Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mechanistic investigation of 3, 5-epimerases / reductases involved in the biosynthesis of GDP-L-Fucose… Lau, Tak Bun Stephen 2009

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

Item Metadata

Download

Media
24-ubc_2009_fall_lau_takbun.pdf [ 3.7MB ]
Metadata
JSON: 24-1.0061230.json
JSON-LD: 24-1.0061230-ld.json
RDF/XML (Pretty): 24-1.0061230-rdf.xml
RDF/JSON: 24-1.0061230-rdf.json
Turtle: 24-1.0061230-turtle.txt
N-Triples: 24-1.0061230-rdf-ntriples.txt
Original Record: 24-1.0061230-source.json
Full Text
24-1.0061230-fulltext.txt
Citation
24-1.0061230.ris

Full Text

Mechanistic Investigation of 3, 5-Epimerases / Reductases Involved in the Biosynthesis of GDP-L-Fucose and GDP-L-Galactose  by  Tak Bun Stephen Lau B.Sc., University of British Columbia, 2003  A THESIS SUBMIEFED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES  (Chemistry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  June 2009 © Tak Bun Stephen Lau, 2009  ABSTRACT  GDP-fucose synthase (GFS), also known as GDP-4-keto-6-deoxy-mannose 3, 5epimerase / 4-reductase (GMER), is the key enzyme in the GDP-L-fucose biosynthetic pathway. The pathway begins with a dehydratase that converts GDP-D-mannose into 6-deoxy-4-ketoGDP-D-mannose. Then GFS performs two consecutive epirnerizations, and a final NADPH dependent reduction to generate GDP-L-fucose. The detailed mechanism of GFS has remained a mystery due to limited mechanistic studies as well as the lack of an x-ray crystal structure with  any bound GDP-sugar. This thesis investigates the mechanism of the reaction as well as the roles of the active site residues in GFS. The results support a sequential ordered epimerization mechanism where the stereocenter at C-3”is inverted first, followed by the inversion of stereochemistry at C-5”. This identifies GDP-6-deoxy-D-altrose as a reaction intermediate for the first time. The findings of this thesis also leads to the assignment of the catalytic residue Cys 109, as the base which deprotonates both epimerization sites, and Hisi 79 as the acid which reprotonates the corresponding enol intermediates.  GDP-mannose 3, 5-epimerase (GME), is the key enzyme that catalyzes the first committed step of vitamin C biosynthesis in plants. It oxidizes the C-4” position of GDP mannose, then sequentially inverts the stereochemistry at C-5” and C-3”, and finally reduces the C-4” carbonyl to produce GDP-L-galactose. The proposed sequential order of epimerization is based on the observation that GDP-L-gulose (the C-5” epimeric product) is produced whereas GDP-D-altrose (the C-3” epimeric product) is not. Structural studies on GME suggest that Cys145 acts as the catalytic base and Lys 217 acts as the catalytic acid in both epimerization  ‘II  steps. However, very limited mechanistic studies on GME have been performed. This thesis provides mechanistic evidence that supports the proposed sequential epimerization order as well as the assignment of roles for the catalytic residues. It also suggests a model whereby protonated Cys 145 is unable to exchange its proton with solvent during the lifetime of the enol(late) intermediate.  iv  TABLE OF CONTENTS ABSTRACT  ii  TABLE OF CONTENTS  iv  LIST OF TABLES  viii  LIST OF FIGURES  ix  LISTS OF SYMBOLS AND ABBREVIATIONS  xv  GLOSSARY  xix  ACKNOWLEDGEMENT  xxii  DEDICATION  xxiv  Chapter 1  Sugar Nucleotide Epimerases  1  1.1  Introduction  2  1.2  Dextro / Levo and a /  1.3  Sugar Nucleotides  7  1.4  Epirnerization at Activated and Unactivated Centers  8  1.5  Short-chain Dehydrogenase Reductase (SDR) Superfamily  9  1.6  Different Strategies used by Sugar Nucleotide Epimerases  11  1.6.1  Direct Oxidation / Reduction Mechanism  11  1.6.2  Glycal Intermediate Mechanism  19  1.6.3  Activated Substrate Epimerization Mechanism  22  1.6.4  Activated Substrate Epimerization / Reduction Mechanism  25  1.6.5  Oxidation / Epimerization / Reduction Mechanism  30  13 Assignment in Monosaccharides  4  1.7  GDP-L-Fucose and its Importance in Mammals and Bacteria  35  1.8  The Role of GDP-Mannose 3,5-Epimerase in Plants  37  V  1.9  Project Goals  .40  Mechanistic Studies with Wild Type GDP-L-Fucose Synthase  42  2.1  Introduction  43  2.2  Cloning, Overexpression, and Purification of GMD and GFS  44  2.3  Enzymatic Synthesis and Purification of GDP-L-Fucose  48  2.3.1  Synthesis of GDP-L-Fucose Using NADPH  48  2.3.2  Synthesis of GDP-L-Fucose Using NADH  50  2.3.3  Enzymatic Synthesis and Revised Purification of GDP-L-Fucose  52  Chapter 2  2.4  Deuterium Wash-in Experiment with GDP-L-Fucose  55  2.5  Kinetic Study of GFS  57  2.5.1  Kinetic Study with C-5” Deuterated Substrate  58  2.5.2  Kinetic Study with C-3” Deuterated Substrate  61  GDP-2-Deoxy-2-Fluoro-D-Mannose as a Mechanistic Probe  65  2.6.1  Introduction and the Experimental Design  65  2.6.2  Chemical Synthesis of GDP-2-Deoxy-2-Fluoro-D-Mannose  69  2.6.3  Reaction of GDP-2-Deoxy-2-Fluoro-D-Mannose with GMD  72  2.6  2.7  Conclusions  73  2.8  Experimental  74  2.8.1  Materials and General Methods  74  2.8.2  Cloning, Overexpression, and Purification of GMD and GFS  75  2.8.3  Enzymatic Synthesis of GDP-6-Deoxy-4-Keto-D-Mannose  77  2.8.4  Enzymatic Synthesis of GDP-L-Fucose  77  2.8.5  Deuterium Wash-in with GDP-L-Fucose  78  vi  Chapter 3  2.8.6  Kinetic Study of GFS  .79  2.8.7  Synthesis of GDP-2-Deoxy-2-Fluoro-D-Mannose  81  2.8.8  Reaction of GDP-2-Deoxy-2-Fluoro-D-Mannose with GMD  85  Mechanistic Studies with GDP-L-Fucose Synthase Mutants  86  3.1  Introduction  87  3.2  Site-directed Mutagenesis of GDP-L-Fucose Synthase  89  3.3  Kinetic Analysis of Mutant Enzymes  90  3.4  Characterization of the CyslO9Ser Mutant  91  3.4.1  Product Analysis with the CyslO9Ser Mutant  91  3.4.2  Deuterium Wash-in Experiment with GDP-L-Fucose  98  3.4.3  Kinetic Isotope Effect Studies with the CyslO9Ser Mutant  3.5  101  Characterization of the Hisl79Gln Mutant  103  3.5.1  Deuterium Wash-in Experiment with GDP-L-Fucose  103  3.5.2  Deuterium Wash-out Experiment with Deuterated Substrates  105  3.6  Future Research  107  3.7  Conclusions  109  3.8  Experimental  115  Site-directed Mutagenesis, Overexpression, and Purification of Mutants  115  3.8.2  Product Analysis with CyslO9Ser and Hisl79Gln  116  3.8.3  Synthesis of 6-Deoxy-GDP-D-Mannose  116  3.8.4  Deuterium Wash-in Experiments with GDP-L-Fucose  117  3.8.5  Kinetic Studies of CyslO9Ser  118  3.8.6  Deuterium Wash-out Experiment with His 179Gm  119  3.8.1  VII  Chapter 4  Mechanistic Studies with GDP-Mannose-3, 5-Epimerase  120  4.1  Introduction  121  4.2  Deuterium Wash-in I Wash-out experiments  123  4.2.1  Deuterium Wash-in Experiment with GDP-L-Fucose  123  4.2.2  Deuterium Wash-out Experiment with [3”H]- or [5”2 Hj-GDP-L2 Fucose  127  4.2.3  Deuterium Wash-in Experiment with GDP-6-Deoxy-D-Mannose  131  4.2.4  Deuterium Wash-out Experiment with [3”H]- or 2 [5 “Hj -GDP-6-Deoxy-4-Keto-D-Mannose 2  4.2.5  Deuterium Wash-in Experiment with GFS-CyslO9Ser Product Mixture  4.3  133  134  Deuterium Wash-in Experiments with GME Mutants  137  4.3.1  Deuterium Wash-in Experiment with Lys2l7Ala Mutant  138  4.3.2  Deuterium Wash-in Experiment with Cysl45Ala Mutant  139  4.4  Future Research  143  4.5  Conclusions  145  4.6  Experimental  148  4.6.1  Over-expression and Purification of GME and Mutants  148  4.6.2  Enzymatic Synthesis of GDP-L-Fucose and GDP-D-Altrose  148  4.6.3  Enzymatic Synthesis of {5”H]-GDP-L-Fucose 2  149  4.6.4  Enzymatic Synthesis of [3”H]-GDP-L-Fucose 2  149  4.6.5  Deuterium Wash-in Experiments  150  4.6.6  Deuterium Wash-out Experiments  151  References  152  vm  LIST OF TABLES Table 2.1  Kinetic parameters obtained with unlabeled and [5”H1-GDP-6-deoxy-4-keto2 D-mamiose and NADPH  Table 2.2  60  Kinetic parameters obtained with unlabeled and [3”Hj-GDP-6-deoxy-4-keto2 D-mannose  Table 3.1  Kinetic parameters for the reaction of Cys 1 O9Ser with unlabeled and [3 “H] 2 GDP-6-deoxy-4-keto-D-mannose  65 -  103  ix  LIST OF FIGURES Figure 1.1  Picture of Hermann Emil Fischer and Fischer projection conversions  4  Figure 1.2  D-glyceraldehyde and L-glyceraldehyde  5  Figure 1.3  Cyclization of glucose and a/f3 assignment  6  Figure 1.4  Generic sugar nucleotide  7  Figure 1.5  A) Direct deprotonationlreprotonation mechanism of epimerization at activated stereocenters. B) Epimerization at unactivated stereocenters requires alternative mechanism (X  =  OH or NH ) 2  8  Figure 1.6  Typical positions and roles of catalytic residues in SDR superfamily enzyme.. ..10  Figure 1.7  Reaction catalyzed by UDP-galactose 4-epimerase (GalE)  Figure 1.8  The UDP-galactose 4-epimerase reaction proceeds via a ketone intermediate.. .12  Figure 1.9  Tritium transfer from the GalE cofactor to a ketone intermediate analogue  Figure 1.10  Occasional release of UDP-4-ketoglucose from the active site causes the  11  .  13  formation of abortive complex. Square brackets represent GalE active site bound species  14  Figure 1.11  Nonstereospecific intramolecular hydride transfer catalyzed by GalE  16  Figure 1.12  Catalytic diad of E. coil UDP-galactose 4-epimerase  18  Figure 1.13  Reaction catalyzed by UDP-N-acetylglucosamine 2-epimerase (RffE)  19  Figure 1.14  Mechanism of UDP-GlcNAc 2-epimerase (RffE). Square brackets represent active site bound species  20  x Figure 1.15  Positional isotope exchange (PIX) experiment with UDP-G1cNAc 2-epimerase  21  Figure 1.16  Biosynthesis of dTDP-L-rhamnose  22  Figure 1.17  Proposed mechanism of the reaction catalyzed by RmlC  23  Figure 1.18  Stereochemical inversions catalyzed by Rm1C, NovW, and EvaD  24  Figure 1.19  The reactions catalyzed by GDP-D-mannose dehydratase (GMD) and GDP-fucose synthase (GFS) in the biosynthesis of GDP-L-fucose  26  Figure 1.20  Mechanism of the reaction catalyzed by GDP-mannose dehydratase (GMD)  27  Figure 1.21  Putative mechanism of the reaction catalyzed by GDP-fucose synthase  28  Figure 1.22  Topological structure of GFS dimer. 56  29  Figure 1.23  Reaction and equilibrium distribution of GDP-mannose 3, 5-epimerase (GME)  31  Figure 1.24  Mechanism of the reaction catalyzed by GME  32  Figure 1.25  Topological structure of GME dimer, 64  33  Figure 1.26  Illustration of the Cys-sugar-Lys sandwich in the active site of GME  34  Figure 1.27  Structure of L-fucose showing both the ‘C 4 and the 4 C, conformations and the structure of GDP-L-fucose  35  Figure 1.28  Structure of L-ascorbic acid (vitamin C)  37  Figure 1.29  Proposed de novo biosynthesis pathway of vitamin C in plants. 64  38  Figure 2.1  Cloning of a gene using the XaJLIC cloning kit (Novagen)  44  xi Figure 2.2  Schematic representation of lac operon mechanism (Access Excellence at the National Health Museum)  Figure 2.3  46  SDS-PAGE showing the purification of GFS. Lanes: 1 and 6) Molecular weight standards carbonic anhydrase (66 kDa) and bovine serum albumin (29 kDa). 2) Crude cell lysate after over-expression. 3 and 4) Ni 2 column flow-through (5 mM and 100 mM imidazole, respectively). 5) Purified GFS  47  Figure 2.4  P NMR Spectrum of GDP-L-fucose produced using NADPH 31  49  Figure 2.5  P NMR spectrum of GDP-L-fucose produced using NADPH followed by 31 4 reduction NaBH  50  Figure 2.6  P NMR spectrum of GDP-L-fucose produced using NADH 31  51  Figure 2.7  P NMR spectra monitoring the hydrolysis of the 3’-phosphate group from 31 NADP by alkaline phosphatase  52  Figure 2.8  Overall purification scheme of GDP-L-fucose  53  Figure 2.9a  ‘H NMR spectrum of purified GDP-L-fucose. Peaks at the 1.2 and 3.0 ppm  are due to triethylammonium counter ions  54  Figure 2.9b  P NMR spectrum of pure GDP-L-fucose 31  54  Figure 2.10  Deuterium wash-in experiment with GDP-L-fucose  55  Figure 2.11  H NMR spectra showing deuterium wash-in with GDP-L-fucose 1  56  Figure 2.12  Synthesis of [5”H]-GDP-6-deoxy-4-keto-D-mannose using GMD 2  58  Figure 2.13  Negative ESI-MS of unlabeled and [5”H]-GDP-6-deoxy-4-keto-D-mannose. .59 2  Figure 2.14  H]-GDP-6-deoxy2 Kinetic plots of the GFS reaction with unlabeled and [5”-  Figure 2.15  4-keto-D-mannose  60  GME catalysis in D 0 buffer 2  61  XI  Figure 2.16  Synthesis of [3”H]-GDP-6-deoxy-4-keto-D-mannose using GMD 2  Figure 2.17  Negative ESI-MS of unlabeled and 2 [3”H ]-GDP-6-deoxy-4-keto-D-mannose. ..63  Figure 2.18  Kinetic plots of GFS with unlabeled and [3 “H]-GDP-6-deoxy-4-keto-D2  62  mannose  64  Figure 2.19  Epimerization via enolate intermediate  66  Figure 2.20  Possible mechanisms of the reaction catalyzed by GFS  66  Figure 2.21  The structure of GDP-2-deoxy-2-fluoro-D-mannose  67  Figure 2.22  Possible products obtained from incubating GDP-2-deoxy-2-fluoro-D-mannose with GFS and NADPH  68  Figure 2.23  Synthesis of GDP-2-deoxy-2-fluoro-D-mannose  70  Figure 2.24  Negative ESI mass spectrum of GDP-2-deoxy-2-fluoro-D-mannose, 11  71  Figure 2.25  P NMR spectrum of GDP-2-deoxy-2-fluoro-D-mannose, 11 31  71  Figure 2.26  GMD reaction with GDP-2-deoxy-2-fluoro-D-mannose 11 monitored by P 31 NMR  72  Figure 3.1  X-ray crystal structures of GFS and GME active sites  87  Figure 3.2  Partial ‘H NMR spectra of wild type and Cl 09S product mixtures  91  Figure 3.3  P NMR spectra of wild type and C1O9S product mixtures 31  92  Figure 3.4  Possible structures of the unknown GDP-sugar produced from CyslO9Ser catalysis  93  Figure 3.5  Enzymatic synthesis of GDP-6-deoxy-D-mannose using GMD  94  Figure 3.6  ‘H NMR comparison of different GDP-6-deoxy-sugars  95  XIII  Figure 3.7  Negative ESI mass spectra of the product mixture obtained in the reaction of H]- or [5 “2 Cys 1 O9Ser with [3 “H] -GDP-6-deoxy-4-keto-D-mannose 2  Figure 3.8  96  Product formation with CyslO9Ser and [3”,5”H]-GDP-6-deoxy-4-keto-D2 mannose  Figure 3.9  97  ‘H NMR spectral comparison of various deuterium wash-in experiments using wild type and ClO9Ser GFS. A) No reaction; B) Reaction with wild type GFS; C) Reaction with CyslO9Ser  Figure 3.10  Proposed mechanism for deuterium wash-in with CyslO9Ser mutant  Figure 3.11  Schematic illustration of the effects of mutation on a non-rate-determining  99 100  step Figure 3.12  101  Kinetic plots of the reaction of CyslO9Ser with unlabeled and [3”H]-GDP-62 deoxy-4-keto-D-mannose  102  Figure 3.13  Deuterium wash-in experiment with His 179Gm mutant  104  Figure 3.14  Negative ESI mass spectra monitoring the deuterium wash-out experiment H]-GDP-6-deoxy-4-keto-D-mannose with His 1 79Gln 2 of [3”, 5 “-  Figure 3.15  105  Negative ESI mass spectra monitoring the deuterium wash-out experiment of H]-GDP-6-deoxy-4-keto-D-mannose with Hisl79Gln 2 H]- and [5”2 [3”-  106  Figure 3.16  Structures of GDP-6-deoxy-L-gulose and GDP-6-deoxy-D-altrose  108  Figure 3.17  The proposed mechanism of GFS and the proposed roles of Cys 109 and His179  Figure 4.1  110  Proposed mechanism of the reaction catalyzed by GME and roles of catalytic residues  121  Figure 4.2  0. .124 2 Expected products from the incubation of GDP-L-fucose with GME in D  Figure 4.3  P NMR spectrum showing the reaction of GDP-L-fucose with GME 31  . .  125  xiv  Figure 4.4  Negative ESI mass spectra showing the GME-catalyzed wash-in of deuterium into GDP-L-fucose  126  Figure 4.5  Enzymatic synthesis of [3”H]-GDP-L-fücose and [5”2 H]-GDP-L-fucose 2  128  Figure 4.6  Partial ‘H NMR spectra showing GDP-L-fiicose, [3”H]-GDP-L-fucose, 2 and [5”H]-GDP-L-fucose 2  Figure 4.7  Negative ESI mass spectra showing deuterium wash-out experiment with H] or [5”2 [3”H]-GDP-L-fucose 2  Figure 4.8  129  130  A) 31 P NMR spectrum of GDP-6-deoxy-D-mannose. B) 31 P NMR spectrum of GDP-6-deoxy-D-mannose after extended incubation with GME in D 0 2  132  Figure 4,9  Structures of GDP-L-gulose and GDP-6-deoxy-D-altrose  135  Figure 4.10  Negative ESI mass-spectra showing deuterium wash-in experiment with Cys 1 O9Ser-GFS product mixture  136  Figure 4.11  Deuterium wash-in studies with Lys2l7Ala mutant  138  Figure 4.12  Deuterium wash-in experiment with Cysl45Ala mutant  140  Figure 4.13  Negative ESI mass spectra of Cysl45Ala-catalyzed deuterium wash-in with GDP-L-fucose  Figure 4.14  H NMR spectra showing Cysl45Ala-catalyzed deuteriurn wash4n using 1 GDP-L-fucose  Figure 4.15  142  Proposed deuterium wash-in studies with Lys2 1 7Ala acting on GDP-L-gulose or GDP-6-deoxy-L-gulose  Figure 4.16  141  143  Proposed deuterium wash-in studies with Cysl45Ala acting on GDP-L-gulose or GDP-6-deoxy-L-gulose  144  xv  LIST OF SYMBOLS AND ABBREVIATIONS A  Adenosine  260 A  Absorbance at 260 nm  340 A  Absorbance at 340 nm  A. thaliana  Arabidopsis thaliana, a species of plant  BSA  Bovine serum albumin  C  Cytidine  CV  Column volume  d  Doublet  Da  Dalton  DMAP  4-Dimethylaminopyridine  DMF  Dimethylformamide Chemical shift (ppm)  E. coil  Escherichia coil, a species of bacteria  ESI-MS  Electrospray ionization mass spectroscopy  EvaD  dTDP-3-amino-2, 3 -trideoxy-3 -C-methyl-D-erythro-4-hexulose 5epimerase  G  Guanosine  xvi  GalE  UDP-galactose 4-epimerase  GDP  Guanosine di-phosphate  GFS  GDP-L-fucose synthase  G1cNAc  N-Acetylglucosamine  GMD  GDP-mannose 4, 6-dehydratase  GME  GDP-D-mannose 3, 5-epimerase  GMER  GDP-4-keto-6-deoxy-D-mannose 3, 5-epimerase  IPTG  Isopropyl -D- 1 -thiogalactopyranoside  J  Coupling constant (Hz)  LB  Luria-Bertani  LIC  Ligation independent cloning  MW  Molecular weight  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)  NAPS  Nucleic acid protein service at UBC  NDP  Nucleoside diphosphate (and nucleotide)  xvii  NMR  Nuclear magnetic resonance  NOE  Nuclear overhauser effect  NovW  TDP-6-deoxy-D-xylo-4-hexulose 3 -epimerase  600 0D  Optical density at 600 nm  PAGE  Polyacrylamide gel electrophoresis  PCR  Polymerase chain reaction  PIX  Positional isotope exchange  RffE  UDP-N-acetylglucosamine 2-epimerase  Rm1B  dTDP-glucose 4, 6-epimerase  Rm1C  dTDP-4-dehydrorhamnose 3, 5-epimerase  RT  Room temperature  SDR  Short-chain dehydrogenase / reductase  SDS-PAGE  Sodium dodecyl sulfate polyacrylamide gel electrophoresis  t  Time  T  Thymidine  U  Uridine  UDP  Uridine diphosphate  Vmax  Maximal reaction velocity (rate)  XVIII  Common Amino Acid Abbreviation A  Ala  Alanine  C  Cys  Cysteine  D  Asp  Aspartate  E  Glu  Glutamate  F  Phe  Phenylalanine  G  Gly  Glycine  H  His  Histidine  I  Tie  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  xix  GLOSSARY  Abortive complex  A complex of nonfunctional enzymes with substrate bound.  This complex is formed often when the cofactor is in the wrong oxidation state (i.e. NAD(P)H) in the case of NAD(P)t dependent enzymes that have a tightly bound cofactor.  Activated stereocenter  A stereocenter that contains an acidic C-H bond (pKa <  30) as a result of a neighboring carbonyl functionality. The acidified proton can be directly removed by the catalysis of enzymes such as epimerase, dehydratase or dehydrogenase.  Alpha helix  In a secondary structure of proteins, the alpha helix is a common  motif. The polypeptide backbone is in a coiled conformation resembling a spring where the R groups of the amino acid residues protrude outward from the helical backbone. In each helical turn, there are 3.6 amino acid residues in which every backbone N-H group donates a hydrogen bond to the backbone C0 group of the amino acid four residues earlier. The cL-helix found in all proteins has a right-handed helical twist.  Beta strand  In a secondary structure of proteins, the beta strand is a common  motif. It consists of stretches of amino acids that are generally 5-10 amino acids long where the peptide backbones are almost fully extended.  xx  Carbohydrates’  This term comes from the stoichiometric formula C(H O), being 2  simple organic compounds such as aldoses and ketoses, which become hydrated. Hence, the term carbohydrates was formed, referring to ‘hydrates of carbon’. In general terms, carbohydrates include monosaccharides, oligosaccharides, and polysaccharides. In addition to the basic sugar groups, carbohydrates include substrates derived from monosaccharides by the reduction of the carbonyl group (alditols), by oxidation of one or more terminal groups to carboxylic acids or by replacement of one or more hydroxyl group(s) by a hydrogen atom, an  amino group, a thiol group or various similar groups.  Nonstereospecific  Enzymes will generally catalyze stereospecific reactions.  However, this term is often used by enzymologists to emphasize the unusual reaction in enzymology and to describe the enzyme catalyzed reactions that are not stereospecific (such as epimerizations).  Stereospecific  A stereospecific reaction is when the starting materials differ in  their configuration converted to stereoisomerically distinct products. This definition means that a stereospecific process must be stereoselective but it does not have to true in the reverse scenario. It is a term applied in processes involving a chiral catalyst (such as enzyme) or chiral reagent where the configuration of the product of the reaction depends specifically on the configuration of the catalyst or reagent. For example, the process becomes reversed when a catalyst or reagent of the opposite configuration is used.  xxi  Rossmann fold  A protein structural motif found in proteins that bind nuclëotides,  especially the cofactor NAD or NADP. It is a structure composed of a minimum of three parallel beta strands that are linked by two alpha helices. The helices topological order is in a beta-alpha-beta-alpha-beta pattern. Since each Rossmann fold can only bind one nucleotide, the binding domains for di-nucleotides such as NAD or NADP consists of two paired Rossman folds that each binds one nucleotide moiety of the cofactor molecule.  Tightly bound  As applied in the context of this thesis, it means the cofactor is  strongly attached through non-covalent binding to the enzyme and it is non-exchangeable. The cofactors that are tightly bound by the enzyme are often co-purified with the enzyme. Furthermore, the cofactor will not dissociate from the enzyme during dialysis, centrifugal filtration, or during catalysis unless the enzyme is denatured.  xxii  Acknowledgments  I owe the majority of my gratitude to my supervisor, Professor Martin Tanner, who has a combination of leadership, knowledge, enthusiasm, and has a great sense of humor. I would like  to thank Martin for accepting me into his research group in the first place even though I did not have any research experience. I also need to thank him for his guidance, patience, and friendship over the last few years. I am honored to be one of his students, as he provided me an opportunity to deepen my knowledge during my graduate studies. I also feel I am so lucky to have Martin as my supervisor where he not only shared his wisdom and valuable knowledge when it was needed, but also provided me encouragement and freedom to pursue my own ideas. I would also like to thank the past and current members of the Tanner lab, it is my pleasure to work with all of them and I must state that I’m very fortunate to have had such talented, helpful, and comedic companions. In particular, I would like to thank Dr. Wayne Chou for his patience and guidance during the first few months of my research, Dr. Jay Read for valuable hands-on instructions and tips on chemical synthesis, as well as Dr. James Morrison for 0”- H 2 guidance on site-directed mutagenesis and his genius “capillary D 0- NMR technique. In 2 the last few years, the “chemistry” between all the members: Feng Liu, Pavel Glaze, Alain Mayer, Louis Luk, Timin Hadi, Xu Li, Jennifer Johnson, and Jackie Bassirri, created an endless supply of entertaining moments both inside and outside the lab. I’m certain that the resonance of laughs from all the lab members will still echo in my head for years. I would like to acknowledge the shops and services in the chemistry department at UBC. In particular, I would like to thank Dr. Elena Polishchuk, Helen Wright, Candice Martins, and Jie (Jessy) Chen for maintaining the Biological Services Laboratory and for providing me, not only  XXIII  a clean and friendly working area, but also helpful advices when I experienced difficulties during my research. I would like to thank Dr. Yun Ling and the technicians of the UBC Mass Spectrometry Center in the Chemistry department for their help, advice, and guidance throughout the years. Without either the facilities or their help, I would not be able to accomplish this much. I need to thank Professor Stephen Withers and his group for allowing me to use their spectrometer on my kinetics studies. I would also like to thank Dr. James Naismith for donating the recombinant plasmids encoding GME and its’ mutants. Last but not least, I would like to thank my parents, my girlfriend Miss Angela Young, and my family for their endless support. Although the material in this thesis is very likely a stranger to them, this thesis would not even exist without their contribution.  xxiv  Dedicated to  My parents & Angela Young  1  Chapter 1 Sugar Nucleotide Epimerases  2  1.1  Introduction  This thesis describes mechanistic studies on two enzymes that come from two very distinct species; one from bacteria and the other from plants. Despite the origin of the enzymes, they both operate on very similar substrates and employ a similar mechanistic strategy in the synthesis of their unusual sugar products. Besides serving as metabolic fuel and molecules of energy storage, carbohydrates also play important roles in cellular structure, signaling, and communication. In order to accommodate such diverse functions, Nature utilizes a huge class of widely varying carbohydrate structures. While many of these carbohydrates are oligomers or polymers of common sugars, like glucose and mannose, with different chain length and connectivity, other essential carbohydrates contain more exotic sugars, such as L-fucose. These exotic sugars are very often evolved in signaling and recognition events and they have increasingly become the focus of modern biological studies. The biosynthetic pathways that Nature uses in generating carbohydrates are often quite complex. Nature may generate common sugars from simple starting materials (de novo biosynthesis), as in the process of gluconeogenesis which starts with pyruvate and produces glucose. However, with exotic sugars, Nature often uses a more energetically efficient approach that begins with abundant common sugars and directly modifies them. In many situations, these modifications take place on activated forms of sugars known as sugar nucleotides. The chemical strategies that Nature employs to modify sugar nucleotides are often completely distinct from the approach that a chemist would devise to achieve the same transformation. Therefore, the elucidation of these strategies and mechanisms merits investigation.  3 Two important classes of sugar-modifying enzymes are dehydratases and epimerases. 2 Dehydratases catalyze an overall elimination of water from their starting materials and produce deoxy sugars. Epimerases catalyze the inversion of stereochemistry at a single stereocenter in molecules having more than one stereocenter. The term “epimerase” has also been applied to enzymes that invert two stereocenters in molecules bearing more than two stereocenters and two such enzymes will be presented in this thesis. Together, dehydratases and epimerases provide access to a diverse set of deoxy and epimeric sugars. Examples of some of the best understood sugar nucleotide epimerases are discussed in this opening chapter with an emphasis on studies of their chemical mechanisms and roles of catalytic residues. Many of these enzymes belong to the short-chain dehydrogenase / reductase (SDR) family and therefore a brief description of this family of enzymes is included. The two enzymes investigated in this thesis are introduced near the end of this chapter; GDP4-keto-6deoxy-D-marmose 3, 5-epimerase / reductase (GMER) or GDP-L-fucose synthase (GFS) from E. coli and GDP-D-mannose 3, 5-epimerase (GME) from A. thaliana. This chapter will establish a connection between GFS and GME, showing that they are unique in this family of enzymes and both employ similar mechanistic strategies even though they catalyze different reactions and are from organisms that belong to different Kingdoms.  4  1.2  Dextro I Levo and a I  13 Assignment in Monosaccharides  Single carbohydrate units, such as glucose, are called monosaccharides and have the generic formula CH O. They are polyhydroxylated carbon chains with one of the carbons 2 existing as an aldehyde or ketone. Polymers of monosaccharides, such as starch, are called polysaccharides. The suffix —ose indicates a molecule is a carbohydrate, and prefixes such as tn-, pent-, and hex- are used to describe the number of carbons. In 1891, the German Nobel Prize winner Hermann Emil Fischer devised a two-dimensional representation of a three-dimensional organic molecule by projection: the Fischer projection. For simplicity, the concept is illustrated using glyceraldehye.  GyceraIdehyde 3D represntation  Figure 1.1  cHO H’C’OH OH 2 CH  Convert to Fischer projection  CHO H——OH OH 2 CH  Glyceraldehyde Fischer projection  Picture of Hermann Emil Fischer and Fischer projection conversions.  In a Fischer projection, all horizontal bonds project towards the viewer while vertical bonds project away from the viewer. One may notice that glyceraldehyde has one stereocenter and the one illustrated in Figure 1.1 is of the “R” configuration. Although the R, S configuration system  5  is widely accepted today as a standard for designating stereochemistry, the configuration of carbohydrates, as well as amino acids, is commonly designated by the D, L nomenclature system. In Fischer’s time, it was known that one enantiomer of glyceraldehyde rotated plane-polarized light by +13.5° and the other rotated the same polarized light by -13.5°. However, it was not known which configuration of glyceraldehyde was responsible for which rotation. With a 50/50 chance of being correct, Fischer arbitrarily assigned the R-glyceraldehyde as D-glyceraldehyde for dextrorotary (rotation to the right) and S-glyceraldehyde as L-glyceraldehyde for levorotatory (rotation to the left). The uncertainty remained as a mystery for over haifa century until 1952 when x-ray diffraction analysis proved Fischer’s 50/50 chance assignment was actually correct. CHO H  CHO OH  HO  OH 2 CH  D-Glyceraldehyde  Figure 1.2  D  H OH 2 CH  L-Glyceraldehyde  D-glyceraldehyde and L-glyceraldehyde.  and L-glyceraldehyde serve as the reference point for the assignment of the relative  configuration to all other aldoses and ketoses. The stereocenter furthest from the carbonyl group is called the penultimate carbon since it is next to the last carbon of the chain. Therefore, a D monosaccharide has the same configuration at its penultimate carbon as D-glyceraldehydes since its —OH is on the right side of the Fischer projection. However, when this nomenclature was extended to the rest of the monosaccharides, the dextro- or levo- assignment no longer reflected the true optical properties of the sugar. For example, D-glucose rotates plane-polarized light to  6 the right but D-fructose rotates it to the left. For this reason, the sign of rotation is often listed directly after the configurational description as in D(+)-glucose and D(-)-fructose. Another nomenclature system in carbohydrate chemistry is the ci. and 13 assignment. Monosaccharides can exist in both open chain and cyclic forms. The cyclization involves one of the hydroxyl groups adding to the carbonyl functionality and forming two possible hemiacetal products as illustrated below for glucose (Figure 1.3). Anomeric Carbon  CHO H  OH CH,OH  HO  CH OH  H Redraw  CH,OH  13.D-Glucose  a-D-Glucose  D-Glucose  Figure 1.3  Cyclization of glucose and a/3 assignment.  A common way of representing the cyclic structure of monosaccharides is to use the Haworth projection named after a British chemistry Nobel Prize winner Sir Walter N. Haworth. In this projection, the monosaccharide is represented as a planar hexagon or pentagon lying perpendicular to the plane of the paper. Groups are attached either above or below the plane to indicate the appropriate stereochemistry. The new stereocenter created by formation of the ring is called the anomeric carbon. When the —OH (or —OR) group at the anomeric carbon is trans to the OH group at C5, it is call the cK-anomer; similarly, if the —OH (or —OR) group at the 2 —CH OH group at C5, it is called the f3-anomer. After half a century 2 anomeric carbon is cis to the —CH of chemistry, the simple name of glucose became a. or 13-D-(+)-glucose.  7  1.3 Sugar Nucleotides  A sugar nucleotide is a nucleotide linked via phosphate to the anomeric position of a sugar (Figure 1.4). The nucleotide contains a diphosphate group (sometimes a monophosphate) linked to the 5’-oxygen of either ribose or 2’-deoxyribose, which in turn is 13-linked to a base. Common bases are the purines (adenine or guanine) or the pyrimidines (cytosine, uracil, or thymine).  Sugar of Interest  R  =  H or OH  Base  Figure 1.4  =  Adenine, Guanine, Cytosine, Uracil, or Thymine  Generic sugar nucleotide.  The nucleotide portion of a sugar nucleotide has several important functions. The negatively charged phosphates reduce membrane permeability of the molecule and keep the exotic sugars within the cell. This allows the sugar to be transferred to glycosyl acceptor molecules that include other sugars, proteins, DNA, lipids, and a variety of secondary metabolites such as macrolide antibiotics. Furthermore, the nucleotide can serve as an excellent leaving group in transferase 2 reactions. Finally, the identity of the nucleotide can also serve to differentiate between pools of sugars and ensure that the correct sugars are transferred to the appropriate acceptors.  8  1.4  Epimerization at Activated and Unactivated Stereocenters  Generally, there are two types of epimerases that can be classified according to the nature of the substrate . The first type of epimerization occurs at stereocenters that have acidic C-H 3 bonds. These are known as ‘activated’ stereocenters and they bear a proton with a relatively low pKa (generally  <  30). In these cases, a direct deprotonationlreprotonation mechanism is  employed and a carbonyl functionality stabilizes the resulting carbanion (Figure 1 .5A). The second type of epimerization occurs at ‘unactivated’ stereocenters where the C-H bonds are nonacidic (pKa> 30) (Figure 1 .5B). Since enzymes cannot directly deprotonate non-acidic C-H bonds, a different mechanistic strategy would have to be used in these cases.  A  0 1 R  2 R  x HB  \  E nz  R  R2 HX  Figure 1.5  HB\  HBB Enz Enz  Mechanism (  Enz  R  Enz  R2 XH  A) Direct deprotonation/reprotonation mechanism of epimerization at activated stereocenters. B) Epimerization at unactivated stereocenters requires alternative mechanism (X =  OH or NH ). 2  9  1.5  Short-chain Dehydrogenase Reductase (SDR) Superfamily  Short-chain dehydrogenase/reductase (SDR) enzymes are a large class of proteins that includes about 3000 members (based on sequence homology) and 63 SDR enzymes have been identified in the human genome. 5 SDR enzymes function as NAD- or NADP- dependent ’ 4 oxidoreductases and perform metal ion-independent redox chemistry. Enzymes in this SDR superfamily act on a variety of substrates ranging from steroids, alcohols, sugars, and aromatic 45 compounds to xenobiotics. SDR enzymes have a core structure of typically 250-3 50 residues in length, but may also possess N- or C-terminal transmembrane domains, signal peptides, or domains responsible for 4 Another feature of SDR enzymes is that even though the formation of multi-enzyme complexes. they have may low sequence identity (typically 15-30% in pair-wise comparisons), there are three conserved catalytic residues present in almost all SDR superfamily members  —  tyrosine,  lysine, and serine. Tyrosine is a strictly conserved residue, serine is sometimes replaced with threonine, and the lysine is almost always present. 4 Another signature of SDR enzymes are the positional patterns of the conserved catalytic residues. The serine (or threonine) is typically 12 residues upstream from the tyrosine which is found four residues upstream from the lysine. Finally, the generally conserved Rossmann fold containing a characteristic signature of Gly-X X-X-Gly-X-Gly, where X can be any residue, is usually found near the N-terminus of the enzyme. Furthermore, based on all the crystallographic data available, SDR enzymes share highly related tertiary structures even though the sequence homology is low. They all reveal the coenzyme binding region composed of an c/f3-fold that is related to a Rossmann-fold, and an  10  active site and substrate binding region with similar relative positioning of the catalytic triad. 5 ’ 4 Figure 1.6 summarizes the typical positions and roles of the catalytic triad side-chains, where the tyrosine residue is the general acid/base, found in proximity to the nicotinamide ring of the cofactor. 4-8 The positive charge of the oxidized nicotinamide ring lowers the pKa of the O of the tyrosine to about 7. The lysine residue hydrogen bonds to the hydroxyls of the ribose and is also thought to activate the oxidized NAD cofactor by electrostatic repulsion. It also serves to lower the pKa of the Tyr-OH, however it does not hydrogen bond with the tyrosine directly. The serine is in proximity to the carbonyl/hydroxyl of the substrate and the 0” of the tyrosine and serves to position the substrate and the tyrosine during catalysis. H NAD(P)H  substrate  H  Serine  Figure 1.6  Typical positions and roles of catalytic residues in SDR superfamily enzyme.  11  1.6  Different Strategies Used by Sugar Nucleotide Epimerases  Nature employs several different strategies in catalyzing epimerizations that modify pre existing sugar nucleotides. These strategies will be discussed in this section and highlighted with some of the well-studied examples.  1.6.1 Direct Oxidation / Reduction Mechanism  Of all the known sugar nucleotide-modifying enzymes, perhaps the best understood in both mechanistic and structural sense is UDP-galactose 4-epimerase (GalE) which interconverts UDP-galactose and UDP-glucose (Figure 1.7). GalE is one of the three enzymes in the Leloir pathway which is the only pathway where glucose and galactose are interconverted biologically. Luis F. Leloir discovered GalE in 1951, and since then scientists have tried to elucidate the mechanism of the reaction. However, a generally accepted chemical mechanism was not established until the  197Os10,h1  GalE exists in both mammals and bacteria, but the most  extensive studies have been performed on the UDP-galactose 4-epimerase from E. coli. OH  OH  OH  U D P-galactose 4-epimerase HO OUDP  UDP-galactose  Figure 1.7  Reaction of UDP-galactose 4-epimerase (GalE).  OUDP  UDP-glucose  12  In the 1950’s when the studies of GalE were still in the early stages, the enzyme was named galacto-waldenase, implying a mechanism involving nucleophilic displacement of the C4” hydroxyl by water. 12 However, early mechanistic studies showed that solvent derived 2 H- and 0-atoms were not incorporated into the epimeric substrates, even upon an extended period of 18 incubation with the enzyme. Therefore the Waldenase mechanism was quickly discounted since it would require an exchange of the C4” oxygen with solvent.’ 36 It was also established early on that GalE is a dimeric enzyme and is isolated with one tightly bound NAD cofactor per subunit. Furthermore, the binding of substrate results in the reduction of cofactor to NADH. It was also known that this cofactor is regenerated upon each catalytic cycle and cannot be removed from the active site without denaturing the enzyme. While several possible mechanisms could be proposed that would account for all these observations, Elizabeth S. Maxwell, in 1957, suggested that GalE oxidizes its substrate at C-4” to generate a ketone intermediate in the course of the reaction (Figure 1.8).12  OUDP  U DP-galactose  Figure 1.8  OUDP  OUDP  U DP-4-keto-glucose  U DP-glucose  UDP-galactose 4-epimerase proceeds via a ketone intermediate.  After this keto-intermediate mechanism was proposed, an increasing amount of evidence was observed in support of it. The first strong evidence came from studies with 4-keto-UDPsugars. The epimerase was treated with NaB 4 to generate the reduced form of the enzyme H 3 (containing NAD H) and it was shown that this reduced enzyme was able to deliver its tritium 3  13  label to UDP-4-keto-6-deoxyglucose to produce a tritium-labeled mixture of UDP-6deoxyglucose and UDP-6-deoxygalactose (Figure 1.9). 19  UDP-Galactose 4Epimerase [4-f3H]NADH 3  OUDP  U DP-4-keto-6-deoxy-g lucose Figure 1.9  U DP-6-deoxy-glucose  OUDP  U DP-6-deoxy-galactose  Tritium transfer from the GalE cofactor to a ketone intermediate analogue.  An even more conclusive piece of evidence came when it was found that extended incubation of GalE with UDP-glucose resulted in the accumulation of UDP-4-ketoglucose. ° 2 This accumulation of UDP-4-ketoglucose was due to the reversible substrate-induced inactivation of the enzyme. 22 The reversible substrate-induced inactivation of the enzyme ’ 21 results from the premature release of the 4-keto intermediate, leaving the tightly bound cofactor in its reduced (inactive) form (Ga1ENADH) (Figure 1.10). When Ga1ENADH binds either UDP-glucose or UDP-galactose, no further catalysis is possible. The complex of reduced enzyme with either substrate or product is called an abortive complex. The released UDP-4-ketoglucose could be chemically reduced by NaB 4 to give a mixture of UDP-[4H 3 H]-glucose and UDP-[43 H]-galactose (Figure 1.10). 3  14 NADH  EA::H  :zO  OUDP  OUDP  OUDP  U DP-galactose  U DP-glucose  U DP-4-keto-glucose  GalE  Accidental Release of Intermediate 0 OH  HO  +  GaIENADH  OH OUDP U DP-4-keto-glucose  UDP-HexoseNj  H <OH 3  OH O\ 3 H HO1j OUDP  Figure 1.10  UDP-Hexose  GaIENADHUDP-Hexose  HO0\ HO_.j  (Abortive Complex)  OUDP  Occasional release of UDP-4-ketoglucose from the active site causes the formation of abortive complex. Square brackets represent GalE active site bound species.  Finally, the direct C-4” oxidation mechanism is supported by the observation of a kinetic isotope effect in the epimerization of UDP-[4H] glucose. 2 24 This observation strongly implies ’ 23 that the C-4” carbon-hydrogen bond is broken during catalysis and this step is at least partially rate determining. Having elucidated the chemical mechanism, enzymologists started to study how GalE is able to achieve catalysis. In particular, it was difficult to understand how the enzyme is able to  15  present either face of the carbonyl of the ketone intermediate to the cofactor. This caught the enzymologists’ attention because most dehydrogenases catalyze stereospecific hydride transfer to and from one particular face of a molecule. However, in the case of GalE, the ketone intermediate is reduced without any preference for stereochemistry and this has been called nonstereospecific hydride transfer. Interestingly, while the hydride is transferred nonstereospecifically to the substrate, the hydride is actually stereospecifically transferred to the cofactor. As with all SDR superfamily members, the hydride is transferred to and from the pro-S position on the B side of the nicotinamide ring. 19 One could propose two different possible mechanisms for the GalE reaction based on all these observations. In the first mechanism the ketone intermediate is released into solution and then rebinds with the other side of the carbonyl exposed to the cofactor. In the second mechanism the ketone intermediate occupies the active site throughout the reaction but undergoes a conformational reorientation within the active site. To differentiate between these two scenarios, an isotope scrambling experiment was performed using a deuterium labeled substrate. A mixture of UDP-glucose-d 7 and UDP-glucose was incubated with GalE and the products were analyzed using mass spectrometry, 25 The first mechanism would result in isotope scrambling because the ketone intermediate would not always rebind to the same active site during catalysis. Therefore epimerization would be an intermolecular process where mass spectrometry signals corresponding to d 1 and d 6 species would be detected. Conversely, the second possible mechanism will not result in isotope scrambling since it is an intramolecular process. Therefore, only signals that correspond to 0 d and 7 d species would be detected. In fact, the experimental result indicated that there was no isotope scrambling during epimerization, implying that epimerization occurs within a single active site and the same NAD cofactor mediates the reaction.  16  Furthermore, studies on the binding energies of different portions of the sugar nucleotide provided insights on how GalE can achieve this nonstereospecific hydride transfer in a single active site. It was revealed that the enzyme actually binds the UDP moiety much more tightly than the sugar moiety. 27 The standard free energy for the binding of the UDP moiety to ’ 26 GalENAD is about -5 kcal/mol, whereas, the standard free energies for the binding of the sugar is about zero. Comparatively, the standard free energy for the binding of UDP-4-ketoglucose to Ga1ENADH is about -7 kcal/mol, whereas, the standard free energy for binding of 4-ketoglucose is about -2 kcal/mol. The values suggest that the UDP moiety anchors the substrates and the intermediate in the active site and leaves the sugar moiety some freedom to rotate inside the active site. By gathering all the accumulated evidence, Frey finally developed a model explaining the nonstereospecific GalE catalysis in 1975 (Figure 1.11)26 It was proposed that the sugar moiety of UDP-4-ketoglucose is able to rotate around the bonds linking the UDP moiety to the 4ketoglucose moiety. A rotation of 1800 will allow the enzyme to present either face of the substrate carbonyl group to a stationary NADH cofactor. NADH  -  UDP-glucose  OH  1HOUDP  U DP-4-keto-glucose 180° rotation NAD  nz H HO 0 OH PDP  B: HO  UDP OH  U DP-4-keto-glucose  Figure 1.11  0  U DP-galactose  Nonstereospecific intramolecular hydride transfer catalyzed by GalE.  17  More recent work has focused on crystallographic studies of GalE. The main goal of the studies was to observe the conformation of substrate bound in the active site and to determine if the sugar moieties of UDP-galactose and UDP-glucose were indeed bound as predicted. To answer this question, two important structures, each with either UDP-glucose bound or UDP galactose bound, were required. The abortive structure of E. coil Ga1ENADH with UDP-glucose bound was solved and the conformation of the bound sugar was illustrated for the first time. 28 This structure indicated that both UDP-glucose and NADH were positioned in a manner consistent with the proposed C-4” oxidation mechanism. However, no conclusive statement could be drawn regarding the 1800 rotation until later when a double mutant of GalE, S124AY149F, was analyzed. Ultimately, it was possible to solve the complex of its abortive 29 The double mutation used here NADH form with either UDP-galactose or UDP-glucose. eliminated any residual activity that could equilibrate the epimers and lead to the crystallization of the favored NADHUDP-glucose complex. By comparing the orientations of the sugar moieties in these two crystal structures, it was apparent that they had rotated dramatically relative to each other, as predicted. This confirmed that the intermediate does undergo rotation during epimerization and that the active site is indeed large enough to accommodate such a motion. GalE is a member of the SDR superfamily and the conserved residues of the catalytic triad for the E. coil enzyme are lysine 153, tyrosine 149, and serine 124.283031 Lysine 153 is hydrogen-bonded to the hydroxyl functionality of the ribose part of the tightly bound NAD and it enhances the chemical reactivity of NAD upon the binding of uridine nucleotides in the active 32 The crystal structure of GalNADHUDP-glucose also suggests that serine 124 is located site. very close to tyrosine 149. While serine 124 is hydrogen-bonded to the C-4” hydroxyl of glucose, tyrosine 149 is located near the nicotinamide ring of NAD and possibly assists serine  18  124 in deprotonating the C-4” hydroxyl during catalysis. 28 The pKa of tyrosine 149 is lowered to 6.08 in wild-type GalE as compared to a pKa of about 10 for free tyrosine. This pKa lowering effect is due to the presence of positive electrostatic field created by NAD and lysine 153 as well as the hydrogen bonding with serine 124. Serine 124 is very close in proximity to the C-4” oxygen of the glucose (2.6  A) which suggests that serine 124 may be the catalytic base as  opposed to tyrosine 149 which is further away  28 Furthermore, mutating either tyrosine (4.3A).  149 to phenylalanine or serine 124 to alanine causes a 10,000 fold or 3,000 fold decrease in activity, respectively. 33 Overall, these observations suggest that tyrosine 149 and serine 124 act together as a catalytic diad. Tyrosine provides the driving force for general acid/base catalysis while serine shuttles protons between the hydroxyl group of glucose and the tyrosine (Figure 1.12). However, studies on the human GalE present a slightly different scenario. The crystal structure of human GalE was obtained using Ga1ENADWUDP-glucose and it shows that tyrosine 157 is the analogous conserved residue. Interestingly, the orientation and the close proximity (3.1A) of tyrosine 157 suggests that it is able to deprotonate the C-4”-OH of the 34 Furthermore, the human GalE active site is also different from that of E. coil glucose directly. GalE in other fundamental ways such as in its ability to bind UDP-GlcNAc. 36 Nevertheless, ’ 35 tyrosine 149 in E. coil GalE and tyrosine 157 in human GalE both appear to be the ultimate conserved basic residues and both enzymes share the same mechanism.  2 NH NAD  Tyrl  Sen 24  Figure 1.12  O—UDP UDP-glucose  Catalytic diad of E. coil UDP-galactose 4-epimerase.  19  1.6.2 Glycal Intermediate Mechanism  UDP-N-Acetylglucosamine 2-epimerase (RffE) is one of the two known enzymes which carry out a rare C-2” epimerization. It catalyzes the interconversion of UDP-N acetylgiucosamine (UDP-G1cNAc) and UDP-N-acetylmannosamine (UDP-ManNAc) (Figure 1.13). This enzyme provides bacteria with a source of UDP-ManNAc which is then used in the biosynthesis of different cell surface polysaccharides. 38 A UDP-hydrolyzing version of this ’ 37 enzyme that generates free ManNAc and UDP was also found to be responsible for controlling the extent of cell surface sialylation in mammals. 39 Sialic acid is biosynthesized from ManNAc  and plays key roles in cellular recognition, signal transduction, and tumorogenesis. The epimerization catalyzed by RffE occurs at an unactivated stereocenter, but, interestingly, is independent from the presence of any cofactor. 4 ’ 40 OH  OH NHAC  HO  -  RffE  HO  AcHN OUDP UDP-GIcNAc  Figure 1.13  OUDP UDP-MariNAc  Reaction catalyzed by UDP-N-acetylglucosamine 2-epimerase (RffE).  Early experimental results showed that deuterium is incorporated at C-2” during epimerization in buffered 42 0. In addition, epimerization of the deuterium labeled substrate, 2 D Hj-G1cNAc, is slowed by a primary kinetic isotope effect. 2 UDP-[2’ A telling mechanistic study 4 showed that upon extended incubation, the stable intermediates UDP and 2-acetamidoglucal  20  accumulated in solution. Ultimately, it was proposed that epimerization occurs via an antielimination to generate 2-acetamidoglucal and UDP, followed by a syn-addition to generate the final product (Figure 1.14).  HO°  UDP-ManNAc  UDP-GIcNAc  HO_—k AcHN  HHzidhe  H  o_  I  2-Acetamidoglucal  RifE  Released  2-Acetamidoglucal  Figure 1.14  +  UDP  Mechanism of UDP-GlcNAc 2-epimerase (RffE). Square brackets represent active site bound species.  Further strong support for this mechanism came from the observation of a positional isotope exchange (PIX) of the oxygen bridging the sugar to UDP (Figure 1.1 5)40 41 When UDP GlcNAc with 180  180  at the anomeric position was incubated with the enzyme, it was found that the  label scrambled into the two non-bridging positions on the 3-phosphate of UDP. This  observation is consistent with a mechanism where the C-0 bond is broken during epimerization to generate 2-acetamidoglucal and UDP. These stable intermediates have a long enough life time for the f3-phosphate of UDP to rotate and, therefore, the C-0 bond can reform with any of the three chemically equivalent oxygen atoms.  H10i:  o_  UDP-GIcNAc  [ UDP-ManNAc  Figure 1.15  UDP-GIcNAc 18(160  AcHN 2-Acetamidoglucal  o_  I  I  0  0 —  —  UDP-ManNAc  RifE  181160  Positional isotope exchange (PIX) experiment with UDP-G1cNAc 2-epimerase.  RIlE is an unusual example of epimerase that is cofactor-independent and acts on an unactivated stereocenter. It is able to promote epimerization since its substrate bears a good leaving group at the anomeric position and an elimination reaction can occur via a transition state with oxocarbenium ion character. In a sense the C-2” position of the substrate is “activated” for elimination even though it is unactivated in terms of the pKa of the C-H bond.  22  1.6.3 Activated Substrate Epimerization Mechanism  Epimerization of sugar nucleotides via this strategy requires a series of three separate enzymes acting in a sequential fashion. The first enzyme activates the substrate for epimerization by installing a ketone functionality via a dehydration or oxidation reaction. Subsequent enzymes catalyze the inversions and reduction of the carbonyl group. One of the best understood examples of such an epimerase is dTDP-4-dehydrorhamnose 3,5-epimerase (Rm1C). This is the second of the three enzymes responsible for the biosynthesis of TDP-L-rhamnose (Figure 1.1 OH O HO  C 3 I-4  RmIB H  OTOP  OTDP  TDP-glucose RmIC 0  o I  OH 3 CH  RmID NADH HOE21OTDP 7 HO  f  OH  TDP-L-rhamnose  Figure 1.16  Biosynthesis of TDP-L-rhaninose.  A dehydratase, RmlB, first converts TDP-glucose into TDP-4-keto-6-deoxyglucose and introduces the keto functionality at C-4”, then the epimerase Rm1C epimerizes the stereochemistry at C-3” and C-5” to give TDP-4-keto-rhamnose. Finally a reductase, RmlD, reduces the ketone at C-4” to give TDP-L-rhamnose. The epimerization steps catalyzed by Rm1C  23  are thought to involve en(di)ol intermediates formed as a result of deprotonation at C-3” or (Figure 1.17).  HO\H,C  C 3 O\\H  RmIC HO  HO  OHI  “OHI  OTDP  OTDP  ‘  \ RmIC  TDP-6-deoxy-4-keto-o-glucose  OH  HO\  /“RmIC  RmIC OH  OTDP  OTDP  OH  OTDP  TDP-4-keto-L-rhamnose  Figure 1.17  Proposed mechanism of the reaction catalyzed by RmIC.  There are three pieces of evidence supporting this mechanism. ° First it was observed 47 that solvent-derived deuterium was incorporated at both C-3” and C-5” during the epimerization process. Second, the epimerization of substrates labeled with deuterium at either C-3” or C-5” are slowed by a primary kinetic isotope effect indicating that C-H bond cleavage is at least partially rate determining. Lastly, the order of inversion of stereocenters is thought to take place first at C-3” and then at C-5”. Evidence in support of this ordering comes from the observation of partial exchange of solvent derived deuterium into TDP-6-deoxy-4-keto-glucose which it is the more stable epimer at equilibrium. Deuterium incorporation at C-5” was only seen in molecules that had already undergone exchange at C-3”, suggesting C-3” is the first to be deprotonated. Two other enzymes are closely related to RmlC (Figure 1.18). They are dTDP-6-deoxyD-xylo-4-hexulose 3 -epimerase (NovW) 46 and dTDP-3-amino-2,3 -trideoxy-3 -C-methyl-D  24  erythro-4-hexulose 5-epimerase (EvaD). 47 Early work assigned 3,5-epimerase activity to both NovW and EvaD, however it has subsequently been showed that NovW only has kinetically significant 3-epimerase activity and EvaD only has 5-epimerase activity. 49 ’ 48  RmIC I  HO  OH OTDP TDP-6-deoxy-D-xylo-4-hexulose  O  3,5-Epimerase  C 3 H  CH OH 3 OH OTDP TDP-6-d eoxy-L-Iyxo4-hexu lose  o  C 3 H  NovW HO OH OTDP TDP.6-deoxy.D.xylo.4-hexu lose  3-Epimerase  OH OTDP TDP-6-deoxy-D-rIbo-4-hexulose OH  H3C 5: OTDP NH 2 TDP-3-amino-2,3,6-trideoxy-3-C-methyl-D.erythro-4.hexulose  Figure 1.18  erase 2 NH  OTDP  TDP-3-amino-2,3,6-tiideoxy-3-C-methyl-L-threo-4-hexulose  Stereochemical inversions catalyzed by RmlC, NovW, and EvaD.  Rm1C, NovW, and EvaD have been grouped together to form the tentatively named RmlC family of epimerases due to their high sequence identity (3O5O%).48 Even though crystal structures of these enzymes have been solved, only limited mechanistic studies have been 49 However, the presence of C-4” ketone in all of the substrates and products implies a performed. direct deprotonation / reprotonation mechanism is at play. This mechanism is supported by the incorporation of deuterium into substrate and product at the sites of epimerization when the reaction is performed in 475 0. Another reason why these enzymes form a class by themselves 2 D ° 5 is that they do not utilize any cofactor and are not members of the SDR superfamily.  Lastly,  25  the x-ray structures of these enzymes reveal four conserved active site residues: a histidine, an aspartate, a lysine, and a tyrosine. The histidine and aspartate are well positioned to operate as a diad in deprotonating the substrate. The positive charge of lysine stabilizes the negative charge at the C-4” oxygen, and the tyrosine is thought to protonate the enolate intermediates in generating the epimeric product. A 1000-fold decrease in activity has been observed upon mutating any of these four residues in Rm1C from Streptococcus suis, supporting their proposed role in ’ 5 catalysis.  1.6.4 Activated Substrate Epimerization / Reduction Mechanism  Epimerization of sugar nucleotides via this strategy requires the sequential action of two enzymes. The first enzyme activates the substrate for epimerization via dehydration or oxidation. The second enzyme carries out the inversion of stereochemistry and also reduces the carbonyl. One of the best known examples of an enzyme using this strategy is the subject of this thesis, GDP-fucose synthase (GFS) (Figure 1.19). This enzyme was first isolated and characterized in the mid-1970’s by Antonio De Flora and it is sometimes known as GDP-4-keto-6-deoxy’ 54 mannose 3 ,5-epimerase/4-reductase (GMER).  Interestingly, the name “synthetase” has  historically been applied to this enzyme but changes in the rules of nomenclature now dictate that enzymes that do not consume ATP during catalysis should be defined by the term “synthase”. 56-58  26 OH OH -O  HO  OGDP G DP-D-man nose  Figure 1.19  Q  GMD  C 3 H  OH  OGDP G DP-6-deoxy-4-keto-D-man nose  OH G DP-L-fucose  The reactions catalyzed by GDP-D-mannose dehydratase (GMD) and GDP-fucose synthase (GFS) in the biosynthesis of GDP-L-fucose.  GFS is responsible for the second step of the de novo biosynthesis of GDP-L-fucose from GDP-D-mannose (Figure 1.19). This pathway was first proposed by Victor Ginsburg in 1960.7981 The first enzyme is a dehydratase, GDP-D-mannose dehydratase (GMD), which converts GDP D-mannose to GDP-4-keto-6-deoxy-D-mannose and it is found in variety of bacterial species, mammals. 80-85 In E. coil, the gene for GMD lies immediately ’ plants, invertebrates, and 56 85 The crystal structure of E. coil GMD reveals that GMD ’ 84 upstream of the gene encoding GFS. is a homodimeric enzyme and is a member of the SDR superfamily with one tightly bound NADPcofactor in each monomer. 80-85 The mechanism of the reaction catalyzed by GMD is illustrated in Figure 1.20. An initial NADP- dependent oxidation of the C-4” hydroxyl forms the GDP-4-keto-D-mannose intermediate and a tightly-bound NADPH cofactor. A subsequent deprotonation at C-5” promotes the elimination of water to give the enone intermediate. Finally, a reduction by the enzyme bound NADPH occurs at C-6” to produce the substrate for GFS, GDP-6-deoxy-4-keto-D-mannose.  27 NADPH  OGDP  GDP-D-mannose  OGDP  GDP-4-keto-D-mannose  NADPH  NADP  OGDP  enone intermediate  Figure 1.20  GDP-6deoxy-4-keto-D-mannose  Mechanism of the reaction catalyzed by GDP-mamiose dehydratase (GMD).  Experiments also demonstrate that GDP-L-fucose is a very effective competitive inhibitor of GMD and it is a classic example of negative feedback inhibition by the final product of the 85 GFS then inverts the stereochemistry of GDP-6-deoxy-4-keto-D-mannose at both C-3” pathway. and C-5” and finally it catalyzes a reduction at C-4” that consumes one equivalent of NADPH. GFS is remarkable in the sense that it is able to perform three distinct tasks within just one active site. Even though the overall reaction catalysis by GFS is not an epimerization, but rather, a diastereomerization followed by reduction, GFS does catalyze two individual epimerization steps during catalysis. The detailed mechanism of GFS has not been fully elucidated; however, the epimerization mechanism is thought to be analogous to that of the reaction catalyzed by Rm1C with the involvement of en(edi)ol intermediates (Figure 1.21). Since the C-3” and C-5” positions are activated by the C-4” ketone functionality, it is very likely that GFS catalyzes the stereochemical inversions via a deprotonation / reprotonation mechanism.  28  H-  (Q3  H Enz-...y.  2 lB  .—  Enz  OGDP / Enz  GDP-6-deoxy-4-keto-D-mannose  enol intermediate  1 B  OH  OGDP  G DP-6-deoxy-4-keto-D-altrose  Enz-BH  HOL OH  OGDP  GD P-L-fucose  Figure 1.21  OH  OGDP  G DP-6-deoxy-4-keto-L-galactose  OH enol intermediate  Putative mechanism of the reaction catalyzed by GDP-fucose synthase.  The order of the epimerization steps in the GFS reaction is not known and is tentatively assigned as C-3” first, then C-5” in the following discussion and Figure 1.21. An initial deprotonation at C-3” generates an enol intermediate and a subsequent reprotonation on the opposite face of the molecule gives the C-3” epimer, GDP-6-deoxy-4-keto-altrose. Then a similar deprotonation and reprotonation sequence inverts the stereochemistry at C-5”to yield GDP-6-deoxy-4-keto-L-galactose. Finally an NADPH-dependent reduction converts the C-4” ketone back into a hydroxyl group and generates GDP-L-fucose (shown in its less stable 4 1 C conformation in Figure 1.21). There is some evidence to support this mechanism. First, it was observed that GDP-6-deoxy-4-keto-mannose bearing a tritium label at C-3” lost its label to solvent during catalysis. 59 This indicates that the inversions of stereochemistry occur via a deprotonation / reprotonation process accompanied by solvent isotope incorporation. Second, it  29  has been observed that GFS is capable in catalyzing the epimerization steps in the absence of the cofactor NADPH. ° Finally, it has been shown that GFS catalyzes the stereospecific transfer of 6 the pro-S hydride from the NADPH cofactor to the C-4” ketone of the sugar nucleotide. This is consistent with the stereospecificity seen in all member of the SDR superfamily. 60 When an NADPH cofactor that had been selectively deuterated on the pro-S face was tested with GFS, it was shown that the reaction was slowed by a kinetic isotope effect (KIE) on the value of kcat of kH/kD  1  =  460  indicating that reduction was only partially rate determining in the overall catalysis.  Two groups have independently solved the x-ray crystal structures of E. coli GFS in a complex with either NADP or NADPH (Figure 1.22). However, no structures with a bound GDP-sugar are 6 58 The structures clearly show that GFS is a member of the SDR 57 available. ’ superfamily and its protein fold closely resembles that of UDP-galactose-4-epimerase, an enzyme that also catalyzes redox chemistry at the C-4” position of a sugar nucleotide (Section l.6.1).63  Figure 1.22  6 Topological structure of GFS dimer.D  30  GFS is a homodimeric enzyme and contains two domains; one for NADPH binding and one for sugar nucleotide binding. GFS also contains the characteristic serine-tyrosine-lysine catalytic triad which is a “fingerprint” of the SDR family members and is responsible for promoting the hydride transfer step. 7 Based on both literature precedence and structural comparison of GFS with other sugar nucleotide-modifying SDR enzymes, GFS Tyr 136 of this triad is strongly implicated as the catalytic acid which protonates the C-4” ketone during the fmal reduction step (B 3 in Figure 1.21). 33,34,63 By examining the crystal structure, two additional active site residues, CyslO9 and His 179, are identified which are not normally found in SDR family members. Modeling studies suggested that they may provide the acids and bases involved in promoting the inversion of stereochemistry at both the C-3” and C-5” positions (B 1 and B 2 in Figure 1.21 )57 58,61 However, in the absence of an x-ray crystal structure with bound substrate, this assignment is tentative. In support of these assignments, site-directed mutations of either Tyr136, CyslO9, or His179 residues resulted in a dramatic decrease in activity suggesting that all these residues are crucial players for efficient catalysis. 61  1.6.5 Oxidation I Epimerization I Reduction Mechanism  In this strategy for epimerization of sugar nucleotides, a single enzyme catalyzes oxidation, epimerization, and reduction of a substrate within a single active site. An example of such an enzyme is found in GDP-mannose 3,5-epimerase (GME). It catalyzes the interconversion  31  of GDP-cL-D-mannose with GDP-f3-L-gulose and GDP--L-ga1actose to give the equilibrium ratio of 80% to 5% to 15%, respectively (Figure 1.23).  HO  GME OGDP  G DP-a-D-man nose Figure 1.23  HO\o;:OGDP  OH  G DP-3-L-gu lose  GME OH  G DP--L-galactose  Reaction and equilibrium distribution of GDP-mannose 3, 5-epimerase (GME).  GME is found in plants and the conversion of GDP-c-D-mannose to GDP-f3-L-galactose is thought to be the first committed step in the de novo biosynthesis of L-ascorbic acid or vitamin 6467 The reaction also produces a small amount of GDP-f3-L-gulose which is also thought to be C. transformed into vitamin C though a distinct, but as yet unknown, pathway. 66 The GME catalyzed epimerization is thought to occur via a transient oxidation at the C-4” position (Figure 1.24).  32 NAD  OGDP  GDP-a-D-mannose  OGDP G DP-cc-4-keto-D-mannose  NADH  1 G DP-3-4-keto-L-guIose  NAD  GDP-3-L-guIose  NADH  NAD OH  HOL OH G DP-3-L-gaactose  Figure 1.24  G DP-3-4-keto-L-gaIactose  Mechanism of the reaction catalyzed by GME.  The ketone functionality at C-4” enhances the acidity of the protons on C-3” and C-5”  and allows epimerization to occur, first at C-5” and then at C-3”, through deprotonation /reprotonation mechanisms. Finally the tightly-bound NADH cofactor reduces the C-4” ketone to regenerate the alcohol and completes this overall redox neutral transformation. There are several observations that support this mechanism. It was observed that GME isolated from a green alge Chiorellapyrenoidosa was able to incorporate tritium from 3 0 into both the C-3” and C-5” 2 H position of substrates and products. 67 The ordering of the two epimerizations was deduced by analyzing the product mixture produced by GME. The observation that GDP-f-L-gu1ose was formed (C-5” epimerization only), but that GDP-D-altrose was not formed (C-3” epimerization only) supports the motion that C-5” epimerization occurs first (Figure 1 .24).64 66  33  During the course of the studies described in this thesis, several x-ray crystal structures of the Arabidopsis thaliana GME were successfully obtained by Naismith and coworkers. 64 These structures include several mutants of GME, with different bound substrates at the active site (GDP-c-D-mannose, GDP-3-L-galactose, and a mixture of GDP-f3-L-gulose with GDP--L-4keto-gulose).  Figure 1.25  Topological structure of GME dimer. 64  GME is also a homodimer and is a member of the SDR superfamily bearing the conserved catalytic serine-tyrosine-lysine triad and the hallmark Rossmann fold with a tightly bound NAD cofactor (Figure 1.25). The crystal structure reveals the positioning of the catalytic triad which, as predicted, is consistent with the typical configuration and function of these residues in SDR enzymes (Figure 1.6). Tyrosine 174 is positioned appropriately to deprotonate the C-4” hydroxyl during oxidation at C-4”. This is also supported by the finding that a tyrosine mutant Yl 74F is unable to catalyze epimerization. The crystal structure also reveals two  34  additional catalytic residues in the active site; cysteine 145 and lysine 217 (Figure 1.26). These two residues are found to be on opposite sides of the active site such that the sugar moiety of the sugar nucleotide is “sandwiched” in between them. They are located in the appropriate positions to act as the acid and base pair that catalyze the epimerizations at C-3” and C-5” via deprotonation I reprotonation mechanisms. Lys217 +  2 NH  /  OGDP  H S Cys145  Figure 1.26  Illustration of the Cys-sugar-Lys sandwich in the active site of GME.  Mutation of these residues (C145S, C145A, K127R, and K127A) had a dramatic effect on activity, supporting their proposed roles. Since Cys 145 and Lys217 are the only acid / base residues found in the vicinity of C-3” and C-5”, they are presumably responsible for both epimerizations and must be able to undergo some minor lateral movement within the active site. The work on GME is also relevant to an understanding of GFS. The two enzymes share 24% sequence identity and the position of the active site residues is similar. Lys 217 of GME aligns with His 179 of GFS and Cys 145 aligns with Cys 109.  35  1.7  GDP-L-Fucose and its Importance in Mammals and Bacteria  —_sL..  COH 3 H  L-fucose  L-fucose  1 Conformation C 4  4 Conformation C 1  0  0  2 NH  1 H,CO—p——0-——p——O———  GDP-L-fucose  Figure 1.27  N  OH  OH  Structure of L-fucose showing both the ‘C 4 and the “C 1 conformations and the structure of GDP-L-fucose.  L-Fucose, also known as 6-deoxy-L-galactose, is widely found in microorganisms, plants, and animals (Figure 1.27). It is a monosaccharide that is a common component of many N- and 0-linked glycans and glycolipids produced by mammalian cells. In bacterial cells, it is found in polysaccharides of the cell wall and in lipopolysaccharides of some Gram-negative micro 68 Two notable features distinguish fucose from other common hexoses. These are the organisms. lack of a hydroxyl group at C-6 and the L- configuration of the sugar. 69 Although L-fucose predominantly exists in the ‘C 4 conformation, for the ease of visualizing stereochemical changes, it will often be shown in the less stable “C 1 conformation in this thesis (Figure 1.27). Frequently, fucose is found as a terminal modification of glycan structures and fucose is incorporated onto these structures via the action of a fucosyltransferase. In all known cases, the substrate for the fucosyltransferase is GDP-L-fueose. 70 Fucosylated glycoproteins and glycolipids can be found in cell membranes or secreted into biological fluids. A variety of functions in biological processes have been established for  36  fucose residues. ° In mammals, L-fucose is a terminal sugar in the glycans that comprise the 7 human ABO blood group antigen, connected via an c-l, 2-linkage. ’ It is also an essential 7 component of the cell surface ligands for the selectin proteins. 72 The Lewis’ and sialyl Lewis’ selectin ligands contain cL-i, 3- and cL-i, 4-fucosylated modifications and play key roles in the inflammation response. The binding of endothelial selectins (a family of cell adhesion molecules on the cell surface) to the fucose containing LewisX and sialyl LewisX structures on the surface of leukocytes is responsible for initiating the inflammatory response. The initial attachment of leukocytes to the endothelium and the rolling of the leukocyte are both mediated by selectins. Humans unable to fucosylate proteins suffer from the immune disorder known as leukocyte adhesion deficiency type ii.’ 0-linked fucose residues (fucose directly linked to hydroxyl groups of serine and theronine residues) are present on the epidermal growth factor (EGF) domains of the mammalian Notch receptors that are required for the proper functioning of Notch signaling pathway. Notch receptors are a family of signaling transmembrane proteins which have important roles in the normal development of an organism. These roles include neurogenesis, angiogenesis, and lymphoid development. Evidence demonstrates that increased levels of fucosylation in the Notch signaling pathway is related to cancer. 75 Furthermore, it has been shown that fucosylation also ’ 74 plays important roles in fertilization and programmed cell death (apoptosis). 69 In recent years, enzymes in the GDP-fucose biosynthetic pathway have become the targets for drug development as GDP-fucose plays an essential metabolic role in many bacteria and parasites. For example, fucose biosynthesis is essential for the life cycle of the human parasite Trypanosoma brucel 76 L-fucose can also be found widely in many strains of which causes African sleeping sickness. bacteria as a component of lipopolysaccharide and polysaccharide capsules. Through evolution, bacteria manipulate these cell membrane glycans in an effort to improve host evasion by means  37  of molecular mimicry. 78 A good example of bacteria using this strategy is the pathogen ’ 77 Helicobacter pylon that causes peptic and duodenal ulcers in humans. It expresses the LewisX trisaccharide as the major component of its lipopolysaceharides to improve endothelial adhesion in the human digestive system. 78  1.8  The Role of GDP-Mannose 3, 5-Epimerase in Plants  L-Ascorbic acid (Figure 1.28), also called L-threo-hex-2-enono- 1 ,4-lactone and commonly known as vitamin C, was first discovered and isolated in late 1920’s by Hungarian physiologist Albert Szent-Gyorgyi.  H(  L-Ascorbic acid  Figure 1.28  Structure of L-ascorbic acid (vitamin C).  Vitamin C is a major carbohydrate in plants and an essential micronutrient for animals. Vitamin C plays a variety of important roles in a wide range of processes in plants. These metabolic  38  events include antioxidant defense, photosynthesis, cell division, and growth regulation. 8688 In addition, Vitamin C is also an essential cofactor for enzymes such as hydroxylases and ° The expression of genes involved in defense and hormone signaling pathways 899 dioxygenases. is affected by the ascorbic acid level in leaves. 91 It is also found that vitamin C plays an important role in methyl-jasmonate stress response pathway and GME is a late methyl-jasmonate response gene involved in this stress response of plants. 99 Methyl-i asmonate is a substance used in plant defense and is produced by plants in response to stress. Methyl-jasmonate mediates plant responses to many biotic and abiotic stresses by triggering a transcriptional reprogramming that allows cells to cope with pathogens and 99 stress. 100 Studies showed that treatment of methyl ’ jasmonate in Arabidopsis and tobacco cells increased their de novo synthesis of vitamin C (Figure 1.29) as well as the transcription of their GME gene.’°° It is not surprising that GME is involved in this stress response since vitamin C is an important primary metabolite of plants. Last, but not least, plant vitamin C is the major source of dietary vitamin C for humans. The biosynthesis pathway of vitamin C in plants was determined in 1998.92 However, it is still far from clear what factors control the synthesis and turnover of vitamin C in plant tissues. The proposed biosynthetic pathway of vitamin C is described below in Figure 1.29. ATP  Gic  ADP  1  GTP  GIc-6-P  _  2  Fru-6-P  3  Man-6-P  Man-I -P 5  4  ?  L-GuI-l,4-Lac—-GuI  PP  ?  GDP-D-Man 6 GDP-L-GuI 6  Ascorbic Acid  NADH  NAD  L-GaI-I-P NADH  Figure 1.29  NAD  P  GMP  Proposed de novo biosynthesis pathway of vitamin C in plants. 64  GDP-L-GaI  39 The D-mannose/L-galactose pathway is in black, the proposed L-gulose pathway is in gray, and the reaction catalyzed by GME is boxed. It is found that the origin of vitamin C can be traced all the way from glucose. First a hexokinase (#1 in Figure 1.29) converts glucose (Gic) into glucose-6-phosphate (Glc-6-P), and then phosphoglucose isomerase (#2 in Figure 1.29) converts it to fructose-6-phosphate (Fru-6-P). Fructose-6-phosphate is later converted into mannose-6-phosphate (Man-6-P) and mannose-1-phosphate (Man-1-P) by the actions of phosphomannose isomerase (#3 in Figure 1.29) and phosphomannose mutase (#4 in Figure 1.29) respectively. Then a GMP is attached to mannose-1-phosphate by GDP-mannose phosphorylase (#5 in Figure 1.29) to yield GDP-D-mannose (GDP-D-Man). GME (#6 in Figure 1.29) takes GDP-D-mannose and converts it to GDP-L-gulose (GDP-L-Gul) and GDP-L-galactose (GDP-L Gal) in this vitamin C biosynthesis committing reaction step. 92 The GDP-L-galactose is then converted into L-galactose-1-phosphate (L-Gal-1-P) and later L-galactose (L-Gal) by the action of GDP-L--galactose pyrophosphatase (#7 in Figure 1.29) and L-galactose- 1-phosphate phosphatase (#8 in Figure 1.29) respectively. 94 Finally the L-galactose is converted into L-galactono- 1,4’ 93 lactone (L-Gal- 1 ,4-Lac) and eventually ascorbic acid by L-galactose dehydrogenase (#9 in Figure 1.29) and L-galactono- 1 ,4-lactone dehydrogenase (#10 in Figure 1 .29)92 95-98 GDP-L-gulose is also thought to be converted into ascorbic acid but the route is less clear. It is thought that the GDP-L-gulose is hydrolyzed by an unknown process to yield L-gulose (L-Gul in Figure 1.29). Then the L-gulose is taken by the same L-galactose dehydrogenase (#9 in Figure 1.29) to produce L-gulono-1,4-lactone (L-Gul-1,4-Lac), and finally a L-gulono-1,4-lactone dehydrogenase (#11 in Figure 1.29) converts L-gulono- 1 ,4-lactone to vitamin C. 96 In either pathway GME plays the important role of introducing the required L-stereochemistry in vitamin C biosynthesis. In addition, GME is a remarkable enzyme that attracts enzymologists due to its capability to perform four sequential catalysis in one single active site.  40  1.9  Project Goals  The aim of this thesis is to study two similar enzymes that catalyze quite distinct reactions: GDP-fucose synthase (GFS) and GDP-mannose 3,5-epimerase (GME). The first goal of this research is to investigate the mechanism of the GFS reaction and to elucidate the order of epimerizations. In addition, we wished to identify the acid and base residues responsible for these stereochemical inversions. The assignment of the catalytic residues as Cys 109 and His 179 is tentative due to the absence of an x-ray crystal structure with bound substrate. Furthermore, there have only been limited mechanistic studies performed on GFS. The only piece of evidence supporting the deprotonation / reprotonation mechanism was a single C-3” tritium label washout experiment. Surprisingly, no information on C-5” solvent derived deuterium incorporation has been reported for this enzyme. Crystallography studies on GFS suggest that only two catalytic residues are responsible for both epimerizations, yet it is conceivable that as many as four different catalytic residues could be involved. In an attempt to identify the role of the catalytic residues, site-directed mutagenesis will be performed to generate both Cys and His mutants of GFS. The reactions of both mutants will be studied and the products will be analyzed. The mutation of a major catalytic residue may cripple the enzyme and premature reduction of the intermediate could give a new product that would indicate the sequential order of the epimerizations. In addition, the role of both potential catalytic residues could be assigned by monitoring deuterium wash-in / wash-out experiments from either the substrate (keto sugar) or the product (GDP-L-fucose) using the mutants. Finally, kinetic isotope effect (KIE) studies on the mutants using either C-3” or C-5”  41  deuterated substrates may not only provide information on the order of epimerizations, but also on the relative barrier heights of the individual steps of catalysis. The other goal of this research is to investigate the mechanism of GME and try to provide mechanistic evidence to support the conclusions that were drawn from previous structural studies. The identity and roles of putative catalytic residues were assigned only based only on an examination of x-ray crystal structures, which only present a static picture of the active site and may not reflect the dynamic reaction mechanism. In order to study the roles of these residues, a series of deuterium wash-in experiments will be performed on both wild-type and mutant forms of GME. To study wash-in during the forward reaction, the substrate GDP-D-mannose will be employed. To study the reverse reaction, the product analogue, GDP-L-fucose will be used. It is anticipated that the results of these mechanistic studies will come to a converging conclusion with the sequential epimerization mechanism that has been proposed for the GME reaction.  42  Chapter 2 Mechanistic Studies with Wild Type GDP-L-Fucose Synthase  43  2.1  Introduction  At the opening of this chapter, the cloning of the genes and the purification of enzymes involved in the biosynthesis of GDP-L-fucose are presented. The early sections in this chapter discuss the synthesis and purification of GDP-L-fucose. Although GDP-L-fucose can be easily made from GDP-mannose using the enzymes GMD and GFS, the purification of the product away from nicotinamide cofactors proved to be problematic. Ultimately, pure GDP-fucose was obtained and a solvent-derived deuterium wash-in experiment was performed. The middle sections in this chapter describe kinetic studies of the GFS reaction. Both C-3” and C-5” deuterated substrates were prepared and used to probe whether kinetic isotope effects are observed during catalysis. The work described in this portion of the chapter lays down the foundation for future studies on the GFS mutants described in Chapter three. Finally, GFS, as introduced in the previous chapter, promotes two epimerizations during catalysis. However, it remains a mystery whether these epimerizations proceed in a strictly ordered sequence, and if so, which center is epimerized first. In order to differentiate between these possibilities, a mechanistic probe, GDP-2-deoxy-2-fluoro-D-mannose, was chemically synthesized and tested. The last portion of Chapter two explains the logic behind this mechanistic probe and presents the results of this experiment.  44  2.2  Cloning, Overexpression, and Purification of GMD and GFS  Both recombinant GMD and GFS were cloned from E. coil K12 genomic DNA using the Xa ligation independent cloning (LIC) kit from Novagen. To begin the cloning process, primers for both GMD and GFS with typical lengths between 27-30 base pairs were designed. One end of the primers was specified by the cloning kit and was complementary to the vector, while the other end of the primers was designed to complement either end of the target gene. Then the gene encoding the target enzyme was amplified by polymerase chain reaction (PCR) with the use of the primers. The final PCR product was a blunt-ended linear piece of double stranded DNA containing the target gene (Figure 2.1).  TARGET GENE  TAA  S. to .r ra ihruh  R (orary) G I E 5’ GGTATTGAGGGTCGCAflI 3’ CCATAACTCCCACGThC  J. XXXGGCTCTAACTCTCCTCT 3’ CC5AGATTGAGAGBAA 5  T4DNApo+dGTPonIy G R I (ary GGTATTGAGGGIC6CATG GCGTAC  T  GET  ENE  CC4GATTAGAGAGA  L!CXa insert C it TAA CT C TC C it T  AGGCCATAACTCCCA  LIC plasmid Annealing, transtormation G  P  <oran) I E AT G AGGCCATAACTCCCAGCG TAC TCCTTCGC  ARGETGENE  XXXGGCTCTPACT CTCCTCT CCGAGATT GAAGAGA  r9c3mDinan1 plasmid  Figure 2.1  Cloning of a gene using the XaILIC cloning kit (Novagen).  45  The commercially available vector from the cloning kit has single stranded DNA at both ends for the purpose of efficient ligation and is termed “sticky ended” (LIC plasmid in Figure 2.1). This vector also contains kanamycin resistant gene, lac operon sequence, and a hexa histidine tag sequence for the ease of selection, over-expression, and purification, respectively. In order to dock the target gene into the vector, complementary “sticky ends” on the amplified gene are required and were generated using T4 DNA polymerase. T4 DNA polymerase not only has 5 ‘-3’ polymerase activity but also has 3 ‘-5’ exonuclease activity. By saturating the reaction mixture with dGTP, the T4 DNA polymerase is able to cleave DNA bases from both 3 ‘-ends of the amplified gene (this PCR product is purified and any dNTP is removed using the PCR cleanup kit from Novagen) until it reaches the first guanosine in the sequence. The PCR primers are designed to ensure that the resulting overhangs are complementary to those on the vector. The polymerase digested amplified gene is then annealed with the vector to produce a doubly nicked plasmid. This doubly nicked plasmid is then transformed directly into competent E. coil cells. The plasmid is ligated within the cell and is subsequently reproduced as the cells divide. Since the vector itself contains a kanamycin resistant gene, it is possible to select for cells containing the desired plasmid. Plasmids containing the target gene are then identified by PCR. The sequence of the gene is verified at the Nucleic Acid Protein Service Unit (NAPS) at UBC to ensure no mutations or errors are introduced throughout the overall process. The plasmid is then transformed into E. coil BL2 1 (DE3) competent cells for protein overexpression and the cells are grow1 in the presence of kanamycin. IPTG (isopropyl f3-D-1thiogalactopyranoside) is added to induce the overexpression of protein. This induction is possible because the vector used contains a lac operon upstream from the target gene. The mechanistic principle of lac operon is briefly explained here (Figure 2.2).  46  Operon off (÷ Glucose, Lactose) -  —  DNA polymerase cannot bind Repressor bound to operator  \‘“  •-  Active repressor  --  IF  Operator  /  --  —-__________  Gene._____ DNA  0 0  Operon on(- Glucose, + Lactose)  0  Inducer (lactose)  Inducer bound to repressor  DNA polynierase 1 Repressor cannot bind to operator  Operator Transcription  Operator  Figure 2.2  LeIle mRNAtranip,.._  Gene  Schematic representation of lac operon mechanism (Access Excellence at the National Health Museum).  In the absence of lactose, the expression of the gene is inhibited due to the binding of a repressor at the gene’s operator. This repressor prevents T7 RNA polymerase from binding to the T7 promotor region of the gene and, thus, no transcription can occur. When lactose or IPTG (a potent lactose analogue) is introduced into the system, it binds to the repressor and inactivates it. T7 Polymerase can now bind to the promotor to initiate transcription and the gene encoding the desired protein is expressed. Under normal circumstances, the expression of the gene is regulated as the level of lactose is adjusted by the cell. However, since IPTO is resistant to cellular metabolism, the expression of the gene cannot be turned off. After induction with IPTG, the cells are grown for another 4 hours or overnight for GMD or GFS, respectively, and then harvested and lysed. The soluble portion of the lysate is then applied to a nickel affinity column. Due to the presence of an N-terminal hexahistidine tag, the  47  desired protein remains bound to the resin while the rest of the lysate passes through the column. Finally, a buffer containing 0.5 M imidazole is used to elute the enzyme. Fractions containing the enzyme are exchanged into a storage buffer containing 10% glycerol and stored at -80 °C. GMD and GFS can be stored for 3 months and 6 months, respectively, without significant loss in activity. As SDS-PAGE analysis shows that GFS obtained in this manner is >90 % pure (Figure 2.3)  E-66 kDa  E—29  Lane# 1 Figure 2.3  2  3  4  5  kDa  6  SDS-PAGE showing the purification of GFS. Lanes: 1 and 6) Molecular weight standards carbonic anhydrase (66 kDa) and bovine serum albumin (29 kDa). 2) Crude cell lysate after over-expression. 3 and 4) Ni 2 column flow-through (5 mM and 100 mM imidazole, respectively). 5) Purified GFS.  48  2.3  Enzymatic Synthesis and Purification of GDP-L-Fucose  In order to ensure that both the GMD and GFS enzymes were active and to produce an authentic sample of GDP-L-fucose, the enzymatic synthesis of GDP-L-fucose from GDP-D mannose was performed. Samples of GDP-L-fucose are also required to perform solvent deuterium wash-in experiments for reactions run in the reverse direction (the principle of this experiment will be discussed in later a section in this thesis). It was found that the recombinant GMD and GFS were quite active in producing GDP-L-fucose, however the purification of GDP L-fucose was problematic.  2.3.1 Synthesis of GDP-L-Fucose Using NADPH  Since the substrate of GFS, GDP-2-deoxy-4-keto-D-mannose, was known to be relatively unstable and could easily decompose by loss of the GDP group,’° 6 the synthesis of GDP-L fucose was performed by incubating both GMD and GFS together with GDP-D-mannose in the presence of excess NADPH. One problem was that the reaction done in this way required a long time to complete due to the fact that the final product of this pathway, GDP-L-fticose, is a competitive, negative feedback inhibitor of GMD. 85 Nevertheless, monitoring the reaction by negative ESI mass spectrometry indicated that it could be driven to completion. Attempts to  purify the product mixture by anion exchange chromatography were unsuccessful as it was found  49  that the one equivalent of NADP produced in the reaction always co-eluted with GDP-L-fucose (Figure 2.4).  NADP’diphosphate + GDP-L-fucose cx phosphate  A  NADP GDP-L-Fucose  ppm  Figure 2.4  —1050  -11.00  —11.50  -1250  P NMR Spectrum of GDP-L-fucose produced using NADPH. 31  After many attempts with different elution gradient methods, no improvement in the purification was made. In the later trials, the one equivalent of NADP was reduced by NaBH 4 (after the removal of enzyme) to give a final mixture of GDP-L-fucose and NADPH. This reaction was quenched by the addition of acetone and the mixture was subsequently lyophilized prior to anion exchange chromatographic purification. It was anticipated that NADPH could easily be separated since it has a greater negative charge than GDP-L-fucose. This was partially successful, however, a small amount of NADP still present in the product (Figure 2.5).  50 NADPdiphosphate + GDP-L-fucose a phosphate  GDP-L-Fucose  / ppm  Figure 2.5  -11.0  -12.0  -13.0  P NMR spectrum of GDP-L-fucose produced using NADPH followed by 31 4 reduction. NaBH  It was postulated that either the reduction of NADP did not proceed to completion or that 4 the NADPH was oxidized during the handling and purification of the mixture. Repeated NaBH reduction was also attempted on “purified” product but no improvement was made.  2.3.2 Synthesis of GDP-L-Fucose Using NADH  From the literature, it was known that GFS is capable of utilizing NADH instead of NADPH, however the reaction proceeds at a much slower rate. 60 Therefore, the synthesis was  51  repeated using NADH and an extended incubation time. The reaction did proceed to completion as analyzed by negative ESI-MS, however, it caused another problem in purification. The excess NADH (signal at -10.7 ppm in the 31 P NMR spectrum) also bears two negative charges and co elutes with the product in ion exchange chromatography (signals at -10.7 and -12.5 ppm in the P NMR spectrum) (Figure 2.6). Attempts to alter the eluting buffers and gradients did not 31 improve the separation.  GDP-L-Fucose + NADH  I  I  -9_o  -100  GDP-L-Fucose  -itO  -12.0  -13.0  ppm  Figure 2.6  P NMR spectrum of GDP-L-fucose produced using NADH. 31  -140  52  2.3.3 Enzymatic Synthesis and Revised Purification of GDP-L-Fucose  Finally, the synthesis and purification of GDP-L-fucose was achieved through the use of alkaline phosphatase. It is known that alkaline phosphatase can hydrolyze the 5’ phosphate group from sugar nucleotides.’° 7 However, it is also capable of hydrolyzing the 2’ phosphate group from sugar nucleotides such as NADP. This was confirmed with a simple 31 P NMR experiment by monitoring the release of free phosphate when NADP was incubated with the phosphatase in a pH 8 triethanolammonium buffer (Figure 2.7).  2’ Phosphate  Before P hos p hata se Addition  I  After Phosphatase Addition  :  Free Phosphate  LMJi ppm  Figure 2.7  ,  5.0  —  0.0  -5.0  -10.0  P NMR spectra monitoring the hydrolysis of the 2’-phosphate group from 31 NADP by alkaline phosphatase.  The use of alkaline phosphatase also solved some other difficulties in the original GDP L-fucose syntheses. In previous attempts, GMD and GFS were both present at the same time and GMD was inhibited by the GDP-L-fucose produced. As a result, the reaction required a lengthy incubation to reach completion. Also it was noticed that GMD can slowly exchange its bound  53 NADP cofactor with exogenous NADPH and reduce the GDP-6-deoxy-4-keto-D-mannose product to give GDP-6-deoxy-D-mannose. By separating the GMD and GFS/NADPH treatment into two sequential steps both problems are alleviated. Finally, the phosphatase digestion step in the new synthesis eliminates the problem of GDP formation from the decomposition of the 4-keto intermediate. Any GDP formed will readily be hydrolyzed to give guanosine which is easily separated. GDP-D-mannose was incubated overnight with GMD in order to guarantee a complete conversion to GDP-6-deoxy-4keto-D-mannose. Then GMD was removed and GFS and excess NADPH were added. Upon the completion of the synthesis and removal of GFS, anion exchange chromatography was used to remove excess NADPH. This gave a mixture of containing both one equivalent of NADP and GDP-L-fucose. This mixture was then buffered to pH 8 and alkaline phosphatase was added. After overnight incubation, this mixture was re-purified with a second anion exchange chromatography to remove the NAD and yield pure GDP-L-fucose that was estimated to be 90% pure by ‘H NMR analysis. Figure 2.8 shows the overall purification scheme and Figure 2.9a and 2.9b show the ‘H NMR and 31 P NMR spectra of purified GDP-L-fucose, respectively.  GFS NADPH  C 3 H NADP  0  OGDP OH  HO  +  OH  GDP-D-mannose  GDP-6-deoxy4-keto-D-mannose  +  G DP-L-fucose  I eq. NADP Excess NADPH st Anion 1  exchange chromatography COOGDP 3 H  1H0 OH  GDP-L-fucose  HO’ Anion OH exciiange GDP-L-fUCOSe chromatography +  1 eq. NAD Figure 2.8  COOGDP 3 H  COOGDP 3 H  Overall purification scheme of GDP-L-fucose.  1H0 OH priospnoiase GDP-L-fucose digestion  Alkaline i.  L  +  I eq. NADP  54  80  7.0  8.0  5.0  4.0  3.0  2.0  1.0  ppm  Figure 2.9a  ‘H NMR spectrum of purified GDP-L-fucose. Peaks at the 1.2 and 3.0 ppm are due to triethylammonium counter ions.  -100 ppml  Figure 2.9b  ’P NMR spectrum of pure GDP-L-fucose. 3  -150  55  2.4  Deuterium Wash-in Experiment with GDP-L-Fucose  In the past studies, it was demonstrated that the GFS reaction proceeds with the incorporation of solvent-derived deuterium at C-3  It remains to show that GFS also catalyzes  the exchange at C-5”. One way of showing this is to run the GFS reaction backward using GDP L-fucose and NADP in D O-exchanged buffer. This type of experiment will also prove to be 2 useful in chapter three using mutant forms of GFS. Even though the equilibrium of the GFS reaction strongly favors the final product GDP-L-fucose, it was anticipated that GFS would still oxidize GDP-L-fucose at C-4”, and epimerize both the C-3” and C-5” positions. If the reverse rate is significant the net observation would be solvent deuterium exchanged into the pooi of GDP-L-fucose (Figure 2.10).  HO OH  OGDP  GDP-L.fucose NADP  ..j  0 2 )f D  NADPH  Or? D  OGDP  H]..GDP4.. 2 L3, 5’.. deoxy.4-keto.Dmannose  OH  D  OGDP  OGDP  H].GDP.6. 2 [3’, 5”deoxy-4.keto-L-gulose  [3•, 5”. H1..GDP.6. 2 deoxy-4-keto-o.altrose I  NADPH NADP  HOO  OH  L3  Figure 2.10  OGDP  HFGDP-1-4ucose 2 5”-  Deuterium wash-in experiment with GDP-L-fucose.  OH  OGDP  HJ.GDP.4.keto.. 2 C3”, 5”. L-fucose  56 It was also conceivable that the order of the epimerizations could be determined by monitoring the rate of deuterium wash-in at the two sites of epimerization. A sample of GDP-L-fucose was incubated with GFS in buffered D 0 and the reaction was monitored by ‘H NMR spectroscopy 2 (Figure 2.11). The peaks corresponding to the H2” 5” signals of GDP-L-fucose are shown. The -  disappearance of the signals at 3.66 ppm and 3.76 ppm indicate that deuterium is incorporated at C-3” and C-5”. The rate of wash-in at the C-5” position appears to be about 2-fold faster than that at the C-3” position (Figure 2.11). The mass of the di-deuterated GDP-L-fucose was also confirmed by negative ESI-MS.  t=Omin  H4  H3  H5  H2  H3C_jOOGDP  GDP-L-Fucose t= 15 mm  ppm  Figure 2.11  3.80  3.70  3.60  3.50  ‘H NMR spectra showing deuterium wash-in with GDP-L-fucose.  This experiment confirms that the GFS reaction proceeds with solvent exchange at both C-3” and C-5” indicating a deprotonation / reprotonation mechanism is at play in both cases.  57  Even though the relative rates of wash-in are not dramatically different, one can see that the H-5” signal disappears about two times faster than H-3” signal. This suggests that if GFS employs an ordered mechanism, then in the forward reaction direction it catalyzes a C-3” epimerization first followed by a C-5” epimerization. Later studies in this thesis confirm that it is, indeed, the correct order of epimerization.  2.5  Kinetic Study of GFS  Additional studies on wild-type GFS involved the measurement of kinetic constants and kinetic isotope effect (KIE) to determine if deprotonation of the C-H bonds were involved in rate determining steps. Previous studies showed that the reaction of [4H]-NADPH was slowed by a 2 kinetic isotope effect (KIE) on kcat of kH/kD = 1  460  The small value of this primary KIE indicates  that the hydride transfer step is only partially rate-determining in catalysis. However, it is also possible that either one of the individual epimerizations or even both epimerizations are also partially rate-determining. In this section of the thesis, kinetic studies on the GFS reaction using unlabeled and [5”H]- or [3”2 H]-GDP-6-deoxy-4-keto-D-mannose are presented. The use of 2 deuterium labeled substrates will probe whether deprotonation steps are rate-limiting and may also suggest the relative energies difference between the epimerization and reduction steps.  58  2.5.1 Kinetic Study with C-5” Deuterated Substrate  The first set of kinetic isotope effect studies were performed on [5”Hj-GDP-6-deoxy-42 keto-D-mannose due to the relative ease of obtaining this compound. This substrate can be made by simply running the GMD reaction in D 0-exchanged buffer (Figure 2.12). 2 NADP  OGDP  GDP-D-mannose  OGDP  ketone intermediate 2 H NADP—’W  NADP D20  OGDP  enone intermediate  Figure 2.12  OGDP  GDP-6-deoxy-4-keto-D-mannose  Synthesis of [5”H]-GDP-6-deoxy-4-keto-D-mannose using GMD. 2  In the GMD reaction, oxidation and dehydration first produces an enone intermediate. In the final step the reduced cofactor transfers a hydride to C-6” and C-5” is protonated by an acidic residue bearing a solvent-derived deuterium to give the final product. To measure GFS kinetics, a sample of GDP-D-mannose was treated with GMD in buffer either prepared from H 0 or D 2 0. 2 The reactions were monitored by negative ESI-MS to ensure that they proceeded to completion (Figure 2.13). After removal of the enzyme, the reactions were lyophilized and resuspended in 0. Negative ESI-MS confirmed that no significant decomposition occurred during 2 H lyophilization (lack of GDP peak) and that non-enzymatic wash-out of the label did not readily  59  occur. Due to the sensitive nature of this keto sugar, the substrates were used immediately in kinetic studies without further purification. The concentration of the substrates was determined using UV spectroscopy.  581Q  10  5059  ---  580  5  584  Figure 2.13  586  588  590  592  594  506  508  nz  __________j  580  5  584  585  588  500  862  594  506  598  Negative ESI-MS of unlabeled and [5”H]-GDP-6-deoxy-4-keto-D-mannose. 2  The kinetic studies were performed by monitoring the consumption of NADPH which absorbs at a wavelength of 340 nm. The amount of NADPH was fixed at a saturating level (100 tM) while the concentration of the keto substrates were varied. All the compounds were incubated in pH 7 sodium phosphate buffer at 25 °C and the GFS reaction was initiated by the addition of enzyme. The initial velocities obtained in the studies were plotted against the substrate concentration and the graph was fitted to Michealis-Menten kinetics (Figure 2.14).  60  03  0 V  iM!min) 1 ( 0.2  0 20  0  40  60  [Substrate] Figure 2.14  80  (riM)  Kinetic plots of the GFS reaction with unlabeled and [5”H]-GDP-6-deoxy-42 keto-D-mannose.  The Michealis-Menten parameters of the GFS reaction were calculated and it was found that both substrates reacted at the same rate (Table 2.1). Thus, no KIE was observed in the reaction of GDP-6-deoxy-4-keto-D-mannose indicating that the C-5 “- epimerization, or more specifically the C-5 “- deprotonation, is not rate-determining in catalysis. Unlabeled-Substrate Km  2.2 -i-f- 0.2uM 0.65  Table 2.1  +1- 0.02 s  H] -Substrate 2 [5 “2.2 0.64  +1- 0.luM -‘-I- 0.01 s’  NADPH 1.5  -i-I- 0.luM  0.74 -i-f- 0.02 s’  Kinetic parameters obtained with unlabeled and [5 “H]-GDP-6-deoxy-4-keto-D2 mannose and NADPH.  61  A kinetic study with NADPH as the variable substrate and GDP-6-deoxy-4-keto-Dmannose held at a fixed concentration of 150 jiM was also performed. The resulting kinetics (graph is not shown) provided a value of Km for NADPH (Table 2.1).  2.5.2 Kinetic Study with C-3” Deuterated Substrate  To complete the story, the kinetic isotope effect was also investigated using [3 “H] 2  -  GDP-6-deoxy-4-keto-D-mannose. Even though C-5” epimerization is not rate-determining, it is still possible that C-3” epimerization could be. In order to gain access to [3”H]-GDP-6-deoxy2 4-keto-D-mannose, GDP-mannose 3, 5 epimerase (GME) was employed. An overexpression plasmid encoding for the enzyme was donated to us by Prof. James H. Naismith and was used to obtain the enzyme. 64 As discussed in Chapter one, section 1.6.5, this enzyme will convert GDP D-mannose into an 80: 5: 15 mixture of GDP-D-mannose, GDP-L-gulose, and GDP-L-galactose, respectively. When this reaction is run in D 0 all species will contain deuterium at C-3” and C-5” 2 after equilibration (Figure 2.15).  HO1  I::lcEIII.  HOEL0  HO+ oHOGDP  GDP-D-Mannose  Figure 2.15  G DP-D-Man nose 80%  GME catalysis in D 0 buffer. 2  G DP-L-Gulose  G DP-L-Galactose  5%  15%  62  It was anticipated that only the major species, GDP-D-mannose (80%), would be accepted by GMD in a subsequent reaction and converted into substrate. The remaining 20% would remain as an impurity in the sample. The reaction of GDP-D-mannose with GME in D 0 proceeded as expected and negative 2 ESI-MS confirmed that all species bore two deuterium atoms. This mixture containing 80% [3”, H]-GDP-D-mannose was purified by size exclusion chromatography and was then incubated 2 5”with GMD in H 0 buffer. Approximately 80% of the material underwent dehydration as 2 monitored by negative ESI-MS, suggesting that GMD selectively processed the [3”, 5 “H]2 GDP-D-mannose and left the other two species in the mixture untouched. Since the GMD reaction washes out isotope at C-5”, the C-3” deuterium remained to yield a 80% pure sample of [3”, 5”H]-GDP-D-mannose for kinetic analysis (Figure 2.16). 2 NADP  OH OH  NADPH GMD  H( H(  HO D  OGDP GDP-D-mannose  D  OGDP  HOD, NADP  NADPH  HO HO 0  Figure 2.16  OGDP  D G DP-6-deoxy-4-keto-D-man nose  Synthesis of [3”H1-GDP-6-deoxy-4-keto-D-mannose using GMD. 2  As a control, the unlabeled substrate was also made under identical conditions with the exception of using buffered H 0 in both enzymatic reactions. This ensures that the 20% 2  63  impurities present in the labeled sample will also be present in the unlabeled sample and will not result in the misinterpretation of a TUE because of inhibition. The completion of the GMD reaction and the identity of the substrates were confirmed by negative ESI-MS (Figure 2.17).  HH D  587 9 589 0  OGDP  GDP-L-gulose + GDP L galactose  OH  kHL  H  .  587.0 0  580.0 582.5 585.0 587.5 590.0 592.5 595.0 597.5 600.0 602.5 605.OmIz  Figure 2.17  0  56.0  OH  587 9  OGDP  GDP-L-gulose + GDP L galactose  9  580.0 582.5 585.0 587.5 590.0 592.5 595.0 597.5 600.0 602.5 605.OmIz  Negative EST-MS of unlabeled and [3”H]-GDP-6-deoxy-4-keto-D-mannose. 2  It was noted that in the production of [3”H]-GDP-6-deoxy-4-keto-D-mannose, a very 2 small signal due to unlabeled GDP-6-deoxy-4-keto-D-mannose (mlz = 586) was always observed. This indicates that the C-3” deuterium may gradually undergo a non-enzymatic exchange with solvent. Nevertheless, it was anticipated that it would have minimal influence on the kinetic studies due to its low abundance (<5%). The kinetic studies were performed under the identical conditions as in the C-5” kinetic isotope effect study and the graph of initial velocities vs. substrate concentrations were fitted to Michaelis-Menten kinetics as before (Figure 2.18).  64 012  0.10  008 0 V  (tMImin)  0.06  0M4  0.02  0  [Substrate] (saM)  Figure 2.18  Kinetic plots of GFS with unlabeled and [3 “H]-GDP-6-deoxy-4-keto-D2 mannose.  The Michealis parameters of the GFS reaction using unlabeled and [3”H]-GDP-62 deoxy-4-keto-D-mannose were calculated (Table 2.2). It was found that there was no KIE observed and, thus, the C-3” deprotonation is not rate determining in the overall catalysis. The H]-GDP-6-deoxy-4-keto-D2 kcat values were similar to those obtained in the study with [5”mannose, however the KM values were somewhat lower. The variability could be due to differences in the technique for substrate preparation and to the presence of 20% GDP-L galactose or GDP-L-gulose. Nevertheless, the deuterated and non-deuterated samples prepared under identical conditions showed indistinguishable kinetics, indicating there was no KIE observed during catalysis.  65  Unlabeled-Substrate  Table 2.2  [3 “H] -Substrate 2  Km  0.45  +1- 0.O3uM  0.43  kcat  0.71  1 ÷/ 0.01 S•  0.69 +/ 0.01  +1- 0.O2uM 51  Kinetic parameters obtained with unlabeled and [3”H]-GDP-6-deoxy-4-keto-D2 mannose.  2.6  GDP-2-Deoxy-2-Fluoro-D-Mannose as a Mechanistic Probe  2.6.1 Introduction and the Experimental Design  As discussed previously, the detailed catalytic mechanism of GFS remains unknown due to the lack of x-ray crystal structures with bound GDP-hexoses. Since both sites of epimerization on GDP-4-keto-6-deoxy-D-mannose are already activated by the presence of C-4” carbonyl, it is assumed that the epimerization occurs through deprotonation / reprotonation steps via an enolate intermediate (Figure 2.19). This is supported by the deuterium wash-in experiments described in section 2.4.  66  R  B  Figure 2.19  BH  Epimerization via enolate intennediate.  Three possible scenarios could describe the order of epimerization: 1) C-3” may be inverted first, 2) C-5” may be inverted first, or 3) the ordering may be random (Figure 2.20).  C-3”  Epimerization  HO-  NADPH NADP C-5’ Epimerizatior  GDP-L-fucose  Figure 2.20  Possible mechanisms of the reaction catalyzed by GFS.  In order to address this uncertainty, a substrate analogue GDP-2-deoxy-2-fluoro-Dmannose was chemically synthesized and tested with the enzyme (Figure 2.21).  67  Figure 2.21  The structure of GDP-2-deoxy-2-fluoro-D-mannose.  Since fluorine is slightly smaller than a hydroxyl group, it was anticipated that the first enzyme in the pathway, GMD, would still recognize GDP-2-deoxy-2-fluoro-D-mannose as a substrate and catalyze the oxidation at C-4’ as well as the dehydration to produce GDP-2,6 -deoxy-2fluoro-4-keto-D-mannose. This substrate analogue could then be incubated with GFS and an analysis of the resulting product should help in elucidating the sequence of epimerizations. Three possible scenarios were expected from this experiment (Figure 2.22).  68 OH 2 CH  3 CH  0 OH  GDP-2-deoxy-2-fluoro-4keto-6-deoxy-D-mannose  GMD  F -  0 OGDP GDP-2-deoxy-2-  GDP  0 2 H  GDP-2-deoxy-2-fluoro-4keto-6-deoxy-L.-glucose  fluoro-D-mannose  0:  HF 0  3 CH  1  NADPH OGDP  * 3 H  1-10  NADP  *  0 3 CH  OGDP  0 4  2  F  OH  GDP-2-deoxy-2fluoro-L-fucose  Figure 2.22  Possible products obtained from incubating GDP-2-deoxy-2-fluoro-D-mannose with GFS and NADPH.  If C-3”is normally epimerized first, deprotonation at C-3” could result in the elimination of HF across C-2” and C-3” to generate compound 1 (Path A, Figure 2.22). A subsequent tautomerization would form a new carbonyl at C-3” and give compound 2. Since compound 2 has two carbonyl groups in the sugar ring and adopts a very different conformation than the normal substrate or intermediate does, it is anticipated that it would be released from the active site before the second epimerization or the final reduction steps could take place. Conversely, if C-5”is normally epimerized first, this should readily occur to generate a 4-keto sugar nucleotide with L-stereochemistry. Subsequently, a deprotonation at C-3” and the elimination of fluoride would give compound 3 (Path B, Figure 2.22). A subsequent tautomerization would result in the  69  formation of compound 4 that would likely be released from the active site. Finally, it is also possible that GFS would simply accept GDP-2,6 -deoxy-2-fluoro-4-keto-D-mannose in the same manner as the normal substrate to give GDP-2-deoxy-2-fluoro-D-fucose, thus providing a new chemo-enzymatic synthesis of this compound.  2.6.2 Chemical Synthesis of GDP-2-Deoxy-2-Fluoro-D-Mannose  The chemical synthesis of GDP -2-deoxy-2-fluoro-D-mannose was performed using a previously published synthetic route (Figure 2.23).’°’ The commercially available tri-o-acetyl-D glucal was fluorinated using selectfluor in a mixture of water and DMF to give the 2-fluoro sugar 5 as a mixture of diastereomers.’° 2 The anomeric hydroxyls were acetylated by stirring in Ac 0 2 to yield a mixture of epimers 6 and 7. A flash chromatographic purification gave the desired manno- epimer 7. The controlled treatment of 7 with dimethylamine in acetonitrile selectively deprotected the anomeric hydroxyl to yield 8. Compound 8 was then phosphorylated using diphenyl chlorophosphate and DMAP in methylene chloride to give 9.  70 OAc  OAc -OAc  Selectfluor AcO  0, DMF 7 H  F  AcO0\ AcO  OH  5 ÷ OAc  OAc F  AcO  (PhO)-,P(O)CI DMAP, CH,CI,  NH 2 Me  AcO  acetonitriie 90%  85%  9  OP(0)(OPh)  1) H,, PtO, N, H,O 3 Et  8  HO  15%  10  2.23  OH  pyridine GMP-morpholidate 1H-tetrazole  OH  85%  Figure  OAc  11  OPO 2 3 NH), 3 (Et  OGDP  Synthesis of GDP-2-deoxy-2-fluoro-D-mannose.  Hydrogenation of 9 at atmospheric pressure in the presence of platinum oxide as a catalyst deprotected the phosphate group. Then, without further purification, the compound was deacetylated in a mixture of water and triethylamine to yield Treatment of 10 with GMP-morpholidate GDP-2-deoxy-2-fluoro-D-mannose  10 as a triethylammonium salt.  and 1H-tetrazole in pyridine gave the final product  (11). The masses of all the numbered compounds above were  confirmed by negative electrospray ionization (ESI) mass spectrometry.  The ESI mass spectrum and 31 P NMR spectrum of GDP-2-deoxy-2-fluoro-D-mannose  are  shown in Figure 2.24  and Figure 2.25, respectively. An impurity is seen in the latter spectrum  that is attributed to a  GMP dimer formed during the morphorlidate coupling step of the synthesis.  It  was not possible to remove this impurity using ion exchange chromatography. Subsequent  71  studies using this chemically synthesized GDP-2-deoxy-2-fluoro-D-mannose 11 were therefore completed in the presence of this GMP dimer.  6059  j  OGDP 5C  G1 580  Figure 2.24  ! 585  590  595  690  605  6O  615  62  025  Negative ESI mass spectrum of GDP-2-deoxy-2-fluoro-D-mannose, 11.  G DP-2-deoxy-2-fI uoro-D-mannos€  ‘‘GMPDimer  pn  Figure 2.25  50  00  -5O  -100  -150  P NMR spectrum of GDP-2-deoxy-2-fluoro-D-mannose, 11. 31  72  2.6.3 Reaction of GDP-2-Deoxy-2-Fluoro-D-Mannose with GMD  In order to use GDP-2-deoxy-2-fluoro-D-mannose as a mechanistic probe with GFS, it was first necessary to convert it into the corresponding 6-deoxy-4-keto compound using the enzyme GMD. The reaction of GDP-2-deoxy-2-fluoro-D-mannose with GMD in 50 mM phosphate buffer at pH 7 was monitored by 31 P NMR spectroscopy (Figure 2.26).  t=Oh 2F analogue, 11.  t= 1 h  t=2 h  ‘1’Ai  \Ji  t=3h  i  ,  r  —  -s-a  /  -1aM  -15.0  ppm  Figure 2.26  GMD reaction with GDP-2-deoxy-2-fluoro-D-mannose 11 monitored by P NMR. 31  73  The 31 P NMR signals attributable to GDP (-6.0 and -9.5 ppm) appeared even at early incubation times suggesting that the 6-deoxy-4-keto compound was formed but was unstable and decomposed to release GDP (Figure 2.26). Further studies involved adding both GMD and excess GFS/NADPH at the same time with the aim of trying to intercept the keto sugar before it decomposed. However, all these experiments led to a similar decomposition of the keto sugar and no conclusive statement could be obtained using this mechanistic probe. The instability of this compound was not completely unsuspected as the parent compound, GDP-6-deoxy-4-ketoD-mannose is also sensitive to decomposition and will generate GDP upon prolonged storage.’° 6 The 2-fluoro substrate seems to have further decreased in stability although the mechanistic rationale for this is not entirely clear.  2.7  Conclusions  In conclusion, E. coli GMD and GFS were cloned and used to obtain pure GDP-L-fucose with the aid of alkaline phosphatase. A deuterium wash-in experiment on GDP-L-fucose showed that deuterium did wash-in to both sites of epimerization and that the C-5” proton signal disappeared faster than the C-3” proton signal. This result suggests that in the forward reaction direction C-3” epimerization occurs first followed by C-5” epimerization and a final reduction give GDP-L-fucose. This ordered mechanism is, indeed, verified by the later experiments in this thesis with the use of GFS mutants. Lastly, both C-3” or C-5”- deuterium labeled GDP-6-deoxy4-keto-D-mannose were made with the use of GMD or a combination of GME and GMD,  74  respectively, in different D 0/H 2 0 buffers. Kinetic studies with these two sets of labeled 2 substrates showed no kinetic isotope effect suggesting that neither C-3” nor C-5” deprotonations are rate-determining in the overall catalysis. However, when these kinetic isotope effect studies are revisited with the use of a GFS mutant in chapter three of this thesis, interesting results were observed which also supported the proposed ordered epimerization mechanism. A mechanistic probe of GFS, GDP-2-deoxy-2-fluoro-D-mannose, was successfully synthesized. However, due to the decomposition of this compound upon treatment with GMD, it could not be used to probe the GFS mechanism.  2.8  Experimental  2.8.1 Materials and General Methods  GDP-D-Mannose, NADPH, and alkaline phosphatase (from bovine calf intestine) were purchased from Sigma-Aldrich. D 0 (99.9%) was purchased from Cambridge Isotope 2 Laboratories, Inc. GDP-mannose 3,5-epimerase was prepared by overexpression of the pEHISTEV-GME plasmid in E. coli Rosetta (DE3) cells and purified as described previously. 64 Protein concentrations were determined by the method of Bradford using bovine serum albumin 4 ‘H NMR spectra were obtained on a Bruker AV300 spectrometer at a field as the standard.’° strength of 300 MHz. Proton decoupled 31 P NMR spectra were recorded on the same  75 spectrometer at 121.5 MHz. Mass spectra were acquired at the Mass Spectrometry Centre at the University of British Columbia (UBC) by electrospray ionization (ESI-MS) using a Bruker EsquireLC ion trap mass spectrometer. Oligonucleotides were synthesized by the Nucleic Acid Protein Service (NAPS) facility at UBC.  2.8.2 Cloning, Over-expression, and Purification of GMD and GFS  The GMD gene was PCR amplified from E. coli K12 genomic DNA using Platinum Taq polymerase (Invitrogen) and the following two primers: 5’GGTATTGAGGGTCGCATGTCAAAAGTC-3’ (forward sequence) and 5’AGAGGAGAGTTAGAGCCTTATGACTCCAGCG-3’ (reverse sequence). The amplified DNA was purified using a SpinPrep TM PCR Clean-up kit (Novagen) according to the manufacturer’s directions. The PCR product was cloned into a pET-30 Xa/LIC vector (Novagen) according to the manufacturer’s instructions to yield pT7GMD. The pT7GMD plasmid was amplified in NovaBlue GigaSingles competent E. coli cells (Novagen). This plasmid encodes for the expression of GMD bearing an N-terminal 43-residue peptide that terminates in a hexahistidine tag. An identical procedure was used to clone the E. coli GFS gene and produce pT7GFS. PCR amplification employed the following two primers: 5’GGTATTGAGGGTCGCATGAGTAAACAACG-3’ (forward sequence) and 5’AGAGGAGAGTTAGAGCCTTACCCCCGAAAG-3’ (reverse sequence). To generate GMD, the pT7GMD plasmid was transformed into chemically competent BL21 (DE3) E. coli. Cells were grown on Luria-Bertani (LB) agar plates containing kanamycin  76 (30 mg/L) at 37 °C for 18 h. A single colony was used to inoculate 10 mL of LB medium containing 30 mg/L kanamycin and this was incubated for 12 h at 37 °C with shaking at 225 rpm. This culture was poured into 500 mL of LB medium containing 30 mg/L kanamycin and grown with shaking at 225 rpm and 37°C until an 0D 600 between 6.0 and 7.0 was reached. At that point, isopropyl-1-thio-f3-D-galactopyranoside (IPTG) was added to a final concentration of 0.3 mM and the cells were grown for an additional 4 hours before harvesting by centrifugation (6 000 rpm, 20 mm at 4 °C). Cell pellets were flash frozen in liquid N 2 and stored at -80 °C. The cells were resuspended in 15 mL of lysis buffer (20 mM sodium phosphate, pH 8.0, containing 2 mM 13-mercaptoethanol, 1 .tg/mL aprotinin, and 1 pg/mL pepstatin A) and lysed by passage through a French Pressure cell at 12 000 p.s.i. The lysate was clarified by spinning in a Sorvall SLA-1500 rotor at 6 500 rpm for 1 h followed by filtration (0.2 m pore, 33 mm syringe driven filter, Millipore). A column (20 mL) of Chelating Sepharose Fast Flow resin (GE Healthcare) was packed and flushed with distilled water. The column was charged with 2 column volumes , followed by washing with 2 CV of distilled water and 3 CV of start 4 (CV) of 100 mM NiSO buffer (20 mM sodium phosphate, pH 8.0, containing 0.5 M NaC1, and 5 mM imidazole). The clarified cell lysate was loaded at a rate of 2 mL/min, and start buffer was passed through the column at a rate of 3 mL/min until no more flow-through protein eluted as determined by monitoring . 280 To remove nonspecifically bound proteins, a mixture of 80% start buffer and A 20% elution buffer (20 mM sodium phosphate, pH 8.0, containing 0.5 M NaCl, 10% v/v glycerol, and 500 mM imidazole) was applied until no more protein eluted. Finally the histidine-tagged GMD was eluted with 100% elution buffer. The protein solution was then concentrated using Amicon Ultra centrifugal filter devices (Millipore). The concentrated enzyme solution was then divided into aliquots and flash frozen in liquid N . Enzyme could be stored at -80 °C for at least 2 2 months without significant loss in activity. Typically 50-60 mg of enzyme was purified from  77  500 mL culture. GFS (from pT7GFS) was generated and purified in an identical manner with the exception that the cell cultures were incubated for a total of 12 h following induction by IPTG. GFS could be stored at -80 °C for at least 3 months without significant loss in activity. Typical protein yields obtained from a 500 mL culture was 110 mg.  2.8.3 Enzymatic Synthesis of GDP-6-Deoxy-4-Keto-D-Mannose  To a mixture of 20 mg GDP-D-mannose, 0.3 mg NADP, and 1 tL f3-mercaptoethanol in  2 mL of sodium phosphate buffer (10 mM, pH 7.0, containing 10 mM NaC1) was added 20 mg of GMD (previously exchanged into 800 p.L of the same buffer using Amicon Ultra centrifugal filter devices (Millipore)). The reaction mixture was incubated for 12 h at rt and monitored by negative electrospray ionization mass spectrometry (-ESI-MS, GDP-mannose m/z = 604, GDP 6-deoxy-4-keto-D-mannose m!z  =  586) to ensure the reaction reached completion. Protein was  removed using Amicon Ultra centrifugal filter devises (Millipore), and the filtrate was lyophilized to yield GDP-6-deoxy-4-keto-D-mannose in a mixture with phosphate buffer salt. This material was used directly in all experiments.  2.8.4 Enzymatic Synthesis of GDP-L-Fucose  A sample containing 20 mg of the lyophilized GDP-6-deoxy-4-keto-D-mannose was re dissolved in 2.0 mL of distilled water, and 30 mg of NADPH and 20 mg of GFS (in 1.0 mL of  78  storage buffer) were added. The reaction was incubated at 28 °C for 2 h and monitored by negative electrospray ionization mass spectrometry (-ESI-MS, GDP-fucose m!z = 588, GDP-6deoxy-4-keto-D-mannose mlz  =  586) to ensure it reached completion. The protein was removed  using Amicon Ultra centrifugal filter devices (Millipore) and the filtrate was loaded onto a Dowex DE52 anion exchange resin column (1.5 cm diameter x 28 cm length). Product was eluted with a linear gradient of 0-0.3 M triethylammonium bicarbonate buffer, pFl 7.5. Eluent fractions were monitored at A 260 and GDP-L-fucose eluted at 0.18 M buffer as a mixture with NADP. NADPH eluted immediately afterwards. All the NADPH-free fractions were pooled and lyophilized. This mixture of NADP and GDP-L-fucose was re-dissolved in 10 mL of 20 mM Trien-HC1 buffer, pH 8.0, and 80 units of alkaline phosphatase was added. The resulting solution was incubated for 12 h at 30 °C and then lyophilized. A second anion exchange column was run using identical conditions, and the fractions containing GDP-L-fucose (well separated from those containing NAD) were pooled and lyophilized. The residue was re-dissolved in 15 mL of distilled water and lyophilized again to GDP-L-fucose (typical overall yield was 18-20 mg, as a bis-triethylammonium salt. GDP-L-fucose prepared in this fashion showed identical 03 —ESI-MS, 31 P NMR and ‘H NMR spectra to those previously reported in the literature.’  2.8.5 Deuterium Wash-in with GDP-L-Fucose  Deuterium wash-in experiments were performed using GFS-WT. A sample of deuterated phosphate buffer (10 mM, pD 7.4) was prepared by repeated (3X) lyophilization of a phosphate buffer (10 mM, pH 7.0) and dissolution in an equal volume of D 0. Sample containing 2.0 mg 2 GDP-fucose, 0.1 mg NADP, and 3 mg of enzyme (previously exchanged into deuterated buffer  79  using Amicon Ultra centrifugal filter devices (Millipore)) was prepared in the deuterated buffer (final volume 800 iL). The sample was incubated at 28 °C and the progress of deuterium washin was monitored using  LH  NMR spectroscopy. The mass of the final product was confirmed by  —ESI-MS upon completion of the deuterium wash-in process.  2.8.6 Kinetic Study of GFS  The [5”Hj-GDP-6-deoxy-4-keto-D-mannose was prepared in an identical fashion to the 2 unlabeled compound except that the GMD reaction was carried out in buffered D 0 and only 2 2 mg of GDP-D-mannose was used while keeping the amount of enzyme used the same (20-25 mg) (A control sample of unlabeled GDP-6-deoxy-4-keto-D-mannose was also prepared simultaneously in buffered H 0 under otherwise identical conditions). A deuterated sodium 2 phosphate buffer ((10 mM, pD 7.4, containing 10 mM NaC1) was prepared by lyophilizing a sample of non-deuterated buffer and re-dissolving it in an equal volume of D 0. This was 2 repeated three times to ensure complete isotopic exchange. The GMD reaction was run as described above (with previous exchange of GMD into the same deuterated buffer) and negative electrospray ionization mass spectrometry indicated that the reaction proceeded to completion and that> 97% of the product contained a single deuterium (-ESI-MS, GDP-mannose mlz = 604, [5 “H] -GDP-6-deoxy-4-keto-D-mannose mlz 2  =  587).  The [3”Hj- labeled substrate was prepared from a mixture containing [3, 5”2 H]-GDP-D2 mannose (80%), [3, 5”H]-GDP-L-gulose (5%) and [3, 5”2 H]-GDP-L-galactose (15%) that was 2 generated using GDP-mannose 3,5-epimerase in buffered D 0. A sample of Trien-DC1 buffer in 2  80  0 (20 mM, pD 7.4) was prepared by repeated (3X) lyophilization of a Trien-HC1 buffer (20 2 D mM, pH 7.0) and dissolution in an equal volume of D 0. To 2.0 mL of this buffer was added 25 2 mg of GDP-mannose, 1.0 1 iL of 13-mercaptoethanol, and 25 mg of GDP-mannose 3,5-epimerase (previously exchanged into 800 jiL of the same buffer using Amicon Ultra centrifugal filter devices (Millipore)). This reaction mixture was incubated for 12 h at 37 °C and monitored by negative electrospray ionization mass spectrometry to ensure that two atoms of deuterium were incorporated into >95% of the material (mlz  606). The protein was then removed using  Amicon Ultra centrifugal filter devices (Millipore). The filtrate was lyophilized, redissolved in 1.5 mL of distilled water, and loaded onto a size exclusion chromatography column (Biogel P-2, 200-400 mesh, 2.5 cm diameter x 70 cm length). The mixture of di-deuterated sugar nucleotides was eluted with distilled water and fractions absorbing at 260 nm were pooled and lyophilized. Non-deuterated control samples for use in KIE determinations were also prepared in buffered 0 under otherwise identical conditions. 2 H In order to obtain [3 “H]-GDP-6-deoxy-4-keto-D-mannose, the mixture of di-deuterated 2 epimeric GDP-hexoses was treated with GMD in buffered H 0 as described for the synthesis of 2 GDP-6-deoxy-4-keto-D-mannose. The unlabeled control sample for use in KIE studies was prepared in a similar fashion. The reactions were monitored by -ESI-MS and observed to stop after approximately 80% of the total material had been consumed, indicating that only GDP mannose served as a substrate for the GMD reaction. Mass analysis of the mono-deuterated products indicated that> 90% of the material contained an atom of deuterium (-ESI-MS, [3”H1-GDP-6-deoxy-4-keto-D-mannose mlz = 587). 2 Kinetic constants for the GFS reaction were obtained using a continuous spectroscopic assay that directly monitors the conversion of NADPH to NADP by following absorbance  81 changes at 340 nm. In the case of keto-substrate kinetics, assay mixtures contained 50 mM sodium phosphate buffer, pH 7.0, 100 .tM NADPH (saturating), and 0.25-10 jM GDP-6-deoxy4-keto-D-mannose in a total final volume of 1.0 mL. In the case of NADPH kinetics, assay mixtures contained 50 mM sodium phosphate buffer, pH 7.0, 150 tM GDP-6-deoxy-4-keto-Dmannose (saturating), and 0.25-10 .tM NADPH in a total final volume of 1.0 mL. The concentration of GDP-6-deoxy-4-keto-D-mannose stock solutions was determined by measuring 260 and using an A  E260  1 cm . Reactions were initiated by addition of 10 p1 of 1 of 11.8 mM  enzyme solution (containing 0.1 .tg of WT-GFS or 60 tg of C1O9S-GFS) and the decrease in 340 was monitored at 25 °C. Initial velocities were calculated using an A  834Q  . 1 of 6 220 M’ cm  Kinetic parameters were determined from fitting the initial velocities to the Hill equation using the program GraFit.  2.8.7 Synthesis of GDP-2-Deoxy-2-Fluoro-D-Mannose  Tetra-O-acetyl-2-deoxy-2-fluoro-a-D-mannopyranose F0 19 H 14 (C ) 9  2.2 g (8.0 mmol) of commercially available tri-O-acetyl-D-glucal was dissolved in 0!DMF (1 OOmL, 1:1) and 5.7 g (16 mmol) of freshly powdered SelectfluorTM was added to 2 H this solution. The mixture was then stirred at 40 °C for 8 hours. The solution mixture was then concentrated in vacuo followed by a usual workup in ethyl acetate to yield a colorless syrup.  82  This product was used without further purification. 10 mL of Ac 0 and 300 mg of ‘2 were added 2 to this syrup and this reaction mixture was then stirred at room temperature for 12 hours. Upon completion, the mixture was concentrated in vacuo. Flash chromatography (30% EtOAc / petroleum ether) was used to purify the product (0.9 g, 35%) and the spectroscopic data (‘H NMR) was consistent with those reported.’°’  Tri-O-acetyl-2-deoxy-2-fluoro-a-D-mannopyranose 8 F0 1 H 12 (C ) 7  0.9 g (2.7 mmol) of tetra-O-acetyl-2-deoxy-2-fluoro-a-D-mannopyranose was dissolved in 18 mL of freshly distilled acetonitrile. Then 10 mL of dimethylamine solution was slowly added slowly and the mixture was stirred at room temperature while the progress of the reaction was constantly monitored by TLC (40% EtOAc /petroleum ether). Upon the completion of reaction, the solvents were removed and the mixture was concentrated in vacuo. Flash chromatography (40% EtOAc /petroleum ether) yielded the hemiacetal product as a colorless oil (1.0 g, 90%) and the spectroscopic data (‘H NMR) was consistent with those reported.’°’  Diphenyl (Tri-O-acetyl-2-deoxy-2-fluoro- a-D-mannopyranosyl) phosphate F0 26 H 24 (C P) 11  0.76 g (2.5 mmol) of tri-O-acetyl-2-deoxy-2-fluoro-x-D-mannopyranose and 0.55 g of DMAP were dissolved in 20 mL of methylene chloride. This mixture was cooled to -10 °C (cooling bath consisted of 85% acetone, 15% water, and dry ice) and 1.0 mL (4.8 mmol) of  83 diphenyl chiorophosphate was added dropwise. The final mixture was stirred under gradual warming to 10 °C for four hours. Upon completion of the reaction, the mixture was further diluted with 20 mL methylene chloride and washed sequentially with cold H 0, 0.5M HC1, and 2 saturated NaHCO 3 solution, dried with MgSO , and concentrated in vacuo. Flash 4 chromatography (20% EtOAc / petroleum ether) of the residue yielded the title phosphate as a colorless syrup (1.1g. 85%) and the spectroscopic data (‘H NMR) was consistent with those reported.’°’  2-Deoxy-2-fluoro- a-D-mannopyranosyl-bis(tributylammonium) phosphate 30 (C P 8 O 2 HFN )  400 mg of diphenyl (Tri-O-acety1-2-deoxy-2-fluoro-c-D-mannopyranosyl) phosphate was dissolved in 15 mL distilled methanol and 40 mg of Pt0 2 was added to the mixture. This mixture was stirred very rigorously under an atmospheric pressure of hydrogen gas at room temperature for 48 hours. This mixture was then filtered, diluted with 2 mL of water and 2 mL of triethylamine and the final mixture was left at 4 °C for 10 hours. The mixture was concentrated in vacuo, then re-dissolved in 5 mL of water. The solution was then passed through a column (1 cm diameter x 5 cm length) of Dowex 50 W-X8 (H) resin into a constantly stirring mixture of water and excess tributylamine. Subsequent lyophilization yielded the title phosphate as a white powder (400 mg, 85%) and the spectroscopic data (‘H NMR) was consistent with those reported.’°’  84  Guanosine 5 ‘-(trihydrogendiphosphate, P’-(2-deoxy-2-fluoro-a-D-mannopyranosyl) ester 25 H 16 (C 1 0 5 FN ) 2 P 4  160 mg (350 jimol) of 2-deoxy-2-fluoro-a*D-mannopyranosyl-bis(tributylammonium) phosphate, 300 mg (410 jimol) of 4-morpholine-N,N’-dicyclohexylcarboxamidinium guanosine 5’-rnonophosphomorpholidate and about 50 mg (0.7 mmol) of 1H-tetrazole were dissolved in 5 mL freshly distilled pyridine (The mixture was repeatedly concentrated in vacuo and re dissolved in 5 mL of freshly distilled pyridine for three times). The mixture was stirred under Ar protection at room temperature for 7 days. Upon completion of the reaction (suggested by depletion of 4-morpholine-N,N’-dicyclohexylcarboxamidinium guanosine 5,monophosphomorpholidate from TLC analysis), the reaction mixture was concentrated in vacuo and was loaded onto a column (2.5 cm diameter x 70cm length) of DE-52 anion exchange resin. The column was run with a gradient of 0.2-0.5 M triethylammonium bicarbonate buffer and the eluent was monitored under liv absorption. Fractions containing the desired product were pooled and lyophilized to yield guanosine 5 ‘-(trihydrogendiphosphate, P’-(2-deoxy-2-fluoro-cL-Dmannopyranosyl) ester (35 mg, 10-15%). The identity of this final product was confirmed by 31 P NMR as well as —ESI-MS. 1 H NMR indicated that the material was only 75% pure with the major impurity being the GMP dimer; however, it was used in preliminary studies without further purification.  85  2.8.8 Reaction of GDP-2Deoxy-2-F1uoro-D-Mannose with GMD  A sample of deuterated phosphate buffer (50 mM sodium phosphate, 50 mM NaC1, pD 7.4) was prepared by repeated (3X) lyophilization of a phosphate buffer (50 mM sodium phosphate, 50 mM NaC1, pH 7) and dissolution in an equal volume of D 0. A sample containing 2 2.3 mg of GDP-2-deoxy-2-fluoro-D-mannose, 0.5 mM NADP, and 5 mg GMD (previously exchanged into deuterated buffer using Amicon Ultra centrifugal filter devices (Millipore)) was prepared in the deuterated buffer (final volume of about 1.0 mL). The sample was incubated at room temperature and the progress of the reaction was monitored using 31 P NMR spectroscopy.  86  Chapter 3 Mechanistic Studies with GDP-L-Fucose Synthase Mutants  87  3.1  Introduction  After the characterization of wild-type GFS, the focus of the research became the study of GFS mutants. Previous literature suggested that two catalytic residues were responsible for the inversions of stereochemistry in the reaction catalyzed by GFS: CyslO9 and 6 58 This 57 His179. ’ notion is also supported by a recent characterization of a related enzyme, GDP-mannose-3,5epimerase (GME). 64 Examination of the structure of GME in complexes with various GDP hexoses showed that Cys 145 (which is homologous to Cys 109 in GFS) and Lys2 17 (which is homologous to His179 in GFS) are positioned appropriately to fill these roles (Figure 3.1).  K283hhh1Z C145  4Keto  1 bYi74  1(178  4  GDP-D-Mannose I G-Iacse  $143  H1Th NAUPH  GFS active site with GDP-keto substrate modeled into active site  Figure 3.1  K217  Superpositioning of GME active sites in a complex with GDP-D-mannose/ GDP-L-galactose  X-ray crystal structures of GFS and GME active sites.  88 A modeling study of GME also suggested that Cys 145 is responsible for the deprotonations at both C-3” and C-5” stereocenters whereas the Lys2 17 is responsible for the corresponding reprotonations that complete the inversions of stereochemistry (Figure 3.1). Since the orientation of the two active site residues in GFS is highly similar to those in GME, it is likely that GFS also utilizes Cys 109 to deprotonate both stereocenters and uses His 179 for the reprotonation steps. In order to study the roles of the putative catalytic residues in GFS, sitedirected mutagenesis was performed and two mutants (CyslO9Ser and Hisl79Gln) were obtained. A brief explanation of site-directed mutagenesis is included in the beginning of this chapter. The middle section of this chapter discusses the characterization of the Cys 1 O9Ser mutant. It was discovered that this mutant not only produces GDP-L-fucose during catalysis, but also produces an epimeric product that provides a clue towards the order of the epimerization steps. In addition, deuterium wash-in experiments with GDP-L-fucose support this ordered mechanism. Lastly, kinetic studies with Cys 1 O9Ser show that a kinetic isotope effect slows the reaction of a C-3” deuterated substrate, strongly supporting the proposed role of this residue in catalysis. The later section of this chapter discusses the characterization of the His 1 79Gln mutant. Despite that fact that this mutant demonstrates a 10,000 fold decrease in activity, it is still able to wash deuterium into the C-3 position of substrate. Together, these experiments paint a consistent picture that outlines the order of epimerization and the roles of these residues in catalysis.  89  3.2 Site-directed Mutagenesis of GDP-L-Fucose Synthase  Two conservative mutants of GFS (CyslO9Ser and Hisl79Gln) were produced using the Quikchange Site-Directed Mutagenesis Kit from Stratagene. This is a polymerase amplificationbased mutagenesis technique that relies on the feature that plasmids isolated from E. coil are methylated at the N-6 position of deoxyadeno sine residues (m A). This characteristic 6 differentiates the parental DNA template from newly formed DNA generated by linear amplification. The addition of a restriction enzyme Dpn 1 will hydrolyze the parental DNA template at the target sequence of 5 ‘-Gm ATC-3’ (there are three such cut sites in the sequence of 6 GFS) but leave the newly synthesized DNA intact. Complementary primers were designed to span the site of mutation and incorporate one or two mismatched bases that introduce the desired mutation. Amplification yields the mutant plasmid in a nicked form and a subsequent Dpn 1 digestion destroys the parental DNA. The resultant nicked plasmid is then transformed into E. coil cells where ligation occurs. This technique was used to generate plasmids encoding for CyslO9Ser and His l79Gln GFS. The plasmids were then amplified, purified and sequenced to ensure no other errors were introduced other then the desired mutation. Overexpression of the mutant plasmids and purification of the mutant proteins were performed in an identical fashion to that described for the wild type enzyme. Typically 80 mg of CyslO9Ser and 60 mg of Hisl79Gln were purified from a 500 mL culture.  90  3.3  Kinetic Analysis of Mutant Enzymes  The importance of the tentative catalytic residues Cys 109 and His 179 was examined using a kinetic analysis. As described in Chapter two, the wild type enzyme follows Michaelis Menten kinetics with values of kcat  0.65  ±  0.02 s and KM(GDPM)  =  2.2  ±  0.2 jtM (saturating  NADPH). It was found that the CyslO9Ser mutant also follows Michaelis-Menten kinetics with values of kcat = 0.0012  ±  0.0001 s 1 and KM(GDPM)  =  0.88  ±  0.1 i.tM (saturating NADPH) (Figure  shown in section 3.4.3). The 500-fold decrease in the maximal activity of the CyslO9Ser mutant strongly suggests that Cys 109 plays a crucial role in the acid/base catalysis. In a later section of this chapter, a kinetic isotope effect is reported when [3 “H] -substrate is used with this mutant 2 indicating that Cys 109 is the residue responsible for deprotonation at C-3”. The other GFS mutant studied in this project was Hisl79Gln. Originally, His l79Asn was also one of the target mutants, but problems were encountered during the preparation of this mutant. Nevertheless, when His l79Gln was subjected to kinetic analysis, only very low activity was detected and the kcat of this mutant was estimated to be about 10,000-fold lower than that of the wild type. This dramatic loss of activity indicates that His 179 is a key catalytic residue and this is consistent with the literature where the His 1 79Asn mutant demonstrated a 1,000 fold decrease in the value of kcat. ’ In a later section of this chapter it is shown that even though the 6 His l79Gln mutant is almost completely inactive, it is still able to catalyze a wash-out of deuterium from substrate in an ordered fashion.  91  3.4  Characterization of the CyslO9Ser Mutant  3.4.1 Product Analysis with the CyslO9Ser Mutant  Given that the CyslO9Ser mutation had a significant effect on the rate of catalysis, it became important to carefully analyze the nature of the products that were formed. A sample of GDP-6-deoxy-4-keto- D-mannose and NADPH was incubated with both the wild type and mutant GFS enzymes, and the resulting products were isolated and identified by ‘H / 31 P NMR spectroscopy and negative ESI-MS. Repeated experiments showed that this mutant actually produces a mixture of products as shown in Figure 3.2 and Figure 3.3.  1”H of GDP-L-FUCOSe Wild Type GFS Reaction  CyslO9Ser GFS Reaction  ppm  Figure 3.2  5.50  5.00  Partial ‘H NMR spectra of wild type and C1O9S product mixtures.  92  G DP-L-Fucose  Wild Type GFS Reaction  CyslO9Ser GFS Reaction  ppm  Figure 3.3  9.O  -10.0  -11.0  42.0  -13.0  P NMR spectra of wild type and C1O9S product mixtures. 31  The reaction of CyslO9Ser is quite distinct from that of the wild type reaction which cleanly produces GDP-L-fucose as the only product. The ‘H NMR spectrum of the wild type reaction product shows a signal due to the anomeric proton of a sugar nucleotide (C-i “) at a chemical shift corresponding to that of GDP-L-fucose (doublet of doublet at 4.90 ppm in Figure 3.2). 108,  109  This upfield shift of the anomeric signal is consistent with a sugar nucleotide that  adopts a 1 4 conformation in solution. The formation of a single product by wild type GFS also C demonstrates the high specificity in the reduction step of the reaction. It suggests that wild type GFS will not reduce the carbonyl of either the starting material or the singly epimerized intermediate at any significant rate but only delivers a hydride once both epimerizations are completed.  93  However, the scenario is very different in the case of the Cysi O9Ser mutant reaction where it is clear that a new product has formed. Even though the product mixture contains 3:1 ratio of the unknown compound to GDP-L-fucose, it showed a single signal when analyzed by negative ESI mass spectrometry. The observed mass was identical to that of GDP-L-fucose (mlz 588) which implies this unknown GDP-sugar is a diastereomer of GDP-L-fucose that is generated by premature reduction of the C-4” carbonyl at some point in the reaction pathway. The anomeric signal of this unknown compound is also shifted downfield. There are three reasonable possibilities for the identity of the unknown GDP-sugars: GDP-6-deoxy-D-mannose (direct reduction of starting material), GDP-6-deoxy-L-gulose (reduction of C-5” epimer) and GDP-6-deoxy-D-altrose (reduction of C-3” epimer) (Figure 3.4).  C 3 H  C 3 H  HQ1 OGDP  GDP-6-deoxy-D-mannose  Figure 3.4  OGDP  GDP-6-deoxy-L-gulose  OH  OGDP  GDP-6-deoxy-D-altrose  Possible structures of the unknown GDP-sugar produced from CyslO9Ser catalysis.  It would be possible to rule out one or more of these structures if authentic samples were available for comparison. In past studies with the dehydratase, GMD, we had found that extended incubation of GDP-D-mannose with excess NADPH would generate GDP-6-deoxy-Dmannose. The enzyme normally bears a tightly bound NADP cofactor in its active site and generates GDP-6-deoxy-4-keto-D-mannose in a redox neutral fashion (Figure 3.5). Upon extended incubation, however, the tightly bound NADP cofactor can exchange with NADPH in solution, and when it rebinds its reaction product it can reduce the C-4” carbonyl to generate the  94  unnatural product GDP-6-deoxy-D-mannose. If this unnatural side reaction is allowed to continue to completion, a clean sample of GDP-6-deoxy-D-mannose can be obtained. NADP  NADP  OGDP  OGDP  0 2 H  GDP-D-mannose  GDP-6-deoxy-4-keto-D-mannose NADPH  NADP NADPH  NADP  OGDP  GD P-6-Deoxy-D-ma nnose  Figure 3.5  OGDP  G DP-6-deoxy-4-keto-D-man nose  Enzymatic synthesis of GDP-6-deoxy-D-mannose using GMD.  ‘H NMR analysis GDP-6-deoxy-D-mannose showed an anomeric proton signal shifted downfield from that of the unknown isomer (H-i”, 5.43 ppm, Figure 3.6). This downfield chemical shift is similar to that of GDP-D-mannose and is consistent with the anomeric signal of a sugar nucleotide in a “C 1 conformation. Therefore, the unknown isomer does not result from a direct reduction of the substrate and the identity of this compound should be either GDP-6deoxy-D-altrose or GDP-6-deoxy-L-gulose. GDP-L-fucose preferentially adopts the ‘C 4 chair conformation as it minimizes the number of axial substituents whereas GDP-6-deoxy-D-mannose (which is highly similar to GDP  95 D-mannose) adopts a 4 1 chair conformation. Therefore, the chemical shifts of the anomeric C proton NMR signal from these compounds set the relative reference points for the expected signal of the unknown GDP-6-deoxy-sugars discussed here. Since either of the two chair conformations of GDP-6-deoxy-D-altrose or GDP-6-deoxy-L-gulose positions an equal number of substituents in both axial and equatorial orientation, the preferred conformation of this unknown GDP-sugar is difficult to predict. The possibility of populating two conformers complicates the structural elucidation via coupling constants analysis or NOE measurement. Since the conformations of this unknown isomer will rapidly interconvert on the ‘H NMR time scale, the chemical shift of its anomeric proton should land somewhere in between that of GDP 6-deoxy-D-mannose and GDP-L-fucose. A comparison of ‘H NMR spectra confirmed this postulation (Figure 3.6).  HO  OH  OGDP  GDP-L.-fucose  ppm  Figure 3.6  5.50  5.00  ‘H NMR comparison of different GDP-.6-deoxy-sugars.  96  In order to determine the identity of this new compound, the CyslO9Ser reaction was further investigated using [3”H]-, [5”2 H]-, and [3”, 5”2 H]-GDP-6-deoxy-4-keto-D-mannose. 2 The mass of the product mixture was analyzed with negative ESI mass spectrometry. It was found that the [3”H1-substrate gave a single signal at m!z 588 corresponding to unlabeled 2 products; whereas both [5 “H] -and [3”, 5 “2 H] -GDP-6-deoxy-4-keto-D-mannose produced two 2 signals at mlz 588 and m/z 589 with a ratio of approximately 1:3 respectively (Figure 3.7).  + H  111-CDP-6-deoxy- GD1-6-dcoxyD-aUrose 2 13— 4-kete-D-•mose  DP-L4cose  (X3JP  OGIF  15H 2 1-GDP-6-deozy4 .kto-D-•mise  11i-GDP2 15’--  cii  (0P  ow  L  6deoxy D 5888  .c.,DP4-deoxy-D--aJtirese  + GDP-L-fcos  I  I I ______J 585  586  Figure 3.7  587  SSS  \  588  589  15-’HJGDP 6 dc.xyD.Irosc  I  cii  \  :  -  op  GDPL4ucose 880  591  892  593  594  595  Ritz  886  587  588  589  )U  8H  5G’  6E)  mu  Negative ESI mass spectra of the product mixture obtained in the reaction of CyslO9Ser with [3”H]- or [5”2 H]-GDP-6-deoxy-4-keto-D-mannose. 2  Since the deuterium was retained only when incorporated at C-5”, the unknown epimer must have an inverted stereochemistry at C-3”, and is therefore assigned as GDP-6-deoxy-Daltrose. These observations indicated that the epimerization catalyzed by GFS is indeed a sequential ordered process where the stereochemistry of C-3”is inverted first by a deprotonation / reprotonation processes followed by epimerization at C-S” using the identical mechanism (Figure 3.8). The introduction of the mutation at Cys 109 cripples the effectiveness of the second  97  epimerization, thus extending the lifetime of the GDP-6-deoxy-4-keto-D-altrose (3”-epimeric intermediate). Since the mutation only affects the epimerization steps without disturbing the function of the catalytic triad, a premature NADPH-dependent reduction process now competes with the second epimerization to yield GDP-6-deoxy-D-altrose (3”-epimeric product) in addition to the normal product GDP-L-fucose. O  0 H C 3  C 3 H  OH HO\  C-3” Epimerization  HO—  D  D  0 2 H  OGDP  Premature Reduction  D OH  GD P-6-deoxy-4-keto-D-man nose  H OH HO1O  OH  H0\  NADPH NADP OGDP  GDP-6-deoxy-4-keto-D-altrose  1i  HO1  OH  OGDP  G DP-6-deoxy-D-altrose  C-5” Epimerization  Reduction  OGDP NADP NADPH OH OGDP GDP-L-fucose GDP-6-deoxy-4-keto-L-galactose  Figure 3.8  Product formation with CyslO9Ser and [3”,5”H1-GDP-6-deoxy-4-keto- D 2 mannose.  The C-3” deuterium labels of [3 “H] and [3 “,5 “2 H] -GDP-6-deoxy-4-keto-D-mannose 2 -  are lost during catalysis because both products of the mixture have the stereochemistry at C-3” inverted. However, the C-5” deuterium labels of [5”HJ- and [3”,5”2 H]-GDP-6-deoxy-4-keto2 D-mannose are only lost in the fraction of the material that is converted into GDP-L-fucose. In addition, the proposed mechanism is consistent with the postulation that the NADPH dependent reduction of GFS requires the substrate / intermediate to adopt a ‘C 4 chair  98 conformation. No GDP-6-deoxy-D-mannose was produced in this reaction presumably because the substrate (GDP-6-deoxy-4-keto-D-mannose) always adopts its more stable 4 1 conformation. C However, after the first epimerization occurs, the intermediate (GDP-6-deoxy-4-keto-D-altrose) can adopt either the 4 1 or the ‘C C 4 confirmation; thus providing a window for the “premature” reduction to yield GDP-6-deoxy-D-altrose. Finally, the GDP-4-keto-L-fucose intermediate adopts the ‘C 4 confirmation upon the completion of the second epimerization and permits the effective reduction in either the crippled catalysis of the Cys 1 O9Ser mutant or the normal catalysis of wild-type GFS.  3.4.2 Deuterium Wash-in Experiment with GDP-L-Fucose  In the previous chapter, it was demonstrated that the wild-type enzyme is capable of washing solvent derived deuterium into both the C-3” and C-5” positions of the product GDP-L fucose when incubated with NADP in D 0. The same experiment was performed with the 2 CyslO9Ser mutant and provided further evidence for the proposed sequential order of epimerizations and the role that CyslO9 plays in the catalysis. Monitoring the process by 1 H NMR spectroscopy showed that the CyslO9Ser mutant only washes deuterium into the C-5” position even upon extended incubation (Figure 3.9).  99  A)  HO  3 H OH  No Redction  OGDP  H4  H5’  (‘Is  H3  A  B)  Hr  d kJu  D OH HOjj\  WT D-wsh-in OH  OGDP  I C)  1\f  HOI  T’ 1 OH  OGDP  A ppm  Figure 3.9  3.80  3.60  1k  f  3.80  ‘H NMR spectral comparison of various deuterium wash-in experiments using wild type and ClO9Ser GFS. A) No reaction; B) Reaction with wild type GFS; C) Reaction with CyslO9Ser.  In the wild type reaction, the signal for H-3” and H-5” both disappear upon incubation (Figure 3.9 B) whereas with CyslO9Ser only the signal for H-5” is lost (Figure 3.8C).  100 NADPH  NADP  17 His Os\  OH  H,  179 H— His  OH  CH OH  OGDP  NADPH  OH  OGDP  GDP-L-Fucose Proton Exchange With Solvent  No C-5 Epimerization Due to CyslO9Ser Mutation  Enz NADPH  NADP  NADPH HO  DOH  ..çc  OH  Ho’L CH OH  OGDP  OH  OGDP  OH  OGDP  H]GDP-L-Fucose 2 [5”-  Figure 3.10  Proposed mechanism for deuterium wash-in with CyslO9Ser mutant.  It is thought that in the reverse reaction direction the general acid residue Hisi 79 (B 2 in Figure 1.21) must act as a general base and is responsible for the deprotonation at C-5” (Figure 3.10). This deprotonation yields the corresponding enol(ate) intermediate but the reverse reaction stops here because the subsequent steps would require protonation on the opposite face by CyslO9 (B 1 in Figure 1.21) which has been mutated to serine. It is postulated that this mutation sufficiently slows the reprotonation at C-5” so that the enol(ate) intermediate simply interconverts with GDP-L-fucose. If the proton on His179 can exchange with solvent during the lifetime of the enol(ate) intermediate, a deuterium will be delivered to C-5” of GDP-L-fucose upon reprotonation of the enol(ate). The mass of the mono-deuterated product was also confirmed by negative ESI mass spectrometry (figure not shown). Therefore, the studies with the  101  CyslO9Ser mutant support a sequential order of epimerization taking place at C-3” first and then C-5” (C-5” first and then at C3”in the reverse direction) and identify the catalytic role of Cys 109 as a base that is responsible for the deprotonation steps at both C-3” and C-S” in the formation of GDP-L-fucose.  3.4.3 Kinetic Isotope Effect Studies with the CyslO9Ser Mutant  The lack of any kinetic isotope effect (KIE) in the wild-type GFS reaction using either C3” or C-5” deuterated substrates indicated that deprotonation at either of these positions was not a rate-determining step. However, it was anticipated that when a catalytic residue responsible for the deprotonation at these two stereocenters is mutated, the relative energy barrier(s) for the deprotonations would be increased and would become rate-determining (Figure 3.11).  •Step is not rate-determining •No kinetic isotope effect observed  Reaction Coordinate •Step affected by mutation becomes rate-determining •Kinetic isotope effect observed  E Reaction Coordinate  Figure 3.11  Schematic illustration of the effects of mutation on a non-rate-determining step.  102  In the case of the CyslO9Ser mutant, both products (GDP-L-fucose and GDP-6-deoxy-Daltrose) are formed via a process that involves deprotonation at C-3 “. Therefore, if Cysi 09 acts as a base, one might expect to see a KIE during the reaction of [3”H]-GDP-6-deoxy-4-keto-D2 mannose. Kinetic studies were performed using unlabeled and [3”H]-GDP-6-deoxy-4-keto-D2 maimose as described in Chapter two. The amount of NADPH was fixed at a saturating level (100 .tM) while the concentration of the keto substrates were varied. All the compounds were incubated in pH 7 sodium phosphate buffer at 25 °C and the GFS reaction was initiated by the addition of enzyme. The initial velocities obtained in the studies were plotted against the substrate concentration and the graph was fitted to Michealis-Menten kinetics (Figure 3.12).  0 V  (J.LMlmin)  0  0.2  0.4  0.8  0.8  1  1.2  1.4  [Substrate] (jtM)  Figure 3.12  Kinetic plots of the reaction of CyslO9Ser with unlabeled and [3”H]-GDP-62 deoxy-4-keto-D-mannose.  1 6  103  A KIE on the value of keat of kH/kD = 1.5 was observed. Therefore, C-3” deprotonation has become a partially rate-determining step of catalysis as the result of the mutation. The Michealis parameters for the reaction of GFS with unlabeled and [3”H]-GDP-6-deoxy-4-keto2 D-mannose are shown in Table 3.1.  Table 3.1  Unlabeled-Substrate  H]-Substrate 2 [3 “-  Km  0.16+/0.OluM  0.21+/0.O1UM  kcat  (9.7+/ 0.1) X i0 54  (6.3÷/ 0.1)  X  i0 S 1  Kinetic parameters for the reaction of CyslO9Ser with unlabeled and [3”Hj2 GDP-6-deoxy-4-keto-D-mannose.  The observation of a KIE on the reaction of [3 “H] -GDP-6-deoxy-4-keto-D-mannose 2 with CyslO9Ser but not with the wild type enzyme strongly suggest that Cys 109 is the catalytic residue responsible for the deprotonation at C-3”.  3.5  Characterization of the His 179Gm Mutant  3.5.1 Deuterium Wash-in Experiment with GDP-L-Fucose  Despite the fact that the overall reaction catalyzed by the His 1 79Gln mutant is extremely slow, it may still be possible to carry out solvent isotope incorporation studies and observe  104  partial reactions catalyzed by this mutant. Therefore, GDP-L-fucose was treated with the His 179Gm mutant and NADP in D 0 to test for deuterium incorporation. According to the 2 proposed mechanism and catalytic role of His 179 residue, it is expected that no deuterium would wash-into either the C-3” or the C-5” position of the molecule (Figure 3.13). In the reverse reaction direction, C-4” oxidation could still occur, however the base required for C-5 deprotonation would be missing. Z 179 GIn  NADPH  NADP  O  H  NADPH  OH  HGin EIiz , 179 OH  HOL OH  CH OH  OGDP  OGDP  No C-5” Deprotonation Due to Hisl79GIn Mutation  GDP-L-fucose  NADPH  109 Cys  NADPH  H  OGDP  Enz-  Enz  NADPH Enz 179 Gin  109  OGDP  GDP-6-deoxy-4-keto-D-man nose  Figure 3.13  Deuterium wash-in experiment with Hisl79Gln mutant.  Repeated experiments showed that, as expected, no deuterium is incorporated into GDP L-fucose even upon extended incubation, consistent with the proposed catalytic role of His 179.  105  3.5.2 Deuterium Wash-out Experiment with Deuterated Substrates  Despite the extremely low activity of this mutant, a very interesting phenomenon was observed that demonstrated the mutant was still properly folded and capable of catalyzing partial steps of the overall reaction. It was found that during the incubation of [3”, 5”H]-GDP-62 deoxy-4-keto-D-mannose with His 179Gm and NADPH, both deuterium labels of the substrate were washed-out prior to any formation of product and one of the deuterium labels disappeared faster than the other. This was apparent upon monitoring the reaction using negative ESI mass spectrometry (Figure 3.14). \  \  04-I  Hisl79GIn  fb  .-  1  HO  NADPH  000P D rnfz=588  OH  HISI79GIn  fj 1 OGDP H  HOLS  T 1  NADPH  OGDP H mtz=586  mfz=587 5881  T=o 1’ I  \5801  588O ¶0  T=l5min  j  J  T =45 mm 584  Figure 3.14  585  586  587  588  585  880  881  582  688  oyz  Negative ESI mass spectra monitoring the deuterium wash-out experiment of [3”, HJ-GDP-6-deoxy-4-keto-D-mannose with His 1 79Gm. 2 5 “-  106  In order to distinguish which deuterium was washed-out faster, the experiment was repeated with separate incubations of [3 “Hj- and [5 “2 H] -GDP-6-deoxy-4-keto-D-mannose. 2 Figure 3.15 shows the result of this experiment, noting that half the amount of enzyme was used (4.5 mg versus 9.0 mg in the previous experiment) to allow better monitoring of the deuterium wash-out process. It is clear from the results that the label at C-3” washes out approximately 18 times faster that the label at C-S” (Figure 3.15).  C 3 O•\&1  C\ 3 O\\H  OH °  OGOP  TA 1  NADPH  H H  H mlz  m!z = 587  mlz=586  mlz=587  OH  Hisl79GIn  Hisl79GIn NADPH  H D  O\  IJ  000P  =  586  587.1  T=O  V  J  \eo  T=5mn /  /  J  /  jq  /•\  T9Omin  I I ILL 583 584 585 586 587 588 589 590 SglrnIz  Figure 3.15  —  I  I  I  I  I  I  I  I  583 584 585 586 587 588 589 590 591  n!Z  Negative ESI mass spectra monitoring the deuterium wash-out experiment of [3”H]- and [5”2 H1-GDP-6-deoxy-4-keto-D-mannose with Hisl79Gln. 2  107  The ability of this mutant to bind the substrates and catalyze the wash-out of deuterium indicates that this mutant is still properly folded. The observation strongly supports the proposed order of epimerizations where stereocenter at C-3” is inverted first and the stereocenter at C5” is inverted second. It is interpreted that the C-3” deuterium gets washed-out faster because this is the site where the first deprotonation normally takes place and is more easily accessed by the cysteine base. It is clear that the enzyme is also able to deprotonate the substrate at C-5” but this is a much slower process since this normally occurs after the C-3” position has been inverted. This result provides strong evidence in support of the proposed catalytic role of both Cys 109 and His179. Since the mutation (Hisl79Gln) only affects the reprotonation step of the epimerization, CyslO9 can still abstract a deuterium, exchange it with a solvent-derived proton, and deliver the proton back onto the stereocenter. The fact that the deuterium at both stereocenters can be washed-out strongly suggests that CyslO9 is indeed the catalytic base that is responsible for the deprotonation steps at both sites.  3.6  Future Research  Despite the reported difficulties in obtaining an x-ray crystal structure of GFS in a complex with a GDP-sugar, it is anticipated that the mechanistic studies described in this thesis can be supported by future crystallographic examination. Wild-type or mutant GFS originating from different strains of bacteria (or even mammals) may behave differently and be amenable to crystallization in a complex with GDP-(keto)-sugar(s) (preferentially GDP-D-mannose and GDP  108  L-fucose). The orientation of substrates with respect to the active site architecture could be revealed and would be expected to agree with our proposed mechanism as well as the proposed role of active site residues. In addition, a GFSGDP-L-fucose crystal structure would not only provide vital evidence to support the postulated 1 4 conformation of GDP-(4-keto)-L-fucose in C the active site, but would also provide a static picture that correlates the necessity of the 1 4 C conformation with respect to the final hydride transfer to GDP-4-keto-L-fucose. This would be useful in explaining how GFS is capable of producing a single product with high specificity. In addition, a great deal of information could be obtained by showing the crystal structure of Hisl79Gln-GFS in a complex with GDP-(4-keto)-D-altrose which is the intermediate of the reaction. The inactivity of the mutant enzyme would prevent any epimerization from taking place, yet C-4” oxidation may occur to generate the 4-keto- intermediate in the active site of the enzyme. In the absence of crystallography, it may still be possible to provide further evidence regarding the nature of the singly epimerized intermediate. If GDP-6-deoxy-L-gulose and GDP 6-deoxy-D-altrose were chemically synthesized in a pure form (Figure 3.16), their ability to bind to His 1 79Gln could be measured.  HQI H  OGDP  GDP-6-deoxy-L-gulose  Figure 3.16  HO1° OH  OGDP  GDP-6-deoxy-D-altrose  Structures of GDP-6-deoxy-L-gulose and GDP-6-deoxy-D-altrose.  109  In each case, C-4 oxidation could occur to generate the corresponding ketone intermediate and it is expected that much tighter binding would be observed in the case of GDP-6-deoxy-(4-keto)-Daltrose.  3.7  Conclusions  The results of the studies with wild-type GFS, CyslO9Ser, and Hisl79Gln converge to support a mechanism for the GDP-fucose synthase reaction which involves an ordered sequence of epimerizations catalyzed by two active site acid / base residues. It is proposed that the first epimerization takes place at the C-3” stereocenter where Cysi 09 serves as a base to deprotonate the substrate and His 179 serves as an acid to reprotonate the stereocenter on the opposite face. Then, the second epimerization process repeats at the C-5” stereocenter where both Cys 109 and His179 play analogous roles in inverting the stereochemistry at this position (Figure 3.17).  110 179 HHjs  Enz  Enz  G DP-6-deoxy-4-keto-D-mannose  g Enz 17 HHIS  5 109 Cys  GDP-6-deoxy-4-keto-D-altrose  OH C_jOGDP 3 H  IHO  0  OH  NADP  GDP-L-fucose  Figure 3.17  Epimerization  H,2 jOGDP OH  Ring Flip  NADPH  OH  OGDP  G DP-6-deoxy-4-keto-L-galactose  The proposed mechanism of GFS and the proposed roles of CyslO9 and His179.  In order for the same catalytic residue to act as an acid or base in two consecutive epimerizations, it is anticipated that there must be rapid proton shuttling to and from the active site of the enzyme during the lifetime of the GDP-6-deoxy-4-keto-D-altrose intermediate. This would be necessary to regenerate the appropriate protonation states of the catalytic residues for the second epimerization step. Finally, an NADPH-dependent reduction occurs at C-4” that is catalyzed by the conserved Tyr136 residue found in all SDR superfamily members. Interestingly, wild-type enzyme only reduces the C-4” carbonyl after both epimerization steps have occurred and produces GDP-L-fucose as the only product. Since the intermediate just before the final reduction is GDP-6-deoxy-4-keto-L-galactose and it preferentially adopts the ‘C 4 conformation (see ring flip in Figure 3.17), an interesting explanation would be that the reduction only proceeds when a molecule in the ‘C 4 conformation.  111  In the studies with CyslO9Ser, it is found that this mutant produces GDP-6-deoxy-Daltrose as the major product and this supports the proposed order of epimerizations where the stereochemistry of C-3”is inverted first. The crippled second epimerization also reflects that CyslO9 is involved in the subsequent epimerization process. According to the proposed mechanism, the Cys 109 mutation should slow down the deprotonation step in both epimerization processes. When GDP-6-deoxy-4-keto-D-mannose binds to the mutant enzyme, it adopts the 4 1 C conformation is not reduced. The mutated residue SerlO9 slowly deprotonates at C-3” and the resulting enol(ate) is readily protonated by His 179 to form the GDP-6-deoxy-4-keto-D-altrose intermediate. This intermediate is then either subsequently epimerized at C-5” and reduced to yield GDP-L-fucose, or prematurely reduced to form GDP-6-deoxy-D-altrose (Figure 3.8). Since the mutation also affects the subsequent deprotonation at C-5”, the GDP-6-deoxy-4-keto-Daltrose intermediate may have a long enough lifetime for the minor 1 4 conformation to be C prematurely reduced in competition with the second epimerization step. Presumably during normal catalysis with wild type enzyme the GDP-6-deoxy-4-keto-D-altrose intermediate preferentially adopts a 4 1 conformation in the active site of wild-type GFS and, therefore, is not C reduced prematurely. The experiment involving solvent-derived deuterium wash-in with GDP-L-fucose and the CyslO9Ser mutant strongly supports the proposed order of the epimerizations and the proposed role of Cys 109 (it demonstrated that Cys 109 is responsible for the deprotonation at C-5”in the forward reaction). In the presence of a catalytic amount of NADP, the wild-type enzyme is capable of incorporating deuterium into both the C-3” and C-5” positions of GDP-L-fucose due to the freely reversible nature of the reaction. However, the Cys 1 O9Ser mutant selectively incorporates deuterium into the C-5” position. This observation is consistent with the proposed order of epimerization where the C-5” position is expected to be deprotonated first in the reverse  112  reaction direction. The oxidation of GDP-L-fucose gives the GDP-6-deoxy-4-keto-L-galactose intermediate as the mutation does not affect the redox architecture of the enzyme. Hisi 79 (which acts as a base in the reverse reaction) then deprotonates the C-5” position of the intermediate but no inversion of stereochemistry can occur due to the mutation of Cys 109 (which acts as an acid in the reverse reaction) (Figure 3.9). During the lifetime of the enol(ate) intermediate, the protonated His 179 exchanges its proton with a solvent-derived deuterium and then delivers this deuterium back to the C-5” position. Finally, the observation of a kinetic isotope effect slowing the reaction of [3 “H]-GDP2 6-deoxy-4-keto-D-mannose strongly suggests that CyslO9 is the base involved in the first deprotonation step at the C-3” position. No KIE is observed in the wild type reactions with either H]-GDP-6-deoxy-4-keto-D-marmose or [3 “2 [5 “H]-GDP-6-deoxy-4-keto-D-mannose, 2 indicating that neither of the deprotonation steps are rate determining in the overall catalysis. This is in agreement with previous studies employing deuterated NADPH that showed a KIE of 1.4 and demonstrated that the final reduction step is partially rate-limiting. 60 In the case of the CyslO9Ser mutant, a mixture of GDP-D-altrose and GDP-L-fucose is produced, but C-3” deprotonation / epimerization must occur in the formation of either product. A KIE of about 1.5 is observed with the use of [3”H]-GDP-6-deoxy-4-keto-D-mannose demonstrating that the 2 mutation of Cys 109 raises the energy barrier of the C-3” deprotonation step and now this step becomes partially rate-determining. Therefore, Cys 109 is the catalytic residue that is responsible for the first deprotonation at C-3”. Gathering the results from the Cys 1 O9Ser deuterium wash-in experiment and KIE measurement, Cys 109 is assigned as the catalytic base which deprotonates both C-3” and C-5” in the ordered epimerization mechanism. Lastly, the other mutant of GFS studied is His 1 79Gln. The 10,000 fold decrease in activity indicates that the mutation of His 179 severely cripples the enzyme. However, this  113  mutated enzyme is believed to be still properly folded as demonstrated by its ability to carry partial steps of the overall catalysis. The deuterium wash-in experiment with this mutant and GDP-L-fucose further supports the proposed catalytic role ofHisl79. As expected, no deuterium was incorporated into GDP-L-fucose by this mutant even upon extended incubation. Presumably GDP-L-fucose is oxidized by the un-disturbed redox catalytic triad, however, the mutated His179 (Gin) is not capable of deprotonating the C-5” position. Eventually, the GDP-6-deoxy-4-keto-Lgalactose intermediate is reduced back to GDP-L-fucose without any deuterium incorporation. Interestingly, this mutant is capable of exchanging deuterium from [3”,5”H]-GDP-6-deoxy-42 keto-D-mannose with solvent in a preferential order (one deuterium washed-out faster than the other). Further examination using [3 “H] -GDP-6-deoxy-4-keto-D-mannose and [5 “2 H] -GDP-62 deoxy-4-keto-D-mannose demonstrates that it is the C-3” deuterium exchanged faster. It is interpreted that the mutation does not affect the ability of Cys 109 to act as a base and it can still readily deprotonate both sites of epimerization. It is anticipated that the first site of epimerization / deprotonation would be accessed by CyslO9 more frequently due to the normal ordering of the mechanism and thus, the C-3” deuterium exchanges faster than the C-5” deuterium. This finding gives further strong evidence for the proposed ordered epimerization mechanism as well as the proposed catalytic role of CyslO9. Interestingly, the use of an ordered epimerization mechanism and the proposed catalytic roles of CyslO9 and His179 are similar to the reaction catalyzed by GDP-mannose 3, 5epimerase, even though the two enzymes adopt an opposite order of epimerization. The opposite order of epimerization with GME is suggested by the fact that GDP-L-gulose is one of the products of the enzyme where the C-5” position is inverted and the stereochemistry of C-3” remains the same. Crystal structures of GME bound with various GDP-sugars have been solved and these sugars include GDP-D-mannose, GDP-L-galactose, and a mixture of GDP-L-gulose  114  and GDP-4-keto-L-gulose. The studies of these structures provide valuable clues about the conformational changes during catalysis as well as the positions of the catalytic residues relative to the bound sugar moiety. All these structures suggest that the active site Cys 145 (which is homologous to CyslO9 in GFS) is positioned appropriately to act as a base for deprotonation at either the C-3” or the C-5” stereocenter (in the GDP-D-mannose to GDP-L-galactose reaction direction). At the same time, the active site Lys217 (which is homologous to His 179 in GFS) is located appropriately to protonate the enolate intermediates. Therefore, the structural studies of GME are in agreement with the findings of this thesis in terms of the roles of active site catalytic residues in the GFS reaction. In conclusion, GDP-fucose synthase is a very fascinating enzyme that catalyzes a complex reaction in a single active site. Despites the complexity in a series of consecutive actions in the active site, GFS remarkably generates GDP-L-fucose as the only product. The studies in this thesis support a mechanism involving an initial epimerization at C-3”, followed by an epimerization at C-5”, and finally a NADPH dependent reduction at the C-4” carbonyl. The active site catalytic residue CyslO9 and His179 has been assigned as the basic and acidic residues involved in both epimerization steps, respectively.  115  3.8  Experimental  3.8.1 Site-directed Mutagenesis, Over-expression, and Purification of Mutants  Plasmids encoding for the GFS mutants, CyslO9Ser and His 179Gm, were generated using the following two sets of oligonucleotides (non-matching nucleotides are underlined): for CyslO9Ser, 5’-CTCGGATCGTCCGCATCTACCCGAAAC-3’ (forward sequence) and 5’GTTTCGGGTAGATGCTGGACGATCCGAG-3’ (reverse sequence); for Hisl79Gln, 5’CACCCGAGTAATTCGCAGTGATCCCAGCATTGC-3’ (forward sequence) and 5’GCAATGATGGGATCACCTGCGAATTACTCGGGTG-3’ (reverse sequence). Mutations were introduced into the pT7GFS vector using the Quikchange Site-Directed Mutagenesis Kit (Stratagene) according to manufacturer’s instructions. The GFS genes in the resulting plasmids were fuily sequenced to ensure that only the desired mutations had been introduced during PCR amplification. Finally, the plasmids for both mutants were transformed, over-expressed, and purified using identical methods and conditions as described for those of wild-type GFS  116  3.8.2 Product Analysis with CyslO9Ser  Lyophilized GDP-6-deoxy-4-keto-D-mannose (20 mg, prepared as described previously) and 30 mg of NADPH were dissolved in 2.0 mL of water. To this solution was added 20 mg of Cys 1 O9Ser (in 1.0 mL of sodium phosphate buffer) and the reaction was incubated at 28 °C for 4 h. The progress of the reaction was monitored by —ESI-MS and judged to be complete by the disappearance of the GDP-6-deoxy-4-keto-o-mannose signal (m!z  =  586). The GDP-hexose  products were purified as described for the enzymatic synthesis of GDP-L-fucose. The final product mixture was isolated as bis-triethylammonium salts with an overall yield of 65% (16 mg). The reactions of deuterated substrates were carried out under analogous conditions and monitored by -ESI-MS.  3.8.3 Synthesis of 6-Deoxy-GDP-D-Mannose  To a mixture of 25 mg GDP-D-mannose, 0.38 mg NADP, and 1.3 L 3-mercaptoethanol in 2.5 mL of sodium phosphate buffer (10 mM, pH 7.0, containing 10 mM NaC1) was added 20 mg of GMD (previously exchanged into 800 p.L of the same buffer using Amicon Ultra centrifugal filter devices (Millipore)). The reaction mixture was incubated for 12 h at rt and monitored by negative electrospray ionization mass spectrometry (-ESI-MS, GDP-D-marmose  mlz = 604, GDP-6-deoxy-4-keto-D-mannose mlz = 586) to ensure the reaction reached  117  completion. To this mixture was added 35 mg of NADPH and the reaction was incubated for a further 12 h at room temperature. The progress of the reaction was judged to be complete by the consumption of GDP-6-deoxy-4-keto-D-mannose (mlz of 586) and the production of GDP-6deoxy-D-mannose (mlz of 588) as monitored by —ESI-MS. The enzyme was removed and product was purified as described in the enzymatic production of GDP-L-fucose. The final product was isolated as a bis-triethylammonium salt with an overall yield of 60% (20 mg).). ‘H 0, as triethylammonium salt, pH 7.4): 6 (ppm) 2 NMR (400 MHz, D ’ 2 Ji’,  6 Hz, 1H, 1’-H), 5.46 (br, d,  8.13 (s, 1H, 8-H), 5.95 (d,  6.8 Hz, 1H, 1”-H), 4.80 (dd, J2’1’ 6 Hz, 3 ’= 2 J ’ , 4.8 Hz,  ’,’ 4.8 Hz, 4 3 1H, 2’-H), 4.54 (dd, J ’= 3 J ’ , 3.6 Hz, 1H, 3’H), 4.37 (m, 1H, 4’-H), 4.23 (m, 2H, 5a’H, 5b’-H), 4.06 (br, m, 3 ”, 3.2 Hz, 1H, 2”-H), 3.94 (dd, 2 2 J ” ”, 3.2 Hz, 3 J ” 3”H), 3.90 (dd,  10 Hz,  10 Hz, 1H,  ” 10 Hz, 3 3.2 Hz, 1H, 5”-H), 3.31 (t, J”,  10 Hz,  1H, 4”-H), 1.25 (d, J ”,” 3.2 Hz, 3H, 6”-H); 31 6 P {‘H} NMR (121.5 MHz, D 0, as 2 triethylammonium salt, pH 7.4): 6 (ppm)  =  -10.88 (d,  Pp). MS (ESF) for P C 2 H 1 0 5 N 2 6 (free acid, 589.3) 5  =  20 Hz, Pa), -13.31 (d, J,= 20 Hz,  rn/z 588.2 [M-HJ.  3.8.4 Deuterium Wash-in Experiments with GDP-L-Fucose  Deuterium wash-in experiments were performed using Cys 1 O9Ser, and His 1 79Gm. A sample of deuterated phosphate buffer (10 mM, pD 7.4) was prepared by repeated (3X) lyophilization of a phosphate buffer (10 mM, pH 7.0) and dissolution in an equal volume of D 0. 2 Three samples containing 2.0 mg GDP-fucose, 0.1 mg NADP, and enzyme (7 mg of Cysi O9Ser  118  or 10 mg of His 1 79Gm, previously exchanged into deuterated buffer using Amicon Ultra centrifugal filter devices (Millipore)) were prepared in the deuterated buffer (final volume 0.8 mL). The samples were incubated at 28 °C and the progress of deuterium wash-in was monitored using ‘H NMR spectroscopy. The mass of the final product was confirmed by —ESI MS upon completion of the deuterium wash-in process.  3.8.5 Kinetic Studies of CyslO9Ser  Kinetic constants for the Cys 1 O9Ser reaction were obtained using a continuous spectroscopic assay that directly monitors the conversion of NADPH to NADP by following absorbance changes at 340 nm. Assay mixtures contained 50 mM sodium phosphate buffer, pH 7.0, 100 .tM NADPH (saturating), and 0.25-10 jiM GDP-6-deoxy-4-keto-D-mannose in a total final volume of 1.0 mL. The concentration of GDP-6-deoxy-4-keto-D-mannose stock solutions was determined by measuring A 260 and using an  £260  of 11.8 mM’ cm . Reactions were initiated 1  by addition of 10 iL of enzyme solution (containing 60 ig of CyslO9Ser) and the decrease in 340 was monitored at 25 °C. Initial velocities were calculated using an A  £340  of 6 220 M’ cm . 1  Kinetic parameters were determined from fitting the initial velocities to the Hill equation using the program GraFit. KIE studies on [3”H]-GDP-6-deoxy-4-keto-D-mannose were performed as 2 described in section 2.8.6 with the exception that 60 ig of the CyslO9Ser was used to initiate each reaction.  119  3.8.6 Deuterium Wash-out Experiment with Hisl79Gln  Samples containing 5 mg of substrate ([3”, 5”H1-GDP-6-deoxy-4-keto-D-mannose, [3”2 H]-GDP-6-deoxy-4-keto-D-mannose, or [5”2 H]-GDP-6-deoxy-4-keto-D-mannose), 9 or 4.5 mg 2 of His 179 Gin (previously exchanged into the same buffer using Amicon Ultra centrifugal filter devices (Millipore)), and 5 mg NADPH were prepared in 10 mM sodium phosphate buffer (pH 7.0, 10 mM, 0.75 mL final volume). The samples were initiated by the addition of enzyme, incubated at room temperature, and monitored by —ESI-MS in 15 minute intervals.  120  Chapter 4 Mechanistic Studies With GDP-Mannose-3,5-Epimerase  121  4.1  Introduction  GDP-mannose 3,5-epimerase (GME), as discussed in chapter one, has many similarities with GFS including sequence homology and the nature of the catalytic reaction. It is responsible for the biological conversion of GDP-D-mannose into GDP-L-gulose and GDP-L-galactose and it plays a key role in the biosynthesis of vitamin C in plants. 6467 Like GFS, x-ray crystal structures of GME have been extensively studied but only very limited mechanistic studies have been performed. However, unlike GFS, crystal structures of GME with bound GDP-sugars have been obtained and provide information on the static orientation of the putative catalytic residues. 64 As a result, Cys145 (analogous to CyslO9 in GFS) has been assigned the role of general base and Lys217 (analogous to His179 in GFS) has been assigned as the role of general acid in both epimerization steps. TM OH  NAD  NADH  NADH  OH  —Enz 7 l 2 N—LyS  Oxidation  OGDP  -s 145 Cys  GDP-D-mannose  OH  OGDP  OH  OGDP Epimerization  GDP4-keto-D-mannose  NADH  Reduction  DP GDP-L-gulose  GDP-4-keto-L-gulose _Enz NADH  NADH  NAD  ,‘  NAYs217 3 H  O  HOL OH  OHOGDP  GDP-L-galactose  Figure 4.1  C-3” Epimerization  Reduction  OH  OHOGDP  OH  OGDP  GDP-4-keto-L-galactose  Proposed mechanism of the reaction catalyzed by GME and roles of catalytic residues.  122  Differences between the reactions catalyzed by GME and GFS include the fact that GME catalyzes an overall redox neutral conversion with no net consumption of the NAD(H) factor. In addition, unlike GFS which produces GDP-L-fucose as the only product, GME generates an equilibrium mixture of starting material and two products. Thus, GME displays much less specificity in terms of promoting hydride transfer to each of the three interconverting 4-keto intermediates. Finally, the presence of GDP-L-gulose (C-5” epimeric product) and the absence of GDP-D-altrose (C-3” epimeric product) in the product mixture suggest that the C-5” stereocenter is inverted first in the GME reaction, which is the opposite order to that of GFS. The goal of this chapter was to provide mechanistic studies that support the proposed order of epimerizations and catalytic roles of Cys145 and Lys217 in the GME reaction. 6467 Mechanistic studies on GME are complicated by the fact that GME generates an equilibrium mixture of GDP-D-mannose, GDP-L-gulose, and GDP-L-galactose. This places certain restrictions on the design of mechanistic studies and makes product analysis / detection more challenging. Building on our foundation of GFS work, the early section of this chapter describes mechanistic studies on wild-type GME with a series of deuterium wash-in / wash-out experiments using GDP-D-mannose, GDP-6-deoxy-D-mannose, GDP-L-fucose, and GDP-D altrose. Later sections of this chapter investigate the epimerization order and the role of the proposed catalytic residues by performing deuterium wash-in experiments with two site-specific mutants (Cysl45Ala and Lys2l7Ala) acting on GDP-D-mannose, GDP-6-deoxy-D-mannose, and GDP-L-fucose. Ultimately our results support the proposed order of epimerization (C-5” first then C-3”) as well as the proposed role of both catalytic residues (Cys145 is the base which is responsible for deprotonation at both sites and Lys217 is the acid responsible for reprotonation),  123  4.2  Deuterium Wash-in I Wash-out Experiments  Since the OME reaction produces a mixture of products that may complicate both the design and analysis of mechanistic studies, we attempted to probe the sequential order of epimerization by comparing the relative progress of the wash-in or wash-out of selectively positioned deuterium labels. In order to avoid analyzing complicated NMR spectra of GDP-sugar mixtures, the experiments performed in this chapter were designed in such a way that the results can either be analyzed solely by negative ESI mass spectrometry or via NMR analysis of the spectrum of a single GDP-sugar.  4.2.1 Deuterium Wash-in Experiment with GDP-L-Fucose  One of the difficulties in studying the GME reaction is that it would be quite difficult to obtain a pure sample of the minor product, GDP-L-galactose. Since GDP-L-fucose is GDP-6deoxy-L-galactose and only differs from GDP—L-galactose only by lacking the hydroxyl group at C-6”, it was anticipated that GME would recognize GDP-L-fucose as an alternate substrate. If this were the case, then incubating GDP-L-fucose in D 0 should results in the formation of a 2 mixture of GDP-6-deoxy sugars each bearing deuterium at both C-3” and C-5” positions (Figure 4.2).  124 NADH  NAD  OH  OGDP  OH  [3”, 5”H]-GDP2 4-keto-L-fucose  [3”, 5”H]-GDP-L-fucose 2  NADH  NAD  NAD D  Reduction  GME  HOL  0 2 D D  GDP-L-fucose  OH  3 CH  OGDP  OGDP D [3”, 5”H]-GDP2 6-deoxy-L-gulose  [3”, 5”H]-GDP-4-keto2 6-deoxy-L-gu lose  NADH  NAD C 3 H  OH  HO°  D  OGDP  [3”, 5”H]-GDP-4-keto2 6-deoxy-D-mannose  Figure 4.2  GME  OGDP D H]-GDP2 [3”, 5”6-deoxy-D-mannose  Expected products from the incubation of GDP-L-fucose with GME in D 0. 2  To test GDP-L-fucose as an alternate substrate, it was incubated with GME in D 0 buffer 2 and the resulting reaction was monitored by 31 P NMR spectroscopy and negative ESI mass spectrometry (Figure 4.3 and Figure 4.4). While GDP-L-fucose was found to serve as a substrate, the reaction was extremely slow and it was difficult to force it to completion even upon extended incubation. Analysis of the 31 P NMR spectrum indicated that signals due to GDP-6-deoxy-Dmannose appeared during the incubation, indicating that inversions at both C-3” and C-5” were taking place (Figure 4.3). However, GDP-6-deoxy-D-mannose never became the major species in the mixture, suggesting that the process had not reached equilibrium. Signal resulting from GDP—  125  6-deoxy-L-gulose were not directly observed; however, from our studies on GFS, it is anticipated that they would coincide with 31 P NMR signals from GDP-L-fucose.  OH HO-Q  OH  +  3 OH  Ho\L 3 OH  OH  000P  OGDP  GDP-6-deoxy-L-gulose  GDP-6-deoxy-D-maimose  I  -90  I  I -10.0  I  I  I  I -11.0  I  I  I  I  -110  -13.0  l  I  1 i  -14.0  -15.0  ppm  Figure 4.3  P NMR spectrum showing the reaction of GDP-L-fucose with GME. 31  It was found that GME not only tends to precipitate during enzyme concentration and buffer exchange as described in the literature , but that it also precipitates during the extended 64 incubations required for these deuterium wash-in and wash-out experiments. Therefore it is difficult to control the amount of active enzyme present in these incubations and it creates unpreventable variations between otherwise identical experiments. This prevents the quantification of the rates of reaction and deuterium wash-in. Nevertheless, the relative rate of  126  deuterium wash-in at the C-3” and C-5” positions should be independent of enzyme concentration and can still provide meaningful data (Figure 4.4).  H .0  HO OH  GME OGUP  Mono-deuterated Product  GME  Di-deuterated Product  Slow  RaPad  mlz=588  mi=589  m!z=590  588  GDP-L-fucose ‘589 i  \590  GDP-L-fucose + GME After 4 Hours  /  GDP-L-fucose + GME OvernightTrial#1  I  GDP-L-fucose-i-GME OvernightTnal#2  586  Figure 4.4  588  590  592  594  m,’z  Negative ESI mass spectra showing the GME-catalyzed wash-in of deuterium into GDP-L-fucose.  127  Monitoring the deuterium wash-in by negative ESI-MS also shows that GDP-L-fiacose serves as a substrate for GME (Figure 4.4). The rate of incorporation is biphasic and GDP-L fucose (m!z  =  588) is converted into a mono-deuterated species (mlz = 588) relatively quickly.  The incorporation of the second deuterium (m!z = 590) occurs much more slowly and requires overnight incubation. This suggests that deuterium is incorporated into one of the positions much faster than the other, very likely implying that an ordered epimerization takes place here. Later experiments in this chapter suggest that the C-S position has a slower deuterium wash-in / wash out rate, reflecting the proposed order of epimerization. It should be mentioned that almost complete incorporation of two deuterium atoms could be obtained (see Figure 4.4, Trial #2) even though the epimerization steps had not reached equilibrium (Figure 4.3 represents a corresponding time point). This indicates that deuterium wash-in can occur without epimerization at one or both of the positions. This can be explained if a polyprotic residue, such as lysine, reversibly deprotonates a position and delivers a solvent-derived deuterium faster than reprotonation on the opposite face occurs.  4.2.2 Deuterium Wash-out Experiment with [3”H]- or [5”2 H]-GDP2 L-Fucose  In order to further investigate the difference in the rate of deuterium incorporation observed in 4.2.1, a deuterium wash-out experiment using [3”H]- or [5”2 H]-GDP-L-fucose was 2 performed. A deuterium wash-in experiment was not considered here because it is easier to observe the appearance of the “lost mass” peaks in mass spectra as they do not overlap with the  128  M+1 peak (due to the natural abundance of ‘ C) present in the spectrum of the starting material. 3 H]-GDP-L-fiicose were made from the sequential action of the wild-type 2 The 2 [3”H ]- and [5”GFS and the CyslO9Ser GFS mutant combined with the use of either H 0 or D 2 0 buffer at 2 different stages of the syntheses (Figure 4.5). The molecular mass of these two compounds were confirmed by negative ESI mass spectrometry (m/z  =  589, data not shown) and the positions of  the deuterium were confirmed by ‘H NMR spectroscopy (Figure 4.6).  D  GFS NADPH 0 2 D GDP-6-deoxy4-keto-D-man nose oH C 3 \  H D  D  C1O9S-GFS H3CjpjOGDP HO OH  NADP 0 2 H  [3”, 5’H]-G DP-L-fucose 2  GFS  HO OGDP  GD P-6-deoxy-  NADPH 0 2 H  HO OH  H]-G DP-L-fucose 2 [3”-  H OH -O  Cj9OGDP 3 H  D H  HO H3C4jOGDP OH  GDP-L-fucose  CIO9S-GFS NADP 0 2 D  H Cjp1OGDP 3 H HO OH  H]-G DP-L-fucose 2 [5-  4-keto-D-mannose  Figure 4.5  Enzymatic synthesis of {3”H]-GDP-L-fucose and [5”2 Hj-GDP-L-fucose. 2  129 H G11GIJP 3 H  H3 K2  H5  V  3 800  3.750  GDP-i..--Fucose  3.700  3.650  3.600  3.550  3500  ppm  Figure 4.6  Partial ‘H NMR spectra showing GDP-L-fucose, [3”H]-GDP-L-fucose, and [5”2 H]-GDP-L-fucose. 2  These two selectively deuterated compounds were incubated with GME in H 0 buffer 2 under identical conditions and the progress of the deuterium wash-out was monitored using negative ESI mass spectrometry (Figure 4.7).  130 DOH  GME  HO9  GME  Unlabeled  Unlabeled  Compounds OH  miz  OGDP  [3’H]-GDP-L-fucose m!z = 589  =  Compounds mlz = 588  588 [5”H 2 ]-GDP-L-fucose mlz = 589  589  589  i  i0  J  J  ,  I\  T = 15 mm  J  T=3Ommn  1,  7,  :1  I’ I  I ‘  T=9Omn  585 586 587 588 589 590 591 592 593 mlz  Figure 4.7  /  \  /  585 586 587 588 589 590 591 592 593 nVz  Negative ESI mass spectra showing deuterium wash-out experiment with [3 “Hj 2 or [5 “Hj-GDP-L-fucose. 2  The results show that solvent isotope exchange occurs at the C-3” stereocenter of GDP-L fucose much faster than the C-5”. This observation is consistent with the proposed order of  131  epimerization (C-5” first then C-3” in the forward reaction direction). However, it is known that the missing hydroxyl group at C-6” dramatically decreases the efficiency of catalysis, possibly because it slows the deprotonation and the subsequent epimerization at C-5”. Therefore, one could argue that normally the epimerization is non-ordered and the observed difference in rates of deuterium wash-out is solely due to the lack of the hydroxyl group at C-6”. Therefore, a deuterium wash-out experiment using GDP-6-deoxy-D-mannose was performed in an attempt to eliminate the possibility of a non-ordered epimerization mechanism.  4.2.3 Deuterium Wash-in Experiment with GDP-6-Deoxy-D-Mannose  A deuterium wash-in experiment was performed by incubating GDP-6-deoxy-D-mannose with GME in D 0 buffer. However, it was found that only a very limited amount of GDP-62 deoxy-D-mannose was converted into other epimeric products (less than 5%) as suggested by 31 P NMR spectroscopy (Figure 4.8). Furthermore, the GDP-6-deoxy-D-mannose (m!z = 588) remained non-deuterated as indicated by negative ESI mass spectrometry (Figure not shown).  132  A) 6 DP-6-deoxy-D-mannose  B)  G DP-6-deoxy-D-mannose  Epimeric Product(s)  tdMFi’SL .10.0  -15.0  ppm  Figure 4.8  A) 31 P NMR spectrum of GDP-6-deoxy-D-mannose. B) 31 P NMR spectrum of GDP-6-deoxy-D-mannose after extended incubation with GME in D 0. 2  It is postulated that the C-6” hydroxyl group is crucial for the C-5” epimerization, which is the first committed step of catalysis. The absence of any significant C-3” epimerization is consistent with the proposed ordered epimerization mechanism where the inversion of C-5” stereochemistry must occur first. The inability of the enzyme to incorporate deuterium at C-5 of GDP-6-deoxy-D-mamiose could either be due to its inability to deprotonate at C-5 or its inability to reprotonate the opposite face of the resulting enolate intermediate. In the latter case, one would have to invoke a sequestered active site acid / base residues that could not exchange its proton with bulk solvent during the lifetime of the enolate intermediate. This is possible if the catalytic base is monoprotic such as a cysteine. A similar phenomenon has been reported in the case of glutamate racemase. 105 However, such a constraint is not observed when the deuterium  133  incorporation is performed in the reverse reaction using GDP-L-fucose. It is observed that after extensive incubation deuterium is incorporated onto the C-5” position (Figure 4.4) while GDP-L fucose remains as the major product (Figure 4.3). Thus isotopic exchange into starting material is faster than product formation and likewise deprotonation is faster than reprotonation on the opposite face of the enolate intermediate. The incorporation of solvent deuterium accompanying deprotonation does not necessarily mean exchange with solvent isotope is fast. With a polyprotic base such as lysine, solvent-isotope exchange can be observed even within a sequestered active site. The previous assignment that Cys145 is responsible for the deprotonation of GDP-4-keto-Dmannose and Lys217 is responsible for the deprotonation of GDP-4-keto-L-galactose is consistent with these observations.  4.2.4 Deuterium Wash-out Experiment with [32H] or [5”H]-GDP2 6-Deoxy-4-Keto-D-Mannose  In parallel with the studies using GDP-6-deoxy-D-mannose, the GME reaction was also examined using GDP-6-deoxy-4-keto-D-mannose. Structural studies on GME reported that a Lysl78Arg mutant (Lys178 is part of the catalytic triad of SDR superfamily enzymes), crystallized as a mixture of abortive complexes containing either GDP-4-keto-L-gulose / NAD or GDP-L-gulose / NADH. 64 Since this finding demonstrates that the active site of GME can accommodate a “keto” form of the GDP-sugar even when the enzyme bound cofactor is in the “wrong” oxidation state, it is anticipated that GME containing NAD can bind GDP-6-deoxy-4keto-D-mannose. It is also anticipated that the epimerization process, similar to that of GFS, 6 ’ 60  134  is independent from the “redox” machinery of the enzyme and GME may still be able to catalyze epimerization even though the cofactor is in the incorrect oxidation state. Therefore, a deuterium wash-out experiment similar to the one previously described in the His 179Gln-GFS study was performed using [3”H]-GDP-6-deoxy-4-keto-D-mannose and GME. However, it 2 H]- or [5”2 was found that no deuterium is washed-out from either [3”.. H]- or [5”2 H]-GDP-6-deoxy-4-keto2 D-mannose upon incubation with GME (Figure not shown). The absence of wash-out with [5”H]-GDP-6-deoxy-4-keto-D-mannose agrees with the notion that the lack of the C-6” hydroxyl 2 inhibits the corresponding epimerization at C-S” and if a deuterium is abstracted from C-5” by Cys145, it cannot be exchanged with solvent without epimerization. The absence of wash-out with [C-3”Hj-GDP-6-deoxy-4-keto-D-mannose is consistent with the proposed order of 2 epimerization where the inversion at C-3” will not happen unless the stereocenter at C-5”is already inverted. It is also possible that GDP-6-deoxy-4-keto-D-mannose is simply not compatible with the architecture of GMENAD active site and, therefore, no deuterium is exchanged with solvent.  4.2.5 Deuterium Wash-in Experiment with GFS-CyslO9Ser Product  Mixture  Lastly, the ability of the GME active site to accept compounds of differing stereochemistry is further explored with the use of a product / intermediate analogue: GDP-6deoxy-D-altrose. In the GME reaction, one of the products is GDP-L-gulose and its  135  stereochemistry at both the C-3” and C-5” positions are opposite to GDP-6-deoxy-D-altrose (Figure 4.9).  HQI H  OH  OGDP  GDP-L-g u lose  Figure 4.9  HOj°  OH  OGDP  G DP-6-deoxy-D-altrose  Structures of GDP-L-gulose and GDP-6-deoxy-D-altrose.  Even though GME would not produce GDP-D-altrose in a mechanistic scheme involving C-5” epimerization followed by C-3” epimerization, its active site may be capable of accommodating GDP-D-altrose if a non-ordered mechanism were possible. It is anticipated that if GME can bind GDP-D-altrose, the enzyme should also be able to bind GDP-6-deoxy-D-altrose and the previous studies on the CyslO9Ser-GFS mutant provides a source of GDP-6-deoxy-Daltrose mixed with GDP-L-fucose (3:1 ratio). A deuterium wash-in experiment was performed using this 75% pure GDP-6-deoxy-D-altrose in order to test whether GME can process it (Figure 4.10).  136  HCOH  HOH Hoj0  HO  GME 0 2 D  + OH  Cl-I  0601’  0601’  OH  :  0601’  mlz = 590  mlz = 588  miz = 588  i  HOL\  3  I GDP-L-fucose\ GDP-6-deoxy-o-altrose  I  /  Inl:3 Ratio \  \.  590  GME Added 586  Figure 4.10  588  590  592  594  mhz  Negative ESI mass-spectra showing deuterium wash-in experiment with CyslO9Ser-GFS product mixture.  The results are similar to these obtained in the GDP-L-fucose deuterium wash-in experiment (in that extended incubation leads to an incorporation of two deuterium atoms into the majority of the GDP-sugars). It indicates that GME can actually bind GDP-6-deoxy-D-altrose and incorporate deuterium into both the C-3” and C-5” positions. Therefore, GDP-6-deoxy-D-  137  altrose (presumably GDP-D-altrose as well) is indeed a substrate of GME suggesting that while the enzyme preferentially follows a mechanism involving C-5” epimerization followed by C-3” epimerization, the opposite order can also be follow at a slower rate.  4.3  Deuterium Wash-in Experiments with GME Mutants  Having confirmed the preferred order of epimerization, subsequent studies on GME focused on supporting the roles of the proposed catalytic residues: Cys145 and Lys217. In the conversion of GDP-D-mannose to GDP-L-galactose, Cys145 is proposed to act as the base in both deprotonation steps and Lys217 is proposed to act as the acid. In these studies, a deuterium wash-in approach is employed in which only one pure GDP-sugar is present in each experiment allowing for simple product analysis. Previous structural studies performed by Professor Naismith and his co-workers provide a vital clue in the design of these experiments. They identified two inactive mutants (Lys2l7Ala and Cysl45Ala) that are incapable of performing any epimerization, yet are still properly folded as demonstrated by a study of their crystal 64 With the kind donation of plasmids encoding for the mutants by Professor Naismith, structures. deuterium wash-in experiments were performed and the roles of these two proposed catalytic residues were revealed.  138  4.3.1 Deuterium Wash-in Experiment with Lys2l7AIa Mutant  To probe the catalytic role of Lys2 17, deuterium wash-in experiments were carried out by incubating Lys2 1 7Ala with either GDP-D-mannose (forward reaction) or GDP-L-fucose (backward reaction) in D 0 buffer and analyzing the results by negative ESI mass spectrometry. 2 It is anticipated that in the forward reaction direction, Cys 145 can still deprotonate GDP-D mannose at C-5”. Since Lys217 is mutated to Ala, no further catalysis can occur and the enol intermediate would simply be converted back to the keto-intermediate (Figure 4.11, top). If, during the lifetime of the enol intermediate, Cys 145 can exchange its proton with solvent-derived deuterium, then this deuterium would be incorporated back onto C-5” (note that it has been postulated that the proton will not be exchanged in the absence of epimerization based on studies with wild-type GME) (Figure 4.11). NADH  NAD  NADH HO  OH  V..AIa217  Due to Lys217 Mutation OH  GDP-D-mannose  NADH  NAD OH  HOL CH 3 OH  OGDP  GDP-L-fucose  Figure 4.11  Oxidation  Z 7 i 2 Aa  NADH  OH  0 2 D 3 CH OH  OGDP  Due to Lys217 Mutation  -SH 145 Cys En  Deuterium wash-in studies with Lys2l7Ala mutant.  OGDP  139  In the case of the reverse reaction (with GDP-L-fucose), it is expected that no deuterium is incorporated into C-3” position as the mutation prevents deprotonation from occurring (Figure 4.11, bottom). The combined results from deuterium wash-in experiments in both directions should provide vital clues to the catalytic role of Lys217. It was found that no deuterium was incorporated in either experiment (Figures not shown). Since the mutation terminates the epimerization at C-5”, the absence of deuterium incorporation at C-5” is consistent with the hypothesis that the Cys 145 does not exchange its proton with solvent unless the epimerization occurs. In addition, the absence of deuterium incorporation at C-3” in the reverse reaction with GDP-L-fucose is consistent with the proposed catalytic role of Lys2 17 (it would serve as the base for deprotonation in the reverse reaction). Of course it is not possible to rule out the notion that mutation to Lys217 may have prevented oxidation at C-4” or impaired the functioning of Cys 145; however, our previous studies with GFS suggest that this would not be the case.  4.3.2 Deuterium Wash-in Experiment with Cysl45Ala Mutant  When probing the catalytic role of Cys145, an approach similar to that used in the Lys2l7Ala investigation was employed. Cysl45Ala was incubated in D 0 with either GDP-D 2 mannose (for the forward direction) or GDP-L-fucose (for the reverse direction). However, a different result is expected here when compared to the Lys2 1 7Ala studies. Since Lys2 17 is functional in the Cysl45Ala mutant, deprotonation of oxidized GDP-L-fucose should still occur  140 and since lysine is polyprotic, deuterium incorporation should be observed (Figure 4.12, bottom). In the incubation with GDP-D-mannose, no deprotonation or deuterium incorporation would be expected (Figure 4.12, top).  NAD  NADH  NADH  217 Enz -Lys  HO\  OH  OGDP  GDP-D-mannose  NAD  Due to Cys145 Mutation OH  -O  Oxidation  HOL OH  0 2 D  OGDP  OGDP  OGDP  GDP-L-fucose  Figure 4.12  Deuterium wash-in experiment with Cysl45Ala mutant.  It was found that Cysl45Ala does not incorporate deuterium into GDP-D-mannose as expected whereas it does incorporate deuterium into GDP-L-fucose as indicated by negative ESI mass spectrometry (Figure 4.13).  141  HOH HO9  Cyst 45AIa  Ho90  0 2 D OH  OGDP  mlz = 588  OH  OGDP  m!z = 589  Before enzyme addition  After enzyme addition  Figure 4.13  Negative ESI mass spectra of Cysl45Ala-catalyzed deuterium wash-in with GDP L-fucose.  In order to confirm the position of deuterium incorporation, the same reaction mixture was then analyzed by ‘H NMR spectroscopy. The ‘H NMR spectrum confirms that essentially no epimerization occurred as GDP-L-fucose is the only sugar nucleotide present. In addition, the deuterium has largely been incorporated into the C-3” position as predicted (Figure 4.14).  142  HOH  Cysl45AIa  H0j  HO°  020 OH  OGDP  rnlz= 588  OH  m!z  H4  OGDP  589  H3 H5  H2  No Enzyme  With Cysl45AIa  I  aoo  Figure 4.14  I 3.7w  I 3.700  I 3000  I 3.000  I 3.500  I 3.5)5)  ‘H NMR spectra showing Cysl45Ala-catalyzed deuterium wash-in using GDP-L fucose.  Overall, the deuterium incorporation studies with Cys 145 Ala provide the strongest piece of evidence supporting the proposed roles of the catalytic bases. Cys145 is responsible for the C3” and C-5” deprotonations in the GDP-L-galactose producing direction, and Lys217 is responsible for the corresponding protonations.  143  4.4  Future Research  The mechanistic studies performed so far only probe the mechanism either from the beginning of the reaction (with GDP-D-mannose) or from the very end (with GDP-L-fucose). A possible future deuterium wash-in experiment can be performed upon the availability of GDP-L gulose or GDP-6-deoxy-L-gulose (Figure 4.15 and Figure 4.16).  NAD  HO10 3 or CH RCH OH 2 H  Reduction  Lys2l7AIa  OGDP  Oxidation  EflZ’•_AIa  Due to Lys217 Mutation  NADH  NADH Lys27AIa  Enz  e —S 145 .—Cys  HH R OGDP D  Due to Lys217 Mutation  Direction to GDP-D-mannose  Figure 4.15  Direction to GDP-L.galactose  Proposed deuterium wash-in studies with Lys2l7Ala acting on GDP-L-gulose or GDP-6-deoxy-L-gulose.  It is anticipated that once again with the Lys2l7Ala mutant, no deuterium incorporation would be observed. Oxidation and C-3” deprotonation may take place; however, since Cys 145 is sequestered from exchange of its proton with solvent, no incorporation would occur.  144 H  IN1LJ  OH -  HO HO  3 or CH R=CH OH 2 R H  Reduction  NADH  1  OGDP  Due to Cys145 Mutation  Oxidation  ys145AIa  ,)(.  Enz- Ala 145  ,Direction to GDP-D-mannose  Figure 4.16  Cysl45AIa  NADH  x— ?çt EflZ___Aa145  OGDP  fr H  1  OGDP  Due to Cys145 Mutation Direction to GDP-L-gaIactose  Proposed deuterium wash-in studies with Cysl45Ala acting on GDP-L-gulose or GDP-6-deoxy-L-gulose.  However, in the case of the Cysl45Ala mutant, oxidation and deprotonation at C-5” could occur and, because lysine is polyprotic, deuterium wash-in would be observed at C-5”. The observation of isotopic incorporation at C-3” with Cysl45Ala and GDP-L-fucose and at C-5” with Cysl45Ala and GDP-L-gulose would strongly support the proposed mechanism. In addition, x-ray crystal structures of wild-type OME and mutants in a complex with GDP-L-fucose would solidify the significance of the deuterium wash-in / wash-out experiments described in this chapter. A comparison of crystal structures in a complex with GDP-L-fucose, GDP-D-mannose, GDP-L-gulose, or GDP-L-galactose may provide information about the importance of the C-6” hydroxyl group in related to the epimerization at C-5”. Finally, GDP-L fucose can also be tested as a competitive inhibitor of GME.  145  4.5  Conclusions  The studies of wild-type GME, Lys2l7Ala, and Cysl45Ala converge to support a mechanism for the GDP-mannose 3,5-epimerase reaction which involves an ordered sequence of epimerizations catalyzed by two active site acid / base residues. It is proposed that upon the formation of ketone at C-4” by oxidation, the first epimerization takes place at the C-5” stereocenter where Cys145 serves as a base to deprotonate the substrate and Lys217 serves as an acid to reprotonate the enol(ate) on the opposite face. Then the intermediate is either reduced to generate GDP-L-gulose or it is moved on to a second epimerization. The second epimerization process repeats at the C-3” stereocenter where both Cys145 and Lys217 play analogous roles in inverting the stereochemistry at the C-3” position. Similar to GFS, in order for the same catalytic residues to play the identical role of either acid or base in two consecutive epimerizations, it is anticipated that there must be a rapid proton shuttling to and from the active site of the enzyme during the lifetime of the GDP-6-deoxy-4-keto-L-gulose intermediate to regenerate the catalytic residues in their appropriate protonation states. Finally a reduction occurs at C-4” and is facilitated by the same SDR superfamily conserved Tyr178 residue. In the studies on wild-type GME, a series of deuterium wash-in and wash-out experiments were performed. Since pure GDP-L-galactose was not available, GDP-L-fucose generated by GFS was used in the mechanistic studies. The deuterium wash-in experiment with GDP-L-fucose demonstrates that deuterium is preferentially incorporated onto one of the epimerization sites more rapidly than the other. The order of the sequential deuterium wash-in is then further investigated by performing deuterium wash-out experiments on [3”H] and [5”2 H]2 GDP-L-fucose. This experiment demonstrates that in the reverse reaction, it is the C-3”  146  stereocenter inverted first followed by the C-5” epimerization. Later, deuterium wash-in experiments with GDP-6-deoxy-D-maimose showed some interesting results. It was observed that GDP-6-deoxy-D-mannose still remains as the major compound in the mixture upon extended incubation. Therefore it is postulated that the C-6” hydroxyl functionality is crucial for the C-5” epimerization and further catalysis steps cannot occur due to the nature of the sequential catalysis. It is also observed that no deuterium is incorporated onto GDP-6-deoxy-D-mannose and an interesting explanation is that the proton abstracted by Cys 145 cannot be exchanged with solvent derived deuterium without the corresponding epimerization. This “epimerization-restricted” proton exchange phenomenon is not new in enzymology and similar observations have been reported in the mechanistic studies of glutamate racemase. 10 Lastly, the specificity of the wildtype GME active site is explored with GDP-6-deoxy-D-altrose. It is thought that even though GME does not produce GDP-D-altrose, it may serve as an alternate substrate. The deuterium wash-in experiment with GDP-6-deoxy-D-altrose shows that GME can bind, oxidize, and deprotonate both C-3” and C-5” positions of GDP-D-altrose there by incorporating deuterium into both sites of epimerization. This indicates that although the epimerizations are ordered, some reaction via the opposite ordering is possible. In the later stage of the study, the mechanistic roles of the proposed catalytic residues, Lys217 and Cys 145, are examined. The experiment used here is strategically designed for simple product analysis. Base on the literature, Lys2l7Ala and Cysl45Ala are properly folded inactive mutants of GME. 64 However, the other catalytic residue in each of these mutants are not affected. Therefore the studies of these mutants not only provide clues about the catalytic role of the mutated residue, but more importantly also demonstrate the capability of the other residues to assist in catalysis. In addition, the inability of these mutants to possess a functional acid / base pair prevents any epimerization from occurring, thus simplifying the analysis.  147  The deuterium wash-in studies with Lys2 1 7Ala shows no deuterium incorporation into either GDP-D-mannose or GDP-L-fucose. Although it is anticipated that with GDP-D-mannose oxidation and deprotonation still occurs at C-5”, the inability of Cys145 to exchange its proton with solvent during the lifetime of the enol(ate) intermediate means no incorporation takes place. In the experiment with GDP-L-fucose, no deuterium is incorporated since the mutated Lys2 17 is proposed to be the base that is responsible for the deprotonation at C-3” in this reaction direction. More interesting results are found in the deuterium wash-in studies with Cysl45Ala. As expected, no deuterium is incorporated into C-5” of GDP-D-mannose because the mutation of Cys 145 stops catalysis at the first deprotonation step. However deuterium is incorporated into the C-3” position of GDP-L-fucose when the experiment is run in the reverse reaction. The studies with Cysl45Ala not only strongly support the proposed catalytic roles of Cys145 and Lys217, but also indicate that the epimerizations are sequential because no deuterium is washed into the C-5” of GDP-L-fucose without C-3” epimerization first occurring. A possible future experiment with GDP-L-gulose can be performed to complete the missing piece of these mechanistic studies. Deuterium wash-in studies using Lys2 1 7Ala and Cysl45Ala with this compound will not only solidify the proposed catalytic roles of Lys217 and Cys145, but will also provide further insight into the sequestered nature of Cys145 towards exchange of its proton with solvent (Figure 4.15). GDP-mannose 3,5-epimerase is a remarkable enzyme that catalyzes a very complex reaction involving an oxidation, two epimerizations, and a reduction. Unlike GFS which only produce one product, GME establishes an equilibrium mixture of GDP-D-mannose, GDP-L gulose, and GDP-L-galactose. The studies in this thesis support the previously proposed mechanism involving an initial oxidation at C-4”, then first sequential epimerization at C-5”,  148  followed by an epimerization at C-3”, and a final reduction of the C-4” hydroxyl. The roles of the basic and acidic residues involved in both of the epimerization steps have been assigned to Cys 145 and Lys2 17, respectively, which are in agreement with the literature.  4.6  Experimental  4.6.1 Over-expression and Purification of GME and Mutants  Plasmids donated by Prof. Naismith were transformed into host cells as instructed, and the enzymes were over-expressed, purified, and stored according to previous published procedures without any modification.M  4.6.2 Enzymatic Synthesis of GDP-L-Fucose and GDP-D-Altrose  GDP-L-fucose was made from GDP-D-mannose as described in section 2.8.3 and 2.8.4. GDP-D-altrose was obtained in a mixture with 25% GDP-L-fucose. This mixture of GDP-sugars was also made from GDP-D-mannose as described in section 2.8.3 and 2.8.4 with the exception of incubating with 20 mg of CyslO9Ser-GFS for 4 hours under otherwise identical conditions.  149  4.6.3 Enzymatic Synthesis of [5 ct..2H] -GDP-L-Fucose  A sample containing 20 mg of the lyophilized GDP-6-deoxy-4-keto-D-mannose (GDP-6deoxy-4-keto-D-mannose was obtained as described in section 2.8.3) was re-dissolved in 2.0 mL of distilled water, and 30 mg of NADPH and 20 mg of GFS (in 1.0 mL of storage buffer) were added. The reaction was incubated at 28 °C for 2 h and monitored by negative electrospray ionization mass spectrometry (-ESI-MS, GDP-fucose m/z mamiose rn/z  =  =  588, GDP-6-deoxy-4-keto-D-  586) to ensure it reached completion. The protein was removed using Amicon  Ultra centrifugal filter devices (Millipore) and the filtrate was lyophilized. This lyophilized fraction was then re-dissolved in 2.0 mL of D 0 and 20 mg of CyslO9Ser-GFS (previously 2 exchanged into deuterated buffer using Amicon Ultra centrifugal filter devices (Millipore)) was added to the mixture. The reaction was incubated at room temperature for 10 hours and the deuterium incorporation was confirmed by —ESI-MS (-ESI-MS, {5”H]-GDP-fucose mlz = 589). 2 Enzyme was removed using Amicon Ultra centrifugal filter devices (Millipore) and the 2 [5”H ]GDP-L-fucose was then purified as described in section 2.8.4. The position of the deuterium label was confirmed using ‘H NMR spectroscopy.  4.6.4 Enzymatic Synthesis of [3”H]-GDP-L-Fucose 2  A sample containing 20 mg of the lyophilized GDP-6-deoxy-4-keto-D-mannose (GDP-6deoxy-4-keto-D-mannose was obtained as described in section 2.8.3) was re-dissolved in 2.0 mL of D 0, and 30 mg of NADPH and 20 mg of GFS (previously exchanged into deuterated buffer 2  150  using Amicon Ultra centrifugal filter devices (Millipore)) were added. The reaction was incubated at 28 °C for 2 h and monitored by negative electrospray ionization mass spectrometry (-ESI-MS, [3”, 5”H]-GDP-fucose mlz 2  590, GDP-6-deoxy-4-keto-D-mannose mlz  =  586) to  ensure it reached completion. The protein was removed using Amicon Ultra centrifugal filter devices (Millipore) and the filtrate was lyophilized. This lyophilized fraction was then re dissolved in 2.0 mL of distilled water and 20 mg of Cysi O9Ser-GFS (in 1.0 mL of storage buffer) was added to the mixture. The reaction was incubated at room temperature for 10 hours and the deuterium incorporation was confirmed by —ESI-MS (-ESI-MS, [3”H]-GDP-fucose m!z 2  589).  Enzyme was removed using Amicon Ultra centrifugal filter devices (Millipore) and the {3”H]2 GDP-L-fticose was then purified as described in section 2.8.4. The position of the deuterium label was confirmed using 1 H NMR spectroscopy.  4.6.5 Deuterium Wash-in Experiments  Deuterium wash-in experiments were performed using wild-type GME, Lys2 1 7Ala, and Cysl45Ala. A sample of deuterated triethanolammonium chloride buffer (10 mM, pD 8.4) was prepared by repeated (3X) lyophilization of a phosphate buffer (10 mM, pH 8.0) and dissolution in an equal volume of D 0. Samples containing 2.0 mg of substrate (GDP-L-fucose, GDP-D 2 mannose, GDP-6-deoxy-D-mannose, or GDP-L-fucose / GDP-D-altrose mixture) and 2 mg of enzyme (previously exchanged into deuterated buffer using Amicon Ultra centrifugal filter devices (Millipore)) were prepared in the deuterated buffer (final volume 0.8 mL). The samples were incubated at 30 °C and the progress of deuterium wash-in was monitored using negative ESI mass spectrometry. In the deuterium wash-in experiment with Cysl45Ala and GDP-L  151  fucose, the position of deuterium incorporation was confirmed by examining the reaction mixture with ‘H NMR spectroscopy.  4.6.6 Deuterium Wash-out Experiments  Samples containing 2 mg of substrate ([3”H]-GDP-6-deoxy-4-keto-D-mannose, or [5”2 H]-GDP-6-deoxy-4-keto-D-mannose, [3”2 H]-GDP-L-fucose, or [5”2 H]-GDP-L-fucose 2  ), 3 mg  of GME (previously exchanged into the same buffer using Amicon Ultra centrifugal filter devices (Millipore)), were prepared in 10 mM triethanolanimonium chloride buffer (pH 8.0, 10 mM, 0.8 mL fmal volume). The samples were initiated by the addition of enzyme, incubated at room temperature, and monitored by negative ESI mass spectrometry in 15 minute intervals.  152  References 1  Moss, G. P.; Smith, P. A. S.; Tavemier, D.; Pure & Appi. Chem. 1995, 67, 13071375.  2  Tanner, M. E. Curr. Org. Chem. 2001, 5, 169-192.  3  Tanner, M. E. Acc. Chem. Res. 2002, 35, 237-246.  4  Jörnvall, H.; Persson, B.; Krook, M.; Atrian, S.; Gonzalez-Duarte, R.; Jeffery, J.; Ghosh, D. Biochemistry 1995, 34, 6003-6013.  5  Oppermann, U.; Filling, C.; Hult, M.; Shafqat, N.; Wu, X.; Lindh, M.; Shafqat, J.; Nordling, E.; Kallberg, Y.; Persson, B.; Jömvall, H. Chem. -Biol. Interact. 2003, 143-144, 247-253.  6  Blankenfeldt, W.; Kerr, I. D.; Giraud, M. F.; McMiken, H. J.; Leonard, G.; Whitfield, C., Messner, P.; Graninger, M.; Naismith, 3. H. Structure 2002, 10, 773-786.  7  Filling, C.; Bemdt, K. D.; Benach, J.; Knapp, S.; Prozoroviski, T.; Nordling, E.; Ladenstein, R.; Jornvall, H.; Oppermann, U. J Biol. Chem. 2002, 277, 2567725684.  8  Naismith, J. H. Chem. Soc. Rev. 2006, 35, 763-770.  9  Frey, P. A. FASEBJ. 1996, 10, 461-470.  10  Leoir, L. F. Arch. Biochem. Biophys. 1951, 33, 186-190.  11  Frey, P. A. Complex pyridine nucleotide-dependent transformations. In Pyridine Nucleotide Coenzymes: Chemical, Biochemical and Medical Aspects, ed.; Dolphin, D.; Poulson, R.; Avramovic, 0. Eds. John Wiley & Sons: New York, 1987; Vol. 2B, pp 461-5 11.  12  Maxwell, E. S. .1 Biol. Chem. 1957, 229, 139-151.  13  Kowalsky, A.; Koshland, D. E. Jr. Biochim. Biophys. Acta 1956, 22, 575-577.  14  Bevill, R. D.; Nordin, J. H.; Smith, F.; Kirkwood, S. Biochem. Biophys. Res. Commun. 1963, 12, 152-156.  15  Anderson, L.; Landel, A. M.; Diedrich, D. F. Biochim. Biophys. Acta 1956, 22, 573-575.  16  Kaicker, H. M.; Maxwell, E. S. Biochim. Biophys. Acta 1956, 22, 5 88-589.  153  17  Wilson, D. B.; Hogness, D. S. I Biol. Chem. 1969, 244, 2132-2136.  18  Bertland, A. U.; Kalckar, H. M. Proc. Nat!. Acad. Sc US.A. 1968, 61, 629-635.  19  Nelsestuen, G. L.; Kirkwood, S. I Biol. Chem, 1971, 246, 7533-7543.  20  Maitra, U. S.; Ankel, H. Proc. Nati. Acad. Sd. US.A. 1971, 68, 2660-2663.  21  Wee, T. G.; Frey, P. A. J. Biol. Chem. 1973, 248, 33-40.  22  Wee, T. G.; Davis, J.; Frey, P. A. I Biol. Chem. 1972, 247, 1339-1342.  23  Nelsestuen, G. L.; Kirkwood, S. Biochim. Biophys. Acta 1970, 220, 633-635.  24  Adair, W. L.; Gabriel, 0.; Ulirey, D.; Kaickar, H. M. I BioL Chem. 1973, 248, 4635-4639.  25  Glaser, L.; Ward, L. Biochim. Biophys. Acta 1970, 198, 613-615.  26  Kang, U. G.; Nolan, L. D.; Frey, P. A. I Biol. Chem. 1975, 250, 7099-7105.  27  Wong, S. S.; Frey, P. A. Biochemistry 1977, 16, 298-305.  28  Thoden, J. B.; Frey, P. A.; Holden, H. M. Biochemistry 1996, 35, 5137-5 144.  29  Thoden, J. B.; Holden, H. M. Biochemistry 1998, 37, 11469-11477.  30  Thoden, J. B.; Frey, P. A.; Holden, H. M. Biochemistry 1996, 35, 2557-2566.  31  Thoden, J. B.; Frey, P. A.; Holden, H. M. Protein Sci. 1996, 5, 2149-2161.  32  Swanson, B. A.; Frey, P. A. Biochemistry 1993, 32, 1323 1-13236.  33  Liu, Y.; Thoden, J. B.; Kim, J.; Berger, E.; Gulick, A. M.; Ruzicka, F. J.; Holden, H. M.; Frey, P. A. Biochemistry 1997, 36, 10675-10684.  34  Thoden, J. B.; Wohiers, T. M.; Fridovich-Keil, J. L.; Holden, H. M. Biochemistry 2000, 39, 5691-5701.  35  Thoden, J. B.; Wohiers, T. M.; Fridovich-Keil, J. L.; Holden, H. M. I Biol. Chem. 2001, 276, 15131-15136.  36  Thoden, J. B.; Henderson, J. M.; Fridovich-Keil, J. L.; Holden, H. M. I Biol. Chem. 2002,277, 27528-27534.  37  Kawamura, T.; Ishimoto, N.; Ito, E. I Biol. Chem. 1979, 254, 8457.  38  Kawamura, T.; Kimura, M.; Yamamori, S.; Ito, E. I Biol. Chem. 1978, 253, 3595.  154  39  Keppler, 0. T.; Hinderlich, S.; Langner, J.; Schwartz-Albiez, R.; Reutter, W.; Pawlita, M. Science 1999, 284, 1372.  40  Sala, R. F.; Morgan, P. M.; Tanner, M. E. I Am. Chem. Soc. 1996, 118, 30333034.  41  Morgan, P. M.; Sala, R. F.; Tanner, M. E. I Am. Chem. Soc. 1997, 119, 1026910277.  42  Salo, W. Biochim. Biophys. Acta 1976, 452, 625.  43  Tanner, M. E.; Kenyon, G. L. in Comprehensive Biological Catalysis, Sinnott, M., Ed. Academic Press: San Diego, 1998, Vol. II, pp 7-41.  44  Glaser, L. The Enzymes 1972, 6, 335.  45  Graninger, M.; Nidetzky, B.; Heinrichs, D. E.; Whitfield, C.; Messner, P. 1 Biol. Chem. 1999, 274, 25069.  46  Kirkpatrick, P. N.; Scaife, W.; Hallis, H. S.; Liu, H. S.; Spencer, J. B.; Williams, D. H. Chem. Commun. 2000, 1565-1566.  47  Tello, M.; Jakimowicz, P.; Errey, J. C.; Freel Meyers, C. L.; Walsh, C. T.; Buttner, M. J.; Lawson, D. M.; Field, R. A. Chem. Commun. 2006, 1079-1081.  48  Chen, H. W.; Thomas, M. G.; Hubbard, B. K.; Losey, H. C.; Walsh, C. T.; Burkart, M. D. Proc. Nati. Acad. Sci. US.A. 2000, 97, 11942-11947.  49  Dong, C.; Major, L. L.; Srikannathasan, V.; Errey, J. C.; Giraud, M. F.; Lam, J. S.; Graninger, M.; Messner, P.; McNeil, M. R.; Field, R. A.; Whitfield, C.; Naismith, J. H. I Mol. Biol. 2007, 365, 146-159.  50  Melo, A.; Glaser, L. I Biol. Chem. 1968, 243, 1475-1478.  51  Dong, C.; Major, L. L.; Allen, A.; Blankenfeldt, W.; Maskell, D. J.; Naismith, J. H. Structure 2003, 11,715-723.  52  Merkel, A. B.; Major, L. L.; Errey, 3. C.; Burkart, M. D.; Field, R. A.; Walsh, C. T.; Naismith, J. H. I Biol. Chem. 2004, 279, 32684-32691.  53  Jakimowicz, P.; Tello, M.; Freel Meyers, C. L.; Walsh, C. T.; Buttner, M. J.; Field, R. A.; Lawson, D. M. Proteins: Struct. Funct. Bioinf 2006, 63, 261-265.  54  De Flora, A.; Morelli, A.; Benatti, U.; Giuliano, F. Arch. Biochem. Biophys. 1975, 169, 362-363.  55  De Flora, A.; Morelli, A.; Benatti, U.; Frascio, M.; Giuliano, F. FEBSLett. 1977, 82, 223-226.  155  56  Sullivan, F. X.; Stahl, M.; Kumar, R.; Kriz, R.; Stahl, M.; Xu, G.-Y.; Rouse, J.; Chang, X.; Boodhoo, A.; Potvin, B.; Cumming, D. A. J Biol. Chem. 1998, 273, 8193-8202.  57  Rizzi, M.; Tonetti, M.; Vigevani, P.; Sturla, L.; Bisso, A.; De Flora, A.; Bordo, D.; Bolognesi, M. Structure 1998, 6, 1453-1465.  58  Somers, W. S.; Stahl, M. L.; Sullivan, F. X. Structure 1998, 6, 1601-1612.  59  Chang, S.; Duerr, B.; Serif, G. I Biol. Chem. 1988, 263, 1693-1697.  60  Menon, S.; Stahl, M.; Kumar, R.; Xu, G. Y.; Sullivan, F. I Biol. Chem. 1999, 274, 26743-26750.  61  Rosano, C.; Bisso, A.; Izzo, G.; Tonettis, M.; Strula, L.; De Flora, A.; Bolognesi, M. IMol. Biol. 2000, 303, 77-91.  62  Jörnvall, H.; person, B.; Krook, M.; Atrian, S.; Gonzàles-Duarte, R.; Jeffery, J.; Ghosh, D. Biochemistry, 1995, 34, 6003-6013.  63  Gerratana, B.; Cleland, W. W.; Frey, P. A. Biochemistry, 2001, 40, 9187-9195.  64  Major, L. L.; Wolucka, B. A.; Naismith, J. H. I Am. Chem. Soc. 2005, 127, (51), 18309-18320.  65  Wolucka, B. A.; Persiau, G.; Van Doorsselaere, J. V.; Davey, M. W.; Demol, H.; Vandekerckhove, J.; Van Montagu, M.; Zabeau, M.; Boerjan, W. Proc. Nati. Acad. Sci. US.A. 2001, 98, 14843-14848.  66  Wolucka, B. A.; Van Montagu, M. I. Biol. Chem. 2003, 278, 47483-47490.  67  Barber, G. A. J. Biol. Chem. 1979, 254, 7600-7603.  68  Flowers, H. M. Advances in Carbohydrate Chem. and Biochem. 1981, 39, 279-345.  69  Becker, D. J., Lowe, J. B. Glycobiology 2003, 13-7, 41R-53R.  70  Albermann, C.; Distler, J.; Piepersberg, W. Glycobiology 2000, 10-9, 875-881.  71  Watkins, W. M. Transfus. Med. 2001, 11, 243-265.  72  Lowe, J. B. Curr. Opin. Cell. Biol. 2003, 15, 53 1-538.  73  Yakubenia, S.; Wild, M. K. FEBSI 2006, 273, 4390-4398.  74  Miyoshi, E.; Moriwaki, K.; Nakagawa, T. I Biochem. 2008, 143, 625-729  75  Stanley, P. Curr. Opin. Struct. Biol. 2007, 17, 530-535.  156 76  Turnock, D. C.; Izquierdo, L.; Ferguson, M. A. J. .J Biol. Chem. 2007, 282, 2885328863.  77  Maki, M.; Reniconen, R. Glycobiology 2004, 14, 1R-15R  78  Wirth, H. P.; Yang, M. 4605.  79  Ginsburg, V. I Biol. Chem. 1960, 235, 2 196-2201.  80  Ginsburg, V. .1 Biochim. Biophys. 1960, 54, 376-378.  81  Ginsburg, V. I Biol. Chem. 1961, 236, 2389-2393.  82  Lamrabet, Y.; Bellogin, R. A.; Cubio, T.; Espuny, R.; Gil, A.; KrishMan, H. B.; Megias, M.; Ollero, F. 3.; Pueppke, S. G.; Ruiz-Sainz, J. E. Mol. Plant Microbe. Interact. 1999, 12, 207-2 17.  83  Ohyama, C.; Smith, P. L.; Angata, K.; Fukuda, M. N.; Lowe, J. B.; Fukuda, M. I Biol. Chem. 1998, 273, 14582-14587.  84  Roos, C.; Koimer, M.; Mattila, P.; Renkonen, R. I Biol. Chern. 2002, 277, 3 168-3175.  85  Somoza, J. K.; Menon, S.; Schmidt, H.; Joseph-McCarthy, D.; Dessen, A.; Stahl, M. L.; Somers, W. S.; Sullivan, F. X. Strucutre Fold Des. 2000, 8, 123-135.  86  Noctor, G.; Foyer, C. H., Annu. Rev. Plant Physiol. Plant Mo!. Biol. 1998, 49, 249279.  87  Smirnoff, N. Ann. Bot. 1996, 78, 66 1-669.  88  Smirnoff, N. Curr. Opin. Plant Rio!. 2000, 3, 229-23 5.  89  Loewus, F. A.; Loewus, M. W. CRC Crit. Rev. Plant Sci. 1987, 5, 10 1-119.  90  Arrigoni, 0.; De Tullio, M. C. Biochem. Biophys. Acta 2002, 1569, 1-9.  91  Pastori, G. M.; Kiddie, G.; Antoniw, 3.; Bernard, S.; Veijovic-Jovanovic, S.; Verrier, P J.; Noctor, G.; Foyer, C. H. Plant Cell 2003, 15, 939-95 1.  92  Wheeler, G. L.; Jones, M. A.; Smirnoff, N. Nature 1998, 393, 365-369.  93  Hancock, R. D.; McRae, D.; Haupt, S.; Viola, R. BMC Plant Biol. 2003, 3, 7.  94  Laing, W. A.; Bulley, S.; Wright, M.; Cooney, J.; Jensen, D.; Barraclough, D.; MacRae, E. Proc. Nat!. Acad. Sci. US.A. 2004, 101, 16976-16981.  95  Gatzek, S.; Wheeler, G. L.; Smirnoff, N. Plant I 2002, 30, 541-553.  Q.; Karita, M.; Blaswer, M. J. Infect.  Imumn. 1996, 64, 4598-  157  96  Oba, K., Fukui, M.; Imai, Y.; Iriyama, S.; Nogami, K. Plant Cell Physiol. 1994, 35, 473-478.  97  Ostergaard, J.; Persiau, G.; Davey, M. W.; Bauw, G.; van Montagu, M. .1 BioL Chem. 1997, 272, 30009-300 16.  98  Imai, T.; Karita, S.; Shiratori, G.; Hattori, M.; Nunome, T.; Oba, K.; Hirai, M. Plant Cell Physiol. 1998, 39, 1350-1358.  99  Stevenson, G.; Andrianopoulos, K.; Hobbs, M.; Reees, P. R. I Bacteriol. 1996, 178, 48 85-4893.  100  Wolucka, B. A.; Goossens, A.; Inzé, D. I Exp. Bot. 2005, 56, 2527-253 8.  101  Stick, R. V.; Watts, A. G. Monatshefte für Chemie 2002, 133, 541-554.  102  Lal, G. S.; Pez, G. P.; Syvret, R. G. Chem. Rev. 1996, 96, 173 1-1755.  103  Timmons, S. C.; Jakeman, D. L. Org. Lett. 2007, 9, 1227-1230.  104  Bradford, M. M. Anal. Biochem. 1976, 72, 248-254.  105  Glavas, S.; Tanner, M. E. Biochemistry 1999, 38, 4106-4113.  106  Rentmeister, A.; Hoh, C.; Weidner, S.; Drager, G.; Elling, L.; Liese, A.; Wandrey, C. Biocatalysis and Biotransformation 2004, 22, 49-5 6.  107  Garen, A.; Levinthal, C. Biochim. Biophys. Acta. 1960, 38, 470-83.  108  Gokhale, U.; Hindshaul, 0.; Palcic, M. M. Can. I. Chem. 1990, 68, 1063-1071.  109  Schmidt, R. R.; Wegmann, B.; Jung, K. H. Liebigs Ann. Chem.1991, 121-124.  

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:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0061230/manifest

Comment

Related Items