Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A detailed kinetic and mechanistic study of a retaining a[alpha]-1,4-galactosyltransferase from Neisseria… Ly, Hoa D. 2002

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

Item Metadata

Download

Media
831-ubc_2002-750450.pdf [ 8.93MB ]
Metadata
JSON: 831-1.0061316.json
JSON-LD: 831-1.0061316-ld.json
RDF/XML (Pretty): 831-1.0061316-rdf.xml
RDF/JSON: 831-1.0061316-rdf.json
Turtle: 831-1.0061316-turtle.txt
N-Triples: 831-1.0061316-rdf-ntriples.txt
Original Record: 831-1.0061316-source.json
Full Text
831-1.0061316-fulltext.txt
Citation
831-1.0061316.ris

Full Text

A DETAILED KINETIC AND MECHANISTIC STUDY OF A RETAINING a-l,4-GALACTOSYLTRANSFERASE FROM Neisseria meningitidis by Hoa D. Ly B.Sc, University of Waterloo, 1996 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 standards: THE UNIVERSITY OF BRITISH COLUMBIA September 2002 © Hoa D. Ly, 2002 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. The University of British Columbia. Vancouver, Canada Department DE-6 (2/88) ABSTRACT Lipopolysaccharyl a-l,4-galactosyltransferase C (EC 2.4.1.x, LgtC) from Neisseria meningitidis is an important enzyme in the biosynthesis of lipooligosaccharides in this bacterium. The reaction catalyzed by this enzyme is the transfer of a galactosyl moiety from uridine diphosphogalactose (UDPGal) to a lipooligosaccharide acceptor containing a galactose residue at the nonreducing terminus, forming a Gal-a-l,4-Gal linkage in the process. Because the reaction proceeds with net retention of stereochemistry at the centre undergoing substitution, LgtC is classified as a "retaining" glycosyltransferase. By analogy with the well studied retaining glycosidases, LgtC is therefore believed to follow a double displacement mechanism involving the formation and subsequent breakdown of a covalent glycosyl-enzyme intermediate. A detailed kinetic and mechanistic characterization of LgtC was carried out. This enzyme was found to be highly specific for the donor substrate, UDPGal, binding this compound with micromolar affinity. As for the acceptor substrate, the enzyme is far less selective, as a range of compounds was found to be capable of fulfilling this role. In the absence of suitable glycosyl acceptors, water was also found to be capable of functioning in such a capacity, with the result being hydrolysis of UDPGal to yield UDP and galactose. Since the reaction catalyzed by LgtC involves the binding of two substrates and the release of two products, an understanding of the order in which these events occur is also important to the characterization of this enzyme. On the basis of kinetic studies together with a UDPGal/UDP exchange study, LgtC was deduced to follow a sequential kinetic mechanism, which requires that all substrates be bound to the enzyme before any chemical catalysis and release of products can occur. Further clarification of ii the order of substrate binding and product release was then achieved through a series of inhibition studies with both substrate analogues and product. On the basis of these studies, LgtC was concluded to follow an ordered Bl Bl kinetic mechanism in which UDPGal binds first, followed by lactose. After catalysis, the trisaccharide is the first product to be released, followed by UDP. In our quest to understand the chemical mechanism of LgtC, much effort was directed towards the attempted trapping of the putative covalent glycosyl-enzyme intermediate and identification of the nucleophile involved in its formation. When these endeavours were unsuccessful, alternatives to the double displacement mechanism, wherein the nucleophile is not enzymic were then sought. With the aid of the three-dimensional structure of this enzyme, all candidate nucleophiles were identified and assessed for their ability to fulfill such a role. Unfortunately, none appeared to be likely to function in such a capacity. Another possibility is that LgtC does not employ a double displacement mechanism, but rather proceeds through an Swi-like mechanism involving the approach of the nucleophile from the same side of the reaction centre as the departing leaving group. However, such a mechanism is not well precedented and no concrete evidence is available to support its occurrence in LgtC. Therefore, at this point in time, the mechanism by which LgtC catalyzes its glycosyl transfer reaction is still a mystery. iii TABLE OF CONTENTS Abstract ii Table of contents iv List of tables ix List of figures x List of schemes xiii List of abbreviations xiv Acknowledgements xvii Dedication xviii 1 Introduction 2 1.1 GENERAL INTRODUCTION 2 1.2 GLYCOSYL TRANSFER 3 1.2.1 Classification of glycosidases and glycosyltransferases 3 1.2.2 Mechanistic implications of inverting and retaining glycosidases and glycosyltransferases 5 1.2.3 Glycosyl transfer in retaining glycosidases - key features of the double displacement mechanism 8 1.2.3.1 The catalytic residues..'. 8 1.2.3.2 Oxocarbenium ion-like transition states 12 1.2.3.3 Covalent glycosyl-enzyme intermediate 16 1.2.4 Mechanistic studies of glycosyltransferases 19 1.2.4.1 Mutagenic investigations of glycosyltransferases 20 1.2.4.2 X-ray crystallographic investigations of glycosyltransferases 21 iv 1.2.4.3 Mechanistic investigations of retaining glycosyltransferases 23 1.3 LLPOPOLYSACCHARYL OC-L,4-GALACTOSYLTRANSFERASE C 25 1.4 AIMS OF THIS THESIS 27 2 Kinetic characterization of LgtC 31 2.1 ACCEPTOR SUBSTRATE SPECIFICITY 31 2.1.1 Synthesis of various lactosides to function as alternate acceptor substrates : 31 2.1.2 Kinetic evaluation of various alternate acceptor substrates 32 2.1.3 Hydrolytic activity of LgtC in the absence of saccharide acceptors 34 2.2 DONOR SUBSTRATE SPECIFICITY 35 2.3 OC-GALACTOSYL FLUORIDE AS AN ALTERNATE DONOR SUBSTRATE 35 2.3.1 Synthesis of a-galactosylfluoride 36 2.3.2 Kinetic evaluation of a-galactosyl fluoride as an alternative donor substrate 37 2.4 DETERMINING THE KINETIC MECHANISM OF L G T C 40 2.4.1 UDPGal/UDP exchange as a diagnostic test of a ping pong kinetic mechanism 43 2.4.2 Inhibition studies with a substrate analogue 46 2.4.3 The use of product inhibition studies to distinguish between ordered and random BI BI kinetic mechanisms 48 3 An investigation of the chemical mechanism of LgtC 54 3.1 ATTEMPTS TO TRAP A COVALENT GLYCOSYL-ENZYME INTERMEDIATE 54 3.1.1 Incompetent acceptor substrate approach 54 3.1.1.1 Synthesis of various 4 '-deoxylactose derivatives 55 3.1.1.2 Kinetic evaluation of incompetent acceptor substrates and their employment for the trapping of the putative covalent glycosyl-enzyme intermediate 59 3.1.2 Fluorosugar approach 60 3.1.2.1 Synthesis of 5-fluoro-a-D-galactopyranosyl fluoride 61 v 3.1.2.2 Attempts to trap the covalent intermediate of LgtC with 5-fluoro-oc-D-galactopyranosyl fluoride 64 3.1.2.3 Synthesis of uridine 5 '-diphospho-(2 "-deoxy-2 "-fluoro)-a-D-galactopyranose 65 3.1.2.4 Kinetic evaluation of UDP-2FGal and its employment for the trapping of the putative covalent glycosyl-enzyme intermediate 66 3.2 REVELATIONS FROM THE X-RAY CRYSTAL STRUCTURE AND ITS IMPLICATIONS FOR CATALYSIS 68 3.2.1 Evaluation of the 3'- and 6 '-hydroxyls of the acceptor lactose as possible nucleophiles 73 3.2.1.1 Synthesis of 6'-deoxylactose 73 3.2.1.2 Synthesis of the putative intermediates, Gal-(3-l,6-Lac and Gal-p-1,3-Lac 74 3.2.1.3 Evaluating the likelihood that the catalytic nucleophile is a hydroxyl group from the acceptor substrate •. 77 3.2.2 Evaluation of Glnl89 as the possible nucleophile 79 3.2.2.1 Kinetic analyses of Glnl89 mutants of LgtC 83 3.3 INVESTIGATION OF A POTENTIAL ANHYDROSUGAR INTERMEDIATE 86 3.3.1 Synthesis of (3-1,4- and/3-1,6-anydrogalactose 87 3.3.2 Evaluating the possibility that the intermediate in the mechanism of LgtC is either a (3-1,4-anhydrogalactose or a (3-1,6-anhydrogalactose species 88 3.4 PROBING THE EXISTENCE OF A LONG-LIVED INTERMEDIATE BY POSITIONAL ISOTOPE EXCHANGE 90 3.4.1 Synthesis of uridine 5'-diphospho-(l "-180)-a-D-galactopyranose 91 3.4.2 Search for Positional Isotope Exchange in the LgtC catalyzed reaction.. 94 3.5 A N ALTERNATIVE TO THE DOUBLE DISPLACEMENT MECHANISM - THE S NI MECHANISM 96 3.6 CONCLUDING REMARKS 102 Materials and methods 107 4.1 GENERAL 107 VI 4.2 SYNTHESIS 107 4.2.1 Lactoside acceptor substrates (Scheme 2.1) 107 4.2.2 cc-Galactosyl fluoride (Scheme 2.2) 114 4.2.3 Incompetent acceptor substrates (Schemes 3.1 and 3.2) 116 4.2.3.1 4'-Deoxylactose 116 4.2.3.2 Benzyl 4'-deoxy-4'-fluorolactoside 120 4.2.3.3 Benzyl 4'-deoxylactoside 126 4.2.4 Fluorinated donor substrate analogues (Schemes 3.3 and 3.4) 129 4.2.4.1 5-Fluoro-a-D-galactopyranosyl fluoride 129 A.2A.2 Uridine 5'-diphospho-(2"-deoxy-2"-fluoro)-a-D-galactopyranose.. 131 4.2.5 6'-Deoxylactose (Scheme 3.5) 135 4.2.6 Gal-pJ-Lac trisaccharides (Scheme 3.6) 136 4.2.7 1,6-Anhydrogalactose (Scheme 3.7) 142 4.2.8 180-UDPGal (Scheme 3.8) 143 4.3 E N Z Y M A T I C 146 4.3.1 Purification of LgtC 146 4.3.2 Enzyme kinetics using a spectrophotometry assay 147 4.3.3 Enzyme Kinetics with a-GalF as the donor substrate 148 4.3.4 UDPGal/UDP Exchange 148 4.3.5 Trapping experiments 149 4.3.5.1 Incompetent acceptor substrate approach 149 4.3.5.2 Fluorosugar approach 150 4.3.6 Evaluation of various compounds as potential intermediates of the LgtC reaction 150 4.3.6.1 Gal-P-Lactoside intermediates 150 4.3.6.2 Anhydrosugar intermediates 151 4.3.7 Positional Isotope Exchange 151 Appendix A - Fundamentals of enzyme kinetics 154 A-l A HISTORICAL PERSPECTIVE 154 A-2 B A S I C EQUATIONS 154 vii A-3 REVERSIBLE INHIBITION 158 A-4 BISUBSTRATE REACTIONS 162 Appendix B - Graphical representation of kinetic data 166 References 171 viii LIST OF TABLES Table 2.1. Kinetic parameters for the utilization of'various acceptor substrates by LgtC at a fixed, saturating concentrations of the donor substrate, UDPGal 33 Table 2.2. Expected patterns of inhibition for analogues of substrates A and B in ping pong, ordered and random Bl Bl kinetic mechanisms. C = competitive, NC = noncompetitive, UC = uncompetitive [92] 47 Table 2.3. Expected product inhibition patterns in an ordered, rapid equilibrium random and nonrapid equilibrium random Bl Bl kinetic mechanisms. C = competitive, NC = noncompetitive, NI = no inhibition, UC = uncompetitive [92] 50 Table 3.1. Inhibition constants for various acceptor analogues with respect to varying concentrations of lactose and a fixed, saturating concentration (300 pM) of UDPGal 59 Table 3.2. Kinetic parameters of wild type LgtC and the various Gin 189 mutants for the substrates lactose and UDPGal 84 ix LIST OF FIGURES Figure 1.1. Stereochemical outcome of glycsosyltransferase-catalyzed reactions resulting in either (a) inversion or (b) retention of anomeric configuration 4 Figure 1.2. Proposed chemical mechanism for (a) inverting and (b) retaining glycosyltransferases 7 Figure 1.3. Chemical rescue of activity in a (a) nucleophile mutant and an (b) acid/base mutant glycosidase by azide 10 Figure 1.4. Examples of various transition state analogue inhibitors of glycosidases 14 Figure 1.5. The use of positional isotope exchange as a means of detecting the formation of a transient intermediate in the reactions of sucrose synthetase and glycogen synthetase. The darkened circles are representative of 1 8 0 24 Figure 1.6. The reaction catalyzed by Lipopolysaccharyl a-l,4-galactosyltransferase C (LgtC) 26 Figure 2.1. Plot of rate as a function of a-GalF concentration for the LgtC-catalyzed transfer of galactose from a-GalF in the presence of a fixed (65 mM) concentration of lactose 38 Figure 2.2. Cleland notation for describing the order of substrate binding and product release in a (a) ping pong (b) ordered and (c) random Bl Bl kinetic mechanism 40 Figure 2.3. Expected patterns in the double reciprocal plot for a (a) ping pong and (b) either an ordered or random Bl Bl kinetic mechanism 42 Figure 2.4. Proton decoupled 3 1P NMR spectrum showing the doublet arising from the P-phosphorus of 180-UDPGal (at ~ 85% isotope incorporation) (a) before incubation with LgtC, (b) after incubation with LgtC and (c) with unlabeled UDPGal added to the sample that had been incubated with LgtC 45 Figure 2.5. Double reciprocal plot for the inhibition of LgtC by UDPGlc at varying concentrations of lactose and a fixed concentration (250 pM) of UDPGal. The concentrations of UDPGlc were 0 (•), 50 (O ), 100 (•) and 200 pM (•) 46 x Figure 2.6. Double reciprocal plot showing the inhibition of LgtC by the product Gal-oc-1,4-Lac at varying concentrations of UDPGal and a fixed concentration (120 mM) of lactose. The concentrations of the product trisaccharide were as follows: 0 (•), 15 (O ), 20 (•), 30 (•), and 40 mM (•) 49 Figure 3.1. Conformations of the (a) 5-fhioro-D-galacto epimer and the (b) 5-fluoro-L-altro epimer. Haworth projections are shown along the C-4, C-5 axis 63 Figure 3.2. Double reciprocal plot showing the inhibition of LgtC by UDP-2FGal at varying concentrations of (a) the donor substrate UDPGal and (b) the acceptor substrate lactose. When the UDPGal concentration was varied, the concentrations of UDP-2FGal were 0 (•), 6.25 (O ), 12.5 (•) and 25 pM (•). When the concentration of lactose was varied, the concentrations of UDP-2FGal were 0 (•), 37.5 (O ), 75 (•), 150 (A) and 300 pM (•) 67 Figure 3.3. Structural view of the LgtC-UDP-2FGal-4dLac ternary complex showing the large N-terminal catalytic domain and the smaller C-terminal membrane attachment domain 69 Figure 3.4. A molecular surface representation of LgtC showing the deeply buried donor substrate analogue (UDP-2FGal) and the more solvent exposed acceptor substrate analogue (4dLac) 70 Figure 3.5. A (a) schematic representation of the active site of LgtC showing the interactions between the enzyme and the substrate analogues and (b) a view of residues within close proximity to the CI" reaction centre 72 Figure 3.6. The catalytic mechanism of A'-acetylhexosaminidases from glycosidase families 18 and 20 80 Figure 3.7. The reaction of glycogen phosphorylase 81 Figure 3.8. An illustration of the proposed role of the coenzyme phosphate as an electrophile in the postulated catalytic mechanism of glycogen phosphorylase [124] 82 Figure 3.9. A possible anhydrosugar mediated double displacement mechanism where the role of the nucleophile in the first step of the reaction is played either by the (a) 4"-OH or the (b) 6"-OH of UDPGal 87 xi Figure 3.10. Cleavage of the CI "-Ol" bond in the substrate UDPGal followed by rotation of the p-phosphate and re-formation of the bond will result in the scrambling of the 1 8 0 (represented by the darkened atom) from the bridging to a nonbridging position in the p-phosphate 91 Figure 3.11. A proposed route for the production of the triphosphate-linked uridine dimer in the tetrazole aided coupling of I80-galactopyranosyl phosphate with UMP-morpholidate 93 Figure 3.12. The proposed ionization steps associated with the SNI mechanism for the decomposition of certain alkyl chlorosulfites to yield alkyl halides with net retention of stereochemistry 98 Figure 3.13. Proposed transition state for the solvolysis of oc-glucosyl fluoride by trifluoroethanol to yield a glucoside product of retained anomeric configuration [143] 99 Figure 3.14. A depiction of a speculative SNi-like mechanism for the LgtC-catalyzed reaction 101 Figure 3.15. A view of the two substrates in the three-dimensional structure of LgtC in complex with UDP-2FGal and lactose. The distance (in angstroms) from the nucleophilic 4'-OH of lactose to the reactive CI" centre and the exocyclic O l " are also indicated 102 xii LIST OF SCHEMES Scheme 2.1. An outline of the synthesis of various lactosides to serve as potential acceptor substrates for LgtC 32 Scheme 2.2. An outline of the synthesis of a-galactopyranosyl fluoride (a-GalF, (2.13)). 36 Scheme 3.1. An outline of the synthesis of 4'-deoxylactose (4dLac, (3.6)) 56 Scheme 3.2. An outline of the synthesis of benzyl 4'-deoxy-4'-fluoro-p-lactoside (Bn4FLac, (3.13)) and benzyl 4'-deoxy-p-lactoside (Bn4dLac, (3.17)) 58 Scheme 3.3. An outline of the synthesis of 5-fluoro-a-D-galactopyranosyl fluoride (5FGalF, (3.20)) 62 Scheme 3.4. An outline of the synthesis of uridine 5'-diphospho-(2"-deoxy-2"-fluoro)-a-D-galactopyranose (UDP-2FGal, (3.24)) 66 Scheme 3.5. An outline of the synthesis of 6'-deoxylactose (6dLac, (3.27)) 74 Scheme 3.6. An outline of the synthesis of Gal-p-1,6-Lac (3.33) 76 Scheme 3.7. An outline of the synthesis of P-l ,6-anhydrogalactose (3.37) 88 18 Scheme 3.8. An outline of the synthesis of uridine 5 '-diphospho-(l "- O)-0C-D-galactopyranose (18Q-UDPGal, (3.40)) 92 xiii LIST OF ABBREVIATIONS 4dLac 4 '-deoxylactose 5FGalF 5-fluoro-a-D-galactopyranosyl fluoride A C 2 O acetic anhydride AcOH acetic acid AIBN 2,2'-azobisisobutyronitrile Bn4dLac benzyl 4 '-deoxylactoside Bn4FLac benzyl 4 '-deoxy-4 '-fluorolactoside BnOH benzyl alcohol Bu3SnH tributyltin hydride CAN eerie ammonium nitrate DAST diethylaminosulfur trifluoride DBU 1,8-diazabicyclo[5.4.0]undec-7-ene DCE dichloroethane DMAP 4-dimethylaminopyridine DMF dimethyl formamide DTT dithiothreitol EC Enzyme Commission (classification number) of the International Union of Biochemistry ESMS electrospray mass spectrometry EtOH ethanol Gal galactose a-GalF a-galactopyranosyl fluoride Glc glucose Glc-l-P glucose-1-phosphate HF-pyridine hydrogen fluoride-pyridine HPLC high performance liquid chromatography kDa kiloDalton KDO 2-keto-3-deoxyoctulosonic acid LgtC lipopolysaccharyl a-l,4-galactosyltransferase xiv LOS lipooligosaccharide MeCN acetonitrile MeOH methanol NaOAc sodium acetate NaOMe sodium methoxide NBS iV-bromosuccinimide N H 4 O A C ammonium acetate NMO 4-methylmorpholine iV-oxide NMR nuclear magnetic resonance Pi phosphate PIX positional isotope exchange PLP pyridoxal phosphate TASF tris(dimethylamino)sulfur (trimethylsilyl)difluoride Tf^O trifluoromethanesulfonic anhydride THF tetrahydrofuran TLC thin layer chromatography . • TsOH /7-toluenesulfonic acid UDP uridine diphosphate UDP-2FGal uridine 5 '-diphospho-(2' '-deoxy-2' '-fluoro)-a-D-galactopyranose UDPGal uridine 5 '-diphosphogalactose UDPGlc uridine 5 '-diphosphoglucose UMP uridine monophosphate xv Amino Acid Abbreviations Ala A alanine Arg R arginine Asn N asparagine Asp D aspartic acid Cys C cysteine Glu E glutamic acid Gin Q glutamine Gly G glycine His H histidine He I isoleucine Leu L leucine Lys K lysine Met M methionine Phe F phenylalanine Pro P proline Ser S serine Thr T threonine Trp W tryptophan Tyr Y tyrosine Val V valine XVI ACKNOWLEDGEMENTS It is with the greatest sincerity that 1 express my thanks to my supervisor, Professor Stephen G. Withers for all of his support and encouragement throughout my time in his laboratory. His insights and enthusiasm were a constant source of inspiration for me and I consider it a great priviledge to have developed myself as a scientist under his tutelage. Without the provision of enzyme from Dr. Warren W. Wakarchuk and Dr. Manuela Dieckelmann of the National Research Council of Canada, none of this work would have been possible. For that, I am greatly in their debt. In addition, I would also like to extend my gratitude to Professor Natalie C. J. Strynadka and Dr. Karina Persson of the Department of Biochemistry and Molecular Biology at the University of British Columbia for a fruitful collaboration, which resulted in the solution of the three-dimensional structure of the enzyme. The generous gift of UDPGal from Professor Monica M . Palcic of the University of Alberta and of p-l,4-anhydrogalactose from Professor Martin E. Tanner of the University of British Columbia are also gratefully acknowledged. I would also like to extend a warm, heartfelt thank you to all my fellow labmates, both past and present, for all their assistance and more importantly, their friendship over the years. The lab was a fun place to work because of all of you. The assistance of Dr. Shouming He, Ms. Karen Rupitz, Dr. Nick Burlinson, Ms. Marietta Austria and Ms. Lianne Diarge for all their technical support is also greatly appreciated. To all my family and friends, especially my parents, my brother and my sister, thank you all for your love, support and unwavering belief in me. And finally to my darling Emily, thank you for the joy you have brought to my life when you chose to become a part of it. "Delight yourself in the Lord And He will give you the desires of your heart" Psalm 37:4 xvii p entente , (Suae and fDlcfi £p X V l l l CHAPTER 1 INTRODUCTION Chapter I Introduction 2 1 INTRODUCTION 1.1 General Introduction In terms of biomass, carbohydrates are by far the most abundant of all macromolecules. In addition, these complex macromolecules are also remarkably diverse in their structures. This diversity is due in large part to the fact that these compounds are not synthesized according to a template, as is the case with other biopolymers such as proteins and nucleic acids. Instead, it is often the substrate specificities of the individual enzymes involved in their synthesis that dictate the final structure of the carbohydrate. The vast assortment of unique structural arrays thus generated makes these biopolymers highly amenable to a wide range of biological functions. Traditionally, carbohydrates are well known for their role in the storage of energy and in the provision of structural integrity. Structures such as starch and glycogen serve as important reservoirs of glucose in plants and animals while others such as cellulose and chitin are integral components in the cell walls of higher plants and the exoskeletons of insects respectively. More recently, the importance of carbohydrates in other biological functions, especially those involving molecular recognition, is also being recognized. In particular, these complex macromolecules often play key roles in many intercellular communication and signal transduction events, which are vital to the development of cells and the regulation of cellular activity [1]. In addition, many pathogens have also evolved to recognize and bind to specific carbohydrates displayed on the surface of cells as a means of gaining entry into the host cell. As a result of their biological importance and potential therapeutic value, carbohydrates, along with the Chapter I Introduction 3 enzymes responsible for their metabolism, are now the subjects of intensive research efforts. 1.2 Glycosyl Transfer Glycosidases and glycosyl transferases are enzymes that are responsible for the transfer of glycosyl moieties from a donor sugar to an acceptor. In the case of glycosidases this acceptor is water, the result being hydrolysis. In the case of other transferases this acceptor is typically an alcohol functionality from another sugar, but can also be from a lipid, an aryl moiety or a range of other components of glycoconjugates. Transfer can also occur to nitrogen or sulfur nucleophiles, as in, for example, nucleotides and plant thioglycosides. While the term glycosyltransferase therefore strictly includes glycosidases it is generally reserved for enzymes catalyzing glycosyl transfer to an acceptor other than water, while the term glycosidase or glycosyl hydrolase is used for enzymes catalyzing the hydrolytic process. The non-hydrolytic glycosyltransferases can be further classified into two groups, those that utilize nucleoside phosphosugars as the donor substrate (NPsugar-dependent) and those that use oligosaccharides as donors. Since the latter group is best identified as transglycosylases, then the use of the term glycosyltransferases in the remainder of this thesis will generally be in reference to the NPsugar-dependent enzymes. 1.2.1 Classification of glycosidases and glycosyltransferases Glycosidases and glycosyltransferases have traditionally been classified on the basis of three criteria: glycone specificity, anomeric configuration of the glycosidic linkage that is formed or broken, and the stereochemical outcome of the reaction catalyzed. With Chapter 1 Introduction 4 regards to glycone specificity, classification is based on the sugar unit (or glycone) for which the enzyme is most reactive. For example, galactosyltransferases would be most active towards the transfer of galactose residues while glucosyltransferases would prefer to transfer glucose units. In addition to glycone specificity, glycosidases and glycosyltransferases are also classified as either " a " or "P" according to the anomeric configuration of the linkage that is broken in the substrate (for glycosidases) or the linkage that is formed in the product (for glycosyltransferases). Finally, the terms "inverting" and "retaining" are also very commonly used to categorize glycosidases and glycosyltransferases. This system of classification is based on the relative stereochemical outcome of the reaction. In reactions catalyzed by retaining enzymes, the anomeric configuration of the product is identical to that of the substrate while inverting enzymes generate products of opposite anomeric configuration in comparison to that of the substrate (Figure 1 . 1 ) . (a) + R'OH OR OR' Figure 1.1. Stereochemical outcome of glycsosyltransferase-catalyzed reactions resulting in either (a) inversion or (b) retention of anomeric configuration. Chapter 1 Introduction 5 To complement these traditional criteria of classification, which tend to focus on the function of the enzyme, glycosidases and glycosyltransferases have also been grouped into distinct families based on similarities in their amino acid sequences. For glycosidases, this method of classification has now placed well over 5000 sequences into 87 different families [2] while for glycosyltransferases, this number now includes more than 2500 sequences and 52 distinct families [3]. Details of these "carbohydrate active enzyme" families are available through the World Wide Web at http://afrnb.cnrs-mrs.fr/~cazy/CAZY/index.html. The discriminatory power of this classification method has been confirmed by the similarity of the three-dimensional structures [4] and conserved chemical mechanisms [5] amongst enzymes within the same family. A significant advantage of this method of classification is that it allows for the logical grouping of enzymes of different EC numbers into specific families thereby offering insights into the divergent evolution of these enzymes [6]. On the other hand, enzymes that have been grouped by function but belong to several different families may reflect convergent evolution [6]. 1.2.2 Mechanistic implications of inverting and retaining glycosidases and glycosyltransferases From a mechanistic standpoint, the classification of glycosidases and glycosyltransferases based on the stereochemical outcome of the reaction is most useful. The reaction catalyzed in all cases is a substitution at a chiral acetal or ketal centre and can occur with either inversion or retention of stereochemistry. These two different stereochemical outcomes are a direct reflection of the two distinct mechanisms by which these two classes of enzymes operate [7]. In the case of inverting enzymes, glycosyl Chapter I Introduction 6 transfer is believed to take place via a direct displacement mechanism where the donor sugar (ie. the glycosyl moiety being transferred) is attacked directly by the acceptor (ie. the moiety to which the donor sugar is transferred) (Figure 1.2a). The reaction proceeds through an oxocarbenium ion-like transition state and is assisted by two important catalytic residues. One residue acts as a general acid to aid in the departure of the aglycone portion of the donor sugar while the second functions as a general base to deprotonate the nucleophilic hydroxyl of the incoming acceptor substrate. The result of this single step mechanism is a product of inverted anomeric configuration relative to the donor substrate. For retaining enzymes, catalysis is believed to proceed via a double displacement mechanism involving the formation and subsequent breakdown of a covalent glycosyl-enzyme intermediate with both steps proceeding through oxocarbenium ion-like transition states (Figure 1.2b). Once again, two important catalytic residues are involved but each has slightly different roles in this case. In the first step of the reaction, one residue functions as a nucleophile, attacking the donor substrate and displacing the aglycone portion of the molecule. This departure of the aglycone leaving group is assisted by a second catalytic residue, which provides general acid catalysis. After functioning as the general acid catalyst for the first step, this same residue is then poised to act as a general base to assist in deprotonating the nucleophilic hydroxyl of the acceptor substrate in the second step of the mechanism. As a consequence of inverting the stereochemistry of the anomeric center twice through two displacement reactions, the overall result is the generation of a product of retained anomeric configuration relative to that of the donor substrate. Chapter 1 Introduction 7 Figure 1.2. Proposed chemical mechanism for (a) inverting and (b) retaining glycosyltransferases Chapter 1 Introduction : 8 1.2.3 Glycosyl transfer in retaining glycosidases - key features of the double displacement mechanism As alluded to earlier, some of the key features of the double displacement mechanism include: the two catalytic residues (a nucleophile and a general acid/base catalyst), oxocarbenium ion-like transition states and the formation of a covalent glycosyl-enzyme intermediate. It is therefore not surprising that mechanistic studies of retaining glycosidases and glycosyltransferases have focused primarily on these particular aspects of the mechanism. From kinetic isotope effects to X-ray crystallography, support for the double displacement mechanism is well documented for retaining glycosidases. Unfortunately, the same cannot be said of retaining glycosyltransferases. Mechanistic studies of these enzymes have been largely hindered in the past by the limited availability and stability of the proteins. As a result, much of the insight into the mechanism of glycosyltransferases is drawn primarily from the extensive body of work surrounding glycosidases. However, this situation is currently changing as the issues of availability and stability surrounding glycosyltransferases have been addressed through high-level recombinant expression of these enzymes. 1.2.3.1 The catalytic residues An important aspect of any enzyme mechanistic study is the identification of residues that are essential to catalysis. In this regard, a great deal of effort has gone into the search for and identification of the two catalytic residues (the nucleophile and acid/base catalyst) involved in the double displacement mechanism of glycosyl transfer. Through sequence alignments of enzymes within a family, conserved amino acid residues were identified as possible candidates to function in such a capacity. Taking advantage Chapter 1 Introduction 9 of the ability of site-directed mutagenesis to make changes to a specific amino acid within a protein, the importance of these conserved residues to catalysis was probed by replacing them with chemically inert substitutes. Kinetic analyses of mutants thus generated have been instrumental in identifying the nucleophile and the acid/base catalytic residues of numerous retaining glycosidases ([8] and references therein). With the exception of some hexosaminidases, which have been shown to follow a slightly different mechanism [9, 10], a pair of carboxylic acid residues (either aspartic acid - Asp or glutamic acid - Glu) was seen to fulfill these roles in all cases. As expected, mutation of either residue is often accompanied by an almost complete loss of activity in the resulting enzyme [8]. As a means of assigning the respective functions of the two catalytic residues from mutagenesis studies, a method based on chemical rescue of activity in mutant enzymes was developed [11]. In this strategy, conserved Asp and Glu residues are individually mutated to an alanine (Ala) residue and the resulting mutants are screened for activity with an activated substrate both in the absence and presence of azide anion. Mutants showing dramatically increased activity in the presence of azide are prime candidates to function in one of these two catalytic roles and are then further differentiated by isolating and characterizing the glycosyl azide product. In the case of the nucleophile mutant, the anomeric configuration of the azide product will be inverted relative to that of the substrate since the azide will simply occupy the space vacated by the catalytic nucleophile and displace the aglycone leaving group of the activated substrate (Figure 1.3a). By contrast, the acid/base mutant will be the one that has made the glycosyl azide of retained anomeric configuration as the azide is accommodated in Chapter 1 Introduction 10 the space previously occupied by the acid/base catalyst and attacks the covalent glycosyl-enzyme intermediate formed in the first step of the reaction (Figure 1.3b). This strategy has been used with much success in assigning the catalytic roles of Asp and Glu residues in a number of fj-retaining glycosidases ([8] and references therein). Once the catalytic residues are assigned for one member of a family, sequence alignment can then be used to identify these important residues in other members of the same family. Figure 1.3. Chemical rescue of activity in a (a) nucleophile mutant and an (b) acid/base mutant glycosidase by azide. In addition to site-directed mutagenesis studies, which have on occasion resulted in wrongful assignments of the catalytic residues [12, 13], another often-used means of predicting the essential residues in an enzyme is X-ray crystallography. The power of this technique is its ability to provide a high resolution, three-dimensional image of the a) Chapter I Introduction protein. In the many structures of retaining glycosidases that have been obtained thus far, the two residues that are positioned to function as the nucleophile and acid/base catalyst are indeed shown to be either aspartic or glutamic acids [14, 15]. The structural data also reveal that the distance between these two carboxylic acids is only approximately 0.55 nm. This proximity of the two residues to each other and therefore to the substrate is consistent with a need for the direct attack of the nucleophile on the substrate. While X-ray crystallography has provided many useful insights into the structures and functions of a diverse number of glycosidases, this technique is not always available for every enzyme. Moreover, it is also sometimes difficult to identify the important catalytic residues unless a substrate, a substrate analogue or an inhibitor is bound in the active site. Another popular and reliable approach to identifying and assigning the roles of the catalytic residues is the use of mechanism-based inactivators. For this purpose, fluorosugars have proved to be invaluable tools, particularly in the identification of the catalytic nucleophile [16, 17]. These mechanism-based inactivators function by forming relatively stable glycosyl-enzyme intermediates. The presence of a highly electron withdrawing fluorine substituent adjacent to the center of developing positive charge will inductively destabilize the oxocarbenium ion-like transition states through which both the glycosylation and deglycosylation steps proceed. As a consequence, both steps of the double displacement mechanism are slowed. However, through incorporation of a good leaving group at C-l , the glycosylation rate can be increased, allowing the fluoroglycosyl-enzyme to be trapped. The identity of the nucleophilic amino acid residue involved in the formation of the intermediate is then determined by mass spectrometric methods following proteolysis of the protein. This approach has led to the unequivocal Chapter 1 Introduction 12_ assignment of the catalytic nucleophile in a large number of retaining glycosidases belonging to many different families [18, 19]. Once again, the identity of this residue is revealed to be a carboxylic acid in all cases. 1.2.3.2 Oxocarbenium ion-like transition states A second feature of the double displacement mechanism is the structure of the oxocarbenium ion-like transition states through which both the glycosylation and deglycosylation steps are believed to proceed. Bond cleavage in each step of the reaction leads to significant positive charge development at both Cl and 05 of the sugar ring (Figure 1.2b). To accommodate an effective overlap between the lone pair of electrons on the oxygen with the resulting vacant p orbital at the reaction center, the pyranose ring is distorted into a half-chair conformation. These characteristics of the oxocarbenium ion-like transition states have been probed extensively for retaining glycosidases through kinetic isotope effects and inhibition studies with transition state analogues. In the double displacement mechanism of glycosyl transfer, much of the evidence for positive charge development in the transition state has come primarily from measurements of secondary a-deuterium kinetic isotope effects. Through the careful selection of substrates with varying degrees of reactivity at the aglycone, these measurements have been reported for both the glycosylation and deglycosylation steps of several retaining glycosidases [20-23]. In all cases, the values obtained were found to be greater than unity, suggesting that positive charge development is indeed present in the transition states of both steps. With regards to the possible distribution of this charge between Cl and 05, recent solvolysis studies of 2,4-dinitrophenyl f}-D-glycopyranosides Chapter 1 Introduction 1_3_ have shown that, at least in solution, the formal positive charge in a glycosyl cation resides mainly with the oxygen atom [24]. Validation of the oxocarbenium ion-like nature of the transition state has also been achieved through inhibition studies with analogues designed to mimic this structure. The concept of transition state analogy, which was first advocated by Pauling in 1946 [25] and more rigorously defined by Lienhard [26] and Wolfenden [27], assumes that interactions between an enzyme and its substrate are both increased and optimized at the transition state. As such, it is generally accepted that analogues mimicking the transition state will bind tightly to the enzyme although a quantitative basis for making such a distinction remains clouded [28]. The necessary features to mimic an oxocarbenium ion-like transition state are a positive charge at the positions corresponding to either 05 or C l , an sp2 hybridized center at the position corresponding to Cl and a half-chair conformation where the atoms corresponding to C5, 05, Cl and C2 are coplanar. Analogues that possess some or all of these characteristics and are also bound tightly by the enzyme provide supporting evidence for the existence of oxocarbenium ion-like transition states (Figure 1.4). Gluconolactone(l.l) Gluconolactam (1.2) Nojirimycin (1.3) 1-Deoxynojirimycin (1.4) Nagstatin(l.lO) Glucotetrazole (1.11) 1,4-Glucoimidazole (1.12) 1,2,3-Glucotriazole (1.13) Figure 1.4. Examples of various transition state analogue inhibitors of glycosidases. Gluconolactone (1.1) and gluconolactam (1.2) were the first glycosidase inhibitors to be classified as transition state analogues. However, despite the similarities of their resonance charge distributions and half-chair conformations to an oxocarbenium ion, both compounds were found to bind with only 2 to 3 orders of magnitude greater affinity to retaining glycosidases than does the parent glycoside [ 2 9 ] . A possible explanation for the lack of inhibitory potency is that the positive charge of the oxocarbenium ion is not well mimicked by the resonance charge distribution in either of the compounds. The ability to imitate this positive charge appears to be extremely Chapter 1 Introduction 15 important, as compounds that can do this effectively tend to bind with much greater affinity to the target enzymes. Examples of these compounds include members of the nojirimycin class of inhibitors and also members of the isofagomine class of inhibitors. In nojirimycin (1.3) and 1-deoxy-nojirimycin (1.4), the endocyclic oxygen of the sugar ring is replaced by a nitrogen atom. This substitution presumably confers a positive charge on the nitrogen upon protonation allowing it to emulate the developing positive charge on 05 of the oxocarbenium ion-like transition state. Both these compounds have indeed been shown to function as potent inhibitors of a- and p-glucosidases [29], However, a rigorous study of the inhibition of Agrobacterium sp. P-glucosidase by these two compounds has raised doubts about their classification as transition state analogues [30]. Compounds of the isofagomine family (1.5,1.6,1.7) also contain a nitrogen atom in the sugar ring, but in this case, the nitrogen occupies the position corresponding to the anomeric carbon atom. As such, protonation of the nitrogen atom in these compounds will allow them to mimic the developing positive charge on Cl of the oxocarbenium ion-like transition state. To date, compounds of this class are amongst the most potent inhibitors of retaining P-glycosidases, often binding 10000 to 100000 times more tightly than the parent sugars that they emulate [31, 32]. In an effort to improve upon the inhibitory potency of nojirimycin and isofagomine type compounds, attempts to incorporate other features of the oxocarbenium ion-like transition state into similar compounds have been made. Derivatives of compounds such as glycohydroximolactams (1.8) and gluconamidine (1.9) not only contain a basic endocyclic nitrogen atom but the half-chair conformation and sp hybridized center at Cl of an oxocarbenium ion are also featured in these molecules. It is Chapter 1 Introduction 16 therefore not surprising that these compounds are strong inhibitors of retaining glycosidases [33-37]. Likewise, bicyclic inhibitors such as nagstatin (1.10) and the nojirimycin tetrazoles (1.11), imidazoles (1.12) and triazoles (1.13), which also contain these features, are also reported to be highly effective inhibitors of many different glycosidases [38]. 1.2.3.3 Covalent glycosyl-enzyme intermediate An important point of contention with regards to the double displacement mechanism has always been whether the intermediate is a covalent glycosyl-enzyme or a stabilized oxocarbenium ion pair. The root of this controversy is historically based and involves the study of hen egg white lysozyme, which was the first enzyme to have its tertiary structure determined through X-ray crystallography [39]. Subsequent model building studies, along with a second structure of the enzyme in a nonproductive complex with the oligosaccharide chitotriose, led to the proposal of a mechanism involving an oxocarbenium ion intermediate [40, 41]. Despite the large amount of evidence -including the extremely short lifetimes of such species [42] - that has accumulated against such an intermediate since that time, a stabilized oxocarbenium ion intermediate has remained the paradigm for the mechanism of retaining glycosidases. As a means of addressing the nature of the intermediate, the ability of secondary oc-deuterium kinetic isotope effects to provide insights about changes in the geometry of the reaction center in the rate-limiting step of the reaction as it proceeds from the ground state to the transition state can be an invaluable tool. A change in the hybridization geometry from sp3 to sp2 is usually characterized by an isotope effect whose value for knlko falls between 1.1 and 1.3. Conversely, a change from sp2 to sp3 geometry will Chapter 1 Introduction 77 result in an inverse isotope effect and the value of k^/ko will be less than unity. These isotope effects are therefore ideal probes for examining the nature of the intermediate in the double displacement mechanism. Secondary a-deuterium kinetic isotope effects using substrates for which the deglycosylation step was rate limiting have been reported for a number of retaining glycosidases [20-23, 43]. In all the cases studied, the values for kH/kD were found to be greater than unity. This result is indicative of a change in the geometry of the anomeric center from sp3 to sp2 as the reaction proceeds from the intermediate to an oxocarbenium ion-like transition state during the second step of the reaction. The sp3 geometry of the intermediate therefore lends strong support to the argument that the intermediate (at least for these particular glycosidases) is indeed covalent and not an ion pair. More concrete evidence for the existence of a covalent glycosyl-enzyme intermediate has come from the actual trapping of this species through a variety of different methods. Early successes were achieved through the use of low temperature and rapid quench techniques to quickly denature the enzyme after its incubation with substrate. Through these methods, the covalent intermediates of several enzymes including porcine oc-amylase [44], A. wentii p-glucosidase [45] and sucrose phosphorylase [46, 47] were trapped and identified. However, the need to denature the protein in order to trap the intermediate raised concerns about the true nature of the intermediate in its native state. In place of such harsh conditions, subsequent strategies for the trapping of the covalent glycosyl-enzyme intermediate were developed using methods that would selectively slow down the deglycosylation step of the reaction. One of the more effective means of accomplishing this task is with the use of mechanism-Chapter 1 Introduction 18 based inactivators such as fluorosugars (vide supra). The success of this approach in trapping the covalent fluoroglycosyl-enzyme intermediate of retaining glycosidases is unparalleled [18]. X-ray crystal structures of several such trapped complexes have also been solved to reveal unequivocally the covalent nature of the intermediate [48-51]. The one argument against this strategy however is that it is biased towards the formation of a covalent adduct since the destabilizing effects of the fluorine substituent would cause an oxocarbenium ion intermediate to collapse. As an alternate means of trapping the intermediate through slowing the deglycosylation step, some success has been achieved through mutation of key active site residues. Using this approach, the covalent glycosyl-enzyme intermediate of a xylanase from C. fimi was the first to be trapped without the need for any modifications on the sugar moiety [52]. The three dimensional structure of this glycosylated protein was also solved to reveal once again that the intermediate is indeed covalent. The combination of mechanism-based inactivators, site-directed mutagenesis and X-ray crystallography has played a critical role in providing direct evidence for a covalent glycosyl-enzyme intermediate for a large number of retaining glycosidases. However, the entrenched idea of an oxocarbenium ion intermediate from the historical work on hen egg white lysozyme remains. The lingering of this sentiment over all these years was due in large part to an inability to trap the intermediate on lysozyme. This was all changed recently when Vocadlo et. al. were able to successfully trap not only a covalent 2-fluorochitobiosyl-enzyme intermediate on the wild-type enzyme but the nonfluorinated chitobiosyl-enzyme intermediate was also trapped on the acid/base mutant of the enzyme [53]. A three-dimensional structure of the 2-fluorochitobiosyl-enzyme Chapter 1 Introduction 19 intermediate of the acid/base mutant was also obtained, providing definitive evidence that the mechanism of hen egg white lysozyme proceeds through a covalent glycosyl-enzyme intermediate and not a stabilized oxocarbenium ion pair. While much of the experimental evidence has therefore pointed to a covalent glycosyl-enzyme as the predominant intermediate in the mechanism of retaining glycosidases, the existence of short-lived, stabilized oxocarbenium ion intermediates along the reaction pathway will nonetheless remain a real possibility. 1.2.4 Mechanistic studies of glycosyltransferases The similarity of the reactions catalyzed by glycosyltransferases to that of glycosidases has led to the tacit assumption that the two classes of enzymes employ similar mechanisms in carrying out their respective tasks. The implication of this hypothesis is that inverting glycosyltransferases will follow a direct displacement mechanism similar to that of the inverting glycosidases while retaining glycosyltransferases will follow the double displacement mechanism involving the formation of a covalent glycosyl-enzyme intermediate that is utilized by retaining glycosidases. However, in contrast to the vast amount of mechanistic data that is available for the glycosidases, our understanding of the mechanisms of glycosyltransferases pales in comparison. As mentioned previously, this naivete is due in large part to the limited availability of these enzymes for study. Fortunately, this problem has been largely overcome through high level recombinant expressions of these proteins. The subsequent organization of their sequences into a database along with the generation of a sequence-based classification system [3] will undoubtedly result in an explosion of new insights into the catalytic machinery of these biologically important enzymes. Such Chapter 1 Introduction 20_ insights will either support or refute our preconceived notions (eg. a covalent glycosyl-enzyme intermediate from studies of retaining glycosidases) about the mechanisms of glycosyl transfer for both inverting and retaining glycosyltransferases. 1.2.4.1 Mutagenic investigations of glycosyltransferases Although the gene sequences of many glycosyltransferases are now readily available, the ease of access to this information has not translated into a great deal of insightful information about the mechanism of these enzymes. Instead, the primary focus of many mutagenesis studies has tended to revolve around the investigation of substrate specificities and substrate binding interactions. With regards to the donor substrate, the work of Seto et. al. on changing the substrate specificities of the blood group A and B enzymes is particularly noteworthy [54, 55]. The blood group A enzyme, a-l,3-7V-acetylgalactosaminyltransferases (GalNAcT) catalyzes the transfer of GalNAc from UDPGalNAc to the acceptor Fuc-a-l,2-Gal-P-OR to yield the A determinant GalNAc-a-l,3-Fuc-a-l,2-Gal-p-OR structure. The generation of the B deteminant on the other hand is the responsibility of an a-l,3-galactosyltransferase (GalT), which catalyzes the transfer of Gal from UDPGal to the same acceptor to yield a Gal-a-l,3-Fuc-a-l,2-Gal-(3-OR structure. The specificity of each enzyme for its corresponding substrate was determined to be the result of a single amino acid difference at position 266 where this residue is a leucine in GalNAcT and a methionine in GalT [55]. By manipulating several other key residues, mutant enzymes capable of catalyzing both reactions were also generated [54]. Similar work pertaining to acceptor substrate specificity has also been carried out on three highly homologous a-l,3-fucosyltransferases (FucT). These enzymes are responsible for catalyzing the fiicosylation of cell surface oligosaccharides, including Chapter 1 Introduction 27 several that mediate the adhesion of leukocytes to endothelial cells during inflammation. Through domain-swapping and truncation studies, the differences in acceptor specificity (Gal-p-l,3-GlcNAc or Gal-P-l,4-GlcNAc) amongst these enzymes was localized to a hypervariable region in the N-terminus of the proteins [56, 57]. The changing ofa single amino acid from Tip to Arg was later identified as the determinant in the acceptor preference of the different fucosyltransferases [58]. The incredible diversity of glycosyltransferases in terms of their amino acid sequences has made it difficult to pinpoint specific residues that may be important to catalysis. The only motif that has been identified thus far as being highly conserved amongst glycosyltransferases is that of a short DXD sequence [59, 60]. This motif is represented in many of the existing glycosyltransferase families and is present in both inverting and retaining enzymes. Mutation of the Asp residues in this motif has been shown to result in complete loss of activity in the enzyme, thereby validating the importance of these residues to catalysis [59, 61-63]. Evidence from the available three-dimensional structures of glycosyltransferases (vide infra) along with metal activation [62] and labeling studies [61] has strongly implicated the importance of this DXD motif in the binding of both the metal ion and also the phosphoryl group of the donor substrate. 1.2.4.2 X-ray crystallographic investigations of glycosyltransferases The recent ability to recombinantly express large quantities of protein has made it much more feasible to obtain X-ray crystal structures of glycosyltransferases. This is evidenced by the large number of three-dimensional structures that have become available over the last three years [64-73]. Prior to this time, the only available structure of a glycosyltransferase was that of a (3-glucosyltransferase from bacteriophage T4, Chapter I Introduction 2 2 which was apparently unrelated in amino acid sequence to any other enzyme [74]. Although glycosyltransferases of different families appear unrelated from the standpoint of their amino acid sequences, the three-dimensional structures of these enzymes reveal a surprisingly different picture. An examination of the available structures shows an adoption of only two main folds by these enzymes [15]. The first is known as the GT-A fold and is represented by proteins belonging to six different families of glycosyltransferases. This fold comprises two dissimilar domains in which one domain is involved in the binding of the nucleotide while the other binds the acceptor substrate. The second fold is designated the GT-B fold and is adopted by proteins representing four different families of glycosyltransferases. This particular fold is characterized by two similar nucleotide binding domains known as Rossmann folds. As for families for which a crystal structure is not yet available, threading analyses has revealed that most will also adopt either the GT-A or GT-B folds [15]. However, the resistance of some families to threading suggests that other folds may also exist. Of the currently available structures of glycosyltransferases, most are of the inverting enzymes. Recalling that these enzymes are believed to operate via a direct displacement mechanism (Figure 1.2a), evidence from the structures of these enzymes does appear to support this postulate. Extrapolating from the work on inverting glycosidases, an essential feature of this mechanism is the catalytic base that is needed to deprotonate the nucleophilic hydroxyl group of the acceptor molecule. For all glycosidases, this residue is always found to be a carboxylic acid (either Asp or Glu). From the structures of all the inverting glycosyltransferases that have been solved to date, a carboxylic acid is indeed observed to be correctly positioned within the active site to Chapter I Introduction 25 fulfill such a role [64-68, 71]. In the case of a glucosyltransferase involved in the biosynthesis of vancomycin group antibiotics, mutation of this putative base catalyst to an alanine residue was shown to result in a 250-fold decrease in activity [71]. This seeming employment of the same mechanism by inverting glycosidases and glycosyltransferases naturally led to a greater confidence in the assumption that the same is also true of the respective retaining enzymes. Unfortunately, support for the double displacement mechanism in the reactions of retaining glycosyltransferases has remained elusive, despite the recent solution of the three-dimensional structures of two such enzymes [69, 70, 72]. 1.2.4.3 Mechanistic investigations of retaining glycosyltransferases Mechanistic studies of retaining glycosyltransferases to date have revolved mainly around glycogen synthetase. This enzyme catalyzes the transfer of a glucose residue from UDP-glucose to the C4 hydroxyl group of a growing glycogen primer, forming an a-1,4 linkage in the process. Measurements of secondary a-deuterium kinetic isotope effects for this enzyme have shown the kn/ko values to be greater than unity on both V m a x and Vm^/Km [75]. This observation is once again suggestive of a change in hybridization from sp3 geometry in the ground state to sp2 geometry in the transition state. However, because it is not known whether the glycosylation step or the deglycosylation step is rate limiting in this reaction, it is difficult to know whether this change in hybridization reflects that of the substrate as it proceeds towards the first transition state or whether it is of the intermediate progressing towards its transition state. Therefore, the only conclusion that can be drawn from this result is that the reaction proceeds through a highly dissociative transition state with substantial oxocarbenium ion Chapter 1 Introduction 24 character. Support for the oxocarbenium ion-like nature of this transition state has come from the inhibition of this enzyme by the transition state analogues, D-gluconolactone and deoxynojirimycin [76]. As a means of detecting the potential formation of transient intermediates in reactions catalyzed by retaining glycosyltransferases, positional isotope exchange (PIX) experiments have been carried out for both sucrose synthetase and glycogen synthetase [75, 77]. To carry out these experiments, 180-labeled UDPGlc was first synthesized with the heavy isotope incorporated into the nonbridging positions of the p-phosphate (Figure 1.5). Should a transient intermediate be formed from cleavage of the Cl"-OT" bond upon incubation of this substrate with enzyme, then scrambling of the lsO-label from the nonbridging to the bridging position should occur, provided that the lifetime of the intermediate is greater than the time required for rotation of the P-phosphate. Unfortunately, no scrambling of the heavy isotope was observed for either glycogen synthetase or sucrose synthetase. The absence of any PIX in both cases was attributed to the inability of the phosphate to rotate freely within the active site of each enzyme. Figure 1.5. The use of positional isotope exchange as a means of detecting the formation of a transient intermediate in the reactions of sucrose synthetase and glycogen synthetase. The darkened circles are representative of l 8 0 . Chapter 1 Introduction 25_ 1.3 Lipopolysaccharyl oc-l,4-Galactosyltransferase C Lipopolysaccharyl a-l,4-galactosyltransferase C (EC 2.4.1.x, LgtC) from Neisseria meningitidis is the third product of a gene cluster that encodes for several glycosyltransferase enzymes in this organism. Together, these enzymes are responsible for building up the saccharide portion of lipooligosaccharides, which constitute the major glycolipids found on the cell surface of N. meningitidis and several other Gram-negative mucosal pathogens including Haemophilus, Moraxella, Bordetella and Campylobacter. Lipooligosaccharides are composed primarily of a lipid A-linked 2-keto-3-deoxyoctulosonic acid (KDO) core that is decorated with various oligosaccharides. The function of these structures is to mimic the glycolipids on the surfaces of human cells, thereby allowing the pathogens to evade detection by the immune system during the course of an infection. The reaction catalyzed by LgtC is the transfer of a galactosyl moiety from uridine diphosphogalactose (UDPGal) to the terminal lactose of a lipooligosaccharide structure (Figure 1.6). This reaction proceeds with overall retention of stereochemistry at the anomeric center and the result is the formation of a Gal-a-1,4-Lac moiety at the terminus of the lipooligosaccharide. This trisaccharide structure, which is also known as the Pk antigen is an integral component of the P blood group glycolipids in humans. Chapter I Introduction 26 A = A-aectyl-D-glucosamine ^ = phosphoctiwiolaminc = Lipid A Figure 1.6. The reaction catalyzed by Lipopolysaccharyl oc-l,4-galactosyltransferase C (LgtC). From a therapeutic standpoint, LgtC is therefore a potentially valuable target for the treatment of bacterial infections. By inhibiting the action of this enzyme, the ability of pathogens to evade the immune system during an infection will be compromised, as they will no longer be able to camouflage themselves from detection. In addition to its value as a potential therapeutic target, LgtC, which catalyzes the formation of a linkage that is difficult to achieve by chemical means, could also be extremely useful in the rapidly growing field of oligosaccharide synthesis. The current limiting factor in such applications however, is the high cost of nucleoside phosphosugars. One method of overcoming this problem has been the development of complex recycling schemes, which allows for the regeneration of the nucleoside phosphosugar [78]. However, this strategy is often very labour intensive and therefore difficult to apply to large scale syntheses. A better solution is to find more cost effective substrates that can act as a substitute for the natural nucleoside diphosphosugars. On the basis of sequence similarities, LgtC is assigned to family 8 of glycosyltransferases [3]. Like most glycosyltransferases, LgtC is naturally a membrane bound protein. The difficulties associated with the handling of such proteins have been Chapter I Introduction 27 one of the major hindrances to their study. The solution to this problem has been to remove the transmembrane domains by cloning and expressing only the soluble portions of these proteins. In LgtC, the transmembrane domain responsible for anchoring the protein to the membrane is located at the C-terminus. Removal of this domain through deletion of up to the last twenty-eight amino acids of the C-terminus has been shown to have no effect on the in vitro activity of the resulting enzyme [79]. Therefore, for practical purposes, only the soluble, truncated form of the protein will be studied. 1.4 Aims of this Thesis The primary aim of the work to be presented in this thesis is to provide a detailed mechanistic study of lipopolysaccharyl oc-l,4-galactosyltransferase C from Neisseria meningitidis. Since very little is known about the mechanisms of retaining glycosyltransferases in general, insights from our study of LgtC could provide a starting point towards a greater understanding into the intricate workings of this class of enzymes. Our work on LgtC will be basically divided into two parts. In the first half of the thesis, much of the focus will be on the detailed kinetic characterization of the recombinant protein, which is missing the nonessential transmembrane domain at the C-terminus. The second half of the thesis will then focus on deciphering the chemical mechanism that is employed by this enzyme in carrying out its task. An important component of the kinetic characterization of LgtC is the determination of the specificity of this enzyme for its substrates. To investigate this aspect of the enzyme, a variety of potential substrates will be synthesized and the kinetic parameters for their turnover will be examined. The information gained from these studies will not only provide insights about the enzyme-substrate interactions in LgtC, Chapter I Introduction 28 but it could also be of great value in the application of this enzyme for synthetic purposes. In addition to the substrate specificities of the enzyme, another fundamental aspect of the glycosyl transfer reaction that needs to be addressed is the order in which the substrates are bound and the products released. For the LgtC catalyzed reaction, which involves two substrates and two products and hence is given the designation BI BI [80, 81], a number of kinetic mechanisms are conceivable. Discernment between the possible ordered, random and ping pong BI BI kinetic mechanisms will be achieved through a series of inhibition studies with both substrate analogues as well as with products. An understanding of the kinetic mechanism is a valuable asset for it will greatly aid in the design of experiments to probe the chemical mechanism of LgtC. The chemical mechanism that has been proposed for the LgtC catalyzed glycosyl transfer reaction is a double displacement mechanism similar to that employed by many retaining glycosidases. As a means of providing experimental evidence for this mechanism, much of our efforts will be directed at the development of strategies to trap the putative covalent glycosyl-enzyme intermediate on LgtC and identify the catalytic nucleophile. Borrowing from the work on retaining glycosidases and transglycosylases, the strategies that will be adopted for this purpose will include both the fluorosugar approach and also the incompetent acceptor substrate approach. Since both strategies require the employment of modified substrates, our first goal is therefore to synthesize these compounds. Once obtained, these compounds will then be evaluated as probes of the enzyme not only in terms of their kinetic behaviour but more importantly, in their ability to trap the much sought after covalent glycosyl-enzyme intermediate. Through collaborative efforts, an X-ray crystal structure of this enzyme will also be sought. The Chapter I Introduction 29 availability of such a structure will be a definite asset in our efforts to understand the workings of this enzyme and thereby of the whole class of retaining glycosyltransferases. CHAPTER 2 KINETIC CHARACTERIZATION OF LIPOPOLYSACCHARYL oc-1,4-GALACTOSYLTRANSFERASE Chapter 2 Kinetic Characterization of LgtC 31 2 KINETIC CHARACTERIZATION OF LGTC 2.1 Acceptor Substrate Specificity Lipopolysaccharyl a-galactosyltransferase C (LgtC) naturally catalyzes the transfer of a galactosyl moiety from UDPGal to a complex lipooligosaccharide (LOS), which is composed of a lipid A moiety linked to an outer-core oligosaccharide through a 2-keto-3-deoxyoctulosonic acid residue (KDO) (Figure 1.6). This outer core oligosaccharide to which LgtC transfers the galactose moiety terminates in a non-reducing end lactose unit. For quantitative assays, the use of LOS is problematic because of its limited solubility and structural heterogeneity. The solution to this problem has been to simply use lactose as the acceptor substrate [82]. However, as implied by the high value of K m (60 mM), the affinity of the enzyme for this acceptor is quite low. In an effort to find an alternative substrate with improved affinity for the enzyme, various lactoside derivatives were prepared in the hope that additional binding interactions with the enzyme would be afforded by the aglycone. These could include 7t-stacking interactions in the case where the aglycone is an aromatic moiety, hydrophobic interactions for aliphatic aglycones and also hydrogen bonding interactions for aglycones that bear hydroxyl or amino functionalities. 2.1.1 Synthesis of various lactosides to function as alternate acceptor substrates Syntheses of the various lactosides were achieved through the Koenigs-Knorr glycosylation procedure involving the condensation of a glycosyl bromide with an Chapter 2 Kinetic Characterization of LgtC 52 alcohol in the presence of silver (II) carbonate [83]. The starting glycosyl bromide (2.2) was easily prepared by treatment of the per-O-acetylated sugar (2.1) with a solution of hydrobromic acid in acetic acid. After glycosylation, the acetate protecting groups were then removed under Zemplen conditions [84] to yield the corresponding lactosides. In this way, the benzyl (2.7), pentenyl (2.8) and allyl lactosides (2.9) were prepared as outlined in Scheme 2.1. Preparation of the 2,3-dihydroxypropyl derivative (2.10) was afforded by oxidation of the double bond in the allyl derivative (2.5) with osmium tetroxide. Scheme 2 . 1 . An outline of the synthesis of various lactosides to serve as potential acceptor substrates for LgtC. OH Q NaOAc Ac ,0 OH °H 0Ac—OAo -°. OAc O. (2.1 ) HBr/AcOH .OAc C H 2 C 1 2 (2.2 ) ROH, Ag 2 C 0 3 CH2CI2 4 A sieves OAc OAc t BuOH, acetone| 2. NaHSO, OAc O NaOMe MeOH R = CH 2 Ph ( 2.3 ) R = (CH 2),CH=CH 2 ( 2.4 ) R = CH 2CH=CH 2 ( 2.5 ) C H 2 C H O H C H 2 O H ( 2.6 ) R = C H 2 P h ( 2.7 ) R = (CH 2) 3CH=CH j ( 2.8 ) R = C H 2 C H = C H , ( 2.9 ) R = C H 2 C H O H C H 2 O H ( 2.10 ) 2.1.2 Kinetic evaluation of various alternate acceptor substrates The kinetic parameters of LgtC for various acceptor substrates are shown in Table 2.1. As can be seen from the high K m values, the enzyme does not bind these compounds very tightly. This low affinity for the acceptor suggests that there are few critical Chapter 2 Kinetic Characterization of LgtC 3_3 interactions with this substrate, a result that could also explain the ability of LgtC to transfer galactose to such a broad range of lactose derivatives. Amongst the different acceptors tested, the monosaccharide galactose was found to be a much poorer substrate than the disaccharide lactose or any of its glycoside derivatives. The second order rate constant for transfer to galactose was 500 - 1200 fold lower than that of the disaccharide substrates, indicating the importance of a second sugar binding site. However, there was very little variation in the kinetic parameters between lactose and its various derivatives. These results imply that the acceptor binding site is composed of at least two important subsites designed to recognize a lactose moiety at the terminus of the LOS. While the importance of other potential subsites was not explored, results from the various lactoside derivatives do show that binding was not significantly enhanced by longer non-sugar substrates. Table 2.1. Kinetic parameters for the utilization of various acceptor substrates by LgtC at a fixed, saturating concentrations of the donor substrate, UDPGal. Substrate K m a p p (mM) kc*a p p(s)-' ( k c a t / K ^ t m M - y 1 ) Galactosea 0.0011 Lactose 60 34 .0.57 Pentenyl lactoside b 1.0 AUyl lactoside 33 35 1.1 Benzyl lactoside 22 25 1.2 2,3-Dihydroxypropyl lactoside 26 35 1.4 a no saturation was obtained up to 250 mM b no saturation was obtained up to 20 mM and problems with solubility prevented the use of higher concentrations * errors in the data range from 5% - 20% Chapter 2 Kinetic Characterization of LgtC 34 2.1.3 Hydrolytic activity of LgtC in the absence of saccharide acceptors In the absence of any glycosyl acceptors, turnover of UDPGal was still observed, albeit at a substantially reduced rate. The value of k c a t /K m for UDPGal under this condition was found to be 0.0039 plVT's"1, some 120 times lower than when lactose was present. Presumably, when suitable glycosyl acceptors are absent, water is capable of acting as an acceptor resulting in hydrolysis of the nucleoside diphosphosugar to yield galactose and UDP. Evidence for the formation of galactose was obtained through TLC analysis of the reaction mixture in which UDPGal alone was incubated in the absence or presence of enzyme overnight. After this time, a spot with a similar Rf value to that of galactose was indeed observed, but only for the mixture in which enzyme was added. Mass spectral analysis of products resulting from the hydrolysis of 180-labeled UDPGal 18 • by the enzyme (with O incorporated at the bridging position between the galactose and UDP) revealed that the 1 K O remained with the UDP. This observation proves that the attack by water on the UDPGal substrate occurs at the expected anomeric carbon of the galactosyl portion of the substrate rather than at the (3-phosphorus. In this regard, the hydrolysis of UDPGal by LgtC is unlike that of the Nudix family of enzymes, which have all been shown to hydrolyze nucleoside diphosphate derivatives via nucleophilic substitutions at phosphorus [85, 86]. The ability of glycosyltransferases to behave as hydrolases is not without precedent since several glucosyltransferases including toxins A and B from Clostridium difficile and T4 phage P-glucosyltransferase have all been shown to be capable of hydrolyzing UDP-Glc in the absence of saccharide acceptor [87-89]. Chapter 2 Kinetic Characterization of LgtC 35_ 2.2 Donor Substrate Specificity In the LgtC-catalyzed glycosyl transfer reaction, UDPGal is the donor substrate and the source of the galactose moiety that is transferred onto a lactose unit at the terminus of a growing lipooligosaccharide. The enzyme appears to be highly specific for its donor substrate as can be seen in comparing the second order rate constants for the turnover of UDPGal and UDPGlc. These two compounds are epimers differing only in the position of the hydroxyl group at the 4 "-position. However, in terms of transferring their respective glycosyl moieties onto the acceptor lactose, the k c at /K m value for UDPGlc (0.0026 pM-'s"1) was found to be some 200 times lower than that of the natural substrate UDPGal (0.47 plvf's"1). This difference in specificity is primarily the result of the much higher kc a t value that is observed for UDPGal (6.2 s"1) than for UDPGlc (0.0029 s"1). As for the affinity of the enzyme for these two compounds, LgtC appears to bind UDPGlc (Km = 1 pM) an order of magnitude more tightly than UDPGal (Km =13 pM). 2.3 oc-Galactosyl fluoride as an alternate donor substrate Complex oligosaccharides are currently in great demand as the field of glycobiology continues to grow at an astounding rate. However, the chemical synthesis of these polyhydroxylated structures is a nontrivial task as it often involves exhaustive protection and deprotection schemes. Adding to the challenge is the need to control stereochemistry in molecules laden with stereogenic centers. In an effort to avoid these difficulties that must be faced by chemical syntheses, enzymatic methods have become an attractive alternative. For this purpose, Nature's own glycosyltransferases are ideal candidates since they naturally catalyze the formation of glycosidic bonds. One Chapter 2 Kinetic Characterization of LgtC 36 drawback to the use of these enzymes however, is the high cost of the nucleoside diphosphosugar substrates that are utilized by these biocatalysts. As a means of circumventing this problem, complex recycling schemes have been developed to regenerate these nucleoside diphosphosugars [78]. A better solution however, is to find readily available alternative substrates that can be utilized by these enzymes in place of the natural nucleoside diphosphosugars. One such possibility is the use of glycosyl fluorides. These compounds have been found to be good substrates for glycosidases and as such, may also be useful as substrates for glycosyltransferases. 2.3.1 Synthesis of a-galactosyl fluoride a-Galactosyl fluoride was synthesized from galactose in three fairly straightforward steps as outlined in Scheme 2.2. The introduction of fluorine at the anomeric position was accomplished by treating per-O-acetylated galactose (2.11) with HFpyridine [90], yielding strictly the oc-anomer (2.12). The acetate protecting groups were then removed using sodium methoxide [84] to yield the desired product (2.13). Scheme 2.2. An outline of the synthesis of oc-galactopyranosyl fluoride (a-GalF, (2.13)). O H . OH O A c , OAc O A c , OAc OH 0 H F F (2.11) (2.12) (2.13) Chapter 2 Kinetic Characterization ofLetC 5 7 2.3.2 Kinetic evaluation of oc-galactosyl fluoride as an alternative donor substrate The application of oc-galactosyl fluoride as an alternate donor substrate for LgtC was initially investigated by Lougheed [82]. Through analysis of the products formed upon incubation of LgtC with a-galactosyl fluoride (a-GalF) and the acceptor lactose in the presence of a catalytic amount of UDP, the expected Gal-a-l,4-Lac trisaccharide was indeed shown to be formed. This reaction was however found to be extremely slow and no saturation of the enzyme with increasing concentrations of a-GalF could be observed. Instead, a slight upward curvature was noticed as the reaction rate was plotted as a function of a-GalF concentration. The explanation that was put forth to rationalize this behaviour of a-GalF was that in addition to functioning as the donor substrate, this compound was also functioning as the acceptor substrate [82]. Such an explanation is not without merit since galactose has been shown to be a viable acceptor substrate for LgtC (Table 2.1). However, further validation of this dual role of a-GalF was needed, as there were insufficient data points at the higher concentrations to clearly conclude that there was an upward curvature in the plot of rate versus a-GalF concentration that led to this hypothesis. To verify the possibility of a-GalF functioning as both the donor and acceptor substrates, it was necessary to first substantiate the kinetic behaviour of this compound at high concentrations. As can be seen in Figure 2.1, an upward curvature in the plot of rate as a function of a-GalF concentration was indeed observed as the concentration of this substrate was increased beyond 400 mM while the concentration of lactose was held at a Chapter 2 Kinetic Characterization of LgtC 55 fixed concentration. At concentrations lower than this value, a linear dependence was observed from which the value of the second order rate constant was calculated to be 2.3 x l ( V 4 m M V . [GalF] (mM) Figure 2.1. Plot of rate as a function of oc-GalF concentration for the LgtC-catalyzed transfer of galactose from a-GalF in the presence of a fixed (65 mM) concentration of lactose. Support for the possibility that a-GalF could function as both donor and acceptor substrate was provided by the observation that turnover of this compound still occurred even in the absence of lactose or any other saccharide acceptors. In fact, the value of kcat/Km for this process was also found to be very comparable to that measured when lactose was present as the acceptor. However, since LgtC has been shown to be capable of hydrolyzing UDPGal in the absence of suitable glycosyl acceptors (see Section 2.1.3), it is possible that the observed rate for the turnover of a-GalF in the absence of lactose could also be simply the result of hydrolysis of this compound. While this scenario is certainly possible, evidence for the formation of at least some of the disaccharide, Gal-a-1,4-GalF upon incubation of enzyme with a-GalF and UDP has been provided by TLC Chapter 2 Kinetic Characterization of LgtC 39 and HPLC analysis of the reaction products [82]. In an attempt to further confirm the dual role of a-GalF as donor and acceptor substrate, 4-deoxy-a-D-galactosyl fluoride (4d-a-GalF) was also tested for turnover by LgtC in the absence of lactose. The lack of a nucleophilic hydroxyl group at the 4-position of this compound should eliminate its capacity to function as an acceptor substrate. When 4d-a-GalF was incubated with LgtC in the presence of UDP but in the absence of lactose or any other suitable glycosyl acceptors, no turnover of this compound could be detected above its spontaneous rate of hydrolysis. Unfortunately, the same was true even when lactose was present. Given that a-GalF is at best a poor substitute for the natural donor substrate, UDPGal, it is therefore not unreasonable to expect that 4d-a-GalF would be even less able to function in such a capacity. As such, not only would 4d-oc-GalF be incapable of functioning as an acceptor substrate in the LgtC catalyzed reaction but it appears that this compound is also incapable of functioning as the donor substrate for this enzyme. In summary, as an alternative to the natural donor substrate, UDPGal, a-GalF has also been shown to be capable of functioning as a galactosyl donor in the LgtC-catalyzed reaction. Unfortunately, the reaction rates are extremely low and the enzyme has very little affinity for this compound. In addition, the kinetic analysis of the reaction is further complicated by the ability of a-GalF to serve as both glycosyl donor and acceptor for LgtC. Nonetheless, for large scale syntheses of oligosaccharides, glycosyl fluorides may prove to be a viable alternative to the more expensive nucleoside phosphosugar substrates, especially if the reaction rates can be improved through the creation of mutant enzymes. Chapter 2 Kinetic Characterization of LgtC 40 2.4 Determining the Kinetic Mechanism of LgtC In the LgtC catalyzed reaction, two substrates are converted by the enzyme into two products. In accordance with the notation developed by Cleland for the classification of multisubstrate enzymes [80], LgtC is therefore designated as a BI BI enzyme. With regards to the order of substrate binding and product release in BI BI systems, three basic kinetic mechanisms are possible. These mechanisms have been assigned the names ping pong, ordered and random, and the Cleland notation for describing the order of substrate binding and product release in each is shown in Figure 2.2. A B p Q Figure 2.2. Cleland notation for describing the order of substrate binding and product release in a (a) ping pong (b) ordered and (c) random BI BI kinetic mechanism. In the case of a ping pong kinetic mechanism, the conversion of the first substrate to product followed by the subsequent release of this product from the enzyme is independent of the binding of the second substrate. During the reaction with the first substrate, the enzyme itself is modified and it is this modified form of the enzyme that then proceeds to bind the second substrate and catalyze its conversion to product. This alternating sequence of substrate binding and conversion to product is unique only for a ping pong mechanism as both the ordered and random mechanisms require that both Chapter 2 Kinetic Characterization of LgtC 41 substrates be bound to the enzyme before any catalysis can occur. For an ordered mechanism, the sequence of substrate binding and product release is also clearly defined whereas in the case of a random mechanism, this order is inconsequential. Amongst the three kinetic mechanisms, the ping pong mechanism is often the easiest to identify. The double reciprocal form of the initial rate equations for a ping pong, ordered and random mechanism are represented by the linear Equations 2.1 - 2.3 respectively. ]_ v V I V ( Y A \ 1 ( m — + v \ max J [A] { K„ • + -1 <KAKJ K. + V [B] V max I -1 r \ max max J [A] • + max J K.. • + -yVmJB] Vx max J cxK K aK V 1 Bl V V max t- J ma max / [A] 1 V^maxC^] K (Equation 2.1) (Equation 2.2) (Equation 2.3) max / An examination of the rate equation for a ping pong mechanism (Equation 2.1) shows that when the reciprocal rate is plotted as a function of the reciprocal concentration of one of the substrates (A), the term defining the slope in this equation is independent of the concentration of the other substrate (B). Therefore, when substrate B is employed at a series of different fixed concentrations, the result will be a family of parallel lines (Figure 2.3a). In contrast to that of a ping pong mechanism, the slope term in the equations describing both the ordered and random cases is dependent on the concentration of the second substrate (Equations 2.2 and 2.3). The double reciprocal plot at different fixed concentrations of B will therefore result in a family of intersecting lines for both the ordered and random cases (Figure 2.3b). On the basis of initial velocity kinetics and the Chapter 2 Kinetic Characterization of LgtC 42 resulting appearance of the double reciprocal plot, a ping pong kinetic mechanism is therefore easily distinguished from that of an ordered or random mechanism. Figure 2.3. Expected patterns in the double reciprocal plot for a (a) ping pong and (b) either an ordered or random BI BI kinetic mechanism. In the case of LgtC, initial velocity studies of this enzyme were found to yield a family of parallel lines in the double reciprocal plot [82]. This result naturally led to the conclusion that LgtC follows a ping pong kinetic mechanism. However, an inherent difficulty with this conclusion is in deciding whether the lines are actually parallel or only appear parallel because they intersect at a point far from the origin. For instance, when the dissociation constant, K A is much smaller than the value of K m A in an ordered BI BI system (Equation 2.2), the resulting family of lines in the double reciprocal plot will intersect far to the left of the 1/v axis and far below the 1/[A] axis [91]. As such, the "best fitting" lines through the data will appear to be parallel. This situation is also true of random BI BI systems when the binding of one substrate strongly inhibits the binding of the other substrate (ie. a » 1 in Equation 2.3) [91]. Given that UDPGal is bound so tightly by LgtC and the acceptor lactose is not, neither ordered nor random BI BI 1/[A] 1/[A] Chapter 2 Kinetic Characterization of LgtC 43 mechanisms can be ruled out based solely on the previously observed parallel lines in the double reciprocal plot [82]. 2.4.1 UDPGal/UDP exchange as a diagnostic test of a ping pong kinetic mechanism A ping pong kinetic mechanism is essentially two independent half reactions of substrate binding followed by conversion to product. Therefore, in the absence of the second substrate, the enzyme will still be capable of binding to and reacting with the first substrate. In the process of the reaction, the enzyme is modified by a portion of this substrate while the other portion is released as the first product of the reaction. However, in the absence of the second substrate, the second half reaction will be unable to occur. The only available option then is for the first product that was released to re-bind to and react with the modified form of the enzyme to regenerate the starting substrate and free enzyme. A common procedure for detecting this process is to start with a substrate that is isotopically labeled in the part of the molecule that is released as the product when the substrate is acted upon by the enzyme. This substrate is then incubated with the enzyme in the presence of excess unlabeled product. After reaction with the enzyme, the isotopically labeled product that is released will be diluted by the pool of excess unlabeled product and the result will be the loss of the label in the re-isolated substrate. The occurrence of any such exchange is therefore diagnostic of a ping pong kinetic mechanism. To investigate the validity of the results obtained from earlier initial velocity studies, which pointed to a ping pong kinetic - mechanism for LgtC, a UDPGal/UDP exchange experiment was carried out using 18Q-labeled UDPGal (with the 1 8 0 Chapter 2 Kinetic Characterization of LgtC 44 incorporated at the bridging position between galactose and UDP). Details regarding the synthesis of this compound are discussed in Section 3.4.1. The exchange experiment was carried out by incubating , 80-UDPGal with LgtC in the presence of excess unlabeled UDP but in the absence of any suitable saccharide acceptors. After 4 hours of incubation, the UDPGal was re-isolated and analyzed by both mass spectrometry and 3 1P NMR. The 18 expected result for a ping pong kinetic mechanism is a washing out of the O heavy atom in the re-isolated UDPGal due to UDP exchange with the excess unlabeled species following cleavage of the Cl"- 1 8 OT" bond. From the results of the mass spec and 3 1P NMR, no UDP exchange was evident as only 180-labeled UDPGal was observed. Data obtained from mass spec and 3 1P NMR revealed no change in the 1 8 0 content of the donor substrate. Shown in Figure 2.4a is the signal (appearing as a doublet) corresponding to the P-phosphorus in the proton decoupled 3 1P NMR spectrum of UDPGal. The smaller, downfield peaks are of the minor amount of non-isotopically labeled compound while the larger peaks are from those of lsO-labeled UDPGal. As can be seen in the spectrum of the re-isolated UDPGal after incubation with LgtC (Figure 2.4b), no loss of the 1 8 0 isotope from the donor substrate was observed. Confirmation that the major peaks in Figure 2.4b are indeed from 180-UDPGal was achieved by spiking the sample with unlabeled UDPGal and once again recording the proton-decoupled 3 1P NMR spectrum (Figure 2.4c). The absence of any washing out of the isotope therefore strongly disfavours a ping pong kinetic mechanism, suggesting that the family of lines observed in the initial velocity studies were in fact not parallel but rather must have intersected at a point far from either of the axes. By process of elimination, the kinetic mechanism of LgtC must therefore be either ordered Bl Bl or random Bl Bl. Chapter 2 Kinetic Characterization of LgtC 45_ -11.4 -11.5 -11.6 -11-4 -11.5 ^ Yyy7 (C) -11.4 -11.5 -11.6 Figure 2.4. Proton decoupled 3 1 P NMR spectrum showing the doublet arising from the P-phosphorus of 1 8 0-UDPGal (at ~ 85% isotope incorporation) (a) before incubation with LgtC, (b) after incubation with LgtC and (c) with unlabeled UDPGal added to the sample that had been incubated with LgtC. Chapter 2 Kinetic Characterization of LgtC 46 2.4.2 Inhibition studies with a substrate analogue Inhibition studies with substrate analogues can be an invaluable tool in deciphering the kinetic mechanism of enzymes. In the case of LgtC, such a study would provide a means of corroborating the results of the UDPGal/UDP exchange study, which suggested the kinetic mechanism of LgtC to not be ping pong. A common hindrance to such studies however, is the availability of suitable substrate analogues. Fortunately, this is not a problem for LgtC as UDPGlc quite adequately addresses this need. By virtue of its 200-fold lower rate of turnover by LgtC in comparison to the natural substrate, this 4 "-epimer of UDPGal will behave essentially as an inhibitor that will compete with UDPGal for the donor binding site of the enzyme. When the inhibition of LgtC by UDPGlc was investigated at a saturating concentration of UDPGal but varying concentrations of lactose, the resulting double reciprocal plot yielded a noncompetitive pattern of inhibition (Figure 2.5). 1/[lactose] (mM)"1 Figure 2.5. Double reciprocal plot for the inhibition of LgtC by UDPGlc at varying concentrations of lactose and a fixed concentration (250 pM) of UDPGal. The concentrations of UDPGlc were 0 (•), 50 (O ), 100 (•) and 200 pM (•). Chapter 2 Kinetic Characterization of LgtC 47_ According to Table 2.2, such an inhibition pattern is inconsistent with the uncompetitive pattern that would have been expected for a ping pong mechanism, hence corroborating the results of the previously discussed UDPGal/UDP exchange study. Unfortunately, no discernment between an ordered mechanism and a random mechanism can be made from this outcome as both mechanisms can potentially yield noncompetitive inhibition patterns in such a study (Table 2.2). However, in the event that the kinetic mechanism of LgtC is ordered, the results of this particular inhibition study do dictate that UDPGal will bind first, followed by the acceptor substrate. Table 2.2. Expected patterns of inhibition for analogues of substrates A and B in ping pong, ordered and random B l B l kinetic mechanisms. C = competitive, NC = noncompetitive, UC = uncompetitive [92]. Kinetic Mechanism Inhibitor Competitive for Substrate Varying A Varying B Ping Pong Bl B l A C U C B UC C Ordered B l Bl A c NC B UC C Random B l Bl A c NC B NC C Chapter 2 Kinetic Characterization of LgtC 48 2 A3 The use of product inhibition studies to distinguish between ordered and random Bl Bl kinetic mechanisms In both ordered and random Bl Bl kinetic mechanisms, the binding of both substrates is a mandatory requirement prior to any chemical catalysis. As their names would indicate, an ordered kinetic mechanism follows a defined sequence of substrate binding and product release whereas in the case of a random mechanism, no such discrimination occurs. From a kinetic standpoint, these two mechanisms are indistinguishable on the basis of initial velocity kinetics since they both yield similar patterns in their respective double reciprocal plots. Inhibition studies with a substrate analogue also proved to be a fruitless endeavour in that respect, as the results did not uniquely pinpoint either an ordered or a random mechanism. Therefore, in an attempt to distinguish between these two kinetic mechanisms, product inhibition studies were employed. In the LgtC catalyzed transfer of galactose from UDPGal to lactose, the products of the reaction are UDP and the trisaccharide Gal-oc-l,4-Lac. To carry out the product inhibition study, the trisaccharide product Gal-a-l,4-Lac was first synthesized enzymatically from UDPGal and lactose using LgtC. The inhibition pattern exhibited by this product was found to be noncompetitive when UDPGal was the varied substrate while lactose was held at a fixed but nonsaturating concentration (Figure 2.6). Chapter 2 Kinetic Characterization of LgtC 49 T — i — i — I — i — i — i — | — i — i — i — | — r 1 /[UDPGal] (|iM)-1 Figure 2.6. Double reciprocal plot showing the inhibition of LgtC by the product Gal-a-1,4-Lac at varying concentrations of UDPGal and a fixed concentration (120 mM) of lactose. The concentrations of the product trisaccharide were as follows: 0 (•), 15 (O ), 20 (•), 30 (•), and 40 mM (•). This observation strongly disfavours a rapid equilibrium random BI BI mechanism since only competitive inhibition patterns or no inhibition are expected in that case (Table 2.3). The data therefore pointed to either an ordered BI BI or a nonrapid equilibrium random BI BI kinetic mechanism. Kinetically, the differentiation of these two possibilities would require further inhibition studies either with the other product, UDP, or again with Gal-a-1,4-Lac but at saturating concentrations of lactose. Unfortunately, neither study is practical. An inhibition study with UDP is problematic because this product is used in the coupled assay to monitor enzyme activity and as such cannot be added. With regards to an inhibition study with Gal-a-l,4-Lac at saturating levels of lactose, the combination of the limited solubility of lactose and its low affinity for the enzyme as well as its propensity to cause substrate inhibition greatly restricts the employment of this substrate Chapter 2 Kinetic Characterization of LgtC 50 at high concentrations. As a result, the unequivocal assignment of a kinetic mechanism for LgtC becomes a nontrivial task. Table 2.3. Expected product inhibition patterns in an ordered, rapid equilibrium random and nonrapid equilibrium random B l Bl kinetic mechanisms. C = competitive, NC = noncompetitive, NI = no inhibition, UC = uncompetitive [92]. Mechanism Product varying A varying B Inhibitor unsat. B sat. B unsat. A sat. A Ordered Bl Bl P NC UC NC NC Q C C NC NI Rapid Equilibrium P c NI C NI Random Bl Bl Q c NI c NI Nonrapid Equilibrium p NC NC NC NC Random Bl Bl Q NC NC NC NC Fortunately, the X-ray crystal structure of the enzyme in complex with analogues of both the donor and acceptor substrates was able to provide some important insights into this matter (see Section 3.2). From the structural data, it is clear that UDPGal is bound to the enzyme in a site buried within the protein. This binding of the donor substrate appears to trigger a conformational change involving the closing of two loops over the active site, which leads to the formation of a large open pocket into which the acceptor substrate is then bound. This observation therefore disfavours any kind of random kinetic mechanism since the binding site for the acceptor substrate is not formed until the donor substrate has been bound. Taken together with the fact that crystallization of the enzyme was only achieved in the presence of an analogue of the donor UDPGal Chapter 2 Kinetic Characterization of LgtC 5J_ but not in the presence of an acceptor analogue, the data from the crystal structure therefore appear to favour an ordered BI BI kinetic mechanism. In addition to its favouring of such a mechanism, the structural data also strongly suggest that the order of substrate binding must be UDPGal followed by lactose. This observation is fully consistent with the results of the inhibition study with UDPGlc (see Section 2.4.2), which suggested that this must be the case if the kinetic mechanism of LgtC is ordered. In light of this additional information, then the only scenario that could have given rise to the noncompetitive pattern that was observed for the inhibition of LgtC by the product Gal-ea-1,4-Lac at varying concentrations of the first bound substrate, UDPGal (Figure 2.6) is if Gal-oc-l,4-Lac was the first product released. Based on all the available evidence, LgtC is therefore concluded to follow an ordered BI BI kinetic mechanism (Figure 2.2b) in which UDPGal binds first, followed by lactose. Product release then involves first the loss of the tri saccharide, Gal-oc-l,4-Lac followed by UDP. The ordered kinetic mechanism that is followed by LgtC is certainly consistent with the apparent need for this and related enzymes to bind both substrates prior to any chemical catalysis. However, in terms of whether an ordered or random kinetic mechanism is employed, no common trend is yet apparent amongst these enzymes. For example, while the inverting galactosyltransferase from bovine milk was found to utilize a random kinetic mechanism [93], substrate binding and product release in other inverting enzymes such as human fucosyltransferase V [94] and rabbit liver N-acetylglucosaminyltransferase I [95] were found to follow an ordered mechanism. In the case of retaining glycosyltransferases, glycogen synthetase was found to follow a random mechanism [96-98] while substrate binding and product release in LgtC is shown to Chapter 2 Kinetic Characterization of LgtC 52. follow an ordered process. Even the rarely observed Theorell-Chance kinetic mechanism has been reported. This special case of an ordered Bl Bl mechanism, which excludes the formation of the ternary complex of the enzyme with its two substrates was demonstrated to occur with the blood group B glycosyltransferase [99]. 53 CHAPTER 3 AN INVESTIGATION OF THE CHEMICAL MECHANISM OF LIPOPOLYSACCHARYL oc-1,4-GALACTOSYLTRANSFERASE Chapter 3 An Investigation of the Chemical Mechanism of LgtC 54 3 AN INVESTIGATION OF THE CHEMICAL MECHANISM OF L G T C 3.1 Attempts to Trap a Covalent Glycosyl-Enzyme Intermediate In the reaction catalyzed by lipopolysaccharyl a-l,4-galactosyltransferase, the transfer of galactose from UDPGal to the lipooligosaccharide acceptor occurs with net retention of stereochemistry at the anomeric centre (Figure 1.6). By analogy with the well-studied retaining glycosidases, LgtC is therefore believed to follow a double displacement mechanism (Figure 1.2b). A key feature of this mechanism is the formation and subsequent breakdown of a covalent glycosyl-enzyme intermediate. While the existence of such an intermediate has been unequivocally demonstrated for numerous glycosidases, no such species has yet been identified in the handful of retaining glycosyltransferases studied thus far. In an attempt to trap this putative intermediate in LgtC, a number of borrowed strategies from the work on retaining glycosidases and tranglycosylases will be employed. 3.1.1 Incompetent acceptor substrate approach In the double displacement mechanism that LgtC is presumed to employ, the breakdown of the putative covalent glycosyl-enzyme intermediate occurs when this species is attacked by the nucleophilic hydroxyl group of the acceptor substrate during the second step of the reaction. In order to trap the intermediate, the enzyme must therefore be allowed to carry out the first step of the reaction to form the intermediate but Chapter 3 An Investigation of the Chemical Mechanism of LgtC 55 it must then be prohibited from completing the reaction, which involves the transfer of the galactosyl moiety onto the acceptor substrate. In terms of designing experiments for such a purpose, it is therefore important to keep in mind that LgtC follows an ordered Bl Bl kinetic mechanism (Section 2.4) and as such, no chemical catalysis will occur until both substrates are bound by the enzyme. One approach to inhibiting the second step of the double displacement mechanism and also meeting the requirements for substrate binding in an ordered Bl Bl kinetic mechanism so that the first step can occur is through the use of incompetent acceptor substrates. In this strategy, the nucleophilic hydroxyl group of the acceptor substrate is removed thus creating an analogue that would be incapable of causing the breakdown of the putative galactosyl-enzyme intermediate. At the same time, such a compound should still be capable of binding into the acceptor binding site and fulfill the substrate binding requirements of an ordered kinetic mechanism. Incubation of LgtC with UDPGal and such an incompetent acceptor substrate might therefore result in an accumulation of the covalent galactosyl-enzyme intermediate as the first step of the reaction can occur while the second step cannot. 3.1.1.1 Synthes i s o f va r i ou s 4 - deoxy l a c t o s e der i va t i ves For the purpose of using the incompetent acceptor substrate approach to trap the putative covalent glycosyl-enzyme intermediate on LgtC, several 4'-deoxy derivatives of lactose were synthesized. The most elementary of these compounds is 4'-deoxylactose (4dLac, (3.6)) in which the nucleophilic hydroxyl group has simply been removed from the acceptor lactose. This deoxygenation of the 4'-position of lactose was accomplished in a manner similar to that reported for the preparation of 4 '-deoxy-oc-maltosyl fluoride [100] and is outlined in Scheme 3.1. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 56 Scheme 3.1. An outline of the synthesis of 4'-deoxylactose (4dLac, (3.6)). Ph OH OH OH OH O l.PhCH(OMe) 2 NaCNBH TsOH, DM F AGO. OH 'OH 6 Q 0 C 2. Ac 2 0, pyridine / E t , 0 ° A c A ^ o l (3.1) (3.2) 1. T f 2 0, pyridine CH 2C1 2 -20 "Ctor.t. . 2. NaLDMF OH „ . _^OBn ,C 0 X-°AC cyclohexene S^Q ^ % d ( O H ) 2 / C ~ $ ^ £ S ^ ACQ! „ ACO I Bu3SnH, AIBN AcO. C(;H6, reflux OAc EtOH, reflux ——• WDM „ . ^ O A c (3.5) NaOMe, MeOH (3.4) (3.3) (3.6) A foreseeable concern associated with the use of this compound however, lies in its capacity to bind to LgtC since lactose itself binds only poorly to the enzyme. As a means of addressing this potential problem, the analogues benzyl 4'-deoxylactoside (Bn4dLac, (3.17)) and benzyl 4'-deoxy-4'-fluorolactoside (Bn4FLac, (3.13)) were also synthesized. On the basis of earlier results (see Section 2.1.2), the presence of a benzyl aglycone was shown to improve the binding of lactose to LgtC by approximately three-fold. The enhancement in binding that is afforded by the benzyl aglycone may therefore provide some compensation for any binding interactions that might be lost from removal of the 4'-hydroxyl group. In the case of Bn4FLac, the additional presence of an axial fluorine substituent at the 4'-position should further help to mimic the electronic environment normally associated with having a hydroxyl group at that position. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 57 The syntheses of Bn4dLac and Bn4FLac from cellobiose octaacetate are outlined in Scheme 3.2. The installation of the benzyl group was accomplished through the reaction of a per-O-acetylated lactosyl bromide (3.7) with benzyl alcohol according to the well established Koenigs-Knorr glycosylation procedure [83]. The benzyl cellobioside (3.9) that was produced after deprotection was then deoxygenated at the 4'-position in a manner similar to that used in the synthesis of 4dLac to yield Bn4dLac (3.17). The synthesis of Bn4FLac (3.13) on the other hand proved to be more challenging than expected as numerous attempts were made to fluorinate the 4'-position without much success. Treatment of (3.11) with the commonly used fluorinating reagent diethylaminosulfur trifluoride (DAST) resulted in the formation of multiple products as observed by TLC. Unfortunately, none turned out to be the desired compound. In place of the one pot procedure offered by DAST, attempts to fluorinate (3.11) were also made by first activating the 4'-OH with a good leaving group and then attempting to displace this group with a fluoride source. After making the triflate of (3.11), attempts to displace this group with tetrabutylammonium fluoride (TBAF) were unsuccessful, as once again, multiple products were observed but none proved to be the desired compound. The successful displacement of the triflate was finally carried out using tris(dimethylamino)sulfur (trimethylsilyl)difluoride (TASF) as the fluoride, source. As an unexpected but pleasant surprise, the /?-methoxybenzyl protecting group at the 6'-position was also removed under the reaction conditions, yielding (3.12) in the process. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 58 Scheme 3.2. An outline of the synthesis of benzyl 4'-deoxy-4'-fluoro-(3-lactoside (Bn4FLac, (3.13)) and benzyl 4'-deoxy-(3-lactoside (Bn4dLac, (3.17)). OAc O O AcO OAc O HBr / AcGH AcO OAc OAc OAc O ^ O A c 0 - V - ° v AcO-A^-—r-\ AcO I (3.7) Br BnOH, A g 2 C 0 3 CH 2 C1 2 , 4 A sieves OAc O O AcO (3.8) NaOMe, MeOH OAc O OAC p-OMe-Ph - ^ - 0 - ^ \ - - O A c OAc V (3.10) 1. PhCH(OMc) 2 * TsOH, DMF 60 °C 2. A c 2 0 , pyridine OH O OH O HO. (3.9) OH OPMB O O AcO N a C N B H 3 T H F HC1 / Et,0 OAc O 1. T f 2 0 , pyridine (3.11) OBn CH 2 C1 2 , -20 °C O A c 2. NaT, DMF 1. Tf 2 0 , pyridine, CH 2 C1 2 , -20 °C 2. TASF, CH 2 C1 2 , reflux A C 0 - ^ V ^ A C O OAc O OAc (3.12) I NaOMe, MeOH F OH HO - \^—^A - - 0 OH O OBn (3.13) j — OPMB A C O \ ^ 2 ^ 0 0 OAc OAc O (3.14) Bu 3SnH, AIBN C 6 H 6 , reflux — OPMB ^ V - o A c O X — - ^ - \ ^ Q OAc O AcO OBn OAc (3.15) | C A N , MeCN, H 2 0 OH A c o V ^ O ^ O ^ ^ O B n OAr. » I' (3.16) NaOMe, MeOH K • UM OH O (3.17) Chapter 3 An Investigation of the Chemical Mechanism of LgtC 59_ 3.1.1.2 Kinetic evaluation of incompetent acceptor substrates and their employment for the trapping of the putative covalent glycosyl-enzyme intermediate Kinetic analyses of the incompetent acceptor substrates showed all three to inhibit LgtC in a noncompetitive manner with respect to lactose; the kinetic parameters are shown in Table 3.1. This type of inhibition suggests that, in addition to being able to compete with lactose for the acceptor binding site, each of these compounds is also capable of binding to the enzyme-lactose complex. This result is not surprising since many potential subsites for binding to the natural LOS acceptor are left unoccupied when the much smaller lactose is used as the substrate. In the case of Bn4dLac and Bn4FLac, the presence of the benzyl group appears to cause a slight improvement in the affinity of these compounds for the lactose binding site as evidenced by their slightly lower K g i values (Table 3.1). Table 3.1. Inhibition constants for various acceptor analogues with respect to varying concentrations of lactose and a fixed, saturating concentration (300 pM) of UDPGal. Acceptor Analogue K e i a p p * (mM) K e i s a p p " (mM) 4'-Deoxylactose (4dLac) 16 33 Benzyl 4'-deoxylactoside (Bn4dLac) 6 102 Benzyl 4'-deoxy-4'-fluorolactoside (Bn4FLac) 6 27 K e j a p p refers to the apparent dissociation constant of the inhibitor from the lactose-free enzyme. ** K e i S a p p refers to the apparent dissociation constant of the inhibitor from the enzyme - lactose complex. "* error range in data is from 5% - 10% Chapter 3 An Investigation of the Chemical Mechanism of LgtC 60 In an attempt to trap and observe a covalent glycosyl-enzyme intermediate, LgtC was incubated with UDPGal in the presence of each of these incompetent acceptor substrates. At time intervals of 5, 10, 20, 40, 80 and 160 minutes, an aliquot of the reaction mixture was removed and quenched by the addition of an equal volume of 6 M urea. The sample was then quickly frozen at -78 °C to await analysis by electrospray mass spectrometry (ESMS). Unfortunately, no accumulation of such an intermediate could be detected. One explanation for this observation is that the reaction does not proceed via such an intermediate. However, other factors may have also contributed to the inability to trap the intermediate using this approach. For instance, elimination of the nucleophilic hydroxyl group in the acceptor substrate could have disrupted important interactions needed to properly orient key active site residues for catalysis. As such, no reaction would be initiated in its absence and no intermediate would be formed. Another possibility is that a covalent glycosyl-enzyme intermediate was formed but was then hydrolyzed under the conditions of analysis. As mentioned previously (Section 2.1.3), LgtC is capable of hydrolyzing UDPGal in the absence of suitable glycosyl acceptors. Since each of the incompetent acceptor substrates is bound only weakly by the enzyme, their ability to shield the putative intermediate from attack by water would therefore be severely limited. 3.1.2 Fluorosugar approach In the double displacement mechanism that has been proposed for LgtC, both steps of the reaction are believed to proceed via oxocarbenium ion-like transition states with substantial positive charge character at both C I " and the endocyclic oxygen of UDPGal. For this reason, fluorosugars could prove to be invaluable tools for trapping the Chapter 3 An Investigation of the Chemical Mechanism of LgtC 6J_ putative covalent glycosyl-enzyme intermediate on this enzyme. The incorporation of an electron-withdrawing fluorine substituent at positions adjacent to the site of developing positive charge will destabilize the oxocarbenium ion-like transition states through which each step proceeds. As a consequence, both steps of the reaction will be slowed. However, if the first step, formation of the intermediate, is faster than the second step, then an accumulation of the intermediate will result. Otherwise, the rate of this first step would have to be selectively increased through the introduction of a good leaving group at the anomeric centre. 3.1.2.1 Synthesis of 5-fluoro-cc-D-galactopyranosyl fluoride By virtue of its 5-fluoro substituent, which will destabilize any developing positive charge at 05, and also by having fluoride as a good leaving group at the anomeric centre, 5-fluoro-a-D-galactopyranosyl fluoride (5FGalF, (3.20)) is a potentially useful tool for trapping the putative covalent glycosyl-enzyme intermediate of LgtC. Related compounds of this type have been successfully used to trap the covalent intermediates of a number of oc-retaining glycosidases [17, 101-103]. The synthesis of this compound from 2,3,4,6-tetra-O-acetyl-a-D-galactopyranosyl fluoride (2.12) is outlined in Scheme 3.3. A key step in the sequence was the free radical photobromination of (2.12) to yield (3.18). This reaction is known to be reasonably regio- and stereoselective, with reaction occurring primarily at the tertiary carbon of the 5-position of per-O-acetylated glycopyranosyl fluorides [104-106] to yield the kinetically and thermodynamically favoured product in which the bromine is in the axial orientation [106]. Fluorination of the resulting 5-bromogalactosyl fluoride (3.19) followed by deacetylation afforded the desired compound (3.20). Chapter 3 An Investigation of the Chemical Mechanism of LgtC 62 Scheme 3.3. An outline of the synthesis of 5-fluoro-oc-D-galactopyranosyl fluoride (5FGalF, (3.20)). <2-12> A g B F 4 I R = Br (3.18) (3.20) toluene R = F ( 3 1 9 ) Initial attempts to fluorinate (3.18) using silver fluoride were unsuccessful, despite previously reported successes with the corresponding D-gluco and D-manno-epimers [17, 102]. This fluorination was eventually accomplished using silver tetrafluoroborate in toluene yielding (3.19) as the product of retained configuration at C5 with no detectable formation of the 5-fluoro-L-«/^ra epimer, which would be the product of inverted configuration at C5. The assignment of stereochemistry at C5 is based on the assumption that the favoured conformation of (3.19) is a 4 Ci chair while that of the 5-fluoro-L-tf/fr-o epimer is most likely a boat (Figure 3.1). In each case, the conformation is favoured by the anomeric effect of the fluorine substituent at C5 and it also allows the bulky 6-O-acetyl group of each epimer to adopt a sterically less demanding equatorial position. In the boat conformation, H4 and F5 of the 5-fluoro-L-a/^ra epimer should have a large coupling constant of 20-30 Hz due to their transdiaxial relationship. Since no H4, F5 coupling was observed in the product of the fluorination reaction, it was concluded that the sole epimer formed in this case was that which had the D-galacto configuration at C5. Support for this conclusion was later provided by the results of performing the fluorination in diethyl ether. Under these conditions, a mixture of the D-galacto and L-altro epimers in an approximately 1:1 ratio ( 'H and 1 9 F NMR) were obtained. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 63 Unfortunately, these two products could not be separated, but from the NMR data one set of peaks was identical to those of (3.19). In the second set of peaks, a large coupling constant of 29.7 Hz was indeed observed for H4, thereby validating the assumption that the favoured conformation of the L-altro epimer is a boat. Figure 3.1. Conformations of the (a) 5-fluoro-D-ga/acto epimer and the (b) 5-fluoro-L-altro epimer. Haworth projections are shown along the C-4, C-5 axis. Two explanations are possible for the observed retention of configuration when (3.18) is fluorinated in toluene in contrast to that seen for sugars with an equatorial acetate substituent at C-4 when mixtures of the two epimers are typically obtained [107]. One possibility is that neighbouring group participation by the axial 4-O-acetyl group of (3.18) yields a 4,5-acetoxonium ion intermediate that can only undergo attack at C5 from the bottom face, yielding the D-galacto epimer. Alternatively, it may be that (3.19) is simply the thermodynamically favoured product from the fluorination of (3.18) and that equilibration of the epimers occurs in this case. The boron trifluoride present in the reaction mixture could certainly act as the Lewis acid catalyst for this epimerization reaction. As mentioned earlier, when diethyl ether was used as the solvent, in place of a F b F Chapter 3 An Investigation of the Chemical Mechanism of LgtC 64 toluene, both the D-galacto and L-altro products were indeed obtained. Presumably, when ether is used as the solvent, all of the boron trifluoride is complexed by the excess ether, thereby decreasing its capacity to function as a Lewis acid catalyst. In contrast, no complexation of boron trifluoride occurs in toluene. As a result, this Lewis acid is free to catalyze the aforementioned epimerization reaction to yield the most stable product. However it is not clear why equilibration would occur more rapidly in the galacto than in the gluco series, thus participation of a neighbouring group seems to be at least partially responsible for the outcome. 3.1.2.2 Attempts to trap the covalent intermediate of L g t C with 5-fluoro-cx-D-galactopyranosyl fluoride To trap the putative covalent glycosyl-enzyme intermediate of LgtC, 5-ffuoro-oc-D-galactopyranosyl fluoride (3.20) was incubated with the enzyme in the presence of UDP. Aliquots of the reaction mixture were removed over the course of three hours and quenched by the addition of 6 M urea followed by storage at -78 °C. The samples were subsequently analyzed by electrospray mass spectrometry to look for an increase in the mass of the protein, which would correspond to the attachment of a 5-fluorogalactosyl moiety. Unfortunately, no accumulation of such an intermediate could be detected, even at high concentrations (50 mM) of 5FGalF. This result was not entirely unexpected however, since oc-D-galactosyl fluoride is itself a very poor substrate for the enzyme in terms of both binding and turnover, even in the presence of exogenous UDP (Section 2.3). Therefore, it is very likely that 5FGalF also binds poorly to the enzyme and is not turned over at all. Indeed, no release of fluoride was observed when 5FGalF (5 mM) was Chapter 3 An Investigation of the Chemical Mechanism of LgtC 65 incubated with high concentrations of LgtC (2.5 mg/mL) nor was any inhibition of the enzyme observed in the presence of 5 mM of this compound. 3.1.2.3 Synthesis of uridine 5-diphospho-(2"-deoxy-2 "-fluoro)-a-D-galactopyranose In light of the poor binding and turnover by LgtC of substrates that lack a UDP moiety, it is therefore crucial that this important recognition element be retained in analogues of the donor substrate. Such a requirement is a major obstacle in the application of the fluorosugar strategy for trapping the covalent glycosyl-enzyme intermediate of this enzyme since the reactivity of the leaving group is not easily modified. The success of compounds such as uridine 5'-diphospho-(2"-deoxy-2"-fluoro)-a-D-galactopyranose (UDP-2FGal, (3.24)) in this endeavour would therefore rest heavily on the hope that the leaving group ability of UDP is sufficient to ensure that the first step of the reaction proceeds faster than the second step in the presence of the 2-fluoro substituent. The preparation of UDP-2FGal from galactal is outlined in Scheme 3.4, with the key intermediate being the 2-deoxy-2-fluoro-a-D-galactopyranose-l-phosphate (3.23). This compound was synthesized by first fluorinating the triester of galactal with Selectfluor® in the presence of water to yield the corresponding acetate-protected 2-fluoro hemiacetal of galactose (3.21). After protection of the anomeric hydroxyl as a (3-acetate, the resulting per-O-acetylated 2-fluorogalactopyranose (3.22) was then phosphorylated according to the procedure of MacDonald [108, 109] in neat phosphoric acid. Following the removal of the acetates and the subsequent conversion to the mono(tri-n-octyl)ammonium salt, (3.23) was then reacted with UMP-morpholidate Chapter 3 An Investigation of the Chemical Mechanism of LgtC 66 according to the procedure described by Wittmann and Wong [110] to yield UDP-2FGal. While this reaction for the synthesis of UDPGal was reported to be complete within a few days when 1H tetrazole was present, we found that a significantly longer period of time was required for this coupling to occur between 2-deoxy-2-fluoro-a-D-galactopyranose-1-phosphate (3.23) and UMP-morpholidate. Scheme 3.4. An outline of the synthesis of uridine 5'-diphospho-(2"-deoxy-2"-fluoro)-oc-D-galactopyranose (UDP-2FGal, (3.24)). O A c . OAc 0 A S - OAc 0 A S - OAc \X o Selectfluor ®^ | A ^ 0 1 • Ac 2 Q, pyridine | V 0 A C 0 ^ ^ 50 °C * AC°~^SOH 2. HBr/AcOH, CH 2 C1* ^ ^ ^ ^ H 2 0 / DMF 3 . Hg(OAc),, AcOH (3.21) 6 V n (3.22) 1. H3PO4 2. LiOH 3. N ( C 8 H 1 7 ) 3 9 H - O H HO " • ^ " V ^ O O f ' N H V I II II I O - P - O - P - O . ^ s . 1 1 M 2NH4 OH OH 1. UMP-morpholidate L . ^ ^ 0 ^ • HO ss\ "r\\ "^ 1 r ^ ^ " * 0 ^ 1H tetrazole, pyridine X |T| ?\ Q® 0 0 © O L,°>J 'VI O - P - 0 NB(C,Hn) _ . , 2. N H 4 O A C I © \ I 4 OH OH OH (3.24) (3.23) 3.1.2.4 Kinetic evaluation of UDP-2FGal and its employment for the trapping of the putative covalent glycosyl-enzyme intermediate Kinetic evaluation of UDP-2FGal found this compound to be neither a time-dependent inactivator nor a slow substrate of LgtC, even at high concentrations (250 p,g/mL) of enzyme. Instead, LgtC was found to be reversibly inhibited by UDP-2FGal. With respect to the donor substrate UDPGal (Km = 13 pM), inhibition by UDP-2FGal Chapter 3 An Investigation of the Chemical Mechanism of LgtC (57 was shown to be competitive (Figure 3.2a) with an apparent K ; value of 2 pM. When the inhibition study was repeated using lactose as the varying substrate, the double reciprocal plot yielded a noncompetitive pattern (Figure 3.2b) from which the values for K e i and K e i S were calculated to be 230 pM and 75 pM respectively. The inability of UDP-2FGal to inactivate LgtC or function as a slow substrate for this enzyme suggests that a stabilized covalent fluoroglycosyl-enzyme intermediate was most likely not formed. Indeed, no trace of such an intermediate could be detected by ESMS when LgtC was incubated with UDP-2FGal either alone or in the presence of lactose or 4dLac. This failure to trap the intermediate once again raises doubts about its possible existence. However, given that no turnover of UDP-2FGal was observed, the more likely explanation is that the reactivity of UDP as a leaving group was insufficient after all to overcome the destabilizing effects of the 2-fluoro substituent on the transition state of the first step. As such, it may not have been possible for the intermediate to even form. 1/[UDP-Gal] (uM)"' 1/[lactose] (mM)'1 Figure 3.2. Double reciprocal plot showing the inhibition of LgtC by UDP-2FGal at varying concentrations of (a) the donor substrate UDPGal and (b) the acceptor substrate lactose. When the UDPGal concentration was varied, the concentrations of UDP-2FGal were 0 (•), 6.25 (O), 12.5 (•) and 25 p M (•). When the concentration of lactose was varied, the concentrations of UDP-2FGal were 0 (•), 37.5 (O), 75 (A), 150 (A) and 300 p M (•). Chapter 3 An Investigation of the Chemical Mechanism of LgtC 68_ 3.2 Revelations from the X-ray Crystal Structure and its Implications for Catalysis Attempts to obtain an X-ray crystal structure of LgtC were done in collaboration with the laboratory of Professor Natalie C. J. Strynadka in the Department of Biochemistry and Molecular Biology at the University of British Columbia. Crystallization of this enzyme was made possible only in the presence of the tight binding inhibitor UDP-2FGal, yielding crystals that diffracted to better than 2 A resolution. In addition to this binary complex of LgtC with UDP-2FGal bound, a structure of the ternary complex of the enzyme with analogues of both substrates was also obtained when the incompetent acceptor substrate, 4dLac was soaked into the LgtC - UDP-2FGal crystals. This structure of LgtC is the first reported structure of a retaining glycosyltransferase and it is also the first structure of any glycosyltransferase in which the sugar moieties of both the donor and acceptor substrates are clearly visible in the electron density map. The structure of LgtC is monomeric and is organized into two domains. The large N-terminal domain is composed of a mixed ot/p fold and contains the active site of the enzyme (Figure 3.3). The smaller C-terminal domain is predominantly a-helical and is postulated to mediate the attachment of the enzyme to the cell membrane (Figure 3.3). From the structure, the UDP-2FGal appears to be held tightly in place by two loops that fold over this substrate upon its binding to the enzyme. These two loops are most likely disordered in the absence of the donor substrate, hence explaining the inability to crystallize the enzyme in the absence of UDP-2FGal. The closure of these two loops over the donor substrate also leads to the formation of the binding site for the acceptor Chapter 3 An Investigation of the Chemical Mechanism of LgtC 69 substrate. In conjunction with the information garnered from kinetic studies of the enzyme (Section 2.4), the data provided by the structure was invaluable in determining the kinetic mechanism of LgtC to be ordered BI BI, with the donor substrate UDPGal binding first followed by the acceptor substrate, lactose. Membrane Attachment Domain Figure 3.3. Structural view of the LgtC-UDP-2FGal-4dLac ternary complex showing the large N-terminal catalytic domain and the smaller C-terminal membrane attachment domain. From the structure of the binary LgtC - UDP-2FGal complex, the donor substrate analogue is found to be almost completely buried within the enzyme. Inside the confines of the active site, the UDP-2FGal analogue is shown to adopt an unusual folded conformation in which the UDP moiety is bound in an extended conformation but is tucked back underneath the pyranose ring of the galactose. A single well-ordered Mn 2 + ion is shown to coordinate the two phosphate oxygens of UDP and also the side chain atoms of three amino acid residues that are conserved amongst all family 8 Chapter 3 An Investigation of the Chemical Mechanism of LgtC 70 glycosyltransferases (His244, Asp 103 and Asp 105). Binding of the acceptor substrate was shown to result in very little change in the LgtC - UDP-2FGal structure. A molecular surface representation of this ternary complex is shown in Figure 3.4, revealing that the acceptor sugar is much more solvent accessible than is the donor sugar. The occluded nature of the donor sugar is perhaps necessitated by the need to shield this substrate or any reactive intermediates that may be formed from it from unwanted hydrolysis. Figure 3.4. A molecular surface representation of LgtC showing the deeply buried donor substrate analogue (UDP-2FGal) and the more solvent exposed acceptor substrate analogue (4dLac). From a mechanistic standpoint, the availability of the three-dimensional structure of LgtC is a valuable resource for identifying residues that may be important to catalysis. In the double displacement mechanism that has been proposed for LgtC, a key residue in Chapter 3 An Investigation of the Chemical Mechanism of LgtC 77 the reaction is the catalytic nucleophile, which is responsible for attacking the reactive anomeric centre of the donor substrate in the first step of the mechanism. Since previous attempts to trap the resulting intermediate and identify the catalytic nucleophile have failed, the crystal structure of LgtC in complex with analogues of both the donor and acceptor substrates provides an alternative means of identifying candidates that may be capable of functioning in such a capacity. A representation of the active site of LgtC showing the interactions with the substrate analogues is shown in Figure 3.5a. An examination of residues in close proximity to the reactive Cl "-position of the donor sugar revealed three candidate nucleophiles (Figure 3.5b). The first candidate is the carbonyl oxygen from the amide side chain of Glnl89. The remaining two candidates are not enzymic residues but are instead from the acceptor substrate, namely the 6'- and 3'-hydroxyl groups. The conspicuous absence of a carboxylic acid residue amongst this group of candidate nucleophiles was somewhat unexpected since the nucleophiles of almost all retaining glycosidases identified to date have been shown to be a carboxylate, either from an aspartic acid or a glutamic acid residue. Nonetheless, each of the candidate nucleophiles was individually evaluated for their ability to function in such a role. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 72 Figure 3 . 5 . A (a) schematic representation of the active site of LgtC showing the interactions between the enzyme and the substrate analogues and (b) a view of residues within close proximity to the C I " reaction centre. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 75 3.2.1 Evaluation of the 3'- and 6'-hydroxyls of the acceptor lactose as possible nucleophiles The prospect of having either the 3-OH or the 6'-OH of the acceptor lactose functioning as the nucleophile is certainly intriguing in that such a mechanism would inherently demand the formation of a ternary complex prior to any chemical catalysis. This scenario would not only minimize the unwanted hydrolysis of any potentially reactive intermediates, but it would also be fully consistent with the ordered BI BI kinetic mechanism that has been determined for LgtC (see Section 2.4). From the perspective of a chemical mechanism, attack at the reactive CI" centre of the donor substrate by either the 6'-OH or the 3'-OH of the acceptor lactose will lead to the formation of the respective Gal-p-l,6-Lac or Gal-|3-1,3-Lac intermediate. A subsequent intramolecular transglycosylation of the galactosyl moiety to the hydroxyl of the 4'-position of either intermediate will result in the formation of the retained Gal-oc-l,4-Lac product. To test the validity of such a mechanism, several experimental options are available. One method is to remove the alleged nucleophilic hydroxyl group from the acceptor substrate and then evaluate the behaviour of the resulting compound in the LgtC catalyzed transfer reaction. Another test of this potential mechanism is to chemically synthesize the putative intermediate(s) and then to look for its conversion to the expected product upon incubation with the enzyme. 3.2.1.1 Synthesis of 6'-deoxylactose To investigate the possibility that the 6'-OH of lactose is the catalytic nucleophile, a 6'-deoxy analogue of this acceptor substrate was synthesized (Scheme 3.5). Removal of the hydroxyl group from the 6'-position was achieved by first replacing it with a Chapter 3 An Investigation of the Chemical Mechanism of LgtC 74 bromine substituent through a ring opening reaction of the benzylidene acetal of (3.1) with A -^bromosuccinimide (NBS) [111, 112]. The resulting 6'-bromo derivative (3.25) was then reacted with tributyltin hydride in a free radical reaction to yield the ester-protected 6'-deoxylactose (3.26). The final deprotected compound (6dLac, (3.27)) was obtained when the benzoyl and acetyl groups were removed in a dilute solution of sodium methoxide in methanol. Scheme 3.5. An outline of the synthesis of 6'-deoxylactose (6dLac, (3.27)). Ph (3.27) (3.26) 3.2.1.2 Synthesis of the putative intermediates, Gal-P-1,6-Lac and Gal-P-1,3-Lac The putative intermediate in the case where the 6'-OH of lactose is the catalytic nucleophile is a Gal-f3-l,6-Lac species. The chemical synthesis of this trisaccharide was accomplished by first condensing the selectively protected lactose (3.29) with a trichloroacetimidate of galactose (3.31) followed by deprotection as shown in Scheme 3.6. The isolation of the 6'-OH of lactose for the glycosylation reaction was achieved Chapter 3 An Investigation of the Chemical Mechanism of LgtC 7 5 through a series of protection/deprotection steps beginning with the reductive ring opening of the benzylidene acetal of (3.1) with sodium cyanoborohydride and anhydrous hydrochloric acid to yield (3.2), as was done previously for the synthesis of 4dLac (Scheme 3.1). Acetylation of the 4'-OH of this compound followed by hydrogenolysis of the benzyl group at the 6'-position afforded (3.29) with the 6'-OH free for glycosylation. The galactopyranosyl trichloroacetimidate (3.31) was prepared by first selectively deprotecting the anomeric acetate of per-O-acetylated galactose (2.11) with hydrazine acetate. The resulting hemiacetal (3.30) was then reacted with trichloroacetonitrile in the presence of DBU to yield the corresponding imidate (3.31). The chemical synthesis of Gal-f3-l,3-Lac (3.34) to investigate the possibility that the 3'-OH of lactose is the nucleophile is a much more challenging endeavour since isolation of the hydroxyl group at this position is a nontrivial task. To avoid the exhaustive protection/deprotection schemes that would be needed for such an undertaking, an enzymatic means for the synthesis of this trisaccharide was sought. To this end, the talents of a (3-1,3-galactosidase from Xanthamonis manihotis were employed. This p-retaining glycosidase normally catalyzes the hydrolysis of Gal-p-1,3-linkages but transglycosylation is also known to occur with this enzyme in the presence of a reactive donor and suitable acceptor to form Gal-p-l,3-linkages [113, 114]. By incubating this enzyme with lactose and the activated donor, /?-nitrophenyl P-D-galactopyranoside, the trisaccharide Gal-P-1,3-Lac (3.34) was easily obtained without the need for any protecting group chemistry. Confirmation of the linkage was achieved by analysis of the 'H NMR spectrum of the acetylated material. Chapter 3 An Investigation of the Chemical Mechanism of LgtC Scheme 3.6. An outline of the synthesis of Gal-|3-1,6-Lac (3.33). 76 OH OBn Q A c OAC ASI OAc OAc, QAc AcO - A ^ — O A c OAc (3.2) Ac 20, pyridine (2.11) H 2NNH 2-HOAc DMF OAc,OBn A c o i i ^ r ^ c o OAc OAc O AcO I OAc. OAc 1 ^ 0 AcO » A ^ - - r - - \ _ OAc^ OH OAc (3.28) H 2 , Pd/C EtOAc/MeOH AcOH (cat.) (3.30) DBU, C13CCN CH2C12 -40 °C to -20 °C OAc, OH A c o S ^ ^ £ 0 ^ . 0 A C AD ^-OAc (3.29) OAc OAc, OAc AcO-A— AcO I O ^ N H (3.31) T 1 C C h BF 3-Et 20, DCE 4 A sieves 1 -OAc A c 0 ^ O A c OAc A c 0 ^ ^ o A c A c 0 X > ^ : OAc OAc (3.32) NaOMe, MeOH H 0 -OH 0 H ,OH OH H 0 - ^ \ 0 H HO OH OH (3.33) Chapter 3 An Investigation of the Chemical Mechanism of LgtC 77 3.2.1.3 Evaluating the likelihood that the catalytic nucleophile is a hydroxyl group from the acceptor substrate To address the possibility of the catalytic nucleophile being the 6'-OH of lactose, the LgtC catalyzed transfer reaction was investigated using 6'-deoxylactose as the acceptor substrate. When this experiment was carried out, no turnover product was observed, which is consistent with the 6'-OH having an important role in catalysis. Unfortunately, this compound also failed to act as an inhibitor of LgtC, suggesting that it was unable to bind to the active site. As a consequence of this latter observation, results from using 6dLac as a substrate are rendered meaningless in terms of providing evidence for the 6'-OH being the nucleophile. These results do however indicate an important binding interaction at this position of the acceptor substrate. On the basis of the crystal structure, this interaction is primarily through a hydrogen bond with Asp 130 of the enzyme. An alternate means of investigating the possibility that the 6'-OH of lactose is the nucleophile is to examine the ability of LgtC to catalyze the turnover of the putative intermediate that would be formed if such was the case. To carry out this experiment, LgtC was incubated with Gal-(3-l,6-Lac and UDP in the presence of an a-1,4-galactosidase from green coffee bean. This particular oc-galactosidase is a retaining exoglycosidase that specifically hydrolyzes oc-l,4-linked galactose residues from glycoconjugates. Consequently, if Gal-(3-l,6-Lac is converted by LgtC to the Gal-a-1,4-Lac product, then this latter trisaccharide will be cleaved by the a-galactosidase to yield galactose and lactose, the formation of which would be easily monitored by TLC or HPLC. In a similar manner, the putative intermediate, Gal-p-l,3-Lac that would be Chapter 3 An Investigation of the Chemical Mechanism of LgtC 78 formed from the 3'-OH of lactose acting as the nucleophile was also examined for turnover by LgtC. Unfortunately, when both reaction mixtures were analyzed, no cleavage products were observed, suggesting that neither Gal-P-1,6-Lac nor Gal-P-1,3-Lac were converted to Gal-a-l,4-Lac. On the other hand, both galactose and lactose were indeed observed when aliquots of the control experiment in which LgtC was incubated with Gal-a-l,4-Lac were analyzed by TLC and HPLC. While the notion of having either the 3'-OH or the 6'-OH of lactose be the nucleophile is therefore strongly disfavoured by the outcome of these experiments, the plausibility of such a mechanism cannot be completely ruled, as follows. A danger associated with investigating the kinetic competence of a chemically synthesized intermediate is that the rate of reaction of this compound, which is normally not released by the enzyme, is dependent on several factors. The most obvious concern is the relative affinities of the enzyme for the substrate and the intermediate. To avoid the accidental release of the intermediate from the active site, the conformation that is adopted by the enzyme to bind the intermediate may be very different from that used to bind the substrate. In such a scenario, the free enzyme, in its ground state conformation, would therefore have a very low affinity for the chemically synthesized intermediate. In addition to the enzyme's relative affinities for the substrate and intermediate, the turnover rate of the externally added intermediate is also further dependent upon the equilibrium constants for the formation of this moiety both on the enzyme and in solution [115]. In the case where the formation of the intermediate is far more favourable on the enzyme than in solution, then the commonly held view that the intermediate will react as fast or faster than the normal substrate will be true. However, if the equilibrium constants for Chapter 3 An Investigation of the Chemical Mechanism of LgtC 79 intermediate formation are very similar regardless of whether it occurs on the enzyme or in solution, then the reaction rate for the turnover of this species will be much slower. For these reasons, it is therefore not possible to completely rule out either Gal-(3-l,3-Lac or Gal-P-1,6-Lac as intermediates of the LgtC-catalyzed reaction even though no turnover of these compounds was observed. 3.2.2 Evaluation of Glnl89 as the possible nucleophile On the basis of the crystal structure, the only enzymic residue that is suitably positioned to function as the catalytic nucleophile is the side chain amide of Gin 189. While amides are notorious for being poor nucleophiles, their ability to still function in such a capacity is not without precedent. An enzymatic example of this is found within the mechanism of JV-acetylhexosamihidases from glycosidase families 18 and 20. Unlike all other retaining glycosidases studied to date, which have been unequivocally shown to employ an enzymic carboxylate in the role of the catalytic nucleophile, these hexosaminidases instead utilize the 2-acetamido group of the substrate itself for this purpose (Figure 3.6) ([116, 117] and references therein). The result is the formation of an oxazolinium ion intermediate rather than the typical covalent glycosyl-enzyme intermediate, as the carbonyl oxygen of the amide attacks at the reaction centre to displace the aglycone. In the case of LgtC, it is therefore conceivable that the carbonyl oxygen of the side chain amide of Gin 189 could perform a similar task of attacking the Cl"-position of the UDPGal substrate to displace the U D P leaving group. In terms of both distance and trajectory, the carbonyl oxygen of this side chain amide is ideally situated for attacking the CI "-position of UDPGal. It is located 3.5 A from the reaction centre and is positioned above the galactose ring such that it is almost in-line with C I " Chapter 3 An Investigation of the Chemical Mechanism of LgtC 8JJ and OI". In this position, this amide oxygen is poised for an SN2-like attack to displace the UDP leaving group from the backside (Figure 3.5b). The advantage that would be gained by LgtC in having a neutral amide nucleophile as opposed to a negatively charged carboxylate is that charge repulsion will be minimized between the attacking nucleophile and the negatively charged UDPGal substrate. Figure 3.6. The catalytic mechanism of A'-acetylhexosaminidases from glycosidase families 18 and 20. Aside from the aforementioned advantages, the notion of having Gin 189 be the catalytic nucleophile of LgtC is also seductive because of its possible tie in with the mechanism of glycogen phosphorylase. Glycogen phosphorylase is the enzyme responsible for the phosphorolytic cleavage of starch and glycogen to produce a-D-glucose-1-phosphate (Glc-l-P) and a shortened polysaccharide (Figure 3.7). However, because the equilibrium constant of the reaction is very close to unity, the reverse reaction is catalyzed equally well by the enzyme. In that case, a glucose moiety from glucose-1-phosphate is added onto a growing chain of glucose residues via a-1,4-linkages. Although glycogen phosphorylase is therefore not a nucleoside diphosphosugar-dependent glycosyltransferase, it is nonetheless a retaining a-glycosyltransferase. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 81 Figure 3.7. The reaction of glycogen phosphorylase. While much work has been done on glycogen phosphorylase, its catalytic mechanism has unfortunately remained unclear. Of particular intrigue has been the role of the essential cofactor pyridoxal phosphate (PLP), which is covalently bound to the enzyme. While the function of this cofactor in phosphorylase is quite different from that of other enzymes, its precise role is still a mystery. From 3 1P NMR studies in conjunction with kinetic studies using glycosidic substrates that lack a phosphate moiety, this cofactor was postulated to function in the role of an acid/base catalyst [118-121]. However, pH dependent studies with analogues of PLP [122, 123] have since negated this idea. While PLP may not be involved in an essential proton transfer during catalysis, this cofactor may nonetheless have an important role in aiding the departure of the phosphate leaving group. Evidence in support of this notion has come from a study in which pyridoxal(5')diphospho(l)-a-D-glucose (PLPPGlc), a covalently linked analogue of both Glc-l-P and PLP, was shown to be capable of transferring its glucosyl moiety to acceptor oligosaccharides with net retention of anomeric configuration [124]. From this study, the highly coordinated phosphate moiety of PLP was therefore proposed to act as an electrophile, producing an electrophilic pull on the substrate phosphate and thereby labilizing the glycosidic bond for reaction (Figure 3.8). Whatever the chemical basis, the arrangement of PLP plus Glc-l-P is reminiscent of way in which the donor sugar is Chapter 3 An Investigation of the Chemical Mechanism of LgtC 82 bound in a glycosyltransferase, suggesting a similar mechanism and possibly a similar evolutionary origin. Figure 3.8. An illustration of the proposed role of the coenzyme phosphate as an electrophile in the postulated catalytic mechanism of glycogen phosphorylase [124]. In terms of the stereochemical outcome of the phosphorylase-catalyzed reaction, the retention of anomeric configuration in the reaction products suggests that glycogen phosphorylase employs a double displacement mechanism similar to that used by retaining glycosidases. However, no catalytic nucleophile has been yet identified in this enzyme despite the availability of numerous three-dimensional structures. Of particular interest however are two recent structures of this enzyme in which one is in complex with a transition state analogue inhibitor [125] and the other is a ternary complex with a thiooligosaccharide and phosphate [126]. In both cases, the bound PLP cofactor was found to be positioned underneath the glucose residue in much the same way that the UDP moiety of UDP-2FGal was found to be tucked back underneath the pyranose ring of Chapter 3 An Investigation of the Chemical Mechanism of LgtC 83_ the 2-fluorogalactose in LgtC. From these two structures of glycogen phosphorylase, the most likely candidate to function as the catalytic nucleophile, by virtue of its close proximity to the reaction centre (3.2 A), is the carbonyl oxygen in the amide backbone of His345. Intriguingly, the position of this atom in relation to the reaction centre of the glycogen phosphorylase catalyzed reaction is almost identical to that of the carbonyl oxygen in the amide side chain of Gin 189 in LgtC. Given the parallels in the reactions of these two enzymes as well as the similarities in the structure of glycogen phosphorylase to other transferases [127, 128], it is therefore possible that the amide in the side chain of Glnl89 and in the backbone of His345 play similar roles in their respective enzymes; that being the catalytic nucleophile. 3.2.2.1 Kinetic analyses of Glnl89 mutants of LgtC To explore the possibility that the side chain amide of Gin 189 is the catalytic nucleophile in LgtC, various mutants of the enzyme at this position were generated and subjected to kinetic analysis. The kinetic parameters of the wild type (WT) enzyme along with the alanine (Ala), glutamic acid (Glu) and asparagine (Asn) mutants are summarized in Table 3.2. For both of the substrates, lactose and UDPGal, all three mutants showed approximately the same level of reduction in activity (approximately 25-fold based on kcat values) relative to that of the wild type enzyme. While the affinity of the enzyme for UDPGal remained relatively unchanged upon mutation of Gin 189, this was not so in the case of lactose, as the K m values for this substrate were considerably higher (7-fold to 9-fold) for the three mutants in relation to that observed for the WT enzyme. This observation is consistent with the formation of an important hydrogen bond between this residue and the 6'-OH of the acceptor substrate, as can be seen from Chapter 3 An Investigation of the Chemical Mechanism of LgtC 84_ the three-dimensional structure (Figure 3.5a). Such a difference in the lactose K m values between the mutants and the wild type enzyme confirms that the approximately 4% residual activity that is observed in these mutants is in fact genuine and not a result of wild type contamination. In addition to this evidence, both the Glnl89Glu and Glnl89Asn mutants also showed substrate inhibition at high concentrations of UDPGal, whereas this was not observed with the wild type enzyme. Table 3.2. Kinetic parameters of wild type LgtC and the various Gin 189 mutants for the substrates lactose and UDPGal. Lactose UDPGal kcat K m k c a t / K r n kcat K m k c a t / K m K j (sT1 (mM) CmM-sV1 (sT1 (uM) fuM-sV1 (pM) Wild Type 24 20 1.2 14 18 0.81 — Gin 189 Ala 1.0 136 0.0074 0.43 25 0.017 — Glnl89Glu 1.1 185 0.0056 0.68 23 0.030 265 Glnl89Asn 1.1 136 0.0062 0.62 6 0.10 1220 error range in data is from 5% - 20% Given the important role of the catalytic nucleophile in a classical double displacement mechanism, the high residual activity that is observed in the Gin 189 mutants of LgtC certainly cast doubts upon the possibility that this residue functions in such a capacity. As a comparison, mutation of the catalytic nucleophile in retaining glycosidases typically results in a 105-fold decrease in the value of k c a t [116, 117]. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 85_ Nonetheless, Gin 189 is clearly important to catalysis if not essential. For example, instead of functioning as a nucleophile, this residue may simply be acting to stabilize the developing positive charge on the oxocarbenium ion through a charge-dipole interaction. Such a mechanism involving an ion pair-like intermediate of some sort would offer more flexibility in that the need for the precise positioning of a nucleophile would be minimized. As a consequence, it is more probable that significant levels of activity may be retained when conservative mutations such as those of the Glnl89Glu and Glnl89Asn mutants are made, as stabilization of the oxocarbenium ion might still be afforded by both Glu and Asn. The high residual activity of the Glnl89Ala mutant on the other hand is more difficult to explain. One possibility is that the activity of this mutant is simply an artifact of the assay, which essentially monitors the production of UDP. Since mutation of Gin 189 to a much smaller alanine residue creates a large cavity into which a water molecule can be accomodated, the activity that is observed with the Glnl89Ala mutant may therefore simply reflect UDPGal hydrolysis rather than galactose transfer. However, when the reaction products of this mutant were analyzed, only transfer products were observed, with no detectable formation of galactose. The addition of exogenous nucleophiles such as azide, formate, acetate, formamide and acetamide also did not result in any rate enhancements (see Section 1.2.3.1). This observation suggests that although a cavity is presumably present in the Glnl89Ala mutant, the binding of water or other small nucleophiles into this space does not occur. Since the activity of the alanine mutant appears to be genuine, one possible explanation that cannot be ruled out is that an adjacent residue may serve as a substitute for Gin 189 within this mutant, acting to stabilize the oxocarbenium ion but with much lower efficiency. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 86 3.3 Investigation of a Potential Anhydrosugar Intermediate Our search for a catalytic nucleophile in LgtC has thus far rested on the notion that the initial displacement reaction in the double displacement mechanism is an intermolecular process. With the aid of the three-dimensional structure, this assumption has led to the identification of three candidate nucleophiles (Gin 189 along with the 3'-and 6'-hydroxyls of lactose), which have since been shown to be unlikely to function in such a capacity. Prior to the availability of the three-dimensional structure of LgtC, other candidate nucleophiles were also evaluated, including the 4"- and 6"-hydroxyls of UDPGal. In the event that one of these hydroxyl groups is the nucleophile, the intermediate that would be formed as a result would be either a p-l,4-anhydrosugar or a P-l,6-anydrosugar (Figure 3.9). Each intermediate could then proceed to react with the acceptor substrate in the second step of the mechanism to yield a product of retained anomeric configuration. To evaluate the likelihood that the mechanism of LgtC proceeds through an intramolecular displacement of the UDP leaving group, each of the putative anhydrosugar intermediates was assessed for its ability to be turned over by the enzyme. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 87 a) Figure 3.9. A possible anhydrosugar mediated double displacement mechanism where the role of the nucleophile in the first step of the reaction is played either by the (a) 4" - O H or the (b) 6"-OH of UDPGal. 3.3.1 Synthesis of (3-1,4- and |3-l,6-anydrogalactose One means of evaluating the possibility that the mechanism of LgtC proceeds through a P-l,6-anhydrogalactose intermediate is to investigate the catalytic competence of this compound as an enzyme substrate. Starting from galactose, the synthesis of P-1,6-anhydrogalactose (3.37) was accomplished according to Scheme 3.7. The key intermediate in this sequence is the per-O-acetylated p-galactosyl fluoride (3.36), which was obtained through a displacement reaction of the corresponding a-galactosyl bromide (3.35) with silver (I) fluoride. Upon deprotection of (3.36) under basic conditions, the P-1,6-anhydrogalactose (3.37) was formed [129, 130]. The mechanism of this reaction is believed to involve the transient formation of a 1,2-epoxide intermediate. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 88 Scheme 3.7. An outline of the synthesis of p-l,6-anhydrogalactose (3.37). AcO-O A c , OAc OAc HBr/AcOH CH 2C1 2 AcO. O A c , OAc AgF, C H 3 C N AcO O A c , OAc NaOMe, MeOH HO Br OH (2.11) (3.35) (336) (3.37) To investigate the possibility that the 4"-OH of UDPGal is the catalytic nucleophile, the intermediate that would result from such an event (p-1,4-anhydrogalactose) would also be needed. This compound was obtained as a generous gift from Professor Martin E. Tanner at the University of British Columbia. 3.3.2 Evaluating the possibility that the intermediate in the mechanism of LgtC is either a |3-l,4-anhydrogalactose or a P-l,6-anhydrogalactose species In the event that the 4"-OH or the 6"-OH of UDPGal is the catalytic nucleophile of LgtC, the intermediate that would be formed in either case would be an anhydrosugar. If this is the case, then the corresponding anhydrosugar should be able to function as a substrate for the enzyme. To carry out this experiment, the enzyme was first incubated with either P-l,4-anhydrogalactose or p-l,6-anhydrogalactose in the presence of UDP and lactose; the reaction mixtures were then analyzed by TLC and HPLC for the formation of the expected trisaccharide product. Unfortunately, none was detected with either anhydrosugar nor were any hydrolysis products observed. While this result does not completely eliminate the potential involvement of an anhydrosugar intermediate in the mechanism of LgtC, it does strongly disfavour this possibility. The recent report of Chapter 3 An Investigation of the Chemical Mechanism of LgtC 89 turnover of both UDP-4 "-deoxygalactose and UDP-6"-deoxygalactose by LgtC also argues against such an intramolecular mechanism [131]. Furthermore, a re-examination of the three-dimensional structure of this enzyme in complex with analogues of both the donor and acceptor substrates also reveals no obvious clues to suggest the involvement of either the 4"-OH or the 6"-OH in catalysis. With both substrate analogues bound in this structure, it is not unreasonable to expect that all residues involved in catalysis would be correctly positioned to carry out their task. However, neither the 4"-OH nor the 6"-OH appears to be poised to function as the catalytic nucleophile to displace UDP in the first step of the double displacement mechanism (Figure 3.5). From the structure, the 6"-OH is oriented away from the anomeric centre and is held in place through a hydrogen bond with the side chain carboxylate of the conserved Asp 188 residue. As for the 4"-OH, this residue is also ill positioned for the role of the nucleophile since the 4 Ci conformation adopted by the galactose ring of the donor substrate places this hydroxyl group far from the reactive anomeric centre. Given the constraints of the binding site, particularly the coordination of the pyrophosphate moiety by the enzyme-bound manganese ion, distortion to a boat conformation to facilitate attack by the 4"-OH is highly improbable. These structural observations therefore strongly support the chemical data, which suggests that an anhydrosugar intermediate is unlikely to be involved in the mechanism of LgtC. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 90 3.4 Probing the Existence of a Long-Lived Intermediate by Positional Isotope Exchange The failure of our attempts to directly observe and/or isolate an intermediate in the LgtC catalyzed reaction therefore necessitated the use of more indirect approaches to probe the mechanism of LgtC and also the nature of any likely intermediates. Since cleavage of UDP from the donor substrate UDPGal is expected to proceed with the breaking of the exocyclic Cl"-01" bond, a positional isotope exchange study was carried out not only to confirm this idea but more importantly as a means of demonstrating the possible existence of a long-lived intermediate after bond cleavage. In such an experiment, 180-labeled uridine 5'-diphospho-a-D-galactopyranose (with the 1 8 0 incorporated at the bridging position between galactose and UDP) would be incubated with LgtC in the presence of an incompetent acceptor. Without a suitable saccharide 18 substrate to accept the donor galactose, any bond cleavage in the donor O-UDPGal substrate may be simply followed by the reformation of the cleaved bond. However, if bond cleavage occurs between CI" and 01" and the UDP thus formed has a lifetime comparable to, or greater than, that required for bond rotation around the terminal phosphate, then scrambling of the isotope into the nonbridging phosphate positions 18 should be observed in the donor substrate (Figure 3.10). This scrambling of the O from a bridging position to a nonbridging position can be readily monitored by the slight up field shift in the 3 1P NMR signal of the (3-phosphate resulting from the increased bond order of the P- 1 80 bond [132]. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 91 Figure 3.10. Cleavage of the Cl "-01" bond in the substrate UDPGal followed by rotation of the p-phosphate and re-formation of the bond will result in the scrambling of the 1 80 (represented by the darkened atom) from the bridging to a nonbridging position in the p-phosphate. 3.4.1 Synthesis of uridine 5'-diphospho-(l"-180)-oc-D-galactopyranose The synthesis of uridine 5'-diphospho-(l "-180)-a-D-galactopyranose (180-UDPGal) is outlined in Scheme 3.8 and involves two essential components. The first is the incorporation of an lsO-label at the anomeric position of galactose and the second is the condensation of the 180-labeled a-galactopyranosyl phosphate (3.39) with UMP-morpholidate to yield the desired compound (3.40). The incorporation of the heavy atom at the anomeric centre was achieved by heating the hemiacetal of 2,3,4,6-tetra-O-acetyl-D-galactose (3.30) with l80-labeled water at 105 °C in the presence of acidic resin. After phosphorylation and subsequent removal of the protecting groups, the resulting galactopyranosyl phosphate (3.39) was then condensed with UMP-morpholidate in accordance with the procedure of Wittmann and Wong [110], except that 1H tetrazole was excluded from the reaction mixture. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 92 Scheme 3.8. An outline of the synthesis of uridine 5'-diphospho-(l "-180)-a-D-galactopyranose ( l x O-UDPGal, (3.40)). OAS-- OAc A c O - \ ^ V - A 1. H2"#, CH3CN resin, 105 °C OAC NOH 2.(PhO)2POCl DMAP, CH2C12 A c O - A - — Q AcO J I' 1. H2, PtO,, MeOH » 2. LiOH, THF O A c - O A c (3.30) (3.38) - P — O P h I OPh ^Sl S ©©-A c 0 , „ 0 / = \ (3.39) 0 H 1. N(C„H17)3 2. UMP-morpholidate pyridine, 4 A sieve: OH OH o #— p—o— p—o- \ K o H O I 1 1 1 1 -p—o—p—o 0 ° 0 ° © 2NII 4 (3.40) The presence of 1H tetrazole was reported to significantly increase the rate of the coupling reaction by providing acid catalysis to the departure of the morpholino leaving group and also by nucleophilic catalysis via the formation of a highly reactive phosphotetrazolide intermediate. In our hands however, the addition of this catalyst was shown by 3 I P NMR to result in the significant formation of a major side product. This compound was later isolated and identified to be a dimer of uridine, linked by a triphosphate moiety. Mass spectral analysis of this compound also found it to contain the 1 8 0 isotope. One plausible explanation for the formation of this species is illustrated in Figure 3.11. After the coupling of galactopyranosyl phosphate with UMP-morpholidate to yield l sO-UDPGal, this desired product can potentially undergo a further displacement reaction at the anomeric centre in the presence of a nucleophile to yield a galactose derivative and a UDP moiety in which the (3-phosphate is labeled with an 1 8 0 isotope. This 1 80-UDP can then proceed to react with UMP-morpholidate to yield the 1 80-Chapter 3 An Investigation of the Chemical Mechanism of LgtC 93 containing, triphosphate-linked uridine dimer. The breakdown of the 180-UDPGal is presumably enhanced in the presence of 1H tetrazole as this compound can potentially provide both acid catalysis to the departure of the 1 80-UDP leaving group and also function as the necessary nucleophile. The involvement of some oxocarbenium ion-like character in the transition state of this reaction is also likely since such problems with the decomposition of the uridine diphosphosugar were not encountered when 2-deoxy-2-fluoro-oc-galactopyranosyl phosphate was coupled with UMP-morpholidate (Section 3.1.2.3). In that case, the presence of an electron withdrawing fluorine substituent at the 2-position presumably disfavours the reaction by destabilizing the cationic transition state. o o • II II II 0 Ur — O — P — O — P — O — P — O — Ur / \ M i e A© o© O N — P — O — U r U " U N u c \ / o 0 + O H O H V 0 0 _ \ O O O \ V II II 0 V II II II t - V - O ^ _ w _ U r _ 0 _ p _ 0 _ p _ - , v - ' ^ _ W . U r _ 0 - p - 0 - P - » - P - 0 - U r H O - ^ ^ \ o © o© X o© o© o© H O | • U D P O H , Q H / V + O NH « , 0 0 0 Nuc \ 1 II II II Ur — O — P — O — P — O — P — O — U r o© o© •© Figure 3.11. A proposed route for the production of the triphosphate-linked uridine dimer in the tetrazole aided coupling of 180-galactopyranosyl phosphate with UMP-morpholidate. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 9_4 3.4.2 Search for Positional Isotope Exchange in the LgtC catalyzed reaction In carrying out the positional isotope exchange (PIX) experiment, 180-UDPGal was first incubated with LgtC along with the buffer components that are required by the enzyme for activity. The incompetent acceptor 4dLac was also included in this mixture since the kinetic mechanism of this enzyme demands that the binding sites of both the donor and acceptor substrates be occupied before any catalysis can occur. After an incubation period of 7 hours, the 180-UDPGal was then re-isolated and its proton-decoupled 3 1P NMR spectrum was recorded. Had scrambling of the l s O isotope occurred during the incubation period, two sets of signals corresponding to the (3-phosphorus of 180-UDPGal would have been expected. One would be the result of the 1 8 0 being incorporated back into the bridging position after bond cleavage and reformation while the other would be due to the ' 8 0 being scrambled into the nonbridging positions around this (3-phosphorus. Statistically, the intensities of these two signals would be in a ratio of 1:2 with the latter signal being shifted slightly up field. Upon examination of the spectrum, it was clear that no positional isotope exchange had occurred as the signal corresponding to the P-phosphorus was found to be unchanged when compared to the spectrum of 180-UDPGal prior to the addition of LgtC. 18 To account for the absence of any scrambling of the O isotope, three important facets of the PIX experiment are considered. First and foremost is the cleavage of the Cl"-01" bond, which may not have occurred in the absence of an appropriate acceptor substrate. While the incompetent acceptor 4dLac can occupy the acceptor binding site and meet the substrate binding requirements of an ordered BI BI kinetic mechanism, the Chapter 3 An Investigation of the Chemical Mechanism of LgtC 9_5 correct orientation of key residues for catalysis may not have occurred in the absence of the nucleophilic 4'-hydroxyl group. As such, no bond cleavage would have been initiated. Even had bond cleavage taken place, the observation of isotopic scrambling is further dependent on the ability of the p-phosphate in the resulting UDP to rotate freely in the active site. However, from the crystal structure, the mobility of this phosphate appears to be limited as it is tightly bound by the enzyme via hydrogen bonds to the amide backbone of Gly 247 and the side chain of His 78 as well as through coordination to the enzyme bound Mn 2 + . Such restrictions in the motion of this p-phosphate would therefore limit the likelihood of scrambling of the 1 8 0 isotope from a bridging to a nonbridging position. The inability to observe any PIX as a result of an immobilized phosphate is a common problem that has plagued such studies of other enzymes. Examples that have illustrated this predicament include sucrose synthetase [77], glycogen synthetase [75], famesyl pyrophosphate synthetase [133], argininosuccinate synthetase [134] and bornyl synthetase [135]. In all cases, no PIXs were observed despite the fact that an intermediate was shown to be formed in each case. Finally, in the case where cleavage of the Cl"-01" bond has occurred and rotation of the p-phosphate has followed, the last requirement for the observation of positional isotope exchange is the reformation of the Cl"-01" bond. In this regard, the ability of LgtC to hydrolyze UDPGal is problematic, as the liberated UDP would have to compete with water for reaction with the intermediate in order to reform UDPGal. Failure to do so would result in the hydrolysis of the intermediate rather than scrambling of the isotope. Unfortunately, hydrolysis of UDPGal was indeed observed, as peaks corresponding to those of l sO-UDP did slowly appear in the proton-decoupled 3 1P NMR spectra taken of Chapter 3 An Investigation of the Chemical Mechanism of LgtC 96_ the reaction mixture. However, substantial amounts (> 60 % by NMR) of the UDPGal were re-isolated such that had any scrambling of the l s O isotope occurred, it would have been detected. Given the number of conditions that must be met in order for any PIX to occur, the absence of such an observation in LgtC is therefore not entirely conclusive that an intermediate is not formed. 3.5 An Alternative to the Double Displacement Mechanism -the SNi Mechanism The reaction catalyzed by LgtC is known to proceed with net retention of anomeric configuration. While such a stereochemical outcome is most often the consequence of two stepwise displacement reactions involving an intermediate, all of our efforts to date have failed to provide any concrete evidence for the existence of such a species. Attempts to trap a covalent glycosyl-enzyme intermediate using a variety of different strategies have been unsuccessful and the possible involvement of nonenzymic intermediates was also found to be unlikely. In addition to these stable, isolatable intermediates, the potential existence of relatively long-lived but unstable intermediates was also investigated through the use of positional isotope exchange studies. Unfortunately, that too turned out to be a fruitless endeavour, as none could be detected. Although reasonable explanations were proposed in each case to rationalize the negative outcomes, the prospect that there simply may not be a detectable intermediate and that the reaction of LgtC does not proceed through a double displacement mechanism is also a distinct possibility that must also be addressed. Chapter 3 An Investigation of the Chemical Mechanism of LgtC £7 Although it is not a common occurrence, reactions that proceed through a single step displacement mechanism are also capable of yielding products with net retention of stereochemistry. In most cases, the observed stereochemical outcome is merely the consequence of neighbouring group participation, but examples where such an effect cannot take place have also been known to produce the same result. The most notable illustration of this latter phenomenon is the reaction of alcohols with thionyl chloride to yield alkyl halides of retained stereochemical configuration. This reaction was proposed by Hughes, Ingold and coworkers to proceed through an alkyl chlorosulfite intermediate that subsequently decomposed with net retention of stereochemistry through what they termed an SNi mechanism [136, 137]. Evidence in support of this hypothesis was later provide by Lewis and Boozer who showed that under certain conditions, the decomposition of alkyl chlorosulfite was indeed first ordered, leading to the formation of stereochemically retained alkyl halides [138]. While the involvement of a concerted, four-centre transition state in this reaction mechanism would be strictly forbidden according to the Woodward-Hoffmann rules [139], it is conceivable that the reaction can occur by way of a multistep ionization mechanism involving intimate ion pairs. Studies on the effects of solvents [140] and substituents [141] along with NMR-aided computational investigations [142] have certainly supported the involvement of ionic species in the SNI mechanism. The feature distinguishing an SNi mechanism from an SNI mechanism however is that the leaving group can undergo further decomposition at a rate that is faster than that at which a potential anion can react at the rear of the atom undergoing substitution. The resulting nucleophilic species that is formed as a Chapter 3 An Investigation of the Chemical Mechanism of LgtC 98 consequence of this decomposition can then attack the reactive centre from the front side, resulting in retention of stereochemistry in the product (Figure 3.12). Figure 3.12. The proposed ionization steps associated with the S]\ii mechanism for the decomposition of certain alkyl chlorosulfites to yield alkyl halides with net retention of stereochemistry. The significance of the SNI mechanism is its illustration that single step displacement reactions can also lead to net retention of stereochemistry at a reaction centre. Such reactions are however very rare because it requires that the nucleophile be more suitably positioned to attack from the front than from the rear of the atom undergoing substitution. In the case involving the unimolecular decomposition of alkyl chlorosulfites, this hard-to-meet requirement is achieved as a consequence of the further decomposition of the chlorosulfmyl leaving group to yield the chloride anion, which can approach the reaction centre from the front side (Figure 3.12). For bimolecular reactions, this front side approach of the nucleophile is more difficult to envision, but is nonetheless possible. In the solvolysis of a-glucosyl fluoride with trifluoroethanol, the predominant product of the reaction was found to be the retained trifluoroethyl a-glucoside [143]. This reaction was proposed to proceed through a highly dissociative transition state in which the stereochemical outcome is attributed to a favourable association of the trifluoroethanol with the 2-OH of a-glucosyl fluoride. This interaction of the two CI Chapter 3 An Investigation of the Chemical Mechanism of LgtC 99 reactants leads ultimately to the positioning of the alcohol to approach the C-l reaction centre from the same side as that from which the fluoride leaving group departed (Figure 3.13). From these examples, it can therefore be concluded that the most important feature of single step displacement reactions, which allows them to yield products of retained stereochemistry is the proper positioning of the nucleophile in relation to that of the centre undergoing substitution. Figure 3.13. Proposed transition state for the solvolysis of a-glucosyl fluoride by trifluoroethanol to yield a glucoside product of retained anomeric configuration [143]. In light of the available evidence that single step displacement reactions can yield products of retained stereochemistry in solution, it is thus not unreasonable to expect that such reactions may be even more prevalent within the confines of an enzyme active site. After all, the primary purpose of these biological catalysts is to provide a scaffold for the positioning of substrates for reaction. Contrary to this expectation however, the only report, of which we are aware thus far of an enzyme that might employ such a single step, SiMi-like mechanism is glycogen phosphorylase [144]. A difficulty associated with the proposal of such a mechanism though, is that it is extremely difficult to probe. As a result, this SNi-like mechanism through which glycogen phosphorylase has been Chapter 3 An Investigation of the Chemical Mechanism of LgtC 100 proposed to proceed is to date only speculative, as there is no real evidence to support this claim. In the reaction catalyzed by LgtC, a net retention of anomeric configuration is observed. Although this enzyme is therefore believed to follow a double displacement mechanism, evidence in support of this hypothesis has remained elusive. Alternatively, a single step, Swi-like mechanism similar to that depicted in Figure 3.14 is also possible for this enzyme. In this scenario, the nucleophilic hydroxyl of the acceptor lactose would have to be situated very close to the reactive CI" centre and on the same side of the galactose ring as the UDP moiety. In this position, it is conceivable that this reactive hydroxyl could also interact with the exocyclic O l " of UDPGal through a hydrogen bond. Upon cleavage of the CI "-01" bond, the departing UDP could then act as a general base to deprotonate the nucleophilic 4'-OH of lactose, thereby facilitating its front side attack of the reaction centre. As with the case for the decomposition of alkyl chlorosulfites, it is very plausible that such a reaction catalyzed by LgtC would also occur via intimate ion pair intermediates. From the standpoint of minimal participation by the protein, such a mechanism would be the epitome, as the nucleophile, general acid catalyst and general base catalyst are all self-contained within the substrates. In essence, the role of LgtC would therefore simply be to serve as a scaffold to properly orient, align and if appropriate distort the two substrates for reaction. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 101 Figure 3.14. A depiction of a speculative S^- l ike mechanism for the LgtC-catalyzed reaction. The idea that the role of LgtC is simply to position the substrates for reaction is certainly appealing in that it would also help to explain the residual activities that are observed in the mutants of Glnl89 (Table 3.2). While mutations at this position were shown to result in significant loss of activity, the amount of activity that was retained was disconcertingly high for this residue to function as the essential catalytic nucleophile in a double displacement mechanism. Furthermore, the activity that was retained in the various mutants (Gin 189Ala, Glnl89Asn, Glnl89Glu) also appeared to be independent of the substitutions made. These observations are therefore more consistent with a more passive role for Gin 189 in which it functions simply to hold the donor UDPGal substrate in place for reaction with the acceptor substrate lactose. As such, any mutations of this residue would lead to a misalignment of the donor substrate, resulting in the same loss in catalytic efficiency. At this time, the only indication that LgtC may employ an SNi-like mechanism is from our most recent three-dimensional structure of this enzyme, which was co-crystallized with UDP-2FGal but with lactose soaked into the crystals. In comparison to the structure obtained with UDP-2FGal and the incompetent acceptor, 4dLac (see Section 3.2), there was very little difference between the two ternary complexes. From this latest Chapter 3 An Investigation of the Chemical Mechanism of LgtC 102 structure however, the position of the nucleophilic 4-OH in the lactose acceptor is clearly defined and is shown to be situated on the same side of the galactose ring as the UDP moiety (Figure 3.15). This nucleophile is indeed very close to the reactive C I " centre (2.9 A) and also appears to form a strong hydrogen bond (2.6 A) to the exocyclic 01" atom. While this observation is by no means evidence for an SNi-like mechanism, it certainly does lend a great deal of appeal to the idea. Figure 3 . 1 5 . A view of the two substrates in the three-dimensional structure of LgtC in complex with UDP-2FGal and lactose. The distance (in angstroms) from the nucleophilic 4 ' - 0 H of lactose to the reactive C I " centre and the exocyclic 0 1 " are also indicated. 3.6 Concluding Remarks On the basis of similarities in the reactions catalyzed, retaining glycosyltransferases are believed to follow a double displacement mechanism similar to that employed by the well-studied retaining glycosidases. The key feature of this Chapter 3 An Investigation of the Chemical Mechanism of LgtC 103 mechanism is the formation and subsequent breakdown of a putative covalent glycosyl-enzyme intermediate. While a great deal of effort was spent on our part in trying to demonstrate the existence of such an intermediate in LgtC, none of the strategies employed to trap this species and identify the catalytic nucleophile was successful. Even with the solution of the three-dimensional structure of this enzyme, our understanding of the chemical mechanism of LgtC continues to remain clouded, as the only identifiable enzymic nucleophile turned out to be an amide from the side chain of a glutamine residue. Subsequent mutagenesis studies revealed that although this residue is important to catalysis, the exact nature of its role remains unclear, as the amount of residual activity in the mutant enzymes was disturbingly high for this residue to function in such a critical capacity as the catalytic nucleophile. Although the reaction of LgtC does not appear to involve a covalent glycosyl-enzyme intermediate, a claim about the existence of such an intermediate has been made for a structurally unrelated retaining a-l,3-galactosyltransferase based on the recently solved three-dimensional structure of that enzyme in complex with the donor substrate UDPGal [70]. From the structure, electron density extending from Glu317 of the enzyme was interpreted by the authors as being that of a galactose moiety covalently bound to the carboxylate side chain of this residue. However, in the absence of a higher resolution structure and a more defined electron density map, it is difficult to ascertain whether or not the galactose residue is actually bonded covalently to Glu317. In light of the possibility that this a-l,3-galactosyltransferase may also possess some hydrolytic activity as has been demonstrated in several other glycosyltransferases, including LgtC (see Section 2.1.3), this observed electron density may simply be due to a galactose residue Chapter 3 An Investigation of the Chemical Mechanism of LgtC 104 which has remained bound to the enzyme after its hydrolysis from UDPGal. Moreover, the importance of Glu317 to catalysis in this a-l,3-galactosyltransferase has also not been confirmed, as mutagenesis data for this particular residue were not presented. A recent determination of the structure of the same a-l,3-galactosyltransferase, but at a much better resolution [72] has also cast further doubts about the conclusions of the work of Gastinel et. al. concerning a covalent glycosyl-enzyme intermediate in that enzyme [70]. In this latest structure of the a-l,3-galactosyltransferase, the C-terminal region, which was previously observed to be disordered in the structure of Gastinel et. al. [70], is now clearly defined [72]. This ordering of the loop region was shown to be an important part of catalysis, as it appears to cause the enzyme to adopt an active conformation. In light of this revelation, the structure solved by Gastinel et. al. is most likely not a good representation of the enzyme in its active conformation. As such, any mechanistic interpretations based on that structure are most likely invalid. As an alternative to a covalent glycosyl-enzyme intermediate, the possibility that the proposed double displacement mechanism of LgtC proceeds through a nonenzymic intermediate was also explored. With the aid of the three-dimensional structure, candidate nucleophiles that would be capable of carrying out the first displacement reaction were identified. The putative intermediates that would be formed after this first displacement reaction were then chemically synthesized and evaluated as substrates for LgtC. In each of the cases studied, no turnover of the externally added, putative intermediates was observed. While these results do not completely eliminate the possibility that such intermediates are involved in the proposed double displacement mechanism of LgtC, they certainly suggest that this is very unlikely to be the case. Chapter 3 An Investigation of the Chemical Mechanism of LgtC 105 Having exhausted the search for potential nucleophiles to carry out the first step of the double displacement mechanism, it is quite conceivable that LgtC may not utilize such a mechanism. The net retention of stereochemistry that is observed in the reaction of this enzyme may in fact be achieved through other means. A possibility, but one that is not well precedented, is that the mechanism of LgtC is SNi-like. In this mechanism, the direct approach of the nucleophilic hydroxyl of the acceptor from the same side of the reaction centre as the departing UDP leaving group would also result in a net retention of stereochemistry in the product, but without the need of an intervening nucleophile. At this point in time however, no experimental evidence is available to support such a hypothesis. As such, the mechanism utilized by LgtC to catalyze its glycosyl transfer reaction with net retention of stereochemistry in the product remains a mystery. CHAPTER 4 MATERIALS AND METHODS Chapter 4 Materials and Methods 107 4 MATERIALS AND METHODS 4.1 General All buffer chemicals and other reagents were obtained from Sigma/Aldrich Chemical Co. unless otherwise noted. All 'H NMR spectra were recorded at either 200, 300 or 400 MHz using a Bruker AC-200, AV-300 or WH-400 spectrometer. All 1 9 F NMR spectra were recorded at 188 MHz using the Bruker AC-200 spectrometer with CF3CO2H as the reference and all 3 1 P NMR spectra were recorded at either 81 or 121 MHz using a Bruker AC-200 with dimethylphosphate as the reference or a Bruker AV-300 with phosphoric acid as the reference. Microanalyses were performed in house by Mr. Peter Borda of the Department of Chemistry at the University of British Columbia. Recombinant LgtC was cloned and expressed by Dr. Warren W. Wakarchukover of the National Research Council of Canada and was provided to us as a cell lysate suspended in 6% polyethylene glycol. Figures presented as double reciprocal plots are for pattern recognition only as all numerical data values were determined by direct fit of the initial rate data into the appropriate equations using the computer program Grafit 3.0. 4.2 Synthesis 4.2.1 Lactoside acceptor substrates (Scheme 2.1) 1,2,2',3,3',4',6,6'-Octa-O-acetyl-p-D-lactose (2.1) Sodium acetate (10 g, 121.9 mmol) was dissolved in acetic anhydride (90 mL) and the mixture was brought to a boil. The heat source was then removed and lactose (20 g, 55.5 mmol) was then added in small amounts so as to maintain the reaction at boiling. After Chapter 4 Materials and Methods 108 1.5 h, the addition of lactose was completed. Once the reaction had subsided, the solution was again brought to a boil before allowing it to cool. Chilled water (400 mL) was then added and the reaction was stirred overnight. Methylene chloride (200 mL) was added to dissolve the precipitate and the layers were separated. The aqueous layer was extracted with additional CH2CI2 (2 x 200 mL) and the combined organic layers were then washed with water (250 mL), aq. NaHC03 (2 x 200 mL) and brine (250 mL), dried over MgS04 and the solvent was evaporated in vacuo. Crystallization from EtOAc/Hex yielded (2.1) (28.5 g, 76%) as a white crystalline solid. ] H NMR (CDC13, 400 MHz): 5 5.65 (d, 1 H, Ji,2 8.2 Hz, H-l), 5.31 (dd, 1 H, J 4 , 3 3.3, J 4-, 5 0.8 Hz, H-4'), 5.20 (dd, 1 H, J 3 , 2 9.2, J 3 , 4 9.1 Hz, H-3), 5.06 (dd, 1 H, J 2 j 3 10.4, J 2 ,r 7.9 Hz, H-2'), 5.00 (dd, 1 H, J 2 , 3 9.2, J2 ji 8.3 Hz, H-2), 4.91 (dd, 1 H, J 3 , 2 10.4, J 3 - > 4 . 3.3 Hz, H-3'), 4.44 (d, 1 H, J r > 2 - 7.9 Hz, H-l'), 4.41 (dd, 1 H, J6a,6b 12.1, J 6 a , 5 1.9 Hz, H-6a), 4.00 - 4.13 (m, 3 H, H6b, H6a', H6b'), 3.83 (ddd, 1 H, J 5 , 6 b 7.1, J 5 , 6 a 6.4, J 5 , 4 0.8 Hz, H-5'), 3.80 (dd, 1 H, J 4 , 5 9.8, J 4 , 3 9.1 Hz, H-4), 3.72 (ddd, 1 H, J 5 , 4 9.8, J5>6b 4.7, J 5 ) 6 a 1.9 Hz, H-5), 2.11, 2.08, 2.06, 2.03, 2.01, 2.00, 1.99 (s, 24 H, 8 x OAc). 2,2',3,3',4',6,6'-Hepta-O-acetyl-cc-D-lactosyl bromide (2.2) To a 0 °C solution of (2.1) (4.7 g, 6.87 mmol) in anhydrous CH2C12 (15 mL) under nitrogen was added 45% HBr/AcOH (5.3 mL). The reaction vessel was then sealed and the solution was allowed to stir at room temperature. After 2.5 h, the reaction mixture was poured into chilled water (80 mL) and diluted with CH2CI2 (60 mL). Solid NaHC0 3 was added to neutralize the excess acid and the layers were separated. The aqueous layer was further extracted with CH2CI2 (2 x 80 mL) and the combined organic extracts were Chapter 4 Materials and Methods 109 washed with water (2 x 80 mL), aq. NaHC03 (50 mL) and brine (50 mL). Removal of the solvent under reduced pressure after drying over MgS0 4 yielded (2.2) (4.5 g, 93%) as a white brittle solid. *H NMR (CDC13, 400 MHz): 5 6.5 (d, 1 H, J,, 2 4.0 Hz, H-l), 5.53 (dd, 1 H, J 3 , 2 9.6, J 3 , 4 9.6 Hz, H-3), 5.33 (dd, 1 H, J 4 > 3 3.4, J 4 j 5 0.9 Hz, H-4), 5.11 (dd, 1 H, h;y 10.4, J 2 - , 7.9 Hz, H-2'), 4.94 (dd, 1 H, J 3 > 2 10.4, J 3 ; 4 3.4 Hz, H-3'), 4.74 (dd, 1 H, J 2 > 3 9.6, J 2 j l 4.0 Hz, H-2), 4.49 (d, 1 H, J, > 2 7.9 Hz, H-l'), 4.47 (dd, 1 H, J 6 a,6b 12.0, J 6 a , 5 1.7 Hz, H-6a), 4.02 - 4.23 (m, 4 H, H-5, H-6b, H-6a', H-6b'), 3.86 (ddd, 1 H, J 5 , 6 b 7.3, J 5,6a' 6.4, J 5 , 4 0.9 Hz, H-5'), 3.83 (dd, 1 H, J 4 j 5 9.8, J 4 , 3 9.6 Hz, H-4), 2.13, 2.10, 2.06, 2.04, 2.03, 2.02, 1.94 (s, 21 H, 7 x OAc). Benzyl 2,2',3,3',4',6,6'-hepta-0-acetyl-j5-D-lactoside (2.3) (2.2) (2.6 g, 3.69 mmol) was stirred in anhydrous CH2C12 (25 mL) containing 4 A sieves under an atmosphere of nitrogen when benzyl alcohol (1.9 mL, 18.47 mmol) and AgC0 3 (2.0 g, 7.39 mmol) were added along with a crystal of iodine. The reaction was covered and stirred at room temperature overnight before it was filtered through Celite®. The filtrate was evaporated in vacuo and the residue was chromatographed over silica gel (PE:EtOAc, 3:2 to 1:1). Crystallization from PE/EtOAc yielded (2.3) (2.0 g, 75%) as a white solid. ! H NMR (CDC13, 400 MHz): 6 7.22 - 7.35 (m, 5H, Ar), 5.32 (dd, 1 H, J 4 > 3 3.3, J 4 ; 5 0.8 Hz, H-4'), 5.16 (dd, 1 H, J 3 , 2 9.3, J 3 , 4 9.2 Hz, H-3), 5.08 (dd, 1 H, J 2 , 3 10.4, J 2 ; , 7.9 Hz, H-2'), 4.99 (dd, 1 H, J 2 > 3 9.3, J2,, 7.9 Hz, H-2), 4.93 (dd, 1 H, J 3 ; 2 10.4, J 3 , 4 3.3 Hz, H-3'), 4.86 (d, 1 H, J 12.3 Hz, PhCH), 4.58 (d, 1 H, J 12.3 Hz, PhCH), 4.50 (d, 1 H, Ji ,2 7.9 Hz, H-l'), 4.48 (d, 1 H, J,, 2 7.9 Hz, H-l), 4.48 - 4.52 (m, 1 H, H-6a), 4.00 -4.15 (m, 3 H, H-6b, H-6a', H-6b'), 3.84 (ddd, 1 H, J 5 ; 6 b 7.4, J 5 > 6 a 6.3, J 5 , 4 - 0.8 Hz, h-5'), 3.79 (dd, 1 H, J 4 , 5 9.6, J 4 , 3 9.2 Hz, H-4), 3.56 (ddd, 1 H, J 5 > 4 9.6, J5,6b 5.0, J 5 , 6 a 2.0 Hz, H-Chapter 4 Materials and Methods 110 5), 2.12, 2.11, 2.02, 2.01, 2.00, 1.98, 1.94 (s, 21 H, 7 x OAc). Anal. Calcd for C33H42O18: C, 54.54; H, 5.83. Found: C, 54.50; H, 5.93. 4-Pentenyl 2,2',3,3',4', 6,6 '-hepta-0-acetyl-f3-D-lactoside (2.4) (2.2) (0.5 g, 0.72 mmol) was stirred in anhydrous CH2CI2 (5 mL) containing 4 A sieves under an atmosphere of nitrogen when 4-penten-l-ol (0.37 mL, 3.60 mmol) and AgC0 3 (0.3 g, 1.08 mmol) were added along with a crystal of iodine. The reaction was covered and stirred at room temperature for 13.5 h before it was filtered through Celite®. The filtrate was evaporated in vacuo and the residue was chromatographed over silica gel (PE:EtOAc, 3:2 to 1:1) to yield (2.4) (0.21 g, 41%) as a colourless gum. ] H NMR ( C D C I 3 , 400 MHz): 5 5.75 (m, 1 H, CH=CH2), 5.32 (dd, 1 H, J 4 , 3 3.4, j 4 , 5 0.8 Hz, H-4'), 5.17 (dd, 1 H, J3,2 9.4, J 3 , 4 9.2 Hz, H-3), 5.08 (dd, 1 H, J 2 > 3 10.4, J 2 , , 7.9 Hz, H-2'), 4.98 (ddd, 1 H, J t r a n s 17.1, J 3.4, J g e m 1.7 Hz, CH=CHtrariS), 4.93 (dd, 1 H, J r , 2 10.4, J 3 , 4 3.4 Hz, H-3'), 4.92 - 4.96 (m, 1 H, CH=CHcis), 4.87 (dd, 1 H, J 2 , 3 9.4, J2,i 8.0 Hz, H-2), 4.46 (d, 1 H, J, ; 2 7.9 Hz, H-l'), 4.43 (d, 1 H, J,, 2 8.0 Hz, H-l), 4.42 - 4.46 (m, 1 H, H-6a), 4.02 - 4.14 (m, 3 H, H-6b, H-6a', H-6b'), 3.87 (dd, 1 H, J 4 , 5 9.7, J 4 > 3 9.2 Hz, H-4), 3.84 (m, 1 H, H-5'), 3.81 (dt, 1 H, J 9.8, J 6.2 Hz, OCH), 3.57 (ddd, 1 H, J 5 , 4 9.7, J5,6b 5.1, J 5 , 6 a 2.0 Hz, H-5), 3.45 (dt, 1 H, J 9.8, J6.7 Hz, OCH), 2.14,2.12, 2.09, 2.03, 2.02, 2.01, 1.94 (s, 21 H, 7 x OAc), 1.52 - 1.70 (m, 4 H, CH 2CH 2). Allyl2,2',3,3',4',6,6'-hepta-0-acetyl-/3-D-lactoside (2.5) (2.2) (0.5 g, 0.69 mmol) was stirred in anhydrous CH2CI2 (5 mL) containing 4 A sieves under an atmosphere of nitrogen when allyl alcohol (0.24 mL, 3.46 mmol) and AgC03 Chapter 4 Materials and Methods 111 (0.4 g, 1.38 mmol) were added along with a crystal of iodine. The reaction was covered and stirred at room temperature for 9 h before it was filtered through Celite®. The filtrate was evaporated in vacuo and the residue was chromatographed over silica gel (PE:EtOAc, 3:2 to 1:1) to yield (2.5) (0.35 g, 76%) as a white solid. ! H NMR (CDC13, 400 MHz): 5 5.81 (m, 1 H, OCH2CH=CH2), 5.32 (dd, 1 H, J 4 , 3 3.4, J 4 , 5 0.9 Hz, H-4'), 5.23 (ddd, 1 H, J t r a n s 17.3, J 3.2, J g e m 1.6 Hz, OCH2CH=CH t r a n s), 5.17 (dd, 1 H, J 3 , 2 9.3, J 3 , 4 9.2 Hz, H-3), 5.17 (ddd, 1 H, J c i s 10.5, J 2.8, J g e m 1.6 Hz, OCH2CH=CHc l s), 5.08 (dd, 1 H, J 2 , 3 10.4, J 2 , i 7.9 Hz, H-2'), 4.93 (dd, 1 H, J 3 , 2 10.4, J 3 > 4 3.4 Hz, H-3'), 4.90 (dd, 1 H, J2,3 9.3, J 2 , i 7.9 Hz, H-2), 4.50 (d, 1 H, J, j 2 7.9 Hz, H-l'), 4.46 (d, 1 H, J,, 2 7.9 Hz, H-l), 4.44-4.50 (m, 1 H, H-6a), 4.27 (ddt, 1 H, J 13.2, J 4.9, J 1.5 Hz, OCH), 4.01-4.14 (m, 4 H, H-6b, H-6a', H-6b', OCH), 3.84 (ddd, 1 H, J 5 , 6 b 7.2, J 5 , 6 a 6.4, J 5 > 4 0.9 Hz, H-5'), 3.78 (dd, 1 H, J 4 , 5 9.7, J 4 , 3 9.2 Hz, H-4), 3.57 (ddd, 1 H, J 5 , 4 9.7, J 5 j 6 b 5.0, J 5 > 6 a 2.0 Hz, H-5), 2.13, 2.10, 2.03, 2.01, 1.94 (s, 21 H, 7 x OAc). 2,3-Dihydroxypropyl 2,2 ',3,3 ',4',6,6'-hepta-0-acetyl-fl-D-lactoside (2.6) N-methylmorpholino N-oxide (0.02 g, 0.18 mmol) was dissolved in a solution of 4:1 acetone:water (1.5 mL) under an atmosphere of nitrogen at 0 °C when a catalytic amount of osmium tetroxide in t-butanol was added. To this was added a solution of (2.5) (0.11 g, 0.16 mmol) in acetone (0.5 mL) and the reaction was stirred overnight. Sodium bisulfite (0.06 g, 0.55 mmol) in water (1 mL) was then added to the reaction and stirring was continued for 1 h. The reaction was poured into brine (10 mL) and extracted with CH2CI2 (2 x 10 mL). Evaporation of the combined organic layers after drying over MgS04 yielded (2.6) (0.11 g, 96%) as a white solid. *H NMR (CDC13, 400 MHz) Chapter 4 Materials and Methods 112 selected data only: 5 5.35 (d, 1 H, J 4 , 3 3.3 Hz, H-4'), 5.18 (dd, 1 H, J 3 , 2 9.5, J 3 , 4 9.1 Hz, H-3), 5.08 (dd, 1 H, J2-,3 10.4, J 2,i 7.9 Hz, H-2'), 4.94 (dd, 1 H, J 3 , 2 10.4, J 3 j 4 3.3 Hz, H-3'), 4.87 (dd, 1 H, J 2 , 3 9.5, J 2 ) , 8.0 Hz, H-2), 4.53 (ddd, 1 H, J 12.1, J 5.3, J 2.1 Hz, OCH), 4.44 (d, 1 H, J r , 2 7.9 Hz, H-l'), 4.42 (d, 1 H, J,, 2 8.0 Hz, H-l), 2.17, 2.13, 2.05, 2.03, 2.02, 1.94 (s, 21 H, 7 x OAc). Benzyl fi-D-lactoside (2.7) To a solution of (2.3) (1.24 g, 1.71 mmol) in anhydrous MeOH (30 mL) under nitrogen was added a catalytic amount of sodium methoxide until the pH of the solution was around 10. The reaction was then stirred overnight at room temperature before it was neutralized with acidic Amberlyte® resin. After evaporation of the solvent, crystallization of the resulting residue from MeOH/EtOAc yielded (2.7) (0.64 g, 87%) as a white solid. 'H NMR (D20, 400 MHz) selected data only: 5 7.30 - 7.60 (m, 5H, Ar), 4.90 (d, 1 H, J 11.4 Hz, PhCH), 4.52 (d, 1 H, J,->2- 8.0 Hz, H-l'), 4.41 (d, 1 H, J 1 > 2 7.8 Hz, H-l), 3.95 (dd, 1 H, J 6 a , 6 b 12.3, J 6 a , 5 2.1 Hz, H-6a), 3.88 (d, 1 H, J 4 , 3 3.3 Hz, H-4'), 3.32 (dd, 1 H, J 4 j 5 8.7, J 4 > 3 8.4 Hz, H-4). Anal. Calcd for Ci 9 H 2 8 0n: C, 52.77; H, 6.53. Found: C, 52.47; H, 6.63. 4-Pentenyl /3-D-lactoside (2.8) To a solution of (2.4) (0.45 g, 0.63 mmol) in anhydrous MeOH (20 mL) under nitrogen was added a catalytic amount of sodium methoxide until the pH of the solution was around 10. The reaction was then stirred overnight at room temperature before it was neutralized with acidic Amberlyte® resin. After the solvent was evaporated in vacuo, Chapter 4 Materials and Methods 113 chromatography of the resulting residue over silica gel (EtOAc:MeOH:H20, 15:4:1) yielded (2.8) (0.17 g, 67%) as a white powder. ] H NMR (D20, 400 MHz) selected data only: 5 5.89 (m, 1 H, CH=CH2), 5.06 (dd, 1 H, J ^ s 17.3, }gem 1.0 Hz, CH=CH trans), 5.00 (dd, 1 H, JC I S 9.3, J g e m 1.0 Hz, CH=CHcis), 4.45 (d, 1 H, J, ,2 7.8 Hz, H-l'), 4.42 (d, 1 H, J 1 > 2 7.6 Hz, H-l), 2.07 - 2.17 (m, 2H, CH2-CH=CH2), 1.65 - 1.75 (m, 2H, OCH 2-CH 2). Anal. Calcd for CyH.oOnV^O: C, 48.68; H, 7.45. Found: C, 48.98; H, 7.14. Allyl p-D-lactoside (2.9) (2.5) (0.14 g, 0.21 mmol) was dissolved in anhydrous MeOH at 0 °C under an atmosphere of nitrogen when gaseous ammonia was bubbled into the solution. After 5 min, both the ammonia source and ice bath were removed and the reaction was stirred overnight at room temperature. Evaporation of the solvent in vacuo followed by chromatography over silica gel (EtOAc:MeOH:H20, 15:4:1) yielded (2.9) (47.5 mg, 60%) as a white solid. ] H NMR (D20, 400 MHz) selected data only: 8 5.95 (m, 1 H, OCH2CH=CH2), 5.35 (dd, 1 H, J , r a n s 17.3, J g e m 1.4 Hz, CH=CHtrariS), 5.26 (d, 1 H, JC 1 S 8.4 Hz, CH=CHcis), 4.50 (d, 1 H, J, > 2 8.0 Hz, H-l'), 4.42 (d, 1 H, Ji,2 7.7 Hz, H-l), 4.37 (m, 1 H, OCH), 4.20 (m, 1 H, OCH), 3.95 (dd, 1 H, J 6 a , 6 b 12.2, J 6 a > 5 1.7 Hz, H-6a), 3.89 (d, 1 H, J 4-> 3. 3.1 Hz, H-4'), 3.52 (dd, 1 H, J 4 j 5 9.8, J 4 , 3 7.8 Hz, H-4). Anal. Calcd for Ci5H260ir!/2H20: C, 46.03; H, 6.95. Found: C, 46.52; H, 6.85. 2,3-Dihydroxypropyl /3-D-lactoside (2.10) To a solution of (2.6) (0.10 g, 0.1.5 mmol) in anhydrous MeOH (20 mL) under nitrogen was added a catalytic amount of sodium methoxide until the pH of the solution was Chapter 4 Materials and Methods 114 around 10. The reaction was then stirred overnight at room temperature before it was neutralized with acidic Amberlyte® resin. After evaporation of the solvent, crystallization of the resulting residue from MeOH yielded (2.10) (34.5 mg, 57%) as a white solid. ! H NMR (D20, 400 MHz) selected data only: 5 4.47 (d, 1 H, J i > 2 8.0 Hz, H-1), 4.42 (d, 1 H, J, ; 2 8.0 Hz, H-l'). Anal. Calcd for C, 5 H 2 gOi3: C, 43.27; H, 6.78. Found: C, 43.30; H, 6.92. 4.2.2 a-Galactosyl fluoride (Scheme 2.2) 1,2,3,4,6-Penta- O-acetyl-fi-D-galactopyranose (2.11) After a solution of sodium acetate (25 g, 310 mmol) and acetic anhydride (350 mL, 3710 mmol) was brought to a boil, the heat source was removed and galactose monohydrate (50 g, 250 mmol) was slowly added so as to maintain the reaction at boiling. Three hours after the complete addition of galactose, the reaction mixture was reheated to boiling and then allowed to cool to room temperature. At this time, the reaction mixture was poured into ice water (800 mL) and then allowed to stir at room temperature overnight. The desired compound was then precipitated from the solution at 0 °C to yield (2.11) as an off-white solid (35.7 g, 43%). *H NMR (CDC13, 400 MHz): 5 5.70 (d, 1 H, J 1 > 2 8.3 Hz, H-l), 5.40 (dd, 1 H, J 4 , 3 3.3, J 4 , 5 0.8 Hz, H-4), 5.31 (dd, 1 H, J 2 j 3 10.4, J2,i 8.3, H-2), 5.10 (dd, 1 H, J 3 > 2 10.4, J 3 > 4 3.3, H-3), 4.00 - 4.18 (m, 3 H, H-5, H-6a, H-6b), 1.99, 2.05, 2.12, 2.17 (s, 15 H, 5 x OAc). 2,3,4,6-Tetra-O-acetyl-a-D-galactopyranosylfluoride (2.12) A solution of (2.11) (12 g, 30.7 mmol) in HFpyridine (48 mL) was stirred at 0 °C under an atmosphere of argon in a polypropylene bottle for 2 h. After this time, the reaction Chapter 4 Materials and Methods 115 mixture was allowed to warm to room temperature and stirring was continued for another 4 h. When judged to be complete by TLC, the reaction mixture was poured into aq. NaHC03 (300 mL) and extracted with CH2C12 (3 x 250 mL). The combined organic extracts were then washed with aq. NaHC03 (3 x 250 mL), water (250 mL) and brine (250 mL). After the solvent was dried over MgS04 and removed under reduced pressure, the residue was chromatographed over silica gel (PE:EtOAc, 3:1) to yield (2.12) (8.6 g, 80%) as a colourless gum. ! H NMR (CDC13, 400 MHz): 5 5.77 (dd, 1 H, J , ) F 53.3, Ji,2 2.7 Hz, H-l), 5.50 (dd, 1 H, J 4 ; 3 3.1, J 4 j 5 1.1 Hz, H-4), 5.34 (dd, 1 H, J 3 > 2 10.9, J 3 ; 4 3.1 Hz, H-3), 5.16 (ddd, 1 H, J 2 ; F 23.7, J 2 ( 3 10.9, J2,, 2.7 Hz, H-2), 4.38 (ddd, 1 H, J5,6a 7.0, J5,6b 6.6, J 5 , 4 1.1 Hz, H-5), 4.05-4.17 (m, 2 H, H-6a, H-6b), 2.02, 2.05, 2.12, 2.14 (s, 12 H, 4 x OAc). 1 9 F NMR (CDC13, 188 MHz): 5 -75.0 (dd, JF,i 53.3, J F , 2 23.7 Hz). a-D-Galactopyranosylfluoride (2.13) (2.12) (2.7 g, 7.71 mmol) was dissolved in anhydrous MeOH (45 mL) and the solution was placed under an atmosphere of argon. A catalytic amount of sodium methoxide was added and the reaction mixture was stirred at room temperature for 3 h. After neutralizing with acidic Amberlyte® resin, the solvent was then evaporated in vacuo. The resulting residue was crystallized from MeOH/Et20 to yield (2.13) (1.2 g, 83%) as a white crystalline solid. ] H NMR (D20, 400 MHz) selected data only: 5 5.56 (dd, 1 H, Ji ( F 55.4, Ji s 2 2.6 Hz, H-l), 3.80 (ddd, 1 H, J 2 , F 30.1, J2,3 10.1, J 2,, 2.6 Hz, H-2). 1 9 FNMR (D20, 188 MHz): 5 -76.3 (dd, JF,i 55.4, J F > 2 30.1 Hz). Anal. Calcd for C 6H,,F0 5: C, 39.56; H, 6.09. Found: C, 39.36; H, 6.23. Chapter 4 Materials and Methods 116 4.2.3 Incompetent acceptor substrates (Schemes 3.1 and 3.2) 4.2.3.1 4 -Deoxylactose 1,2,2 ',3,3', 6-Hexa-0-acetyl-4', 6 '-O-benzylidene-cc-lactose (3.1) To a suspension of lactose (45 g, 124.90 mmol) in DMF (110 mL) was added benzaldehyde dimethyl acetal (20.6 mL, 137.40 mmol) followed by a catalytic amount of p-toluenesulphonic acid (0.47 g, 2.5 mmol). The reaction was then stirred under reduced pressure (20 mm Hg) for 4 d at 60 °C. After this time, water (200 mL) was added to the reaction and unreacted benzaldehyde dimethyl acetal was extracted with EtOAc (2 x 200 mL). The aqueous layer was evaporated in vacuo following which, pyridine (200 mL) and acetic anhydride (100 mL) were then added to the resulting residue. The reaction was allowed to stir overnight before the volume was decreased by evaporation under reduced pressure. To the remaining residue was added ice water (300 mL) and the crude product was then extracted with CH2O2 (2 x 250 mL). The combined organic layers were washed with 10% v/v HC1 (3 x 200 mL), water (2 x 200 mL) and brine (1 x 250 mL), dried over M g S 0 4 and the solvent was evaporated under reduced pressure. Crystallization from EtOAc/Hexane yielded (3.1) as a slightly yellowish solid (8 g, 10%). 'H NMR (CDCI3, 200 MHz): 5 7.30 - 7.55 (m, 5 H, Ar), 6.45 (d, 1 H, J 1 > 2 3.7 Hz, H-l), 5.48 (s, 1 H, PhCH), 5.45 (dd, 1 H, J 3 , 2 10.4, J 3 , 4 9.9 Hz, H-3), 5.35 (dd, 1 H, J 2 j 3 10.3, h;v 7.9 Hz, H-2'), 5.05 (dd, 1 H, J 2 , 3 10.4, J 2 , i 3.7 Hz, H-2), 4.87 (dd, 1 H, J3,2 10.3, Jr,4-3.5 Hz, H-3'), 4.45 (d, 1 H, J, ,2 7.9 Hz, H-l'), 4.30-4.55 (m, 1 H, H-5), 4.15-4.45 (m, 2 H, H-6a, H-6b), 4.13 (dd, 1 H, J6a,6b 8.2, J 6 a , 5 - 4.1 Hz, H-6a'), 4.08 (dd, 1 H, J4-,3- 3.5, J 4 , 5 1.6 Hz, H-4'), 3.95 - 4.05 (m, 2 H, H-5', H-6b'), 3.80 (dd, 1 H, J 4 > 3 9.9, J 4 > 5 9.3 Hz, H-4), 2.17, 2.11, 2.15, 2.14, 2.00 (s, 18 H, 6 x OAc). Chapter 4 Materials and Methods 117 1,2,2 ',3,3 ',6-Hexa-0-acetyl-6'-O-benzyl-a-lactose (3.2) (3.1) (3.5 g, 5.13 mmol) was dissolved in anhydrous THF (120 mL) when sodium cyanoborohydride (3.2 g, 51.27 mmol) was added. A saturated solution of HC1 in diethyl ether was then cannulated into the reaction mixture in portions until the evolution of gas had ceased. Within 0.5 h, the reaction was judged to be complete by TLC (PE:EtOAc, 1:1). At this time, the reaction mixture was added to water (100 mL) and the crude product was extracted with CH2CI2 (2 x 100 mL). The combined organic layers were washed with aq NaHC03 (2 x 100 mL) and water (100 mL), dried over MgS0 4 and evaporated in vacuo. Chromatography over silica gel (PE:EtOAc, 4:5) yielded (3.2) as a white solid (2.66 g, 76%). 'H NMR (CDCI3, 200 MHz) selected data only : 5 7.30 -7.45 (m, 5 H, Ar), 6.25 (d, 1. H, J,, 2 3.7 Hz, H-l), 5.43 (dd, 1 H, J 3 , 2 10.2, J 3 , 4 9.1 Hz, H-3), 5.19 (dd, 1 H, Jr,r 10.1, J2,1 7.8 Hz, H-2'), 5.02 (dd, 1 H, J 2 > 3 10.2, J 2,, 3.7 Hz, H-2), 4.89 (dd, 1 H, J 3 , 2 10.1, J 3 > 4 3.1 Hz, H-3'), 4.52 (s, 2 H, PhCH2), 4.45 (d, 1 H, J r , 2 7.8 Hz, H-l'), 2.15, 2.10, 2.07, 2.04, 2.01, 2.00 (s, 18 H, 6 x OAc). 1,2,2 ',3,3',6-Hexa-0-acetyl-6'-0-benzyl-4'-deoxy-4'-iodo-a-cellobiose (3.3) A stirring solution of (3.2) (1.32 g, 1.93 mmol) in anhydrous CH2CI2 (15 mL) under argon was cooled to - 20 °C before pyridine (5 mL) and triflic anhydride (0.88 mL, 5.20 mmol) were added. The reaction was then warmed to room temperature and stirred for 2 h prior to the addition of aq NaHC03 (50 mL). The crude material was extracted with CH2C12 (2 x 35 mL) and the combined organic extracts were subsequently washed with water (2 x 80 mL). Coevaporation of the organic layer with MeCN under reduced Chapter 4 Materials and Methods 118 pressure yielded a yellow foamy gum which was then taken up in anhydrous DMF (25 mL). After the addition of Nal (1.44 g, 9.63 mmol), the reaction was allowed to stir at room temperature overnight under an atmosphere of argon. The reaction mixture was then added to aq NaHC0 3 (100 mL) and extracted with CH2C12 (3 x 125 mL). After the combined organic layers were washed with water (2 x 75 mL), dried over MgS04 and evaporated under reduced pressure, the resulting residue was crystallized from EtOAc/Hexane to yield (3.3) (0.94 g, 62%) as a cotton-like solid. 'H NMR (CDC13, 200 MHz) selected data only : 5 7.30 - 7.45 (m, 5 H, Ar), 6.24 (d, 1 H, J,, 2 3.7 Hz, H-l), 5.42 (dd, 1 H, J 3 , 2 10.3, J 3 , 4 9.3 Hz, H-3), 5.25 (dd, 1 H, J3,2 9.2, J 3 , 4 11.0 Hz, H-3'), 5.00 (dd, 1 H, J 2 ) 3 10.3, J2,i 3.7 Hz, H-2), 4.78 (dd, 1 H, J 2 j 3 9.2, J 2 8 . 1 Hz, H-2'), 4.50 (s, 2 H, PhCH2), 4.48 (d, 1 H, J r , 2 ' 8.1 Hz, H-l'), 2.12, 2.09, 2.06, 2.01, 2.00 (s, 18 H, 6 x OAc). 1,2,2r,3,3 ' ,6-Hexa-0-acetyl-6'-0-benzyl-4'-deoxy-a-lactose (3.4) To a solution of (3.3) (0.90 g, 1.13 mmol) in anhydrous benzene (30 mL) under argon was added tributyltin hydride (1.65 g, 5.67 mmol) and a catalytic amount of AIBN. The reaction was then refluxed for 45 min after which time the solvent was evaporated in vacuo. The residue was dissolved in MeCN (120 mL) and washed with hexane (3 x 70 mL). Evaporation of the MeCN layer followed by crystallization from EtOAc/Hexane yielded (3.4) (0.87 g, 85%) as a white fluffy solid. ! H NMR (CDC13, 400 MHz): 5 7.25 - 7.40 (m, 5 H, Ar), 6.24 (d, 1 H, J,, 2 3.7 Hz, H-l), 5.42 (dd, 1 H, J 3 j 2 10.1, J 3 , 4 9.4 Hz, H-3), 4.99 (dd, 1 H, J 2 , 3 10.1, J 2 > 1 3.7 Hz, H-2), 4.91 (ddd, 1 H, J 3 j 4 a x 11.3, J 3-, 2 9.6, J 3 , 4 e q 5.4 Hz, H-3'), 4.80 (dd, 1 H, J 2 ,3 9.6, h;v 7.8 Hz, H-l'), 4.50 (s, 2 H, PhCH2), 4.43 (dd, 1 H, J6a,6b 12.2, J 6 a , 5 2.0 Hz, H-6a), 4.36 (d, 1 H, J r > 2 - 7.8 Hz, H-l'), 4.12 (dd, 1 H, J 6 b, 6a Chapter 4 Materials and Methods 119 12.2, J 6 b , 5 4.3 Hz, H-6b), 3.97 (ddd, 1 H, J 5 ; 4 10.1, J5,6b 4.3, J 5 > 6 a 2.0 Hz, H-5), 3.78 (dd, 1 H, J 4 , 5 10.1, J 4 , 3 9.4 Hz, H-4), 3.60 - 3.67 (m, 1 H, H-5'), 3.56 (dd, 1 H, J 6 a-, 6b' 9.8, J 6 a > 5 5.3 Hz, H-6a'), 3.45 (dd, 1 H, J 6 b ' , 6 a ' 9.8, J 6 b', 5' 4.8 Hz, H-6b'), 2.14, 2.08, 2.01, 1.98 (s, 18 H, 6 x OAc), 2.00 - 2.10 (m, 1 H, H-4'eq), 1.50- 1.64 (m, 1 H, H-4'ax). I, 2,2',3,3',6-Hexa-0-acetyl-4'-deoxy-cc-lactose (3.5) (3.4) (0.23 g, 0.34 mmol) was dissolved in EtOH (4 mL) when cyclohexene (1.39 mL, 13.76 mmol) and 20% Pd(OH)2/C (0.06 g) were added and the reaction mixture was refluxed. After 2 h, the catalyst was removed by filtration through Celite® and the filtrate was evaporated in vacuo. Crystallization from EtOH yielded (3.5) (0.13 g, 63%) as a white fluffy solid. ] H NMR (CDC13, 400 MHz): 5 6.24 (d, 1 H, J 1 > 2 3.7 Hz, H-l), 5.45 (dd, 1 H, J 3 , 2 10.0, J 3 , 4 9.2 Hz, H-3), 5.01 (dd, 1 H, J 2 ) 3 10.0, J2,i 3.8 Hz, H-2), 4.94 (ddd, 1 H, J 3 , 4 a x 11.5, J 3 ) 2 9.5, J 3 ) 4 'e q 5.4 Hz, H-3'), 4.81 (dd, 1 H, J 2 - 3 9.5, h;v 7.7 Hz, H-2'), 4.47 (d, 1 H, J, ,2 7.7 Hz, H-l'), 4.44 (dd, 1 H, J 6 a > 6 b 12.5, J 6 a , 5 2.0 Hz, H-6a), 4.08 (dd, 1 H, J 6 b , 6 a 12.5, T6 b,5 4.5 Hz, H-6b), 4.00 (ddd, 1 H, J 5 > 4 10.0, J 5 ) 6 b 4.5, J 5 , 6 a 2.0 Hz, H-5), 3.83 (dd, 1 H, J 4 , 5 10.0, J 4 > 3 9.2 Hz, H-4), 3.52 - 3.68 (m, 3 H, H-5', H-6a', H-6b'), 2.15,2.08, 2.06,2.04, 1.98, 1.97 (s, 18 H, 6 x OAc), 2.00 - 2.10 (m, 1 H, H-4'eq), 1.50 -1.64 (m, 1 H, H-4'ax). 4'-Deoxylactose (3.6) To a stirring solution of (3.5) (0.50 g, 0.87 mmol) in anhydrous MeOH (8 mL) under argon was added a catalytic amount of sodium methoxide until the pH of the solution was slightly basic. The reaction was allowed to stir at room temperature overnight prior Chapter 4 Materials and Methods 120 to being neutralized with Amberlyte® IR-120 acidic resin. After filtration, evaporation of the solvent under reduced pressure yielded (3.6) (0.24 g, 85%) as a white solid. lH NMR (D20, 400 MHz) selected data only for (3 anomer : 5 4.63 (d, 1 H, J,, 2 8.0 Hz, H-1), 4.40 (d, 1 H, J, ,2 7.9 Hz, H-l'), 3.22 - 3.29 (m, 1 H, H-5), 3.19 (ddd, 1 H, J 5 , 4 a x 11.2, J 5,6a 8.0, J5-,6b- 3.2 Hz, H-5'), 1.95 (dd, 1 H, J ^ - a x 12.0, J 4 e q ; 3 4.4 Hz, H-4'eq), 1.42 (ddd, 1 H, J 4 ax ,4eq 12.0, J 4 a x , 3 11.9, J 4 ax.s 11.2 Hz, H-4'ax). Anal. Calcd for C12H22O10: C, 44.17; H, 6.80. Found: C, 43.87; H, 6.91. 4.2.3.2 Benzyl 4'-deoxy-4'-fluorolactoside 2,2',3,3',4',6,6'-Hepta-0-acetyl-cc-cellobiosyl bromide (3.7) A solution of octa-O-acetylcellobiose (10.95 g, 16.10 mmol) in anhydrous CH2CI2 (35 mL) under an atmosphere of nitrogen was cooled to 0 °C prior to the addition of 45% w/v HBr/AcOH (13 mL). The reaction vessel was then sealed and the reaction was allowed to warm to room temperature. After 4 h, the reaction was diluted with CH2CI2 (100 mL) and the solution was added to ice water (200 mL) whereupon it was neutralized with solid NaHCCb. Upon separation of the two phases, the aqueous layer was further extracted with CH2CI2 (2 x 100 mL) and the combined organic extracts were washed with aq. NaHC03 (2 x 75 mL), water (75 mL) and brine (75 mL). Removal of the solvent under reduced pressure after drying over MgS04 yielded (3.7) as a white solid (11.12 g, 99%). 'H NMR (CDCI3, 400 MHz): 5 6.50 (d, 1 H, J,, 2 4.1 Hz, H-l), 5.50 (dd, 1 H, J 3 , 2 9.7, J 3 , 4 9.7 Hz, H-3), 5.13 (dd, 1 H, J 3 , 4 9.3, J 3 , 2 9.2 Hz, H-3'), 5.05 (dd, 1 H, J 4 > 5 9.7, J 4 , 3 - 9.3 Hz, H-4'), 4.91 (dd, 1 H, J 6 a ,6b 12.5, J 6 a \ 5 4.5 Hz, H-6a'), 4.11 - 4.23 (m, 2 H, H-5, H-6b), 4.03 (dd, 1 H, J 6 b ) 6a 12.5, J 6 b j 5 2.3 Hz, H-6b'), 3.81 (dd, 1 H, J 4 ; 3 9.7, J 4 , 5 9.7 Chapter 4 Materials and Methods 121 Hz, H-4), 3.65 (ddd, 1 H, J 5 j 4 9.7, J 5 ; 6 a 4.5, J 5 , 6 b 2.3 Hz, H-5'), 2.11, 2.06, 2.01, 2.00, 1.98, 1.96 (s, 21 H, 7 x OAc). Benzyl 2,2 r,3,3",4',6,6' -hepta-O-acetyl-fi-cellobioside (3.8) To a solution of (3.7) (9.99 g, 14.29 mmol) in anhydrous CH 2 C1 2 (100 mL) containing 4 A molecular sieves under an atmosphere of nitrogen was added benzyl alcohol (7.39 mL, 71.44 mmol) and silver carbonate (7.88 g, 28.58 mmol). A crystal of iodine was added and the reaction vessel was shielded from the light and stirred at room temperature overnight. After this time, the reaction mixture was filtered through Celite® and washed with CH 2 C1 2 . The solvent was then evaporated in vacuo and the residue was crystallized from EtOAc/Hex to yield (3.8) (6.91 g, 66%) as a white fluffy solid. ] H N M R (CDC1 3, 400 MHz): 8 7.20 - 7.35 (m, 5 H, Ar), 5.11 (dd, 2 H, J 3 , 2 / 3 , 2 9.3, J 3 , 4 / r , 4 - 9.3 Hz, H-3, H-3'), 5.03 (dd, 1 H, J 4 , 5 9.8, J 4 , 3 9.3 Hz, H-4'), 4.94 (dd, 1 H, J 2 , 3 9.6, h;v 7.9 Hz, H-2), 4.89 (dd, 1 H, J 2 > 3 9.3, J 2,, 8.0 Hz, H-2), 4.83 (d, 1 H, J 12.3 Hz, PhCH). 4.56 (d, 1 H, J 12.3 Hz, PhCH), 4.51 (dd, 1 H, J6 a,6b 12.0, J 6 a , 5 2.0 Hz, H-6a), 4.49 (d, 1 H, J,, 2 8.0 Hz, H-1), 4.48 (d, 1 H, J r > 2 - 7.9 Hz, H- l ' ) , 4.33 (dd, 1 H , J 6 a , 6 b 12.5, J 6 a , 5 4.5 Hz, H-6a'), 4.08 (dd, 1 H, J 6 b ,6 a 12.0, J6 b,5 5.0 Hz, H-6b), 4.01 (dd, 1 H, J 6 b ; 6 a ' 12.5, J 6 b , 5 2.3 Hz, H-6b'), 3.77 (dd, 1 H, J 4 , 5 9.6, J 4 ; 3 9.3 Hz, H-4), 3.63 (ddd, 1 H, J 5 , 4 9.8, J 5 , 6 a 4.5, J 5,6b 2.3 Hz, H-5'), 3.54 (ddd, 1 H, J 5 , 4 9.6, J 5 > 6b 5.0, J 5 , 6 a 2.0 Hz, H-5), 2.12, 2.05, 2.00, 1.98, 1.97, 1.95 (s, 21 H, 7 x OAc). Chapter 4 Materials and Methods 122 Benzyl fi-cellobioside (3.9) (3.8) (6.33 g, 8.71 mmol) was suspended in anhydrous MeOH under nitrogen when a catalytic amount of sodium methoxide was added until a basic pH was obtained. After 24 h, the reaction was neutralized with acidic Amberlyte® resin. Removal of the solvent under reduced pressure yielded (3.9) (3.91 g) quantitatively as a white solid. 'H NMR (D20, 400 MHz) selected data only: 5 7.35 - 7.50 (m, 5 H, Ar), 4.90 (d, 1 H, J 11.6 Hz, PhCH), 4.75 (d, 1 H, J 11.6 Hz, PhCH), 4.56 (d, 1 H, J,, 2 8.0 Hz, H-l), 4.47 (d, 1 H, J r , 2 -7.8 Hz, H-l'), 3.96 (dd, 1 H, J 6 l l f 6 b 12.1, J 6 a > 5 2.0 Hz, H-6a), 3.88 (dd, 1 H, J 6 a , 6 b 12.6, J 6 > 5 2.2 Hz, H-6a'), 3.79 (dd, 1 H, j 6 b i 6 a 12.1, J6b,5 5.0 Hz, H-6b), 3.70 (dd, 1 H, J 6 b , 6 a 12.6, J6b,5 5.6 Hz, H-6b'). Benzyl2,2',3,3',6-penta-0-acetyl-4',6'-O-p-methoxybenzylidene-/3-cellobioside (3.10) To a suspension of (3.9) (3.64 g, 8.42 mmol) in DMF (100 mL) was added p-anisaldehyde dimethyl acetal (1.72 mL, 10.10 mmol) followed by a catalytic amount of />toluenesulphonic acid (0.03 g, 0.17 mmol). The reaction was then stirred under reduced pressure (20 mm Hg) for 10 d at 60 °C. After this time, solid NaHC0 3 was added to neutralize the reaction and the solvent was evaporated in vacuo. To the resulting residue was added pyridine (30 mL) and acetic anhydride (15 mL) and the reaction was stirred overnight. The volume of pyridine and acetic anhydride was then decreased by evaporation under reduced pressure and the resulting syrup was taken up in 10% v/v HC1 and extracted with CH2C12 (3 x 80 mL). The combined organic extracts were washed with 10% v/v HC1 (2 x 50 mL), aq. NaHC03 (50 mL) and water (50 mL), dried over MgS04 and the solvent was evaporated in vacuo. The residue was crystallized Chapter 4 Materials and Methods 123 from EtOAc/Hexane to yield (3.10) (3.72 g, 58%) as a white solid. ' H N M R (CDC1 3, 400 MHz): 8 7.20 - 7.35 (m, 7 H, Ar), 6.80 - 6.90 (m, 2 H, Ar), 5.40 (s, 1 H, MeOPhCH), 5.22 (dd, 1 H, J 3 , 2 9.4, J 3 , 4 9.3 Hz, H-3), 5.11 (dd, 1 H, irA- 9.3, J 3 , 2 9.2 Hz, H-3'), 4.94 (dd, 1 H, J 2 > 3 9.4, J 2 , , 7.9 Hz, H-2), 4.88 (dd, 1 H, J 2 , 3 9.2, J 2 ; 1 7.8 Hz, H-2'), 4.84 (d, 1 H, J 12.3 Hz, PhCH), 4.57 (d, 1 H, J l j 2 7.9 Hz, H- l ) , 4.56 (d, 1 H, J 12.3 Hz, PhCH), 4.50 (dd, 1 H, J 6 a ) 6 b 11.9, J 6 a > 5 1.9 Hz, H-6a), 4.49 (d, 1 H, J, > 2 7.8 Hz, H- l ' ) , 4.31 (dd, 1 H, J4,3" 9.3, J 4 ; 5 4.9 Hz, H-4'), 4.07 (dd, 1 H, J 6 b , 6 a 11-9, J6b,5 4.7 Hz, H-6b), 3.75 - 3.90 (m, 4 H, H-4, OCH 3 ) , 3.66 (dd, 1 H, J 6 a, 6b 10.2, J 6 a > 5 9.6 Hz, H-6a'), 3.64 (dd, 1 H, J 6 b ; 6 a 10.2, J6b-,5- 9.3 Hz, H-6b'), 3.53 (ddd, 1 H, J 5 , 4 9.8, J 5 , 6 b 4.7, J 5 , 6 a 1.9 Hz, H-5), 3.43 (ddd, 1 H, J 5 > 6 a 9.6, SS;w 9.3, J 5 . 4 - 4.9 Hz, H-5'), 2.12, 2.03, 2.00, 1.99, 1.97 (s, 15 H, 5 xOAc) . Benzyl 2,2 ',3,3',6-penta-0-acetyl-6'-0-p-methoxybenzyl-pl-cellobioside (3.11) (3.1.0) (1.17 g, 1.54 mmol) was dissolved in anhydrous D M F (12 mL) when N a C N B H 3 (0.49 g, 7.72 mmol) was added. A solution of TFA (1.20 mL, 15.43 mmol) in anhydrous DMF (8 mL) was then added dropwise to the reaction mixture and the solution was allowed to stir at room temperature overnight. The reaction was then filtered through Celite® and washed with CH 2 C1 2 . The filtrate was added to aq N a H C 0 3 (100 mL) and extracted with additional portions of CH 2 C1 2 (3 x 75 mL). The combined organic extracts were then washed with aq NaHC0 3 (100 mL), water (2 x 100 mL) and brine (100 mL), dried over M g S 0 4 and the solvent was removed under reduced pressure. Chromatography over silica gel (PE:EtOAc, 1:1) afforded (3.11) (0.85 g, 72%) as a white solid. ' H N M R (CDC1 3, 400 MHz): 8 7.15 - 7.35 (m, 7 H, Ar), 6.80 - 6.90 (m, 2 H, Ar), Chapter 4 Materials and Methods 124 5.09 (dd, 1 H , J 3 , 2 9.3, J 3 , 4 9.2 Hz, H-3), 4.95 (dd, 1 H , iy,r 9.4, J 3 , 4 9.3 Hz, H-3'), 4.93 (dd, 1 H , J 2 , 3 9.3, J 2 ,i 7.9 Hz, H-2), 4.82 (d, 1 H , J 12.3 Hz, PhCH), 4.80 (dd, 1 H , J 2 ) 3 9.4, J 2 7 . 9 Hz, H-2'), 4.55 (d, 1 H , J 12.3 Hz, PhCH), 4.50 (dd, 1 H , J 6 a ,6b 12.0, J 6 a , 5 2.0 Hz, H-6a), 4.47 (d, 1 H , J l 1.2 Hz, MeOPhCH), 4.46 (d, 1 H , J U 2 7.9 Hz, H- l ) , 4.44 (d, 1 H , J, > 2 7.9 Hz, H- l ' ) , 4.41 (d, 1 H, J 11.2 Hz, MeOPhCH), 4.06 (dd, 1 H , J 6 b j 6 a 12.0, J 6 b > 5 5.0 Hz, H-6b), 3.67 - 3.76 (m, 3 H , H-4, H-4', H-6a'), 3.64 (dd, 1 H , J 6 b , 6 a 9.9, J 6 b , 5 4.8 Hz, H-6b'), 3.52 (ddd, 1 H , J 5 > 4 9.8, J 5 , 6 b 5.0, J 5 , 6 a 2.0 Hz, H-5), 3.39 - 3.45 (m, 1 H , H -5'), 2.10, 2.03,2.00, 1.97, 1.94 (s, 15 H , 5 x OAc) . Benzyl 2,2 ',3,3', 6-penta-0-acetyl-4 '-deoxy-4 '-fluoro-pZ-lactoside (3.12) A solution of (3.11) (0.11 g, 0.15 mmol) in anhydrous C H 2 C 1 2 (5 mL) under an atmosphere of argon was cooled to -20 ° C before triflic anhydride (66.4 pL , 0.40 mmol) was added dropwise to the solution. The reaction was stirred at room temperature for 1 h and was then diluted with 10% v/v HC1 (20 mL). The mixture was extracted with C H 2 C 1 2 (3 x 20 mL) and the combined organic layers were washed with an additional portion of 10% v/v HC1 (20 mL), water (2 x 20 mL) and brine. After drying over M g S 0 4 , the solvent was evaporated under reduced pressure. The resulting residue was then taken up in anhydrous C H 2 C 1 2 and the solution was cooled to -10 ° C prior to the addition of tris(dimethylamino)sulfur (trimethylsilyl)difluoride (0.12 g, 0.44 mmol). The reaction was then refluxed for 0.5 h after which time water (20 mL) was added and the reaction was extracted with C H 2 C 1 2 (3 x 20 mL). The combined organic extracts were washed with water (2 x 20 mL), dried over M g S 0 4 and the solvent was removed in vacuo. Chromatography over silica gel (PE:EtOAc, 2:3) yielded (3.12) (40.3 mg, 43%) Chapter 4 Materials and Methods 125 as a white solid. ! H NMR (CDC13, 500 MHz): 5 7.25 - 7.35 (m, 5 H, Ar), 5.15 (dd, 1 H, J 2 , 3 10.3, J 2,i 8.0 Hz, H-2'), 5.12 (dd, 1 H, J 3 j 2 9.2, J 3 , 4 9.2 Hz, H-3), 4.96 (dd, 1 H, J 2 , 3 9.4, J2,i 7.8 Hz, H-2), 4.89 (ddd, 1 H, J 3 , F 27.6, J 3 j 2 10.4, J 3 ) 4 2.7 Hz, H-3'), 4.84 (d, 1 H, J 12.1 Hz, PhCH), 4.80 (dd, 1 H, J 4 ; F 50.3, J 4 , 3 2.7 Hz, H-4'), 4.57 (d, 1 H, J12.3 Hz, PhCH), 4.52 (d, 1 H, J r > 2 . 7.4 Hz, H-l'), 4.51 (dd, 1 H, J6a,6b 11.9, J 6 a , 5 2.2 Hz, H-6a), 4.49 (d, 1 H, h,2 7.8 Hz, H-l), 4.07 (dd, 1 H, J 6 b ) 6 a 11.9, J 6 b , 5 5.2 Hz, H-6b), 3.87 (ddd, 1 H, J 6 a , 6 b 11.5, 7.5, J&-.F 1.0 Hz, H-6a'), 3.85 (dd, 1 H, J 4 , 3 9.3, J 4 > 5 9.3 Hz, H-4), 3.72 (dd, 1 H, J 6 b , 6 a 11.5, J 6 b > 5 4.9 Hz, H-6b'), 3.61 (ddd, 1 H, J 5 > F 26.4, J 5 , 6 a 7.5, J5 ) 6b 4.9 Hz, H-5'), 3.55 - 3.59 (m, 1 H, H-5), 2.11,2.06, 2.04, 2.03, 1.98 (s, 15 H, 5 x OAc). 1 9 F NMR (CDC13, 188 MHz): 8 -140.0 (ddd, J F ; 4 50.3, J F , 3 27.6, J F , 5 26.4 Hz). Anal. Calcd for C 2 9 H 3 7 FO: C, 54.04; H, 5.79. Found: C, 54.32; H, 5.78. Benzyl 4 '-deoxy-4'-fluoro-fi-lactoside (3.13) A solution of (3.12) (0.09 g, 0.14 mmol) in anhydrous MeOH (5 mL) under argon was made basic through the addition of a catalytic amount of sodium methoxide. The reaction was stirred at room temperature overnight before it was neutralized with acidic Amberlyte® resin. Removal of the resin by filtration followed by evaporation of the solvent in vacuo gave a quantitative yield of (3.13) (60 mg) as a white solid. 'H NMR (D20, 400 MHz) selected data only: 5 7.35 - 7.55 (m, 5 H, Ar), 4.91 (d, 1 H, J 11.6 Hz, PhCH), 4.54 (d, 1 H, J, ,2 7.8 Hz, H-l'), 4.51 (d, 1 H, Ji> 2 6.0 Hz, H-l), 3.98 (dd, 1 H, J6a,6b 12.3, J 6 a , 5 1.9 Hz, H-6a), 3.33 (dd, 1 H, J 4 > 3 8.6, J 4 , 5 8.4 Hz, H-4). 1 9 F NMR (D20, 188 MHz): 5 -141.2 (ddd, J F , 4 51.0, J F j 3 30.0, J F > 5 30.0 Hz). Anal. Calcd for Ci9H2 7FO,o: C, 52.53; H, 6.26. Found: C, 52.33; H, 6.31. Chapter 4 Materials and Methods 126 4.2.3.3 Benzyl 4'-deoxylactoside Benzyl 2,2 ',3,3 ',6-penta-0-acetyl-4'-deoxy-4'-iodo-6'-0-p-methoxybenzyl-/3-lactoside (3.14) To a stirred solution of (3.11) (1.22 g, 1.60 mmol) in anhydrous CH2C12 (20 mL) under argon at -20 °C was added pyridine (1.1 mL, 13.47 mmol) followed by triflic anhydride (0.73 mL, 4.33 mmol). After 5 min, the reaction was allowed to warm to room temperature and then was stirred at this temperature for an additional hour. After dilution with 10% v/v HC1 (65 mL), the reaction was then extracted with CH2CI2 (3 x 65 mL). The combined organic extracts were washed with 10% v/v HC1 (65 mL), water (65 mL) and brine (65 mL), dried over MgS04 and the solvent was removed under reduced pressure. The residue was taken up in anhydrous DMF (20 mL), Nal (1.20 g, 8.02 mmol) was added and the reaction was stirred at room temperature overnight. After diluting with water (75 mL) and extracting with CH2CI2 (3 x 75 mL), the combined organic layers were then washed with water (4 x 75 mL) and brine (75 mL), dried over MgSC»4 and the solvent was evaporated in vacuo. The residue was chromatographed over silica gel (PE:EtOAc, 3:2) and the desired fractions were pooled and recrystallized from EtOAc/Hex to yield (3.14) (0.54 g, 38%) as a white solid. \H NMR (CDCI3, 400 MHz): 5 7.20 - 7.35 (m, 7 H, Ar), 6.85 - 6.90 (m, 2 H, Ar), 5.15 (dd, 1 H, h,y 9.9, J 2 ,1 7.8 Hz, H-2'), 5.10 (dd, 1 H, J 3 > 2 9.5, J 3 , 4 9.3 Hz, H-3), 4.95 (dd, 1 H, J 2 , 3 9.5, h,\ 7.9 Hz, H-2), 4.83 (d, 1 H, J 12.3 Hz, PhCH), 4.64 (dd, 1 H, J 4 , 3 4.2, J 4 , 5 1.1 Hz, H-4'), 4.56 (d, 1 H, J 12.3 Hz, PhCH), 4.47 (d, 1 H, J, > 2 7.8 Hz, H-l'), 4.43 - 4.49 (m, 2 H, MeOPhCH2), 4.43 (d, 1 H, J l j 2 7.9 Hz, H-l), 4.43 (dd, 1 H, J 6 a, 6b 11.9, J 6 A ,5 2.0 Hz, H-6a), 4.25 (dd, 1 H, J3,2 9.9, J3-,4- 4.2 Hz, H-3'), 4.05 (dd, 1 H, J 6 b , 6 a 11.9, J 6 b , 5 5.0 Hz, H-6b), 3.72 - 3.80 (m, 4 H, Chapter 4 Materials and Methods 727 H-4, OCH3), 3.62 (dd, 1 H, J 6 a , 6 b 9.4, J 6 a , 5 5.5 Hz, H-6a'), 3.53 (ddd, 1 H, J 5 , 4 9.9, J5 ;6b 5.0, J 5 , 6 a 2.0 Hz, H-5), 3.45 (dd, 1 H, J 6 b - ( 6 a - 9.4, J 6 b j 5 7.0 Hz, H-6b'), 2.91 (ddd, 1 H, J5,6b- 7.0, J5-,6a- 5.5, Jy,4' 1.1 Hz, H-5'), 2.09, 2.05, 2.01, 1.98, 1.97 (s, 15 H, 5 x OAc). Benzyl2,2',3,3 ',6-penta-0-acetyl-4'-deoxy-6'-O-p-methoxybenzyl-p-lactoside (3.15) (3.14) (0.50 g, 0.57 mmol) was dissolved in anhydrous benzene (15 mL) when tributyltin hydride (0.83 g, 2.87 mmol) and a catalytic amount of AIBN were added and the reaction was refluxed under an atmosphere of argon. After 7 h, hexane (60 mL) was added and the reaction was extracted with acetonitrile (80 mL). The acetonitrile layers were washed with two additional portions of hexane (60 mL) before the solvent was evaporated under reduced pressure. Chromatography over silica gel (PE:EtOAc, 2:1 to 3:2) yielded (3.15) (0.36 g, 84%) as a colourless gum. 'H NMR (CDC13, 400 MHz): 8 7.15 - 7.35 (m, 7 H, Ar), 6.84 - 6.89 (m, 2 H, Ar), 5.11 (dd, 1 H, J 3 ) 2 9.4, J 3 , 4 9.2 Hz, H-3), 4.94 (dd, 1 H, J 2 , 3 9.4, J2 >i 7.9 Hz, H-2), 4.89 (ddd, 1 H, J3,4ax 11.5, J 3 , 2 9.7, J 3 , 4 - e q 5.4 Hz, H-3'), 4.82 (d, 1 H, J 12.3 Hz, PhCH), 4.76 (dd, 1 H, J 2 j 3 9.7, h;v 7.8 Hz, H-2'), 4.55 (d, 1 H, J 12.3 Hz, PhCH), 4.49 (dd, 1 H, J 6 a , 6 b 12.0, J 6 a , 5 2.0 Hz, H-6a), 4.47 (d, 1 H, J,, 2 7.9 Hz, H-l), 4.41 (s, 2 H, MeOPhCH2), 4.35 (d, 1 H, J, j 2 7.8 Hz, H-l'), 4.10 (dd, 1 H, J 6 b ,6a 12.0, J 6 b , 5 4.9 Hz, H-6b), 3.73 - 3.80 (m, 4 H, H-4, OCH3), 3.51 - 3.64 (m, 2 H, H-5, H-5), 3.51 (dd, 1 H, J 6 a ,6b- 9.8, 5.1 Hz, H-6a'), 3.40 (dd, 1 H, J 6 b , 6 a 9.8, ' J 6 b-, 5- 5.0 Hz, H-6b'), 2.09, 2.01, 1.97, 1.95 (s, 15 H, 5 x OAc), 1.47- 1.63 (m, 2 H, H-4'ax, H-4'eq). Chapter 4 Materials and Methods 128 Benzyl 2,2 ',3,3',6-penta-0-acetyl-4 '-deoxy-j3-lactoside (3.16) Ceric ammonium nitrate (0.55 g, 1.00 mmol) was added to a solution of (3.15) (0.34 g, 0.46 mmol) in 9:1 acetonitrile/water (4 mL) and the reaction was stirred at room temperature for 6 h. After this time, the reaction was added to aq. NaHCCh (15 mL) and extracted with C H 2 C I 2 (3x15 mL). The combined organic extracts were washed with aq. NaHCOa (15 mL), water (15 mL) and brine (15 mL), dried over MgSC»4 and the solvent was removed under reduced pressure. The resulting residue was crystallized from EtOAc/Hex to yield (3.16) (232 mg, 81%) as a white needle-like solid. ! H NMR ( C D C I 3 , 400 MHz): 5 7.20 -7.35 (m, 5 H, Ar), 5.14 (dd, 1 H, J 3 , 2 9.2, J 3 > 4 9.1 Hz, H-3), 4.94 (dd, 1 H, J 2 , 3 9.2, J 2 ) i 7.8 Hz, H-2), 4.88 - 4.96 (m, 1 H, H-3'), 4.83 (d, 1 H, J 12.3 Hz, PhCH), 4.78 (dd, 1 H, J 2 , 3 9.4, J 2 - 1 7.7 Hz, H-2'), 4.57 (d, 1 H, J 12.3 Hz, PhCH), 4.51 (dd, 1 H, J 6 a , 6 b 11-8, J 6 a , 5 2.1 Hz, H-6a), 4.49 (d, 1 H, J , , 2 7.8 Hz, H-l), 4.46 (d, 1 H, J , > 2 7.7 Hz, H-l'), 4.08 (dd, 1 H, J 6 B > 6 A 11.8, J 6 b , 5 5.3 Hz, H-6b), 3.82 (dd, 1 H, J 4 > 5 9.5, J 4 , 3 9.1 Hz, H-4), 3.51 - 3.64 (m, 4 H, H-5, H-5', H-6a', H-6b'), 2.11, 2.04, 2.03, 1.98, 1.97 (s, 15 H, 5 x OAc), 1.46 - 1.60 (m, 2 H, H-4'^, H-4'eq). Anal. Calcd for C 2 9 H 3 8 0 , 5 : C, 55.59; H, 6.11. Found: C, 55.91; H, 6.16. Benzyl 4'-deoxy-/3-lactoside (3.17) A solution of (3.16) (0.21 g, 0.33 mmol) in anhydrous MeOH (10 mL) under argon was made basic through the addition of a catalytic amount of sodium methoxide. The reaction was stirred at room temperature overnight before it was neutralized with acidic Amberlyte® resin. Removal of the resin by filtration followed by evaporation of the solvent in vacuo gave a quantitative yield of (3.17) (147 mg) as a white solid, m.p 169 -Chapter 4 Materials and Methods 129 171 °C. 'H NMR (CD3OD, 400 MHz) selected data only: 5 7.22 - 7.45 (m, 5 H, Ar), 4.91 (d, 1 H, J 11.8 Hz, PhCH), 4.66 (d, 1 H, J 11.8 Hz, PhCH), 4.39 (d, 1 H, J,, 2 7.8 Hz, H-l), 4.34 (d, 1 H, J, ; 2 7.8 Hz, H-l'), 3.93 (dd, 1 H, J6a,6b 12.1, J 6 a , 5 2.4 Hz, H-6a), 3.86 (dd, 1 H, J 6 b , 6 a 12.1, J 6 b , 5 4.2 Hz, H-6b), 3.13 (dd, 1 H, J 4 > 5 8.7, J 4 , 3 8.1 Hz, H-4), 1.89 (dd, 1 H, J 4 e q , 4 a x 12.7, J 4 e q ,3 5.0 Hz, H-4'eq), 1.41 (ddd, 1 H, J 4 a X ) 4 e q 12.7, J 4 a x , 3 11.9, J 4 a x > 5 11.9 Hz, H-4',«). Anal. Calcd for C 1 9H 2 SOio: C, 54.80; H, 6.78. Found: C, 54.50; H, 6.72. 4.2.4 Fluorinated donor substrate analogues (Schemes 3.3 and 3.4) 4.2.4.1 5-Fluoro-a-D-galactopyranosyl fluoride 2,3,4,6-Tetra-0-acetyl-5-bromo-a-D-galactopyranosylfluoride (3.18). A suspension of 2,3,4,6-tetra-O-acetyl-a-D-galactopyranosyl fluoride (2.12) (2.0 g, 5.7 mmol) and A'-bromosuccinimide (4.1 g, 22.8 mmol) in anhydrous CC14 (50 mL) was heated to reflux under nitrogen by means of two 200-watt household light bulbs. After 9 hours, the reaction was allowed to cool and then filtered. The filtrate was washed with aqueous NaHC03 (2x50 mL) and water (2x50 mL), dried (MgS04) and concentrated in vacuo. The residue was chromatographed over silica gel (CH2Cl2:EtOAc, 69:1) to yield (3.18) (1.28 g, 52%) as a colourless gum. 'H NMR ( C D C I 3 , 200 MHz): 5 6.03 (dd, 1 H, J3,211.2, J3)42.9 Hz, H-3), 5.95 (dd, 1 H, J,, F 53.4, Ji,2 3.3 Hz, H-l), 5.78 (d, 1 H, J4>i2.9 Hz, H-4), 5.26 (ddd, 1 H, J2,F23.4, J2>311.2, J2,i 3.3 Hz, H-2), 4.52 (d, 1 H, j 6 a , 6 b 11.9 Hz, H-6a), 4.32 (d, 1 H, J 6 b , 6 a 11.9 Hz, H-6b), 2.00-2.12 (s, 12 H, 4 x OAc). 1 9 F NMR ( C D C I 3 , 188 MHz): 5 -70.5 (dd, JF,i 53.4, J F , 2 23.4 Hz). HRMS-DCI: Calc. for Chapter 4 Materials and Methods 130 C, 4 H 1 8 0 9 F 7 9 Br + NH 4 + : 446.04619, Found: 446.04645; Calc. for Ci 4 H 1 8 0 9 F 8 1 Br + NH 4 + : 448.04415, Found: 448.04343. 2,3,4,6-Tetra-0-acetyl-5-fluoro-cc-D-galactopyranosylfluoride (3.19). To a solution of (3.18) (0.25 g, 0.58 mmol) in anhydrous toluene (5 mL) and 4 A sieves was added silver tetrafluoroborate (0.27 g, 1.37 mmol) and the reaction was stirred under an atmosphere of nitrogen. After 40 minutes, the reaction was filtered through Celite and rinsed with EtOAc. The filtrate was then washed with aqueous NaHC03 (25 mL) and brine (25 mL), dried (MgS04) and the solvent evaporated in vacuo. Purification by silica gel chromatography (CH2C12) yielded (3.19) (29 mg, 14%) as a colourless gum. 'H NMR (CDC13, 200 MHz): 5 5.85 (dd, 1 H, J,, F 1 53.1, J,, 2 3.0 Hz, H-l), 5.60-5.67 (m, 2 H, H-3, H-4), 5.23 (ddd, 1 H, J2,Fi 23.1, J 2 , 3 10.8, J2>i 3.0 Hz, H-2), 4.42 (dd, 1 H, J6a,F5 25.5, J6a,6b 12.3 Hz, H-6a), 4.06 (dd, 1 H, J6b,F5 12.3, J6b,6a 12.3 Hz, H-6b), 2.00-2.11 (s, 12 H, 4 x OAc). 1 9 F NMR (CDCI3, 188 MHz): 5 -38.1 (ddd, JF5,6a25.5, J F 5 > F 1 23.9, J F 5 )6b 12.3 Hz, F-5), -64.9 (ddd, J F , , , 53.1, J Fi, 2 23.1, J F ] , F 5 23.9 Hz, F-l). HRMS-DCI: Calc. for Ci 4 Hi 8 0 9 F 2 + NH 4 + : 386.12627, Found: 386.12690. 5-Fluoro-a-D-galactopyranosylfluoride (3.20). A solution of (3.19) (46 mg, 0.13 mmol) in anhydrous methanol under an atmosphere of nitrogen was cooled to 0 °C. Ammonia gas was bubbled into the reaction for 5 minutes, after which the reaction vessel was sealed and allowed to warm to room temperature. After 2.5 hours, the reaction was shown to be complete by TLC. The solvent was subsequently evaporated in vacuo and the residue was chromatographed over silica gel Chapter 4 Materials and Methods 131 (EtOAc:MeOH, 49:1) to yield (3.20) (14 mg, 57%) as a colourless syrup. ] H NMR (CD3OD, 400 MHz): 5 5.58 (dd, 1 H, J, , F 1 54.4, J 1 > 2 3.2 Hz, H-l), 4.13 (dd, 1 H, J4,F5 2.7, J4)31.9 Hz, H-4), 4.02 (ddd, 1 H, J3,210.5, J 3 , F 5 2.4, J 3 , 4 1.9 Hz, H-3), 3.93 (ddd, 1 H, J 2 , F 1 25.0, J 2 > 3 10.5, h,\ 3.2 Hz, H-2), 3.79 (dd, 1 H, J 6 a , F 5 23.3, J6a,6b 12.2 Hz, H-6a), 3.59 (dd, 1 H, J 6 b > F5 14.2, J 6 b , 6 a 12.2 Hz, H-6b). 1 9 F NMR (CDC13, 188 MHz): 5 -44.82 (ddd, J F 5,6 a 23.3, J F 5 , F 1 22.5, J F 5 j 6 b 14.2 Hz, F-5), -66.98 (ddd, J F 1 > , 54.4, J F U 2 25.0, J F 1 , F 5 22.5 Hz, F-l). HRMS-DCI: Calc. for C 6H,o0 5F 2 + NH 4 + : 200.08400, Found: 200.08346. 4.2.4.2 Uridine 5 -diphospho-(2' -deoxy-2' -fluoro)-a-D-galactopyranose 3,4,6-Tri-0-acetyl-2-deoxy-2-fluoro-D-galactopyranose (3.21) 3,4,6-Tri-O-acetyl-D-galactal (0.912 g, 3.349 mmol) was dissolved in DMF (15 mL) when water (6 mL) and Selectfluor™ (N-fluoro-N-chloromethyltriethylenediamine bis(tetrafluoroborate), 4.20 g, 11.80 mmol) from Air Products and Chemicals Inc. were added and the reaction was stirred at 50 °C. After 24 h, the reaction was shown to be complete by TLC (PE:EtOAc, 3:2). To the reaction was added water (40 mL) and this was then extracted with CH2C12 (3 x 50 mL). The combined organic layers were washed with water (3 x 50 mL), dried over MgS04 and the solvent was removed under reduced pressure. Chromatography of the resulting residue over silica gel (PE:EtOAc, 3:2 to 1:1) yielded (3.21) (0.36 g, 35%) as a colourless gum. ] H NMR (CDC13, 400 MHz) for the a anomer: 5 5.52 (d, 1 H, J 1 > 2 3.8 Hz, H-l), 5.38 - 5.50 (m, 2 H, H-3, H-4), 4.75 (ddd, 1 H, J 2 j F 49.9, J 2 ; 3 10.0, J 2 > 1 3.8 Hz, H-2), 4.47 (m, 1 H, H-5), 4.00-4.15 (m, 2 H, H-6a, H-6b), 2.10, 2.00, 1.99 (s, 9 H, 3 x OAc). 1 9 F NMR(CDC13, 188 MHz): 5-131.1 (dd, J F ; 2 49.9, J F , 3 14.5 Hz). Chapter 4 Materials and Methods 132 1,3,4,6-Tetra-0-acetyl-2-deoxy-2-fluoro-/3-D-galactopyranose (3.22) To a solution of (3.21) (0.88 g, 2.84 mmol) in pyridine (7 mL) was added acetic anhydride (3.5 mL) and the reaction was stirred at room temperature overnight. The pyridine and acetic anhydride were then removed by evaporation under reduced pressure. The residue was then taken up in 10% v/v HC1 (80 mL) and extracted with CH2CI2 (3 x 70 mL). The combined organic extracts were washed with 10% v/v HC1 (70 mL), aq. NaHC03 (70 mL), water (70 mL) and brine (70 mL), dried over MgS04 and the solvent was removed in vacuo. The residue was then dissolved in anhydrous CH2CI2 (10 mL) under an atmosphere of argon and the temperature was brought to 0 °C. To this was added a solution of 45% HBr/AcOH after which the argon source was removed and the reaction vessel was sealed and allowed to warm to room temperature. After 4 h, the reaction was poured into ice water (80 mL) and diluted with CH2CI2 (80 mL). Solid NaHCCh was added to neutralize the excess acid and the layers were separated. The aqueous layer was further extracted with CH2CI2 (2 x 70 mL). The organic layers were subsequently combined and washed with aq. NaHCCh (2 x 70 mL), water (100 mL) and brine (70 mL) and dried over MgSC»4. Evaporation of the solvent under reduced pressure yielded a beige gum to which acetic acid (35 mL) and Hg(OAc)2 (1.87 g, 5.97 mmol) were added. The reaction was allowed to stir at room temperature under an atmosphere of argon. After 3 h, the reaction was poured into water (100 mL) and then extracted with CH2CI2 (3 x 80 mL). The combined organic extracts were washed with aq. NaHCCh (3 x 80 mL), water (80 mL) and brine (80 mL). After drying over MgSC»4, the solvent was remove in vacuo and the residue was chromatographed over silica gel (PE:EtOAc, 3:1 to 5:2) to yield (3.22) (0.83 g, 83%) as a white solid. *H NMR (CDC13, 400 MHz): 5 5.77 Chapter 4 Materials and Methods 133 (dd, 1 H, J 1 > 2 8.1, JI,F 4.1 Hz, H-l), 5.43 (m, 1 H, H-4), 5.15 (ddd, 1 H, J 3 , F 13.2, J 3 > 2 9.8, J 3 ; 4 3.6 Hz, H-3), 4.62 (ddd, 1 H, J 2 , F 51.6, J 2 , 3 9.8, h,i 8.1 Hz, H-2), 4.05 - 4.20 (m, 3 H, H-5, H-6a, H6b), 2.18, 2.14, 2.05, 2.03 (s, 12 H, 4 x OAc). 1 9 F NMR (CDC13, 188 MHz): 5-132.1 (dd, J F ; 2 51.6, J F ) 3 13.2 Hz). 2-Deoxy-2-fluoro-CC-D-galactopyranose-l-phosphate, mono(tri-n-octyl)ammonium salt (3.23) Anhydrous H3P04 (0.3lg, 3.17 mmol) was dried under vacuum for 24 h before it was melted at 50 °C. (3.22) (0.14 g, 0.40 mmol) was then added and the reaction was stirred under reduced pressure (20 mm Hg) for 9 h. After this time, THF (1 mL) and a solution of 2 M LiOH (6 mL) were added and the reaction was allowed to stir at room temperature overnight. After filtering through Gelite® and washing with 0.01 M LiOH, the solvent was evaporated in vacuo. The residue was then taken up in water and passed through a column of Bio-Rad AG 50W-X2, 200-400 mesh, sulfonic acid cation exchanger (H+ form). The desired fractions were pooled and the solvent volume was decreased by evaporation under reduced pressure. Tri-n-octylamine (0.14 g, 0.40 mmol) was added and the solution was lyophilized yielding (3.23) (0.28 g) as a colourless syrup. 'H NMR (CDC13, 400 MHz): 5 5.74 (dd, 1 H, J i , P 5.9, J , , 2 3.6 Hz, H-l), 4.63 (m, 1 H, H-2), 4.23 (m, 1 H, H-3), 3.72 - 4.15 (m, 4 H, H-4, H-5, H-6a, H-6b), 2.80 (m, 6 H, NCH2), 1.65 (m, 6 H, NCCH2), 1.30 (m, 30 H, CH2), 0.85 (t, 9 H, CH3). 1 9 F NMR (CDC13, 188 MHz, proton decoupled): 5 -132.4. 3 1 P NMR (CDC13, 81 MHz, proton decoupled): 5 0.01. Chapter 4 Materials and Methods 134 Uridine 5 '-diphospho-(2-deoxy-2-fluoro)-a-D-galactopyranose, di-ammonium salt (3.24) To (3.23) (0.25 g, 0.40 mmol) was added anhydrous pyridine (5 mL), which was then evaporated. This procedure was repeated twice before UMP-morpholidate (0.33 g, 0.48 mmol) was added. Evaporation with anhydrous pyridine (5 mL) was again repeated three times. 1H tetrazole (0.07 g, 1.01 mmol) and anhydrous pyridine (3 mL) were then added and the reaction was stirred at room temperature. An aliquot of the reaction mixture was transferred into an NMR tube containing a capillary of DMSO-d6 so that the progress of the reaction could be monitored via 3 1P NMR. After 27 days, the reaction was diluted with water and evaporated under reduced pressure. After repeating this four times, the residue was taken up in 100 mM N H 4 H C O 3 (5 mL) and the tri-n-octylamine was extracted with diethyl ether (3x5 mL). The aqueous layer was lyophilized to yield the crude product. Purification was afforded by size exclusion chromatography through a column of Bio-Gel P2 extra fine resin (1 x 45 cm) using a Beckman Biosepra ProSys Workstation. The product was eluted with 250 mM NH4HCO3 at a flow rate of 0.1 mL/min. The desired fractions were pooled and lyophilized to yield (3.24) (90.0 mg, 37%) as a white powdery solid. 'H NMR (D20, 400 MHz) selected data only: 5 7.91 (d, 1 H, J 6 , 5 6.1 Hz, H-6), 5.93 (m, 2 H, H-l', H5), 5.76 (dd, 1 H, J, ,P 7.1, J r > 2 3.6, H-l"). 1 9 F NMR (D20, 188 MHz): 5 -132.4 (dd, J 2 • F 49.9, J 3 ,F 11.1 Hz). 3 1P NMR (D20, 81 MHz, proton decoupled): 5-9.10 (d, JPp,Pa 19.9 Hz, Pp), -10.8 (d, JP a ( Pp 19.9 Hz, Pa). Anal. Calcd for C 1 5 H 2 9 FN 4 Oi6P2: C, 29.91; H, 4.85; N, 9.30. Found: C, 30.37; H, 5.34; N, 9.89. Chapter 4 Materials and Methods 135 4.2.5 6 '-Deoxylactose (Scheme 3.5) 1,2,2',3,3',6-Hexa-0-acetyl-4'-O-henzoyl-6'-hromo-6'-deoxylactose (3.25) N-bromosuccinimide (0.06 g, 0.29 mmol) and barium carbonate (0.12 g, 0.59 mmol) were added to a solution of (3.1) (0.20 g, 0.29 mmol) in anhydrous CCU (9 mL) and the mixture was refluxed under an atmosphere of argon. After 1.5 h, the reaction was cooled to 0 °C and filtered. Evaporation of the filtrate under reduced pressure followed by chromatography of the resulting residue over silica gel (PE:EtOAc, 3:2 to 9:7) yielded (3.25) (156.3 mg, 70%) as a colourless gum. *H NMR (CDC13, 400 MHz) oc-anomer: 5 7.45 - 8.10 (m, 5 H, Ar), 6.29 (d, 1 H, J , , 2 3.7 Hz, H-l), 5.75 (d, 1 H, J 4-,y 2.8 Hz, H-4'), 5.50 (dd, 1 H, J 3 , 2 9.9, J 3 , 4 9.6 Hz, H-3), 5.20 (dd, 1 H, J 2 , 3 - 10.2, J 2 x 7.9 Hz, H-2'), 5.07 (dd, 1 H, J 3 > 2 10.2, J 3 , 4 2.8 Hz, H-3'), 5.02 (dd, 1 H, J 2 ) 3 9.9, J 2 , i 3.7 Hz, H-2), 4.58 (d, 1 H, Jr ; 2' 7.9 Hz, H-l'), 4.47 (dd, 1 H, J 6 a, 6b 12.2, J 6 a > 5 1.8 Hz, H-6a), 4.13 (dd, 1 H, J 6 b > 6 a 12.2, J 6 b , 5 4.1 Hz, H-6b), 4.06 (ddd, 1 H, J 5 , 4 10.2, J 5 , 6 b 4.1, J 5 , 6 a 1.8 Hz, H-5), 3.92 (t, 1 H, J 5 j 6 6.5 Hz, H-5'), 3.88 (dd, 1 H, J 4 , 5 10.2, J 4 j 3 9.6 Hz, H-4), 3.38 (d, 2 H, J 6 ; 5 6.5 Hz, 2 x H-6'), 2.17, 2.13, 2.04, 2.03, 2.00, 1.92 (s, 18 H, 6 x OAc). I, 2,2 ',3,3',6-Hexa-0-acetyl-4'-O-henzoyl-6'-deoxylactose (3.26) A solution of (3.25) (0.15 g, 0.19 mmol) in anhydrous benzene (5 mL) was stirring at room temperature under argon when tributyltin hydride (0.28 g, 0.97 mmol) and a catalytic amount of AIBN were added. The reaction was refluxed for 1 h before the solvent was evaporated in vacuo. The residue was then taken up in acetonitrile (30 mL) and washed with hexane (3 x 20 mL). Evaporation of the acetonitrile under reduced pressure followed by chromatography of the resulting residue over silica gel (PE:EtOAc, Chapter 4 Materials and Methods 136 5:4) yielded (3.26) (113.0 mg, 86%) as a colourless gum. ] H NMR (CDC13, 400 MHz) a-anomer: 8 7.45 - 8.15 (m, 5 H, Ar), 6.27 (d, 1 H, J 1 > 2 3.7 Hz, H-l), 5.48 (dd, 1 H, J 3 , 2 10.0, J 3 , 4 9.4 Hz, H-3), 5.43 (d, 1 H, J 4 - 3 - 2.9 Hz, H-4'), 5.19 (dd, 1 H, J 2 > 3 10.3, J 2 , , 7.8 Hz, H-2'), 5.04 (dd, 1 H, J 3 > 2 10.3, J 3 > 4 2.9 Hz, H-3'), 5.02 (dd, 1 H, J 2 ) 3 10.0, J2>i 3.7 Hz, H-2), 4.50 (d, 1 H, Ji > 2 7.8 Hz, H-l'), 4.45 (dd, 1 H, J6a,6b 12.2, J 6 a , 5 2.0 Hz, H-6a), 4.12 (dd, 1 H, J 6 b j 6 a 12.2, J 6 b , 5 4.4 Hz, H-6b), 4.01 (ddd, 1 H, J 5 , 4 10.2, J 5 j 6 b 4.4, J 5 , 6 a 2.0 Hz, H-5), 3.85 (q, 1 H, J 5 > 6 6.4 Hz, H-5'), 3.82 (dd, 1 H, J 4 , 5 10.2, J 4 , 3 9.4 Hz, H-4), 2.17, 2.12,2.03,2.01,2.00, 1.91 (s, 18 H, 6 x OAc), 1.25 (d, 3 H, J 6 , 5 6.4 Hz, 3xH-6). 6'-Deoxylactose (3.27) To a stirring solution of (3.26) (0.10 g, 0.15 mmol) in anhydrous MeOH (1.5 mL) under argon was added a catalytic amount of sodium methoxide until the pH of the solution was slightly basic. The reaction was stirred at room temperature for 5 h before Amberlyte® IR-120 acidic resin was added to neutralize the solution. Removal of the resin by filtration followed by evaporation of the solvent in vacuo gave a quantitative yield of (3.27) (48.5 mg) as a white solid. ] H NMR (D20, 400 MHz) selected data only: 5 5.18 (d, 1 H, J l a , 2 3.0 Hz, H-la), 4.65 (d, 1 H, J ] p > 2 7.8 Hz, H-lp), 4.38 (d, 1 H, J r , 2 - 7.7 Hz, H-l'), 3.48 (ddd, 1 H, J 5 , 4 7.2, J 5 ) 6 b 7.2, J 5 j 6 a 2.5 Hz, H-5), 1.25 (d, 3 H, J 6 ; 5 6.3 Hz, 3 x H-6'). Anal. Calcd for C ^ O i o ^ O : C, 42.98; H, 6.91. Found: C, 42.98; H, 7.00. 4.2.6 Gal-P-Lac trisaccharides (Scheme 3.6) 1,2,2 ',3,3 ',4',6-Hepta-0-acetyl-6'-O-benzyllactose (3.28) A solution of (3.2) (0.57 g, 0.83 mmol) in pyridine (4 mL) and acetic anhydride (2 mL) Chapter 4 Materials and Methods 137 was stirred at room temperature overnight. The solvent was evaporated in vacuo and the residue was taken up in 10% v/v HC1 (40 mL) and extracted with CH2CI2 (3 x 40 mL). The combined organic extracts were washed with additional portions of 10% v/v HC1 (2 x 40 mL), aq. NaHC03 (40 mL) and water (40 mL), dried over MgS04 and the solvent was removed under reduced pressure. Chromatography over silica gel (PE:EtOAc, 5:4 to 1:1) yielded (3.28) (504.2 mg, 83%) as a white solid. 'H NMR (CDC13, 400 MHz) a -anomer: 5 7.20 - 7.40 (m, 5 H, Ar), 6.25 (d, 1 H, J,, 2 3.7 Hz, H-l), 5.43 (dd, 1 H, J 3 , 2 10.0, J 3 , 4 9.3 Hz, H-3), 5.42 (d, 1 H, J 4 , 3 2.8 Hz, H-4'), 5.07 (dd, 1 H, J 2 j 3 10.4, i2;r 7.8 Hz, H-2'), 4.98 (dd, 1 H, J 2 ) 3 10.2, J2,i 3.7 Hz, H-2), 4.93 (dd, 1 H, J 3 j 2 10.4, J 3-A- 2.8 Hz, H-3'), 4.50 (d, 1 H, J 11.9 Hz, PhCH), 4.44 (d, 1 H, J, j 2 7.8 Hz, H-l), 4.42 (dd, 1 H, J 6 a > 6 b 12.2, J 6 a , 5 2.0 Hz, H-6a), 4.40 (d, 1 H, J 11.9 Hz, PhCH), 4.07 (dd, 1 H, J 6 b j 6 a 12.2, J 6 b , 5 4.3 Hz, H-6b), 3.98 (ddd, 1 H, J 5 > 4 10.0, J 5 > 6 b 4.3, J 5 , 6 a 2.0 Hz, H-5), 3.80 (dd, 1 H, J 4 , 5 10.0, J 4 , 3 9.3 Hz, H-4), 3.73 - 3.85 (m, 1 H, H-5'), 3.51 (dd, 1 H, J 6 a , 6 b 9.3, J 6 a , 5 5.7 Hz, H-6a'), 3.41 (dd, 1 H, J 6 b , 6 a 9.3, J 6 b ) 5 7.2 Hz, H-6b'), 2.14, 2.08, 2.04, 2.02, 1.98, 1.96, 1.94 (s, 21 H, 7 x OAc). 1,2,2'',3,3 ',4', 6-Hepta- O-acetyl- a-lactose (3.29) (3.28) (0.49 g, 0.67 mmol) was dissolved in a 1:1 solution of EtOAc/MeOH (20 mL) and hydrogenated at 1 atm in the presence of 5% w/w Pd/C (54 mg) and a drop of acetic acid. After 4 d, the reaction was filtered through Celite® and the solvent was evaporated under reduced pressure. The residue was crystallized from EtOAc/Hex to yield (3.29) (304 mg, 71%) as a white solid. 'H NMR (CDC13, 400 MHz): 5 6.25 (d, 1 H, J 1 > 2 3.7 Hz, H-l), 5.46 (dd, 1 H, J 3 > 2 9.9, J 3 > 4 9.3 Hz, H-3), 5.35 (d, 1 H, J 4 > 3 3.3 Hz, H-4'), 5.16 (dd, 1 H, Chapter 4 Materials and Methods 138 J 2 , 3 10.4, J2-,v 7.9 Hz, H-2'), 5.03 (dd, 1 H, J 2 > 3 9.9, J 2 >, 3.7 Hz, H-2), 5.00 (dd, 1 H, J 3 > 2 10.4, J 3,4 3.3 Hz, H-3'), 4.55 (d, 1 H, J, ;2- 7.9 Hz, H-l'), 4.47 (dd, 1 H, J6a,6b 12.1, J 6 a , 5 2.0 Hz, H-6a), 4.08 (dd, 1 H, J 6 b ,6 a 12.1, J6b,5 4.4 Hz, H-6b), 4.03 (ddd, 1 H, J 5 , 4 10.0, J 5 ; 6 b 4.4, J 5 , 6 a 2.0 Hz, H-5), 3.86 (dd, 1 H, J 4 > 5 10.0, J 4 > 3 9.3 Hz, H-4), 3.65 - 3.76 (m, 2 H, H-5',H-6a'), 3.50 (dd, 1 H, j 6 b ) 6 a 13.8, J 6 b ; 5 9.8 Hz, H-6b'), 2.19, 2.18, 2.12, 2.07, 2.06, 2.00, 1.98 (s,21 H, 7 x OAc). 2,3,4,6- Tetra-O-acetyl-D-galactopyranose (3.30) Per-O-acetylated p-D-galactose (1.46 g, 3.75 mmol) was dissolved in anhydrous DMF (20 mL) under argon when hydrazine acetate (0.38 g, 4.13 mmol) was added. The reaction was stirred at room temperature and carefully monitored by TLC (PE:EtOAc, 1:1) until the starting material was depleted. Water (70 mL) was then added to the reaction mixture, which was subsequently extracted with CH2C12 (3 x 70 mL). The combined organic extracts were washed with aq. NaHC0 3 (70 mL) and water (2 x 70 mL), dried over MgS04 and the solvent was removed in vacuo. Purification was afforded by silica gel chromatography to yield (3.30) (1.12 g, 86%) as a white gummy solid. 'H NMR (CDC13, 400 MHz) a-anomer: 5 5.48 (d, 1 H, J,, 2 3.5 Hz, H-l), 5.44 (dd, 1 H, J 4 ; 3 3.3, J 4 , 5 1.2 Hz, H-4), 5.38 (dd, 1 H, J 3 , 2 10.8, J3>4.3.3 Hz, H-3), 5.12 (dd, 1 H, J 2 , 3 10.8, J2,i 3.5 Hz, H-2), 4.44 (ddd, 1 H, J 5 j 6 a 6.6, j 5 j 6 b 6.6, J 5 ) 4 1.2 Hz, H-5), 4.09 (dd, 1 H, J 6 a , 6 b 11.4, J 6 a , 5 6.6 Hz, H-6a), 4.05 (dd, 1 H, J 6 b > 6 a 11.4, J 6 b , 5 6.6 Hz, H-6b), 2.11, 2.07, 2.02, 1.96 (s, 12 H, 4 x OAc). Chapter 4 Materials and Methods 139 2,3,4,6-Tetra-O-acetyl-a-D-galactopyranosyl trichloroacetimidate (3.31) To a solution of (3.30) (0.85 g,. 2.43 mmol) in anhydrous CH2CI2 (7 mL) under argon at -40 °C was added trichloroacetonitrile (2.43 mL, 24.30 mmol) followed by DBU (36 |iL, 0.24 mmol). The reaction was kept between -20 °C and -40 °C for 3 h before the solvent was evaporated under reduced pressure. Chromatography over silica gel (PE:EtOAc, 3:1 to 2:1) yielded (3.31) (0.98 g, 82%) as a white solid. 'H NMR (CDC13, 400 MHz): 5 8.63 (s, 1 H, NH), 6.60 (d, 1 H, J,, 2 3.3 Hz, H-l), 5.54 (dd, 1 H, J 4 , 3 3.0, J 4 , 5 1.1 Hz, H-4), 5.41 (dd, 1 H, J 3 , 2 10.8, J 3 > 4 3.0 Hz, H-3), 5.34 (dd, 1 H, J 2 > 3 10.8, J2,i 3.3 Hz, H-2), 4.42 (ddd, 1 H, J 5 , 6 a 6.7, J5,6b 6.7, J 5 , 4 1.1 Hz, H-5), 4.15 (dd, 1 H, J 6 a , 6 b 11.3, J 6 a , 5 6.7 Hz, H-6a), 4.06 (dd, 1 H, J 6 b , 6 a 11.3, J 6 b > 5 6.7 Hz, H-6b), 2.14, 2.00, 1.99, 1.98 (s, 12 H, 4 x OAc). 2",3",4",6"-Tetra-0-acetyl-D-galactopyranosyl-j3-(l,6)-0-(l,2,2 ',3,3',4',6-hepta-O-acetyl)-a-lactose (3.32) To a solution of (3.29) (0.29 g, 0.46 mmol) and (3.31) (0.28 g, 0.58 mmol) in anhydrous dichloroethane (6 mL) under argon in the presence of 4 A molecular sieves was added boron trifluoride diethyl etherate (30 pL, 0.23 mmol) and the reaction was stirred at room temperature. After 3.5 h, pyridine (0.5 mL) was added and the solvent was evaporated in vacuo. The resulting residue was chromatographed over silica gel (PE:EtOAc, 1:1 to 2:3) to yield (3.32) (77.4 mg, 18%) as a white solid . 'H NMR (CDC13, 400 MHz): 8 6.25 (d, 1 H, Ji> 2 3.7 Hz, H-l), 5.44 (dd, 1 H, J 3 , 2 9.9, J 3 ) 4 9.6 Hz, H-3), 5.37 (d, 1 H, J 4 ,3 2.7 Hz, H-4"), 5.35 (d, 1 H, J4-,3 3.4 Hz, H-4'), 5.13 (dd, 1 H, J 2 1 0 . 4 , h;v 7.9 Hz, H-2"), 5.07 (dd, 1 H, J2,3 10.4, J 2,i 7.9 Hz, H-2'), 5.04 (dd, 1 H, J 3 > 2 10.4, J 3 - A 2.7 Hz, H-3"), 5.02 (dd, 1 H, J3,2 10.4, J 3 , 4 3.4 Hz, H-3'), 4.93 (dd, 1 H, J 2 > 3 9.9, J2,i 3.7 Hz, H-2), Chapter 4 Materials and Methods 140 4.56 (d, 1 H, J , , 2 ' 7.9 Hz, H-l"), 4.45 (d, 1 H, J, ; 2 7.9 Hz, H-l'), 4.42 (dd, 1 H, J6a,6b 11.9, J6a,5 1-7 Hz, H-6a), 4.17 (dd, 1 H, J 6 b , 6a 11.9, J 6 b , 5 6.5 Hz, H-6b), 4.04 - 4.13 (m, 2 H, H-6a", H-6b"), 3.99 (ddd, 1 H, J 5 > 4 10.3, J 5 , 6 b 6.5, J 5 , 6 a 1.7 Hz, H-5), 3.91 (ddd, 1 H, Js",6a- 6.7, J 5 j6b" 6.7, J 5 ) 4 0.5 Hz, H-5"), 3.81 (dd, 1 H, J 4 > 5 10.3, J 4 > 3 9.6 Hz, H-4), 3.71 -3.78 (m, 3H, H-5', H-6a', H-6b'), 2.16, 2.14, 2.12,2.10, 2.04,2.02, 1.98, 1.95, 1.93 (s, 33 H, 11 x OAc). D-Galactopyranosyl-j3-(l,6)-0-lactose (3.33) A solution of (3.32) (72 mg, 0.07 mmol) in anhydrous MeOH (20 mL) was made basic by the addition of a NaOMe/MeOH (14 pL of 1 M) solution. The reaction was allowed to stir at room temperature overnight before it was neutralized with acidic Amberlyte® resin. The solvent was removed under reduced pressure and the crude solid was purified by HPLC using a Tosohaas amide-80 (21.5 mm x 30 cm) column, eluting with acetonitrile/water (65:35) at a flow rate of 7 mL/min. The desired fractions were pooled and lyophilized to yield (3.33) (28.5 mg, 76%) as a white solid. 'H NMR (D20, 400 MHz) selected data only: 5 5.23 (d, 1 H, Ji a , 2 3.5 Hz, H-la), 4.66 (d, 1 H, J ] p , 2 8.4 Hz, H-lp), 4.47 (d, 1 H, J, j 2 7.9 Hz, H-l"), 4.46 (d, 1 H, J, >2- 7.9 Hz, H-l'). Anal. Calcd for C i 8 H 3 2 O i 6 1!/2H 20: C, 40.68; H, 6.35. Found: C, 40.40; H, 52.95. D-Galactopyranosyl-(5-(l,3)-0-lactose (3.34) Lactose monohydrate (0.11 g, 0.31 mmol) was dissolved in 100 mM sodium phosphate buffer (2.5 mL) at pH 6.0 when /?-nitrophenyl (3-D-galactopyranoside (0.025 g, 0.10 mmol) and 50 pL of 2.42 mg/mL p-l,3-galactosidase from X. manihotis were added. Chapter 4 Materials and Methods 141 Additional aliquots of p-nitrophenyl p-D-galactopyranoside (0.015 g, 0.06 mmol) were subsequently added every hour for the next four hours. After this time, the reaction was quenched by the addition of 2 M NaOH (450 pL). The enzyme was then removed from the reaction mixture by passage through a 10 000 Da molecular weight cutoff membrane filter. The filtrate was neutralized with acidic Amberlyte® resin and purified by HPLC using a Tosohaas amide-80 (21.5 mm x 30 cm) column, eluting with acetonitrile/water (65:35) at a flow rate of 7 mL/min. The desired fractions were pooled and lyophilized to yield (3.34) (15 mg, 10%) as a white solid. Anal. Calcd for C,8H320 1 6-2H 20: C, 40.00; H, 6.71. Found: C, 39.98; H, 6.66. To confirm that the correct linkage had formed, a small portion of (3.34) (3 mg) was acetylated by stirring in acetic anhydride (0.5 mL) and pyridine (1 mL) overnight. After the reaction was worked up in 10% v/v HQ (5 mL) and extracted with CH2CI2 (3x4 mL), the combined organic layers were dried over MgSC»4 and the solvent was evaporated in vacuo. ] H NMR (CDCI3, 400 MHz) selected data only of the cc-anomer. 5 6.25 (d, 1 H, J,, 2 3.7 Hz, H-l), 5.37 (d, 1 H, J 4 ,3 3.3 Hz, H-4"), 5.34 (dd, 1 H, J 4 , 5 3.0, J 4 ,3 2.7 Hz, H-4'), 5.22 (dd, 1 H, J 3 , 2 10.4, J 3 > 4 9.0 Hz, H-3), 5.14 (dd, 1 H, J 2 " ( 3 " 10.4, J 2 ,i 7.9 Hz, H-2"), 5.06 (dd, 1 H, J 2 ) , 8.0, J 2-, 3- 7.9 Hz, H-2'), 5.02 (dd, 1 H, J 2 , 3 10.4, J 2 ( i 3.7 Hz, H-2), 4.91 (dd, 1 H, J 3 » 2 » 10.4, J 3 » 4 » 3.5 Hz, H-3"), 4.56 (dd, 1 H, J3,2' 7.9, J3-A- 2.7 Hz, H-3'), 4.45 (d, 1 H, J , 8 . 0 Hz, H-l'), 4.44 (d, 1 H, J r • 2 -7.9 Hz, H-l"), 4.24 (dd, 1 H, J 6 a > 6 b 12.0, J 6 a , 5 4.3 Hz, H-6a), 4.14 (dd, 1 H, J 6 b ,6 a 12.0, J 6 b , 5 5.8 Hz, H-6b), 3.91 (m, 1 H, H-5"), 3.83 (dd, 1 H, J4,5 9.7, J4,3 9.0 Hz, H-4), 1.92 -2.17 (s, 33 H, 11 x OAc). Chapter 4 Materials and Methods 142 4.2.1 1,6-Anhydrogalactose (Scheme 3.7) 2,3,4,6-Tetra-O-acetyl-a-D-galactopyranosyl bromide (3.35) To a solution of (2.1) (2.22 g, 5.69 mmol) in anhydrous CH2C12 (13 mL) under an atmosphere of argon at 0 °C was added 45% w/v HBr/AcOH (4 mL). The reaction vessel was then sealed and the reaction was allowed to warm to room temperature. After stirring for 3 h, the reaction mixture was added to 30 mL of chilled aq. NaHC03 and extracted with CH2C12 (3 x 25 mL). The combined organic layers were washed with additional aliquots of aq. NaHC03 (3 x 35 mL) followed by water (35 mL) and brine (35 mL). Removal of the solvent in vacuo after drying over MgS04 yielded (3.35) (2.23 g, 95%) as a slightly yellowish gum. 'H NMR (CDCI3, 400 MHz): 8 6.69 (d, 1 H, J,, 2 3.9 Hz, H-l), 5.50 (dd, 1 H, J 4 , 3 3.2, J 4 ; 5 1.1, H-4), 5.39 (dd, 1 H, J 3 , 2 10.6, J 3 , 4 3.2 Hz, H-3), 5.04 (dd, 1 H, J2>3 10.6, J 2 ; 1 3.9 Hz, H-2), 4.47 (ddd, 1 H, J 5 , 6 a 6.6, J5,6b 6.5, J 5 , 4 1.1 Hz, H-5), 4.16 (dd, 1 H, J 6 a,6b 11.4, J6a,5 6.6 Hz, H-6a), 4.08 (dd, 1 H, J 6 b,6 a 11.4, J 6 b , 5 6.5 Hz, H-6b), 1.99, 2.04, 2.10, 2.13 (s, 12 H, 4 x OAc). 2,3,4,6-Tetra-0-acetyl-p3-D-galactopyranosylfluoride (3.36) Silver (I) fluoride (4.13 g, 32.6 mmol) was added to a solution of (3.35) (2.23 g, 5.43 mmol) in anhydrous CH 3CN (50 mL) under an atmosphere of argon and the reaction mixture was covered and stirred at room temperature for 1.5 h. The reaction mixture was then filtered through Celite® and the filtrate was evaporated in vacuo. The resulting residue was dissolved in EtOAc (60 mL) and washed with brine (3 x 30 mL). The organic layer was then dried over MgS0 4 and the solvent was removed under reduced pressure. Chromatography over silica gel (PE:Et20, 1:1) yielded (3.36) (1.35 g, 71%) as Chapter 4 Materials and Methods 143 a white solid. *H NMR (CDC13, 400 MHz): 5 5.40 (m, 1 H, H-4), 5.29 (dd, 1 H, J 3 , 2 10.0, J 3 , 4 3.2 Hz, H-3), 5.25 (dd, 1 H, Ji,F 54.6, J,, 2 7.0 Hz, H-l), 5.02 (m, 1 H, H-2), 4.6 - 4.21 (m, 2 H, H-6a, H-6b), 4.04 (m, 1 H, H-5). 1 9 F NMR (CDC13, 188 MHz): 5 -65.6 (dd, JF>i 54.6, J F > 2 9.2 Hz). Anal. Calcd for C 1 4H,9F0 9: C, 48.00; H, 5.47. Found: C, 48.24; H, 5.53. fi-1,6-Anhydrogalactose (3.37) A solution of (3.36) (0.27 g, 0.76 mmol) in anhydrous MeOH (3.5 mL) was made basic by the addition of a catalytic amount of NaOMe. The resulting reaction mixture was stirred at room temperature for 6.5 h before it was neutralized with acidic Amberlyte® resin. The crude material was then purified by silica gel chromatography to yield (3.37) (65 mg, 47%) as a white solid. ! H NMR (CD3OD, 400 MHz): 5 5.25 (s, 1 H, H-l), 4.33 (d, 1 H, J 6 a , 6 b 7.0 Hz, H-6a), 4.28 (dd, 1 H, J 5 j 6 b 6.0, J 5 ; 4 4.5 Hz, H-5), 3.95 (dd, 1 H, J 4 , 5 4.5, J 4 , 3 4.4 Hz, H-4), 3.83 (d, 1 H, J 3 , 4 4.4 Hz, H-3), 3.64 (s, 1 H, H-2), 3.53 (dd, 1 H, J6b,6a 7.0, J 6 b , 5 6.0 Hz, H-6b). 4.2.8 lsO-UDPGal (Scheme 3.8) Diphenyl (2,3,4,6-tetra-0-acetyl)-fl-!80]-a-D-galactopyranosylphosphate (3.38) To a solution of 2,3,4,6-tetra-O-acetyl-D-galactose (3.30) (0.49 g, 1.41 mmol) in anhydrous acetonitrile (2 mL) in a thick walled bomb was added 97% 180-enriched water (0.5 mL). A few beads of Amberlyte® IR-120 acidic resin were added and the chamber was flooded with argon and sealed. The reaction mixture was then heated to 105 °C for 24 h. After this time, the solvent was removed in vacuo and the residue was Chapter 4 Materials and Methods 144 chromatographed over silica gel (PE:EtOAc, 3:2 to 1:1). This l-180-labeled 2,3,4,6-tetra-O-acetyl-D-galactopyranose along with DMAP (0.25 g, 2.03 mmol) were then dissolved in anhydrous CH2CI2 and stirred at room temperature for 20 min under an atmosphere of argon. Diphenyl chlorophosphate (0.54 g, 2.03 mmol) was then added to the reaction mixture and stirring was continued for 3 h, when it was then added to a 10% v/v solution of HC1 (40 mL) and extracted with CH2CI2 (3 x 40 mL). The combined organic extracts were washed with aq. NaHCCb (35 mL) and water (3 x 35 mL), dried over MgSC»4 and the solvent was evaporated under reduced pressure. Chromatography over silica gel (PE:EtOAc, 12:7 to 3:2) yielded (3.38) (316.1 mg, 39%) as a colourless gum. 'H NMR ( C D C I 3 , 400 MHz): 5 7.15 - 7.45 (m, 10 H, Ar), 6.10 (dd, 1 H, J i , P 6.4, J , , 2 3.3 Hz, H-l), 5.47 (dd, 1 H, J 4 , 3 3.1, J 4 , 5 1.1 Hz, H-4), 5.37 (dd, 1 H, J 3 ; 2 10.9, J 3 , 4 3.1 Hz, H-3), 5.23 (ddd, 1 H, J 2 ; 3 10.9, J 2 , i 3.3, J 2 , P 3.0 Hz, H-2), 4.32 (ddd, 1 H, J 5 ) 6 a 6.6, J 5 , 6 b 6.5, J 5 , 4 1.1 Hz, H-5), 4.06 (dd, 1 H, J 6 a ,6b H-3, J 6 a , 5 6.6 Hz, H-6a), 3.91 (dd, 1 H, J 6 b j 6 a 11.3, J 6 B , 5 6.5 Hz, H-6b), 2.12, 1.97, 1.90, 1.83 (s, 12 H, 4 x OAc). 3 1P NMR (CDC13, 81 MHz, proton decoupled): 5 -13.70. LR-LSIMS: calcd. for C26H29013P: 581. Found: 581 (C 2 6 H 2 9 1 6 0,3P) , 583 (C 2 6H2 9 l 60 1 2 1 8OP), 1 6 0/ , 8 0 = 20/80. oc-D-Galactopyranosyl-[ l-18 O]-phosphate monopyridinium salt (3.39) (3.38) (0.31 g, 0.54 mmol) was dissolved in 1:1 EtOAc:MeOH (6 mL) when Pt02 (0.10 g) was added and the reaction mixture was hydrogenated at 6 atm. After 2 d, this mixture was filtered through Celite® and chromatographed over silica gel (EtOAc to EtOAc:MeOH:H20, 27:2:1 to 7:2:1). The desired fractions were pooled, concentrated and redissolved in THF (1 mL). To this was then added 2 M LiOH (2 mL) and the Chapter 4 Materials and Methods 145 reaction mixture was allowed to stir at room temperature overnight. The reaction volume was then reduced and eluted through a Bio-Rad® AG 50W-X2, 200 - 400 mesh sulfonic acid cation exchange column (pyridinium form). The desired fractions were pooled and lyophilized to yield 32 (126 mg, 69%) as a white fluffy solid. *H NMR (D20, 300 MHz): 5 7.90 - 8.70 (m, 5 H, pyridine), 5.40 (dd, 1 H, J,, P 7.0, J,, 2 3.5 Hz, H-l), 4.00 (dd, 1 H, J5>6a 6.3, J 5 , 6 b 6.3 Hz, H-5), 3.88 (d, 1 H, J 4 , 3 2.9 Hz, H-4), 3.77 (dd, 1 H, J 3 > 2 10.3, J 3 ; 4 2.9 Hz, H-3), 3.68 (ddd, 1 H, J 2 > 3 10.3, J2,i 3.5, J 2 , P 3.0 Hz, H-2), 3.53 - 3.63 (m, 2 H, H-6a, H-6b). 3 1P NMR (D20, 121 MHz, proton decoupled): 5 0.09. LR-LSIMS: calcd for C 6 H 1 2 0 9 F : 259. Found: 259 (C6Hi2 , 609P"), 261 (C 6 H 1 2 1 6 0 8 1 8 OF), 1 60/ 1 80 = 17/83. Uridine 5'-diphospho-[l "-180]-(X-D-galactopyranose, diammonium salt (3.40) To a solution of (3.39) (0.08 g, 0.24 mmol) in water (1.5 mL) was added tri-n-octylamine (0.10 mL, 0.24 mmol) and the mixture was lyophilized. The resulting residue along with UMP-morpholidate (0.18 g, 0.26 mmol) was dried over P2Os overnight. The two reagents were then dissolved in anhydrous pyridine (2.5 mL) and the reaction mixture was stirred at room temperature in the presence of 4 A molecular sieves under an argon atmosphere. An aliquot of the reaction mixture was transferred into an NMR tube containing a capillary of DMSO-d6 so that the reaction progress can be monitored by 3 I P NMR. After l i d , the mixture was filtered and the filtrate was added to 50 mM NH 4 HC0 3 . The tri-n-octylamine was extracted with Et20 (3x15 mL) and the aqueous phase was lyophilized to yield the crude product. Purification was afforded by anion exchange chromatography on a DEAE Sephacel column (26 mm x 12.5 cm, 50 - 500 mM NH 4HCQ 3, 1.5 mL/min) followed by size exclusion chromatography on a Bio-Gel Chapter 4 Materials and Methods 146 P2, extra fine column (16 mm x 55 cm, 50 mM N H 4 H C O 3 , 0.15 mL/min) using a Beckman Biosepra ProSys Workstation. The desired fractions were pooled and lyophilized to yield 3 3 (25 mg, 18%) as a white powder. 'H NMR (D20, 300 MHz): 5 7.83 (d, 1 H, J 6 ; 5 8.2 Hz, H-6), 5.30 - 5.40 (m, 2 H, H-5, H-l'), 5.53 (dd, 1 H, J r > P 7.2, Ji-,2- 3.6 Hz, H-l"), 4.23 - 4.28 (m, 2 H, H-2', H-3'), 4.07 - 4.20 (m, 3 H, H-4', H-5a', H-5b'), 4.06 (dd, 1 H, J 5» 6 a» 6.2, J 5 , 6 b 6.2 Hz, H-5"), 3.92 (d, 1 H, J4-,y 3.0 Hz, H-4"), 3.80 (dd, 1 H, J 3 > 2 10.2, J 3 ,4 3.0 Hz, H-3"), 3.68 (ddd, 1 H, J 2 ,3 10.2, h;v 3.6, J 2 ,P 3.0 Hz, H-2"), 3.55 - 3.65 (m, 2 H, H-6a", H-6b"). 3 1P NMR (D20, 121 MHz, proton decoupled): 5 -9.95 (d, IP, J P p, P a 20.1 Hz, Pp), -11.51 (d, IP, J P a , P p 20.7 Hz, Pa). HR-LSIMS: calcd. for C i 5 H 2 3 N 2 1 6 O i 6 1 8 O P 2 " : 567.0513. Found: 567.0515. LR-LSIMS: 1 6OV 1 80 = 15/85. Anal. Calcd. for C i 5 H 2 2 N 2 1 6 O i 6 1 8 O P 2 2 " • 2NH4 +: C, 29.86; H, 5.18; N, 9.29. Found: C, 30.21; H, 5.16; N, 8.97. 4.3 Enzymatic 4.3.1 Purification of LgtC Recombinant LgtC was obtained from W.W. Wakarchuk of the NRC as a cell lysate suspended in 6% polyethylene glycol. The initial purification of this enzyme was achieved through anion exchange chromatography using a Pharmacia XK26 column (2.6 x 3.5 cm) packed with Source 15Q resin. LgtC was eluted from the column after approximately 60 minutes using a linear gradient of 0 - 30% NaCl solution in 20 mM TrisHCl, pH 7.5 at a flow rate of 2 mL/min on a Beckman Biosepra Prosys Workstation. The individual fractions were analyzed by gel electrophoresis on a Pharmacia Chapter 4 Materials and Methods 147 PhastSystem using PhastGel Gradient 8-25. The gels were developed with a silver staining solution and the LgtC-containing fractions were pooled and concentrated using Centricon® filters equipped with a 10 kDa molecular mass cutoff membrane. The LgtC-containing retentate was then subjected to further purification by size exclusion chromatography using a pre-packed Sephacryl S-100 HR column (1.6 x 60 cm) from Pharmacia, which had been equilibrated with 50 mM NH4OAc, pH 7.0. The protein was eluted from the column at a flow rate of 0.5 mL/min on a Beckman Biosepra Prosys Workstation. Once again, the fractions were analyzed by gel electrophoresis and the desired ones were pooled and concentrated. 4.3.2 Enzyme kinetics using a spectrophotometric assay Steady state kinetic studies were performed using a continuous coupled assay similar to that described by Gosselin et. al. [145] in which UDP release is coupled to the oxidation of NADH (k = 340 nm, e = 6.22 mM"'cm''). Standard reaction conditions were as follows: 20 mM HEPES, pH 7.5, 0.1 % bovine serum albumin, 50 mM KCl, 5 mM MnCl2, 5 mM DTT, 0.7 mM phosphoenolpyruvate, 300 pM NADH, 375 units of pyruvate kinase and 500 units of lactate dehydrogenase. The concentrations of substrates added depended upon the experiment. In most cases, the concentration of the fixed substrate was usually held as close to saturation levels as possible (200 - 300 pM for UDPGal and 100 - 120 mM for lactose) while the concentration of the variable substrate usually ranged from 0.5 Km to 10 K m for UDPGal and 0.1 K m to 2 Km for lactose. Reactions were initiated by the addition of LgtC (0.25 - 1.0 pg/mL), bringing the final volume to 200 pL in the quartz reaction vessels. The change in absorbance at 30 °C was Chapter 4 Materials and Methods 148 monitored by means of a UNICAM 8700 UV-Vis spectrophotometer equipped with a circulating water bath. 4.3.3 Enzyme Kinetics with oc-GalF as the donor substrate Kinetic parameters for the transfer of galactose from a-galactosyl fluoride to lactose by LgtC were determined using a continuous assay in which fluoride release was monitored using an Orion ion selective fluoride electrode interfaced to a pH/ion selective meter from Fischer Scientific. For standard assays, a solution containing 100 mM HEPES (pH 7.5 or 7.0), 2.5 mM MnCl2 , 5 mM DTT and 2 mM UDP plus appropriate concentrations of lactose and a-galactosyl fluoride was incubated at 30°C. The spontaneous hydrolysis of a-galactosyl fluoride was monitored for 5 minutes before LgtC was added to initiate the reaction, giving a final assay volume of 300 pi. The reaction was then followed for 10 minutes and the reaction rate corrected for the background hydrolysis. 4.3.4 UDPGal/UDP Exchange LgtC (150 pg/mL) was incubated with 180-UDPGal (5 mM) and excess unlabeled UDP (50 mM) in the presence of 20 mM HEPES, pH 7.5, 5 mM MnCl2 and 5 mM DTT to give a total reaction volume of 600 pL. After 4 h at room temperature, the enzyme was removed by centrifugation through a Centricon® filter equipped with a 10 kDa molecular weight cutoff membrane. The filtrate was then purified by anion exchange chromatography using a Pharmacia HiTrap Q HP column (5 mL) and a Beckman Biosepra Prosys Workstation. Elution was afforded by a 50 - 500 mM gradient of NH4HCO3 at a flow rate of 1 mL/min. The peaks corresponding to UDP and UDPGal Chapter 4 Materials and Methods 149 were difficult to resolve and were collected together. After lyophilization, the sample was then analyzed by both electrospray mass spectrometry and 3 1P NMR. Analysis by electrospray mass spectrometry was performed on a single quadrupole instrument built in the laboratory of Professor Don Douglas in the department of chemistry at the University of British Columbia. An approximately 50 pM of sample (in terms of UDPGal) was prepared by dissolving the lyophilized powder in a 10% solution of water in methanol. This sample was injected into the mass spec at a flow rate of 1 pL/min. The instrument was run in negative ion mode at a voltage of 3500 V. The remainder of the lyophilized sample was dissolved in 500 pL of chelex-treated D 2 0 containing approximately 10 mM of EDTA and a proton decoupled 3 1P NMR (121 MHz) was obtained. Acquisition parameters were as follows: sweep width = 5482 Hz, acquisition time - 2 s, delay between pulses = 2 s and pulse width = 10.0 ps. 4.3.5 Trapping experiments 4.3.5.1 Incompetent acceptor substrate approach LgtC (1 mg/mL) was incubated with UDPGal (1 mM) and either 4dLac, Bn4FLac or Bn4dLac (40 mM) in a solution of 20 mM HEPES, pH 7.5, 50 mM KC1, 5 mM MnCl2 and 5 mM DTT. The total volume of the reaction mixture was 150 pL. After time intervals of 5, 10, 20, 40, 80 and 160 minutes, an aliquot of the reaction mixture (20 pL) was added to an equal volume of 6 M urea. The sample was then quickly frozen at -78 °C to await analysis by electrospray mass spectrometry. This procedure was carried out by Dr. Shouming He using a Sciex AP1-300 mass spectrometer interfaced with a Michrom UMA HPLC system. The sample was introduced through a microbore PLRP Chapter 4 Materials and Methods 150 column (1 x 50 mm) and eluted with a gradient of 20 - 100% solvent B at a flow rate of 50 mL/min over 5 min (solvent A: 0.06% TFA, 2% CH 3CN in water; solvent B: 0.05% TFA, 90% C H 3 C N in water). The mass spectrometer was scanned over a range of 400 -2000 Da with a step size of 0.5 Da and a dwell time of 1 ms. 4.3.5.2 Fluorosugar approach The procedure for this approach is essentially the same as that of the incompetent acceptor substrate approach except that UDP-2FGal (1 mM) took the place of UDPGal. The experiment was also performed with either lactose as the acceptor or with 4dLac as an incompetent acceptor. The reaction mixtures were quenched and analyzed in the same manner as previously described. 4.3.6 Evaluation of various compounds as potential intermediates of the LgtC reaction 4.3.6.1 Gal-p-Lactoside intermediates Gal-p-l,6-Lac and Gal-p-l,3-Lac at a concentration of 10 mM were each incubated with LgtC (10 pg/mL) together with the necessary components needed for enzyme activity (20 mM HEPES, pH 7.0, 50 mM KC1, 1 mg/mL BSA, 5 mM MnCl2, 5 mM DTT). Also present in the reaction mixture was an oc-galactosidase from green coffee bean (100 pg/mL) and either the presence or absence of 1 mM UDP to bring the total volume of the reaction mixture to 20 pL. The oc-galactosidase from green coffee bean specifically hydroyzes Gal-a-1,4 linkages. As such, any conversion of the Gal-p-Lac trisaccharides to the Gal-a-l,4-linked product will lead to the hydrolysis of the latter to yield lactose and galactose. The ability of LgtC to convert either Gal-p-1,6-Lac or Chapter 4 Materials and Methods 151 Gai-p-i,3-Lac to the desired Gal-a-l,4-Lac product was monitored by TLC (EtOAc:MeOH:H20, 7:2:1) for the production of this disaccharide and monosaccharide products. 4.3.6.2 Anhydrosugar intermediates Both p-l,4-anhydrogalactose and P-l,6-anhydrogalactose at a concentration of 50 mM were each incubated with UDP (2 mM), lactose (50 mM) and LgtC (0.4 mg/mL) in the presence of 30 mM Tris, pH 7.5, 70 mM KC1, 15 mM MnCl2 and 10 mM DTT. The total volume of the reaction mixtures was 210 pL. The reaction mixtures were allowed to stand at room temperature for 20 h prior to being filtered through Centricon® filters equipped with a 10 kDa molecular mass cutoff membranes. The filtrates were then analyzed for the presence of the Gal-a-l,4-Lac by TLC (EtOAc, MeOH, H 20, 7:2:1) and HPLC using a Tosohaas amide-80 (4.6 mm x 25 cm) column, eluting with acetonitrile/water (65:35) at a flow rate of 1 mL/min on a Waters 600 HPLC system equipped with a Waters 2410 refractive index detector. 4.3.7 Positional Isotope Exchange l sO-UDPGal (1.8 mg) was dissolved in 510 pL of Chelex-treated D 2 0 and the proton decoupled 3 1P NMR spectrum was recorded on a 300 MHz Bruker AV-300 spectrometer. This sample of 180-labeled UDPGal was then removed from the NMR tube and added to a solution of HEPES (20 mM), pD 7.5, 4dLac (10 mM), MnCl2 (5 mM), DTT (5 mM) and LgtC (150 pg/mL) in D 2 0 to bring the final volume to 600 pL. After incubating for 7 h at room temperature, the reaction mixture was centrifuged through a Centricon® filter equipped with a 10 kDa molecular weight cutoff membrane Chapter 4 Materials and Methods 152 to remove the enzyme. The l 80-UDPGal was then re-isolated by anion exchange chromatography using a Pharmacia HiTrap Q HP column (5 mL) and a Beckman Biosepra Prosys Workstation. The compound was eluted with a linear gradient of 50 -500 mM NH4HCO3 at a flow rate of 1 mL/min. After the desired fractions were pooled and lyophilized, this compound was then redissolved in chelex-treated D2O (600 pL) containing EDTA (10 mM) and the proton decoupled 3 1P NMR spectrum was again recorded. Acquisition parameters were as follows: sweep width = 5482 Hz, acquisition time = 2 s, delay between pulses = 2 s and pulse width = 6.8 ps. APPENDIX A FUNDAMENTALS OF ENZYME KINETICS Appendix A Fundamentals of Enzyme Kinetics APPENDIX A - FUNDAMENTALS OF ENZYME 154 KINETICS A-l A Historical Perspective The simplest reaction catalyzed by an enzyme is the conversion of a single substrate into a single product. The rate of such a reaction shows a characteristic dependence on the concentration of the substrate. At low substrate concentrations, the initial velocity is directly proportional to the product of enzyme and substrate concentrations (ie. v = A:[E][S]). At high concentrations of substrate, the initial velocity becomes dependent on enzyme alone (ie. v = k[E]). This biphasic dependence of the rate on substrate concentration was first observed by Brown and Henri in 1902, leading to the proposal of an intermediate complex between enzyme and substrate. Building upon this model, Michaelis and Menten, in 1913, proceeded to develop the first satisfactory mathematical account of the relationship between the rate of catalysis and substrate concentration. Together with the introduction of the steady state concept by Briggs and Haldane in 1925, a mathematical model for describing the behaviour of enzyme catalysis was born. A-2 Basic Equations The general scheme for an enzyme-catalyzed conversion of a single substrate into a single product is illustrated by the following equation, Appendix A Fundamentals of Enzyme Kinetics 155 k l _ k 2 E + S ^ * " = E S • E + P k - i where E is the enzyme, S is the substrate, ES is the enzyme-substrate (Michaelis) complex and P is the product. Under steady state conditions, the rate of change in the concentration of the ES complex is equal to zero and is described by Equation A. 1. ^p- = UE][S] -UES] -k2[ES] =0 ( E f i u a t i o n A 1 ) While the individual concentrations of the enzyme, both in its free form (E) and in complex with the substrate (ES) are unknown, the total amount of enzyme (E0) added to the reaction is quantifiable and is equal to the sum of the concentrations of both free and bound enzyme forms (Equation A.2). [E]o = [E] + [ES] (Equation A.2) Solving for [ES] in Equations A. l and A.2, [ES] = [E],[S] (Equation A.3) [S] k-l+k2 Assuming that the rate limiting step is the breakdown of the ES complex to yield product, then the rate of the reaction (v) can be described by Equation A.4. V=^P~ = k2[ES] (Equation A.4) ot By substituting the expression for [ES] from Equation A.3 into Equation A.4, the result is (Equation A.5) If we proceed to define the Michaelis constant, K m , as the ratio of the rate constants IS] Appendix A Fundamentals of Enzyme Kinetics 156 (k.]+k2)lki and the turnover number, kca! as the rate constant k?, then Equation A.5 can be simplified to the more general form known as the Michaelis-Menten equation (Equation A.6). In this equation, the product of kcat and [E]0 is also referred to as Vnax. This equation describes a rectangular hyperbola such as is shown in Figure A. 1. A close inspection of this equation also reveals that when the initial rate of the reaction is equal to one-half the maximal velocity (ie. v = (kcat\E\J2)), the value of K m becomes equal to the concentration of the substrate. As such, K m is often defined as the concentration at which the initial velocity is half-maximal. Additionally, when the value of fe is much smaller than that of k.i, the value of K m also becomes a good approximation of the dissociation of the substrate from the enzyme. V = K,AS] (Equation A.6) v = V n B X [S] /K n l v=k c a t|E| 0=V, max v r = V, 2 K, in (Substrate) Figure A. 1. A plot showing the typical hyperbolic nature of the Michaelis-Menten equation. Appendix A Fundamentals of Enzyme Kinetics 157 For ease of plotting and subsequent extraction of the kinetic parameters prior to the advent of computer, the Michaelis-Menten equation (Equation A . 6 ) was often rearranged into the form of a linear equation (Equation A . 7 ) . J _ _ Km . _ J _ + 1 (Equation A . 7 ) v v rsn v V V max |_*-» J r max A plot of the reciprocal of the initial velocity as a function of the reciprocal of the substrate concentration (also referred to as the Lineweaver-Burk plot) would therefore yield a straight line (Figure A . 2 ) . The value of K n / V m a x could then be obtained from the slope of the line while the y-intercept would provide the value for 1 / V m a x - The importance of the double reciprocal plot will be more apparent when the kinetics of enzyme inhibition as well as multisubstrate reactions are discussed. Figure A .2 . A typical double reciprocal (Lineweaver-Burk) plot from which the important kinetic parameters of an enzyme catalyzed reaction can be extracted. Appendix A Fundamentals of Enzyme Kinetics 158 A-3 Reversible Inhibition Many substances can combine with an enzyme in a reversible manner to influence the activity of the enzyme. Those that cause a decrease in the activity are referred to as inhibitors. These compounds can be classified as competitive, uncompetitive or mixed inhibitors depending on how they bind the enzyme. L Competitive inhibition A competitive inhibitor is a compound that competes directly with a normal substrate for binding to the enzyme active site. These compounds often resemble the substrate but are unreactive towards the enzyme. In the presence of such a compound, the simple enzyme-catalyzed reaction must therefore be expanded to include a second equilibrium, Kj, which describes the dissociation of the inhibitor from the enzyme-inhibitor complex (Kj = [E][I]/[EI]). k, k, E + S w = ES • E + P k., + I EI The total concentration of enzyme is now given by [E]=[E] +[ES] +[EI] (Equation A.8) Applying the steady state approach (d[ES]/dt = 0) and substituting into Equation A.6 gives Equation A.9. Appendix A Fundamentals of Enzyme Kinetics 159 VJ.S] (Equation A.9) V = K.1 + +[S] As can be seen from Equation A.9, a competitive inhibitor affects only the K m term of the Michaelis-Menten equation, increasing it by a factor of (1 + [I]/Kj). The value of Vmax is unaffected since at high concentrations of substrate, the inhibitor is displaced from the enzyme. A double reciprocal plot of Equation A.9 for different fixed concentrations of inhibitor will yield a family of lines that intersect on the y-axis (Figure A.3). Figure A.3. The pattern observed in the double reciprocal plot for a series of different concentrations ofa competitive inhibitor. ii. Uncompetitive Inhibition An uncompetitive inhibitor is a compound that binds to the enzyme-substrate complex but not to the free enzyme. Increasing [I] l/[Substrate] Appendix A Fundamentals of Enzyme Kinetics 160 E + S ES + I E + P ESI The kinetic description for this type of inhibition is described by Equation A. 10. Vn 1 + v LZ1 K, [S] (Equation A. 10) K„ 1 + [S] In such a scenario, both the values of V m a x and K m are decreased by a factor of (1 + [I]/Kj). A double reciprocal plot of Equation A. 10 for different fixed concentrations of inhibitor will yield a family of parallel lines (Figure A.4). Figure A.4. The pattern observed in the double reciprocal plot for a series of different concentrations of an uncompetitive inhibitor. Appendix A Fundamentals of Enzyme Kinetics 161 UL Mixed/Noncompetitive Inhibitor A mixed or noncompetitive inhibitor is a compound that can bind to both the free enzyme and also the enzyme-substrate complex. E + S + I ES + I E + P EI ESI The Michaelis-Menten equation for mixed/noncompetitive inhibition is defined by Equation A. 11. F,™ [S] (Equation A . I I ) K„ 1 + K, 1 + ui From Equation A . I I , it can be seen that the name mixed inhibition arises from the fact that the denominator has the factor (1 + [I]/Kj) multiplying the K m term as in competitive inhibition and the factor (1 + [I]/Kj') multiplying the [S] term as in uncompetitive inhibition. Mixed inhibitors are therefore effective at both high and low substrate concentrations. A double reciprocal plot of Equation A . I I for different fixed concentrations of inhibitor will yield a family of lines intersecting to the left of the 1/v axis (Figure A.5a). For the special case in which Kj = K j ' , the intersection point will also occur on the 1 /[Substrate] axis (Figure A.5b) Appendix A Fundamentals of Enzyme Kinetics 162 Increasing [I] l/|Substrate| l/|Substratel Figure A.5. The pattern observed in the double reciprocal plot for a series of different concentrations of a mixed/noncompetitive inhibitor where (a) K; ^ Ki and where (b) Kj = Kj'. A-4 Bisubstrate Reactions Enzymes that catalyze reactions involving two substrates and yielding two products account for approximately 60% of known biochemical reactions. In most cases, the reactions catalyzed are either the transfer of a specific functional group from one substrate to the other or oxidation/reduction reactions in which reducing equivalents are transferred between the two substrates. Mathematical treatment of the various modes of substrate binding and product release were worked out by King and Altaian, Alberty, Dalziel, and Frieden in the late 1950's; and by Hanes and Wong, and Cleland in the early 1960's. More qualitative methods, which used the double reciprocal plot as a tool for distinguishing the various possible mechanism were also worked out by Cleland. L Terminology a. Substrates are designated by the letters A, B, C and D in the order that they add to the enzyme. Appendix A Fundamentals of Enzyme Kinetics 163 b. Products are designated P, Q, R and S in the order that they are released from the enzyme. c. Stable enzyme forms are designated E, F and G with E being the free enzyme, if such distinctions can be made. A stable enzyme form is defined as one that by itself is incapable of converting to another stable enzyme form. d. The number of reactants and products in a given reaction are specified, in order, by the terms UNI (one), Bl (two), TER (three), and QUAD (four). it Types of Bl Bl Reactions Enzyme-catalyzed group-transfer reactions fall under two major mechanistic classifications: a. Sequential reactions Reactions in which all substrates must combine with the enzyme before a chemical reaction can occur and products be released are known as Sequential reactions. These reactions can be further subclassified into those with a compulsory order of substrate addition (Ordered mechanism) and those that have no preference for the order of substrate addition (Random mechanism). In the Ordered mechanism, the binding of the first substrate is apparently required for the enzyme to form the binding site for the second substrate, whereas for the Random mechanism, both binding sites are present on the free enzyme. Appendix A Fundamentals of Enzyme Kinetics 164 b. Ping Pong reactions Mechanisms in which one or more products are released before all substrates have been added to the enzyme are known as Ping Pong reactions. In these reactions, a functional group from the first substrate is transferred onto the enzyme, yielding the first product and a stable enzyme that has been modified by the functional group. In the second stage of the reaction, this functional group is then displaced from the enzyme by the second substrate to yield the second product, regenerating the original form of the enzyme in the process. APPENDIX B GRAPHICAL REPRESENTATION OF DATA Appendix B Graphical Representation of Data APPENDIX B - GRAPHICAL REPRESENTATION OF KINETIC DATA 166 0 30 60 90 120 [lactose] (mM) 0 20 40 60 [allyl lactoside] (mM) 0 20 40 60 [2,3-dihydroxypropyl lactoside] (mM) i i i i i i i i i i i i i i i i i i i 0 5 10 15 20 [pentenyl lactoside] 0 20 40 60 [benzyl lactoside] (mM) 0 50 100 150 200 250 [galactose] (mM) Figure B . l . Michaelis-Menten plots for the transfer of galactose from UDPGal (300 uM) to various acceptors by LgtC. Appendix B Graphical Representation of Data 167 l—-i 1 1 r [UDP-Gal] (u.M) Figure B.2. A plot of rate as a function of [UDPGal] for the LgtC-catalyzed turnover of UDPGal in the absence of any glycosyl acceptors. Figure B.3. A plot of rate as a function of [UDPGlc] for the LgtC-catalyzed transfer of glucose to lactose (100 mM). Appendix B Graphical Representation of Data 168 i i i i I i i i i I i i i i I i i i i I i i 1/[lactose] (mM)-1 Figure B.4. Double reciprocal plot for the inhibition of LgtC by 4'-deoxylactose (4dLac, (3.6)) at varying concentrations of lactose and a fixed (300 (J.M) concentration of UDPGal. The concentrations of 4dLac were 3.75 (•), 7.5 (O ), 15 (•) and 30 mM (•). 1/[lactose] (mM) 1 Figure B.5. Double reciprocal plot for the inhibition of LgtC by benzyl 4'-deoxylactoside (Bn4dLac, (3.17)) at varying concentrations of lactose and a fixed (300 u,M) concentration of UDPGal. The concentrations of Bn4dLac were 0 (•), 5 (O ) and 20 mM (•). Appendix B Graphical Representation of Data 169 1/[lactose] (mM)- 1 Figure B.6. Double reciprocal plot for the inhibition of LgtC by benzyl 4'-deoxy-4'-fluorolactoside (Bn4FLac, (3.13)) at varying concentrations of lactose and a fixed (300 pM) concentration of UDPGal. The concentrations of Bn4FLac were 0 (•), 1.5 (O ), 3 (•), 6 (•) and 12mM(A). [lactose] (mM) [UDP-Gal] (pM) Figure B.7. Michaelis-Menten plots of the two substrates for the Glnl89Ala mutant of LgtC. When lactose was the variable substrate, the concentration of UDPGal was held constant at 250 pM. When UDPGal was the variable substrate, the concentration of lactose was held constant at 100 mM. Appendix B Graphical Representation of Data 170 T — i — i — i — i — r — i — i — i — r [lactose] (mM) [UDPGal] (pM) Figure B.8. Michaelis-Menten plots of the two substrates for the Glnl89Glu mutant of LgtC. When lactose was the variable substrate, the concentration of UDPGal was held constant at 250 pM. When UDPGal was the variable substrate, the concentration of lactose was held constant at 100 mM. 0 20 40 60 80 100 0 100 200 300 400 [lactose] (mM) [UDPGal] (pM) Figure B.9. Michaelis-Menten plots of the two substrates for the Glnl89Asn mutant of LgtC. When lactose was the variable substrate, the concentration of UDPGal was held constant at 250 pM. When UDPGal was the variable substrate, the concentration of lactose was held constant at 100 mM. 171 REFERENCES [I] Varki, A., Glycobiology, 3 (1993) 97-130. [2] Henrissat, B., Bairoch, A., Biochem. J., 316 (1996) 695-696. [3] Campbell, J.A., Davies, G.J., Bulone, V. , Henrissat, B., Biochem. J., 326 (1997) 929-942. [4] Catterall, C.F., Lyons, A., Sim, R.B., Day, A.J., Harris, T.J., Biochem. J., 242 (1987) 849-856. [5] Kunnath-Muglia, L . M . , Chang, g.H., Sim, R.B., Day, A.J., Ezekowitz, R.A., Mol. Immunol., 30 (1993) 1249-1256. [6] Hobart, M . I , Fernie, B.A., DiScipio, R.G., J. Immunol., 154 (1995) 5188-5194. [7] Koshland, D.E. , Biol. Rev., 28 (1953) 416-436. [8] Ly, H.D., Withers, S.G., Annu. Rev. Biochem., 68 (1999) 487-522. [9] Knapp, S., Vocadlo, D., Gao, Z., Kirk, B., Lou, J., Withers, S.G., J. Am. Chem. Soc, 118 (1996)6804-6805. [10] van Scheltinga, A.C.T. , Armand, S., Kalk, K.H. , Isogai, A., Henrissat, B., Dijkstra, B.W., Biochemistry, 34(1995) 15619-15623. [II] MacLeod, A . M . , Tull, D., Rupitz, K., Warren, R.A.J., Withers, S.G., Biochemistry, 35 (1996) 13165-13172. [12] Cupples, C.G. , Miller, J.H., Huber, R.E., J. Biol. Chem., 265 (1990) 5512-5518. [13] Planas, A., Juncosa, M . , Lloberas, J., Querol, E. , FEBS, 308 (1992) 141-145. [14] McCarter, J.D., Withers, S.G., Curr. Opin. Struct. Biol., 4 (1994) 885-892. [15] Bourne, Y. , Henrissat, B., Curr. Opin. Struct. Biol, 11 (2001) 593-600. [16] Withers, S.G., Rupitz, K., Street, I.P., J. Biol. Chem., 263 (1988) 7929-7932. [17] McCarter, J.D., Withers, S.G., J. Am. Chem. Soc, 118 (1996) 241-242. 172 [18] Zechel, D.L. , Withers, S.G. (1999) in Comprehensive Natural Products Chemistry (Poulter, C. D., Ed.) pp 279-314, Elsevier, New York. [19] Rye, C.S., Withers, S.G., Curr. Opin. Chem. Biol, 4 (2000) 573-580. [20] Sinnott, M.L. , Souchard, I.J., Biochem. J., 133 (1973) 89-98. [21] Umezurike, G.M. , Biochem. J., 254 (1988) 73-76. [22] Kempton, I.B., Withers, S.G., Biochemistry, 31 (1992) 9961-9969. [23] Tull, D., Withers, S.G., Biochemistry, 33 (1994) 6363-6370. [24] Namchuk, M.N. , McCarter, J.D., Becalski, A., Andrews, T., Withers, S.G., J. Am. Chem. Soc, 122 (2000) 1270-1277. [25] Pauling, L . , Chem. Eng. News, 24 (1946) 1375-1377. [26] Lienhard, G.E. , Science, 180 (1973) 149-154. [27] Wolfenden, R., Annu. Rev. Biophys. Bioeng., 5 (1976) 271-306. [28] Mader, M . M . , Bartlett, P.A., Chem. Rev., 97 (1997) 1281-1301. [29] Legler, G., Adv. Curb. Chem. Biochem., 48 (1990) 319-385. [30] Withers, S.G., Namchuk, M . , Mosi, R. (1999) (Stutz, A. E . , Ed.) pp 188-203, Wiley-V C H , Weinheim. [31] Bols, M.,Acc. Chem. Res., 31 (1998) 1-8. [32] Lundt, I., Madsen, R. (1999) (Stutz, A. E. , Ed.) pp 112-123, Wiley-VCH, Weinheim. [33] Hoos, R., Naughton, A.B., Thiel, W., Vasella, A. , Weber, W., Rupitz, K., Withers, S.G., Helv. Chim. Acta., 76 (1993) 2666-2686. [34] Hoos, R., Vasella, A., Rupitz, K., Withers, S.G., Carbohydr. Res., 298 (1997) 291-298. [35] Papandreou, g., Tong, M.K. , Ganem, G.,J. Am. Chem. Soc, 115 (1993) 11682-11690. [36] Legler, G., Finken, M.T., Carbohydr. Res., 292 (1996) 103-115. 173 [37] Pan, Y.T. , Kaushal, G.P., Papandreou, G., Ganem, B., Elbein, A.D. , J. Biol. Chem., 267 (1992) 8313-8318. [38] Heightman, T.D., Vasella, A.T. , Angew. Chem. Int. Ed., 38 (1999) 750-770. [39] Blake, C.C.F. , Koenig, D.F., Mair, G.A., North, A.C.T. , Phillips, D . C , Sarma, V.R., Nature, 206 (1965) 757-761. [40] Phillips, D . C , Proc. Natl. Acad. Sci. U.S.A., 57 (1967) 484-495. [41] Blake, C.C.F. , Johnson, L .N. , Mair, G.A., North, A.C.T. , Phillips, D . C , Sarma, V.R., Proc. Roy. Soc. B., 167 (1967) 378-388. [42] Amyes, T.L. , Jencks, W.P., J. Am. Chem. Soc, 111 (1989) 7888-7900. [43] Sinnott, M.L. , Withers, S.G., Biochem. J., 143 (1974) 751-763. [44] Tao, B.Y., Reilly, P.J., Robyt, J.F., Biochim. Biophys. Acta, 995 (1989) 214-220. [45] Legler, G., Roeser, K.R., Illig, H.K., Eur. J. Biochem., 101 (1979) 85-92. [46] Voet, J.G., Abeles, R.H., J. Biol. Chem., 245 (1972) 1020-1031. [47] Mieyal, J.J., Simon, M . , Abeles, R.H., J. Biol. Chem., 247 (1972) 532-542. [48] Davies, G.J., Mackenzie, L.F. , Varrot, A., Dauter, M. , Brzozowski, A . M . , Schulein, M . , Withers, S.G., Biochemistry, 37 (1998) 11707-11713. [49] White, A., Tull, D., Johns, K., Withers, S.G., Rose, D.K., Nature Struct. Biol., 3 (1996) 149-154. [50] Sidhu, G., Withers, S.G., Nguyen, N.T., Mcintosh, L.P., Ziser, L. , Brayer, G.D., Biochemistry, 38 (1999) 5346-5354. [51] Notenboom, V., Birsan, C , Warren, R.A.J., Withers, S.G., Rose, D.R., Biochemistry, 37 (1998) 4751-4758. [52] Notenboom, V., Birsan, C , Nitz, M . , Rose, D.R., Wan-en, R.A.J., Withers, S.G., Nat. Struct. Biol., 5 (1998) 812-818. 174 [53] Vocadlo, D.J., Davies, G.J., Laine, R., Withers, S.G., Nature, 412 (2001) 835-838. [54] Seto, N.O.L., Palcic, M . M . , Compston, C.A., Li , H. , Bundle, D.R., Narang, S.A., J. Biol. Chem., 272 (1997) 14133-14138. [55] Seto, N.O.L., Compston, C.A., Evans, S.V., Bundle, D.R., Narang, S.A., Palcic, M . M . , Eur. J. Biochem., 259 (1999) 770-775. [56] Legault, D.J., Kelly, R.J., Natsuka, Y. , Lowe, J.B., J. Biol. Chem., 270 (1995) 20987-20996. [57] Xu, Z., Vo, L. , Macher, B.A., J. Biol. Chem., 271 (1996) 8818-8823. [58] Dupuy, F., Petit, J.M., Mollicone, R., Oriol, R., Julien, R., Maftah, A., J. Biol. Chem., 274(1999) 12257-12262. [59] Wiggins, C.A.R., Munro, S., Proc. Natl. Acad. Sci. USA, 95 (1998) 7945-7950. [60] Breton, C , Bettler, E . , Joziasse, D.H., Geremia, R.A., Imberty, A., J. Biochem., 123 (1998) 1000-1009. [61] Busch, C , Hofmann, F., Selzer, J., Munro, S., Jeckel, D., Aktories, K., J. Biol. Chem., 273 (1998) 19566-19572. [62] Zhang, Y. , Wang, P.G., Brew, K., J. Biol. Chem., 276 (2001) 11567-11574. [63] Ihara, H., Ikeda, Y. , Koyota, S., Endo, T., Honke, K., Taniguchi, N., Eur. J. Biochem., 269 (2002) 193-201. [64] Gastinel, L.N. , Cambillau, C , Bourne, Y. , EMBOJ., 18 (1999) 3546-3557. [65] Charnock, S.J., Davies, G.J., Biochemistry, 38 (1999) 6380-6385. [66] Ha, S., Walker, D., Shi, Y. , Walker, S., Protein Sci., 9 (2000) 1045-1052. [67] Pedersen, L.C. , Tsuchida, K., Kitagawa, H. , Sugahara, K.., Darden, T.A., Negishi, M. , J. Biol. Chem., 275 (2000) 34580-34585. 175 [68] Onligil, U.M. , Zhou, S., Yuwaraj, S., Sarkar, M . , Schachter, H., Rini, J .M., EMBO J., 19 (2000) 5269-5280. [69] Persson, K., Ly, H.D., Dieckelmann, M. , Wakarchuk, W.W., Withers, S.G., Strynadka, N.C.J., Nat. Struct. Biol., 8 (2001) 166-175. [70] Gastinel, L.N. , Bignon, C., Misra, A.K. , Hindsgaul, O., Shaper, J.H., Joziasse, D.H., EMBOJ., 20 (2001) 638-649. [71] Mulichak, A . M . , Losey, H.C., Walsh, C.T., Garavito, R.M. , Structure, 9 (2001) 547-557. [72] Boix, E . , Swaminathan, G.J., Zhang, Y., Natesh, R., Brew, K., Acharya, K.R., J. Biol. Chem., 276 (2001) 48608-48614. [73] Ramakrishnan, B., Balaji, P.V., Qasba, P.K., J. Mol. Biol., 318 (2002) 491-502. [74] Vrielink, A., Ruger, W., Driessen, H.P.C., Freemont, P.S., EMBO J., 13 (1994) 3413-3422. [75] Kim, S .C , Singh, A.N. , Raushel, F .M. , J. Biol. Chem., 263 (1988) 10151-10154. [76] Kim, S .C , Singh, A.N. , Raushel, F .M. , Arch, Biochem. Biophys., 267 (1988) 54-59. [77] Singh, A.N. , Hester, L.S., Raushel, F .M. , J. Biol. Chem., 262 (1987) 2554-2557. [78] Ichikawa, Y. , Lin, Y . C . , Dumas, D.P., Shen, G.J., Garcia-Junceda, E . , Williams, M.A., Bayer, R., Ketcham, C , Walker, L . E . , Paulson, J.C., Wong, C H . , J. Am. Chem. Soc, 114 (1992) 9283-9298. [79] Wakarchuk, W., Cunningham, A., Watson, D., Young, M. , Prot. Engineering, 11 (1998) 295-302. [80] Cleland, W.W., Biochim. Biophys. Acta., 61 (1963) 104-137. [81] Cleland, W.W. (1970) in The Enzymes (Boyer, P. D., Ed.) pp 1-65, Academic Press. [82] Lougheed, B. (1998) in Department of Chemistry, University of British Columbia, Vancouver. 776' [83] Koenigs, W., Knorr, E . , Ber. Dtsch. Chem. Ges., 34 (1901) 957. [84] Zemplen, G., Pacsu, E. , Ber. Dtsch. Chem. Ges., 62 (1929) 1613-1614. [85] Weber, D J . , Bhatnagar, S.K., Bullions, L .C. , Bessman, M J . , Mildvan, A.S., J. Biol. Chem., 267 (1992) 16939-16942. [86] O'Handley, S.F., Frick, D.N., Bullions, L .C. , Mildvan, A.S., Bessman, M J . , J. Biol. Chem., 271 (1996) 24649-24654. [87] Just, I., Wilm, M . , Selzer, J., Rex, G., von Eichel-Streiber, C , Mann, M . , Aktories, K., J. Biol. Chem., 270(1995) 13932-13936. [88] Ciesla, W.P.J., Bobak, D.A., J. Biol. Chem., 273 (1998) 16021-16026. [89] Morera, S., Imberty, A., Aschke-Sonnenborn, U., Riiger, W., Freemont, P.S., J. Mol. Biol., 292(1999)717-730. [90] Hayashi, M . , Hashimoto, S., Noyori, R., Chem. Lett., (1984) 1747-1750. [91] Segel, I.H. (1993) Enzyme Kinetics - Behaviour and analysis of rapid equilibrium and steady-state enzyme systems, John Wiley & Sons Inc., New York. [92] Piszkiewicz, D. (1977) Kinetics of chemical and enzyme-catalyzed reactions, Oxford University Press, New York. [93] Bell, J.E., Beyer, T.A., Hill, R.L., J. Biol. Chem., 251 (1976) 3003-3013. [94] Qiao, L. , Murray, B.W., Shimazaki, M . , Schultz, J., Wong, C.H. , J. Am. Chem. Soc, 118 (1996)7653-7662. [95] Nishikawa, Y. , Pegg, W., Paulsen, H., Schachter, H. , J. Biol. Chem., 263 (1988) 8270-8281. [96] Gold, A . M . , Biochemistry, 19(1980)3766-3772. [97] Salsas, E . , Larner, J., J. Biol. Chem., 250 (1975) 3471-3475. [98] Plesner, L. , W., P.I., Esmann, V., J. Biol. Chem., 249 (1974) 1119-1125. 777 99] Kamath, V.P., Seto, N.O.L., Compston, C.A., Hindsgaul, O., Palcic, M . M . , GlycoconjugateJ., 16 (1999) 599-606. 100] Lindhorst, T.K., Braun, C , Withers, S.G., Carbohydr. Res., 268 (1995) 93-106. 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 McCarter, J.D., Withers, S.G., J. Biol. Chem., 271 (1996) 6889-6894. Howard, S., He, S., Withers, S.G., J. Biol. Chem., 273 (1998) 2067-2072. Ly, H.D., Howard, S., Shum, K., He, S., Zhu, A., Withers, S.G., Carbohydr. Res., 329 (2000) 539-547. Ferrier, R J . , Tyler, P.,J. Chem. Soc, Perkin Trans. 1, (1980) 1528-1534. Praly, J.P., Descotes, G., Tetrahedron Lett., 28 (1987) 1405-1408. Somsak, L., Ferrier, R.S.,Adv. Carbohydr. Chem. Biochem., 49 (1991) 37-92. Wong, A.W. (personal communication). MacDonald, D.L. , J. Org. Chem., 27 (1962) 1107-1109. MacDonald, D.L. , Carbohydr. Res., 3 (1966) 117. Wittmann, V. , Wong, C.H.,y. Org. Chem., 62 (1997) 2144-2147. Hanessian, S., Plessas, N.R., J. Org. Chem., 34 (1969) 1035-1044. Hanessian, S., Plessas, N.R., J. Org. Chem., 34 (1969) 1045-1053. Wong-Madden, S T . , Landry, D., Glycobiology, 5 (1995) 19-28. Taron, C.H. , Benner, J.S., Hornstra, L.J., Guthrie, E.P., Glycobiology, 5 (1995) 603-610. Cleland, W.W., Biochemistry, 29 (1990) 3194-3197. Davies, G. , Withers, S.G., Sinnott, M.L. (1997) in Comprehensive Biological Catalysis (Sinnott, M . L. , Ed.) pp 119-208, Academic Press, London. Zechel, D.L. , Withers, S.G., Acc. Chem. Res., 33 (2000) 11-18. 776? II18] Helmreich, E.J.M., BioFactors, 3 (1992) 159-172. 119] Palm, D., Klein, H.W., Schinzel, R., Buehner, M. , Helmreich, E.J.M., Biochemistry, 29 (1990) 1099-1107. [120] Klein, H.W., Im, M.J., Palm, D., Helmreich, E.J.M., Biochemistry, 23 (1984) 5853-5861. [121] Klein, H.W., Palm, D., Helmreich, E.J.M., Biochemistry, 21 (1982) 6675-6684. [122] Withers, S.G., Shechosky, S., Madsen, N.B., Biochem. Biophys. Res. Comm., 108 (1982) 322-328. [123] Stirtan, W.G., Withers, S.G., Biochemistry, 35 (1996) 15057-15064. [124] Withers, S.G., Madsen, N.B., Sykes, B.D., Takagi, M . , Shimomura, S., Fukui, T., J. Biol. Chem., 256(1981) 10759-10762. [125] Mitchell, E.P., Withers, S., Ermert, P., Vasella, A., Garman, E . , Oikonomakos, N., Johnson, L . , Biochemistry, 35 (1996) 7341-7355. [126] Watson, K.A. , McCleverty, C , Geremia, S., Cottaz, S., Driguez, H., Johnson, L.N. , EMBO J., 18(1999)4619-4632. [127] Artymuik, P.J., Rice, D.W., Poirrette, A.R., Willett, P., Nat. Struct. Biol., 2 (1995) 117-120. [128] Holm, L. , Sander, C., EMBOJ., 14 (1995) 1287-1293. [129] Micheel, F., Klemer, A., Baum, G., Ristic, P., Zumbulte, F., Chem. Ber., 88 (1955) 475-479. [130] Micheel, F., Klemer, A. (1961) in Adv. Carh. Chem. pp 85-103. [131] Sujino, K., Uchiyama, T., Hindsgaul, O., Seto, N.O.L., Wakarchuk, W.W., Palcic, M . M . , J. Am. Chem. Soc, 122 (2000) 1261-1269. [132] Villafranca, J.J., Methods in Enzymology, 177(1989)390-403. [133] Mash, E.A. , Gurria, G .M. , Poulter, C.D., J. Am. Chem. Soc, 103 (1981) 3926-3927. 179 [134] Hilscher, L.W., Hanson, C D . , Russell, D.H., Raushel, F., Biochemistry, 24 (1985) 5888-5893. [135] Croteau, R.B., Shaskus, J.J., Renstrom, B., Felton, N.M. , Cane, D.E. , Saito, A., Chang, C , Biochemistry, 24 (1985) 7077-7085. [136] Cowdrey, W.A., Hughes, E.D., Ingold, C.K., Masterman, S., Scott, A.D. , J. Chem. Soc, (1937)1252-1271. [137] Hughes, E.D., Ingold, C.K., Whitfield, I .C, Nature, 147 (1941) 206-207. [138] Lewis, E.S., Boozer, C.E. , J. Am. Chem. Soc, 74 (1952) 308-311. [139] Woodward, R.B., Hoffmann, R., Angew. Chem., Int. Ed. Engl, 8 (1969) 781-932. [140] Boozer, C.E. , Lewis, E.S., J. Am. Chem. Soc, 75 (1953) 3182-3186. [141] Cram, D.J., J. Am. Chem. Soc, 75 (1953) 332-338. [142] Schreiner, P.R., von Rague Schleyer, P., Hill, R.K., J. Org. Chem., 58 (1993) 2822-2829. [143] Sinnott, M.L . , Jencks, W.P., J. Am. Chem. Soc, 102 (1980) 2026-2032. [144] Klein, H.W., Im, M.J., Palm, D., Eur. J. Biochem., 157 (1986) 107-114. [145] Gosselin, S., Alhussaini, M. , Streiff, M.B., Takabayashi, K., Palcic, M . M . , Anal. Biochem., 220 (1994) 92-97. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0061316/manifest

Comment

Related Items