Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mechanistic studies on UDP-N-acetylglucosamine 2-epimerase Sala, Rafael F. 1999

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

Item Metadata

Download

Media
831-ubc_1999-389707.pdf [ 7.28MB ]
Metadata
JSON: 831-1.0061874.json
JSON-LD: 831-1.0061874-ld.json
RDF/XML (Pretty): 831-1.0061874-rdf.xml
RDF/JSON: 831-1.0061874-rdf.json
Turtle: 831-1.0061874-turtle.txt
N-Triples: 831-1.0061874-rdf-ntriples.txt
Original Record: 831-1.0061874-source.json
Full Text
831-1.0061874-fulltext.txt
Citation
831-1.0061874.ris

Full Text

Mechanistic Studies on UDP-7V-Acetylglucosamine 2-Epimerase by RAFAEL F. SALA B.Sc, Pontificia Universidad Catolica del Peru, 1989 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Department of Chemistry We accept this thesis as conforming To the required standard:  THE UNIVERSITY OF BRITISH COLUMBIA January, 1999 ©Rafael F. Sala, 1999  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, 1 agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  Department of  C /-rfc /^ 7  I^H? /  The University of British Columbia Vancouver, Canada Date  DE-6 (2/88)  1  11  Abstract  The epimerase)  bacterial catalyzes  enzyme the  UDP-iV-acetylglucosamine 2-epimerase interconversion  between  the  sugar  (UDP-GlcNAc nucleotide  2-  UDP-JV-  acetylglucosamine (UDP-GlcNAc) and its C-2" epimer UDP-N-acetylmarinbsamine (UDPManNAc). This enzyme differs from known racemases and epimerases in that it must invert a stereogenic center that bears a non-acidic proton. It must therefore employ a mechanism that is more complicated than a simple inversion by deprotonation followed by reprotonation on the opposite face of the reaction center. Previous studies on the epimerase obtained from natural sources have revealed that the substrate (UDP-GlcNAc) is required as an activator in the reverse reaction (UDP-ManNAc to UDP-GlcNAc), and that the enzyme is highly specific for its normal substrates. In addition, evidence was obtained suggesting that a proton transfer was ultimately responsible for the epimerization.  A mechanism was suggested in which the epimerization was facilitated by  increasing the acidity of the C-2" proton via transient formation of a sugar nucleotide ketointermediate (Kawamura et al., 1978: 1979; Salo, 1976). These results have been corroborated on a recombinant E. coli UDP-GlcNAc 2-epimerase and evidence supporting an alternate mechanism has been obtained. A substrate labeled with 0 at the anomeric position was prepared in order to be used in 18  Positional Isotope Exchange (PIX) experiments. These experiments test whether the anomeric bond is cleaved during the epimerization.  The scrambling of an  O label located at the p-  31  phosphate moiety of UDP-epimers was detected by  P N M R spectroscopy. This provided  strong evidence for a mechanism involving cleavage of the anomeric bond and formation of 2acetamidoglucal and UDP as intermediates.  Ill  Additional studies with alternative substrates were used to gain more information about the mechanism of the epimerase. A 3'-deoxy analog (3'-deoxy-UDP-GlcNAc) was prepared and incubated with the epimerase to test for mechanisms involving the formation of a ketointermediate at this center. The absence of epimerization with this substrate of the epimerase failed to rule out any such mechanisms, but suggested the need of the 3"-hydroxyl group as a necessary element for the recognition of the substrate. A trifluoroacetamido analog (UDPGICNAC-F3) was also prepared and proved to be an alternative substrate of the enzyme with a kcat of 2.6 x 10" s" an apparent K of 3.40 ± 0.37 m M and a Hill coefficient of 1.6 ± 0.2. This 2  1  m  compound was also an acceptable substrate for two sugar transferases. The possible involvement of neighboring group participation of the acetamido group during the epimerization was also tested by incubation of the oxazoline of ManNAc with the epimerase under conditions that permitted the detection of enzyme catalyzed formation of 2acetamidoglucal or sugar nucleotides from this putative intermediate. obtained  in this  experiment  along with the behavior  The negative result  of the epimerase  with the  trifluoroacetamido analog suggests (but does not prove) that neighboring group participation is not involved in the epimerization. The minimal description of the mechanism of the epimerase is unprecedented and involves the anti elimination of UDP from UDP-GlcNAc (syn from UDP-ManNAc) to form 2acetamidoglucal and UDP as reaction intermediates followed by a syn addition (anti if started from UDP-ManNAc) of UDP to the double bond of the glycal intermediate to generate the sugar nucleotide epimer. Precedents for this process as well as a more detailed discussion of the type of elimination involved and future directions are also presented.  iv  Table of Contents Abstract  ii  Table of Contents  iv  List of Figures  ix  List of Tables  xii  Abbreviations and Symbols  xiii  Acknowledgements  xvi  Dedication  xvii  Chapter One: Introduction  1  1.1  Racemases and Epimerases  2  1.2  Enzymatic Epimerization of Carbohydrates  4  1.3  UDP-TV-Acetylglucosarnine 2-Epimerase  9  1.3.1 Mammalian Enzyme  9  1.3.2 Bacterial Enzyme  14  Previous studies with the bacterial UDP-GlcNAc 2-Epimerase  16  1.4.1 Distribution of the bacterial UDP-GlcNAc 2-epimerase  18  1.4.2 Isolation and purification of the enzyme  18  1.4  1.4.3 Effects of enzyme concentration, pH, N A D , cations and nucleotides  19  1.4.4 Requirement of UDP-GlcNAc in the epimerization of UDP-ManNAc  20  1.4.5 Effect of Substrate concentration  21  1.4.6 Substrate specificity  21  1.4.7 Equilibrium Constant  21  1.4.8 Enzyme assay  22  Proposed mechanistic pathways followed by the bacterial UDP-GlcNAc 2-epimerase  23  Aims of this Thesis  26  +  1.5 1.6  V  Chapter Two: Initial studies with cloned UDP-GlcNAc 2-Epimerase  27  2.1  Introduction  28  2.2  Overexpression of cloned UDP-GlcNAc 2-epimerse in E. coli  29  2.3  Purification of cloned UDP-GlcNAc 2-epimerase  31  2.4  Solvent Deuterium Isotope Incorporation  31  2.5  Development of a coupled assay and kinetic characterization of UDP-GlcNAc 2-epimerase  36  Synthesis of UDP-ManNAc and its use in the evaluation of the Regulation of UDP-GlcNAc 2-epimerase  38  2.6  2.6.1 General strategies for the synthesis of glycosyl nucleosides diphosphates  38  2.6.2 Synthesis of UDP-ManNAc  44  2.6.3 Evaluation of the regulation of UDP-GlcNAc 2-epimerase  46  2.7  Conclusions  47  2.8  Experimental Methods  48  2.8.1 General  48  2.8.2 General synthetic procedures  49  (a) Catalytic Hydrogenation over Pd/C  49  (b) Transesterification with sodium methoxide (O-deacetylation)  49  (c) General procedure for the preparation of uridine 5'(2-Acetamido2-deoxy-a-D-hexopyranosyl diphosphates  50  2.8.3 Synthesis of the dilithium salt of uridine 5'(2-Acetamido2-deoxy-a-D-mannopyranosyl diphosphates (UDP-ManNAc))  51  2.8.4 Protein Determination  52  2.8.5 Definition of a Unit  52  2.8.6 Strains and Media  53  2.8.7 Overexpression and purification of the cloned UDP-GlcNAc 2-epimerase . . . 53 2.8.8 Solvent Deuterium Isotope Incorporation  54  2.8.9 Overexpression and purification of the cloned UDP-ManNAc dehydrogenase  55  2.8.10 Kinetic characterization of UDP-GlcNAc 2-epimerase  56  2.8.11 Evaluation of the regulation of UDP-GlcNAc 2-epimerase  57  vi Chapter Three: Positional Isotope Exchange (PIX) Studies with UDP-GlcNAc 2-Epimerase  58  3.1  Introduction  59  3.2  Application of PIX on the study of UDP-GlcNAc 2-Epimerase  66  3.3  Synthesis of labeled substrate for the PIX experiment  67  3.4  PIX experiment under reversible conditions  75  3.5  PIX experiment under irreversible conditions  78  3.6  Mechanistic implications  81  3.7  Experimental Methods  82  3.7.1 General  82  3.7.2 Attempted synthesis of 2-Acetamido-3,4,6-tri-0-acetyl-2-deoxy1- 0-a-D-glucopyranose (11a) by acid-catalyzed hydrolysis of 2- methyl-(3,4,6-tri-0-acetyl-1,2-dideoxy-a-D-glucopyrano)-[2,1 - d]2-oxazoline (14) 18  3.7.3 Preparation of 2-Acetamido-3,4,6-tri-0-acetyl-2-deoxy-[ 1 - O J - a D-glucopyranose (11a)  82  18  83  3.7.4 Preparation of Disodium Uridine-5'-(2"-Acetamido-2"18 deoxy-[l"- 0]-a-D-glucopyranosyl-diphosphate), 18  0-labeled UDP-GlcNAc (10)  83  3.7.5 PIX experiment under reversible conditions  84  3.7.5 PIX experiment under irreversible conditions  85  Chapter Four:  Mechanistic Studies on UDP-GlcNAc 2-Epimerase with Alternative substrates  86  4.1  Introduction  87  4.2  Studies with 3 '-deoxy analogue of UDP-GlcNAc  91  4.2.1 Synthesis of the 3"-deoxy analogue of UDP-GlcNAc  92  4.2.2 Attempted epimerization of 3"-deoxy-UDP-GlcNAc with UDP-GlcNAc 2-epimerase 4.2.3 Mechanistic implications 4.3  4.4  93 98  Possible involvement of neighboring group participation during the enzymatic epimerization  99  Studies with the trifluoroacetamido analogue of UDP-GlcNAc 4.4.1 Introduction  .105 105  vii  4.5  4.6  4.7  4.4.2 Synthesis of the trifluoroacetamido analogue of UDP-GlcNAc  106  4.4.3 Epimerization of the trifluoroacetamido analogue of UDP-GlcNAc with UDP-GlcNAc 2-epimerase  109  4.4.4 Kinetic characterization of the trifluoroacetamido analogue ofUDP-GlcNAc  112  4.4.5 Mechanistic implications  113  Reaction of the oxazoline of ManNAc and UDP with UDP-GlcNAc 2-epimerase  115  4.5.1 Mechanistic implication  119  Use of trifluoroacetamido analogue of UDP-GlcNAc in reactions with sugar transferases  119  Experimental Methods  123  4.7.1 General  123  4.7.2 Synthesis of 3"-deoxy analog of UDP-GlcNAc (17)  124  (a) (b) (c)  Synthesis of Benzyl 2-acetamido-4,6-0-isopropylidene-2-deoxy3-0-phenoxy-thiocarbonyl-a-D-glucopiranoside (21)  124  Synthesis of Benzyl-2-acetamido-4,6-0-isopropylidene2,3-dideoxy-a-D-ribohexo-pyranoside (22)  125  Synthesis of Disodium uridine 5'-(2-acetamido-2,3-dideoxy-a-Dribohexo-pyranosyl diphosphate), 3"-deoxy UDP-GlcNAc (17)  126  4.7.3 Attempted epimerization of 3"-deoxy-UDP-GlcNAc with UDP-GlcNAc 2-epimerase  126  4.7.4 Test of 3"-deoxy-UDP-GlcNAc as an inhibitor of UDP-GlcNAc 2-epimerase  126  4.7.5 Test of 3"-deoxy-UDP-GlcNAc as a specific activator of UDP-GlcNAc 2-epimerase 4.7.6 Synthesis of trifluoracetamide analogue of UDP-GlcNAc (19) (a) (b) (c) (d) (e)  "  127 127  2-Trifluoroacetamido-3,4,6-tri-0-acetyl-2-deoxyct-D-glucose (31)  127  1,5-Anhydro-2-deoxy-2-trifluoroacetamido-3,4,6-tri-0acetyl-D-arabino-hex-l-enitol (32)  128  2-Trifluoromethyl-(3,4,6-tri-0-acetyl-l,2-dideoxy-a-Dglucopyrano)-[2,l-d]- oxazoline (33)  128  Diphenyl (2-trifluoroacetamido-3,4,6-tri-0-acetyl-2deoxy-a-D-glucopyranosyl) phosphate (34)  129  Monotriethylammonium salt of 2-trifluoroacetamido-3,4,6tri-O-acetyl-2-deoxy-a-D-glucopyranosyl phosphate (35)  130  Vlll  (f)  Monopyridinium salt of 2-trifluoroacetamido-2-deoxya-D-glucopyranosyl phosphate (36)  130  (g)  Preparation of 36 by the MacDonald procedure  131  (h)  Dilithium uridine 5'-(2-trifluoroacetamido-2-deoxy-a-Dglucopyranosyl diphosphate) (19)  131  4.7.7 Epimerization of the trifluoroacetamido analogue of UDP-GlcNAc with UDP-GlcNAc 2-epimerase  132  4.7.8 Kinetic characterization of the trifluoroacetamido analogue ofUDP-GlcNAc  132  4.7.9 Preparation of the oxazoline of ManNAc (18)  133  4.7.10 Attempted coupling of the oxazoline of ManNAc and UDP with UDP-GlcNAc 2-epimerase  133  Chapter Five: Conclusions and Future Directions  134  5.1  Introduction  135  5.2  (3-Eliminations  136  5.2.1 Background  136  5.2.1 Precedence in enzymatic reactions  141  5.3  Elimination-addition reactions in UDP-GlcNAc 2-epimerase  146  5.4  Future directions  148  Appendix 1 Regulatory enzymes and Hill equation  153  Appendix 2 Coupled Enzyme Assays  156  References  158  List of Figures Figure 1.1  Schematic representation of (a) a one-base, (b) a two-base mechanism for enzymatic inversion of stereocenters  4  Figure 1.2  Proposed mechanism of UDP-galactose 4-epimerase  6  Figure 1.3  Proposed mechanism for the enzymatic 3,5-epimerization of GDP-D-mannose  7  Proposed mechanism of epimerization of D-ribulose-5-phosphate 3-epimerase  8  A proposed mechanism for the reaction catalyzed by L-ribulose 5-phosphate 4-epimerase  9  Figure 1.4 Figure 1.5 Figure 1.6  Sequence of reactions proposed by Spivak and Roseman for the enzymatic epimerization/hydrolysis of UDP-GlcNAc by the mammalian epimerase . . . 10  Figure 1.7  Mechanism proposed by Salo and Fletcher for the mammalian UDP-GlcNAc 2-epimerase  Figure 1.8  Mechanism proposed by Sommar and Ellis for the mammalian enzyme UDP-GlcNAc 2-epimerase  Figure 1.9  12  14  Reaction catalyzed by the bacterial enzyme UDP-GlcNAc 2-epimerase . . . . 15  Figure 1.10 Enzymatic formation of UDP-ManNAcUA  16  Figure 1.11 Path A: Mechanism proposed by Salo for the bacterial UDP-GlcNAc 2 epimerase  18  Figure 1.12 Alternate mechanistic pathway (Path B) for the enzymatic catalysis of UDP-GlcNAc 2-epimerase  24  Figure 1.13 A proposed mechanistic pathway (Path C) for the enzymatic catalysis of the epimerase involving nucleophilic attack at the p-phosphorus  25  Figure 1.14 A proposed 'hybrid' mechanistic pathway (Path D) for the enzymatic catalysis of the epimerase  26  Figure 2.1 Figure 2.2 Figure 2.3  Figure 2.4 Figure 2.5  Structure of UDP-GlcNAc showing the protons whose *H N M R signals undergo changes during the epimerization reaction at C-2"  32  Expansion of the 'H N M R ( D 2 O , 500 MHz) spectrum acquired during the enzymatic epimerization of UDP-GlcNAc  33  Newman projections showing the antiperiplanar relationship between (a) C-2" and the p-phosphorous atoms, and (b) H-2" and the anomeric oxygen atom  34  Plot of the initial rate of epimerization of UDP-GlcNAc versus its concentration  38  Enzyme-based synthesis of UDP-Glc  39  X Figure 2.6  Methods for the chemical synthesis of glycosyl nucleoside diphosphates . . . 40  Figure 2.7  Preparation of 2-acetamido-glycosyl phosphates by nucleophilic attack at the anomeric center  Figure 2.8  Figure 2.9  Preparation of 2-acetamido-glycosyl phosphates by nucleophilic attack at the phosphorus  43  Procedure for the synthesis of UDP-ManNAc  45  Figure 2.10 Ion-paired reversed-phase HPLC traces of a sample containing Figure 3.1  42  47  (a) Overall reaction catalyzed by Glutamine synthetase (bj Proposed stepwise mechanism for the action of the enzyme  61  Figure 3.2  PIX experiment on ATP labeled substrate  62  Figure 3.3  Reaction catalyzed by AcetylCoA synthetase  63  Figure 3.4  P-NMR spectrum at 145.7 MHz of inorganic phosphate randomly labeled with 44% O and 66% 0 . Test of C - 0 anomeric bond cleavage by PIX during the enzymatic l s  Figure 3.5  1 6  65  epimerization of UDP-GlcNAc  67  Figure 3.6  Synthesis of 0 labeled substrate  68  Figure 3.7  Reported acid-catalyzed hydrolysis of GlcNAc oxazoline  Figure 3.8  Mass spectra fragmentation pattern analysis for incorporation of 0 at the anomeric center Expanded carbonyl resonance region of high resolution C N M R for  69  the product of acid catalyzed hydrolysis of oxazoline of GlcNAc  71  Figure 3.9  1 8  l 8  13  Figure 3.10 Possible pathways for the hydrolysis of GlcNAc oxazoline  72 77  P spectra (D 0) of the p-phosphorus region of 0 labeled UDP-GlcNAc. A: before irreversible enzymatic epimerization. B: after epimerization  80  Formation of intermediates in the reaction catalyzed by UDP-GlcNAc 2- epimerase  89  3 1  Figure 3.12  3 1  Figure 4.2  :  P spectra (D 0) of the (3-phosphorus region of 0 labeled UDP-GlcNAc. A: before enzymatic epimerization. B: after enzymatic epimerization  Figure 3.11  Figure 4.1  68  1 8  2  1 8  2  Alternative pathway for the formation of 2-acetamidoglucal through Path D  90  Figure 4.3  Synthesis of 3 "-deoxy UDP-GlcNAc  92  Figure 4.4  Ion-paired reversed-phase HPLC traces of a sample containing : (A) UDP-ManNAc and 3"-deoxy-UDP-GlcNA c; (B) UDP-ManNAc and UDP-GlcNAc before and after incubation with the enzyme  94  Region of *H N M R spectra showing H - l " and H-6 of 3- deoxy-UDP-GlcNAc  96  !  Figure 4.5  xi Figure 4.6  31  P N M R spectrum of a mixture containing 3"-deoxy-UDP-GlcNAc  and UDP-GlcNAc after incubation with epimerase  97  Figure 4.7  Formation of axazolinium ion during glycosylation reactions  100  Figure 4.8  Formation of oxazolinium ions through SNi mechanisms  101  Figure 4.9  Proposed JV-Acetyl-p-hexosaminidase mechanism  103  Figure 4.10 Proposed mechanism for UDP-GlcNAc 2-epimerase involving oxazoline-type intermediates  104  Figure 4.11 Schematic representation of the synthetic routes for the preparation of UDP-GlcNAc-F  108  3  Figure 4.12 Relevant regions of *H and F N M R spectra during the incubation 19  of UDP-GlcNAc-F3 with epimerase  110  Figure 4.13 Proposed base-catalyzed hydrolysis of 2-trifluoroacetamidoglucal  112  Figure 4.14 A plot of initial velocity of UDP-GlcNAc-F3 epimerization as a function of UDP-GlcNAc-F3 concentration  113  Figure 4.15 *H N M R spectra of a sample containing oxazoline of ManNAc, UDP and UDP-GlcNAc-F , after different incubation periods with epimerase . . . 118 3  Figure 4.16 The reaction catalyzed by galactosyl transferase  120  Figure 4.17 Reaction catalyzed by "Core-2"-GlcNAc-transferase  121  Figure 4.18 Reaction catalyzed by GlcNAc-transferase V  122  Figure 4.19 Removal of the trifluoroacetamido group in tetrasaccharide 40a  123  Figure 5.1  The three types of mechanisms for (3-elimination reactions  138  Figure 5.2  Variable transition state theory of elimination reactions  139  Figure 5.3  More O'Ferrall diagram for description of the mechanism of P-elimination reactions  140  Figure 5.4  Production of alkenes by anti and syn elimination reactions  141  Figure 5.5  Phosphoenolpyruvate transferring enzymes that catalyzes ^n-elimination on an unactivated center, (a) UDP-GlcNAc enolpyruvyl transferase (MurA) (b) 5-enolpyruvyl-shikimate synthase 143  Figure 5.6  Three possible mechanism for the p-elimination of UDP from UDP-GlcNAc  146  xii  List of Tables Table 5.1 Table 5.2  Some 1,2-elimination/addition enzymatic reactions involving sy«-stereochemistry  144  Some 1,2-elimination/addition enzymatic reactions involving a«ri-stereochemistry  145  Abbreviations and Symbols  8  chemical shift  s  extinction coefficient  Abs  absorbance  Ac  acetate  Ad  Adenine  ADP  adenosine 5'-diphosphate  a. k. a.  also known as  ATP  adenosine 5'-triphosphate  B. cereus  Bacillus cereus  CH CN  acetonitrile  CoA  Coenzyme A  d  doublet (NMR)  D2O  deuterium oxide  Da  dalton(s)  dd  doublet of doublets (NMR)  ddd  doublet of a doublet of doublet (NMR)  DEAE  diethylaminoethyl  DNA  deoxyribonucleic acid  DTT  dithiothreitol  E. coli  Escherichia coli  EDTA  ethylene diamine tetraacetate (disodium salt)  Enz  enzyme  Glu  Glutamic acid  Gin  Glutamine  GDP  guanosine 5'-diphosphate  h  hour(s)  3  .  xiv HPLC  high pressure/performance liquid chromatography  Hz  hertz  IPTG  isopropyl-1 -thio-p-D-galactopyranoside  J  coupling constant (in NMR)  k  catalytic rate constant (turnover number)  kDa  kilodalton(s)  cat  *  Kj  dissociation constant for an enzyme-inhibitor complex  K  Michaelis constant  LB  Luria Bertani medium  LSIMS  Liquid Soft Ionization Mass Spectrometry  min  minute(s)  m  NAD  nicotinamide adenine dinucleotide  +  NADH NADP  nicotinamide adenine dinucleotide, reduced form +  nicotinamide adenine dinucleotide phosphate  NADPH  nicotinamide adenine dinucleotide phosphate, reduced form  NMR  nuclear magnetic resonance  m  multiplet (NMR)  PAGE  polyacrylamide gel electrophoresis  Pi  Inorganic phosphate  PLP  pyridoxal phosphate  ppm  parts per million  psi  pounds per square inch  pUSOl  plasmid containing gene encoding for UDP-ManNAc dehydrogenase  rpm  revolutions per minute  s  singlet (NMR); second(s)  S. pneumonaie  Streptococcus pneumoniae  S. typhimurium  Salmonella typhimurium  SDS  sodium dodecyl sulphate  t  triplet (NMR)  T4  a bacteriophage  TBAHS  tetrabutylammonium hydrogen sulphate  Trien  triethanolamine  Tris  tris(hydroxy)amino methane  U  Uridine  UDC  uridine 5'-diphosphate chloroacetol  UDP  uridine diphosphate  UDP-Gal  uridine 5'-diphosphate galactose  UDP-Glc  uridine 5'-diphosphate glucose  UDP-GlcNAc  uridine 5'-diphosphate 7V-acetylglucosamine  UDP-glycidol  uridine 5'-diphosphate 2,3-epoxypropanol  UDP-ManNAc  uridine 5'-diphosphate 7V-acetylmannosamine  UDP-ManNAcUA  uridine 5'-diphosphate A^-acetylmannosaminuronic  UMP  uridine 5'-monophosphate  UV  ultraviolet  Vis  visible  XVI  Acknowledgements I wish to thank my supervisor, Dr. Martin Tanner for his advice, encouragement and particularly for his" patience. I would like to recognize the help and friendship of the members of the Tanner's group especially to Rob Campbell and Suzana Glavas who helped me to make some sense out of my sometimes undecipherable 'Spanglish' writing and for most valuable and stimulating discussions. I would like to thank the people in the Withers' group particularly Renee Mosi and David Vocadlo, the people in the N M R facility specially Dr. Nick Burlinson for their help and interest in my project. Thanks especially to my family and friends for their support and for helping to make my time here an enjoyable one.  For  my family  Chapter One Introduction  1  • 2 1.1  Racemases and Epimerases. One of the main characteristics of biopolymers such as peptides or oligosaccharides, is  the need to possess a correct three-dimensional arrangement for optimal biological function. This requirement of specific spatial distribution is ultimately regulated by the stereogenic centers of the building blocks, mainly amino acids and sugars, which constitute these molecules. Thus, by changing the configuration at one or more of the stereogenic centers in a building block, it is possible to modify the biological properties of a biopolymer. It is not surprising therefore that certain biological processes could require a biopolymer that contains one or more building blocks that differ only in the configuration of one or more of its stereogenic centers from more ubiquitous precursors. Organisms have developed enzymes that invert a stereogenic center by means of a simple single-substrate single-product reaction in order to produce these unusual biosynthetic precursors from more available substrates.  This is a more energetically efficient way to  increase the repertoire of biosynthetic precursors than to synthesize the unusual precursor from more basic components through a long biosynthetic pathway. The enzymes responsible for carrying out such inversion processes are known as racemases and epimerases (Adams, 1976; Glaser, 1972). Racemases act on substrates with only one stereogenic center (enantiomers) while epimerases act on substrates containing more than one stereogenic center (diastereomers).  The different racemases and epimerases are  distributed among a wide variety of biological organisms; some racemases (e.g. lactate racemase, mandelate racemase) are present only in bacteria while others (e. g. UDP-D-glucose 4-epimerase) are found in bacteria, yeast and mammals. Many studies of the mechanism of action of these enzymes on amino acids, peptides, a-hydroxyacids and carbohydrates have been done in detail and comprehensive reviews can be found in the works of Glaser (1972), Adams (1976) and more recently by Tanner and Kenyon (1998).  3 Some racemases and epimerases, particularly proline racemase (Rudnick and Abeles, 1975) and mandelate racemase (Kenyon et al'., 1995), have been the subject of thorough studies aimed at further understanding the means by which an enzyme active site is capable of accommodating opposite stereogenic centers and stabilizing reactive intermediates.  It is  interesting to note that in most enzymatic racemization and epimerization processes, the stereogenic center that undergoes inversion bears a hydrogen that is removed during the transformation. These types of enzymes follow either a deprotonation/reprotonation mechanism or an hydride transfer process using a cofactor, such as NAD . +  The second type of  transformation is best exemplified by UDP-galactose 4-epimerase, as described later on in this chapter. In the case of a deprotonation/reprotonation mechanism, the enzyme simply deprotonates the substrate on one face and then reprotonates the resulting intermediate on the opposite face of the molecule. The enzymes that act in this fashion were further classified considering the number of enzymatic bases used to bring about the 1,1 proton shift (Rose, 1966); the terms 'one-base mechanism' and 'two-base mechanism' were coined to described two possible modes of action (Figure 1.1). A one-base mechanism-type utilizes a single enzymatic residue at the active site that is responsible for both, the removal of a proton from one face of the substrate to form an anionic intermediate and the subsequent reprotonation on the opposite face. This process can occur with or without proton exchange with the solvent depending on the lifetime of the carbanion intermediate and the accessibility of the enzymatic base to exchangeable protons. In the two-base mechanism, the substrate is deprotonated by an enzymatic base and epimerization/racemization occurs by reprotonation on the opposite face by the conjugate acid of a second enzymatic base. derived protons.  In this process all of the product molecules contain solvent  4  Figure 1.1  S c h e m a t i c representation  o f (a) a one-base, (b) a two-base  mechanism for  e n z y m a t i c i n v e r s i o n o f stereogenic centers.  1.2  Enzymatic Epimerization of Carbohydrates T h e generation  o f n e w biosynthetic precursors b y i n v e r s i o n o f a n y o f the several  stereogenic centers present i n a sugar m o l e c u l e is a c o m m o n e n z y m a t i c process.  A variety o f  e n z y m e s have e v o l v e d that catalyze stereochemical inversions o n s i m p l e free sugars as w e l l as o n derivatized sugars s u c h as p h o s p h o r y l a t e d sugars o r sugar n u c l e o t i d e s . have b e e n g r o u p e d into t w o categories: those that u s e N A D  +  1  These enzymes  as a c o f a c t o r to facilitate the  reaction and those that are cofactor-independent.  Nucleotides are phosphoric acid esters of nucleosides. A nucleoside is constituted of a nitrogenous base (purines or pyrimidines) linked to a furanose sugar (D-ribose or 2-deoxy-D-ribose) through a (3-glycosidic linkage. 1  5  Historically, the first carbohydrate epimerase was discovered by Leloir (Leloir, 1951) and is a classical example of a cofactor-dependent enzyme. This enzyme was first known as galactowaldenase, in reference to the Walden inversion (which describes the inversion process that occurs in a typical SN2-type reaction).  The reaction catalyzed is the reversible  epimerization of the C-4" center of UDP-D-glucose (UDP-Glc) to afford UDP-D-galactose (UDP-Gal). This activated form of D-galactose (UDP-Gal) is an important building block in 2  the synthesis of a number of polysaccharides, glycoproteins, and glycolipids that are present in a variety of microorganisms and in mammals. The current working hypothesis for the mechanism used by this enzyme4s portrayed in Figure 1.2.  The reversible epimerization of UDP-Gal to UDP-Glc, is initiated by the binding  of UDP-Gal to the epimerase. This binding induces a conformational change in the enzyme (E -> E*) that enhances the reactivity of NAD as a hydride acceptor. The intermediate UDP-4+  keto-glucopyranoside is then produced at the active site by a transfer of a hydride from C-4" of UDP-Gal to the N A D cofactor. The keto-intermediate then undergoes a spatial reorientation, +  mainly by rotation around the bond linking the anomeric oxygen to the p-phosphorus atom of the UDP moiety. This reorientation allows the opposite face of the 4-keto group to approach 3  NADH and accept a hydride from the reduced nucleotide cofactor forming UDP-Glc and regenerating the cofactor. This mechanism was formulated based on a substantial body of kinetic, stereochemical and binding information (Frey, 1987) and has recently received structural support based on Xray cfystallographic analysis (Thoden et al., 1997).  The notation C-X" or H-X" refers to atoms in the monosaccharide linked to the nucleotide. Atoms in the sugar component of the nucleoside (e.g. ribose) are designated as C-X' or H-X', while atoms in the nitrogenous base moiety of the nucleoside receive no special notation. The phosphorus atoms in a nucleotide sugar are designated a,p, and y, the a phosphorus being attached to the C5' of the nucleoside. UDP contains only a and (3 phosphates. 2  3  6  OH OH  Figure 1.2. Proposed mechanism of UDP-galactose 4-epimerase  Alternatively, stereogenic centers may be inverted by NAD -dependent epimerases +  through the transient oxidation of a hydroxyl adjacent to the site of epimerization.  The  resulting carbonyl group increases the acidity of the a-hydrogens, thus 'activating' the stereogenic center and allowing the occurrence of a 1,1 proton shift. A n example of this type of process occurs in an enzyme found in the green alga Chlorella  pyrenoidosa  (Barber and  Hebda, 1982) which catalyzes the epimerization at two stereogenic centers (Figure 1.3). C. pyrenoidosa  epimerase uses N A D  +  The  to oxidize the hydroxyl group at C-4" of GDP-D-  mannose and proceeds with the epimerization at the 3 and 5 positions of the resulting ketosugar intermediate.  The product, GDP-L-galactose, is formed by the inversion of the  stereogenic centers that are adjacent to the keto group (C-3" and C-5") via enol or ene-diol intermediates followed by a reduction of the transient ketone.  7  GDP-L-Gal  Figure 1.3 Proposed mechanism for the enzymatic 3,5-epimerization of GDP-D-mannose  The majority of cofactor-independent epimerases invert stereogenic centers that are adjacent to a carbonyl or a carboxylate functionality.  As mentioned earlier, the usual  mechanism employed is a deprotonation/reprotonation of the stereogenic center by a base in the active site of the enzyme that exploits the increased acidity of the proton due to the presence of a vicinal 'activating' group. An enzyme in this category that has been thoroughly studied is Dribulose 5-phosphate 3-epimerase. This enzyme interconverts D-ribulose 5-phosphate and Dxylulose 5-phosphate (Figure 1.4). The epimerization occurs by an initial deprotonation at C-3 to form an ene-diol intermediate, which is then reprotonated on the opposite face to form the epimeric product (Adams, 1976; Walsh, 1979a).  1  8  CH OH  S~^W—  /  B" i  CH OH  C — O H II II C — O H  C=0  —  H — C — O H 1  2  1  1  H — C — O H |  2  1  2  2  3  B 1  CH OP0 -  CH OP0 "  CH OP0 ' " 2  H O — C — H ^ * " " ^  H — C — O H ;  1  CH OH 2  2  1 c=o 1 1C — O H  2  3  3  ENZ  ENZ  D-Ribulose 5-phosphate  ene-diol intermediate  D-xylulose 5-phosphate  Figure 1.4 Proposed mechanism of epimerization of D-ribulose-5-phosphate 3-epimerase  There are a few cofactor-independent epimerases that are able to catalyze inversion at seemingly 'unactivated' stereogenic centers, and these must employ a unique mechanism. These include UDP-iV-acetylglucosamine 2-epimerase and L-ribulose-5-phosphate 4-epimerase. The former is the subject of this thesis. The latter is also under study in the laboratories of Dr. Martin Tanner; while its mechanism is not yet fully understood, some advances on the understanding of its mode of action have been made recently (Johnson and Tanner, 1998). L-Ribulose-5-phosphate 4-epimerase  interconverts  L-ribulose-5-phosphate  and D-  xylulose-5-phosphate and is involved in the bacterial metabolism of arabinose. Early studies on this enzyme showed that the epimerization mechanism is quite unique due to the need to catalyze the inversion of the 'unactivated' stereogenic center (C-4). It was found that neither hydrogen nor  oxygen isotopes were exchanged with solvent  during  the  enzymatic  epimerization (McDonough and Wood, 1961) and that an N A D cofactor was not required for +  the reaction (Deupree and Wood, 1970).  In addition, the absence of a significant kinetic  isotope effect when [4- H]-D-xylulose-5-phosphate was used as a substrate (McDonough and 3  Wood, 1961), and the fact that the enzyme had a divalent metal ion requirement (Deupree and Wood, 1972) pointed towards a mechanism that involves an unusual C - C bond cleavage  9 (Figure 1.5).  The current working hypothesis for the mechanism of this enzyme involves an  initial retro-aldol reaction that occurs following the deprotonation of the C-4 hydroxyl group. Cleavage of the C-3 - C-4 bond generates glycoaldehyde phosphate and a metal-bound enolate of dihydroxyacetone. The rotation of the aldehyde carbonyl to expose the opposite face, followed by the aldol addition of the enolate, produces the epimeric sugar (Johnson and Tanner, 1998).  HOH C 1 C = 0 2  ,M  2 +  HO  H  Co-C-H ^  CH OP0 2  2 3  2  X  |-|  2  rotation  X  H C T ^ H  M  2  CH OH  +  2  C = 0  —  HO—C—H  |  H—C—OH  V  "  /  HOH C^.O^,„  H O H z C ^ / C K , , , , ^  2  1  CH OP0 2  2  2 3  "  "0 POH C 3  2  "0 POH C 3  2  ENZ  D-Xylulose-5-phosphate  L-Ribulose-5-phosphate  Figure 1.5 epimerase  A proposed mechanism for the reaction catalyzed by L-ribulose 5-phosphate 4-  1.3  UDP-Af-Acetylgrucosamine-2 Epimerase  1.3.1  Mammalian Enzyme The  mammalian  enzyme  UDP-7V-acetylglucosamine 2-epimerase  catalyzes  the  conversion of UDP-GlcNAc (1) into ManNAc and UDP. This is not strictly an example of a sugar nucleotide epimerase due to the irreversible formation of a non-nucleotide bound sugar (ManNAc). An alternate name for this enzyme, 'UDP-JV-acetylglucosamine-2-epimerase (hydrolyzing)', was coined by Salo (1976) after the discovery of a bacterial enzyme that epimerizes the nucleotide sugar in a reversible fashion (vide infra). classification of the mammalian enzyme would consider it to be a lyase.  A more rigorous  10 The free ManNAc formed by the mammalian epimerase is used in the biosynthetic pathway that ultimately produces sialic acids. Sialic acids are a family of compounds derived from neuraminic acid and are widely distributed in vertebrate tissue and in bacteria.  They  occur as components of complex mucous substances such as mucolipids and mucoproteins, and are also found in some oligosaccharides present in milk. Studies on the mammalian epimerase have been limited due to its instability; nevertheless, mechanisms have been proposed to suggest the means by which this enzyme operates. These proposals will be discussed briefly due to their relevance to the mechanistic pathway followed by the bacterial epimerase, which is the subject of the present work.'This enzyme was designated as an epimerase based on the first proposed mechanism (Spivak and Roseman, 1966) that considered the reversible formation of UDP-ManNAc (2) as the first step of the transformation. This step was followed by an irreversible hydrolysis of the nucleotide sugar to generate free ManNAc and UDP (Figure 1.6).  OH O H  OH O H  1  2  UDP-GlcNAc  UDP-ManNAc  + OH O H  UDP  ManNAc  Figure 1.6 Sequence of reactions proposed by Spivak and Roseman (1966) for the enzymatic epimerization/hydrolysis of UDP-GlcNAc by the mammalian epimerase.  , 11 This mechanism was inconsistent with previous experimental results that found that the C-2 tritiated product, [2- H] ManNAc, was formed from UDP-GlcNAc when the reaction was 3  carried out in tritiated water, but no tritiated substrate could be recovered from an incomplete reaction (Glaser, 1960).  This last experimental result suggested the absence of reversible  proton exchange between the solvent and the substrate that in turn implies that the proton exchange at C-2 must occur with retention of configuration during the irreversible hydrolysis of the intermediate (UDP-ManNAc). The unlikehood of such an exchange makes this mechanism unsound. Later studies on this enzyme (Salo and Fletcher, 1970) involved the use of the putative intermediate, UDP-ManNAc, as an alternate substrate and the use of tritiated water as the solvent. This lead to the proposal of a different mechanism that considered the following items: (1)  The putative intermediate, UDP-ManNAc, was not detected during the enzymatic transformation of UDP-GlcNAc to UDP and ManNAc.  (2)  The tritiated product, [2- H]-ManNAc, was isolated when either UDP-GlcNAc or UDP3  ManNAc was incubated with the enzyme. (3)  The tritiated/non-tritiated  product ratio ([2- H]-ManNAc/ManNAc) obtained when  UDP-GlcNAc was used as the substrate was higher than the ratio obtained when using UDP-ManNAc. (4)  UDP-GlcNAc was not detected during the enzymatic transformation using UDPManNAc as the substrate.  (5)  No evidence of the formation of tritiated substrate was found when the enzyme was incubated with either UDP-GlcNAc or UDP-ManNAc for a brief period of time. Salo and Fletcher ruled out the previously proposed mechanism based on items 4 and 5  which contradict the formation of UDP-ManNAc in a first reversible step, and on item 3, which is inconsistent with the labelling of ManNAc during the second irreversible step.  They  12 proposed a mechanism (Figure 1.7) that considers the irreversible formation of a glycosylenzyme bound intermediate as a first step followed by the formation of a keto-intermediate. Epimerization then occurs via an enol or enolate sugar nucleotide.  UDP-GlcNAc  glycosyl-enzyme (GlcNAc)  glycosyl-enzyme (ManNAc)  Figure 1.7 Mechanism proposed by Salo and Fletcher (1970) for the mammalian UDPGlcNAc 2-epimerase. E represents the enzyme. A wavy bond represents a mixture of a and p anomers  This mechanism could account for the lack of tritium incorporation into the substrates (item 5) and the failure to detect UDP-GlcNAc when UDP-ManNAc was used as substrate (item 4). In addition, UDP-ManNAc might not be an intermediate but rather an alternative substrate that forms the glycosyl-enzyme intermediate and is hydrolyzed in an irreversible reaction with the enzyme (reactions 7 and 6 in Figure 1.7). This is consistent with the inability to detect labeled UDP-ManNAc (item 1) and the lack of tritium incorporation in the partial reaction (item 5). This mechanism might also account for the lower [2- H]-ManNAc/ManNAc  13 ratio, derived from the enzymatic reaction on UDP-ManNAc in tritiated water (items 2 and 3), by noting that the formation of the glycosyl-enzyme (ManNAc) can lead to either a direct hydrolysis with no incorporation of tritium (reaction 6), or to the introduction of the label at C2 by equilibration through reactions 5, 4 and 3. The core of Salo's  mechanism lies in the formation of a keto-intermediate which  requires that the enzyme either contain or utilize a cofactor such as N A D or N A D P . The +  +  occurrence of keto-intermediates is well documented for the epimerization of sugar nucleotides (Glaser, 1972; Tanner, 1998) and, although no enhancement of activity was found upon addition of N A D or N A D P , this could be justified by proposing the presence of a tightly +  +  bound cofactor. An alternative mechanism was proposed by Sommer and Ellis (1972) in which the enzyme does not have any NAD -type cofactor requirement (Figure 1.8). +  The formation of  ManNAc from UDP-GlcNAc is explained by the initial formation of a 2-acetamidoglucal intermediate (3) by a favorable aw/i-elimination of UDP from the nucleotide. This is followed by an irreversible stereospecific hydrolysis of the intermediate to afford ManNAc. This mechanism is also consistent with the experimental observations made earlier. Thus, the formation of  [2- H] ManNAc and the absence of tritium incorporation in the  recovered substrate when the reaction is carried out in tritiated water can be explained by the irreversible stereospecific transfer of an exchangeable proton on the enzyme active site during the addition of water to the 2-acetamidoglucal intermediate. In addition, the formation of [23  H] ManNAc from UDP-ManNAc and the lower [2- FL] ManNAc/ManNAc ratio when UDP3  ManNAc instead of UDP-GlcNAc was used as the substrate can be accounted for by an unfavorable 5y«-elimination of UDP from UDP-ManNAc to give 2-acetamidoglucal.  14  ENZ  H HO' HO' HO-  + UDP AcNH V  3 UDP-GlcNAc  OH Enz ManNAc  2-Acetamidoglucal  Figure 1.8 Mechanism proposed by Sommar and Ellis (1972) for the mammalian enzyme UDP-GlcNAc 2-epimerase.  Evidence to support the proposed mechanism was provided by experiments in which the enzyme was incubated with the putative intermediate 2-acetamidoglucal to yield a compound that could not be properly identified but showed chromatographic and electrophoretic properties similar to those of ManNAc. This compound was not observed when the glucal was incubated with buffer alone. Additionally, kinetic studies of product inhibition by UDP were consistent with an ordered mechanism where UDP was released first followed by the irreversible formation of ManNAc.  1.3.2  Bacterial enzyme The bacterial enzyme UDP-iV-acetylglucosamine 2-epimerase catalyzes the reversible  epimerization of UDP-GlcNAc (1) to UDP-ManNAc (2) (Figure 1.9). The biological role of this enzyme is to provide the bacterium with a source of activated ManNAc residues, in the form of a nucleotide sugar, to be used as a building block in the biosynthesis of the bacterial cell wall polysaccharides.  15  Figure 1.9 Reaction catalyzed by the bacterial enzyme UDP-GlcNAc 2-epimerase.  Several studies have shown that ManNAc residues are important in both Gram-positive and Gram-negative bacteria. In Gram-positive bacteria such as Staphylococcus aureus H. and Bacillus subtilis, a ManNAc residue is a component of the "linkage unit" that serves to attach teichoic acids to the peptidoglycan (Harrington and Baddiley, 1985). Teichoic acids are major cell wall components of these types of bacteria. They are present in cell walls in amounts comparable to or greater than the peptidoglycan, the basic structural component of the cell wall. The biological function of teichoic acids is not very well understood but it is believed that they play a role in the control of cation concentration at the surface of the cell (Baddiley, 1989). Gram-negative bacteria such as Escherichia coli, have A^-acetylmannosaminuronic acid (ManNAcUA) residues in the carbohydrate portion of the enterobacterial common antigen (ECA), a glycolipid produced by all members of the Enterobacteriaceae (Meier-Dieter et al., 1992). The bacteria produce these residues from UDP-GlcNAc by using UDP-GlcNAc 2epimerase and UDP-ManNAc 6-dehydrogenase, an N A D  +  cofactor-dependent enzyme that  catalyzes the two-fold oxidation of UDP-ManNAc to form the corresponding mannosuronic acid, UDP-ManNAcUA (4) (Figure 1.10). A brief description of how these enzymes were discovered and of previous mechanistic studies on the epimerase follows.  16  UDP-ManNAcllA  Figure 1.10 Enzymatic formation of UDP-ManNAcUA. A = UDP-GlcNAc 2-epimerase; B = UDP-ManNAc dehydrogenase  1.4  Previous studies with the bacterial UDP-GlcNAc 2-epimerase Prior to 1970, it was known that A^-acetylmannosaminuronic (ManNAcUA) was one of  the components of the polysaccharide that can be extracted from the cell walls of a bacterium (Micrococcus lysodeikticus). The presence of this monosaccharide suggested the existence of a nucleotide sugar precursor as the activated form of ManNAc used in the biosynthesis of these polysaccharides. Attempts to isolate this precursor, however gave ambiguous results (Rosental and Sharon, 1964; Biely and Jeanloz, 1969). In 1972 it was demonstrated (Anderson et al., 1972) that the sugar precursor of the Af-acetyl-D-mannosaminuronic acid residues was in fact its activated nucleotide form, UDP N-acetyl-D-mannosaminuronic acid (UDP-ManNAcUA). This led to the difficulty of explaining how this unusual nucleotide sugar was made by the bacterium.  It was suggested that the ubiquitous sugar nucleotide, UDP-GlcNAc, was the  precursor and was transformed to UDP-ManNAcUA either by oxidation at C-6 of the sugar moiety followed by epimerization at C-2, or by the same processes occurring in the reverse order. In 1974 the formation of UDP-ManNAcUA from UDP-GlcNAc using a soluble enzyme „ extract from Escherichia coli and radiolabeled substrates was reported (Ichihara et al., 1974), but the exact pathway for this transformation remained unknown.  17 Later studies done by Japanese researchers (Kawamura et al., 1975) unveiled the presence of an 'enzyme system' that was responsible for the conversion of UDP-GlcNAc to UDP-ManNAcUA in Escherichia coli. They were able to separate two components of the enzyme system and proved that the biosynthetic course to make UDP-ManNAcUA proceeded in two steps: the reversible epimerization of UDP-GlcNAc at C-2 brought about by one of the components, followed by the NAD -dependent oxidation of the formed UDP-ManNAc at C-6 +  effected by the other component. The first component was dubbed as a new UDP-GlcNAc 2epimerase, differing from its mammalian counterpart in that it did not cleave off the UDP moiety to give a free ManNAc during the epimerization. After Kawamura's discovery of the bacterial epimerase, Salo used a crude extract of the enzyme to demonstrate the incorporation of tritium from tritium-enriched water into the C-2" position of the nucleotide sugar substrate (Salo, 1976). This result revealed that a proton transfer mechanism was followed, and provided evidence against alternative pathways that the enzyme might follow to bring about the epimerization, in particular the involvement of a hydride transfer or the homolytic cleavage at C-2 by some sort of radical process. Salo (1976) initially proposed a mechanism analogous to that previously presented for the mammalian 'hydrolyzing' epimerase (Salo and Fletcher, 1970). This proposed mechanism involves the formation of a 3"-keto sugar nucleotide intermediate followed by epimerization through a enpl/enolate (Figure 1.11). This is one among other possible mechanisms, which for the purposes of this thesis will be referred as 'Path A ' Further work done by Kawamura and coworkers on the epimerase isolated from Bacillus cereus (Kawamura et al., 1978) and from Escherichia coli (Kawamura et ai,  1979)  described several features of the enzyme but did not lead to the proposal of a mechanism different from the one proposed by Salo. A brief description of the characteristics of this enzyme revealed by these studies follows.  18  Figure 1.11 epimerase.  1.4.1  Path A : Mechanism proposed by Salo (1976) for the bacterial UDP-GlcNAc 2  Distribution of the bacterial UDP-GlcNAc 2-epimerase The enzyme has been found in a variety of Gram-positive and Gram-negative bacteria  (Kawamura et al., 1978). Some of the microorganisms that have shown epimerase activity are: Bacillus cereus, Bacillus subtilis, Bacillus megaterium, Escherichia  coli, Staphylococcus  aureus H and Micrococcus lysodeikticus.  1.4.2  Isolation and purification of the enzyme The first reported isolation of the epimerase was from Escherichia coli 014 K7 H"  (Kawamura et al., 1975). The isolation protocol consisted of the treatment with protamine sulfate on a 20 OOOxg supernatant fraction from cultivated cells followed by precipitation with  19 ammonium sulfate and partition by chromatography on hydroxylapatite (calcium phosphate). This procedure yielded two components that were found to catalyze the transformation of UDP-GlcNAc to UDP-ManNAcUA. Further isolations of the enzyme were performed on extracts from B. cereus (Kawamura et al., 1978) and again from Escherichia coli 014 K7 H" (Kawamura et al., 1979).  The  improved procedures included the use of hydroxylapatite, anion exchange (DEAE-cellulose), size exclusion (Sephadex G-200) and affinity chromatography (ADP and UDP-hexanolamine agarose).  By a combination of these techniques it was possible to separate the fraction  corresponding to the epimerase from the one corresponding to the dehydrogenase and achieve 10 000 and 700 fold purification factors for these enzymes respectively.  1.4.3  Effects of Enzyme concentration, pH, cations, NAD* and nucleotides A linear relationship between the amount of enzyme and the rate of epimerization of the  substrate was observed. The optimal pH was found to be between 7.5 and 8 with 50 mM TrisHC1 or 50 m M potassium phosphate buffers for the enzyme isolated from B. cereus and between 7 and 9 for the Escherichia coli enzyme. The epimerization reaction was not influenced by the presence of the following substances: N A D , N A D P , N A D H and N A D P H at a concentration of 10 mM; 2+  +  mercaptoethanol, dithiothreitol, K , N a , N H +  +  + 4  at 50 mM; E D T A and M g  2 +  at 5 mM; A T P ,  UTP, and UMP at 1 mM. The  epimerase  isolated from Escherichia  coli was  strongly  inhibited  by  p-  chloromercuribenzoate, a reagent specific for thiol groups, up to 90% inactivation with an inhibitor concentration of only 25 u.M.  The inhibition was reversible, as evidenced by the  complete restoration of enzyme activity upon incubation with either dithiothreitol or 2mercaptoethanol.  20 UDP was also found to be a reversible inhibitor, with a Kj of approximately 1 mM, that does not change the Hill coefficient (see Appendix 1) or the apparent K  m  value for the  epimerase.  1.4.4  Requirement of UDP-GlcNAc in the epimerization of UDP-ManNAc When the epimerization reaction was tested in the reverse direction, using UDP-  ManNAc as substrate, the reaction did not proceed at an appreciable rate using normal concentrations of epimerase.  In order to perform these experiments, UDP-ManNAc was  isolated out of the enzymatic epimerization of UDP-GlcNAc and purified by paper chromatography. However, the purified UDP-ManNAc was not entirely free of UDP-GlcNAc. Thus, upon incubation of purified UDP-ManNAc preparations with higher concentration of enzyme, the epimerization reaction took place after a time lag and proceeded in an accelerated fashion. The length of the time lag was found to be dependent on the purity of the substrate preparations suggesting that the enzyme needed UDP-GlcNAc as an activator/modulator in order to perform the reverse reaction. This is a known phenomenon in enzymology and is termed allosteric activation (see Appendix 1). In fact, when a small amount of UDP-GlcNAc was added to the incubation mixture for the reverse reaction, rapid conversion of UDPManNAc to UDP-GlcNAc was observed, consistent with the expected equilibrium position for the epimerase reaction (vide infra).  It was found that the concentration of UDP-GlcNAc  required to get half maximum activation was 0.5 mM (Bacillus cereus) or 0.6 mM (Escherichia coli). In addition, a plot of the rate of epimerization of UDP-ManNAc against UDP-ManNAc concentration in the presence of UDP-GlcNAc as an activator gave a typical hyperbolic curve (normal Michaelis-Menten kinetics) with an apparent K value for UDP-ManNAc of 0.22 m M m  when using the epimerase isolated from Bacillus cereus (with 0.5 m M activator) or 0.5 m M (with 1 m M activator) with the epimerase from Escherichia coli.  21 This evidence suggests that the enzyme is allosterically activated by its own substrate, UDP-GlcNAc but not by the substrate for the reverse reaction, UDP-ManNAc. Several compounds such as D-glucose, GlcNAc, GlcNAc 1-phosphate, UDP-Glc, UDPGalNAc, UDP-ManNAc, A T P , UTP, UDP, UMP, N A D , N A D H , inorganic phosphate or +  diphosphate, 2-mercaptoethanol, dithiothreitol, MgCb, E D T A , etc. were tested as alternative activators at a concentration of 1 m M with unsuccessful results.  1.4.5  Effect of Substrate concentration As expected, a plot of the rate of epimerization of UDP-GlcNAc against the  concentration of this substrate gave a sigmodial curve which is characteristic of an allosteric enzyme and is described by the Hill equation (see Appendix 1). The Hill coefficient («) and apparent K  m  (K') obtained for the epimerase isolated from Bacillus cereus (n = 1.8, K' = 1.1  mM) differ somewhat from those reported for the Escherichia coli enzyme (n = 2.0, K' = 0.63 mM). These results strongly suggest the presence of a modulator-binding site distinct from the catalytic site that seems to be strictly specific for UDP-GlcNAc.  1.4.6  Substrate Specificity In addition to the high specificity of the modulator binding site, the epimerase seems to  be extremely specific for its substrate.  When other potential substrates such as GlcNAc 1-  phosphate, GlcNAc 6-phosphate, UDP-GalNAc, and UDP-Glc were incubated with the epimerase, conversion of the alternative substrates was not detected either in the presence or absence of the activator UDP-GlcNAc.  1.4.7  Equilibrium Constant  22 Using either sugar nucleotide as starting substrate the epimerization reached the same equilibrium ratio of UDP-ManNAc to UDP-GlcNAc of 0.11 (Bacillus cereus enzyme). Other studies on the Escherichia coli epimerase reported a similar equilibrium ratio of UDP-ManNAc to UDP-GlcNAc of 1:9 (Kawamura et al, 1975) and 1: 10 (Salo, 1976 and Kawamura et al, 1982).  1.4.8  Enzyme assay. In order to follow the progress of the reaction for the kinetics experiments, enzymatic  assays for the epimerase were developed.  Most of the early assays used radioactivity to  quantify the isolated products of the reaction after incubation of a UDP-[ C] GlcNAc with the 14  enzyme (Kawamura et al, 1975).  In a first assay the product of the epimerization UDP-  [ C]ManNAc was oxidized, using an extract containing UDP-ManNAc dehydrogenase, to 14  UDP-[ C]-ManNAcUA. l4  The sugar moiety of the latter nucleotide was subsequently  transferred to a polysaccharide using an enzyme extract isolated from Escherichia coli (Ichihara et al,  1974). In this polysaccharide synthesis UDP-glucose is required as a co-substrate and  UDP-ManNAcUA cannot be replaced by UDP-GlcNAcUA, UDP-GlcNAc or UDP-ManNAc. This procedure allowed following of the epimerization with UDP-GlcNAc 2-epimerase by isolation of the formed polysaccharide by paper chromatography and radioactive counting (Kawamura, 1975, 1978).  A second approach used acid hydrolysis of the products of  epimerization followed again by paper chromatography and radioactive counting (Kawamura et al,  1979). A n assay based on the enzymatic incorporation of tritium from H H O into the 3  substrate was developed by Salo (Salo, 1976). Once again purification by anion exchange chromatography and radioactive counting of the column fractions was necessary. All the aforementioned assays took advantage of the easily detectable transformation of radioactive labelled substrates. The drawbacks of these assays were the need to prepare and  23 purify the labelled substrates as well as the fact that they are all stopped assays. Thus, after incubation of the substrate with the enzyme at moderate temperatures, the reaction was normally 'quenched' by boiling at 100 °C for a short period of time and purified by chromatography. A more convenient assay was developed by Kawamura et al. (1979) by using the enzyme UDP-ManNAc dehydrogenase as a coupling enzyme with the epimerase.  The  dehydrogenase transformed any UDP-ManNAc formed by the epimerase into UDPManNAcUA by using 2 moles of N A D per mol of nucleotide oxidized and producing N A D H +  along with the mannosaminouronic acid product (Figure 1.10). The fact that the formation of N A D H can be measured spectrophotometrically (absorbance at 340 nm) allowed the development of assays that avoided the use of radiolabeled substrates. This was still a stopped assay as it required the 'quenching' of the enzymatic reaction prior to the spectroscopy measurement.  1.5  Proposed mechanistic pathways followed by the bacterial UDP-GlcNAc 2epimerase As it has been already mentioned, an interesting feature of the epimerase is the need to  invert an stereogenic center (C-2") that is relatively non-acidic. The proposed mechanism presented by Salo (Path A , Figure 1.11), involves a C-3"-keto intermediate that is formed with the aid of a cofactor (i.e. N A D ) and is needed to increase the acidity of the C-2" proton +  (permitting epimerization via deprotonation/reprotonation process). Due to the experimental outcome of tritium incorporation at C-2", any rational mechanism suggested for the enzyme should ultimately  involve proton transfer steps.  Alternative pathways that employ a hydride transfer or hydrogen atom abstractions at the site of  24 A first proposed alternate mechanistic pathway that will be called Path B in this thesis, involves an elimination reaction in which the proton at C-2" is removed by an enzymatic base and the bond between C - l " and the p-phosphate oxygen of UDP (anomeric oxygen) is cleaved. This will generate a 2-acetamidoglucal intermediate (3) that can be converted to product by the re-addition of UDP with reprotonation on the opposite face of the sugar (Figure 1.12).  Enz  HO'  H  :B  j  HO' HO HO  3  HO' HO  AcHN 0 ^ / 0 -  1  A  O o  ,o~ c  /A  o o_  HO'  O. .  HO HO O.  AcNH  (1  AcNH  o*o o o_ A  / O . / O - u  O Q_ o 0_  Enz  Figure 1.12 Alternate mechanistic pathway (Path B) for the enzymatic catalysis of the UDPGlcNAc 2-epimerase  Another possible mechanism (Path C) involves an enzymatic nucleophilic attack at the P-phosphorus atom of the UDP group with formation of an open chain sugar and a UDPphosphorylated enzyme intermediate. The free sugar can suffer the epimerization at C-2 and proceed towards the product by subsequent ring closure and reattachment to the UDP moiety (Figure 1.13).  25  Figure 1.13 A proposed mechanistic pathway (Path C) for the enzymatic catalysis of the epimerase involving nucleophilic attack at the (3-phosphorus.  One other possible mechanism that the epimerase might follow can be envisioned as an "hybrid" of Path B and the mechanism proposed by Salo (Path A). In this pathway (Path D, Figure 1.14) the enzyme could form the same C-3" keto-intermediate as the one proposed by Salo. This would be followed by the deprotonation at C-2" and the cleavage of the anomeric oxygen to yield a 3-keto-2-acetamidoglucal intermediate 5. This intermediate could evolve to the epimerized product by conjugate addition of the UDP moiety combined with protonation on the opposite stereochemical sense and a subsequent reduction of the ketone at C-3".  26  Figure 1.14 A proposed 'hybrid' mechanistic pathway (Path D) for the enzymatic catalysis of the epimerase.  1.6 Aims of this Thesis The goal of this thesis is to provide experimental evidence that could help to elucidate the mechanism by which the bacterial enzyme UDP-GlcNAc 2-epimerase might operate.  In  order to attain this goal, the chemical synthesis of alternative substrates followed by the analysis of the enzymatic epimerization (or the absence of it) of these new substrates will be used to give an insight into the mechanistic pathway employed by the enzyme.  A more  detailed discussion of the possible mechanism of the epimerase as well as a proposal for further studies will be addressed in the final chapter.  Chapter Two  Initial Studies with Cloned UDP-GlcNAc 2-Epimerase  Z7  28 2.1  Introduction In Chapter One, some preliminary studies done by Japanese researchers (Kawamura et  al, 1979), on the bacterial enzyme UDP-GlcNAc 2-epimerase were described. The epimerase isolated from Escherichia coli was found to be a dimer with a subunit molecular weight of 38 kJJa, which required its substrate, UDP-GlcNAc, as activator for the epimerization. The need of this effector was reflected in the kinetics of UDP-GlcNAc epimerization as a function of substrate concentration which showed positive cooperativity (Appendix 1).  A mechanism  involving deprotonation/protonation at the epimerization center, C-2", was ultimately responsible for the epimerization (rather than anhydride transfer or a radical process). This was evidenced by the observation that solvent derived tritium was incorporated into the C-2" position when the epimerization reaction was performed in tritium enriched water (Salo, 1976). Pure preparations of the enzyme were, however, difficult to obtain since UDP-GlcNAc 2-epimerase represents only a very small fraction of the total protein cellular extract of Escherichia coli. In fact, the reported 10 000-fold purification procedure from 70 g of wet bacterial cells harvested from 25 L of cell culture yielded, after a multistep purification process, 0.24 mg of purified epimerase (Kawamura et al, 1979). Using this methodology, it would be too laborious to obtain the relatively large quantities of homogeneous enzyme needed for extensive mechanistic studies of the enzyme-catalyzed reaction.  In order to overcome this  problem, the cloning and overexpression of the gene encoding for UDP-GlcNAc 2-epimerase in Escherichia coli were undertaken as a separate project by Dr. Paul Morgan in our laboratory. We first wished to verify if the reported observations made on the endogenous enzyme could be reproduced with the recombinant enzyme. The evidence supporting a proton transfer mechanism was tested by performing the reaction in a deuterated medium following the incorporation of deuterium at C-2" by high resolution 'H N M R spectroscopy. The necessity of UDP-GlcNAc as an activator of the epimerization reaction was corroborated by following the  29 reverse epimerization, in the presence and absence of activator, using ion-paired reversed phased HPLC and chemically synthesized UDP-ManNAc. Finally, to test the similarity of the properties of the cloned and the endogenous epimerases, the kinetic constants for the cloned epimerase were evaluated. To do this, a more accurate (continuous) coupled assay was developed involving a second enzyme, UDP-ManNAc dehydrogenase. As large amounts of this second enzyme were necessary, the relevant gene needed to be cloned and overexpressed (work done by Dr. Paul Morgan).  2.2  Overexpression of cloned UDP-GlcNAc 2-epimerase in Escherichia coli The cloning of genes and overexpression of gene products are well known molecular  biology techniques which permit the production of large amounts of a protein by taking advantage of the genetic and metabolic machinery of a microorganism. In order to clone a gene, the genomic D N A of a particular organism is cut into small fragments at specific sites using specialized enzymes. The fragments containing the intact gene are identified and inserted into a 'vector', a double-stranded D N A molecule that acts as a vehicle for the foreign D N A fragment. Commonly used vectors are bacterial 'plasmids' which are small circular D N A molecules that can exist and replicate in a cell independent from the cell's main DNA. It is possible to introduce the vector with its added D N A fragment into an appropriate bacterial host, such as Escherichia coli in order to express the foreign protein.  This can result in the  production of the protein at levels much higher than typically seen for endogenous proteins. A detailed description of the theory of, and the techniques used in gene cloning and overexpression is beyond the scope of this thesis and can be found in more specialized literature (Dale, 1998) The gene encoding UDP-GlcNAc 2-epimerase in Escherichia coli is present in the gene cluster that controls the biosynthesis of the Enterobacterial Common Antigen (ECA) (Meier-  30 Dieter et al, 1992). The E C A is a cell-surface polysaccharide that is present in all members of the Enterobacteriaceae family (a. k. a. gut bacteria) which includes the genera Escherichia, Salmonella, Shigella and Yersinia. E C A is found only in this family and has the same structure in all the members of the family. Finding the exact location of the epimerase gene within the cluster was somewhat problematic, however. The confusion arose because an initial assignment (Daniels et al., 1992) of a gene as encoding the epimerase was incorrectly made. Later studies done in our group (Sala et al., 1996; Morgan et al, 1997) and others (Marolda and Valvano, 1995) showed that another gene within the same cluster was in fact the one encoding for the epimerase. A search of the literature showed that this latter gene had already been cloned and overexpressed in Escherichia coli and its gene product was a cytosolic protein of unknown function that is required for bacteriophage N4 adsorption (Kiino et al., 1993). The expression vector for the epimerase gene (pKI86) was obtained from Dr. Diane R. Kiino and Dr. Lucia B. RothmanDenes. The pKI86 plasmid was transformed into CaC^-competent Escherichia coli cells and the resulting cell culture was allowed to grow to adequate levels before overexpression of the enzyme was induced by adding isopropyl-l-thio-p-D-galactopyranoside (IPTG). As described in previous work (Kiino et al,  1993, Sala et al. 1996), this procedure  gave high levels of protein expression, which was evidenced by a distinct band of the appropriate molecular weight on an SDS-polyacrylamide gel of the crude cell lysate. After purification the expressed protein showed a molecular weight of 42 250 ± 4 Da (obtained by electrospray mass spectroscopy) which was identical, within experimental error, to the one calculated from its gene sequence (42 246 Da). The epimerase activity was initially assayed by following the enzymatic reaction by H !  N M R spectroscopy, as will be described in the next section. However, a more accurate assay  31 was necessary in order to quantitatively evaluate the enzyme activity.  This led to the  development of a coupled assay that will be described in more detail in section 2.5  2.3  Purification of cloned UDP-GlcNAc 2-epimerase In contrast with the bacterial epimerase isolated from natural sources, the overexpressed  epimerase could be purified by a far simpler procedure (Sala et al, 1996; Morgan et al, 1997). Many of the steps formerly necessary to obtain homogeneous protein (hydroxylapatite, size exclusion and affinity chromatography) were no longer needed. The crude cell lysate, prepared by subjecting the cells to passage through a French pressure cell and removing the ruptured cell wall debris by centrifugation, was partially purified by applying the protein mixture on a weak anion exchange column, and eluted by using a linear gradient of NaCI in pH 8.5 buffer. The protein containing fractions were evaluated for the presence of overexpressed epimerase by SDS-polyacrylamide gel electrophoresis. The fractions containing the epimerase were desalted by successive concentration and diluted with the appropriate buffer. The protein extract was further purified by applying it to a strong anion exchange HPLC column and eluting with a linear gradient of NaCI in pH 8.5 buffer. The purified protein was homogeneous when analyzed by SDS-PAGE.  2.4  Solvent Deuterium Isotope Incorporation One of the features of the bacterial epimerase reported by Salo (1976), was the ability of  the enzyme to incorporate tritium in the C-2" position of UDP-GlcNAc when the reaction took place in tritiated water. In order to verify this observation with the recombinant epimerase, the epimerization of UDP-GlcNAc in D 0 was followed by high resolution ' H N M R spectroscopy. 2  Since the epimerization reaction yields an equilibrating mixture of UDP-GlcNAc and UDPManNAc in an approximately 10:1 ratio (Kawamura et al, 1979), the spectra acquired during  32 the course of the reaction were expected to be composed mainly of signals corresponding to UDP-GlcNAc with small peaks corresponding to UDP-ManNAc. Furthermore, the H N M R !  spectra were expected to evolve over the course of the epimerization reaction, as a result of the incorporation of deuterium from the solvent into the C-2" position of UDP-GlcNAc. Specifically, the changes expected were the disappearance of the H-2" signal and modifications in the coupling pattern for H - l " and H-3" (Figure 2.1).  pyranose ring furanose ring Figure 2.1 Structure of UDP-GlcNAc showing the protons whose *H N M R signals undergo changes during the epimerization reaction at C-2"  Indeed, over the course of the enzymatic epimerization the signals corresponding to the protons of the pyranose moiety of the sugar nucleotide underwent several changes (Figure 2.2). A clear analysis of the changes of the 'H-NMR signals corresponding to the protons in the pyranose part of the molecule (3.47 to 3.97 ppm) was feasible using high resolution N M R spectroscopy. Signals due to the ribofuranose part of the molecule (4.12 to 4.35 ppm) did not interfere with this analysis.  33  Figure 2.2 Expansion of the ' H N M R ( D 2 O , 500 MHz) spectrum acquired during the enzymatic epimerization of UDP-GlcNAc. The insets show the signals arising from the anomeric proton of the pyranose moiety. See text for more details.  34 The signals arising from the H-2" of UDP-GlcNAc are observed as a ddd centered at 3.94 ppm. H-2" is coupled to H-1" with J= 3.3 Hz and to H-3" with J=  10.4 Hz. The third  coupling constant of 3.1 Hz arises due to a four-bond coupling between H-2" and the pphosphorous atom of the sugar nucleotide, as result of an in-plane "W" relationship for the H2 " - C - 2 " - C - l " - 0 - l " - P - p fragment. This type of long-range coupling is commonly observed between protons having a 1, 3 diequatorial relationship on a six-membered ring. The presence of this unusual V  i  H  ji  p  along with the value of the coupling constant (7.4 Hz) between H-1" and  the p-phosphorus nucleus (which suggests a dihedral angle of approximately 60°) has been used as evidence to support a rigid conformation of the sugar nucleotide in which the pyranose ring exists exclusively as a C i conformer (as depicted in Figure 2.1) with an antiperiplanar 4  relationship between the p-phosphorus and the C-2" atoms (Sarma and Lee, 1976; Figure 2.3).  Figure 2.3 Newman projections showing the antiperiplanar relationship between (a) C-2" and the p-phosphorous atoms, and (b) H-2" and the anomeric oxygen atom.  As can be observed in Figure 2.2,  the signals corresponding to H-2" decrease in  intensity as the reaction progresses and, after complete exchange with deuterium from the solvent, no signals in that region/are observed. Changes in the coupling patterns due to the exchange of H-2" are also detectable.  The signal arising from H-3" is located in the same  35 region of the spectrum as those arising from one of the diastereomeric protons at C-6" (3.73 to 3.79 ppm) and therefore, the coupling pattern is somewhat obscured. However, the changes that occur in this region can be attributed to the change in the coupling pattern of H-3" from a dd (J= 10.4; 9.7 Hz) to a d (J = 9.7 Hz). A more noticeable change in the coupling pattern is noticed in the signals corresponding to the anomeric proton H-1" (insets in Figure 2.2). This region of the spectrum (5.30-5.70 ppm) is normally free from other proton signals and is therefore useful for evaluation of deuterium incorporation even in N M R spectra taken at lower resolution. The H-1" signal, which initially appears as a doublet of doublets, due to the coupling of the anomeric proton with the p-phosphorus of the UDP moiety (J = 1A Hz) and to H-2" (J= 3.3 Hz), collapses to a doublet as H-2" as it is replaced by deuterium. In addition to the changes observed in the 'H-NMR spectrum of UDP-GlcNAc over the course of the epimerization, some additional signals were also detected. In particular, a small doublet centered at 5.41 ppm appears in the anomeric region of the spectrum. This signal can be attributed to the anomeric proton (H-1") of the other sugar nucleotide formed during the enzymatic epimerization, 2"- H-UDP-ManNAc. Other distinctive signals corresponding to this sugar nucleotide were also identified.  The signal corresponding to H-3" of 2"- H-UDP2  ManNAc can be assigned to a small doublet centered at 4.07 ppm and the signal corresponding to the methyl group of the acetamido moiety of 2"- H-UDP-ManNAc can be ascribed to the 2  presence of a small singlet at 2.00 ppm (0.03 ppm upfield of the analogous methyl group signal of 2"- H-UDP-GlcNAc, not shown in Figure 2.2). 2  Additional evidence for the incorporation of a solvent-derived deuterium isotope into the substrate has also been obtained by mass spectroscopy. The mass spectrum taken of the recovered mixture of 2"- H-UDP-GlcNAc and 2"- H-UDP-ManNAc after the epimerization 2  2  shows the expected increase of one mass unit relative to the non deuterated substrate.  36 2.5  Development of a coupled assay and kinetic characterization of UDP-GlcNAc 2epimerase In Chapter one it was mentioned that several discontinuous or stopped-assays had  previously been employed in order to determine the kinetic parameters of the epimerase. Due to the nature of an epimerization reaction, changes in the concentration of the substrate or product are not easily detected in a more desirable continuous fashion. For example, it is not ? possible to use UV-Vis spectroscopy since there is essentially no change in the absorption of the solution during the epimerization as both the substrate (UDP-GlcNAc) and the product (UDP-ManNAc) share the same UV-Vis spectroscopic features.  In order to circumvent this  problem, a common technique used in enzymology was employed. The reaction progress can be indirectly followed in a continuous fashion by 'coupling' it to another irreversible reaction. In the 'coupled assays' a convenient second enzyme is employed to further transform the product of the first reaction generating a measurable spectroscopic change (Appendix 2). The enzyme that seems to be the best choice for developing a coupled assay for UDPGlcNAc 2 epimerase is UDP-ManNAc dehydrogenase. As mentioned earlier (section 1.3), this NAD -dependent enzyme catalyzes the irreversible two-fold oxidation of UDP-ManNAc to +  UDP-ManNAcUA by using two equivalents of N A D . +  This oxidation generates two  equivalents of N A D H that conveniently shows a characteristic absorption at 340 nm. Therefore, it is possible to follow continuously the progress of this enzymatic reaction by monitoring spectrophotometrically the increase in absorbance at this wavelength. The optimal conditions required for the coupled assay were determined by Dr. Paul Morgan in our laboratory.  It was found that it was necessary to use a large amount of the  dehydrogenase (8.3 mg/mL) for a properly coupled assay due to the low activity of the dehydrogenase under the conditions of the assay.  This required the availability of large  quantities of the dehydrogenase. Since the amounts of dehydrogenase obtained from standard  37 purification protocols were insufficient (Kawamura et al., 1979), it was necessary to pursue the cloning and overexpression.of this enzyme. The details of the cloning, overexpression purification and characterization of the dehydrogenase can be found in Dr. Morgan's Ph.D. Thesis (Morgan, 1998). In the work described in this thesis, an additional purification step with weak anion exchange chromatography was included prior to the final purification by HPLC.  This additional step  proved to be useful in eliminating some of the impurities that produced a background rate in N A D H production.  The purified protein appeared homogeneous when analyzed by SDS-  P A G E and showed a molecular weight of 45 720 ± 4 Da which was identical (within experimental error) to the one reported by Morgan (1998). The coupled assay < was used to generate a plot of the epimerase rate against the concentration of the substrate and the resulting curve was sigmodial (Figure 2.4) which indicates the allosteric nature of the enzyme. A Hill coefficient (ri), which is indicative of the degree of allosteric activation, of 2.20 ± 0.2 and an apparent K  (fC) value was 0.47 ± 0.06  m  mM. These results are similar to those reported by Morgan (n = 2.29 ± 0.2 ; K' = 0.73 ± 0.09 mM;  1998) and by Kawamura et al. ( n = 2.0 ; K' = 0.63 mM ; 1979) for the enzyme purified  from natural sources. The k  value was calculated to be 11.3 s" . The error reported with the 1  cal  data indicates the deviation of the data from the curve-of-best-fit as calculated by the program Grafit (Erithacus Software, 1994). However, values determined from two other independent data sets, performed with different preparations of the epimerase, differed by as much as 20 %. Therefore, these results agree within experimental error to those reported previously.  38  0  0.5  1  1.5  2  2.5  3  3.5  4  [UDP-GlcNAc] (mM)  Figure 2.4 Plot of the initial rate of epimerization of UDP-GlcNAc versus its concentration.  2.6  Synthesis of UDP-ManNAc and its use in the evaluation of the regulation of UDPGlcNAc 2-epimerase  2.6.1  General strategies for the synthesis of glycosyl nucleoside diphosphates Glycosyl nucleoside diphosphates (or sugar nucleotides) have been the subject of recent  interest due to their application as substrates for glycosyl transferases used in enzyme mediated oligosaccharide synthesis (Wong et al,  1995).  Many synthetic preparations of these  compounds have been developed on a small scale (Kochetov and Shibaev, 1973; Gabriel, 1982). Enzyme-based approaches for glycosyl nucleoside diphosphate syntheses have been reported, and some of the procedures are practical in gram quantities (Heidlas et al, Wong et al,  1995).  1992;  The feasibility of an enzymatic approach largely depends on the  availability of the enzymes to be employed in the preparation.  For instance, while UDP-  glucose can be prepared easily starting from glucose by employing three stable and  39 commercially available enzymes (Figure 2.5), the enzymatic synthesis of UDP-GlcNAc have been hampered by the limited availability and laborious preparations of the required enzymes.  OH  OH  o  Figure 2.5 Enzyme-based synthesis of UDP-Glc; (i) UTP, hexokinase (EC 2.7.5.1), (ii) phosphoglucomutase (EC 2.7.5.1), (iii) UTP; UDP-Glc pyrophosphorylase (EC 2.7.7.9)  Enzyme-based preparations of UDP-ManNAc were described by Kawamura et al. (1978, 1979). Initially, they isolated UDP-ManNAc from an incubation of UDP-GlcNAc with UDP-GlcNAc 2-epimerase by using paper chromatography. However, due to the unfavorable formation of UDP-ManNAc under equilibrating conditions and the difficulty to isolate this compound from a mixture containing its epimer (UDP-GlcNAc) an alternate chemo-enzymatic approach was used. UDP-GlcNAc was transformed to UDP-ManNAcUA by incubation with the epimerase, UDP-ManNAc dehydrogenase and N A D . After isolation of the product by +  paper chromatography, the recovered UDP-ManNAcUA was subjected to a carbodiimide mediated reduction of the carboxylate moiety to afford UDP-ManNAc. Due to the low yields and the possibility of contamination of the UDP-ManNAc preparations, a more convenient preparation using chemical synthesis was investigated (Salo et al. 1970). The most widely employed synthetic method for the construction of glycosyl esters of nucleoside pyrophosphates involve the late-stage formation of the diphosphate moiety via coupling of deprotected glycosyl phosphates with nucleoside 5'-phosphoramidates. This  40 method was first reported by Moffat and Khorana (1958) and several modifications on the Nsubstituted analogs of the amidophosphate group have been investigated for the synthesis of a variety of sugar nucleotides. By far, the most popular modification has been the one that exploits nucleoside 5'-phosphoromorpholidates ('Khorana-Moffat' procedure; Roseman et al., 1961). Other procedures such as the 'mixed anhydride' method (Michelson, 1964), the use of carbodiimide (Kampe, 1966) and the phosphinothioic anhydride method (Furusawa et al, 1976) have been employed but have not received as general an acceptance as the KhoranaMoffat procedure (Figure 2.6).  OH OH  2) Mixed anhydride method  3) Phosphinothioic anhydride method  Figure 2.6 Methods for the chemical synthesis of glycosyl nucleoside diphosphates. B represents an organic base (Uracil, Guanine, Adenine, etc.)  41 The  chemical  syntheses of  glycosyl 1-phosphates  can be  divided into  two  fundamentally different approaches. In the first, the anomeric carbon of the sugar acts as the electrophilic center, and undergoes a nucleophilic addition by either phosphoric acid ('MacDonald reaction', MacDonald, 1962), salts of phosphoric acid (Cori et al, 1937) or dialkylphosphates.  The stereoselectivity of the phosphate incorporation at the anomeric  position (a or p) depends on the reaction conditions and on the presence of a neighboring participating group at the C-2 position of the sugar.  In the case of 2-acetamido-2-deoxy  glycosyl derivatives, the participating  group can generate  2-acetamido  an oxazoline  intermediate upon the formation of a carbonium ion at the anomeric center. Under kinetically controlled conditions, the oxazoline intermediate suffers the attack by the nucleophile to give the 1,2-trans linked product.  Under thermodynamically controlled conditions however, the  more stable a-glycosyl phosphate is formed regardless of the presence of a neighboring participating group. In the case of 2-acetamido-2-deoxy-glucopyranosyl derivatives both the stable a-phosphate (1,2-cis-linked product) and the somewhat unstable p-phosphate (1,2 translinked product) have been prepared using appropriate conditions. The latter is formed as the kinetic product and is difficult to isolate (Baluja et al, 1960; Khorlin et al, 1970; Roy et 1991; Busca and Martin, 1998; Figure 2.7 A).  al,  In the case of 2-acetamido-2-deoxy-  mannopyranosyl derivatives the formation of the a-phosphate product is favored based on both kinetic and thermodynamic grounds (Figure 2.7 B). This product is relatively unstable and its isolation is better accomplished after removal of the protecting groups of the phosphate moiety (Salo and Fletcher, 1970, Sabesan and Neira, 1992).  42  OR  1,2-cis thermodynamic product B) 2-Acetamido-mannopyranosyl derivative  1,2-trans kinetic and thermodynamic product  Figure 2.7 Preparation of 2-acetamido-glycosyl phosphates by nucleophilic attack at the anomeric center. A wavy bond denotes a mixture of a and (3 anomers. X represents an electrophilic leaving group (e.g. Cl). R and R' represent protecting groups on the phosphate moiety and the sugar alcohols respectively (e.g. alkyl, acyl, aryl).  In the second approach, the anomeric glycosyl oxygen is activated rather than its carbon. The anomeric oxygen then acts as a nucleophile on an electrophilic form of phosphorus. The most common examples of this methodology are the phosphorylation of the anomeric hydroxyl group of protected sugars by diphenyl phosphorochloridate (Sabesan and Neira,  1992)  or  the  two  step  phosphorylation  via  phosphitylation  with dibenzyl-  phosphoramidites followed by oxidation of the formed glycosyl phosphite (Sim et ah, 1993). The ratio of a to p anomer of the formed glycosyl phosphate depends on the initial anomeric composition of the protected hexopyranoses, the amount of base, the solvent, the temperature,  43 and the amount of phosphorylating agent (Sabesan and Neira, 1992; Sim et al., 1993). Again, the participating 2-acetamido group plays an important role in regulating stereoselectivity by providing a decomposition pathway for the 1,2-trans linked products (Figure 2.8; p-phosphate for GlcNAc, a-phosphate for ManNAc).  1,2-cis B) 2-A'cetamido-mannopyranosyl derivative  0—P=o I  1,2-trans  Figure 2.8 Preparation of 2-acetamido-glycosyl phosphates by nucleophilic attack at the phosphorus. A wavy bond denotes a mixture of a and p anomers. R and R' represent protecting groups of the phosphate moiety and the sugar alcohols respectively (e.g. alkyl, acyl, aryl).  44 2.6.2  Synthesis of UDP-ManNAc In order to test the reported necessity of UDP-GlcNAc in the enzymatic epimerization  of UDP-ManNAc, it was desirable to obtain pure UDP-ManNAc. In the work of Kawamura et al. (1978, 1979), which prepared UDP-ManNAc by an enzymatic procedure, problems arose due to slight impurities of UDP-GlcNAc in their preparations.  The preparation of UDP-  ManNAc by a chemical route was deemed superior since it provides UDP-ManNAc free of its epimer, UDP-GlcNAc.  Due to its increased purity, more reliable results could be obtained  when used to test the need of UDP-GlcNAc as a modulator of the epimerase activity.  In  addition, the evaluation of other potential substrates, as modulators of the epimerase activity was feasible (see Chapter Four). UDP-ManNAc has previously been chemically synthesized by two groups, Salo and Fletcher (1970) and Yamazaki et al. (1981), by coupling the 2-acetamido-2-deoxy-ct-Dmannopyranosyl phosphate (ManNAc a-1-phosphate, 7) with an appropriate activated form of uridine monophosphate.  Salo and Fletcher employed the Khorana-Moffat procedure and  Yamazaky et al. used a modified procedure of the mixed anhydride method. In the present work, it was decided to use the Khorana-Moffat procedure as described by Salo, due to its simplicity and the availability of the reagents. The synthesis of the ManNAc a-1-phosphate 7, started with the condensation of the peracetylated ManNAc oxazoline 5 (Khorlin et al,  1968) with dibenzylphosphate to give a  single product, presumably the labile peracetylated a-mannopyranosyl dibenzylphosphate 6 (Figure 2.9).  Attempts to isolate this product failed due to its decomposition. Instead, the  crude reaction mixture was subjected to catalytic hydrogenation to remove the benzyl groups followed by 0-deacetylation with sodium methoxyde in methanol.  Dowex  50W-X8  (pyridinium form) resin was used to neutralize the reaction in order to avoid acid-catalyzed hydrolysis of the phosphate.  Isolation of ManNAc a-1-phosphate was performed by  45 preparative T L C chromatography. Some decomposition of the sugar phosphate occurred during this purification procedure so that only low overall yields of the desired compound were obtained (18% overall yield).  UDP-ManNAc  HO  OH  Figure 2.9 Procedure for the synthesis of UDP-ManNAc : (i) Dibenzylphosphate, dichloroethane, r.t.(ii) H , Pd/C, MeOH (iii) MeONa, MeOH (iv) UMP-Morpholidate, pyridine, r,t, 5 days. 2  The coupling of 7 with the U M P moiety was accomplished using a slight modification of the standard Khorana-Moffat procedure (Roseman et al, 1961). The sugar phosphate was converted to its more soluble trioctylammonium salt by adding two equivalents of trioctylamine to a solution of the sugar phosphate in pyridine. The solution was then rendered anhydrous by repeated evaporation with dry pyridine and the residue mixed with a similarly treated solution of commercially available UMP-morpholidate. The mixture was allowed to react for 4-5 days at room temperature and the product was purified by anion exchange chromatography. The UDP-ManNAc containing fractions eluted after U M P as a major peak. Incidentally, the U M P  46 dimer, diuridine 5' pyrophosphate (as suggested by ' H and P N M R spectroscopy), produced 31  by coupling of U M P with the morpholidate reagent, was also found as a by-product eluting with or very close to the desired sugar nucleotide.  Elimination of this by-product was  accomplished by size exclusion chromatography. This last procedure also served to remove the inorganic salts present in the solution. The UDP-ManNAc containing fractions were pooled and lyophilized to afford pure UDP-ManNAc.  2.6.3  Evaluation of the regulation of UDP-GlcNAc 2-Epimerase In the study of epimerase from natural sources, Kawamura et al. (1979) reported that  there was an absolute requirement of UDP-GlcNAc in order to maintain active enzyme when UDP-ManNAc is used as the substrate. It was found that 0.6 mM UDP-GlcNAc was required to give half maximal activation of UDP-GlcNAc 2-epimerase. The sample of UDP-ManNAc used in the experiment performed by Kawamura et al. was isolated and purified from the product of the enzymatic epimerization of UDP-GlcNAc and therefore was not absolutely free of UDP-GlcNAc.  A re-examination of this property of the epimerase, was attempted with  purer chemically-synthesized UDP-ManNAc. Identical solutions containing 1.1 mM of UDP-ManNAc were incubated with the epimerase in the presence and absence of 0.35 mM UDP-GlcNAc.  After a pre-determined  period of time, the reaction mixture was analyzed by ion-paired reverse-phase HPLC using the method described by Meynial et al. (1995).  No detectable evidence of epimerization had  occurred in the sample with pure UDP-ManNAc (Figure 2.8 A), whereas the sample containing UDP-GlcNAc showed a 28% conversion of UDP-ManNAc to UDP-GlcNAc (Figure. 2.10 B). A more extensive incubation period produced less than 1% conversion in the sample with no UDP-GlcNAc (Figure 2.10 A), confirming the result that demonstrated the requirement of UDP-GlcNAc for epimerase activity.  47  0.24 -  0.24-  t = 120 min  t = 0 min 0.20  0.20-^ UDP-ManNAc  0.16-^  0.16  CM CO CM  0.12-  \  0.12  0.08  0.08  0.04 ^  0.04 \  0.0  JV  0.0  i i i i i i i i i i i i i i ' i ' I i I i I ' I 0 1 2 3 4 5 6 7 8 9 10 11 12  11' 0 1  111111111111111111111 2 3 4 5 6 7 8 9  Time (min)  10 11 12  Time (min)  B 0.20 H t = 0 m i n E  0.16-J UDP-GlcNAc  UDP-ManNAc  <  0.08 H  J 4  6  \  8  i  0.12  H  t = 10 m i n  CM CO CM  @  0.08 -  ui  -  .0  1  0.0  I  0.16  0.04-  ii ii iIii  0.0 10  12  Time (min)  0  2  4  ' i ' I ' I ' I ' I ' I ' I ' I 6 8 10 12 Time (min)  Figure 2.10 Ion-paired reversed-phase HPLC traces of a sample containing : (A) only UDPManNAc (1.1 mM) and (B) UDP-ManNAc (1.1 mM) and UDP-GlcNAc (0.35 mM) before (t = 0 min) and after incubation with epimerase.  2.7  Conclusions The incorporation of deuterium from the solvent during the enzymatic epimerization of  UDP-GlcNAc confirms the previously reported observation that the epimerase catalyzes solvent derived tritium incorporation at C-2". This experiment supports a mechanism for the enzymatic epimerization involving the cleavage of the carbon-hydrogen bond at C-2" via removal of a proton. Consequently, possible mechanisms involving direct hydride transfer at C-2" carbon-hydrogen bond can be ruled out. All the four possible mechanisms postulated in Chapter one are consistent with this experimental outcome.  48 The requirement for UDP-GlcNAc in the epimerization of synthetic UDP-ManNAc confirms the previously reported observation that the epimerase is tightly regulated by its own substrate. The specificity of the activation process of this dimeric enzyme suggests a model in which one subunit active site must first be occupied by UDP-GlcNAc in order to allow the second subunit to perform the epimerization, this is consistent with the model of positive cooperativity and the observed kinetic behavior of the enzyme. However, it is also possible that the enzyme possesses a modulator site, different from the active site, that is also specific for UDP-GlcNAc. The use of UDP-ManNAc dehydrogenase as an auxiliary enzyme for developing a continuous spectrophotometric assay for the epimerase gave a sigmodial curve for initial velocity kinetics of the epimerase with Hill coefficient of 2.20 ± 0.18, an apparent K  m  0.06 mM and a k  of 0.47 ±  of 11.3 ± 0.2 s" this values are consistent with the positive cooperativity 1  cat  and the kinetics values reported for the epimerase isolated from natural sources.  2.8  Experimental Methods  2.8.1  General UDP-GlcNAc and N A D was purchased from Sigma Chemical Co. (St. Louis, Mo). +  Other chemicals were purchased from Sigma Chemical Co., Fisher Scientific Co. or Boehringer Mannheim  Biochemicals  Inc.  2-Methyl-(2-acetamido-3,4,6-tri-0-acetyl-l,2-dideoxy-p-D-  mannopyrano)-[2,l-if]-oxazoline was prepared according to Khorlin et al. (1968) Thin layer chromatography was performed on aluminum-backed sheets of silica gel 6OF254  (Merck) of thickness 0.2 mm. Compounds were visualized by U V or spraying with a  solution containing H2SO4 (31 mL), ammonium molybdate (21 g), and Ce(S04)2 (1 g) in water (500 mL) and then heating at 110 °C for 3 min.  49 !  H Nuclear magnetic resonance spectra were recorded on Bruker instruments at the  indicated field strengths: an AC-200 at 200 MHz, a WH-400 at 400 MHz, or a AMX-500 at 500 MHz. All chemical shifts were reported using the 8 scale in ppm. 1 3  C N M R spectra were recorded on the following instruments at the indicated field  strengths: a Varian XL-300 at 75 MHz, or a Bruker AMX-500 at 125.76 MHz. Spectra obtained in CDCI3 were referenced to the solvent peak (8 = 77.0 ppm) 31  P N M R spectra were recorded on a Bruker AC-200 at 81.015 MHz. Chemical shifts  were reported in ppm using H3PO4 (8 = 0.00 ppm) as external reference.  2.8.2  General synthetic procedures  (a)  Catalytic Hydrogenation over Pd/C The dibenzyl ester of the peracetylated glycosyl phosphate (0.1 mol) was dissolved in  dry methanol (10-20 mL) in a 50 mL round bottom flask equipped with a three way stopcock which had one of its ends connected to a latex balloon filled with hydrogen. Palladium (10%) on activated carbon (10-20 % of the weight of the sugar phosphate added) was added. The suspension was stirred under an atmosphere of hydrogen, and the debenzylation monitored by T L C . After the reaction was complete the suspension was filtered through a fritted-disc funnel of fine porosity. The resulting solution was concentrated for further analysis or used directly in the O-deacetylation reaction.  (b)  Transesterification with sodium methoxide (O-deacerylation) A solution of peracetylated glycosyl phosphate (0.05 mol) in methanol (10-15 mL) in a  50 mL round bottom flask was cooled in an ice bath. A freshly prepared (IM)  solution of  sodium methoxide in methanol was slowly added until the resulting solution was basic by pH paper.  The solution was stirred until the reaction was complete as judged by T L C . The  50 reaction was stopped by stirring with A G 50W-X8 cation exchange resin (pyridinium form) until neutral pH was achieved. The resin was filtered of and washed with methanol. The combinedfiltrateand washings were evaporated in vacuo using a rotary evaporator.  (c)  General procedure for the preparation of uridine 5'-(2-Acetamido-2-deoxy-a-Dhexopyranosyl diposphates) A solution of commercially available 4-morpholine Af Af'-dicyclohexylcarboxamidinium  uridine 5'-phosphoromorpholidate (1.2 to 1.5 equivalents relative to the amount of sugar phosphate) in dry pyridine (10-15 mL) was concentrated to dryness in vacuo. The process of dissolution and concentration was repeated twice, dry air being admitted into the flask after each concentration. A solution of deprotected glycosyl phosphate in water was converted into its trioctylammonium salt by slowly passing through a column of Dowex 50-X8 (pyridinium) resin, washing the column thoroughly with water, concentrating in vacuo to ~ 5 mL, and adding a solution of trioctylamine (2 equivalents relative to the amount of sugar phosphate) in pyridine. The solution was then concentrated to dryness (in vacuo) in a rotary evaporator and three batches of dry pyridine (5 to 10 mL each) were successively evaporated in vacuo from the residue, dry air being admitted after each concentration The residue was finally redissolved in pyridine and the solution was added to the nucleoside phosphomorpholidate. The solution was again evaporated in vacuo and the mixture was finally kept in dry pyridine (5-10 mL) at room temperature for five days. The solvent was then evaporated in vacuo and the residue was redissolved in water and extracted twice with ether.  The ether extracts were back-extracted  with water and the combined aqueous extracts was concentrated in vacuo to ~1 mL and applied to an ion exchange column (Dowex 1X4-200 1cm x 20 cm; 100-200 mesh) equilibrated with 0.02 M LiCl at pH 3.5. The sugar nucleotide was eluted with a linear gradient developed from 0.02M to 0.5 M LiCl pH 3.5 while monitoring the U V absorption of the eluate at 254 nm using  51 a SPECTRUM Spectra/Chrom Flow Thru U V detector. The sugar nucleotide normally appears in the second peak, with the first peak containing UMP.  An additional peak frequently  appeared immediately after the peak of the desired compound (sometimes not well separated from it).  This peak corresponded to a by-product of the coupling reaction , diuridine-5'  pyrophosphate, that was separable during the desalting procedure.  The sugar nucleotide  containing fractions were pooled, neutralized by addition of pyridine, concentrated in vacuo to ~2 mL and desalted by passage through a 2.5 cm x 70 cm column of Bio-Gel P-2 (200-400 mesh) equilibrated with de-ionazed water and the compound eluted with deionized water. The major fractions exhibiting significant absorbance at 254 nm that contained no chloride ion, as determined by testing with 5% AgNC*3 solution, were pooled, concentrated in vacuo to a small volume which was , lyophilized and stored as a powder at -20 °C.  2.8.3  Synthesis of the Dilithium salt of Uridine 5'-(2-Acetamido-2-deoxy-ct-Dmannopyranosyl diphophosphate (UDP-ManNAc), 2. 2-Methyl-(2-acetamido-3,4,6-tri-0-acetyl-l,2-dideoxy-p-D-mannopyrano)-[2,l-fiT]-  oxazoline (400 mg, 1.22 mmoles) was dibenzylphosphorylated according to the procedure of Yamazaki et al. (1981). The solvent of the crude reaction mixture was removed in vacuo and the resulting mixture was redissolved in dry methanol and immediately subjected to catalytic hydrogenolysis and 0-deacetylation using the general procedures outlined earlier. The crude fully deprotected glycosyl phosphate was purified by preparative thin layer chromatography on 2mm silica plates using chloroform:methanol:H20 (10:10:3) as the eluent. The plate was visualized by spraying a small lateral band with molybdate/ceric sulphate solution followed by hot air drying using a heat gun. The sugar phosphate showed an Rf of approximately 0.15 and the band in this region was scraped off and eluted with the same solvent system. The solvent was evaporated of in vacuo to a small volume which was passed through a small A G 50W-X8  52 (pyridinium form) column an eluted with water. This afforded the monopyridinium salt of 2acetamido-2-deoxy-a-D-mannopyranosyl phosphate after evaporation of the solvent in vacuo (85 mg, 0.22 mmoles , 18% overall yield). The 'H N M R of this compound was identical to that reported in the literature (Yamazaki et al,  1981)  The fully deprotected glycosyl phosphate (85 mg., 0.22 mmoles) was subjected to UMP-Morpholidate coupling using the general procedure described above. The UDP-ManNAc prepared in this fashion (22 mg, 16% yield) showed an identical *H N M R to the one reported in the literature (Yamazaki et al, 1981).  2.8.4  Protein Determination Protein concentration was assayed by the Bradford dye-binding procedure (Bradford,  1976) with protein assay solution purchased from Bio-Rad.  A standard curve was obtained  using bovine serum albumin (Bio-Rad) as the standard.  Different concentrations of the  standard protein were prepared in 50 mM Tris-HCl pH 8.8 buffer and mixed with the assay solution to obtain a plot of OD at 595 nm versus the concentration of the standard.  The  concentrations of the samples were obtained by reading the OD at 595 nm of diluted solutions (in the same buffer) that fitted within the standard curve. All measurements were performed at room temperature.  2.8.5  Definition of a Unit A unit of UDP-GlcNAc 2-epimerase is defined as the amount of epimerase that  produces 1 umol of UDP-ManNAc per minute under the standard conditions of the coupled assay (37 °C, pH 8.8, 2 mM DTT, 4 mM N A D ) in the presence of 4 mM UDP-GlcNAc. The +  specific activity of the epimerase under these conditions was determined to be 16.3 ± 0.5 units/mg.  53 2.8.6  Strains and Media The plasmid containing the cloned epimerase gene (pKI86) was provided by Drs. Lucia  B. Rothman-Denes and Diane R. Kiino (University of Chicago). The plasmid containing the cloned dehydrogenase gene (pUSOl) was provided by Dr. Paul Morgan (UBC). Escherichia coli JM109(DE3) (Promega) was used as the host for the overexpression of both, the epimerase and dehydrogenase genes. Escherichia coli cells were grown in L B media (Difco) that contained ampicillin (100 ixg/mL). General microbiology techniques used are those described by Ausbel et al. (1992).  2.8.7  Overexpression and purification of the cloned UDP-GlcNAc 2-epimerase The pKI86 plasmid was transformed into previously prepared CaCb-competent  Escherichia coli JM109(DE3) cells and inoculated onto LB-agar plates containing 100 ug/mL of ampicillin. After overnight incubation of the plates at 37 °C, a single colony was chosen and used to inoculate 1 L of L B media supplemented with 100 p.g/mL ampicillin. The inoculated media was shaken (260 rpm.) at 37 °C and the cell growth was monitored by observing the optical density (OD) of the media at 600 nm. Upon reaching an OD between 0.8 and 1.2, IPTG was added to the media (0.4 m M final concentration). The cells were left to grow for an additional 3 h, then harvested by centrifugation for 15 min at 5000 rpm and stored at -78 °C. The pelleted cells (1.35 g) were thawed with warm water and resuspended in 50mM triethanolamine-HCl buffer pH 8.5 containing glycerol (10% v/v), E D T A (2mM), dithiothreitol (2mM), pepstatin (1 mg/L), aprotinin (1 mg/L) and phenylmethylsulfonyl fluoride (1.5 mM) (buffer A).  Stock solutions of pepstatin and aprotinin (5 mg/mL) and phenylmethylsulfonyl  fluoride (lOOmM) in ethanol were prepared. Cells were lysed by two passes through a French pressure cell at 20 000 psi. Following ultracentrifugation at 30 000 rpm for 45 min to pellet the  54 cell debris, the resulting lysate (6.2  mL)  was loaded onto a column (15  mL)  of  diethylaminoethyl-cellulose (DE-52, Whatman) that had been pre-equilibrated with buffer A . The column was eluted with a linear gradient of buffer A (200 ml) and the same buffer containing 200 mM NaCI (200 mL). The eluent containing the relevant fractions as judged by SDS-PAGE  were concentrated using an Ultrafree-15  (Millipore) concentrator.  The  concentrated protein solution was desalted with buffer B (50mM triethanolamine-HCl buffer pH 8.5 containing 2mM D T T and 10% v/v glycerol) by repeated dilution and concentration using the same concentrator. The concentrated protein (approximately 2 mL of 36 mg/mL) was frozen in liquid nitrogen and stored at -78 °C. The protein solution was quickly thawed and applied to a Waters AP-1 Protein-Pak Q 8HR HPLC anion exchange column attached to a Waters 625 L C system equipped with a Waters 486 tunable UV/Visible absorbance detector to monitor the appearance of protein by its absorbance at 280 nm.  After application of the  protein, the column was washed isocratically with buffer B (24 mL) and was developed with a linear gradient of NaCI (0 to 0.5 M) in the same buffer with a flow rate of 0.8 mL /min. The relevant fractions were pooled, concentrated and separated in small portions (4 mg/ml) that were frozen with liquid nitrogen and stored at -78 °C.  2.8.8  Solvent Deuterium Isotope Incorporation A sample of potassium phosphate buffer (50 mM, pH 8.8, containing 18.0 mM UDP-  GlcNAc) was lyophilized and resuspended in an equal volume of D 2 O two times. UDPGlcNAc 2-epimerase (0.27 units) was added to 450pL of this deuterated buffer. The reaction was monitored by H N M R (500 MHz, 32 scans) at 25 °C. [  55 2.8.9  Overexpression and purification of the cloned UDP-ManNAc dehydrogenase The pUSOl plasmid was transformed onto previously prepared CaC^-competent  Escherichia coli JM109(DE3) cells and inoculated on LB-agar plates containing 100 p,g/mL of ampicillin. After overnight incubation of the plates at 37 °C, a single colony was chosen and used to inoculate 2 L of L B media supplemented with 100 Lig/mL ampicillin. The inoculated media was shaken (260 rpm) at 37 °C and the cells growth was monitored by observing the optical density (OD) of the media at 600 nm. Upon reaching an OD between 0.8 and 1.2, IPTG was added to the medium (0.4 mM final concentration). The cells were left to grow for an additional 5 h, harvested by centrifugation for 15 min at 5000 rpm and stored at -78 °C. The pelleted cells (~ 4 g) were thawed with warm water and resuspended in 50mM triethanolamineHC1 buffer pH 8.5 containing glycerol (10% v/v), E D T A (2mM), dithiothreitol (2mM), pepstatin (1 mg/L), aprotinin (1 mg/L) and phenylmethylsulfonyl fluoride (1.5 mM) (buffer A). Cells were lysed by two passes through a French pressure cell at 20 000 psi. Following by ultracentrifugation at 30 000 rpm for 45 min to pellet the cell debris, the resulting lysate (6.2 mL) was loaded onto a column (100 mL) of diethylaminoethyl-cellulose (DE-52, Whatman) that had been pre-equilibrated with buffer A and then eluted with a linear gradient of buffer A (200 ml) and the same buffer containing 500 mM NaCl (200 mL). The fractions containing the dehydrogenase, as judged by SDS-PAGE were concentrated using Ultrafree-15 concentrators.  (Millipore)  The buffer of the concentrated protein solution was desalted with buffer B  (50mM triethanolamine-HCl buffer pH 8.5 containing 2mM D T T and 10% v/v glycerol) by repeated dilution and concentration using the same concentrator. The concentrated protein was frozen in liquid nitrogen and stored at -78 °C. The protein solution was quickly thawed and applied to a Waters AP-2 Protein-Pak Q 40HR HPLC anion exchange column attached to a Waters 625 L C system equipped with a Waters 486 tunable UV/Visible absorbance detector to  56 monitor the appearance of protein by its absorbance at 280 nm. After application of the protein, the column was washed isocratically with buffer B (100 mL) and was developed with a linear gradient of NaCI (0 to 0.5 M) in the same buffer with a flow rate of 3.6 mL /min. The fractions containing dehydrogenase eluted at a salt concentration of approximately 0.2 M NaCI. These fractions were pooled, concentrated and separated in small portions that were frozen with liquid nitrogen and stored at -78 °C.  2.8.10 Kinetic characterization of UDP-GlcNAc 2-epimerase The kinetic parameters of the cloned UDP-GlcNAc 2-epimerase were determined in the UDP-GlcNAc to UDP-ManNAc direction with a continuous coupled spectrophotometric assay using UDP-ManNAc dehydrogenase as the auxiliary enzyme.  The initial rates were  determined at 37 °C in assay mixtures (0.6 mL total volume) containing 50 mM Tris-HCl (pH 8.8) buffer, 2 m M DTT, 4 mM N A D and UDP-ManNAc dehydrogenase (5 mg of protein in +  each cuvette) and a variable concentration of UDP-GlcNAc (0.22 to 3 mM). The epimerization was initiated by adding of UDP-GlcNAc 2-epimerase (2.9 x 10" units) and followed by 3  measuring the increase in absorbance at 340 nm due to N A D H formation ( £340 = 6220 M^cm" ) 1  using a Varian Cary 3E spectrophotometer.  The rate data was divided by a factor of 2 to  account for the stoichiometry of the dehydrogenase reaction. The resulting rate was plotted as a function of substrate concentration, and the kinetic parameters were determined by a direct fit of the data to a Hill equation using the computer program Grafit (Erithacus Software, 1994). The program employs a non-linear regression analysis on the data using the method of Marquart(1963).  57 2.8.11 Evaluation of the regulation of UDP-GlcNAc 2-epimerase A solution of UDP-ManNAc (1.1 mM) in potassium phosphate buffer (40 uL, 50 mM, pH 8.1, containing 2 mM dithiothreitol) was incubated for 10 min at 37 °C with the epimerase (2.6 x 10" units). 4  An analogous solution that also contained 0.35 mM UDP-GlcNAc was  treated identically. The reactions were analyzed by ion-paired reversed-phase HPLC using a Waters Radial-Pak 8NVC18 reversed-phase HPLC column pre-equilibrated with 50 mM phosphate buffer (pH 7.0, containing 2.5 mM tetrabutylarnmonium hydrogen sulfate) and eluting with a linear gradient 0-10% C H 3 C N in the same buffer, following the procedure described by Meynial et al. (1996).  Chapter Three  Positional Isotope Exchange (PIX) Studies with UDP-GlcNAc 2-Epimerase  5-g  59 3.1  Introduction One of the main goals in the study of enzyme mechanisms is to understand the pathway  by which existing chemical bonds are broken and new ones are made during the conversion of substrates to products. Several methods have been devised to detect the chemical events that occur during this conversion. In some cases, enzyme-bound intermediates, that might form during enzymatic processes, can be detected by spectroscopic means (UV-visible, NMR, EPR, etc.) or, if they are loosely bound to the enzyme and sufficiently stable, can even be isolated and characterized. In these cases, unambiguous evidence of the reaction mechanism followed by the enzyme may be obtained.  Most often, however, enzyme-bound intermediates are  difficult, if not impossible, to detect or isolate and indirect techniques must be employed. A convenient method to gain valuable information in the study of chemical reactions, particularly organic reaction mechanisms, is through the use of isotopes incorporated in specific places of the molecules under study. This method is also employed in the study of enzymatic reactios. Thus, it is possible to limit the number of reasonable chemical mechanisms proposed for the mode of action of an enzyme by observing the location of a labeled atom in the substrate before and after the enzymatic reaction. Since the early fifties, molecular isotope exchange (MIX) has been a useful technique for detecting and proposing reasonable enzyme-bound intermediates (Douderoff et al, Boyer, 1959). For example, in a enzymatic reaction:  AB +  C  would look for the exchange of the isotope of B (B*): A B +  < *  AC +  B* =*=^  1947;  B , one  AB* +  B in  the absence of C as an evidence for an initial cleavage of the A - B bond by the action of the enzyme E, as shown below: C E + A B -«  ^  E« A B  E  A  -<  >  E  A  + B  ^  —  E  + AC  60 An inherent limitation of this technique is that it requires B to be freely exchangeable in the absence of C after the formation of the enzymatic complex in which A and B are no longer  bonded ( E  B  ). This is not always the case, since B may only be released upon the binding of  c. An interesting situation arises when A and B are connected via a carboxylate or phosphate ester linkage. In these cases, the transfer reaction  E«AB ~*  *"  can be  E  B detected by intramolecular isotope scrambling: O  •  W  \\  C — R  C — R  The occurrence of this process is termed positional isotope exchange (PIX) and does not A depend on product release ( E  A <  *  E  + B ) but, rather, requires the cleaved  intermediate B to have a degree of freedom that allows the equilibration of similar atoms (oxygens) by torsional motion:  O  Therefore, when E  B  returns to A B , the labeled oxygen should be distributed randomly  among the two (carboxylate) or three (phosphate) positions, provided that the torsional equilibration occurs at a rate higher than the overall rate of the enzymatic reaction.  61 The application of PIX has often been used with enzyme-catalyzed reactions involving phosphoryl transfer and comprehensive reviews can be found in the literature (Raushel and Villafranca, 1988; Cohen, 1983, Rose, 1979).  In the first application of the technique,  Midelfort and Rose (1976), showed that y-glutamyl phosphate (y-Glu-P) was an intermediate formed during the glutamine synthetase reaction (Figure 3.1).  (a)  O"  ATP  0" i+ NH  V  -  NH  ADP  NH  2  ^^^A  ^  3  0"  0  +  p  NH3  3  Gin  Glu  (b)  ATP  +  Glu  ADP • y-Glu-P + N H  ««——*•  3  ADP • y-Glu-P  ——*•  ADP  + Gin + p. •  Figure 3.1 (a) Overall reaction catalyzed by Glutamine synthetase. Pj represents inorganic phosphate, (b) Proposed stepwise mechanism for the action of the enzyme.  18  The detection of the movement of one of the  O isotopes in labeled ATP, from a p-y  bridging position (6 in Figure 3.2) to a p nonbridging position (7) during the incubation of labeled ATP and glutamate in the absence of ammonia, was indicative of the reversible formation of the enzyme-bound reaction intermediate ADP (Figure 3.2). Since no chemical or enzymatic methods were available for removing directly the y-P04 unit intact, Midelfort and Rose devised an indirect method to determine the extent of the PIX on the recovered ATP.  62  all PH Y II Ad-0-P-0-P-«-P-«  I o_  I  0_  I  +  II II V»v Ad—O-P-O-P.i I  Glu  o_  •_  6  -Glu  o  y-Glu-P  ADP  ATP  Ad—O-P-O-P-O—PI I I  o  Figure 3.2  o  +  Ad—O-P-O-P~., I \ o 0_  Glu  -Glu  f C  •_  P I X experiment o n A T P labeled substrate. •  denotes  18  0  atom and A d denotes  Adenine  The  p and y p h o s p h o r y l groups o f the r e c o v e r e d A T P were subjected to a s e c o n d  s c r a m b l i n g reaction u s i n g A c e t y l C o A  synthetase.  This enzyme  catalyzes the  reversible  adenylation o f acetate w i t h the transient f o r m a t i o n o f pyrophosphate (Figure 3.3). equilibrating conditions this e n z y m e w i l l generate, besides structures 6 and 7, structures s u c h as 8 and 9 ( F i g u r e 3.3).  Under  labeled A T P  T h e extent o f the P I X i n the o r i g i n a l reaction was then 18  evaluated b y the appearance o f Y - P O 3 groups w i t h o n l y one a t o m o f  O (as i n 9).  The  equilibrated A T P sample was reacted w i t h d i h y d r o x y a c e t o n e and g l y c e r o l kinase i n order to form  dihydroxyacetone  phosphate  containing  the  Y-PO3  group.  The  dihydroxyacetone  phosphate was then converted to m e t h y l g l y o x a l and i n o r g a n i c phosphate b y b a s e - c a t a l y z e d elimination  a n d the  phosphate  was  permethylated  trimethylphosphate b y c o n v e n t i o n a l mass spectroscopy.  and  analyzed  for  O  enrichment  as  63  0  0 II II  0 II II  II -0-P 1 Ad-O-P-0- -pi1 1 1 0_ 0_ o  + H  0  II  3  •  II  1  I  •  Q  3  o  II  Ad-0-P-«-P-«-P-  o  O Acetyl-CoA synthase || O * Ad-O-P-0—^ + \_ CH  0 II C ^ O  0  II  0  •  II  0 0 || || HO-P-0—P-0 \ \  o  II  0  0  .  Ad—0-P-«-P-0—P-« O O  I o  Figure 3.3 Reaction catalyzed by AcetylCoA synthetase.  This rather elaborate derivatization, necessary for GC-MS analysis, was eventually circumvented by employing fast atom bombardment mass spectrometry (Conolly et al. 1984; Hilscher et al, 1985), however the use of mass spectroscopy for the analysis of PIX experiments has largely been displaced by the use of NMR spectroscopy. The application of NMR spectroscopy in the analysis of PIX relies on the appearance of new signals in the spectra of the labeled substrates subsequent to treatment with the enzyme. The new signals are produced as a consequence of the scrambling of isotopes within the molecule and are observable only after adopting conditions for high resolution NMR spectroscopy. The chemical shift of any nuclei resonance depends on the induced magnetic fields of the circulating electrons in its immediate environment. The immediate environment obviously includes the directly attached atoms and therefore, it is expected that the substitution 18  of different isotopes, for example  16  O for  O, will perturb the electronic environment at a  phosphorous or carbon nuclei that is directly bonded to it inducing a measurable change in the chemical shift.  64  The physical basis for this effect has been understood for some time and has been observed in H and C NMR spectra on substitution of a proton in an organic molecule by a !  l 3  deuterium. Ramsey (1952) predicted the effect as resultingfromperturbations of the zero-point vibrational levels by substitution of one isotope for another. Later on, Gutowsky (1959) held that the effect of changes in the vibrational levels manifest themselves at the probe nucleus in an electrostatic fashion, and that the change in shielding should be proportional to the inverse fourth power of the difference between average distances of the two isotopes from the observed nucleus. Batiz-Hernandez and Bernheim (1967) provided a similar qualitative explanation for the isotope induced shift based on the changes in vibrational levels of bonds between the observed nucleus and the isotope. They proposed that the electricalfieldsemanatingfromthe bonded atoms diminish the absolute shielding at the observed nucleus. Thesefieldsarise from the oscillations of the bonded atoms in the applied magnetic field.  Thus, when a heavier  isotope is substituted for a lighter one, its average distance to the observed nucleus is slightly shorter, and its vibrational amplitude is smaller. It then must produce smaller electric fields which, when felt at the probe nucleus, do not deshield it as strongly as would those from the lighter isotope. This accounts for the almost universally observed trend of shift to higher field (lower ppm numbers) on substitution of a heavier for a lighter isotope at the probe nucleus. Gutowsky (1959) predicted that the magnitude of such shifts should be on the order of 0.01 ppm and should reflect the number of such substitutions. The isotopic shift was observed experimentally when Conn and Hu (1978) discovered that the substitution of an O for an O in phosphate and phosphate esters resulted in about a \  i  31  0.02 ppm upfield chemical shift in the  P NMR resonance for each oxygen substituted. This  effect is illustrated in Figure 3.4, in which the phosphate randomly labeled with 44%  P NMR spectrum of a sample of inorganic  O is shown. The signals arising due to the five  possible species containing 0 to 4 atoms of O are clearly separated.  65  16  o o 18  2  16  2  o o 18  3  ppm Figure 3.4 P - N M R spectrum at 145.7 M H z of inorganic phosphate randomly labeled with 44% 0 and 66% 0 (adapted from Cohn and Hu, 1978). 31  1 8  1 6  The magnitude of the isotopically induced chemical shift is well correlated with the bond order between oxygen and phosphorous. This phenomenon has been exploited in studies 1S  on nucleotide phosphate esters. Thus, one can distinguish between  O in a 'bridging' position  to another phosphorous or carbon ( P - 0 bond order of one) from the 'nonbridging' ( P - 0 bond 18  order higher than one) position, and the number of  O incorporated by observing the  magnitude of the isotopically induced chemical shift. This N M R approach has been applied to many enzymatic PIX reactions and several examples can be found in the literature (Cohen, 1983: Raushel and Villafranca, 1988). One of the limitations of this technique is that approximately 5 to 10 umol of labeled substrate is required per analysis. This is significantly more material than the comparable mass  66 spectroscopy methods. In addition, in order to get the maximum possible resolution the sample must be titrated to approximately pH 9 and all of the divalent cations must be sequestered. This is required in order to minimize the exchange broadening which occurs in the N M R signal in the presence of metals and protons (Raushel and Villafranca, 1988; Robitaille et al, 1991).  3.2  Application of PIX in the study of UDP-GlcNAc 2-Epimerase As described in Chapter One, some of the possible mechanisms (Path B and D) that this  enzyme can use to catalyze the epimerization involve the transient cleavage of the bond between the anomeric carbon (C-l") and the anomeric oxygen of the glucopyranose moiety. This cleavage will form enzyme bound UDP along with either a 2-acetamidoglucal (path B) or a 3-keto 2-acetamidoglucal (Path D) as intermediates. In order to probe the transient formation of enzyme bound UDP during the epimerization of UDP-GlcNAc 2-epimerase it was considered relevant to perform PIX studies with this enzyme. The PIX experiments employed synthetic UDP-GlcNAc that contained an 0 isotopic 1 8  label at the pyranose sugar-UDP bridging position (top structure in Figure 3.5).  In the PIX  experiments, the labeled substrate was incubated with the enzyme and the equilibrated products are examined for evidence of isotopic scrambling after a period of time. Provided that UDP is transiently formed and has a life time comparable to, or greater than, that required for the torsional equilibration of the oxygen atoms of the p-phosphoryl group, the  O isotope should  be found in three of the positions of this group ('scrambled') for both of the recovered epimers. If the scrambling is complete, a statistical distribution (2:1 ratio) of the label on the nonbridging position to the bridging position should be observed.  67  H O - V - - M AcNH  I •  O  O  N  II  —P—0—P—O  A- A-  N  u  UDP-GlcNAc  HO  v  HO  "  .  HO"T-^-°.  •  H 0 - X _ - 5 ^ AcNH I II „ H O—P—0—P—O  n  o  O  AcNH  II  o UDP-GlcNAc  A-  •P—0. o  O  0  UDP-ManNAc  II HOI  o  O - J - O - P - C ,  AcNH  HO HO'  0  II  o  II  • —P—O—P—o  I.  I.  N  o o Figure 3.5 Test of C - O anomeric bondUDP-ManNAc cleavage by PIX during the enzymatic epimerization of UDP-GlcNAc. Darkened atoms represents 0 label. 1 8  3.3  Synthesis of labeled substrate for the PIX experiment The strategy for the synthesis of  introduction of the  1 8  O-labeled UDP-GlcNAc (10), started with the  0 label into the anomeric position of the pyranose sugar in order to  produce labeled 2-acetamido-3,4,6-tri-0-acetyl-2-deoxy-a-D-glucopyranose  (11a). Compound  11a was then converted in two steps to the a-dibenzylphosphate derivative 12, by reaction with dibenzyl A/;/Y-diethyl-phosphoramidate in the presence of 1,2,4-triazole, followed by oxidation of the labile glycosyl phosphite. Complete removal of the protecting groups yielded the desired 18  0-labeled GlcNAc phosphate (13), which provided 10 upon UMP-morpholidate coupling  using standard conditions (Figure 3.6).  68 0" N a  OBn  "NHAc  Acer"  |  0)  AcO^r^S-  A c O ^ V AcO°'  (")  OBn V ^ N H A c  OAc  OAc  11a  12  (iii)  HO^^i  (iv)  HO"'  +  |  —p=o  1. 0 N a+ ''NHAc  V  OH  13 (V)  O II - O — P - O  H O - ^ r HO"'"  S  - p S  ''NHAc HO  OH  r v  10  Figure 3.6 Synthesis of 0 labeled substrate (i) Dibenzyl A^N-diethylphosphoramidate, 1,2,4triazole, CH C1 (ii) H 0 , THF, -78 °C (iii) H , 10% Pd/C, C H O H (iv) CH ONa, C H O H (v) UMP-morpholidate, trioctylamine, pyridine, 5 days. 1 8  2  2  2  2  2  3  3  3  An initial attempt to prepare 11a was based on the reported (Warren et al, 1984; Nakabayashi et al,  1988) acid-catalyzed hydrolysis of 2-methyl-(3,4,6-tri-0-acetyl-l,2-  dideoxy-a-D-glucopyrano)-[2,l-d]-oxazoline  (14)  that  generates  2-acetamido-3,4,6-tri-0-  acetyl-2-deoxy-a-D-glucopyranose (Figure 3.7). By heating 14 with 95% enriched 0 water in 1 8  C H C N and catalytic amounts of /7-toluensulfonic acid, the expected anomerically deprotected 3  sugar was isolated as the main product and gave an identical ' H N M R spectrum to the one reported in the literature (Sim et al, 1993).  Figure 3.7 Reported acid-catalyzed hydrolysis of GlcNAc oxazoline  69 The location of the  0 label, however, was found not to be at the anomeric position.  1 8  i o  Evidence for the position of the the mass spectra.  O label came from analysis of the fragmentation patterns in  A typical fragmentation for this particular peracetylated monosaccharide  involves the loss of H 2 O from the anomeric position resulting in the formation of an oxonium ion (Figure 3.8). If the  l s  O label were only incorporated at the anomeric position, a major peak 1R  at m/z =330 should be observed regardless of the percentage of incorporation  O. In contrast,  if the incorporation of 0 isotope occurs at a different site on the molecule, a signal at m/z = 1 8  332 should be observed.  AcO  H 1+ -H  ^  H«  AcO'  2  "'NH  AcO " 4  OAc^J^  n .0.  AcO '^^ 0  c OAc  O  m/z = 350 [M ( 0) + H ] m/z = 348 [M ( 0) + H ] 18  +  16  +  ''NH  A  0  m/z = 330  H U AcO  H0  AcO'  ^  2  vcO^^T AcO  ''NH  AcO"'""NH  OAc  m/z = 350 [M ( 0) + H ] m/z = 348 [M ( 0) + H ] 18  +  16  +  m/z = 332  1  Figure 3.8 Mass spectra fragmentation pattern analysis for incorporation of anomeric center.  s  O at the  The mass spectra of the product isolated after acid-catalyzed hydrolysis of oxazoline 14 gave both signals (m/z 330 and 332) resulting from the loss of H 2 O from the anomeric position.  70 This suggests that in a portion of the sample, the label is in another part of the molecule distinct from the anomeric position. The most likely alternative position for incorporation of the label is the carbonyl of the acetamido group. Further evidence for the incorporation of the was found by careful analysis of the  1 3  18  0 label into the acetamido group of 11  C N M R spectra. When compared to the  1 3  C N M R of an  authentic unlabeled sample of 11 (synthesized as described by Avalos et al., 1993) a slight difference was observed. The carbonyl signal corresponding to the acetamido group (-170.4 ppm) appeared to be shorter and broader than the one from the authentic sample. By increasing the resolution of the spectra using mathematical manipulation in the processing of the FID and acquiring the spectra with a shorter sweep width, the broad peak was resolved into two different signals of similar intensity having very close chemical shifts (170.49 and 170.46 ppm; Figure 3.9).  This outcome is consistent with the reported shifts in the N M R resonance  positions of C nuclei that are attached to oxygen when an 0 isotope substitutes the naturally 1 3  occurring 0 (Risley and VanEtten, 1979; Vederas, 1980). 16  18  71  16  0-11  / / o-n ,8  11 11 | 1 1 11 | 1 1  171.8  II I[ I  171  4  I j M  I I | I I I I | I I I I | I I I  171.0  170.6  1  | I I I I | I I I I | I I II  170.2  | I II  I | I I I I | I I I I | I I I I | I I I I | I I I I | I I  169.8  169.4  169.0  168.6  ppm 13  Figure 3.9 Expanded carbonyl resonance region of high resolution C N M R for the product of acid catalyzed hydrolysis of oxazoline of GlcNAc. The signals arising from the acetamido group for the labeled and unlabeled product are indicated.  The mechanism that can account for this experimental outcome is portrayed in Figure 3.10.  Under acid-catalyzed conditions, oxazolonium 14a will be formed and its hydrolysis  could proceed either by the addition of water to the C-7 acyl carbon to form orthoimidate 15 (route (a)), or by direct nucleophilic attack of water at the anomeric C - l (route (b)) to give 1-p1 8  0 H GlcNAc peracetate ( l i b ) which anomerize under this conditions to produce the most  stable observed product l-cc- OH GlcNAc peracetate (11a). 18  This later route seemed more  reasonable since acid-catalyzed addition of alcohols and other nucleophiles to C - l of the GlcNAc oxazoline peracetate is known to give exclusively p-glycosides (Warren and Jeanloz, 1977; Kunz and Unverzagt, 1988).  72  (a).  .0^ ' .vO  AcO  5  A  1  AcO' OAc  'N H  AcO  AcO'  ' T T X*  AcO°  AC0^-S^""'N V 1 H ^J(c) /  +  Y OAc  OAc 15  14a  ''NH  3  p-TsO"  16  D  B AcO  k/°Y"'  O. AcO AcO ' 1  > ^  ''NH  AcO  I  'NH  ,0.  . 0H  y  'NH  ^  AcO'  OAc  OAc  lib  X  AcO'  0  OAc 11c 11a  Figure 3.10 Possible pathways for the hydrolysis of GlcNAc oxazoline (see text for details).  The orthoimidate 15 could collapse, under the acidic conditions, to the protonated 218  amino-a-tetraacetate glucopyranose 16 (route C) or to the 1-a-OH GlcN- O-Ac peracetate 11c (route D). Studies done subsequent to this work by Jha and Davis (1995), postulate that the hydrolysis of the oxazoline of GlcNAc occurs via route A and C, with formation of aminoester 16 as the kinetic product which further rearranges to give the thermodynamic product 11c, even under the acidic reaction conditions. In fact, Jha and Davis reported the isolation and identification of the amino ester 16 by treating GlcNAc oxazoline 14 with one equivalent of />-toluensulfonic acid and excess water in CH3CN.  The amino ester 16 precipitated as its tosylate-ammonium salt in quantitative yield,  but did not undergo 0-1 -> N-2 acyl migration because of the tosylate insolubility in C H 3 C N . However, when sub-stoichiometric amounts of /?-toluensulfonic acid (0.1 eq.) were used, the  73 rearrangement was observed. It was postulated that the unreacted oxazoline, which is a weak base with pKa - 5 . 5 (Deslongchamps, 1975), was necessary to deprotonate the ammonium salt 16 which has a pKa ~ 8.50, thus making the 0-1  —» N-2 acyl rearrangement to the  hydroxyamide 11c kinetically feasible. The result obtained in our experiment, differs somewhat with the mechanism proposed by Jha and Davis for the hydrolysis of this oxazoline. While they postulated that the hydrolysis of the oxazoline proceed only via route A (and then C or D), our results shows that route B (the direct attack of water to the anomeric carbon), is competitive with route A (formation of orthoimidate 15) and therefore should not be disregarded. In practical terms, this hydrolysis failed to provide the desired l-a- OH-GlcNAc peracetate (11a) with a substantial level of 18  incorporation of 0 isotope at the anomeric position. 1 8  An alternative approach to obtain 11a was therefore devised.  During a study to  compare the rate of mutarotation and the rate of exchange of H2O in the anomeric center of glucose, Rittenberg and Graff (1958) prepared partially labeled glucose by boiling glucose in 30%  1 8  0 water for 18 hours. They reported 0 incorporation in the free sugar somewhat less 18  than one-sixth of the isotope concentration of the water. Although the incorporation of 0 at 1 8  the anomeric center was not satisfactory, many factors could account for this outcome and we sought to improve the conditions for the incorporation of label into GlcNAc. The percentage of 1 8  0 incorporation increased significantly by using 1-a-OH-GlcNAc peracetate (11) instead of  the deprotected monosaccharide and by employing water that was considerably more enriched in  18  0  (-95%).  In addition, the exchange reaction was carried out at slightly lower  temperatures (90-95 °C) in the presence of a non-nucleophilic, polar co-solvent (acetonitrile) that will improve the chances of recovering the desired a-anomer since the anomeric effect is 18  stronger in this solvent than in water.  Under these conditions, the percentage of  O  incorporation at the anomeric position after the exchange ranged from 60 to 90% as determined  74 by mass spectroscopy.  Lowering the temperature of the reaction mixture gave a lower  percentage of incorporation but gave cleaner reaction mixtures. An additional chromatography step was necessary in order to purify the product since the final reaction mixture was a dark 1  o  brown syrup. This gave the desired 1-a- O H GlcNAc peracetate (11a) in acceptable yields 1X  (-70%) with reasonably high levels of O incorporation. The presence of the O at the anomeric position was confirmed by analysis of the mass l s  spectra of the isolated product. In addition to the M + H signal corresponding to the labeled +  (m/z =350) and the non-labeled (m/z '= 348) compound, the signals corresponding to the fragment resulting from elimination of water from the anomeric position of 11a showed no significant signal at m/z = 332 (in contrast to the strong signal at m/z = 330), which indicates that the  1 8  0 label was located as planned.  Once the 0 label was introduced at the anomeric position, 11a was converted to 21 8  18  acetamido-3,4,5-tri-0-acetyl-2-deoxy-[l-  0]-a-D-glucopyranosyl dibenzylphosphate (12) by 18  the procedure of Sim et al. (1993).  During this preparation a small decrease in the  O  enrichment was observed, which presumably occurred during the phosphytilation reaction. Compound 12 was fully deprotected by hydrogenolysis over Pd/C catalyst followed by 18  standard deacetylation with sodium methoxide to yield the desired 1- O-labeled a glycopyranosyl phosphate (13). The  final  product,  disodium  uridine-5'-(2"-acetamido-2"-deoxy-[l"-  0]-ct-D-  glucopyranosyl-diphosphate (10), was produced by the coupling of the pyridinium salt of 13 and commercially available UMP-morpholidate following the modified procedure of Roseman 1  et al. as outlined in the general protocol in Chapter Two.  The  R  O label was found to be  incorporated in 81% isotopic abundance as judged by +LSIMS mass spectrometry.  Further  evidence of its location and extent of 0 enrichment were determined when analyzing the high 1 8  75 resolution P N M R spectra which was acquired and processed in the same fashion as described 31  for the PIX experiment (vide infra).  3.4  PIX experiment under reversible conditions In section 3.2 it was described how a PIX experiment could test for the transient  cleavage of the anomeric C - O bond of UDP-GlcNAc during the epimerization reaction. A first 18  experiment was devised in which the  O-labeled substrate was subjected to enzymatic  epimerization for extended periods under reversible conditions. The extent of the reaction was monitored by observing the change in the coupling pattern of the anomeric proton, H-l". As explained in chapter two, a complete exchange of solvent derived deuterium at C-2" will cause the coupling pattern of H - l " to collapse from a doublet of doublet to a clean doublet. The substrate was incubated for much more than enough time (approximately 30 times longer) to produce a complete deuterium exchange at C-2". This ensured that each molecule of substrate had beeen handled by the enzyme multiple times. The P N M R spectra of the equilibrating 31  mixture was then acquired to evaluate the occurrence and extent of the PIX. In order to obtain the high resolution required to observe the small differences of 31  chemical shift in the  •  P N M R spectra (~ 0.01 ppm), it was necessary to properly set up the  values of the acquisition parameters of the spectrometer. This consisted of reducing the sweep width within the region of interest with concomitant extension of the acquisition time (see experimental section for details). Additional mathematical manipulation when processing the FID helped to detect small variations in the chemical shift. One other factor that had to be considered was the presence of divalent ions that can produce an unwanted broadening of the 31  P N M R signal. It was necessary to eliminate these ions from the solution to be analyzed.  This has been accomplished in other experiments by using a chelating agent such as E D T A . In this particular experiment, a chelating resin (Chelex-100) was employed and proved to be  76 efficient in increasing the resolution of the P N M R spectra. Small amounts of resin were 31  introduced directly into the N M R tube containing the reaction prior to the acquisition of the N M R spectra. It was not necessary to remove the resin from the N M R tube since it completely settled in the bottom of the N M R tube producing no detectable interference with the acquisition of the spectra. The resulting high-resolution-proton-decoupled  31  P N M R spectra of the reaction  mixture before and after equilibration with the enzyme are shown in Figure 3.11.  Only the  region of the spectra that corresponds to the p-phosphate of the sugar nucleotide (-12.4 to -14.0 ppm) is shown. The a-phosphate signals that arise at lower field (-10.5 to -12.0 ppm) show no significant change during the experiment.  Furthermore, all peaks appears as doublets due to  coupling to the adjacent phosphorous nuclei (Jp.p = 21.0 Hz). The P N M R spectra of the initial reaction mixture (Fig 3.11, A) showed two doublets 31  18  in the p-phosphate region, a major one (a) corresponding to the  O-labeled substrate and a  minor one (n), slightly downfield, corresponding to unlabeled substrate.  The difference in  chemical shift between these two doublets is 0.013 ppm that agrees with that expected for a P1 8  0 single bond arising from the presence of the 0 isotope in a bridging position. In addition, 1 8  the integrals of the signals are consistent with those expected for an 81% 0 incorporation. 1 8  After the epimerization was carried out, a new doublet (b) arose in the  P N M R signals  of the UDP-GlcNAc p-phosphate present in the equilibrating mixture (Figure 3.11, B). This new doublet confirmed the occurrence of a PIX. It was shifted 0.029 ppm upfield from that of 18  the unlabeled substrate signal (n) and can be assigned to a torsional isomer with a P- O bond order greater than one. This indicates that the  1 8  0 isotope is now in a nonbridging position of  the phosphate moiety. Furthermore, the integration of the upfield signals (a and b) showed the expected 1:2 ratio which reflects the statistical distribution of the label between the nonbridging (a) and bridging (b) positions.  77  L J  LJ  A.  i—i—i—i—i—|—i—i—i—i—|—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i -12.4 -12.6 12.8 13.0 p 13.2 13.4 13.6 13.8 p  m  i i i i i r14.0  Figure 3.11 P spectra (D 0) of the p-phosphorus region of 0 labeled UDP-GlcNAc. A : before enzymatic epimerization. B: after enzymatic epimerization. Peaks appears as doublets due to the coupling to the adjacent a-phosphorus. A prime indicates signals corresponding to UDP-ManNAc present in the equilibrating mixture after epimerization. n = non-labeled sugar nucleotide, a = sugar nucleotide with 0 label at the bridging position, b = sugar nucleotide with 0 label at nonbridging position. 31  18  2  18  18  Finally, in addition to the signals arising from UDP-GlcNAc, a new series of three doublets had appeared at about one ppm upfield. These new signals can be assigned to the pphosphate of the product of the epimerization reaction, UDP-ManNAc. The relative ratios of these new doublets paralleled those observed for UDP-GlcNAc and were also indicative of full  78 scrambling of the O label between the bridging (a') an non bridging (b') in the formed l s  product. Furthermore, the overall ratio of the integration of the three UDP-GlcNAc doublets to that of the three UDP-ManNAc doublets was approximately 10:1, the expected ratio for the enzymatic epimerization under equilibrating conditions.  3.5  PIX experiment under irreversible conditions The experiment described above successfully demonstrated the PIX for the epimerase  under reversible conditions. This experiment, however, did not provide information on the possibility that only partial torsional equilibration may occur during a single catalytic event. There have been reports of enzymatic reactions in which a bound intermediate capable of exhibiting PIX is formed, normally phosphate or pyrophosphate, yet scrambling is not observed (Mash et al., (1981) for farnesyl pyrophosphate synthetase; Croteau et al. (1985) for geranyl pyrophosphate synthetase  Hilscher et al. (1985) and Ghose and Raushel (1985) for  arginosuccinate synthetase and Mizrahi et al. (1985) for D N A polymerase I).  This  phenomenon is believed to occur due to restricted bond rotation of the phosphate (or pyrophosphate) moiety due to complexation of the oxygen anions by a metal (normally M g ) +2  and/or protein side chains.  Thus, the statistical distribution observed in the experiment  described above could have required that each substrate molecule underwent multiple catalytic events before equilibrium was reached. A experiment was devised in order to assess the possibility of a partial PIX occurring under non-equilibrating conditions.  The experiment relies on the irreversible enzymatic  transformation of the product of the enzymatic epimerization (UDP-ManNAc) to UDPManNAcUA by the enzyme UDP-ManNAc dehydrogenase. The epimerization reaction was performed under "coupled" conditions (see section 2.5) to ensure that the conversion of UDPManNAc to UDP-ManNAcUA was rapid relative to its rate of formation. The reaction was  79 .stopped before completion and analyzed for occurrence and extent of PIX by  3 1  P NMR  spectroscopy. The substrate used for this particular experiment was 64% enriched with O at l s  the anomeric position as judged by both, +LSIMS mass spectrometry and by high resolution 31  P N M R spectroscopy (see Figure 3.12 A).  The enzymatic epimerization was performed  under the same conditions (37 °C, 50 mM Tris-HCl buffer pH 8.8, 2 mM DTT) used for the determination of the kinetic constants of the epimerase (see section 2.5). A calculated amount of epimerase was added so that it would take approximately 120 min to epimerize 30% of the substrate. In order to remain coupled it was necessary to perform the reaction with a sufficient amount of purified UDP-GlcNAc dehydrogenase and N A D in the reaction mixture. In contrast to the PIX experiment under reversible conditions, the enzymatic epimerization was performed in the absence of chelating agents (Chelex-100 resin).  Chelating agents could promote an  increase in the PIX extent by sequestering divalent ions that could restrict the torsional equilibration of the (3-phosphoryl oxygens of the enzyme-bound UDP intermediate. After the determined time had elapsed, the enzyme activity in the reaction mixture was terminated by drastically decreasing the pH of the solution to levels in which the enzyme was no longer active (pH ~ 2). The reaction mixture was then prepared for analysis of the PIX by 31  P N M R spectroscopy by separating the enzyme from the reaction mixture using membrane  filtration. The resulting protein-free solution was neutralized with pyridine and lyophilized to afford a solid residue that was reconstituted in deuterated phosphate buffer. As in the former experiment, it was necessary to modify the acquisition and processing parameters in order to 31  obtain the desired resolution in the N M R spectrum.  31  P N M R spectrum. Figure 3.12 B shows the resulting  P  80  A  B  -|—i—i—i—|—i—i—i—i—|—i—i—i—i—| -12.4  -12.6  i i i i | i i i i | i i i i | i i i i | i i i i | i i i i | i  -12.8  -13.0  -13.2  -13.4  -13.6  -13.8  r  -14.0  ppm Figure 3.12 P spectra (D 0) of the p-phosphorus region of 0 labeled UDP-GlcNAc. A : before irreversible enzymatic epimerization. B : after epimerization. A prime indicates signals corresponding to UDP-ManNAcUA present in the reaction mixture after epimerization. n = non-labeled sugar nucleotide, a = sugar nucleotide with 0 label at the bridging position, b = sugar nucleotide with 0 label at the nonbridging position. 3 1  1 8  2  1 8  1 8  The extent of the enzymatic reaction (-30%) was confirmed by evaluating the integrals of the set of signals for the starting material (UDP-GlcNAc, -12.5 to -13.0 ppm) with the signals corresponding to the product (UDP-ManNAcUA, -13.6 to -14.0 ppm). The pattern of  81 the signals is consistent with complete PIX under these conditions.  Both sets of signals  corresponding to the substrate and the final product showed the expected statistical distribution on 30% turnover conditions consistent with a complete torsional equilibration of the bridging to nonbridging oxygens of the p-phosphoryl moiety. This is most readily apparent in the signals of the product, UDP-ManNAcUA a 1:2 distribution between the bridging (a') and the nonbridging (b') positions are clearly seen.  The signals corresponding to UDP-GlcNAc  showed a pattern consistent with torsional equilibration of only about 30% of the labeled 1  material to give the expected statistical distribution for nonbridging position.  R  O label in the bridging and  This also suggests that the enzyme-bound intermediates partition  forward towards product formation and backward towards substrate regeneration, occur at approximately equal rates.  3.6  Mechanistic implications The observation of isotopic scrambling in the labeled substrate strongly supports the  notion that the epimerization reaction catalyzed by UDP-GlcNAc 2-epimerase proceeds with cleavage of the anomeric C - 0 bond of UDP-GlcNAc to form a enzyme bound UDP as one of the intermediates.  The additional finding that the extent of the scrambling is complete even  under irreversible conditions indicates the absence of a restriction on the motion of the generated UDP intermediate by metals and/or protein side chains. The findings in these experiments could be used to rule out two of the proposed mechanisms postulated for the mode of action of UDP-GlcNAc 2-epimerase: Path A and Path C.  Both proceed without cleavage of the anomeric C - 0 bond.  The two other proposed  mechanisms that are consistent with the findings of this experiment differ in the need of a cofactor (NAD ) necessary to transiently oxidize the C-3" hydroxyl to a ketone (Path D) and in +  82 the structure of the possible intermediate (either a 2-acetamidoglucal, 3 or a 3-keto-2acetamidoglucal, 5). These will be dealt with more detail in the next chapter.  3.7  Experimental Methods  3.7.1  General 95% enriched  1 8  0 water was purchased from Cambridge Isotope Laboratories,  Massachusetts. CH3CN was dried by distillation over Catb and stored over 3A molecular sieves. The ratio of labeled : unlabeled compounds was calculated by comparing the M+H  +  integrals on the +DCI (NH3) or +LSI (for polar compounds) mass spectra. 2-Methyl-(3,4,6-tri0-acetyl-l,2-dideoxy-a-D-glucopyrano)-[2, l-d]-oxazoline commercially  available  (14)  was  prepared  from  2-acetamido-l,3,4,6-tetra-0-acetyl-2-deoxy-p-D-glucopyranose  (Aldrich Co.) by the method of Nakabayashi et al. (1986). 2-Acetamido 3,4,6-tri-0-acetyl-2deoxy-ct-D-glucopyranose (11) was prepared by the method of Avalos et. al. (1993).  C NMR  proton-decoupled spectra were recorded at 125.77 MHz (Bruker-AMX-500). Spectra acquired in CDCI3 were referenced to the solvent peak (8 = 77.00 ppm). Desorption chemical ionization (DCI) mass spectra were recorded on a Delsi Nermag RIO-10C mass spectrometer using NH3 as the chemical ionization gas. High resolution liquid secondary ion (HRLSI) mass spectra were recorded on a Kratos Concept II HQ mass spectrometer  3.7.2  Attempted synthesis of 2-Acetamido-3,4,6-tri-0-acetyl-2-deoxy-l- O-a-Dglucopyranose (11a) by acid-catalyzed hydrolysis of 2-methyl-(3,4,6-tri-0-acetyl1,2-dideoxy-a-D-glucopyrano)-[2,1 -d]-oxazoline (14) 18  To a solution of GlcNAc oxazoline 14 (840 mg, 2.55 mmol) in 30 mL of dry 1 ft  acetonitrile was added p-toluensulfonic acid (49 mg, 0.25 mmol) and 95%  O water (204 uL,  10.20 mmol). A small amount of a white precipitate was formed. The suspension was stirred and refluxed for 6 h and then filtered and concentrated to a dark syrup. This was redissolved in  83 a small amount of EtOAc and purified by passing through a small SiC»2 column (10 g) eluted with a 1:1 mixture of acetone : hexanes. After evaporation of the relevant fractions, 540 mg (60 %) of a slightly yellow, hygroscopic foam was obtained which had H N M R (CDC1 ) ]  3  identical to that reported in the literature (Sim et al, 1993):  l 3  C N M R (CDC1 ) 8 171.33, 3  170.82, 170.46 (broad, resolved into two signals at higher resolution: 170.49 and  170.46),  169.33 (C=0, 4 Ac), 91.52 (C-l) 71.02 (C-5), 68.45 (C-3), 67.55 (C-4), 62.18 (C-6), 52.43 (C2), 22.99, 20.63, 20.62, 20.49 (CH , 4 Ac). +DCIMS (NH ) m/z 350 [M ( 0) + H , 60.04], 348 18  3  +  3  [M ( 0) + H , 14.69), 332 [M ( 0) +H - H 0 , 40.76], 330 [M ( 0) + H - H 16  +  18  +  18  +  2  1 8 2  0, M  ( 0 ) + H - H 0 , 21.53]. 16  +  2  3.7.3  Preparation of (11a)  2-Acetamido-3,4,6-tri-0-acetyl-2-deoxy-[l- O]-cx-D-glucopyranose 18  2-Acetamido-3,4,6-tri-0-acetyl-2-deoxy-l-a-D-glucopyranose (11; 620 mg, 1.8 mmol) was dissolved in a mixture containing 0.4 mL of 95% enriched  1 8  0 water and 1.0 mL dry  C H C N . The solution was placed in a sealed vial, and heated for 17 h at 90-95 °C. The solvent 3  was removed by distillation in vacuo and the dark brown syrup was chromatographed on a small bed of silica (Merck 230-400) using acetone:hexane (1:1) as eluent. Evaporation of the solvent of the relevant fractions gave 11a (445 mg, 71%) with 89.9% 1 3  1 8  0 enrichment. *H and  C N M R spectra were similar to those reported previously. +DCIMS (NH ) m/z 350 [M ( 0) 18  3  + H , 100], 348 [M ( 0 ) + H , 11.3], 330 [M ( 0) + H - H +  3.7.4  l6  +  18  +  1 8 2  0 , M ( 0)+H - H 0 , 71.3] 16  +  2  Preparation of Disodium Uridine-5'-(2"-Acetamido-2"-deoxy-[l"- 0]-a-Dglucopyranosyl-diphosphate), O-labeIed UDP-GlcNAc (10). 18  ls  1R  2-Acetamido-3,4,6-tri-0-acetyl-2-deoxy-[l-  0]-a-D-glucopyranose  was  dibenzyl-  phosphorylated using the method of Sim et al. (1993). The resulting 2-acetamido-3,4,6-tri-018  acetyl-2-deoxy-[l- 0]-a-D-glucopyranosyl  dibenzylphosphate was fully  deprotected by  84 standard protocols and subjected to uridine monomorpholidate coupling as explained in chapter two to give the desired compound with 81% 0 label at the anomeric position. The 'H and C 18  N M R spectra of the final  18  1 3  0 labeled UDP-GlcNAc were identical to those of the commercial  unlabeled product. The location and extent of the 0 label was corroborated by short sweep l 8  width experiments on P N M R spectroscopy. +LSIMS (thioglycerol) m/z 654 [M ( O) + 31  ls  H ,100], 652 [M ( 0)+ H , 23.3); P N M R (D 0) 8 -12.726 (d, J . = 21.0 Hz, 0.2P, 0-P- O), +  16  +  31  18  2  P  P  -12.739 (d, J . P = 21.0.Hz, 0.8P, p-P- 0), -11.014 (d, J . = 21.0 Hz, IP, a-P). 18  P  3.7.5  P  P  PIX experiment under reversible conditions. A solution of 0-labeled UDP-GlcNAc (450 uL, 17 mM, 81% 18  18  0 incorporation) in  deuterated potassium phosphate buffer (50 m M pD 8.0-8.2) was prepared. The sample was placed in a 5 mm N M R tube and Chelex-100 resin was added (20 mg of 200-400 mesh, N a form, previously rinsed with D 0), and both *H and P N M R spectra were collected.  +  A  31  2  solution of UDP-GlcNAc 2-epimerase (1.3 units in 200 pL of the same buffer containing 20 mM of dithiothreitol) was added and the solution was incubated for 8 h at 25 °C. *H and P 31  N M R spectra were collected after this time. T1  The proton-decoupled  P NMR  spectra were obtained on a Bruker AC-200E  multinuclear spectrometer operating at a frequency of 81.015 MHz. All spectra were externally referenced to H3PO4 (0.0 ppm). Typical acquisition parameters were 1620 Hz (20 ppm) sweep width, 10.125 sec acquisition time, 0.4 sec delay between pulses and 3 psec pulse width. Well resolved spectra were obtained after 100 or more scans. Exponential/Gaussian resolution enhancement function.  The FID were apodized using an  85 3.7.6  PIX experiment under irreversible conditions. A solution of 0 -labeled UDP-GlcNAc (550 pX, 17.8 mM, 64% 18  18  0 incorporation) in  50 mM deuterated Tris-HCl buffer (pD 8.6-8.8) was prepared. The sample was placed in a 5 mm N M R tube and Chelex-100 resin was added (20 mg of 200-400 mesh, N a form, previously +  rinsed with D 0), and its P N M R spectrum was collected The solution from the N M R tube 31  2  was diluted to a total volume of 6.12 mL in 50 mM Tris-HCl buffer (pH 8.8) containing 2 mM of dithiothreitol and 4 mM N A D  +  (final concentration of UDP-GlcNAc = 1.60 mM).  The  solution was placed in a 37 °C bath and solutions of UDP-ManNAc dehydrogenase (5.27 mg, 310 pL) and UDP-GlcNAc 2-epimerase (2.95 x 10" units, 45 uX) in the same buffer were 2  added. The reaction mixture was shaken gently at 37 °C for 120 min after which the enzyme activity was suppressed by adding 1 N HC1 until the pH of the reaction mixture was 2. The clear  suspension was  concentrator.  quickly  concentrated  using Ultrafree-15  membrane  The eluted solution was lyophilized and reconstituted in deuterated potassium  phosphate buffer (50 mM pD 8.0-8.2) and submitted for above.  (Millipore)  P N M R spectroscopy as detailed  Chapter Four Mechanistic Studies on UDP-GlcNAc 2-Epimerase with Alternative Substrates  87 4.1  Introduction The PIX experiments described in Chapter Three gave important information about the  possible mechanism of action followed by the enzyme UDP-GlcNAc 2-epimerase.  The  occurrence of a PIX supports a mechanism in which the anomeric C - O bond of UDP-GlcNAc is broken during the enzymatic epimerization generating UDP as one of the enzyme bound intermediates. Among the possible mechanisms postulated in chapter one for the mode of action of the epimerase (section 1.5), only two (Path B and Path D) are in accord with the experimental outcome observed in the PIX experiments. Path B involves the cleavage of the anomeric C - 0 bond via a B-elimination process with the formation of 2-acetamidoglucal and UDP as intermediates. Path D postulates the transient oxidation at C-3" prior to the formation of UDP and a 3-keto-2-acetamidoglucal intermediate. At the same time the work described in this thesis was being carried out, Dr. Paul Morgan in our laboratory was studying the ability of the epimerase to catalyze the formation of either UDP-GlcNAc  or UDP-ManNAc from the putative  intermediates  UDP and 2-  acetamidoglucal. The outcome of the attempted enzymatic condensation was evaluated using an ion-paired reversed-phase HPLC that allows the separation of the sugar nucleotide epimers and UDP. Conveniently, the conjugated Af-acylenamine functionality of 2-acetamidoglucal has a detectable absorbance at the same wavelength (262 nm) as the sugar nucleotides absorb, allowing the observation of all the species present in the reaction mixture by this method. The experiment was designed to test whether  14  C - U D P would be incorporated into a pool of the  equilibrating sugar nucleotides in the presence of added glucal. During the analysis of the incubation mixtures it was observed that the intensity of the UDP and 2-acetamidoglucal peaks actually increased after long incubation periods. It was also noted that in control samples  88 containing only UDP-GlcNAc and enzyme (no 2-acetamidoglucal present), peaks attributable to 2-acetamidoglucal  and UDP appeared after long incubation periods.  Based on these  observations, it was postulated that these putative intermediates were actually being formed during the epimerization and were being released into solution. Indeed, when UDP-GlcNAc was incubated with large amounts of the epimerase for longer periods of time, it was completely converted to UDP and 2-acetamidoglucal. enzymatically-formed  2-acetamidoglucal  Furthermore, the identity of the  was confirmed by isolating the material  comparing its *H N M R spectrum to that of synthetic 2-acetamidoglucal.  and  This experimental  evidence that 2-acetamidoglucal and UDP were accumulating in the solution during long incubation periods with the epimerase, strongly suggests that they are intermediates of the enzymatic epimerization and are occasionally released from the active site of the enzyme as portrayed in Figure 4.1. Additionally, the accumulation of these intermediates implies that the free intermediates are thermodynamically more stable than the equilibrating mixture of UDP-GlcNAc and UDP13  ManNAc (when free in solution). This preference explains why significantly  incorporated  acetamidoglucal  since  the  into  the  sugar nucleotide  coupling of  thermodynamically unfavorable.  these  pool  intermediates  C-labeled UDP was not when  incubated  would  have  with  been  2-  very  89  UDP-ManNAc  UDP-GlcNAc  ENZ  ENZ  «2  AcNH  HO HO HO  HQ HO' HO AcHN  O  AcNH  0~  HO HO HO  P  -r,  60. o  O  O" '  O.  O  ~  P o~ 0.^0.^0. II  o  O  II  O ENZ  ENZ  HO O"  HO HO  +  AcNH 2-acetamidoglucal  II o  II o UDP  Figure 4.1 Formation of intermediates in the reaction catalyzed by UDP-GlcNAc 2-epimerase. Species in brackets are enzyme bound. U = Uridine  Although the evidence presented so far is consistent with the proposed path B mechanism for the enzymatic epimerization, there is still the possibility that the enzyme employs an alternate mechanism that could account for these experimental results. One can envision a mechanism such as that postulated in Path D, in which the accumulation of 2acetamidoglucal after extended incubation could be explained by the reduction of the 3-keto-2acetamidoglucal intermediate prior to the readdition of the UDP moiety. This 'mistake' of the enzyme could account for the generation of the 2-acetamidoglucal intermediate that diffuses out of the active site of the enzyme and accumulates in solution (Figure 4.2) The enzymatic epimerization  of sugar nucleotides via the formation of keto-  intermediates is a known strategy employed by several enzymes (UDP-galactose 4-epimerase, UDP-7V-acetylglucosamine 4 epimerase, UDP-xylose 4-epimerase, GDP-D-mannose epimerase,  90 etc.) that operate at non-acidic stereocenters. These enzymes commonly use a N A D cofactor +  to effect the epimerization by two possible routes. One method is the direct oxidation of the inversion stereocenter to a ketone followed by reduction via hydride transfer to the opposite face to generate the inverted epimer (i.e. UDP-galactose 4-epimerase).  The second route  involves formation of a transient ketone from a hydroxyl group adjacent to the inversion center which thus becomes more acidic and prone to epimerization by a deprotonation/reprotonation mechanism (i.e. GDP-D-mannose epimerase).  Figure 4.2 Alternative pathway for the formation of 2-acetamidoglucal through Path D: (a) normal enzymatic process, (b) early reduction of 3-keto-intermediate leading to 2acetamidoglucal formation.  91 In order to test for the possible formation of a 3-keto 2-acetamidoglucal intermediate (Path D), a substrate analog lacking a hydroxyl group at C-3" was synthesized. This analog, 3"deoxy UDP-GlcNAc (17), was used as a probe to test if a transient oxidation at this center is necessary to perform the epimerization. Experiments using this substrate are described in the next section. The possible occurrence of neighboring group participation by the 2-acetamido group during the enzymatic epimerization will also be addressed in this chapter. This possibility will be probed by testing the oxazoline of ManNAc (18), as an intermediate in a possible stepwise formation of 2-acetamidoglucal. Furthermore, studies on the enzymatic epimerization with a trifluoroacetamido analog of UDP-GlcNAc, UDP-GlcNAc-F (19), will be presented. Finally 3  the utility of compound 19 in replacing UDP-GlcNAc in other enzymatic processes will be presented at the end of the chapter.  4.2  Studies with the 3"-deoxy analog of UDP-GlcNAc 3"-Deoxy-UDP-GlcNAc (17), was used to investigate the possibility that a transient  oxidation at the C-3" center is involved in the epimerization reaction. If this substrate analog is catalytically competent, one could rule out any mechanism that requires formation of a 3"-ketointermediate (Paths A and D).  92 4.2.1  Synthesis of the 3"-deoxy analog of UDP-GlcNAc  The synthesis of 3"-deoxy UDP-GlcNAc (17) (Figure 4.3) was achieved via a slight modification of the preparation reported by Srivastava et al. (1990). The known benzyl 2acetamido-2-deoxy-4,6-0-isopropylidene-a-D-glucopyranoside 20 (Hecker and Minich, 1990)  17  Figure 4.3 Synthesis of 3"-deoxy UDP-GlcNAc: (i) PhOCSCl, 4-(dimethylamino) pyridine, CH CN, r.t; (ii) BuSnH, AIBN, PhCH , 80 °C; (iii) CH COOH-H 0, 40 °C; (iv) Ac 0, pyridine, r.t.; (v) H , EtOH, 5% Pd-C, r.t.; (vi) Et NP(OBn) , 1,2,4-triazol, CH C1 , r.t.; H 0 , THF, -78 °C; (vii) MeONa, MeOH, r.t.; H , MeOH, 5% Pd-C, r.t.; (viii) UMP-morpholidate, pyridine, r.t. 3  3  3  2  2  2  3  2  2  2  2  2  2  2  93 was used to alleviate solubility problems observed with the 4,6-0-benzylidene derivative. The free hydroxyl group at C-3 was thioacetylated with phenylchlorothionocarbonate to produce 21. Radical-induced reductive cleavage of 21 with tributyltin hydride provided the 3-deoxyglycoside 22. Subsequent acid-catalyzed hydrolysis of the isopropylidene group followed by Oacetylation and hydrogenolysis of the anomeric benzyl group provided the 3-deoxy reducing sugar 23.  The phosphytilation/oxidation method of Sim et al. (1993), rather than the reported  method using dibenzyl phosphorochloridate (Srivastava et al,  1990), was used for the  phosphorylation of 23 due to the availability of the reagents. This gave the dibenzylphosphate 3-deoxy derivative 24 that was fully deprotected to produce the corresponding 3-deoxy sugar phosphate 25  and then converted to 3"-deoxy-UDP-GlcNAc (17)  by standard UMP-  morpholidate coupling.  4.2.2  Attempted epimerase  epimerization  of  3"-deoxy-UDP-GlcNAc  with  UDP-GlcNAc  2-  3"-Deoxy-UDP-GlcNAc was initially tested as an alternate substrate by incubation with UDP-GlcNAc 2-epimerase in a deuterated buffer.  The reaction was followed by 'H and P 31  N M R spectroscopy. A careful analysis of the N M R spectra after 12 h showed no evidence of epimerization. No signals, which could be attributed to the formation of a new epimer, were present.  Moreover, solvent-derived deuterium incorporation at C-2" was not observed as  evidenced by the absence of any change in the coupling pattern of H - l " in the H N M R !  spectrum. There were three questions considered relevant in finding a rationale for the lack of epimerization of this analog. First, since the enzyme appears to have an activator site (reflected in its allosteric behavior), was the 3 "-deoxy analog an activator of the epimerase, but not a substrate? Second, if it was not an activator, was it a substrate in the presence of an activator  94 (UDP-GlcNAc)?  Finally, was the 3"-deoxy analog even binding to the active site of the  enzyme, albeit not being epimerized? The first question was addressed by testing the 3"-deoxy analog as an 'activator' in the catalysis of the epimerization of UDP-ManNAc in a similar fashion as described in section 2.6.3. This possibility was tested by incubating the enzyme in a buffer solution containing 3"deoxy-UDP-GlcNAc and UDP-ManNAc followed by analysis of the extent of the epimerization of UDP-ManNAc by ion-paired reversed-phase HPLC chromatography (Figure 4.4).  0.24 - t = 0 m i n 0.20 \ CM CD CN  CO .Q <  (f~^  >  UDP-ManNAc 3"-deoxy analog  0.16 ~  v  0.12 0.08 0.04 0.0 0  ' I i I ' I ' I ' I ' I i I ' I ' I i I ' I ' I ' 2 4 6 8 10 12  4  6  8  10  T i m e (min)  T i m e (min)  B 0.16 H  0 . 2 0 H t = 0min  |  0.16^  UDP-ManNAc  CN  a ui .o <  I UDP-GlcNAc  {I  o.i2H  0.08 H  t = 10 m i n  0.12  CN CD CN  @  0.08-  tri  —  <  0.04-  0.04-1 0.0 0.0  I I I 2  I  M I ' I ' I 4 6  I  I ' I ' I ' I ' I ' I 8 10 12  T i m e (min)  0  i l i l i l i l i l • l > I i l i i i i i i i i 2 4 6 8 10 12 T i m e (min)  Figure 4.4 Ion-paired reversed-phase HPLC traces of a sample containing : (A) UDPManNAc (1.1 mM) and 3"-deoxy-UDP-GlcNAc (0.35 mM); (B) UDP-ManNAc (1.1 mM) and UDP-GlcNAc (0.35 mM) before and after incubation with the enzyme.  95 No detectable traces of UDP-GlcNAc or 3"-deoxy-UDP-ManNAc were observed (Figure 4.4 A). A control reaction containing UDP-GlcNAc instead of the 3"-deoxy compound showed approximately 30% conversion of the UDP-ManNAc under identical conditions (Figure 4.4 B). This experiment showed that the removal of the hydroxyl group at C-3" renders this compound unable to act as an activator of the epimerase. It is quite possible that 3"-deoxy-UDP-GlcNAc acts as a substrate only in the presence of the activator/substrate UDP-GlcNAc.  The test for the epimerization of 3"-deoxy-UDP-  GlcNAc was therefore repeated in the presence of saturating (8 mM) and sub-saturating (0.5 mM) amounts of UDP-GlcNAc as an activator of the epimerase. followed by ' H N M R spectroscopy.  The reaction was again,  The regions of interest for the analysis of the  epimerization are shown in the initial spectra of 3"-deoxy-UDP-GlcNAc (before addition of enzyme or activator) in Figure 4.5 A . One of these regions (around 7.9 ppm) shows the signal that arises due to one of the vinylic protons (H-6) of the uridine moiety.  The other region  (centered on 5.4 ppm), correspond to the anomeric proton (H-l") of the pyranose ring of the sugar nucleotide. After incubation with the epimerase, the H N M R spectrum of the resulting mixture !  showed no signs of epimerization of, or deuterium incorporation into, the 3"-deoxy-UDPGlcNAc. This is particularly noticeable by following the H - l " of the 3"-deoxy analog which does not collapse into a doublet (due to deuteration of the C-2") and by the lack of an additional signal at higher magnetic field in the H - l " region attributable to the formation of 3"-deoxyUDP-ManNAc.  The enzyme was clearly active however, since UDP-GlcNAc was rapidly  epimerized/deuterated  and then gradually converted to the intermediates UDP and 2-  acetamidoglucal (Figure 4.5 B and C). This was evidenced by the appearance of a new doublet in the H-6 region attributable to free UDP, the arising of a new signal around 6.7 ppm  96 corresponding to the C - l vinylic proton of 2-acetamidoglucal and the progressive disappearance of the signal corresponding to the H - l " of the UDP-GlcNAc.  H - 6 region  H -1" region 3"-deoxy-UDP-GlcNAc only  1 6 h after addition of enzyme and UDP-GlcNAc added as activator  B  a  9  j  *  ^^^>JL^^  12 h after  Jl i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—r  8.0  7.5  7.0  6.5  6.0  5.5  PP™  Figure 4.5 Region of *H N M R spectra showing H - l " and H-6 of 3-deoxy-UDP-GlcNAc (*) A) Before addition of enzyme or activator. B) After 6 h incubation with the enzyme and activator (a). Formation of intermediates UDP (u) and 2-acetamidoglucal (g) is observed. C) After 12 h incubation.  Additional evidence of the transformation to the free intermediates, in particular UDP, 31  was corroborated by observing the  4.6).  P N M R spectra of the resulting mixture after 12h (Figure  97  3"-deoxy-UDP-GlcNAc  UDP UDP-GlcNAc  J  Figure 4.6 P N M R spectrum of a mixture containing 3"-deoxy-UDP-GlcNAc and UDPGlcNAc after incubation with epimerase 3 1  These results have shown that the 3"-deoxy analog is not a substrate for the enzyme. However,  further  experiments  showed it  does inhibit  the  epimerase  at  millimolar  concentrations, indicating that it may actually bind to the active site of the enzyme. A thorough analysis of the mode of inhibition was not possible due to insufficient amounts of 3"-deoxyUDP-GlcNAc. A preliminary study, assuming that the 3-deoxy analog behaves as a competitive inhibitor, led to a K\ value of 1 mM. This result is consistent with a similar K; value reported for UDP (Kawamura et ai, 1979).  98 4.2.3  Mechanistic implications The information gained by the experiments employing 3"-deoxy-UDP-GlcNAc does not  therefore discount the possible involvement of a transient oxidation at the C-3" center by a cofactor. The fact that the epimerase does not require any exogenous cofactors does not necessarily rule out the involvement of a cofactor in the epimerization process. Some enzymes such as UDP-galactose 4-epimerase (Wilson and Hogness, 1964) and S-adenosylhomocysteine lyase (Palmer and Abeles, 1979), contain a tightly-bound cofactor within the active site that is regenerated after each catalytic cycle and therefore have no need for a external source of the cofactor. Dr. Paul Morgan in our laboratory, investigated the possibility that a tightly bound cofactor, specifically N A D , might be involved in the epimerization process. The epimerase +  was tested for the transient generation of bound N A D H , which should give a clear absorption in the U V spectrum above 310 nm. No evidence for N A D H formation was found by UV-Vis spectroscopy even when large amounts of epimerase were incubated with saturating amounts of UDP-GlcNAc. Additionally, the epimerase was proteolytically digested in order to release any possible tightly bound N A D cofactor. The digested enzyme pool was assayed for the presence +  of N A D using a known protocol (Palmer and Abeles, 1976). +  Again, no evidence for the  presence of a cofactor was found. Finally, the absence of an N A D binding site in the amino +  acid sequence of UDP-GlcNAc 2-epimerase provided indirect evidence against its participation. It has been found that N A D dependent enzymes share a common motif in their amino acid +  sequence responsible for binding the cofactor (Wierenga et ah, 1985, 1986). The amino acid sequence of UDP-GlcNAc 2-epimerase does not show any of the particular features predicted for such an N A D binding motif. +  The absence of a significant epimerization of the 3"-deoxy analog could be due to a non-productive or 'dystopic' binding with the epimerase.  Apparently, the removal of the  99 hydroxyl group at C-3" produces a significant change in the conformation of the substrate analog or affects crucial interactions between the modified sugar nucleotide and the active site of the enzyme.  This experimental outcome is consistent with the reported high substrate  specificity of the epimerase reported by Kawamura et al. (1979).  4.3  Possible involvement of neighboring group participation during the enzymatic epimerization Among the proposed mechanisms for the mode of action of UDP-GlcNAc 2-epimerase  presented in Chapter One, only Path B (elimination/readdition mechanism) is consistent with the experimental results shown so far.  The formation of 2-acetamidoglucal and UDP  intermediates by a B-elimination of UDP from the substrate UDP-GlcNAc followed by the readdition of the UDP moiety represents a minimal description of the mode of action of the epimerase. A further discussion of the type of elimination mechanism ( E l , E2 or Elcb) will be presented in Chapter Five. When considering a more detailed description of this mechanism, there is one intriguing possibility that must be addressed. The presence of an acetamido group at C-2" introduces the possibility of neighboring group participation during the formation of the intermediates. Participation by JV-acetyl functional groups has been observed in a number of non-enzymatic cases such as in the hydrolysis of esters, amides and alkyl halides (Bruice and Benkovic, 1966). Among the reactions of carbohydrates, non-enzymatic glycosylation reactions with 2acetamido-2-deoxy-glycosyl donors under conditions favoring oxocarbenium intermediates (such as 26 in Figure 4.7) are known to proceed with the formation of oxazolinium ion (14a).  100  HNCOCH  3  27  Figure 4.7 Formation of axazolinium ion during glycosylation reactions. M = promotor for a glycosylation rection. X = leaving group.  The formation of these intermediates has been used to explain the high degree of stereoselectivity for the (3 configuration of the products. This oxazolinium ion can lose the N-H proton to afford a stable oxazoline 14 that has also been used as a glycosyl donor to form 1,2trans glycosides, 27.  Alternatively, 14a can eliminate the C-2 proton to afford the 2-  acetamidoglucal, 28 . (Banoub etal, 1992). The formation of oxazoline or oxazolinium ion has also been used to explain the spontaneous hydrolysis of 2-acetamido-2-deoxy p-D-glucopyranosides across an extended pH range (1.5 to 10.5) (Piszkiewics and Bruice, 1967, 1968). This behavior contrasts with the aanomers that exhibited only specific acid and specific base catalyzed hydrolysis. The trans-1,2  101 diaxal disposition between the acetamido functionality and the leaving group was deemed necessary for the occurrence of the spontaneous catalysis. This stereospecific disposition of the acetamido group allowed the formation of oxazolinium ion or an oxazoline through a SNi (intramolecular nucleophilic substitution) process prior to the stereospecific attack of the nucleophile  (H2O).  These SNi processes are feasible for the p-anomers of 2-acetamido-  glucopyranosides and the a-anomers of 2-acetamido-mannopyranosides derivatives. It should be noted however, that while the p-glucopyranosides derivatives must adopt a boat-type conformation in order to achieve the required 1,2 transdiaxal conformation needed for the intramolecular cyclization, the a-mannopyranoside derivatives can effect the cyclization without a significant distortion of its more stable chair-type conformation (Figure 4.8).  OL = Leaving group OL  Figure 4.8 Formation of oxazolinium ions through SNi mechanisms.  In summary, the formation of oxazolinium ions of 2-deoxy-2-acetamido glycosides can be postulated to occur via two scenarios: First, under conditions that promote oxocarbenium  102 ion formation, the oxazolinium ion can be formed in a stepwise process. The pyranose ring in this intermediate adopts a half-chair or 'sofa' conformation that can suffer a nucleophilic attack by the acetamido moiety (Jones and Kosman, 1980). Second, the formation of an oxazolinium ion can occur, in a more stereolectronically favored fashion, through a somewhat concerted mechanism (an SNi process) when the proper trans-1,2 diaxal conformation between the leaving group and the acetamido moiety is initially present. Oxazoline or oxazolinium ion intermediates  have also been postulated in the  mechanism of a few enzymes. Early studies on the mode of action of lyzozyme proposed the involvement of neighboring group participation of the 2-acetamido group of the substrates to account for the retention of configuration of the hydro lyzed products (Lowe et al, 1967; Lowe and Sheppard, 1968).  Later studies on this enzyme however, disregarded this possibility  (Raftery and Rand-Meier, 1968). A more representative case of the involvement of neighboring group participation in an enzymatic reaction was proposed by Yamamoto (1973a, 1973b, 1974a-c) after a series of studies using substrate analogs designed to prove the specificity for the iV-acetamido group of an Af-acetyl-p-D-glucosaminidase. Finally, Knapp et al (1996) have put forward strong evidence for a mechanism involving neighboring group participation and oxazoline formation in the stereospecific hydrolysis catalyzed by jack hexosaminidase (Figure 4.9).  bean N-acetyl  103  \  \  H HO  O-  OR  CH  OH  CH  3  bound intermediate  CH  3  3  O"  O"  Figure 4.9 1996).  H  o-R  cyclized intermediate  bound product  Proposed A^-Acetyl-(3-hexosaminidase mechanism (Adapted from Knapp et al.,  The formation of an oxazolinium intermediate during the epimerization of UDPGlcNAc with UDP-GlcNAc 2-epimerase is a possibility that requires further analysis. The most likely mode of action of the epimerase is the inversion of the stereochemistry at the stereochemistry at C-2" via a p-elimination of UDP. This elimination produces the observed 2acetamidoglucal intermediate which then undergoes a reprotonation on the opposite face and readdition of the UDP moiety to afford the epimeric sugar nucleotide. Since this enzymatic reaction is reversible, the enzyme must be able to perform two different  p-elimination  processes. In the UDP-GlcNAc to UDP-ManNAc direction, the enzyme must promote an antielimination process, while in the reverse direction, the enzyme must generate the intermediates through a ^-elimination (Figure 4.10)  104  UDP  Figure 4.10 Proposed mechanism for UDP-GlcNAc 2-epimerase involving oxazoline-type intermediates.  The neighboring group participation of the 2-acetamido group is feasible during the cw/7-elimination only if this process has strong E l character. In other words, the UDP leaving group must completely depart, and generate an Oxocarbenium ion, prior to the formation of an oxazoline-type intermediate. On the other hand, the ^-elimination process is more likely to involve neighboring group participation of the 2-acetamido group due to the proper trans-\, 2 diaxial conformation between this moiety and the UDP leaving group. Thus, the formation of a manno-type oxazolinium ion intermediate through an SNi process could facilitate the departure of the UDP moiety. Finally, the formation of the 2-acetamidoglucal intermediate during the enzymatic epimerization can be explained by the elimination of the C-2 proton from either of the proposed oxazolinium ions (Figure 4.10). As described earlier, this elimination process has been observed in non-enzymatic cases (Salo and Fletcher, 1969) Two experiments were envisioned in order to test the possibility of neighboring group participation during the epimerization with UDP-GlcNAc 2-epimerase. First, it was decided to  105 prepare, and test a substrate analog in which the acetyl group of the hexosamine part of the substrate was replaced by trifluoroacetyl. It is well known that the replacement of hydrogen atoms by fluorine atoms significantly changes the electronic properties of the group or groups directly attached to the fluorine atoms due to its strong electronegativity. Thus, the fluorination of an acetamido group will significantly increase the electrophilicity of the carbonyl carbon with concomitant reduction in the nucleophilicity of the carbonyl oxygen. The net result of this fluorine substitution is to significantly reduce the ability of the group at C-2" to stabilize the carbonium ion at C - l through neighboring group participation. Second, the preparation of the oxazolines of GlcNAc and ManNAc were considered desirable for testing the possible formation of oxazoline-type intermediates during the enzymatic epimerization. Although, the preparation of the oxazoline of ManNAc (18) was accomplished following a published procedure (Khorlin et al, 1968), attempts to obtain the oxazoline of GlcNAc were unsuccessful. This latter compound is known to be hydrolytically unstable and prone to decomposition during isolation (Knapp et al 1996; Ballardie et al, 1976). Due to the unavailability of this compound it was not possible to evaluate the possibility of its occurrence as an intermediate in the enzymatic epimerization.  4.4  Studies with the trifluoroacetamido analog of UDP-GlcNAc  4.4.1. Introduction As mentioned above, the neighboring group ability of the 2-trifluoroacetamido group in reactions involving formation of a carbonium ion at C - l of 2-amino-2-deoxy-glycosides contrast with the strong participating nature of the 2-acetamido group. Wolffom and Conigliaro (1969) reported that during the synthesis of purine nucleosides of 2-amino-2-deoxysaccharides, the 2-acetamido derivative gave only the p-anomer while the trifluoroacetamido derivative of  106 2-amino-2-deoxy-D-glucose gave a 1:4 mixture of <x:0 anomers. It was postulated that the 2trifluoroacetamido group participated in a slight extent at C - l since a completely nonparticipating nature would have produced a much higher a:p ratio such as the 6:10 ratio produced by the nonparticipating iV-(2,4-dinitrophenyl) group. Another indication of the diminished neighboring group participation ability of the 2trifluoroacetamido group can be found in the studies of an 7V-acetyl-p-D-glucosaminidase reported by Yamamoto (1973b, 1974c). It was concluded that the 2-acetamido group at C-2 provided anchimeric assistance during the enzyme catalyzed hydrolysis of the substrate. A 2trifluoroacetamido analog of the substrate was prepared and found to have a hydrolytic rate 4 orders of magnitude slower than the normal substrate. Although steric effects and hydrophobic factors were analyzed by synthesizing and testing several substrate analogs, the main rational for the drastic drop in the rate of hydrolysis of the 2-trifluoroacetamido analog was considered to be the electronic effect due to the strongly electron withdrawing trifluoromethyl group. It was therefore desirable to prepare the trifluoroacetamido analog of the substrate for the epimerase, namely UDP-GlcNAc-F3 (19).  4.4.2  Synthesis of the trifluoroacetamido analog of UDP-GlcNAc The synthesis of UDP-GlcNAc-F3 (19; Figure 4.11), began with the preparation of the  glycosyl bromide 30 from an a/p mixture of  2-trifluoroacetamido-1,3,4,5-tetra-0-acetyl-2-  deoxy-D-glucose (29a and 29b; Wolfrom and Conigliaro,1969). oxide in 10:1  acetonitrile:water  gave  Treatment of 30 with silver  2-trifluoroacetamido-3,4,6-tri-0-acetyl-2-deoxy-a-D-  glucose 31 in 57% yield. The relatively low yield was due to the formation of the peracetylated glucal 32 and oxazoline 33 as by-products in yields of 7 % and 12 % respectively. These products have been obtained previously upon treatment of 30 with methanol under Koenigs-  107 Knorr conditions (Meyer zu Reckendorf and Wassiliadau-Micheli, 1970). Compound 31 was phosphorylated by treatment with diphenylchlorophosphate and 4-N, N-dimethylaminopyridine in dichloromethane at -10 °C. These conditions are known to produce anomerically enriched aglycosyl phosphate diphenyl esters for a number of acetylated hexopyranoses (Sabesan and Neira, 1992). Indeed the a-diphenylphosphate 34 was obtained in 52% yield and no B-anomer was isolated upon purification of the reaction product by conventional chromatography. A byproduct (~15%>), whose H N M R and mass spectra were consistent with that expected for the l  corresponding  a-glycosyl  chloride  (2-triflouroacetamido-3,4,6-tri-(9-acetyl-2-deoxy-«-D-  glucopyranosyl chloride), was also obtained.  The formation of glycosyl chlorides has been  observed during the diphenylphosphorylation of sugars when w-butyl lithium was used as the base (Hung and Wong, 1996). The phenyl groups of the phosphate triester were removed by hydrogenation over platinum oxide catalyst (Putman, 1963) to give 35.  Low yields (-30%)  were observed, presumably due to acid catalyzed decomposition of the forming glycosyl phosphate.  Cleavage of the O-acetyl groups in the presence of the base-sensitive  trifluoroacetamido group was accomplished by careful treatment with sodium methoxide in methanol to give 36 (68%).  108  OAc  —-OAc  OL  AcO AcO  CF3COHN I  ---OAc  OAc  29a  HBr  A c O ^ \ ^ — - ° \  Br  Q  AcO AcO  M e C N / H  ^  A c O - ^ - ^ * A  —-OAc  0  A c O - ^ ^ - -  30  OAc  AcO^A^°\  32  2  HOAT A CF3COHN CO-^-TS 1  OAc  ---OAc  CF3COHN Ag 0  (  :  C  F  3  C  \  N  29b  r  CF  2  DMAP, -10°C  ^ O H  CF3COHN  NaOMe O 0  _ N _  ~* 0  MeOH  o-  H , 4 Atm A c O ^ Y ^ ^ ° \ AcO-X-^r*-\  AcO^V-^-^°\  2  AcO-^^A  Pt0  CF3COHN j  H  CF COHN 3 J  I  35 pyH  2  O—P—OH  I 3 6  -—OAc  -—-OAc C  ^  (PhO) POCI  3  2) N a O M e , M e O H  HO~"V^--~ \ H O - A ^ A  N  *Jo  33  1) H3PO4, 6 0 ° C  H  31  AcO-J^T"A  CF3COHN  O  u  x  2  34  0-  +  ,1 , OPO(OPh)  Et NH  +  3  UMP-morpholidate pyridine  --OH HO^A—-°\ HO-X^A CF3COHN  o 0  _  L  p  O 0  _ p _  0  i  o19  Figure 4.11 GlcNAc-F  Li+  L-1 '  r\t  OH  1  A I  1  OH  Schematic representation of the synthetic routes for the preparation of UDP-  3  In further work, a more efficient method for the preparation of compound 36 was developed. Treatment of the peracetylated B-Af-trifluoroacetylglucosamine 29b (Wolfrom and Conigliaro, 1969), with neat H3PO4 at 60 °C (the MacDonald procedure; Warren et al., 1973), followed by deacetylation with sodium methoxide in methanol gave a 27% overall yield of 36.  109 The GlcNAc-F3 a-phosphate 36  was coupled to UMP using standard Khorana  conditions with UMP-morpholidate in dry pyridine (Roseman et al., 1961). The product UDPGlcNAc-F3 (19), was isolated as its dilithium salt in 44% yield following purification by anion exchange and size exclusion chromatographies.  4.4.3  Epimerization of the trifluoroacetamido analog of UDP-GlcNAc with UDP-GlcNAc 2-epimerase In order to test UDP-GlcNAc-F3 as an alternative substrate for UDP-GlcNAc 2-  epimerase, a solution (8.9 mM) of this compound in deuterated buffer was incubated with the epimerase (3.8 units) at 37 °C and monitored by H and F N M R spectroscopy. The changes, l  19  over a 19 h incubation, were observed in relevant regions of the 'H and F spectra and are 19  shown in Figure 4.12.  UDP-GlcNAc-F3 underwent similar changes as the normal substrate  when incubated with the epimerase. The deuterated epimer UDP-ManNAc-F and the 3  intermediates, 2-trifluoroacetamidoglucal and free UDP, were all detected. The ' H N M R signal corresponding to the anomeric proton of deuterated UDP-ManNAc-F3 was observed as a doublet centered at 5.52 ppm while the anomeric proton of the glucal gave rise to a singlet at 6.85 ppm. Free UDP was detected by the appearance of a doublet at 7.95 ppm attributed to the H-6 of the heterocyclic base.  110  1  19  H NMR gf  A)  F NMR  gf  t = Oh  ii gf  t=1 h  B)  gf  mf  m  f  t = 2h C)  glycal-F  JL  A. UDP  D)  glycal-F  3  t= 4h  3  TFA  uUL t= 19 h  E)  1  I ' i'  8.4  ii  JL  i' i' i' i' i' i • i' i' i' i' i • i' i • i' i *  8.2 8.0  7.8 7.6  7.4  7.2 7.0 6.8 6.6 6.4 6.2 6.0 ppm  5.8 5.6 5.4  i' i • i • i • i' i' i • i' i • i • i • i • i -72  -73  -74  5.2  -76  -76  -77  -78  PP"!  Figure 4.12 Relevant regions of H and F N M R spectra during the incubation of UDPGlcNAc-F (gf) with epimerase: mf = UDP-ManNAc-F , TFA= C F C O O ~. H N M R spectra A , B and C were recorded at 200 MHz, spectra D and E at 400 MHz. ]  19  !  3  3  3  Ill The value of F N M R spectroscopy for monitoring the development of the reaction is 1 9  evident, unlike ' H N M R , UDP-GICNAC-F3 showed only one signal in its F N M R spectra. The 19  new signals that appear in the  1 9  F N M R spectrum as the reaction proceeds, confirmed the  formation of the compounds detected in the *H N M R spectra. Two distinct features that contrast with the normal substrate were noticed. First, the glycal was formed at a rate comparable to the epimerization reaction. While in the normal substrate the exchange of the H-2" proton with deuterium occurs much faster than the accumulation of the glycal and free UDP intermediates, in the case of the trifluoroacetamido analog these two processes occur at roughly the same rate. These relative rates are judged by comparing the change in the coupling pattern of the signal corresponding to H - l " and the arising of the signals attributable to free UDP and 2-trifluoroacetamidoglucal (Figure 4.13 D). Second, upon extended incubation the 2-trifluoroacetamidoglucal intermediate underwent base catalyzed hydro lytic decomposition.  One of the products detected by  decomposition was trifluoroacetate (CFsCOO ). -  l 9  F N M R in this  This decomposition was independent of the  presence of the enzyme since samples of synthetic 2-trifluoroacetamido glucal incubated with the same buffer and monitored by  l 9  F N M R spectroscopy gave similar results. A tentative  mechanism for this process is depicted in Figure 4.13.  112  HOHO' HO  +  HOHOHO  H 0 2  :0 f  CF  CF  3  3  HO  further decomposition  Figure 4.13  4.4.4  Proposed base-catalyzed hydrolysis of 2-trifluoroacetamidoglucal  Kinetic characterization of the trifluoroacetamido analog of UDP-GlcNAc Once UDP-GI6NAC-F3 was deemed to be an alternate substrate for the epimerase, its  kinetic constants were obtained by using a coupled assay similar to the one used for the normal substrate was performed. The dehydrogenase was first shown to accept UDP-ManNAc-F3 as a substrate, and large amounts of the enzyme UDP-ManNAc dehydrogenase were employed in order to ensure the validity of the coupled assay.  A verification that the coupling conditions  still held for this substrate was that the initial rate of epimerization doubled when the amount of epimerase added was doubled in assay mixtures containing the same amount of UDP-GlcNAcF and dehydrogenase. 3  The initial rate of epimerization was measured against the concentration of the substrate and the resulting data was processed using the program Grafit. The curve that fitted the data most closely was sigmodial (Figure 4.14), suggesting an allosteric behavior for this alternate substrate similar to the one observed for the normal substrate, UDP-GlcNAc. coefficient for this substrate was 1.6 ± 0.2 and the apparent K  m  2  The calculated k  cat>  1  The Hill  (K'), value was 3.40 ± 0.37 mM. 1  2.6 x 10" s" was significantly lower than with the normal substrate 11.3 s" .  113 The kca/K' value for the alternate substrate, 7.6 x 10" mM" s" was also significantly lower 3  1  1  than that from the normal substrate (24.0 mM" s" ). A direct comparison of the values of k , 1  1  ca  between the normal and the alternate substrate showed that the enzyme epimerized UDPGlcNAc-F 430 times slower than the normal substrate UDP-GlcNAc. 3  0  1  2  3  5  4  6  7  [UDP-GlcNAc-F ] (mM) 3  Figure 4.14 A plot of initial velocity of UDP-GlcNAc-F3 epimerization as a function of UDPGlcNAc-F3 concentration  4.4.5  Mechanistic implications Several factors have to be considered when analyzing the enzymatic epimerization of r'  the alternate substrate UDP-GlcNAc-F3. It has been mentioned that the epimerase is highly specific for its substrate. This specificity is not only in the polar UDP moiety region but seems to extend to the pyranose ring part of the substrate since the epimerase will not accept potential substrate analogs such as UDP-GalNAc, UDP-Glc (Kawamura et al, 1979) or 3"-deoxy-UDP-  114 GlcNAc into their corresponding C-2" epimers.  Changing the N-acyl substituent of the  substrate from methyl to trifluoromethyl, is bound to exert some influence in the binding, electronic and steric interactions between the substrate and the enzyme active site since this group is directly bonded to the reaction center (C-2") and must move from an equatorial position to an axial position during the enzymatic epimerization. A previous study examined the effect that replacing an TV-acetamido group had on the rate of hydrolysis by an ./V-acetyl-p-Dglucosaminidase (Yamamoto, 1974c).  He reported the use of van der Waals volumes  (molecular volumes), calculated by the process described by Bondi (1964), in the evaluation of the steric bulkiness of several ./V-acyl substituents.  The Van der Waals volume (V ) for  trifluoromethyl substituent was reported to be 20.49  w  cmVmol while that for the methyl  substituent was 13.67 cm /mol. The difference in size, however, was not the only significant 3  factor in the drastic change of rate for the fluorinated analog (10 times slower) since other 4  analogs with similar or even higher V  w  than the fluorinated one, showed only a 100 fold  decrease in rate. A noticeable difference was that the rate of enzymatic epimerization of UDP-GlcNAcF3 related to the normal substrate (as approximated by the rate at which the alternate substrate exchanges its H-2" proton with deuterium from the media) was comparable to the rate at which the 2-trifluoroacetamidoglucal and UDP intermediates are released from the active site of the enzyme. In contrast to this, it has been estimated that the rate of release of 2-acetamidoglucal occurs approximately once every 400 turnovers during the epimerization of the normal substrate (Morgan, 1998). Thus, with the normal substrate, much longer incubation periods were required to detect signs of free 2-acetamidoglucal or UDP.  This outcome can be  rationalized by postulating that the 2-trifluoracetamidoglucal intermediate does not bind as well  115 as the 2-acetamidoglucal intermediate in the enzyme active site of the enzyme and therefore is more readily released (Figure 4.1) The experiments with UDP-GlcNAc-F3 could not provide definitive evidence in favor for or against the occurrence of neighboring group participation or for the formation of an oxazoline type intermediate. The 400-fold decrease in the rate of epimerization could indicate an impaired neighboring group participation during the syrc-elimination to form the glucal intermediate if the enzyme follows an El type elimination mechanism, although a more drastic decrase in rate was expected if the epimerization reaction occurs with a neighboring group participation. This decrease in the rate may be just a consequence of steric effects and the high specificity of the epimerase. Further discussion of the possible mechanism of the epimerase and the factors involved in its mode of action will be addressed in the next chapter.  4.5  Reaction of the oxazoline of ManNAc and UDP with UDP-GlcNAc 2-epimerase  In an attempt to obtain evidence for the possible formation of oxazoline-type intermediate during the enzymatic epimerization with UDP-GlcNAc 2-epimerase, an experiment that uses the synthetic oxazoline of ManNAc was devised. The basic strategy was to incubate the oxazoline with UDP and the epimerase and observe whether the enzyme is able to catalyze the coupling between UDP and the oxazoline or alternatively, to transform the oxazoline into the known intermediate 2-acetamidoglucal.  The former is perhaps less likely  since it may be thermodynamically unfavorable. As in the experiment described earlier (section 4.1) for the attempted coupling of 2acetamidoglucal and UDP by epimerase, some important factors must be taken into account. First, the epimerase has an absolute requirement for its substrate, UDP-GlcNAc, as an activator. Therefore, a small amount of the substrate or a substrate analog that can activate the enzyme  116 must be present during the experiment.  Second, UDP has been reported to be a competitive  inhibitor of the enzyme with a Kj of 1 mM (Kawamura et al. 1979). This could be detrimental for the possible binding of the oxazoline in the active site of the enzyme since the UDP moiety could bind prior to the oxazoline and possibly restrict its access. An experiment was designed that took all the above factors into consideration. The small amount of UDP-GlcNAc needed to activate the enzyme could be problematic when evaluating the ability of the epimerase to produce 2-acetamidoglucal from the oxazoline since 2-acetamidoglucal is also produced from UDP-GlcNAc. This problem was circumvented by using the substrate analog UDP-GlcNAc-F3 as the activator. The 2-trifluoroacetamidoglucal formed by incubation with the epimerase has unique 'H N M R signals and therefore any 2-acetamidoglucal formed could easily be detected and attributed to a ring opening of the oxazoline of ManNAc. The concentration of UDP was selected such that it would not cause more than 50% inhibition of the epimerase. The concentration of the oxazoline of ManNAc was 10 fold greater than that of either UDP or UDP-GlcNAc-F in order to promote the binding of the oxazoline 3  with the epimerase. Finally, large quantities of epimerase and long incubation times were used to maximize the chance for detection of either 2-acetamidoglucal or UDP-ManNAc that could be formed during the experiment. The actual experiment was run by incubating a solution containing the synthetic oxazoline of ManNAc (10 mM), UDP-GlcNAc-F3 (1 mM) and UDP (ImM) with epimerase (1.90 units).  The sample was incubated at 37 °C and the reaction monitored by *H N M R  spectroscopy. The region of the spectrum between 5.34 and 8.20 ppm was particularly relevant for detecting the formation of either 2-acetamidoglucal or UDP-ManNAc as signals attributable to the anomeric proton of either 2-acetamidoglucal (6.67 ppm; in Figure 4.15) or UDPManNAc (5.42 ppm; um in figure 4.16) arise within this region. Figure 4.15 shows a series of  117 spectra of the sample taken at various times during the incubation. The initial spectrum showed additional signals (marked with *) due to the presence of impurities in the components of the reaction mixture (commercial UDP, UDP-GlcNAc-F3 and oxazoline of ManNAc).  The  presence of these impurities was considered to be innocuous for the experiment since no detectable change in these signals was observed throughout the experiment. The epimerase was obviously active, despite the presence of the competitive inhibitor UDP.  This is demonstrated by the fact that the alternative substrate, U D P - G l c N A c ^ ,  underwent slow enzymatic replacement of the C-2" proton by deuterium (observed in the changes in the coupling pattern of H - l " (gf in Figure 4.15)), as well as epimerization to deuterated UDP-ManNAc-F (mf). 3  The further transformation to the intermediates 2-  trifluoroacetamidoglucal and UDP can be seen by the appearance of the signal corresponding to the anomeric proton of the glucal intermediate (6.84 ppm;) and the increase of the intensity of the signal corresponding to the H-6 proton of the UDP molecule (7.95 ppm).  118  H-6 region  vinyl proton of glycal region  .—*—«,  UDP  gf  H-1" region  t=0 h gf  *  -JUL.  t = 3h  expected position for 2-acetamidoglucal glycal-F  mf  3  t=6 h  t= 18 h  J  J  111111111111111111111111111111111111111111111111111111111111111111111111  8.2  8.0 7.8 7.6 7.4 7.2 7.0 6.8 6.6 6.4 6.2 6.0 5.8 5.6 5.4 ppm  Figure 4.15 *H N M R spectra of a sample containing oxazoline of ManNAc (signals not shown), UDP and UDP-GlcNAc-F3 (gf), after different incubation periods with epimerase. Shown here: H-6 region with signals for UDP and UDP-GlcNAc-F3, vinylic proton region of the glycal intermediates and anomeric protons region (H-l"). Impurities are shown as *. See text for details.  119 Under the conditions of this experiment, the formation of either UDP-ManNAc or UDPGlcNAc (since the experiment was run under equilibrating conditions) was not detected. The possible formation of the intermediate 2-acetamidoglucal was also not observed.  4.5.1  Mechanistic implications The results obtained in the aforementioned experiment do not completely rule out the  absence of neighboring group participation during the ^n-elimination step involved in the enzymatic epimerization of UDP-ManNAc to UDP-GlcNAc. It is possible that the form of the enzyme that normally binds UDP and the oxazoline is in a 'closed' conformation and therefore, free enzyme will not accept the oxazoline and UDP from solution preventing its transformation to 2-acetamidoglucal or the formation of UDP-ManNAc by coupling with UDP.  A more  simple explanation could be that the oxazoline is not an intermediate of the enzymatic epimerization.  This experiment, along with the result obtained with the trifluoroacetamido  analog, suggest (but not prove) that neighboring group participation is not involved during the enzymatic epimerization.  4.6  Use of trifluoroacetamido analog of UDP-GlcNAc in reactions with sugar transferases The utility of preparing the trifluoroacetamido analog of UDP-GlcNAc was evidenced  not only by its use as an alternate substrate for UDP-GlcNAc 2-epimerase but also as an alternate substrate for two different glycosyltransferases involved in the enzymatic synthesis of oligosaccharides. Glycosyltransferases can catalyze the formation of glycosydic bonds between activated sugar nucleotides and a suitable acceptor molecule (Figure 4.16). This work was done in collaboration with Dr. Monica M. Palcic at the University of Alberta and has recently been published (Sala etal, 1998).  120  HO  UDP  +  galactosyl transferase  HO  :  O  0  glycosyl donor  OH OH  OH  UDP-Galactose  Figure 4.16  HO  O  lactose  The reaction catalyzed by galactosyl transferase  The use of transferase enzymes that employ sugar-nucleotides is a promising method for the elaboration of complex natural and unnatural carbohydrates (Wong et al,  1995;  Dreuckhammer et al., 1991). The main advantage to this approach is that the enzymes are highly regiospecific and stereospecific with regard to directing the activated donor carbohydrate to the appropriate hydroxyl group of the acceptor, thereby eliminating the need for multiple protection-deprotection steps required in non-enzymatic syntheses. One of the drawbacks of this strategy, however, is that the specificity of these enzymes precludes the use of alternate donors that differ significantly from the natural donors making the enzymatic synthesis of some unnatural oligosaccharides difficult to achieve. A large class of transferases utilize UDP-GlcNAc as the activated donor to introduce Nacetylglucosamine residues in natural glycoconjugates. One such enzyme considered in this work is the UDP-D-GlcpNAc:|3-D-Galp-(1^3)-a-D-GalpNAc (GlcNAc to GalNAc) B-(l->6)GlcNAc-transferase that is also referred to as the "core-2"-GlcNAc-transferase (GlcNAcT, E C  121 2.4.1.102). This enzyme is involved in the biosynthesis of O-linked glycoproteins in mammals (Oehrlein et al, 1993) transferring a GlcNAc residue to the 0-6 of GalNAc in the disaccharide P-D-Gal/?-(l->3)-a-D-GalpNAc-0R (37; where -OR can be either a hydroxyl residue of a serine or threonine in a protein, aryl or alkyl group) to form the trisaccharide P-D-Galp-(1—»3)-[P-DGlcpNAc (1 ->6)]-a-D-GalpNAc-OR, 38 (Figure 4.17).  38 38a  X = H X = F  Figure 4.17 Reaction catalyzed by "Core-2"-GlcNAc-transferase. R = O H , alkyl, aryl or serine or threonine residue in a protein.  The second transferase that was tested with UDP-GlcNAc-F3 as an alternate substrate was the UDP-D-GlcpNAc:a-D-Man/?-(l->6)-p-D-Glc/7NAc transferase V (GlcNAcT-V, E C 2.4.1.155). This enzyme transfers a GlcNAc residue to the 0-6 of the mannose residue in the trisaccharide p-D-GlcpNAc-(1^2)-a-D-Manp-(l^-6)-p-D-Glcp-OR (R = (CH ) CH ) (39) to 2  form the tetrasaccharide Glcp-OR, 40  7  3  p-D-GlcpNAc-(1^2)-[p-D-GlcpNAc-(1^6)]-a-D-Manp-(1^6)-/?-D-  (Figure 4.18), and controls the branching pattern of asparagine linked  oligosaccharides (Cummings et al, 1982; Schachter, 1986).  122 UDP-GlcNAc-F3 proved to be a good alternate substrate for both transferases. Incubation of (3-D-Gal/?-(l—>3)-a-D-GalpNAc-OR (37),  with a partially purified extract of  'core-2' transferase, and the donor UDP-GlcNAc-F3, resulted in the formation of the trisaccharide product 38a in a 62% isolated yield. The trifluoro analog was also a good donor when incubated with cloned GlcNAcT-V transferase and acceptor 39.  The resultant  tetrasaccharide 40a was obtained in a 56% isolated yield.  HO v H O ^ ^ - - -0  HO—v  H O ^ A — • -O  HOA-—7 HN  Lo^ji^,  =0  > HO^ HO-  HO-J--7  OH  6  0 H  UDP-GlcNAc-X  =0  >  GlcNAcT-V  3  ^ r ° > o OH  Z.O-  \ HN 7—-—i^r— OH OH  ^(ChyrCHa  HO  HO  CX  H  o-^o^/_.  HO^ HO39  Figure 4.18  OH  HN  3  40  X = H  40a  X = F  Reaction catalyzed by GlcNAc-transferase V  One convenient characteristic of the trifluoro acetamido group is its ease of removal under relatively mild conditions (Bergeron et al., 1984; Imazawa and Eckstein, 1979; Newman, 1965). This effectively permits the introduction of glucosamine residues into oligosaccharides that normally contain GlcNAc by using the UDP-GICNAC-F3 as the substrate for transferases instead of UDP-GlcNAc and removing the trifluoroacetamido group once the oligosaccharide has been formed.  This strategy could lead to the preparation of a variety of unnatural  compounds with potential bioactive properties as well as provide a nucleophilic functionality to which other molecules can be appended using reductive amination chemistry.  The direct  123  transfer of UDP-glucosamine was not expected to be feasible because a charged ammonium group was replacing the acetamido functionality of the natural substrate donor. Indeed, it was shown that the "core-2"-GlcNAcT accepts neither UDP-glucose nor UDP-glucosamine as a donor substrate. This indicates that the acetamido moiety serves as a necessary recognition element in the transferase reaction. In addition, the feasibility of this method was demonstrated by the enzymatic synthesis of the tetrasaccharide 40a followed by the removal of the trifluoroacetyl group to generate the glucosamine containing tetrasaccharide 41 (Figure 4.19).  Figure 4.19 Removal of the trifluoroacetamido group in tetrasaccharide 40a.  4.7  Experimental Methods  4.7.1  General Melting points were recorded on a Reichert melting point apparatus and were  uncorrected. Optical rotations were measured with a Perkin-Elmer 241 M C polarimeter. Flash column chromatography was performed using Silica Gel 60 (230-400 mesh, E. Merck, Darmstadt).  2,2'-Azobis(2-methyl-propionitrile) was from Eastman Kodak.  Phenyl  chlorothionoformate and tributylstannane (as 1M solution in THF) were from Aldrich. Benzyl  124  2-acetamido-2-deoxy-4,6-0-isopropylidene-a-D-glucopyranoside  was prepared according to  Hecker and Minich (1990).. Pyridine, CH CN, CH C1 , and NEt were destilled under N from 3  2  2  3  2  CaH . MeOH was distilled under N from magnesium methoxide. All other reagents were 2  2  commercially available or prepared as indicated in the preceding chapters.  All reactions  performed in organic solvents were carried out under argon. *H NMR spectra were recorded at 400 MHz (Bruker WH-400) C NMR spectra were recorded at 75.43 MHz (Varian XL-300) 1 3  "3 1  and assignments are tentative, and  19  P NMR spectra were recorded at 81.0 MHz (Bruker AC-200E)  F NMR at 188.0 MHz (Bruker AC-200E) with proton decoupling. The P chemical 31  shifts are expressed relative to external H3PO4 (0.00 ppm).  The F chemical shifts are 19  expressed relative to CFCI3 for solutions in CDCI3 (0.00 ppm) or 1M C F 3 C O O H for solutions in D 0 (-76.53 ppm). Signals upfield of CFCI3 were assigned negative values. Microanalysis 2  were carried out by Mr. Peter Borda in the microanalytical laboratory at UBC.  4.7.2  Synthesis of 3"-deoxy analog of UDP-GlcNAc (17)  (a)  Synthesis of Benzyl 2-acetamido-4,6-0-isopropylidene-2-deoxy-3-6>-phenoxythiocarbonyl-cc-D-glucopiranoside (21) Benzyl 2-acetamido-2-deoxy-4,6-0-isopropylidene-a-D-glucopiranoside, 20 (8.2 g ,  23.4 mmol) was dissolved in 120 mL of dry acetonitrile. Phenyl chlorothionocarbonate was added (5.6 mL, 40.7 mmol) and the solution was stirred for 17 h at room temperature. The solvent was removed in vacuo and the resulting yellow solid was dissolved in C H 2 C I 2 (300 mL) and washed with ice-cold 0.1M HC1, saturated aqueous NaHC03, and water. The resulting solution was dried over Na S04, the solvent was removed in vacuo and the residue was 2  chromatographed on silica (CH Cl :MeOH 80:1) to give 21 (5.7 g , 50 %), m.p. 160-162 °C, *H 2  2  NMR (CDCI3) 5 7.38-6.95 (10H, m, 2 Ph), 5.82 (d, 1H, J = 10.2 Hz J , = 10.3 Hz, H-3), 4.89 (d, 1H, J 3  4  1 ; 2  2 > N H  = 9.2 Hz, NH), 5.77 (dd, 1H, J ,  = 3.7 Hz, H-l) 4.73 (d, 1H, J  2  g e m  3  = 12.0 Hz,  125 PhC/ffl), 4.49 (d, 1H, PhCtfH), 4.45 (ddd, 1H, H-2) 3.75-3.95 (m, 4H, H-4, H-5, H-6, H-6'), 1.90 (s, 3H, CH3CO), 1.47 (s, 3H, (CH ) C), 1.41 (s, 3H, (CH ) C); 3  2  3  1 3  2  C NMR (CDC1 ) 5 3  195.76 (C-S), 169.77 (COCH ), 102.62 ((CH ) G), 97.13 (C-l), 80.53 (C-3), 71.73 (C-4), 3  3  2  69.81(CH Ph), 63.85 (C-5), 62.14 (C-6) 52.72 (C-2), 28.94 ((CH ) C), 23.20 (COCH ), 18.89 2  3  ((CH ) C). MS (DCI, NH ) 488 (M + H 3  2  2  3  100%); Anal. Calcd for C25H29O7NS: C, 61.59; H,  +  3  6.00; N, 2.87; S, 6.58. Found: C, 61.34; H, 5.93; N, 2.95; S, 6.55.  (b)  Synthesis of Benzyl-2-acetamido-4,6-0-isopropylidene-2,3-dideoxy-a-D-ribohexopyranoside (22) A solution of 21 (1.74 g, 3.6 mmol) in dry toluene (20 mL) was heated at 80 °C under  argon, and l,r-azobisisobutyronitrile (440 mg, 2.67 mmol) was added, followed by tributyltin hydride (5.7 mL, 21.4 mmol). After 2 h at 80 °C, the mixture was cooled to room temperature and the solvent evaporated in vacuo. The residue was chromotagraphed on silica (CH C1 2  2  MeOH 100:1) to give 22 (1.05 g, 90%) as a white solid: m.p. 44-45 °C; 'H NMR (CDC1 ) 5 3  7.42-7.24 (5H, m, Ph), 5.65 (d, 1H, J , 2  (d, 1H, J  = 9.4 Hz, NH), 4.75 (d, 1H, J ,  N H  ;2  = 3.8 Hz, H-1) 4.73  = 12.0 Hz, PhC/ffl), 4.45 (d, 1H, PhCMf), 4.20-4.30 (m, 1H, H-2), 3.55-3.95 (m,  g e m  4H, H-4, H-5, H-6, H-6'), 1.99 (m, 1H, H-3eq), 1.90 (s, 3H, CH CO), 1.67 (m, 1H, H-3ax), 3  1.47 (s, 3H, (CH ) C), 1.38 (s, 3H, ((CH ) C); C NMR, 8 169.30 (COCH ), 99.39 ((CH ) Q, 13  3  2  3  2  3  3  2  96.19 (C-l), 69.87 (CH Ph), 69.47 (C-5), 65.51 (C-4), 62.75 (C-6) 47.41 (C-2), 31.20 (C-3), 2  29.18 ((CH ) C), 23.28 (COCH ), 19.07 ((CH ) C). MS (DCI, NH ) 3  2  3  3  2  3  336 (M + H  +  100%) Anal.  Calcd for C , 8 H 0 N : C, 64.46; H, 7.51; N, 4.18. Found: C, 64.19; H, 7.59; N, 4.20. 25  5  126 (c)  Synthesis of Disodium uridine 5'-(2-acetamido-2,3-dideoxy-a-D-ribohexopyranosyl diphosphate), 3"-deoxy UDP-GlcNAc (17) C o m p o u n d 22  was deprotected under identical c o n d i t i o n s as those reported for  b e n z y l i d e n e a n a l o g (Srivastava et al,  1990).  the  T h e remainder o f the synthesis was identical to  that described b y Srivastava et al. w i t h the e x c e p t i o n that a two-step d i b e n z y l p h o s p h o r y l a t i o n procedure was e m p l o y e d ( S i m  et.al,  1993).  d e o x y - U D P - G l c N A c s h o w e d identical * H , the literature (Srivastava et al,  4.7.3  Attempted epimerase A  mL)  c o m p o u n d s a n d the final  3"-  C N M R and mass spectra that the ones reported i n  1990).  epimerization  solution (0.6  1 3  T h e intermediate  of  of  3"-deoxy-UDP-GIcNAc  3 " - d e o x y - U D P - G l c N A c (9.6  with  mM)  UDP-GlcNAc  and  2-  U D P - G l c N A c 2-  epimerase (20 units) i n a deuterated p o t a s s i u m phosphate buffer (200 m M , p D 8.2, c o n t a i n i n g 2 mM  dithiothreitol)  p e r i o d o f 12 h.  was incubated at 37 ° C a n d m o n i t o r e d b y * H N M R  A n a l o g o u s solutions  spectroscopy o v e r a  that also contained 8.0 or 0.5 m M U D P - G l c N A c were  a n a l y z e d i n the same manner.  4.7.4  Test of 3"-deoxy-UDP-GlcNAc as an inhibitor of UDP-GlcNAc 2-epimerase T h e initial rates for the e p i m e r i z a t i o n o f a f i x e d amount o f U D P - G l c N A c (0.7  were tested i n the presence o f a variable concentration o f 3 " - d e o x y - U D P - G l c N A c (0.6 m M ) u s i n g the c o u p l e d assay described i n chapter 2.  mM) -  Rates were determined b y f o l l o w i n g the  increase i n absorbance at 340 n m ( N A D H formation) i n assay mixtures that contained 50 T r i s - H C l (0.6 m L , p H 8.8), 2 m M dithiothreitol, 4.0 m M N A D and 3.7  m g o f U D P - M a n N A c dehydrogenase.  6.0  T h e Kj  +  mM  , 2.4 x 10" units o f epimerase 4  was estimated b y d e t e r m i n i n g  the  127 concentration at which the rate of epimerization was approximately 50% of a sample containing no inhibitor.  4.7.5  Test of 3"-deoxy-UDP-GlcNAc as a specific activator of UDP-GlcNAc 2-epimerase. A solution of UDP-ManNAc (11 mM) and 3"-deoxy-UDP-GlcNAc (0.35 mM) in  potassium phosphate buffer (40 pL, 50 mM, pH 8.1, containing 2 m M dithiothreitol) was incubated for 10 min at 37 °C with the epimerase (2.6 x 10 ^ units). A n analogous solution that contained 0.35 mM UDP-GlcNAc instead of the 3 "-deoxy analog was treated identically. Each reaction mixture was analyzed by ion-paired reversed-phase HPLC using the method of Meynal et al. (1995)  4.7.6 (a)  Synthesis of the trifluoroacetamido analog of UDP-GlcNAc (19) 2-Trifluoroacetamido-3,4,6-tri-0-acetyl-2-deoxy-a-D-glucose (31) 2-Trifluoroacetamido-3,4,6-tri-(9-acetyl-2-deoxy-a-D-glucopyranosyl  bromide 30 (1.08  g, 2.3 mmol) was dissolved in 10 mL of dry C H 3 C N containing A g 0 (0.50 g), H 0 (1 mL) was 2  2  added, and the mixture was stirred for 2 h in the dark. The resulting suspension was filtered, concentrated, and separated on a silica gel column using T.3 EtOAcrhexanes. Fractions of higher mobility were segregated for further purification.  Later fractions were pooled and  evaporated to give 31 as a white solid. (0.53 g, 57%) : m.p. 174 °C; [a]? + 20.0° (c 0.5, CHCI3); ' H N M R (CDCI3): 5 6.64 (br d,l H, J  2m  9.1 Hz, NH), 5.34 (dd, 1 H, J  Hz, H-3), 5.33 (d, 1 H, J , , 3.3 Hz, H-l), 5.13 (dd, 1-H, J 2  2 > 3  9.9, J  3 ; 4  9.6  9.6 Hz, H-4), 4.29 (ddd, H-2) 4.25-  4 > 5  4.08 (m, 3 H, H-5, H-6, H-6') 2.08, 2.03, 2.00 (3s, each 3 H, 3Ac);  1 3  C N M R (CDC1 ): 5 3  171.40, 171.05, 169.48 ( C O C H , 3 Ac), 157.30, (q, C O C F , J ,F 37.7 Hz), 115.53 (q, CF ,J ,F 3  3  C  3  C  287.7 Hz), 90.87 (C-l), 70.46 (C-5), 67.82 (C-3), 67.71 (C-4), 61.89 (C-6), 52.76 (C-2), 20.77,  128 20.60, 20.48 ( C H , 3 Ac); F N M R (CDC1 ): 5 -76.6; DCIMS m/z 419 ([M+NH ] 1 9  3  3  4  +  ,100%).  Anal. Calcd. for C14H18O9NF3: C, 41.90; H, 4.52; N, 3.49. Found: C, 42.15; H, 4.32; N, 3.29.  (b)  l,5-Anhydro-2-deoxy-2-trifluoroacetamido-3,4,6-tri-0-acetyl-D-arabino-hex-lenitol (32) This compound was obtained in 7 % yield as a by-product of the chromatographic  purification of 31. Further purification by flash chromatography using benzene:ether 9:1 was 22  required in order to separate this compound from the oxazoline 33. Clear oil, [a]  D  - 46.5° (c 1,  CHC1 ), ' H N M R (CDC1 ): 8 8.19 (br s, 1 H, NH), 7.69 (s, 1 H, H-1), 5.22 (dd, 1 H, J =J , 3  3  3A  3.6 Hz, H-4), 5.15 (dd, 1 H, J , 3  4.35 (dd, 1 H, J  g  e  m  5  4 5  1.1 Hz, H-3), 4.45-4.40 (ddd,l H, J , 7.9, J ,6' 4.1 Hz, H-5), 5  6  5  11.9 Hz, H-6), 4.22 (dd, 1 H, H-6') 2.12, 2.10, 2.08 (3s, each 3 H, 3Ac);  1 3  C  N M R (CDC1 ): 8 172.66, 170.48, 169.54 ( C O C H , 3 Ac), 154.73 (q, J , F 37.5 Hz, COCF ), 3  3  3  C  115.53 (q, J C - F 286.4 Hz, CF ), 141.07 (C-l) 128.34 (C-2) 73.19, 66.40, 66.32, 60.92 (C-3, C3;  4, C-5, C-6) 20.82, 20.79, 20.75 ( C H , 3 Ac); F N M R (CDC1 ): 8 -76.4; DCIMS m/z 401 1 9  3  ([M+NH ] 4  +  3  , 100%). Anal. Calcd. for C i H 0 N F : C, 43.87; H, 4.21; N, 3.65. Found: C, 4  1 6  8  3  43.58; H, 4.28; N, 3.90.  (c)  2-Trifluoromethyl-(3,4,6-tri-0-acetyl-l,2-dideoxy-a-D-glucopyrano)-[2,l-d]oxazoline (33) This compound was obtained as a clear oil as described above. [a]  CHC1 ), *H N M R (CDC1 ): 8 6.27 (d, 1 H, J 3  3), 4.93 (ddd, 1 H, J  3  2A  1>2  7.6 Hz, H-1), 5.31 (dd, 1 H, J  2;  22 D  + 3.6° (c 2.2,  3=^3,4  2.4 Hz, H-  1-3 , J ,s 8.2 Hz, H-4), 4.36 (m, 1 H, H-2) 4.28-4.15 (m, 2H, H-6, H-6'), 4  3.63 (m, 1 H, H-5), 2.10, 2.08, 2.06 (3s, each 3, 3Ac); C N M R (CDC1 ): 8 170.60, 169.33, 1 3  3  169.02 (COCH3, 3 Ac), 156.22 (q, J , F 41.0 Hz, CNCF ), 115.90 (q, J , F 273.0 Hz, C F ) , C  3  C  3  102.22 (C-l) 68.77, 68.60, 67.51, 63.97, 63.20 (C-2, C-3, C-4, C-5, C-6) 20.77, 20.63, 20.62  129 (CH , 3 Ac); 3  19  F N M R (CDC1 ): 8 -71.3; DCIMS m/z 401 ([M+NH ] , 100%). Anal. Calcd. for +  3  4  C i H i 0 N F 3 : C, 43.87; H, 4.21; N, 3.65. Found: C, 44.11; H, 4.19 ; N, 3.57. 6  4  8  (d)  Diphenyl (2-trifluoroacetamido-3,4,6-tri-0-acetyl-2-deoxy-<x-D-glucopyranosyl) phosphate (34) A mixture of 31 (1.24 g, 3.1 mmol) and D M A P (0.87g, 7.1 mmol) in 25 mL of dry  CH2CI2  was stirred at room temperature for 10 min and then cooled to -10 °C.  Diphenylchlorophosphate (1.4 mL, 7.1 mmol) was added dropwise and the solution was stirred at this temperature for 2 h. The mixture was then diluted with CH2CI2 (30 mL) and washed with ice-cold water, ice-cold 0.5M HC1 and saturated NaHC03. After drying the organic phase over M g S 0 , it was concentrated to a small residue which was purified by column chromatography 4  using EtOAc:hexanes 2:3 to afford the title compound as white crystals (1.02 g, 52 %): m.p. 65 °C; [a] D  2 0  + 60.0° (c 1, CHC1 ); ' H N M R (CDC1 ): 5 7.40-7.08 (m, 10H, 2Ph), 6.88 (br d, 1H, 3  3  J ,NH 8.7 Hz, NH), 6.01 (dd, 1 H, J , p 6.0 Hz, J H  2  Hz, H-3), 5.20 (dd, 1 H, J  4 ; 5  1 > 2  3.2 Hz, H-l), 5.32 (dd, 1 H, J  3A  9.9, J ,3 11.3 2  9.7 Hz, H-4), 4.41 (ddd, 1 H, H-2), 4.20-3.88 (m, 3 H, H-5, H-6,  H-6'), 2.03, 2.01, 1.99 (3s, each 3 H, 3Ac); C N M R (CDC1 ): 6 170.92, 170.54, 169.21 13  3  (COCH3, 3 Ac), 157.77 (q, J ,F 38.4 Hz, C O C F ), 129.96, 129.89, 125.85, 125.80, 119.91, 3  C  119.85, 119.82, 119.76 (2 Ph), 115.45 (q, J ,F 287.6 Hz, C F ) , 96.12 (d, J ,p 7 Hz, C-l) 70.03 3  C  c  (C-5), 69.39 (C-3), 67.20 (C-4), 60.91 (C-6), 52.48 (d, J , 8.4 Hz, C-2), 20.53, 20.45, 20.24 2  C  (CH , 3  3 Ac); F N M R (CDC1 ): 5 -76.1; 19  3 1  3  P N M R (CDC1 ): 5 -14.39; DCIMS m/z 651  ([M+NH ] , 100%). Anal. Calcd. for C 2 6 H O i 2 N P F  3  +  4  49.08; H, 4.41; N, 2.22.  27  P  3  : C, 49.30; H, 4.30; N, 2.21. Found: C,  130 (e)  Monotriethylammonium salt of 2-trifluoroacetamido-3,4,6-tri-0-acetyl-2-deoxy-aD-glucopyranosyl phosphate (35) A solution of 34 (250 mg, 0.4 mmol) in EtOAc: MeOH 1:1 (10 mL) was hydrogenated  (4 Atm) in the presence of Pt02 catalyst (25 mg). After 4 hours, the catalyst was removed by filtration and Et N (2 mL) was added to the filtrate. Evaporation of the solvent afforded a syrup 3  that was purified by preparative thin layer chromatography using CHCl3:MeOH:H20 10:10:1 to give the free acid form of 35 as a white hygroscopic solid (60 mg , 32 %), this was converted to the monotriethylammonium salt by passing through a small column (3 cm) of Dowex 50WX8 (triethylammonium form): R 0.6 F  (CHCl :MeOH:H 0 10:10:1); ' H N M R (D 0): 8 5.53 (dd, 3  2  1 H, J , , 7.1 Hz, Ji, 3.1 Hz, H-l), 5.42 (dd, 1 H, J ,i 10.4, P  2  2  2  J  3A  9.8 Hz, H-3), 5.13 (dd, 1 H,  J ,5 4  9.8 Hz, H-4), 4.50-4.14 (m, 4 H, H-2, H-5, H-6, H-6') 2.14, 2.10, 2.02 (3s, each 3 H, 3Ac); C 1 3  N M R (D 0): 8 174.07, 173.43, 173.14 ( C O C H , 3 Ac), 159.64 (q, J ,Y 38.4 Hz, COCF ), 2  3  3  C  115.86 (q, Jc, 284.5 Hz, C F ), 92.79 (d, J , P 5.5 Hz, C-l), 71.17 (C-5), 68.63 (C-3), 68.31 (CF  3  C  J ,P 8.4 Hz, C-2), 20.41, 20.37, 20.20 (CH , 3 Ac);  2  4), 62.05 (C-6), 52.57 (d, 5 -75.4;  3 1  C  19  3  F N M R (D 0): 2  P N M R (D 0): 8 -1.07; HRLSIMS m/z : 480.0522 (calcd. for C i H i 0 i N P F , 2  4  9  2  3  480.0519).  (f)  Monopyridinium phosphate (36)  salt  of  2-trifluoroacetamido-2-deoxy-a-D-glucopyranosyl  Compound 35 (180 mg, 0.4 mmol) was dissolved in 3 mL of dry MeOH. The solution was cooled to 0 °C (ice-water) and 0.50 mL of NaOMe in MeOH (IN) was added.  After  stirring 20 minutes the reaction was quenched using Dowex 50W-X8 resin (pyridinium form, previously washed with MeOH). The resin was filtered off and the methanolic solution was evaporated. The residue was dissolved in a small amount of water, passed through a small column (3 cm) of Dowex 50W-X8 (pyridinium form) resin and lyophilized to give 7 as the  131 monopyridinium salt (120 mg, 68 %): 'H N M R (D 0): 8 5.48 (dd, 1 H, J 2  1 ; P  7.2, J  1 > 2  3.2 Hz, H-  1), 4.04 (dd, 1 H, J , 10.6 Hz, H-2), 3.93-3.72 (m, 4 H, H-3 H-5 H-6 H-6'), 3.53 (dd, 1 H, 2  3  J =J ,5 9.6 Hz, H-4); 3A  4  1 3  C N M R (D 0): 8 159.61 (q, J ,v 37.6 Hz, COCF ), 115.96 (q, J , F 2  3  c  C  284.7 Hz, CF ), 93.12 (d, J . 5.8 Hz, C-l), 73.04 (C-5), 70.20 (C-3), 69.80 (C-4), 60.50 (C-6), 3  C P  54.65 (d, J , p 8.0 Hz, C-2); F N M R (D 0): 8 -75.7; P N M R (D 0): 8 -1.33; HRLSIMS m/z 2  19  3 1  c  2  2  354.0189 (calcd for C H i 0 N P F , 354.0202). 8  (g)  2  9  3  Preparation of 36 by the MacDonald procedure Crystalline phosphoric acid (500 mg, 5.1 mmol) was dried in vacuo over P Os for 12 h. 2  Solid 2-trifluoroacetamido-1,3,4,6-tetra-O-acetyl-2-deoxy-y0-D-glucopyranose  (280 mg, 0.63  mmol) was added and the mixture was heated at 60 °C in vacuo. The formation of a melt and the evolution of acetic acid vapors was observed.  After 2 h, heating was ceased and the  resulting dark black mixture was dissolved in anhydrous THF (5 mL). The solution was cooled to 0 °C and cone, ammonium hydroxide (0.5 mL) was added. The precipitate of ammonium phosphate was filtered off and washed with THF (20 mL). The combined filtrate and washings were evaporated to give a syrupy residue that was applied to two plates of silica gel and eluted with C H C l : M e O H : H 0 10:10:1. Isolation of the corresponding zone (R 0.15) gave crude 6 as 3  2  F  a light brown solid. This material was O-deacetylated and purified as described above to give 7 (73 mg, 27% overall yield).  (h)  Dilithium uridine diphosphate) 19  5'-(2-trifluoroacetamido-2-deoxy-a-D-glucopyranosyl  This compound was prepared following the general procedure outlined in Chapter 2 for UDP hexapyranosyl diphosphates in 44% yield and isolated as its dilithium salt (dihydrate): ' H  132  N M R (D 0): 5 7.92 (d, 1 H, J ,6 8.1 Hz, H-6), 5.93 (d, 1 H, J 2  5  5), 5.57 (dd, 1H, Jr<,p 7.1,  VX  4.9 Hz, H-1'), 5.91 (d, 1 H, H-  J >,T 3.3 Hz, H-1"), 4.34-4.28 (m, 2 H, H-2',  H-3'), 4.25-4.12 (m, 3  V  H, H-4" H-5'a, H-5'b), 3.95-3.72 (m, 4 H, H-2" H-5", H-6"a, H-6"b, H-3"), 3.54 (dd, 1 H, J »4"=^4-,5" 3  9.6 Hz, H-4");  1 3  C N M R (D 0): 5 160.4 (q, 2  J ,F C  37.7 Hz, C O C F ), 167.2 (C-4), 3  152.7 (C-2) 142.5 (C-6), 116.6 (q, J - F 284.4 Hz, C F ) , 103.5 (C-5), 94.6 (d, J - P 5.9 Hz, C-l"), 3  C  C  89.3 (C-l'), 84.1 (d, J , P 9.2 Hz, C-4'), 74.7 (C-3'), 73.9 (C-5"), 71.2 (C-2'), 70.5, 70.3 (C-3", C  J.  2  C-4"), 65.8 (d, Jc,p 5.5 Hz, C-5'), 61.1 (C-6"), 55.3 (d,  C  8.8 Hz, C-2"), F N M R (D 0): 19  P  2  8-  74.9; P N M R (D 0): 8 -11.01 (d, .7 20.3 Hz, Pp), -12.90 (d, J 20.3 Hz, Pa); HRLSIMS m/z 3 1  2  674.07928  (calcd  for  Ci H Oi N P F 7  2 3  7  3  2  Li  3  2  ,  674.0775).  Anal.  Calcd  for  C i H N O i P F L i ' 2 H 0 : C, 28.79; H, 3.70; N, 5.92. Found: C, 28.68; H, 3.79; N, 5.71. 7  22  4.7.7  3  7  2  3  2  2  Epimerization of the trifluoroacetamido analog of UDP-GlcNAc with UDP-GlcNAc 2-epimerase A solution (0.6 mL) of UDP-GlcNAc-F (8.9 mmol) and UDP-GlcNAc 2-epimerase 3  (3.8 units) in a deuterated Tris-HCl buffer (50 mM, pD 8.8) containing 2 m M dithiothreitol) was incubated at 37 °C and monitored by H and F N M R spectroscopy over a period of 19 h. !  4.7.8  19  Kinetic characterization of the trifluoroacetamido analog of UDP-GlcNAc The kinetic parameters of the cloned UDP-GlcNAc 2-epimerase with UDP-GlcNAc-F  3  as substrate were evaluated in a similar fashion as for the normal substrate described in chapter 2.  The initial rates were determined at 37 °C in assay mixtures (0.6 mL total volume)  containing 50 m M Tris-HCl (pH 8.8) buffer, 2 m M DTT, 4 m M N A D  +  and UDP-ManNAc  dehydrogenase (5 mg of protein in each cuvette) and a variable concentration of UDP-GlcNAc (0.15 to 6 mM). The epimerization was initiated by addition of UDP-GlcNAc 2-epimerase (2.1  133 units) and followed by measuring the increase in absorbance at 340 nm due to NADH formation (S340 = 6220 M^cm" ) using a Varian Cary 3E spectrophotometer. The rate data was 1  divided by a factor of 2 to account for the stoichiometry of the dehydrogenase reaction. The resulting rate was plotted as a function of substrate concentration, and the kinetic parameters were determined by a direct fit of the data to a Hill equation using the computer program Grafit (Erithacus Software, 1994). The program employs a non-linear regression analysis on the data using the method of Marquart (1963), and reports an error for the data based on the deviation from the calculated curve-of-best-fit.  4.7.9  Preparation of the oxazoline of ManNAc (18) 2-Methyl-(2-acetamido-3,4,6-tri-0-acetyl-l,2-dideoxy-p-D-mannopyrano)-[2,l-d]-  oxazoline was prepared and O-deacetylated as described by Khorlin et al. (1968). *H NMR (D 0): 8 5.27 (d, J i , = 1 Hz, H-l), 5.24 (dd, J , = 7.5, J , = 3.8 Hz, H-3), 4.53 (dd, H-2), 4.23 2  2  2  3  3  4  (dd, J , = 9.0 Hz, H-4) 3.86-3.75 (m, H-5,H-6), 3.64 (dd, J 4  5  =12.1, H-6'); C NMR (D 0): 8 13  g e m  2  171.78 (C=N), 103.84 (C-l); 85.62, 82.08, 79.01, 71.67 , 65.78 (C-6), 15.21 (CH CN); DCIMS 3  m/z 204 ([M+l], 100%); IR 1672 cm" (C=N). Anal. Calcd. for C H 0 N : C, 47.29; H, 6.45; 1  8  13  5  N 6.89. Found: C, 47.36; H, 6.45; N, 6.83.  4.7.10 Attempted coupling of the oxazoline of ManNAc and UDP with UDP-GlcNAc 2epimerase A solution (0.5 mL total volume) containing oxazoline of ManNAc (10 mM), UDPGlcNAc-F (1 mM) and UDP (ImM) with epimerase (2.08 units) in deuterated potassium 3  phosphate buffer (50 mM, pD 8.2, containing 2 mM dithiothreitol) was incubated at 37 °C and the reaction monitored by *H NMR spectroscopy over a period of 18 hours.  Chapter Five Conclusion and Future directions  134  135 5.1  Introduction Four possible catalytic mechanisms for the mode of action of UDP-GlcNAc 2-  epimerase were presented in Chapter One, based on the information available at time when this project was initiated. Each of these proposed mechanisms attempts to explain how the enzyme inverts a stereocenter that does not bear an acidic proton since a conventional epimerization by deprotonation/reprotonation is not feasible. The main goal of this thesis was to propose the most reasonable mechanism for the epimerase based in the results obtained by evaluating substrate analogues designed to act as probes for the enzyme. The production of large amounts of UDP-GlcNAc  2-epimerase  necessary for  mechanistic studies was facilitated by the use of recombinant D N A technology that allowed cloning and overexpression of the gene encoding for the epimerase in E. coli. Previous studies done with the endogenous epimerase (Kawamura et al,  1978,  1979)  unveiled certain  characteristics of the enzyme that were important for the present study and were corroborated with the recombinant epimerase. These characteristics included the allosteric nature of the enzyme and the occurrence of a proton transfer mechanism during the epimerization (as presented in Chapter two). Some important differences among the four mechanisms served as the basis on which experiments aimed at probing the viability of a particular mechanistic pathway were designed. Paths A and D, for instance, require the presence of an N A D cofactor to perform the transient +  oxidation of the hydroxyl at C-3". Although the presence of this cofactor has been speculated in the literature (Salo, 1976), no efforts were made until recently to prove its existence (Morgan et ah, 1997). Experiments to detect the presence of this cofactor gave negative results, strongly suggesting that the epimerase did not operate with the help of a cofactor (Morgan, 1998). The fate of the anomeric C - 0 bond of the hexosamine moiety of the substrate during the epimerization is another distinct feature among the proposed mechanisms. While Paths A  136 and C describe an inversion of the C-2" stereocenter without breaking and reforming the anomeric C - O bond, Paths B and D do require this cleavage to take place. The positional isotope exchange experiments described in Chapter three demonstrated the occurrence of isotopic scrambling under reversible and irreversible conditions, providing strong evidence for the cleavage of the anomeric bond during the epimerization. The above results allowed us to eliminate Paths A and C as a possibility and favor Path B. ' Further evidence that Path B was the most likely candidate for a minimal description for the mode of action of the epimerase was obtained when the intermediates postulated in this mechanism were in fact observed in solution after long incubation periods. These species were positively identified as 2-acetamidoglucal and UDP by N M R spectroscopy. The best minimal description for the mechanism of action of this enzyme (Path B) involves a p-elimination reaction for the formation of the intermediates from UDP-GlcNAc as the first step towards the formation of the epimeric UDP-ManNAc. A more detailed discussion of p-elimination reactions as well as precedents for its occurrence in enzymatic processes is described in the next section. Finally, an accurate description of the exact reaction pathway used by this enzyme could be subject of further investigation and proposed future directions on the study of this enzyme are presented at the close of the chapter.  5.2  p-Eliminations  5.2.1  Background An elimination reaction occurs when a molecule decomposes by splitting of two  fragments that are not replaced by other atoms or groups. These have been classified depending on the relative placement of the carbon atoms from which elimination occurs. Thus, aeliminations (1,1-eliminations) occur when both groups are on the same carbon and they form  137 unstable divalent carbon species (carbenes).  In B-eliminations (1,2-eliminations)  the two  fragments are lost from adjacent atoms and form a double or triple bond. In y-eliminations (1,3-eliminations)  the loss of fragments occurs between alternate carbons and produces  cyclopropane structures. By far, the most common elimination process is the 3-elimination. This process is usually reversible and is called a 1,2 addition in the reverse direction. B-Eliminations can be further subdivided by closer examination of the mechanisms involved.  Three different  mechanisms are outlined in Figure 5.1 As depicted, there are three classes of heterolytic 0-elimination reactions: E l , Elcb and E2. In E l eliminations, the cleavage of the C-X bond occurs prior to the cleavage of the C-H bond, and produces a carbonium ion intermediate.  The intermediate then loses a proton from  the a-carbon to form the alkene product. This type of mechanism is common in solvolytic organic reactions with substrates that can stabilize the developing carbonium ion, but unusual in enzymatic reactions possibly due to the need to protect the carbonium ion from the ubiquitous nucleophiles present in proteins (Anderson, 1998). A second pathway is referred to as Elcb because the intermediate formed is a carbanion. The presence of an electron-withdrawing functional group bonded to the a-carbon promotes the abstraction of the ot-proton and facilitates the formation of the carbanion intermediate.  This situation is common in enzymatic eliminations, however, the means by  which an enzyme activates the a-proton remains uncertain (Gerlt and Gassman, 1993; Guthrie andKluger, 1993).  138  E l Mechanism  RCH -CHR'  •  2  RCH -CHR'  RCH=CHR'  2  X  + BH  X  Elcb Mechanism  RCH -CHR' 2  +  B  -  RCH—CHR'  X  + BH  *.  RCH=CHR.'  *-  RCH=CHR' +  + X"  X  E 2 Mechanism 5  B. H  RCH -CHR' + B" 2  I  **  X  R" i H  \  BH +  X  X 8"  Figure 5.1 The three types of mechanisms for S-elimination reactions.  E2 mechanisms involve a bimolecular transition state in which abstraction of a proton p to the leaving group is concerted with departure of the leaving group. A better understanding of many features of P-elimination reactions is obtained upon the recognition that these three mechanisms represent variants of a continuum of mechanistic possibilities.  Many p-elimination reactions occur via mechanisms that are intermediate  between the limiting types. This idea is called the variable E2 transition state theory (Bunnet, 1962) and is depicted in Figure 5.2  139  Increasing C-H breaking in the transition state  B  B—H  H  B  \  E1cb  \  H  B  H  X  E1cb-like  Synchronous E2  E1-like  E1  E2 Increasing C-X breaking in the transition state  >  Figure 5.2 Variable transition state theory of elimination reactions. Adapted from Carey and Sandberg(1990)  Another useful way of understanding the ideas of the variable transition state theory, employs a two-dimensional potential energy surface diagram referred to as the More O'Ferral diagram (More O'Ferral, 1970) in which the x-axis represents the extent of the C-H bond cleavage and the_y-axes the cleavage of the C-X bond (Figure 5.3). The More O'Ferral diagram facilitates the understanding of how factors that stabilize an intermediate will alter the relative position of the transition state.  Thus, interactions that help to stabilize the formation of a  carbanion will move the transition state toward the lower-right hand corner, while those that favor a carbonium ion will move the transition state toward the upper-left. The dashed circles in the More O'Ferral diagram represent the regions in which the intermediates, in the E l and Elcb mechanism, are stable. If both of these intermediates are unstable, the area within these circles will collapse and a concerted reaction with a single transition state (E2 reaction) will result. The position of the transition state for this concerted reaction does not necessarily lie along the diagonal of the More O'Ferral diagram (perfectly synchronous). In the E2 transition state  140 represented in Figure 5.3 the cleavage of the C-H bond occurs to a greater degree than that of the C-X bond (E2-'Elcb-like').  Figure 5.3 More O'Ferrall diagram for description of the mechanism of p-elimination reactions. R indicates an activating group X indicates an heteroatom with oxidation state n. n  ac  In addition to the classification as E l , Elcb, or E2 that describes the timing of the cleavage of the C-H and C-X bonds, p-elimination reactions can also be classified in terms of the stereochemical relationship between the cleaved C-H and C-X bonds. Two scenarios are possible (Figure 5.4), either H and X depart from opposite faces of the incipient double bond in an anti elimination, or they are eliminated from the same face of the double bond in a syn elimination. If both a and P carbons are stereogenic, the stereochemistry of the elimination process will determine whether the olefin generated has a Z or E configuration.  141  >  B: H  R  <R  3  \  4  \  h  /  \  B: H  anti elimination  1,2-anf/addition  X  B: H  B:  X"  v A  I I  H  syn elimination  V 4 R  t  R  2  R  1,2-syn addition 3 4  Figure 5.4 Production of alkenes by anti and syn elimination reactions. X represents a leaving group, Ri, R.2, R-3 and R 4 represent other groups.  5.2  Precedents in enzymatic reactions Elimination-addition reactions occur in many enzyme catalyzed processes and generally  involve an Elcb mechanism. The mechanism by which enzymes catalyze these processes are, for the most part, better described in terms of acid-base chemistry. The enzymatic elimination of H X from a substrate normally occurs by the initial proton abstraction from a carbon of substrate with subsequent loss of the leaving group X (= O H , OR, N H , NHR) from an adjacent 2  carbon of the generated carbanionic species to produce an olefin. This process is frequently facilitated by the presence of an 'activating' group such as a carbonyl or carboxylate in the substrate that enhances the acidity of the proton being abstracted. These reactions are all reversible, the reverse process being a 1,2 addition reactions (Figure 5.4).  142 Interestingly, the ease of proton abstraction, which depends on the degree of 'activation' exerted by the adjacent groups, has been correlated with the stereochemistry followed by the elimination reaction (syn or anti). While a«/i-elimination processes are commonly observed in substrates bearing groups that do not contribute in a high degree to the activation of the proton (such as carboxylate), sy«-eliminations are found in cases where the abstracted proton is adjacent to more activating groups such as the carbonyl group of a ketone, aldehyde or thioester.  Some examples of enzymatic ^-elimination reactions on activated  substrates are presented in Table 5.1. There are exceptions to this general rule as syn eliminations have been reported on substrates with relatively non-acidic protons. One such case is the reaction catalyzed by UDPGlcNAc  enolpyruvyl  transferase.  This  enzyme  transfers  the  enolpyruvyl  group  of  phosphoenolpyruvate (PEP) to the 3"-OH of UDP-GlcNAc via an addition-elimination mechanism that involves the formation of a tetrahedral intermediate prior to the elimination of phosphate (Figure 5.5 a).  The stereochemical course of the enolpyruvyl transfer has been  recently established (Skarzynski et al, 1998), and was found to involve an the addition to the double bond of PEP in anti fashion prior to the syn elimination of the phosphate from the tetrahedral  intermediate.  A  similar  phosphoenolpyruvate-transferring  scenario  enzyme,  has  also  been  proposed for  5-enolpyruvylshikimate-3-phosphate  another synthase  (Figure 5.5 b; Lee et al, 1984; Grimshaw et al, 1984; Kim et al, 1996). Precedence for enzyme-catalyzed a/tfz-elimination  reactions are common in the  literature. Such reactions are observed throughout the enolase and fumarase superfamily of enzymes and also in most of the carbon-nitrogen lyases. Some representative examples are shown in Table 5.2.  143  (a) HO  v  HO-^-"7*A UDP-GlcNAc  00c H  HO  HOHO O  buDP  HNAC  3  HNAc  C ^  OUDP  H C=<^  HNAc  2  Q [ J D p  COO  OPO,  coo  1  EP-UDP-GlcNAc  OPO,  PEP  (b) coo COO  coo  OH  -"o,Pcr"  CH-,  OH Shikimate-3phosphate  -"0,PO"  •S^o^OH  OPO,  HPO/  CH, COO  -"O3PO""  Y  "'o  COO  OH  O PO  ^coo  3  5-EP-shikimate-3phosphate  PEP  Figure 5.5 Phosphoenolpyruvate transferring enzymes that catalyzes yyH-elimination on an unactivated center, (a) UDP-GlcNAc enolpyruvyl transferase (MurA) (b) 5-enolpyruvylshikimate synthase.  There are also precedents for the reverse process of (3-eliminations, namely 1,2 additions.  Anti 1,2-additions normally occur in the hydration of glycals by retaining p-  glucosidases (Legler, 1990). The only, known exception for this type of enzymes is the syn addition in the hydration of 2-acetamidoglucal by p-A^-acetylhexosaminidase (Lai and Whithers, 1994). The nonenzymatic acid catalyzed addition of alcohols to glycals are known to proceed preferentially with a syn stereochemical course, however, trans addition product are also observed (Kalia et al., 1992).  144  Enzyme  Chemistry of elimination step HO^  p-Hydroxydecanoylthioester dehydratase (EC 4.2.1.60)  H Schwab et al., 1986  HV H C(CH ) CH 3  2 5  \""COSR H  2  HO"S P-Hydroxybutyryl-CoA dehydratase (EC 4.2.1.55)  H  Sedgwick et al., 1978 COSCoA  HoC HO^v  Methylglutaconyl-CoA hydratase(EC 4.2.1.18)  H  Messner et al., 1975 ~0 CH C - " V \ 2  "  2  3  H  H  COSCoA  C  HO Dehydroquinate dehydratase (EC 4.2.1.10)  Reference  OH  "0,Cv  <3H  Hanson and Rose, 1963  * 0 H  O  O—v  Mazumder et al., 1989  UV Endonuclease V  S-Adenosylhomocysteine hydrolase (EC 3.3.1.1)  £V..H NH  AD  Parry and Asconas, 1985  H  3  OH HO Dehydroquinate synthase (EC 4.6.1.3)  0,C  Widlanski et al., 1987 OH  |J  0  P  H  0  3  CO-  Muconate cycloisomerase (EC 5.5.1.1)  Chari  "0,C>  3-Carboxy-cis-cis-muconate cyclase (EC 5.5.1.5)  etal.,  1987  O COKirby etal., 1975  Table 5.1 Some 1,2-elimination/addition enzymatic reactions involving .sy^-stereochemistry  145 Chemistry of elimination step  Enzyme Fumarase (fumarate hydratase) ( E C 4.2.1.2) R = H Mesaconate hydratase (EC 4.2.1.34) R = CH  H O ^  CO,  0,(5  3  Maleate hydratase (EC 4.2.1.34) R = H Citraconate hydratase (EC 4.2.1.35) R = CH  H  H  H  co,  Reference Farrar et al., 1957 England, 1958 England et al., 1967 Subramanian etal, 1966 England et al., 1967 Subramanian et al., 1966  0,C 3  HO«,  CO,  Aconitase (Aconitase hydratase) (EC 4.2.1.3)  Hanson and Rose, 1963  ,  H O *v  Enolase(EC 4.2.1.11)  H  ^ HO*,  2-Isopropylmalate dehydratase (EC 4.2.1.33)  C 0  2  Cohne/a/., 1970  H H CO,  Cole etal., 1973  (H C) HC 3  2  Aspartate ammonia-lyase (EC 4.3.1.1)  R = CO. H  Histidine ammonia-lyase (EC 4.3.1.3) R=  3  N ^  H  Anet, 1960 England, 1958 Retey etal., 1970 Wightman etal, 1972  N  o,c Phenylalanine ammonia-lyase (EC 4.3.1.5) R=  Argininosuccinate lyase H (EC 4.3.2.1) R= ^ -  N  R-N ^  H  Hoberman et al., 1964 Miller and Buchanan, 1962  R  Argininosuccinate lyase (EC 4.3.2.1) R=  "o,c  OH OH CO,  3-Carboxy-cis-cis-muconate cycloisomerase (EC 5.5.1.2)  Chari et al., 1987  H ^ T "  X  H  Table 5.2 Some 1,2-elimination/addition enzymatic reactions involving arcfr'-stereochemistry  146 5.3  Elimination-addition reactions in UDP-GlcNAc 2-epimerase The minimal description for the mechanism of UDP-GlcNAc 2-epimerase presented  earlier in this thesis, does not address the type of p-elimination that the enzyme employs to produce the 2-acetamidoglucal and UDP intermediates. The three types of p-elimination ( E l , Elcb or E2) can be postulated for the initial a«fr'-elimination of UDP from UDP-GlcNAc (Figure 5.6)  AcNH  Figure 5.6 Three possible mechanisms for the P-elimination of UDP from UDP-GlcNAc  Although the first mechanism (Elcb) is very common in enzymatic reactions, it is unlikely to be the preferred route for the mode of action of UDP-GlcNAc 2-epimerase. In all  147 known cases of enzymatic Elcb elimination, an electron withdrawing group in the substrate, such as an adjacent carbonyl or carboxylate, serves to stabilize the anionic intermediate (Anderson, 1991; Creighton and Murthy, 1990; Anderson , 1998). This is not the case in UDPGlcNAc since the absence of electron withdrawing groups adjacent to the C-2" make the generation of a carbanion at this center a very unlikely process. The second possibility is a concerted E2 mechanism that proceeds without any intermediate.  The only well documented example of an enzymatic E2 elimination is the  reaction catalyzed by enoyl-CoA hydratase (Crotonase). A concerted syn elimination of water has been reported for this enzyme (Bahnson and Anderson, 1991) but the substrate of the reaction, iS-3-hydroxybutyrylpantetheine, possesses unique structural features that make a synelimination chemically feasible. UDP-GlcNAc lacks these particular features which precludes drawing an adequate analogy between the epimerase and this enzyme. The third possibility appears to be the most suitable for the mechanism of the epimerase. Although El-type elimination reactions are unusual enzymatic processes, it has been proposed for the reaction catalyzed by imidazolglycerol phosphate dehydratase (Parker et. al., 1995).  The cationic intermediate, adjacent to the imidazol ring, is stabilized through  conjugation to the heterocycle. A somewhat similar mode of stabilization can be postulated for UDP-GlcNAc 2-epimerase.  The presence of a good leaving group (UDP) and the possible  formation of a relatively stable oxocarbenium ion intermediate are factors that could contribute to stabilize a cationic intermediate (El elimination). Perhaps, the mechanism is partially E l , in other words, the enzyme does not necessarily form a formal carbonium ion intermediate but rather proceeds through a E2-El-like mechanism. In terms of the More O'Ferral diagram, the aforementioned factors could help to stabilize the developing of an anomeric carbon with a cationic character, and therefore will displace the transition state towards the upper left corner of the diagram (Figure 5.3).  Further support for this possibility has been found in the  148 observation of a primary kinetic deuterium isotope effect for the epimerase reaction (Morgan et al,  1997).  The epimerization of C-2"-[ H]-UDP-GlcNAc was slowed by an isotope effect z  (measured under saturating conditions) of 1.8. This result indicates that the cleavage of the C2" proton is at least partially rate-determining. Precedents of primary deuterium isotope effect in nonenzymatic E l elimination reaction can be found in the literature (Baciocchi et al,  1979;  Noyce and Lane, 1962) Regarding the stereochemistry of the addition elimination reactions, it is important to note that UDP-GlcNAc 2-epimerase must employ both, an a«/7-elimination of UDP from UDPGlcNAc and a svn-elimination of UDP from UDP-ManNAc in order to form the intermediates. Furthermore, the enzyme must also catalyze the readdition of the intermediates by two different processes, the syn addition to form UDP- ManNAc and the anti addition to generate UDPGlcNAc. As seen before, there are precedents for both, syn and anti eliminations and additions in enzymatic processes but the use of all of these processes in a single enzyme is unprecedented. The results presented in this study provides a reasonable mechanism for the mode of action of UDP-GlcNAc 2-epimerase, however, more detailed features of the enzymatic mechanism can be the subject of further studies, as proposed in the next section.  5.4  Future directions Most of the well-understood enzymatic mechanisms include a precise identification of  the amino acid residues involved in the catalytic process. Several techniques have evolved that aid in the correct assignment of these residues, including the use of active-site directed irreversible inhibitors or affinity labels, X-ray crystallographic structural data, and more recently the technique of site-directed mutagenesis. These techniques can be applied for further mechanistic studies on UDP-GlcNAc 2-epimerase.  149 The number of catalytic bases involved during the enzymatic epimerization is a question that still remains unanswered. As mentioned in Chapter one, enzymatic epimerization can occur via a one-base mechanism or a two-base mechanism (Figure 1.1). Experiments that use the concept of 'internal transfer' (Cardinale and Abeles, 1968) can be helpful in determining which type of mechanism the epimerase employs.  The fraction of product  containing solvent derived deuterium can be measured after performing the epimerization under irreversible/initial velocity conditions in both reaction directions.  If two separate  enzymatic bases are responsible for the transfer of protons from the epimeric sugars, one would expect that the product would exclusively contain solvent derived isotope. On the other hand, if a single, flexible base shuttles the proton between the epimers, one may expect to see substrate-derived protons in the product even if the epimerization is performed in deuterated media. The use of affinity labels can provide valuable information regarding the precise nature of the active site bases. Preliminary studies in this area, however, failed to specifically label active site residues (Morgan, 1998). Apparently, the active site of the epimerase is stringent for both the UDP moiety and the hexosamine part of the substrate. Inhibitors based on UDP as the only recognizable moiety in their structure (such as uridine 5'-diphosphate chloroacetol 42, or uridine 5'-diphosphate 2,3-epoxypropanol 43) fail to label the active site possibly due to inefficient binding.  42  o  o  OH OH  OH OH  43  150 A great deal of information can be gained by X-ray crystallographic structural data of the enzyme. Initial work in this area is being done by Mr. Robert Campbell in our group in collaboration with Dr. Natalie Strynadka (UBC, Biochemistry). The interpretation of structural data obtained by this technique has an important caveat. The allosteric nature of the epimerase suggests that the enzyme has both a 'resting' (without UDP-GlcNAc) and an 'active' (with UDP-GlcNAc) conformation that must differ significantly. It is desirable to obtain the crystal structure of the enzyme in its 'active' conformation, however the high concentrations of enzyme necessary to grow crystals may result in the formation of the intermediates 2acetamidoglucal and UDP. Crystallization of the active form may be possible by crystallizing the enzyme in the presence of its substrate at a pH at which the activity of the enzyme is negligible.  Alternatively, this problem might be overcome by using a non-epimerizable  substrate analog that could satisfy the allosteric requirement of the enzyme without being turned over.  The 3"-deoxy substrate analog described in Chapter Four was not turned over by  the enzyme even though it appears to bind to the active site, however, it did not work as an activator of the allosteric enzyme and therefore could not be used to this end. candidate for this type of study is the sugar phosphonate analog 44.  The best  The structure of this  compound resembles in all aspects that of the natural substrate of the enzyme with the exception that an 'isosteric' methylene group replaces the anomeric pyranose oxygen. This modification voids the elimination of the UDP moiety by the epimerase and should therefore not be turned over by the enzyme. The synthesis of the precursor phosphonate analog of the GlcNAc a-1-phosphate has been recently reported (Casero et al., 1996) and the coupling of UDP with this sugar phosphonate should afford the desired substrate analog 44.  This  compound is very likely to bind to the epimerase modulator site, therefore 'activating' the enzyme and also to the active site but should not undergo any epimerization.  151  X-Ray crystallographic data could also provide valuable information about the nature of the active site acid/base residues that participate in the transfer of the C-2" proton and could allow a more detailed study by the technique of site directed mutagenesis. The modification of specific residues that might be important in enzymatic catalysis can be achieved by this technique.  One problem with these modifications is that if the activity of the enzyme is  abolished after mutating one residue this does not necessarily mean that the mutated residue is important in catalysis. It may instead be important in binding of the substrate or in defining the proper conformation of the enzyme. On the other hand, if the mutated enzyme catalyzes the formation of the glucal intermediate or causes a PIX such as the one described in Chapter Three, but not the overall epimerization, one could assert that the mutated residue was in fact the catalytic base. Finally, the present study has suggested that neighboring group participation by the acetamido group of the substrate is not significant during the enzymatic epimerization but has not conclusively proven its complete absence. Further studies along this line may be desirable, in particular a further analysis on the specificity of different parts of the acetamido moiety can be of interest. The replacement of the acetyl oxygen by sulfur in both UDP-sugar epimers (45, 46) would permit the formation and detection of a more stable GlcNAc or ManNAc thiazolines  (47, 48)  152  45  46  47  48  Other substrate analogs that might be interesting to evaluate with the epimerase include the acetate analog 49 and the jV-methanesulfonylated analog 50.  Both analogs have  significantly reduced neighboring group participation and can serve as probes for evaluating the specificity of the JV-acetamido group.  49  50  153  Appendix 1 Regulatory enzymes and Hill equation Many enzymes obey what is known as Michaelis-Menten kinetics which occur when a plot of the reaction rate versus substrate concentration is a rectangular hyperbola (or when a plot of reciprocal initial velocity versus reciprocal substrate concentration is linear). The kinetics of some enzymes, however, do not follow this ideal description but rather give velocity versus substrate concentration curves that are sigmoidal. This type of enzyme is called control or regulatory and is usually situated at the beginning or at the branch points of a metabolic pathway.  The activity of these enzymes are usually regulated or modulated by substances  called effectors (or modulators) which can be activators, inhibitors or both. Normally, effector molecules are not structurally similar to the substrate of the enzyme and therefore is not likely for them to bind at the active site but at an alternative site called the allosteric site. If an effector of an enzyme is also the substrate, it is called homotropic effector; if it is a nonsubstrate, it is called heterotropic. Enzymes sensitive to allosteric regulation are usually multisubunit  (oligomeric)  enzymes, and show a complicated kinetic behavior as a consequence of structural interaction or cooperativity between subunits, with each subunit generally having its own active site. For an oligomeric enzyme where each subunit has an independent active site the velocity versus substrate concentration curve will be the usual rectangular hyperbola (Michaelis-Menten kinetics). In other words, one molecule of a tetrameric enzyme with four independent active sites is kinetically indistinguishable from four molecules of a monomeric enzyme, each with its own active site. On the other hand, if the subunits of an oligomeric enzyme interact, with each subunit modulating the responses of others, the kinetics becomes more complicated. In such a case the presence of a substrate molecule (or some other ligand that activates or inhibits) in one site affects either the binding of a subsequent molecule to other vacant sites or the rate of product formation from other occupied sites. This type of enzyme will not follow Michaelis-Menten kinetics, and the enzyme is classified as 'allosteric' (Segel, 1993).  Generally allosteric enzymes give sigmoidal velocity curves. If the binding of one  substrate molecule facilitates the binding of the next substrate molecule by increasing the affinities of the other vacant binding sites the enzyme is said to show positive cooperativity. The tetrameric protein hemoglobin is classical allosteric enzyme that was studied early in the century by Bohr who reported a sigmoidal oxygen saturation curve for this protein. In  1  154 1910 A . V . Hill proposed an empirical equation which gave a reasonably good fit to the observed data. Further improvements on the equation were done by Adair in 1925 and more recently by Koshland (1970) who summarized some diagnostic test to determine whether cooperativity is present and its kind (positive or negative). These considerations can be illustrated by discussing a simplified velocity equation for allosteric enzymes, the Hill equation (Walsh, 1979). Given a high degree of cooperativity in substrate binding for an enzyme with n equivalent binding sites, the velocity ( v ) equation is approximated by :  v=V [S]"/(K'+[S] ) n  max  where n is the number of binding sites per enzyme molecule (e.g., 4 in tetrameric CTP synthetase), and K' is a complex constant (Segel, 1975). K' contains Ks, (the intrinsic ES dissociation constant) as well as terms for subunit interactions that alter the intrinsic Ks. K' is sometimes called 'apparent K ' m  but differs from K  m  in that K'  does not equal that  concentration of substrate producing Pm /2 except when n=\. ax  The Hill equation can be rearranged to give the following linear form:  log {v /(K  max  -v)} = n log [S] - log K'  This form is useful because it allows determination of n. A plot of log {v /(V  max  versus log [S] gives a straight line with slope equal n.  -v)}  Actually, the observed n (called n , m  with the m for 'measured') obtained experimentally will be less than the theoretical n (the number of binding sites), because an assumption of infinite cooperativity was made in deriving the equation (Segel, 1975). The observed value of n  is known as the Hill coefficient; for  m  hemoglobin (which actually contains four binding sites) n  m  = 2.8. Generally, the observed Hill  coefficient is less than the maximum number of sites, representing a combination of the number of sites and the strength of interaction between them (Koshland, 1970). When n  m  = 1, the sites are independent, there is no allosteric interaction, and the  enzyme follows Michaelis-Menten kinetics. cooperativity. When n  m  When n  m  > 1, the enzyme shows positive  < 1, the enzyme shows negative cooperativity. Figure A . l , shows how  these three situations (independent  binding sites, positive cooperativity, and negative  cooperativity) are evidenced in the v vs. [S] plot, the double-reciprocal (Lineweaver-Burk plot), and the Hill plot.  In the v vs. [S] plot note that the sigmoidicity may be hard to detect;  155 sigmoidicity varies with the size of Ks and the value of «. In the double reciprocal plot, positive cooperativity gives a line that is concave upwards; negative cooperativity gives a line that is concave downwards.  [S]  Figure A . l A) Michaelis-Menten, B) double-reciprocal, and C) Hill plots for independent binding, positive cooperativity, and,negative cooperativity. The data are calculated from the Hill equation, in which V = l,K l, and n = 0.5,1, and 2. Adapted from Walsh (1979) ,=  max  156  Appendix 2 Coupled Enzyme Assays  One of the major problems for a significant study of initial-rate kinetics is often the method of assay. If possible, an assay should be accurate, sensitive, and convenient.  In  addition, the ability to continuously monitor a reaction process is of great value. Unfortunately, many reactions dp not produce changes in the spectral or other properties of the reactants and cannot be directly measured. To allow continuous assay of such reaction, the formation of a product can be measured by addition of an auxiliary enzyme that produces a measurable change. These thus termed 'coupled enzyme assay' affords a convenient, reliable method of measuring the steady-state activity of an enzyme. A number of approaches describing ways to ensure valid coupled assays have appeared in the literature (McClure, 1969; Rudolph et al., 1979). The basic problem is to determine what auxiliary enzymes will react at a rate that allows monitoring only of the steady-state concentration of the product(s) P of the reaction being studied. The systems have an inherent lag time prior to the steady state that must be analyzed and minimized. The simplest example of such system is:  +•  A  Primary Enzyme ( E l )  P  »- Q Auxiliary Enzyme (E2)  In order to ensure the validity for continuous coupled assay the reaction conditions required for optimal activity of one of the enzymes must not be detrimental for the performance of the other enzyme. Thus, if the optimal pH for the two enzymes differ significantly, or if one of the co-substrates required for one of the enzymes inhibits the other, then a continuous assay is not feasible. However, if this is not the case, then it is possible to measure the rate of the primary enzyme in a continuous fashion by monitoring the rate of formation of the product of the auxiliary enzyme Q.  157 The general approach to such assays has been to use a large excess of the auxiliary enzyme to ensure steady state conditions. The behavior of such a system was first treated quantitatively by McClure (1969) using the following assumptions: (1) kj, the rate constant for E l , is assumed to be an irreversible zero-order step. To meet this criterion, all substrates for E l must be saturating or only a small fraction of the initial amount of the substrates is turned over during the assay period. Irreversibility is assumed since P is continuously removed by E2 during the assay. (2) The second reaction is irreversible and first order with respect to P (rate constant ki). than the K  m  This necessitates that the steady state concentration of P is significantly lower for P with respect to E2 and that the other substrates for E2 be nearly saturating. If  the equilibrium for the reaction catalyzed by E2 lies to the right or only a small amount of reaction occurs so that the substrates for E2 are not depleted during the assay period, then irreversibility can be assumed.  Proper choice of the auxiliary enzyme will satisfy this  condition. With these assumptions its is possible to calculate the amount of E2 required for a theoretically correct assay.  158  References  Adams, E. (1976) Adv. Enzymol. Relat. Areas Mol. Biol. 44, 69-138 Anderson, J.S., Page, R.L. and Salo, W.L. (1972) J. Biol. Chem. 247, 2480-2485 Anderson, V . E . (1991) In Enzyme Mechanisms from Isotope Effects, Cook, P.F., Ed.; C R C Press: Cleveland, pp 389-417 Anderson, V . E . (1998) in Comprehensive Biological Catalysis: A Mechanistic Reference, Vol. 2: Reactions ofNucleophilic/Carbanionoid Carbon, Sinnott, M.L. Ed.; Academic Press: NewYork. pp. 115-133 Anet, F.A.L. (1960) J. Am. Chem. Soc. 82, 994-995 Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., and Struhl, K. (1992) in Short Protocols in Molecular Biology (A Compendium of Protocols from Current Protocols in Molecular Biology); Green Publishing Associates and John Wiley and Sons: New York. Avalos, M., Reyes, B., Cintas, P., Jimenez, J.L., Palacios, J.C. and Valencia, C. (1993) Tetrahedron Lett. 34, 1359-1362 Baddiley, J. (1989) BioEssays 10, 207-210 Bahnson, B.J. and Anderson, V . E . (1991) Biochemistry 30, 5894-5906 Ballardie, F.W., Capon, B., Dearie, W.M. and Foster, R.L. (1976) Carbohydr. Res. 49, 79-92 Baluja, A., Chase, H.B., Kenner, G.W. and Todd, A . (1960) J. Chem. Soc. 4678-4681 Banoub, J., Boullanger, P. and Lafont, D. (1992) Chem. Rev. 92, 1167-1195 Barber, G.A. and Hebda, P.A. (1982) Methods Enzymol. 83, 522-525 Batiz-Hernandez, H. and Bernheim, R. (1967) Prog. NMR Spect. 3, 63-85 Bergeron, R.J., Garlich, J.R. and Stolowich N.J., (1984) J. Org. Chem. 49, 2997-3001 Biciocchi, E., Clementi, S., and Sebastiani, G.V. (1979) J. Org. Chem. 44, 32-34 Biely, P. and Jeanloz, R.W. (1969) J. Biol. Chem. 244, 4928-4935 Bondi, A (1964) J. Phys. Chem. 68, 441-451 Boyer, P.D. (1959) Arch. Biochem. 82, 387-410  159 Bruice T.C. and Benkovic S.J. (1966) In 'Biorganic Mechanism' Vol. I, W. A . Benjamin, Inc., New York, 187-195 Bunnet, J.F. (1962) Angew. Chem. Int. Ed. Engl. 1, 225-233 Busca, P. and Martin, O.R. (1998) Tetrahedron Lett. 39, 8101-8104 Cardinale, G.J and Abeles R.H. (1968) Biochemistry 7, 3970-3978 Carey, F.A. and Sundberg, R.J. (1990) In 'Advanced Organic Chemistry' 3 Structure and Mechanisms, Plenum Press: New York, pp 368-373  rd  Ed., Part A:  Casero, F., Cipolla, L., Lay, L., Nicotra, F., Panza, L. and Russo, G. (1996) J. Org. Chem. 61, 3428-3432 Chari, R.V.J., Whitman, C P . , Kozarich, J.W., Ngai, K.-L., and Orston, L.N. (1987) J. Am. Chem. Soc. 109,5514-5519. Cohen J.S. (1983) In Magnetic Resonance in Biology, Vol. 2, John Wiley & Sons: New York, pp 1-50 Cohn, M and Hu, A . (1978) Proc. Natl. Acad. Sci. USA 75, 200-203 Cohn, M, Pearson, J.E., O'Connell, E.L. and Rose, L A . (1970) J. Am. Chem. Soc. 92, 40954098 Cole, F.E., Kalyanpur, M.G., and Stevens, C M . (1973) Biochemistry 12, 3346-3350 Conolly, B.A., Eckstein, F. and Grotjahn, L. (1984) Biochemistry 23, 2026-2031 Cori, C.F., Colowick, S.P. and Cori, G.T. (1937) J. Biol. Chem. 121, 465-467 Creighton, D.J. and Murthy, N.S.R.K. (1990) In The Enzymes, Vol. 19, Sigman, D.S. and Boyer, P.D., Eds.; Academic Press: San Diego, pp 397-403 Croteau, R.B., Shaskeu, J.J., Renstrom, B., Felton, N.M., Cane, D.E., Salto, A . and Chang, C. (1985) Biochemistry 24, 7077-7085 Cummings, R.D., Trowbridge, I.S. and Kornfeld, S. (1982) J. Biol. Chem., 257 13421-13427 Dale, J.W. (1998) In Molecular Genetics of Bacteria 3 pp. 221-239  rd  Ed.; John Wiley and Sons: NewYork.  Daniels, D.L., Plunkett, G. Ill, Burland, V . and Blattner, R.R. (1992) Science 257, 771-777 Deslongchamps, P. (1975) Tetrahedron 31, 2463-2490 Deupree, J.D., and Wood, W.A. (1970) J. Biol. Chem 245, 4700-4705  160 Deupree, J.D., and Wood, W.A. (1970)/. Biol. Chem 247, 3093-3097 Douderoff, M., Barker, H.A., and Hassid, W.Z., (1947) J. Biol. Chem. 168, 725-730. Dreuckhammer, D.G., Hennen, W.J., Pedersen, R.L., Barbas III, C.F., Gautheron, C.M., Krach, T. and Wong, C.-H. (1991) Synthesis, 449-525 England, S. (1958) J. Biol. Chem., 233 1003-1009 England, S., Britten, J.S., and Litowsky, I. (1967) J. Biol. Chem., 2255-2259 Farrar, T.C., Gutowsky, H.S., Alberty, R.A., and Miller, W.G. (1957) J. Am. Chem. Soc. 79, 3978-3980 Frey, P. (1987) in Pyridine Nucleotide Coenzymes, Dolphin, D., Poulson, R., and Avramovic, Eds.; John Wiley and Sons: New York. pp. 461-511. Furusawa, K., Sehine, M. and Hata, T. (1976) J. Chem. Soc, Perkin Trans. 1, 1711-1716 Gabriel, O. (1982) Methods Enzymol. 83, 332-353 Ge'rlt, J.A. and Gassman, P.G. (1992) J. Am. Chem. Soc. 114, 5928-5934 Ghose, C. and Raushel, F.M. (1985) Biochemistry 24, 5894-5902 Glaser, L. (1960) Biochem. Biophys. Acta 41, 534-536 Glaser, L. (1972) in The Enzymes, 3 355-380.  rd  ed., Boyer, P.D., Ed.; Academic Press: New York. pp.  Grimshaw, C.E., Sogo, S.G., Copley, S.D. and Knowles, J.R. (1984) J. Am. Chem Soc. 106, 2699-2700 Guthrie, J.P. and Kluger, R. (1993) J. Am. Chem. Soc. 115, 11569-11572 Gutowsky H.S. (1959) J. Chem. Phys. 31, 1683-1684 Hanson, K.R. and Rose, L A . (1963) Proc. Nat. Ac. Sci. 50, 981-988 Harrington, C R . and Baddiley, J. (1985) Eur. J. Biochem. 153, 639-645 Hecker, S. J. and Minich, M.L. (1990) J. Org. Chem., 55, 6051-6054 Heidlas, J.E., Watson, J L., Pale, P. and Whitesides, G.M (1992) J. Org. Chem. 57, 146-151. Hilscher, L.W., Hanson, C D . , Russel, D.H. and Raushel F.M. (1985) Biochemistry 24, 5888-5893  161 Hoberman, H.D., Havir, E.A., Rochovansky, O. and Ratner, S. (1964) J. Biol. Chem. 239, 3818-3820 Hung, S.-C. and Wong, C.-H. (1996) Tetrahedron Lett. 37, 4903-4906. Ichihara, N., Ishimoto, N. and Ito, E. (1974) FEBS Letters 40, 309-311 Imazawa, M. and Eckstein, F. (1979) J. Org. Chem. 44, 2039-2041. Jha, R. and Davis, J.T. (1995) Carbohydr. Res. 277, 125-134 Johnson, A . E . , Tanner, M.E. (1998) Biochemistry 37, 5746-5754 Jones, C S . and Kosman, D.J. (1980) J. Biol. Chem. 255, 11861-11869 Kaila, N., Blumenstein, M., Bielawska, H. and Franck, R.W. (1992) J. Org. Chem. 57, 45764578 Kampe, W (1966) Chem. Ber. 99, 593-599 Kawamura, T., Ichihara, N., Ishimoto, N. and Ito, E. (1975) Biochem. Biophys. Res. Comm. 66, 1506-1512 Kawamura, T., Kimura, T., Yamamori, S. and Ito, E. (1978) J. Biol. Chem. 253, 3595-3601 Kawamura, T., Ishimoto, N. and Ito, E. (1979) J. Biol. Chem. 254, 8457-8465 Kawamura, T., Ishimoto, N. and Ito, E. (1982) Methods Enzymol. 83, 515-519 Kenyon, G. L., Gerlt, J. A., Petsko, G. A . and Kozarich, J. W. (1995) Acc. Chem. Res. 28, 178186 Khorlin, A.Y., Shulman, M.L., Zurabyan, S.E., Privalova, I.M. and Kopaevich,Y.L. (1968) Izv. Akad. NaukSSSR, Ser. Khim. 9, 2094-2098 Kiino, D.R., Licudine, R., Wilt, K., Yang, D.H.C. and Rothman-Denes, L.B. (1993) J. Bacteriol. 175, 7074-7080 Kim, D.H. Tucker-Kellog, G.W., Lees, W.J., and Walsh, C T . (1996) Biochemistry 35, 54355440 Kirby, G.W., O'Loughlin, G.J., and Robins, D.J. (1975) J. Chem. Soc. Chem Commun. 402-403 Knapp, S., Vocadlo, D., Gao, Z., Kirk, B. Lou, J. and Withers, S.G. (1996) J. Am. Chem. Soc. 118, 6804-6805 Kochetov, N K. and Shibaev, V. N. (1973) Adv. Carbohydr. Chem. Biochem. 28, 307-399  162 Koshland, D.E. (1970) In 'The Ezymes' 3 vol.1, pp 341-347  rd  ed., P. Boyer ed., Academic Press : New York  Kunz, H. and Unverzagt, C. (1988) Angew. Chem. Int. Ed. Engl. 27, 1697-1699 Lai, E.C.K. and Withers, S. (1994) Biochemistry 33, 14743-14749 Lee, J.J., Asano, Y., Shieh, T.-L., Spreafico, F., Lee, K. and Floss, H.G. (1984) J. Am. Chem. Soc. 106, 3367-3368 Legler, G. (1990) Adv. Carbohydr. Chem. Biochem. 48, 349-357 Leloir, L.F. (1951) Arch. Biochem. Biophys. 33,186-194 Lowe, G., Sheppard, G., Sinnot, M.L. and Williams, A . (1967) Biochem. J. 104, 893-899 Lowe, G. and Sheppard, G. (1968) J. Chem. Soc. Chem. Commun. 529-530 MacDonald, D.L. (1962) J. Org. Chem. 27, 1107 MacDonough, M.W. and Wood, W.A. (1961) J. Biol. Chem. 236, 1220-1224 Marolda, C L . and Valvano, M.A. (1995) J. Bacteriol. Ill, Marquart, D.W. (1963)  5539-5546  Soc. Ind. Appl. Math. 11, 431 -441  Mash, E.A., Gurria, G.M., and Poulter, C D . (1981) J. Am. Chem. Soc. 103, 3927-3935 Mazumder, A., Gerlt., J.A., Rabow, L., Absalon, M.J. Stubbe, J., and Bolton, P.H. (1989) J. Am. Chem. Soc. 107, 8029-8030 McClure, W.R. (1969) Biochemistry 8, 2782-2786 Meier-Dieter, U., Barr, K., Starman, R., Hatch, L. and Rick, P.D. (1992) J. Biol. Chem. 267, 746-753 Messner, B., Eggerer, H., Cornforth, J.W. and Mallaby, R. (1975) Eur. J. Biochem. 53, 255-257 Meyer zu Reckendorf W. and Wassiliadau-Micheli, N. (1970) Chem. Ber., 103 1792-1796. Meynial, I., Paquet, V. and Combes, D. (1995) Anal. Chem. 67, 1627-1631 Michelson, A . M (1964) Biochem. Biophys. Acta. 91, 1-13 Midelfort, C.G. and Rose, L A . (1976) J. Biol. Chem. 251, 5881-5887 Miller, R.W. and Buchanan, J.M. (1962) J. Biol. Chem. 237, 491-496  163 Mizrahi, V., Henrie, R.N., Marlier, J.F., Johnson, K.A., and Benkovic, S.J. (1985) Biochemistry 24,4010-4018 Moffat J.G. and Khorana H.G. (1958) J. Am. Chem. Soc, 80, 3756-3761 More O'Ferral, R.A. (1970) J. Chem. Soc.(B) 274-277 Morgan, P. (1998) In "The Mechanism of UDP-N-Acetylglucosamine 2-Epimerase". Ph. D. Thesis, University of British Columbia Morgan, P.M., Sala, R.F. and Tanner, M.E. (1997) J. Am. Chem. Soc. 119, 10269-10277 Nakabayashi, S., Warren, C D . and Jeanloz, R.W. (1986) Carbohydr. Res. 150, C4-C10 Nakabayashi, S., Warren, C D . and Jeanloz, R.W. (1988) Carbohydr. Res. 174, 279-289 Newman, H. (1965) J. Org. Chem. 30, 1287-1288 Noyce, D.S. and Lane, C A . (1962) J. Am. Chem. Soc. 84, 1641-1644 Oehrlein, R., Hindsgaul, O and Palcic, M.M. (1993) Carbohydr. Res. 244, 149-159 Palmer, J.L. and Abeles, R.H. (1976) J. Biol. Chem. 251, 5817-5819 Palmer, J.L. and Abeles, R.H. (1979) J Biol. Chem. 254, 1217-1226 Parker, A.R., Moore, J.A., Schwab, J.M. and Davisson, V.J. (1995) J. Am. Chem. Soc. 117, 10605-10613 Parry, R.J. and Asconas, L.J. (1985) J. Am. Chem. Soc. 107, 1417-1418 Piszkiewics, D. and Bruice, T . C (1967) J. Am. Chem. Soc. 89, 6237-6243 Piszkiewics, D. and Bruice, T.C. (1968) J . Am. Chem. Soc. 90, 2156-2163 Putman, E.W. (1963) Methods Carbohydr. Chem. 2, 261-270. Raftery, M.A. and Rand-Meir, T. (1968) Biochemistry. 7, 3281-3289 Ramsey, N.F. (1952) Phys. Rev. 87, 1075-1079 Raushel, F.M. and Villafranca, J.J. (1988) CRC Crit. Rev. Biochem. 23, 1-26 Retey, J., Fierz, H. and Zeylemaker, W.P. (1970) FEBS Lett. 6, 203-205 Risley, J.M. and VanEtten, R.L. (1979) J. Am. Chem. Soc. 101, 252-253 Rittenberg, D. and Graff C. (1958) J. Am. Chem. Soc. 80, 3370-3372  164 Robitaille, P.-M.L., Robitaille, P.A., Brown, G.G. Jr. and Brown, G.G. (1991) J. Magn. Res. 92,73-84 Rose, I.A. (1966) Annu. Rev. Biochem. 35, 23-56 Rose, I.A. (1979) Adv. Enzymol. 50, 361-395 Roseman, S., Distler, J.J., Moffat, J.G and Khorana, H.G. J. Am. Chem. Soc, 83 (1961) 659663 Rosental, S. and Sharon, N. (1964) Biochem. Biophys. Acta 83, 378-380 Roy, R., Tropper, F.D. and Grand-Mitre, C. (1991) CanJ. Chem. 69, 1462-1467 Rudnick, G. and Abeles, R.H. (1975) Biochemistry 14, 4515-4522 Rudolph, F.B., Baugher, B.W. and Beissner, R.S. (1979) Methods Enzymol. 63, 22-41 Sabesan, S. andNeira, S. (1992) Carbohydr. Res. 223,169-185 Sala, R.F., MacKinnon, S., Palcic, M. M. and Tanner, M. E. (1998) Carbohydr. Res., 306,127136 Sala, R.F., Morgan, P.M. and Tanner, M.E. (1996) J. Am. Chem. Soc. 118, 3033-3034 Salo, W.L. (1976) Biochim. Biophys. Acta 452, 625-628 Salo, W.L. and Fletcher H.G. (1969) J. Org. Chem. 34, 3189-3191 Salo, W.L. and Fletcher, H.G. (1970) Biochemistry 9, 882-885 Sarma, R.H. and Lee, C.-H. (1976) Biochemistry. 15, 697-704 Sedgwick, B., Morris, C., and French, S.J. (1978) J. Chem. Soc. Chem. Commun. 193-194 Schachter, H. (1986) Biochem. Cell Biol. 64, 163-181 Schwab, J.M., Klassen, J.B., and Habib, A. (1986) J. Chem. Soc. Chem Commun. 357-358 Segel, LH. (1975) in Enzyme Kinetics: Behaviour and Analysis of Rapid Equilibrium and Steady-State Enzyme Systems; John Wiley and Sons: New York, pp 83-89 Sim, M.M., Kondo H. and Wong C.-H. (1993) J. Am. Chem. Soc. 115,2260-2267. Skarzynski, T., Kim, D.H., Lees, W.J., Walsh, C T . and Duncan, K. (1998) Biochemistry 37, 2572-2577 Sommar, K.M. and Ellis, D.B. (1972) Biochim. Biophys. Acta 268, 581-595 Spivak, C T . and Roseman, S. (1966) Methods Enzymol. 9, 612-615  165 Srivastava, G., Alton, G. and Hindsgaul O. (1990) Carbohydr. Res. 207, 259-276. Subramanian, S.S., Raghavendra, R. and Bhattacharyya, P.K. (1966) Indian J. Biochem. 3, 260267 Tanner, M.E., and Kenyon, G.L. (1998) in Comprehensive Biological Catalysis: A Mechanistic Reference, Vol. 2: Reactions of Nucleophilic/Carbanionoid Carbon, Sinnott, M.L. Ed.; Academic Press: New York. pp. 7-41. , Tanner, M.E. (1998) in Comprehensive Biological Catalysis: A Mechanistic Reference, Vol. 3: Radical Reactions and Oxidation/Reduction, Sinnott, M.L. Ed.; Academic Press: New York. pp. 76-82. Thoden, J.B., Hegeman, A . D . , Wesenger, G., Chapeau, M., Frey, P.A. and Holden, H.M. (1997) Biochemistry 36, 6294-6304 Vederas, J.C. (1980) J. Am. Chem. Soc. 102, 374-376 Walsh, C. (1979a) in Enzymatic Reaction Mechanisms; W.H. Freeman and Co.: San Francisco, pp 570-599 Walsh, C. (1979b) in Enzymatic Reaction Mechanisms; W.H. Freeman and Co.: San Francisco, pp 157-160 Warren, C D . and Jeanloz, R.W. (1977) Carbohydr. Res. 53, 67-84 Warren, C D . , Konami, Y and Jeanloz. R. W. (1973) Carbohydr. Res. 30, 257-279. Warren, C D . , Milat, M. L., Auge C. and Jeanloz. R. W. (1984) Carbohydr. Res. 126, 61-80. Widlanski, T.S., Bender, S.L. and Knowles, J.R. (1987) J! Am. Chem. Soc. 109, 1873-1875 Wierenga, R.K., De Maeyer, M.C.H. and Hoi, W.G.J. (1985) Biochemistry 24, 1346-1357 Wierenga, R.K., Terpstra, P. and Hoi, W.G.J. (1986) J. Mol. Biol. 187, 101-107 Wightman, R.H, Staunton, J., Battersby, A.R., and Hanson, K.R. (1972) J.Chem.Soc.,Perkin Trans. 1 2355-2364 Wilson, D.B. and Hogness, D.S. (1964) J. Biol. Chem. 239, 2469-2481 Wolfrom, M. L. and Conigliaro, P. J.(1969) Carbohydr. Res., 11 63-76. Wong, C.-H. and Whitesides, G.M. (1994) in Tetrahedron Organic Chemistry Series , Vol. 12: Enzymes in Synthetic Organic Chemistry, Baldwin, J.E. and Magnus, P. D. Eds. Elsevier Science, Oxford U.K. pp. 252-274 Wong, C.-H., Halcomb, R. L., Ichikawa, Y. and Kajimoto, T. (1995) Angew. Chem. Int. Ed. Engl. 34, 521- 546  166  Yamamoto, K (1973a) J . Biochem. 73, 631-635 Yamamoto, K (1973b) J. Biochem. 73, 749-753 Yamamoto, K (1974a) J. Biochem. 75, 659-661 Yamamoto, K (1974b) J. Biochem. 75, 931-933 Yamamoto, K (1974c) J. Biochem. 76, 385-390 Yamazaky, T. Warren, C D . , Herscovics, A . and Jeanloz, R . W . 2247-2252  (1981) Can. J. Chem. 59,  

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.831.1-0061874/manifest

Comment

Related Items