acetonitrile gradient. HPLC analysis showed that there were NAD + , NADH, UDP-glucose and UDP-glucuronic acid in the reaction mixture and no UDP-Glc-6-CHO was detected (Figure 4.9). This observation indicates that both reduction and oxidation occur in the incubation of UDP-Glc-6-CHO with the Cys260Ala mutant and NADH. Integration of the HPLC peaks showed that the ratio of UDP-glucose to UDP-glucuronic acid is 4.3:1. This means 81 % of the products in the reaction mixture is UDP-glucose, which is consistent with the ratio of the amount of NADH reduced to the amount of UDP-Glc-6-CHO added. The existence of UDP-134 glucuronic acid indicates that the oxidation of UDP-Glc-6-CHO by the Cys260Ala mutant did occur during the incubation. This would be expected since NAD + was accumulating in the reaction mixture as the aldehyde was reduced. Thus, the majority of the aldehyde was rapidly reduced to the alcohol and "kinetically trapped" in this form; however, 20% was oxidized to the acid. 0.030-0.025-A262 0.015-0.005-I 2 1 4 1 j \ 1 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.0 Time (min) Figure 4.9 HPLC analysis of incubation products of 0.21 mM UDP-Glc-6-CHO with 0.125 mg/mL Cys260Ala and 0.188 mM NADH. Conditions: Section 4.7.6. (1 = UDP-glucose, 2 = NAD + , 3 = UDP-glucuronic acid, 4 = NADH) 4.5.3 HPLC Analysis of the Oxidation Products of UDP-Glc-6-CHO by the Cys260Ala Mutant Since the enzyme catalyzed reduction of UDP-Glc-6-CHO is more rapid than its oxidation, it is likely that some reduction occurs during the extended incubation of UDP-Glc-135 6-CHO with the Cys260Ala mutant and NAD + . HPLC analysis of the reaction mixture of 0.21 mM UDP-Glc-6-CHO, 0.6 mM NAD + and the Cys260Ala mutant after 30 minutes incubation at 30 °C showed that UDP-glucose was produced in addition to UDP-glucuronic acid (Figure 4.10). This is consistent with the observation that the mutant enzyme can catalyze the reduction of the aldehyde as NADH accumulates, and once formed, the alcohol is "kinetically trapped" and not reoxidized. Since initial velocity rates were used to calculate the kinetic constants for the oxidation of UDP-Glc-6-CHO by the Cys260Ala mutant, the results are still valid. Figure 4.10 HPLC analysis of incubation of 0.21 mM UDP-Glc-6-CHO with 0.125 mg/mL Cys260Ala mutant and 0.6 mM NAD + . Conditions: Section 4.7.6. (1 = UDP-Glc-6-CHO, 2 = UDP-glucose, 3 = NAD + , 4 = UDP-glucuronic acid, 5 = NADH) 136 4.6 Mechanistic Implications and Conclusion The dramatic loss of the UDP-glucose dehydrogenase activity in the Cys260Ala mutant indicates that Cys260 is important for catalysis. The fact that the Cys260Ala mutant can catalyze the oxidation of UDP-Glc-6-CHO efficiently suggests that the mechanism of this oxidation step differs from the covalent catalysis employed by the wild type enzyme and the Cys260Ser mutant because the Cys260Ala mutant lacks the necessary nucleophile at the active site. The oxidation of UDP-Glc-6-CHO catalyzed by the Cys260Ala mutant is proposed to proceed via the hydrated form of UDP-Glc-6-CHO that is analogous to the (thio)hemiacetal intermediate. The observation that UDP-Glc-6-CHO can be reduced by the Cys260Ala mutant indicates that the Cys260Ala mutant is able to catalyze the first oxidation step of UDP-glucose. Since the mutant enzyme can efficiently catalyze both halves of the reaction it remains to be explained why UDP-glucose is not oxidized at an appreciable rate. A reasonable explanation is that UDP-glucose is readily oxidized to give the bound aldehyde and NADH. In order for the second oxidation to occur, however the aldehyde must be hydrated (because covalent catalysis is not possible). Since hydration of the bound aldehyde is very slow, it can only occur upon release of the aldehyde into solution. This is a very slow process since the species is tightly bound in the active site. Thus there is no effective way of catalyzing the second step of the oxidation when starting from UDP-glucose. On the other hand, the enzyme can readily oxidize the free aldehyde since it can bind the hydrated form directly from solution. The results from the "preincubation" and deuterium "washout" experiments indicate that both the UDP-Glc-6-CHO and the NADH produced from the first 137 oxidation step are tightly bound in the active site of the Cys260Ala mutant, and are only released upon reverting to starting materials. Previous studies on bovine liver UDP-glucose dehydrogenase also showed that when the essential thiol group was derivatized with cyanide to prevent the formation of the thiohemiacetal, the enzyme still did not release the aldehyde intermediate into the incubation medium.77 It is possible that in the normal mechanism for the wild type enzyme and the Cys260Ser mutant, NADH is only released after the (thio)hemiacetal intermediate is formed. The Cys260Ala mutant has no nucleophile at its active site so no addition can occur and the NADH can therefore not be released. It is likely that the reason for the dramatic loss of the dehydrogenase activity of the Cys260Ala mutant is that there is no mechanism by which the UDP-Glc-6-CHO can be hydrated at the active site of the mutant enzyme and proceed forward in the second oxidation step. It is conceivable that no water molecules can easily get access to the aldehydic carbonyl at this stage of the reaction. This is unlike horse liver alcohol dehydrogenase, which has been observed to be able to catalyze the sequential oxidation of an alcohol to the corresponding carboxylic acid concomitant with the dismutation of the aldehyde intermediate.50 It has been proposed that it is possible for the horse liver alcohol dehydrogenase to catalyze the hydration of an aldehyde at the active site when a hydroxide ion is ligated to the active site zinc. The formation of a hydrate is catalyzed by delivering the zinc-bound hydroxide ion to the bound aldehyde intermediate. However, in the case of UDP-glucose dehydrogenase, no metal ions have been reported to be important for its catalysis. Treatment with EDTA does not affect the 138 dehydrogenase activity.112 Therefore, the Cys260Ala mutant might not be able to catalyze the hydration of UDP-Glc-6-CHO at its active site. It is likely that when the mutant enzyme is incubated with UDP-Glc-6-CHO, it binds the hydrated form directly and oxidizes it to the corresponding carboxylic acid. Site-directed mutagenesis studies showed that cysteine residues are not important for catalysis in the case of histidinol dehydrogenase, which had long been thought to employ a mechanism similar to UDP-glucose dehydrogenase. Interestingly Zn 2 + has been found to bind to histidinol dehydrogenase and is essential for the enzyme's activity." It has also been proposed that histidinol dehydrogenase catalyzes the oxidation of the aldehyde intermediate through a hydrated species.102 We would speculate that histidinol dehydrogenase might catalyze the hydration of the aldehyde intermediate at its active site via a mechanism like that of horse liver alcohol mechanism in which a zinc bound water molecule or hydroxide ion is delivered to the aldehyde intermediate to form a gem-diol. 4.7 Experimental Methods 4.7.1 General The plasmid for the Cys260Ala mutant was provided by Dr. Ivo van de Rijn from the Wake Forest University Medical Center, North Carolina, USA. UDP-glucose, NAD + and NADH were obtained from Sigma. UDP-g/wco-hexodialdose (UDP-Glc-6-CHO) was generously provided by Robert Campbell in our laboratory. The UDP-glucose dehydrogenase wild type enzyme and the Cys260Ser mutant were prepared as described in Chapter Two and Three, respectively. All other chemicals were obtained as described in Chapter Two, unless otherwise noted. 139 4.7.2 Purification of the Cys260Ala Mutant The plasmid that contains the Cys260Ala gene was generated using the pAlter site directed mutagenesis kit from Promega. The plasmid pGAC400 was transformed into the cells of E. coli strain JM 109 and overexpressed in TYPG medium containing 125 pg/mL of ampicillin for selection. The overexpression and purification procedures are analogous to those of the Cys260Ser mutant (Section 3.7.2). The modification was that 125 pg/mL ampicillin was added to both LB-agar plates and TYPG media. The purity of the mutant was judged by 12 % SDS-PAGE. Protein bands were visualized by staining with Coomassie Brilliant Blue. The molecular weight of the Cys260Ala mutant was determined by electrospray ionization mass spectrometry (Perkin-Elmer Sciex API 300) under the identical conditions to those described for the wild type UDP-glucose dehydrogenase in Chapter Two. The electrospray mass spectrometry was performed by Shouming He. 4.7.3 Analysis of the Mutant Enzyme Activity Protein concentration was determined by the method of Bradford, using bovine serum albumin (BSA) as the standard. Both the protein assay solution and the standard protein were purchased from Bio-Rad. The mutant activity was measured under saturating conditions. All assays were performed at 30 °C in 50 mM Trien-HCl buffer, pH 8.7, containing 1 mM DTT, 3 mM UDP-glucose and 3 mM NAD + (0.5 mL in total volume). Generation of NADH was monitored at 340 nm (s = 6220 M"1 cm"1) using a Varian Cary 3E UV-visible spectrophotometer equipped 140 with a circulating water bath. The incubation sample was analyzed using electrospray mass spectrometry. 4.7.4 Kinetic Constants for the Oxidation of UDP-Glc-6-CHO by the Cys260AIa Mutant The kinetic constants for the oxidation of UDP-Glc-6-CHO were obtained using the continuous assay method. The concentration of UDP-Glc-6-CHO stock solution was determined using its absorbance at 262 nm (£262 = 8700 M"1 cm"1). The Cys260Ala mutant (0.05 mg/mL) was added to a 500 pL solution of various concentrations of UDP-Glc-6-CHO and 0.5 mM NAD + in 50 mM Trien-HCl buffer with 2 mM DTT, pH 8.7 to initiate the reaction. The reactions were typically followed for five minutes. Rates were determined by following the increase in absorbance at 340 nm, at eight substrate concentrations ranging from 2.5 pM to 2.0 mM UDP-Glc-6-CHO, using a least squares analysis with Cary3 software (version 3.0). The resulting calculated initial velocities were plotted as a function of substrate concentrations, and the kinetic parameters were determined by a direct fit of the data to an equation using the computer program Grafit version 3.0 (Erithacus Software, 1994). The error values are reported as a deviation of the data from the calculated curve-of-best-fit. 4.7.5 Studies on the Reduction of UDP-Glc-6-CHO 4.7.5.1 UV Analysis UDP-Glc-6-CHO (75 pM) was incubated with 0.15 mM NADH and either wild type UDP-glucose dehydrogenase (0.22 pM) or the Cys260Ala mutant (0.22 pM) in 50 mM Trien-HCl buffer, 2 mM DTT, pH 8.7 at 30 °C. The change of NADH concentration was 141 monitored at 340 nm (e = 6220 M"1 cm"1) using a Varian Cary 3E UV-visible spectrophotometer equipped with a circulating water bath. 4.7.5.2 Kine t ic Analysis of the Reduction of U D P - G l c - 6 - C H O by the Cys260Ala Mutan t The kinetic constants for the reduction of UDP-Glc-6-CHO were obtained using the continuous assay method similar to that used for the oxidation of UDP-Glc-6-CHO. The Cys260Ala mutant (0.01 mg/mL) was added to a 500 pL solution of various concentrations of UDP-Glc-6-CHO and 0.5 mM NAD + in 50 mM Trien-HCl buffer with 2 mM DTT, pH 8.7 to initiate the reaction. The reactions were typically followed over a five-minute period. Rates were determined by following the increase in absorbance at 340 nm, at seven substrate concentrations ranging from 6.9 pM to 1.4 mM UDP-Glc-6-CHO. The kinetic constants (kcat and Km) were obtained using the identical method to that for the oxidation of UDP-Glc-6-CHO. 4.7.5.3 Effect of Preincubation with UDP-glucose on the Oxidat ion of U D P - G l c - 6 - C H O In the preincubation experiment, the initial incubation solution (500 pL in total) contained the Cys260Ala mutant (0.1 mg/mL), 3 mM NAD + , 0.26 mM UDP-glucose, 1 unit of diaphorase, 0.33 mM of />-iodonitrotetrazolium violet (INT) in 50 mM Trien-HCl buffer, pH 8.7. The solution was incubated at 30 °C and the reaction was followed at 500 nm (SINT = 12 800 M"1 cm"1) by a Varian Cary 3E UV-visible spectrophotometer equipped with a circulating water bath. After 10 minutes, UDP-Glc-6-CHO was added to the solution with a final concentration of 0.256 mM. The reaction was continuously monitored by increase in 142 absorbance at 500 nm. The reaction rate was calculated from the slope after UDP-Glc-6-CHO was added using a least squares analysis with Cary 3 software version 3.0. In the control experiment, the initial incubation solution contained the Cys260Ala mutant (0.1 mg/mL), 3 mM NAD + , 1 unit of diaphorase, and 0.33 mM of p-iodonitrotetrazolium violet (INT) in 50 mM Trien-HCl buffer, pH 8.7. It was incubated at 30 °C for 10 minutes before 0.256 mM UDP-Glc-6-CHO was added. The reaction was followed by monitoring the increase in absorbance at 500 nm. The reaction rate was calculated in the identical way as the preincubation experiment. 4.7.5.4 Deuterium "Washout" Experiment Deuterated UDP-glucose (UDP-(6,6-di-2H)glucose) was prepared as described in Section 2.7.9. Deuterated UDP-glucose (2 mM) was incubated with 20 mM NADH, 4 mM NAD + , 2 mM DTT and 0.1 mg Cys260Ala mutant in 2 mL 50 mM Trien-HCl buffer, pH 8.7, at 30 °C for 23 hours. The incubation solution was loaded onto a DE52 column (60 mL) eluting with a linear gradient of 0 - 300 mM ammonium bicarbonate buffer (800 mL in total volume). The eluents were assayed for UDP-sugar by incubation with wild type UDP-glucose dehydrogenase and NAD + . The fractions containing the UDP-sugar were pooled and evaporated under pressure to dryness. The remaining solid was redissolved in water and loaded onto a column of Amberlite ion-exchange resin (IR-120, in Na+ form) and eluted with water. The eluent was evaporated under pressure to dryness. The resulting solid was redissolved in water and loaded onto a Bio-gel P-2 column (2.5 x 45 cm) eluting with water. The eluent was evaporated under pressure to dryness and redissolved in D 20. The ! H NMR 143 spectra were recorded on Bruker WH-400 at 400 MHz. Low resolution liquid secondary ionization mass spectra were recorded on Kratos Concept II HQ Mass Spectrometer. 4.7.6 HPLC Analysis of the Reduction Products The sample contained 0.21 mM UDP-Glc-6-CHO, 0.188 mM NADH and the Cys260Ala mutant (0.12 mg/mL) in 50 mM triethanolamine-HCl buffer with 1 mM DTT, pH 8.7. The reaction mixture was incubated at 35 °C for 30 minutes. An aliquot of the sample was injected onto a Waters Radial-Pak 8NVC18 reversed-phase HPLC column pre-equilibrated with solvent A (Solvent A: 50 mM phosphate buffer, pH 7, with 2.5 mM TBAHS) and eluted with a gradient of 0 - 2 % solvent B in 18 minutes followed by 2 % - 30 % solvent B over 40 minutes (Solvent B: 50 mM phosphate buffer, pH 7, with 50 % (v/v) acetonitrile and 2.5 mM TBAHS). The eluent was monitored at 262 nm. 4.7.7 HPLC Analysis of UDP-Glc-6-CHO Oxidation Products The reaction mixture contained UDP-Glc-6-CHO (0.21 mM), NAD + (0.6 mM) and the Cys260Ala mutant (0.12 mg/mL) and was incubated at 30 °C for 30 minutes. The reaction sample was analyzed by HPLC in a fashion that was identical to the reduction reaction sample. Chapter Five Summary of Conclusions 145 All of the previous studies on bovine liver UDP-glucose dehydrogenase suggest that a thiol is involved in covalent catalysis. Our work on the enzyme from Streptococcus pyogenes also supports this notion. The dramatic activity loss of the Cys260Ser and Cys260Ala mutants indicates that the conserved Cys260 is the catalytically important thiol. The best evidence for covalent catalysis in the mechanism to date is the direct observation of a covalent adduct involving the mutated residue Ser260. The formation of an ester intermediate from the attachment of a UDP-sugar to the serine residue of the Cys260Ser mutant convincingly supports the involvement of a thioester intermediate in the mechanism employed by the wild type enzyme. Therefore we conclude that the wild type enzyme operates via covalent catalysis using Cys260. In the mechanism employed by the enzyme from Streptococcus pyogenes (Figure 5.1, X = SH), the C-6 hydroxyl of UDP-glucose is first oxidized to form an aldehyde intermediate and the first molecule of NADH. The aldehyde is tightly bound to the enzyme. The second oxidation is initiated by the addition of an active site Cys260 to the aldehyde to form a thiohemiacetal intermediate. The first molecule of NADH might dissociate at this point and a second NAD + could bind. A hydride transfer from the thiohemiacetal intermediate to the NAD + produces a covalently bound thioester intermediate and a second molecule of NADH. The above two oxidation steps are reversible. The final irreversible step, hydrolysis of the thioester, gives the product UDP-glucuronic acid. 146 Enzyme NAD+ OUDP UDP-glucose Enzyme T C H 2 HO \ H ^ O H X H O A ^ ^ ^ \ OH I OUDP Hydrated aldehyde l ^ N A D + Enzyme I , C H 2 X ^ NADH H O ^ \ ^ ^ ° v H O - - ^ ^ ^ OH I Aldehyde OUDP intermediate X=SH,OH Enzyme HO X _ C H 2 H A ^ NADH H O - - ^ r ~ ^ - - ' 0 x H O ^ ^ - ^ A OH I (Thio)hemiacetal OUDP intermediate NADH X=H NADH s N A D + H O ^ ^ ^ ^ ° \ H O - \ ^ ^ v A OH 1 OUDP UDP-glucuronic acid X=S I ' ^NADH Enzyme H 2 0 x=o v. slow H O ^ ^ ^ ^ ° \ H O - ~ ! ^ - ^ \ - ^ OH I OUDP (Thio)ester intermediate Figure 5.1 Mechanisms for the wild type enzyme and the Cys260Ser and Cys260Ala mutants. 147 In our solvent isotope incorporation study, the 1 80 isotope-induced chemical shift 13 changes in the C NMR carboxylate signals of the enzymatic product UDP-glucuronic acid indicates that only one solvent oxygen atom is incorporated. This observation is consistent with our proposed mechanism, and argues against the involvement of an imine intermediate unless the water molecule derived from C-6 alcohol is redelivered to hydrolyze the thioester intermediate. The observation that there was no primary kinetic isotope effect in the dehydrogenase reaction using UDP-(6,6-di- H)glucose as the substrate indicates that the two hydride transfer steps are not rate-limiting. The hydrolysis of the thioester intermediate is the best candidate for the rate-limiting step. The proposed mechanism is also consistent with the studies on bovine liver UDP-glucose dehydrogenase. Since UDP-glucose dehydrogenase from Streptococcus pyogenes shares high sequence identity with that from bovine liver as well as other bacterial sources, the proposed mechanism very likely applies to all of them. The Cys260Ser mutant can slowly catalyze both oxidation steps and a corresponding ester intermediate is formed (Figure 5.1, X = OH). However, the hydrolysis of the ester linkage is extremely slow, and therefore the adduct accumulates and can be detected by electrospray mass spectrometry. The mutant enzyme-adduct was subjected to pepsin digestion and the resulting peptide mixture was analyzed by various tandem electrospray mass spectroscopy techniques. The peptides bearing the ester adduct were identified by the technique of neutral loss tandem mass spectrometry. The peptide that was labeled was found to contain Ser260. Identical adducts were formed regardless of whether UDP-glucose or UDP-Glc-6-CHO were employed. 148 The technique of neutral loss tandem mass spectroscopy was very useful in the identification of the residue involved in the covalent catalysis. Our work has broadened the usage of this technique from analyzing acetal linkages in the case of glycosidases to analyzing any species linked to an enzyme via a normal ester linkage. This technique can be applied to other enzymes utilizing covalent catalysis. The covalent intermediate can be trapped by using mutated enzyme or mechanism-based inhibitors. This method is a rapid, sensitive and conclusive alternative to the environmentally unfriendly radioactive method. The observation that the Cys260Ala mutant readily oxidizes the aldehyde intermediate (UDP-Glc-6-CHO) indicates that in the absence of an active site nucleophile, the mutated dehydrogenase is capable of catalyzing the second oxidation step without the involvement of covalent catalysis. The oxidation is very likely to occur through the hydrated form of the aldehyde that resembles the thiohemiacetal intermediate (Figure 5.1, X = H). The observation that the Cys260Ala mutant does not appreciably catalyze the oxidation of UDP-glucose could be explained by postulating that the aldehyde intermediate is tightly bound in the mutant active site and there is no mechanism by which it can be hydrated and proceed forward in the second oxidation step. The X-ray crystal structure of the Streptococcus pyogenes UDP-glucose dehydrogenase with UDP-xylose and NAD + bound was solved by Robert Campbell in our laboratory (personal communication) by the time this thesis had been written. Further investigation of the mechanism of this enzyme will involve mutating the conserved amino acid residues that are located at the active site of the enzyme and studying the kinetics of the mutated enzyme. The candidate residues that might be responsible for delivering the water molecule during the thioester hydrolysis (possibly a glutamate residue) are the best targets to 149 be explored. If the amino acid residue that is involved in the water delivery is mutated to an alanine, the thioester intermediate will accumulate and could be detected by electrospray mass spectrometry. 150 References (1) Fersht A., Enzyme Structure and Mechanism, 2" Ed., W. H. Freeman and Company, 1985, p 390 (2) Oppenheimer, N. J.; Handlon, A. L. "Mechanism of NAD-Dependent Enzymes", The Enzymes, 1992, Vol. XX, p 453 (3) Rossmann, M.G.; Moras, D.; Olsen, K. W. Nature 1974, 250, 194 (4) Wierenga, R. K.; Ho, W. G. J. Nature 1983, 302, 842 (5) Westheimer, F. H. "Mechanism of Action of the Pyridine Nucleotides", Pyridine Nucleotide Coenzymes, Dolphin, D.; Poulson, R.; Avramovic, O. Ed., Wiley: New York, 1987, p 255 (6) Westheimer, F. H.; Fisher, H.; Conn, E. E.; Vennesland, B. J. Am. Chem. Soc. 1951, 73, 2043 (7) Fisher, H. F.; Conn, E. E.; Vennesland, B.; Westheimer, F. H. J. Biol. Chem. 1953, 202, 687 (8) You, K. Methods in Enzymology 1982, 87, 101 (9) Ohno, A.; Ushio, K.; "Redox Reactions of Pyridine Nucleotides", Pyridine Nucleotide Coenzymes, Dolphin, D.; Poulson, R.; Avramovic, O. Ed., Wiley: New York, 1987, p 276 (10) Pettersson, G. Crit. Rev. Biochem. 1987, 21, 349 (11) Persson, B.; Krook, M.; Jornvall, H. Eur. J. Biochem. 1991, 200, 537 (12) Jornvall, H.; Persson, B.; Krook, M.; Atrian, S.; Gonzalez-Duarte, R.; Jeffery, J.; Ghosh, D. Biochemistry 1995, 34, 6003 151 (13) Branden, C. -I.; Jornvall, H.; Eklund, H.; Furugren, B. "Alcohol dehydrogenases" The Enzymes, Vol. 1 IA, 3rd Ed., Academic Press: London, 1975, pl03 (14) Eklund, FL; Branden, C. -I.; Jornvall, H. J. Molec. Biol. 1976,102, 61 (15) Eklund, FL; Nordstrom, B.; Zeppezauer, E.; Soderlund, G.; Ohlsson, I.; Bowe, T.; Branden, C. -I. FEBS Lett. 1974, 44, 200 (16) Eklund, FL, Plapp; B. V.; Samama, J. -P.; Branden, C. -I. J. Biol. Chem. 1982, 257, 14349 (17) Cedergren-Zeppezauer, E.; Samama, J. -P.; Eklund, H. Biochemistry 1982, 21, 4895 (18) Kvassman, J.; Larsson, A.; Pettersson, G. Eur. J. Biochem. 1981,114, 555 (19) Cook, P. F.; Cleland, W. W. Biochemistry 1981, 20, 1805 (20) Wratten, C. C ; Cleland, W. W. Biochemistry 1965, 4, 2442 (21) Shore, J. D.; Gutfreund, FL; Yates, D. J. Biol. Chem. 1975, 250, 5276 (22) Andersson, P.; Kvassman, J.; Linstrom, A.; Olden, B.; Pettersson, G. Eur. J. Biochem. 1981,113, 425 (23) Koivusalo, M.; Baumann, M.; Uotila, L. FEBS Lett. 1989, 257, 105 (24) Conway, A.; Koshland, D. E. Jr. Biochemistry 1968, 7, 4011 (25) Gennis, L. S. Proc. Natl. Acad. Sci. (USA), 1976, 73, 3928 (26) Moras, D.; Olaen, K. W.; Sabesan, M. N.; Buehner, M.; Ford, G. C ; Rossmann, M. G. J. Biol. Chem. 1975, 250, 9137 (27) Biesecker, G.; Harris, J. I.; Thierry, J. C ; Walker, J. E.; Wonacott, A. J. Nature 1977, 266, 328 (28) Trentham, D. R. Biochem. J. 1971,122, 71 152 (29) Soukri, A.; Mougin, A.; Corbier, C ; Wonacott, A.; Branlant, C ; Branlant, B. Biochemistry 1989, 28, 2586 (30) Walker, J. E.; Carne, A. F.; Runswick, M. J.; Bridgen, J.; Harris, J. I. Eur. J. Biochem. 1980, 108, 549 (31) Liu, Z.; Sun, Y.; Rose, J.; Chung, Y.; Hsiao, C ; Chang, W.; Kuo, I.; Perozich, J.; Lindahl, R.; Hempel, J.; Wang, B. Nature Structural Biology 1991,4, 317 (32) Hurley, T. D.; Steinmetz, C. G.; Weiner, H. Enzymology-and Molecular Biology of Carbonyl Metabolism 7, Weiner, H. Ed., 1998, Plenum Press, pi5 (33) Moore, S. A.; Baker, H. M.; Blythe T. J.; Kitson, K. E.; Kitson, T. M.; Baker, E. N. Enzymology and Molecular Biology of Carbonyl Metabolism 7, Weiner, H. Ed., 1998, Plenum Press, p27 (34) Hempel, J.; Nicholas, H.; Lindahl, R. Protein Science 1993, 2, 1890 (35) Perozich, J.; Nicholas, H.; Wang, B.; Lindahl, R.; Hempel, J. Protein Science 1999, 8, 137 (36) Kitson, T. M.; Hill, J. P.; Midwinter, G. G. Biochem. J. 1991, 275, 207 (37) Pietruszko, R.; Ariola, D. P.; Blatter, E. E.; Mukerjee, N. Enzymology and Molecular Biology of Carbonyl Metabolism 4, Weiner, H., Ed., 1993, Plenum Press, p 221 (38) Farres, J.; Wang, T. T. Y.; Cunningham, S. J.; Weiner, H. Biochemistry 1995, 34, 2592 (39) Farres, J.; Wang, X.; Takahashi, K.; Cunningham, S. J.; Wang, T. T.; Weiner, H. J. Biol. Chem. 1994, 269, 13854 (40) Weiner, H.; Hu, J. H. J.; Sanny, C. G. J. Biol. Chem. 1976, 251, 3853 (41) Vallari, R. C ; Pietruszko, R. Arch. Biochem. Biophys. 1981, 212, 9 153 (42) Feldman, R. I.; Weiner, H. J. Biol. Chem. 1972, 247, 260 (43) Wang, X.; Weiner, H. Biochemistry 1995, 34, 237 (44) Vedadi, M.; Szittner, R.; Smillie, L.; Meighen, E. Biochemistry 1995, 34, 16725 (45) Eisses, K. T. Bioorg. Chem. 1989,17, 268 (46) Hara, S.; Yamamoto, H.; Deyashiki, Y.; Nakaya, T.; Oritani, H.; Sawada, H. Biochim. Biophys. Acta 1991,1075, 61 (47) Kendal, L. P.; Ramanthan, A. N. Biochem. J. 1952, 52,430. (48) Hinson, J. A.; Neal, R. A. Biochim. Biophys. Acta 1975, 384, 1 (49) Abeles, R. FL; Lee, H. A. J. Biol. Chem. 1960, 235, 1499 (50) Shearer, B. L.; Kim, K.; Lee, K. M.; Wang, C. K.; Plapp, B. V. Biochemistry 1993, 32, 11186 (51) Henehan, G. T. M.; Oppenheimer, N. J. Biochemistry 1993, 32, 735 (52) Olson, L. P.; Luo, J.; Almarsson, O.; Bruice, T. C. Biochemistry 1996, 35, 9782 (53) Feingold, D. S.; Franzen, J. S. Trends Biochem. Sci. 1981, 6, 103 (54) Dougherby,B. A.; van de Rijn, I. J. Biol. Chem. 1993, 268, 7118 (55) Arrecubieta, C.; Garcia, E.; Lopez, R. J. Bacteriol. 1996,178, 2971 (56) Sieberth, V.; Gordon, G. P.; Robers, I. S.; Jann, K. J. Bacteriol. 1995,777, 4562 (57) Wessels, M. R.; Goldberg, J. B.; Moses, A. E.; DiCesare, T. J. Infect. Immun. 1994, 62, 433 (58) Watson, D. A.; Musher, D. M. Infect. Immun. 1990, 58, 3135 (59) Cross, A. S. Curr. Top. Microbiol. Immunol 1990,150, 87 (60) Moxon, E. R.; Kroll, J. S. Curr. Top. Microbiol. Immunol. 1990,150, 65 154 (61) Crater, D.; Dougherty, B.; van de Rijn, I. J. Biol. Chem. 1995, 270,28676 (62) Veasy, L.; Wiedmeier, S.; Thompson, J. et al N. Engl. J. Med. 1987, 316, 421 (63) Voet, D.; Voet, J. G. Biochemistry, Wiley, 1990, p 258 (64) Pouyani, T.; Harbison, G. S.; Prestwich, G. D. J. Am. Chem. Soc. 1994,116, 7515 (65) Nerbert, D. W.; Gonzalez, F. J. Annue. Rev. Biochem. 1987, 56, 945 (66) Binari, R. C ; Staveley, B. E.; Johnson, W. A.; Godavarti, R.; Sasisekharan, R.; Manoukian, A. S. Development 1997,124, 2623 (67) Lindahl, U.; Feingold, D. S.; Roden, L. Trends Biochem. Sci. 1986,11, 221 (68) Strominger, J. L.; Kalchar, H. M.; Axelrod, J.; Maxwell, E. S. J. Am. Chem. Soc. 1954, 75,6411 (69) Zalitis, J.; Feingold, D. S. Arch. Biochem. Biophys. 1969,132, 457 (70) Dougherty, B. A.; van de Rijn, I. J. Biol. Chem. 1993, 268, 7118 (71) Hempel, J.; Perozich, J.; Romovacek, H.; Hinich, A.; Kuo, I.; Feingold, D. S. Protein Science 1994, 3, 1074 (72) Spicer, A. P.; Kaback, L. A.; Smith, T. J.; Seldin, M. F. J. Biol. Chem. 1998, 273, 25117 (73) Strominger, J. L.; Maxwell, E. S.; Axelrod, J.; Kalckar, H. M. J. Biol. Chem. 1957, 224, 79 (74) Nelsestuen, G. L.; Kirkwood, S. J. Biol. Chem. 1971, 246, 3828 (75) Ridley, W. P.; Kirkwood, S. Biochem. Biophys. Res. Commun. 1973, 54, 955 (76) Ridley, W. P.; Houchins, J. P.; Kirkwood, S. J. Biol. Chem. 1975, 250, 8761 (77) Gainey, P. A.; Pestell, T. C ; Phelps, C. F. Biochem. J. 1972,129, 821 (78) Franzen, J. S.; Ishman, R.; Feingold, D. S. Biochemistry 1976,15, 5665 155 (79) Franzen, J. S.; Marchetti, P. S.; Feingold, D. S.; Biochemistry 1980,19, 6080 (80) Franzen, B.; Carrubba, C ; Feingold, D. S.; Ashcom, J.; Franzen, J. S. Biochem. J. 1981, 199, 603 (81) Franzen, J. S.; Marchetti, P.; Ishman, R.; AshcomJ.; Feingold, D. D. Biochem. J. 1978, 173, 701 (82) Ordman, A. B.; Kirkwood, S. J. Biol. Chem. 1977, 252, 1320 (83) Schiller, J. G.; Bowser, A. M.; Feingold, D. S. Carhohydr. Res. 1972, 25, 403 (84) Ordmann, A. B.; Kirkwood, S. Biochim. Biophys. Acta 1977, 481, 25 (85) Ridley, W. P.; Kirkwood, S. Biochem. Biophys. Res. Commun. 1973, 54, 955 (86) Franzen, J. S.; Kuo, I.; Eichler, A. J.; Feingold, D. S. Biochem. Biophys. Res. Commun. 1973,50,517 (87) Franzen, J. S.; Ishman, R.; Feingold, D. S. Biochemistry 1976, 75, 5665 (88) MacQuarrie, R. A.; Bernhard, S. A. Biochemistry 1971,10, 2456 (89) Stallcup, W. B.; Koshland, Jr. D. E. Biochem. Biophys. Res. Commun. 1972, 49, 1108 (90) Eccleston, E. D.; Thayer, M. L.; Kirkwood, S. J. Biol. Chem. 1979, 254, 11399 (91) Schiller, J. G.; Bowser, A. M.; Feingold, D. S. Biochim. Biophys. Acta 1973, 293, 1 (92) Schiller, J. G.; Lamy, F.; Frazier, R.; Feingold, D. S.; Biochim. Biophys. Acta 1976, 453, 418 (93) Neufeld, E. F.; Hall, C. W. Biochem. Biophys. Res. Commun. 1965,19, 456 (94) Adams, E. J. Biol. Chem. 1955, 217, 325 (95) Grubmeyer, C. T.; Gray, W. R. Biochemistry 1986, 25, 4778 (96) Burger, E.; Gorisch, H. Eur. J. Biochem. 1981,116, 137 156 (97) Davies, D.; Teixeira, A.; Kenworthy, P. Biochem. J. 1972,127, 335 (98) Grubmeyer, C. T.; Insinga, S.; Bhatia, M.; Moazami, N. Biochemistry 1989, 28, 8174 (99) Grubmeyer, C.; Skiadopoulos, M.; Senior, A. E. Arch. Biochem. Biophys. 1989, 272, 311 (100) Nagai, A.; Ohta, D. J. Biochem. 1994,115, 22 (101) Teng, H.; Grubmeyer, C. Biochemistry 1999, 38, 7363 (102) Teng, H.; Segura, E.; Grubmeyer, C. J. Biol. Chem. 1993, 268, 14182 (103) Nagai, A.; Kheirolomoom, A.; Ohta, D. J. Biochem. 1993,114, 856 (104) Warren, W. A. J. Biol. Chem. 1970, 245, 1675 (105) Corbier, C.; Seta, F. D.; Branlant, G. Biochemistry 1992, 31, 12532 (106) De Luca, C.; Lansing, M.; Crescenzi, F.; Martini, I.; Shen, G.-J.; O'Regan, M.; Wong, C.-H. Bioorg. & Med. Chem. 1996, ¥,131 (107) Arrecubieta, C.; Lopez, R.; Garcia, E. J. Bacteriol. 1994,176, 6375 (108) Dillard, J. P.; Vandersea, M. W.; Yother, J. J. Exp. Med. 1995,181, 973 (109) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S.F.; Whitehouse, C. M. Science 1989, 246, 46 (110) Snyder, A. P. Biological and Biotechnological Applications of ESI-MS, American Chemical Society, Washington, DC, 1996 (111) Smith, R. D.; Light-Wahl, K. J. Biol. Mass Spectrom. 1993, 22, 493 (112) Campbell, R. E.; Sala, R. F.; van de Rijn, L; Tanner, M. E. J. Biol. Chem. 1997, 272, 3416 (113) Burger, E.; Gorisch, H. Eur. J. Biochem. 1981,118, 125 157 (114) Grubmeyer, C. T.; Insinga, S. J. Am. Chem. Soc. 1990,112, 5906 (115) Batiz-Hernandez, H.; Bernheim, R. A. Progress in Nuclear Magnetic Resonance Spectroscopy 1967, 3, 63 (116) Risley, J. M.; Van Etten, R. L. J. Am. Chem. Soc. 1979,101, 252 (117) Risley, J. M.; Van Etten, R. L. J. Am. Chem. Soc. 1980,102, 6699 (118) Risley, J. M.; Van Etten, R. L. J. Am. Chem. Soc. 1981,103, 4389 (119) Mega, T. L.; Van Etten, R. L. J. Am. Chem. Soc. 1988,110,6312 (120) Risley, J. M.; Van Etten, R. L. J. Am. Chem. Soc. 1980,102, 4609 (121) Diakur, J.; Nakashima, T. T.; Vederas, J. C. Can. J. Chem. 1980, 58, 1311 (122) Jones, J. B.; Taylor, K. E. Can. J. Chem. 1976, 54, 2969 (123) Karas, M.; Hillenkamp, F. Anal. Chem. 1988, 60, 2299 (124) Campbell, R. E.; Tanner, M. E. Angew. Chem. Intl. Ed. Engl. 1997, 36, 1520 (125) Richards, J. H. The Enzymes, Vol. 2, Boyer, P. Ed, Academic Press: New York, 1992, p 321 (126) Walsh, C. Enzymatic Reaction Mechanisms, W. H. Freeman and Co.: San Francisco, p 109 (127) Snetkova, E. V.; Akulov, G. P.; Gordeeva, L. S.; Kaminskii, Yu. L. Khimiya Prirodnykh Soedinenii 1987, 125 (128) Moss, G. Arch. Biochem. Biophys. I960, 90, 111 (129) Roseman, S.; Distler, J. J.; Moffatt J. G.; Khorana, H. G. J. Am. Chem. Soc. 1961, 83, 659 (130) Bradford, M. M. Anal. Biochem. 1976, 72, 248 158 (131) Ausubel, F. M.; Brent, R.; Kingston, R. E.; Moore, D. D.; Seidman, J. G.; Smith, J. A.; Stuhl, K. Short Protocols in Molecular Biology (A compendium of Protocols from Current Protocols in Molecular biology), Green Publishing Associates and John Wiley and Sons, New York, 1992. (132) Bruins, A.P.; Covey, T.R.; Henion, D. Anal. Chem. 1987, 59, 2642 (133) Friebolin, H. Basic One- and Two-Dimentional NMR Spectroscopy, second, enlarged edition, VCH: New York, 1993, p 90 (134) Tu, G. C ; Weiner, H. J. Biol. Chem. 1988, 263, 1212 (135) Tu, G. C , Weiner, FL, J. Biol. Chem. 1988, 263, 1218 (136) Ghenbot, B.; Weiner, H. FASEB J. 1991, 5, A417 (137) Hughes, M. J. G.; Andrews, D. W. BioTechniques 1996, 20, 188 (138) Miao, S., Ziser, L.; Aebersold, R.; Withers, S. G. Biochemistry 1994, 33, 7027 (139) Withers, W. G.; Aebersold, R. Protein Science 1995, 4, 361 (140) Mackenzie, L. F.; Brooke, G. S.; Cutfield, J. F.; Sullivan, P. A.; Withers, S. G. J. Biol. Chem. 1997,272,3161 (141) Zechel, D. L.; He, S.; Dupont, C ; Withers, S. G. Biochem. J. 1998, 336, 139 (142) Bruins, A. P.; Covey, T. R.; Henion, J. D. Anal Chem. 1987, 59, 2642 (143) Hess, D.; Covey, T. C ; Winz, R.; Brownsey, R.; Aebersold, R. Protein Science 1993, 2, 1342 (144) Meynial, I.; Paquet, V.; Combes, D. Anal. Chem. 1995, 67, 1627 (145) Paladini, A.; Leloir, L. F. J. Biochem. 1952, 51, 426 159 (146) Kochetkov, N. K.; Shibaev, V. N. Advances in Carbohydrate Chemistry and Biochemistry 1973, 28, 307 (147) Comisarow, M. B.; Marshall, A. G. Chemical Physics Letters 1974, 25, 282 (148) Senko, M. W.; Hendrickson, C. L.; Pasa-Tolic, L.; Marto, J. A.; White, F. M.; Guan, S.; Marshall, A. G. Rapid Commun. Mass Spectrom. 1996,10, 1824 (149) Senko, M. W.; Hendrickson, C. L.; Emmett, M. R.; Shi, S. D.-H.; Marshall, A. G. J. Am. Soc. Mass. Spectrom. 1997, 8, 970 (150) Emmett, M. R.; white, F. M.; Hendrickson, C. L.; Shi, S. D.-H.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 1998, 9, 333 (151) Frimpong, K.; Rodwell, V. W. J. Biol. Chem. 1994, 269, 11478 (152) Fersht, A. Enzyme Structure and Mechanism, , W.H. Freeman and Company, New York, 1987, pp 405 (153) Brenner, S. Nature 1988, 334, 528-530 (154) Gorbalenya, A. E.; Donchenko, A. P.; Blinov, V. M.; Koonin, E. V. FEBS lett. 1989, 243, 103-114 (155) Thompson, S.; Mayerl, F.; Peoples, O. P.; Masamune, S.; Sinskey, A. J.; Walsh, C. T. Biochemistry 1989, 28, 5735-5742 (156) Karsten, W. E.; Viola, R. E. Biochim. Biophys. Acta 1992,1121, 234-238 (157) Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry, Part A, Plenum Press, New York, 1993, Third Edition, p 288 160 Appendix 161 A: Sequence Alignment UDPNAMDH St-Hypo SA-UDPGDH GDPMDH Bov-UDPGDH MSFAT MFGIDEVK MK MR MFEIKK 1LPTRAFA-SRQKQ-@LPL-AVEFGKSRQ-SL-GVLLSLQNE-VCAGCLSARGHE-HGPTCSVIAHMCPEIF SIG J^INQHAVDTI^ RGEIHJ;VEPD£ASWKTAVEGGFLRASTTP-- 72 VGF J VNKKRILELKNG--•-VDVNSETTEEELREARYLKFJSEIEK 73 riVIILPSKVDKIgNgLSPiQDEYIEyYLKSKQLS--IKA&DSKA 69 5IGVB VSSTKIDLINQgKSPiVEPG£'EALLQQGRQTGRLSGCTDFKK 72 INESRINAWlSPTLPiYEPGiiKEWESCRGKNLFFSS-NIDD 77 UDPNAMDH --VEADRWlJJA^Ff St-Hypo I-KECNFYIH SA-ODPGDH AYKEAELVl| GDPMDH AVLDSDVS.M. Bov-UDPGDH AIKEADLVFQS: |FKGD-INTY-NYNS--RINYF K K N G D L KTYGMGKGRAA TYVESAARSIAPV — LKKGAUJILE ILTPLIKASETVGTV- - -LNRGDI.^VYE jTQHVETVIKEVLSV NSHATLIIK GYIETVCREIGFAIREKSERHT^WR ILKYIEACARRIVQ- --NSHGYKI^TEK USPVgSTEKMAEWLAEMRPDLTFP 142 /YPGCTEEECVPILARMSGMTFN 144 JIPIGFITEMRQKFQTD 134 /LPQTVNNWIPLIEDCSGKKAG 147 /PVRAAESIRRIFDANTKPNLNL 154 UDPNAMDH QQVGEQADVNIAYCJj St-Hypo QDF YVGYS SA-UDPGDH RIIFSO GDPMDH VDF GVGTNC Bov-UDPGDH Q VS l | EVLPGQVMVELIKNDRVIGgMT PVCSARASELY-KIFLEGEC- - WTNSRTi INPGDKKHRLTNIKKITSgST AQIAELIDEVYQQIISAGTY--KAESIKVi LRESKALYDNLYPSRIIVSCEENDSPKVKADAEKFALLLKSAAKKNNVPVLIMGASEi LRESTAIKDYDFPPMTVIG ELDKQTGDLLEEIYRELDAPIIRKTVEV; FLAEGTAIKDLKNPDRVLIGGDETPEGQRAVQA--LCAVYEHWVPREK--ILTTNTWSSl UDPNAMDH TEfflSFRDVNIAFApLSLICADQGINVWELIRLANRHPRVNI- -LQPG St-Hypo IE TORDLNI ALVffiJLAI IFNRLNIDTEAVLRAAGSKWNFLP FR SA-UDPGDH FA TYLALRVAYFfflELDTYAESRKLNSHMIIQGISYDDRIGMHYNNPS GDPMDH TC VWHAAKVTFApiGNIAKAVGVDGREVMDVICQDHKLNLSRYYMRPj Bov-UDPGDH TA AFLAQRISSIHSISALCEATGADVEEVATAIGMDQRIGNKFLKAS I P W F I V - - A Q N P Q Q A — R L 283 (PYY&T- - H K S Q G I G Y Y P E I 282 K Q | L - - A H Y N N I P Q - - T L 279 ' R A | T Y R A S Q L D V E H - - P M 289 L N ^ V Y L C E A L N L P E V A R Y 298 UDPNAMDH St-Hypo SA-UDPGDH GDPMDH Bov-UDPGDH IRTAR1 ILAGRR: IEAIVSSjj LGSLMRSfi WQQVID1 H-KPFWVIDQVKAAVADCLAATDKRASELKIACFgLi |DN-MGNYVSEQLIKAMIKKGINVEGS SVLILgFTJ -RKSYIAKQIINVLKEQESPVK WGVYRLI1 ISNQ VQK-AFD-LITSHDTRK VGLLgLS, | D Y Q R R R F A S R - I I D S L - F N T V T D K KIAILgFAE PNIDf ENCP^ll SNSDN1 ,GTDj|i IKDTG:' ^PAMEIAELIAQWHSGETLWESN 362 TTRIIDWKE|GKY-SCKVD|FD^W 356 MKDVIDIIKS-KDIKII|YE^M 351 |PLVELAEM1IG-KGYEFR3:;FDRN 355 pJSSIYISKYlsMD- EGAHLH§YDi*K 369 UDPNAMDH IH-QLPKKLTGLWYSGAA 379 St-Hypo VDAEEVRREYGIIPVSEVKSSHYDA-IIVAVGHQQFKQMGSEDIRGFGKDKHVLYDLKYVLPAEQSDVRL 425 SA-UDPGDH LNKLESEDQ--SVLVNDLENFKKQANIIVTNRYDNELQDVKNKVYSRDIFGRD 402 GDPMDH VEYARVHGANKEYIESKIPHVSS— LLVSDLDEWA-SSDVLVLGNGDELFVDLVNKTPSGKKLVDLVGFMPHTTTAQA 431 Bov-UDPGDH VPREQIWDLSHPGVSKDDQVAR---LVTISKDPYEACDGAHAWICTEWDMFKELDYERIHKKMLKPAFIFDGRRVLDG 446 UDPNAMDH 379 St-Hypo 425 SA-UDPGDH 402 GDPMDH EGICW 436 Bov-UDPGDH LHNELQTIGFQIETIGKKVSSK 468 Figure A . l Multiple alignment of UDP-glucose dehydrogenase from Streptococcus pyogenes (SA-UDPGDH) with UDP-glucose dehydrogenase from bovine liver (Bov-UDPGDH), GDP-mannose dehydrogenase from Pseudononas (GDPMDH), UDP-N-acetylmanno-saminuronic acid dehydrogenase from E. coli (UDPNAMDH) and a hypothetical sequence from Salmonella (St-Hypo). Strictly conserved residues are given in reverse-background and identities in 4 of the 5 sequences are hashed. (Reproduced from Hempel, J.; Perozich, J.; Romovacek, H.; Hinich, A.; Kuo, I.; Feingold, D. S. Protein Science 1994, 3, 1074) 162 B: Graphical Representation of Kinetic Data Wild type UDP-glucose dehydrogenase A: Direct assay E i E 0 . 0 2 0 . 0 4 0 . 0 6 0 .08 0.1 0 . 1 2 0 . 1 4 0 . 1 6 [UDPG] m M 0 .2 0 . 4 [NAD] m M B: Coupled assay Figure A.2 Direct plots of kinetic data for the wild type dehydrogenase reaction. Conditions: 50 mM Trien-HCl buffer, pH 8.7, 30 °C. (A) [UDPGDH] = 0.22 pM, [DTT] = 2 mM. The fixed concentration of the second substrate (either UDP-glucose or NAD+) was 0.5 mM. (B) [UDPGDH] = 0.22 pM, [NAD+] = 0.5 mM (with variable [UDPG]), [UDPG] = 0.2 mM (with variable [NAD+]). 163 Cys260Ala mutant A: Oxidation of UDP-Glc-6-CHO 0 0 . 2 0 . 4 0 .6 0 .8 1 1.2 1.4 1.6 1.8 2 [ U D P - G l c - 6 - C H O ] m M B: Reduction of UDP-Glc-6-CHO 0 .2 0 . 4 0 .6 0 .8 [ U D P - G l c - 6 - C H 0 ] m M Figure A.3 Direct plots of kinetic data for the dehydrognease reactions catalyzed by the Cys260Ala mutant. Conditions: 50 mM Trien-HCl buffer, pH 8.7, [DTT] = 1 mM, 30 °C. (A) [Cys260Ala] = 1.1 pM, [NAD+] = 0.5 mM. (B) [Cys260Ala] = 0.22 pM, [NADH] = 0.17 mM. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0061493/manifest

Comment

Related Items