Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Deoxy and deoxyfluoro glycosides as mechanistic probes of Escherichia coli (lacZ) B-galactosidase McCarter, John D 1991

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

Item Metadata

Download

Media
831-UBC_1991_A6_7 M32_5.pdf [ 5.09MB ]
Metadata
JSON: 831-1.0059772.json
JSON-LD: 831-1.0059772-ld.json
RDF/XML (Pretty): 831-1.0059772-rdf.xml
RDF/JSON: 831-1.0059772-rdf.json
Turtle: 831-1.0059772-turtle.txt
N-Triples: 831-1.0059772-rdf-ntriples.txt
Original Record: 831-1.0059772-source.json
Full Text
831-1.0059772-fulltext.txt
Citation
831-1.0059772.ris

Full Text

DEOXY AND DEOXYFLUORO GLYCOSIDES AS MECHANISTIC PROBES OF Escherichia coli (lacZ) (3-GALACTOSIDASE By JOHN D. McCARTER B.Sc, The University of Victoria, 1988 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Chemistry) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA May 1991 ©John D. McCarter, 1991 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada Date DE-6 (2/88) ABSTRACT The mechanism of galactoside hydrolysis by Escherichia coli (lacZ) (5-galactosidase was probed with a series of 2',4' dinitrophenyl deoxy and deoxyfluoro glycopyranosides. A two-step mechanism has been proposed previously for this process: 1) cleavage of the glycosidic bond and formation of a covalent galactosyl-enzyme intermediate ("galactosylation"); 2) hydrolysis of the mtermediate to give free enzyme and galactose ("degalactyosylation"). A series of deoxy and deoxyfluoro analogs of 2',4'-dinitrophenyl-p,-D-galactopyranoside was prepared for this study. The 2-deoxyfluoro derivative was found to be an effective mechanism-based inactivator of E. coli (lacZ) P-galactosidase. This compound thus joins a class of 2-deoxy-2-fluoro glycosides which inactivate glycosidases of this type by the accumulation of a stable glycosyl-enzyme intermediate. The active site-directed nature of this inhibitor was shown by protection against inactivation by a competitive ligand and the near 1:1 stoichiometry of dinitrophenolate release with enzyme inactivation. Furthermore, when freed from excess inactivator the 2-deoxy-2-fluorogalactosyl enzyme intermediate turned over slowly in buffer (ti/2 = 69 h at 25°C) and exhibited enhanced rates of reactivation in the presence of the acceptors methanol or glucose, providing strong evidence that the intermediate is catalytically competent The other deoxy and deoxyfluoro analogs synthesized were substrates for the enzyme although the rates of enzymic hydrolysis were two to four orders of magnitude slower than have been measured for the parent compound. These large rate reductions are thought to result primarily from the loss of important transition state binding interactions due to the substitution of a hydrogen or a fluorine for a hydroxyl at a given position on the galactopyranose ring. These results strongly suggest that much of the catalytic power of ii the enzyme is derived from non-covalent interactions between the enzyme active site and the galactopyranose ring of the substrate. A linear free energy relationship (r = 0.80) was shown to exist between the logarithm of kcat/Km for the enzyme-catalyzed reaction and the logarithm of the first order rate constant for the spontaneous hydrolysis of the same series of deoxy and deoxyfluoro glycopyranosides. Since the spontaneous process has considerable oxocarbonium ion character at the transition state, these data suggest that the enzymic mechanism involves a similar electron deficient transition state. iii TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iv LIST OF FIGURES vi LIST OF TABLES viii LIST OF ABBREVIATIONS AND DEFINITIONS ix ACKNOWLEDGEMENT x CHAPTER I INTRODUCTION 1. Glycosidases and glycoside hydrolysis 1 2. The catalytic mechanism of a "retaining" glycosidase 2 3. Escherichia coli (lacZ) P-galactosidase 2 3.1 Covalent glycosyl enzyme intermediate 4 3.2 Oxocarbonium ion-like transition states 7 3.3 Acid catalysis 9 3.4 Non-covalent interactions 10 4. Modified sugars as transition state probes 13 5. Aims of this study 15 CHAPTER H RESULTS AND DISCUSSION 1. Synthesis of fluorinated analogs of 2',4'-dinitrophenyl |3-D-galactopyranoside 17 1.1 Fluorination at the 2 position 17 1.2 Fluorination at the 3 position 18 1.3 Fluorination at the 4 position 21 1.4 Fluorination at the 6 position 22 2. Synthesis of deoxy analogs of 2',4'-dinitrophenyl (3-D-galactopyranoside 25 2.1 Deoxygenation at the 3 position 25 2.2 Deoxygenation at the 6 position 32 iv 3. Inactivation of E. coli (lacZ) (J-galactosidase with 2-deoxy-2-fluoro-galactopyranosides.. 33 4. Turnover of the E-P complex: hydrolysis and transglycosylation 41 5. Dinitrophenolate burst and active site titration 49 6. Modified sugars as transition state probes 52 7. Conclusions 70 CHAPTER m MATERIALS AND METHODS 1. Synthesis 72 1.1 General methods and materials 72 1.2 General procedures 73 1.2.1 Fluorination 73 1.2.2 Acetylation 74 1.2.3 Selective deacetylation at the 1 position 74 1.2.4 2',4'-Dinitrophenyl glycopyranoside formation 75 1.2.5 Deacetylation with HCV methanol 75 1.3 2',4'-Dinitrophenyl glycopyranosides 76 2. Enzymology 94 2.1 General procedures 94 2.2 Extinction coefficients 94 2.3 Determination of K m and for hydrolyses of 2,'4'-dinitrophenyl (3-D-glycopyranosides by E. coli (lacZ) p-galactosidase 95 2.4 Determination of Kj and kj for 2FpX3alDNP by time-dependent inactivation 97 2.5 Determination of (5-galactosidase concentration by absorbance at 280 nm and 2,4-dinitrophenolate burst 97 2.6 Reactivation of covalently inactivated P-galactosidase 98 REFERENCES 100 v LIST OF FIGURES Figure 1. Reaction catalyzed by a glycosidase 1 Figure 2. Proposed mechanism for hydrolysis of P-galactosides by a "retaining" f$-galactosidase 3 Figure 3. Kinetic model for partitioning of products between water and methanol 4 Figure 4. Affinity labelling of an active site glutamate by conduritol C epoxide 6 Figure 5. Covalent inactivation of pABG5 P-glucosidase by 2',4'-dinitrophenyl 2-deoxy-2-fluoro-p>-D-glucopyranoside 7 Figure 6. Galactosyl pyridinium salt 8 Figure 7. Comparison of structural and electronic similarity of a galactosyl cation with deoxyga/acronojirimycin and ga/acfonojirimycin 8 Figure 8. Stereochemistry of E. coli (lacZ) p*-galactosidase-catalyzed hydration of an octenitol derivative ; 10 Figure 9. Free energy diagram of an enzymic reaction involving: A) Maximum enzyme-substrate complementarity at the ground state; B) Maximum enzyme-substrate complementarity at the transition state 12 Figure 10. Synthetic scheme for the preparation of 2F|5GalDNP 18 Figure 11. Synthetic scheme for the preparation of 3F(5GalDNP 20 Figure 12. Synthetic scheme for the preparation of 4F(3GalDNP 21 Figure 13. Synthetic scheme for the preparation of 6FpXjalDNP 22 Figure 14. Proposed mechanism for reaction of N-(2-chloro-l,l,2-trifluoroethyl)-N,N-diemylamine with 1,2: 3,4-di-O-isopropylidene-a-D-galactopyranose 24 Figure 15. Reaction of methyl 2,3,6-tri-0-benzoyl-4-chloro-4-deoxy-a-D-glucopyranoside with sodium methoxide in methanol at 45°C 28 Figure 16. Proposed mechanism for formation of a 3,6-anhydro-a-D-galactopyranosyl structure from a 4-O-toluenesulphonyl glucopyranose moiety 29 Figure 17. Synthetic scheme for the preparation of methyl 3,4-anhydro-ct-D-galactopyranoside 30 Figure 18. Synthetic scheme for the preparation of 3d(3GalDNP 31 Figure 19. Synthetic scheme for the preparation of 6d(JGalDNP 33 Figure 20. Kinetic model for a "retaining" glycosidase 34 Figure 21. Hypothetical Gibbs free energy diagram for a "retaining" glycosidase showing rate-limiting deglycosylation 36 vi Figure 22. Kinetic model for inactivation via the accumulation of E-P, the glycosyl enzyme intermediate 37 Figure 23. Inhibition of E. coli (lacZ) R-galactosidase by 2Fp*GalDNP 40 Figure 24. Protection against inactivation by fJGalSiPr 41 Figure 25. Formation of allolactose by E. coli (lacZ) pVgalactosidase via transglycosylation 42 Figure 26. A kinetic model for turnover of the E-P complex via transglycosylation 43 Figure 27. Turnover of 2-deoxy-2-fluoro galactosyl enzyme in the presence of various ligands 44 Figure 28. 2,4-Dinitrophenolate burst observed upon incubation of E. coli (lacZ) (3-galactosidase with 2FpGalDNP 51 Figure 29. Stoichiometry of 2,4-dinitrophenol released from pVgalactosidase incubated with 2F|3GalDNP 51 Figure 30. Hydrolysis of pGalDNP 57 Figure 31. Hydrolysis of 6dpGalDNP 57 Figure 32. Hydrolysis of 4d|3GalDNP 58 Figure 33. Hydrolysis of 6FPGalDNP 58 Figure 34. Hydrolysis of 4FpGalDNP 59 Figure 35. Hydrolysis of 3FPGalDNP 59 Figure 36. Hydrolysis of pGluDNP 60 Figure 37. Chair to half-chair transition in the hydrolysis of a P-D-galactopyranoside via an oxocarbonium ion intermediate 62 Figure 38. Linear free energy relationships between kinetic parameters for the P-galactosidase-catalyzed reaction and the spontaneous hydrolysis rates of a series of deoxy and deoxyfluoro 2',4'-dinitrophenyl glycopyranosides. A) Plot of log(kcat/Km) vs. the logarithm of the spontaneous hydrolysis rate constant; B) Plot of log(kcat) vs. the logarithm of the spontaneous hydrolysis rate constant 67 vii LIST OF TABLES Table 1. Partitioning of products between methanol and water in |3-galactosidase-catalyzed solvolysis 5 Table 2. Size comparison of C-H, C-F, C-0(H) and C-OH groups 14 Table 3. Relative solvolysis rates of 1-phenylethyl esters and halides 30 Table 4. Turnover of 2-deoxy-2-fluorogalactosyl enzyme in the presence of various ligands 46 Table 5. Effect of deoxygenation at the 2 position on the hydrolysis of 1-deoxyglycerol-l-yl-P-D-galactopyranoside by E. coli (lacZ) p*-galactosidase 49 Table 6. Kinetic parameters of 2\4'-dinitrophenyl (3-D-glycopyranosides with E. coli (VacZjp'-galactosidase 56 Table 7. Spontaneous hydrolysis of a series of deoxy and deoxyfluoro 2',4'-dinitrophenyl P-D-glycopyranosides 61 viii LIST OF ABBREVIATIONS AND DEFINITIONS Carbohydrate Structure PGalPNP 4'-nitrophenyl P-D-galactopyranoside PGalONP 2'-nitrophenyl P-D-galactopyranoside PGalDNP 2',4'-dinitrophenyl P-D-galactopyranoside PGluDNP 2',4'-dinitrophenyl P-D-glucopyranoside PGalSiPr isopropylthio-P-D-galactopyranoside 2FPGalF 2-deoxy-2-fluoro-p-D-galactopyranosyl fluoride 2FPGluF 2-deoxy-2-fluoro-P-D-glucopyranosyl fluoride 2FpGluDNP 2',4'-dinitrophenyl 2-deoxy-2-fluoro-p-D-glucopyranoside 2FpGalDNP 2',4'-dinitrophenyl 2-deoxy-2-fluoro-p-D-galactopyranoside 3FpGalDNP 2',4l-dinitrophenyl 3-deoxy-3-fluoro-P-D-galactopyranoside 4FPGalDNP 2',4,-dinitrophenyl 4-deoxy-4-fluoro-P-D-galactopyranoside 6FpGalDNP 2',4'-dinitrophenyl 6-deoxy-6-fluoro-P-D-galactopyranoside 2dPGalDNP 2',4'-dinitrophenyl 2-deoxy-P-D-/yxo-hexopyranoside 3dPGalDNP 2',4'-dinitrophenyl 3-deoxy-P-D-xy/o-hexopyranoside 4dPGalDNP 2',4l-dinitrophenyl 4-deoxy-p-D-xy/o-hexopyranoside 6dpGalDNP 2,,4'-dinitrophenyl 6-deoxy-P-D-galactopyranoside Other Abbreviations Ac acetyl Bz benzoyl Ph phenyl Ts p-toluenesulphonyl Physical Constants k Boltzmann constant h Planck's constant T Temperature (K) R Gas constant ix ACKNOWLEDGEMENT I would like to express my thanks to the following people for their advice, help and support during the course of my research: my supervisors, Steve Withers and Mike Adam, for their invaluable guidance and encouragement. My coworkers, especially Mark Namchuk, who provided much useful advice and who I hope enjoyed our discussions as much as I did. And Altaire, whose unwavering belief in the eventual completion of this thesis was remarkable. x Chapter I INTRODUCTION 1. Glvcosidases and glycoside hvdrolvsis Glycosidases, or glycoside hydrolases, constitute a large family of enzymes which catalyze the hydrolysis of glycosidic linkages by cleavage of the C-O bond between the sugar and the aglycone (R, which may be another sugar or an aryl or alkyl group) as shown in Figure 1. Figure 1. Reaction catalyzed by a glycosidase. Glycosidases may be subdivided according to: 1) Anomeric configuration of the substrate. A pVglycosidase will only catalyze the hydrolysis of (i-glycosides while an a-glycosidase is specific for a-glycosides. 2) Stereochemical outcome of the reaction. Glycosidases are termed "inverting" or "retaining" depending on the relative anomeric configurations of the substrate hydrolyzed and the product initially released from the enzyme. Thus a "retaining" (3-glycosidase will catalyze the hydrolysis of (J-glycosides to give P-glycose products. 1 3) Glycone specificity. A given glycosidase usually exhibits maximum activity with a specific type of glycoside, e.g. a galactosidase will be most active against galactosides, in which the C(4) hydroxyl of the sugar is axial, though the hydrolysis of other glycosides may be catalyzed as well (albeit less efficiently). 2. The catalytic mechanism of a "retaining" glycosidase A general mechanism for "retaining" glycosidases, which are the focus of this study, was proposed by D.E. Koshland in 1953 (shown in Figure 2 for a |3-galactosidase).1 The main features of this mechanism, which are consistent with a large amount of experimental data that has accumulated since that time, are summarized below: 1) The reaction proceeds via formation of a covalent glycosyl-enzyme intermediate. 2) Both the formation and breakdown of this intermediate proceed via oxocarbonium ion-like transition states. 3) General acid catalysis may assist aglycone departure though this is not essential for all glycosidases with all substrates. 4) Non-covalent interactions between enzyme and substrate are responsible for most of the rate acceleration. These features will be discussed in some detail below, particularily in relation to the lac(Z) P-galactosidase from Escherichia coli. 3. Escherichia coli (lacZ) ft-galactosidase This large (464 kDa) tetrameric protein, encoded by the Z gene of the lac operon of E. coli, enables the organism to utilize lactose as an energy source. The enzyme exhibits a 2 Figure 2. Proposed mechanism for hydrolysis of p-galactosides by a "retaining" p-galactosidaseJ 3 somewhat liberal specificity for the aglycone and consequently may hydrolyze a wide variety of aryl |3-galactosides with varying degrees of efficiency. In addition to C-O bonds, C-S, C-F and C-N bonds in appropriate substrates may also be cleaved by the enzyme.2 The enzyme requires Na+ and Mg 2 + for full activity with O-galactosides.2 Although first crystallized in 19613, elucidation of the tertiary structure of the protein by X-ray crystallography has not been achieved. Consequently, a detailed picture of enzyme-substrate interactions in the active site has not yet emerged although a great deal of mechanistic information is available. 3.1 Covalent glvcosvl enzvme intermediate Stokes and Wilson4 reported a constant partioning ratio between the hydrolysis and methanolysis products formed in the (3-galactosidase-catalyzed solvolysis of a series of aryl (3-galactosides in mixtures of methanol and buffer (Figure 3 and Table 1). E + S — ES 1 Hp • E-P HX P-OH CH3OH P-OCH3 Figure 3. Kinetic model for partitioning of products between water and methanol. S is the substrate, ES the non-covalent Michaelis complex, E-P the glycosyl enzyme intermediate, HX the aglycone, P-OH the hydrolysis product and P-OCH3 the methanolysis product. 4 Though the rates of enzyme-catalyzed solvolysis vary widely, the constant ratio of products suggests involvement of an enzymic intermediate common to all substrates. However, little ijiformation about the nature of the mtermediate can be inferred from the data. Aglycone Methanolysis/ hydrolysis V m (relative) 2'-nitrophenyl 1.97 1.0 3'-chlorophenyl 2.08 0.5 4-nitrophenyl 1.99 0.2 phenyl 1.94 0.1 4'-bromophenyl 2.02 0.02 Table 1. Partitioning of products between methanol and water in fi-galactosidase-catalyzed solvolysis0 a Data from Stokes, T.M. and Wilson, I.B. (1972) Biochemistry 11,1061 . The existence of an intimate ion pair, as has been proposed for lysozyme5, is a possibility. However, given the instability of glycosyl cations (whose lifetime is estimated to be on the order of 10"10 seconds in water6), it seems unlikely that even a very favourable microenvironment in the active site could allow the leaving group to diffuse out and an acceptor to diffuse in and react before collapse to a covalent species occurred. This suggests that stabilization of the intermediate is occurring covalently via a nucleophilic residue on the enzyme. Herrchen and Legler7 have proposed that a possible candidate for such an enzymic nucleophile is an active site glutamate (Glu-461), identified with the affinity label conduritol C epoxide, which presumably reacts with the enzymic carboxylate group as shown to irreversibly inactivate the enzyme (Figure 4). 5 \ \ Glutamate Glutamate Figure 4. Affinity labeling of an active site glutamate by conduritol C epoxide. Site-directed mutagenesis studies8, in which the glutamate was replaced by a glutamine, produced a mutant enzyme which exhibited some 10"2 of the activity of the wild type. This residual activity with an uncharged glutamine residue in the mutant enzyme suggests that Glu-461, if it is indeed the enzymic nucleophile, acts not electrostatically but covalently to stabilize the developing positive charge at the anomeric centre.9 The covalent nature of a glycosyl-enzyme intermediate in another glycosidase, Agrobacterium pABG5 (5-glucosidase, has been unequivocally demonstrated.10 In the reaction of 2',4'-dintrophenyl 2-deoxy-2-fluoro-|3-D-glucopyranoside with P-glucosidase, the fluorine atom at C(2) of the pyranose ring slows both the glucosylation and deglucosylation steps, while the highly activated leaving group accelerates glucosylation sufficiently to trap the 2-deoxy-2-fluoroglucosyl intermediate and inactivate the enzyme (Figure 5). Evidence for the covalent nature of the intermediate and an a-linkage to the enzyme were obtained by 1 9 F nuclear magnetic resonance spectroscopy. The 2-deoxy-2-fluoroglucosyl enzyme is extremely stable, with a half-life of greater than five hundred hours at 30°C in buffer, but upon incubation with a suitable acceptor the enzyme may be reactivated and a disaccharide product isolated,11 demonstrating that the intermediate is 6 catalytically competent. The identity of the nucleophilic residue, a glutamic acid, was also deterrnined by inactivating the enzyme with a radiolabeled inhibitor, digesting the protein and comparing the sequence of the labeled peptide to the known amino acid sequence of the protein.11 HO HO OH 7 ^ A ^ O - H ^ ) — N 0 2 coo- °>N Glutamate Glutamate Figure 5. Covalent inactivation o/AgrobacteriumpABG5 fi-glucosidase by 2',4'-dinitrophenyl 2-deoxy-2-fluoro-B-D-glucopyranoside. 3.2 Qxocarbonium ion-like transition states Much of the evidence for oxocarbonium ion-like transition states associated with both the galactosylation and degalactosylation steps of (5-galactosidase-catalyzed hydrolysis is derived from kinetic isotope effect studies. By selection of appropriate aglycones it is possible to obtain substrates for which either the galactosylation step or the degalactosylation step is rate-limiting. The isotope effect on each step may then be determined. The P-galactosidase-catalyzed hydrolysis of C(l)-deuterated -^galactosyl pyridinium salts12 (Figure 6) exhibited a-secondary deuterium isotope effects of 1.15 to 1.20, indicating substantial sp2 character at the transition state for formation of the galactosyl-enzyme since galactosylation is rate-cletermining for these substrates. 7 Figure 6. Galactosyl pyridinium salt. Similar large isotope effects are observed when degalactosylation is rate-determining. In the enzyme-catalyzed hydrolysis of 2',4'-dinitrophenyl P-D-galactopyranoside, an a-deuterium isotope effect of 1.25 was reported13, suggesting increased sp2 character at the transition state for hydrolysis of an sp3-hybridized covalent intermediate. P-Glycosidases are strongly inhibited by compounds which resemble an oxocarbonium ion in conformation or charge (Figure 7). HO ^ O H O H ^ O H + HO H 6 > - " HO H O - \ ^ N v H 2 + a galactosyl cation deoxygalactonojirimycin Ki = 0.0125 mM galactonojirimycin Ki = 0.000045 mM Figure 7. Comparison of structural and electronic similarity of a galactosyl cation with deoxygalactonojirimycin and galacvonojirimycin. The former is isoelectronic with a galactosyl cation; the dehydrated form of the latter is isosteric as well. Ki values shown are for E. coli (lacZ) fi-galactosidase.14 8 These transition state analogs bind more tightly to the enzyme than normal, uncharged glycosides that adopt a chair conformation, as would be expected for inhibitors which mimic the positively charged half-chair transition state. The consequent extremely low Kj values for these inhibitors provide supportive evidence for the existence of oxocarbonium ion-like transition states. 3.3 Acid catalysis The role played by an acid catalyst in the catalytic mechanism of E. coli (lacZ) P-galactosidase is presently unclear. However, the efficient hydrolysis of P-galactosyl pyridinium salts (Figure 6), which cannot accept a proton, indicates that acid catalysis is not essential to catalysis by P-galactosidase.12 However, Mg 2 + , which binds to the enzyme in the ratio of one Mg 2 + per protomer15, contributes significantly to catalysis with some substrates. Removal of Mg 2 + from the enzyme has little effect on the kinetic parameters of P-galactosyl pyridinium salts while the kcat values of aryl O-galactosides decrease 4- to 150-fold upon removal of Mg 2 + . 1 6 It was originally suggested by Sinnott that Mg 2 + is required for optimal placement of an acid catalytic group13, accounting for the decreased rates with O-galactosides and the much smaller effect on p-galactosyl pyridinium salts when Mg 2 + was removed from the enzyme. A likely candidate for this acid catalytic group is tyrosine-503. Replacement of Tyr-503 by phenylalanine via siterdirected mutagenesis yields a mutant enzyme with kcat values for nitrophenyl galactosides some 10*3 that of the wild type enzyme.17 Further evidence for a strategically-placed acid catalytic residue has been reported by Lehmann and Schlesselmann.18 The stereochemistry of the product from hydration of the octenitol in Figure 8 indicates that the substrate was protonated from the P face. 9 D Figure 8. Stereochemistry ofE. coli (lacZ) frgalactosidase-catatyzed hydration of an octenitol derivative. Alternatively, it has been more recently proposed (again by Sinnott9) that Mg 2 + , when it can co-ordinate to the aglycone, may act as a simple electrophilic catalyst The site-directed mutagenesis studies of Edwards et al. 1 9 show that Glu-461 is important to Mg 2 + binding and may also be co-ordinated to Mg 2 + . The results of a previous nuclear magnetic resonance study with Mn+2-enzyme which suggested that a divalent metal ion bound too far from the aglycone to effect electrophilic catalysis now appear to be misleading.20-9 The stereochemical results of Lehmann and Schlesselman may be explained by addition of a Mg2+-coordinated water. The notion that the enzyme primarily uses electrophilic rather than general acid catalysis to assist aglycone departure is consistent with observations that Mg 2 + contributes significantly to catalysis even when the aglycone is an acidic phenol which is deprotonated at pH 7. Proton transfer from an acid catalytic residue of higher pKa, such as the simple phenol of tyrosine, would at best be only partial. 3.4 Non-covalent interactions Although stabilization of a developing oxocarbonium ion by an enzymic nucleophile and electrophilic or acid catalysis may contribute to the rate accelerations achieved by p*-galactosidase, it is unlikely that these factors alone account for the full catalytic power of the enzyme. It is now a well-accepted principle of enzymology that enzymes achieve such 10 spectacular rate enhancements by stabilization of the transition states for the steps involved in an enzyme-catalyzed reaction, a concept first suggested by Haldane21 in 1930 and later elaborated by Pauling.22 This stabilization of the transition state and consequent decrease in activation energy may be directly attributed to the increased binding energy realized upon tighter binding of enzyme and transition state (see Reference 23 for a detailed discussion). That enzyme complementarity to the transition state rather than the ground state of the substrate leads to an enhancement in rate is illustrated below for a simple enzyme-catalyzed reaction. When the ground state is stabilized by realizing maximum binding energy on binding the substrate (as in Figure 9 A)), the activation energy represented by AGC* is increased and the rate is therefore reduced. When the transition state is stabilized by realizing maximum binding energy on binding the transition state (as in Figure 9 B)), AG^ is decreased and the rate is enhanced, while the overall activation energy from free enzyme plus free substrate to the transition state (represented by AG-j4 ) is decreased by A G R . Thus by preferentially stabilizing the transition state through specific binding interactions, an enzyme is able to utilize binding energy for catalysis. Ligand/protein interactions are the basis for the specificity of enzyme/substrate binding and antigen recognition by irnmunoglobulins. Such interactions likely take the form of hydrophobic interactions and hydrogen bonding. Given the large number of hydroxyls associated with most carbohydrates, it is likely that hydrogen bonding plays an important role in binding and catalysis by enzymes with carbohydrate substrates. Indeed, the detailed picture of hydrogen bonding interactions between protein and ligand in arabinose binding protein derived from X-ray crystallographic data24 is an indication of the importance of hydrogen bonding in other carlx)hydrate-binding proteins. Elegant work by Fersht25 has shown that tyrosyl t-RNA synthetase utilizes a complex network of hydrogen bonds to achieve ground state specificity and to stabilize the transition state. Using site-directed mutagenesis, mutant enzymes were prepared in 1 1 Figure 9. Free energy diagram of an enzymic reaction involving: A) Maximum enzyme-substrate complementarity at the ground state: B) Maximum enzyme-substrate complementarity at the transition state. 12 which specific enzyme/substrate interactions (identified from the known X-ray crystallographic structure) had been deleted. Comparison of kinetic data for these mutants with that for the wild type enzyme yielded information about the importance of hydrogen bonding at each residue to catalysis. Fersht proposed that such interactions account for virtually all of the rate acceleration provided by this particular enzyme since no other residues which can participate catalytically (e.g. as nucleophiles, acid catalysts etc.) have been identified. 4. Modified sugars as transition state probes The complementary approach to that employed by Fersht in probing hydrogen bonding structure at the transition state involves systematic modification of the substrate rather than the enzyme. This more classical approach is appropriate when X-ray crystallographic data is not available, as is the case for E. coli (lacZ) P-galactosidase, and should provide essentially the same information. Specifically, systematic structural modifications of the glycopyranose ring within a series of substrates can provide information about: 1) interactions between the glycone and the enzyme at the transition state; and 2) the electronic nature of the transition state. Changes in the glycone must be sterically conservative to avoid steric clashes in the active site. The substiruents in the modified substrates must therefore be no larger than a hydroxyl group. Two substiruents which fulfill this criterion are hydrogen and fluorine (Table 2). Deoxy and deoxyfluoro sugars have been shown to be useful probes of hydrogen bonding interactions between an enzyme and substrate in the work of Street et al. with rabbit muscle glycogen phosphorylase.27-28 Note that a hydrogen bond consists of two paired functional groups. One group, termed the "donor", is covalently bonded to a 13 Group Bond Length (A) Van der Waals Radius (A) Total (A) C-H 1.09 120 2.29 C-F 1.39 1.35 2.74 C-0(H) 1.43 1.40 2.83 C-OH 1.43 2.10 3.53 Table 2. Size comparison of C-H, C-F, C-O(H) and C-OH groups.26 hydrogen atom which interacts with the lone pair electrons of the "acceptor" atom (which can be F, O or N). Thus a hydroxyl group on an unmodified sugar can both accept and donate a hydrogen bond while the hydrogen of a deoxy sugar cannot participate in any hydrogen bonding interaction. There is evidence from X-ray crystallographic data29 that fluorine may participate in C-F—H-N hydrogen bonds although the resultant interactions are probably weak since fluorine's lone pair electrons are held tightly by the nucleus. Evidence for C-F—H-0 hydrogen bonds in a fluorinated carboxylic acid has also been reported.30 A deoxyfluoro sugar may therefore be capable of acting as a (weak) hydrogen bond acceptor when an appropriate donor group is available but cannot possibly perform a donor role itself. All hydrogen bonding interactions between the enzyme and the hydroxyl at a particular position on the normal substrate will be abolished in the corresponding deoxy analog. The enzyme should show a significant loss in binding energy for this analog that will be reflected in increased activation energies for the catalytic steps. The deoxyfluoro analog would be expected to bind poorly if the substrate hydroxyl at that position acted as a hydrogen bond donor to a group on the enzyme. If, on the other hand, the sugar hydroxyl 14 normally accepts a hydrogen bond from the enzyme then the deoxyfluoro sugar may still bind appreciably. In this way, information about hydrogen bond polarity may be obtained. Substitution of a hydroxyl by fluorine or hydrogen also has significant effects on the electronic nature of the transition state. The rate of uncatalyzed spontaneous glycoside hydrolysis is decreased by electron-withdrawing substituents on the glycopyranose ring and accelerated by electron-rich groups because of the primarily electronic effects of these substituents on the developing positive charge at the transition state. Similar effects observed for the enzyme-catalyzed hydrolyses would, for example, constitute evidence that the enzymic transition states also have substantial oxocarbonium ion character. 5. Aims of this study It was proposed to probe the role of non-covalent binding interactions in transition state stabilization by E. coli (lacZ) P-galactosidase with a series of deoxy and deoxyfluoro glycopyranosides. In this way, information about the the importance to catalysis of enzyme-substrate interactions at each position on the galactopyranose ring and possibly hydrogen bond polarities may be obtained. The choice of a highly activated aromatic leaving group, 2,4-dinitrophenolate, as the aglycone permits convenient determination of enzymic reaction rates by measuring the rate of aglycone release spectrophotometrically. In addition, an analog incorporating a fluorine at C(2) and a good leaving group as the aglycone may permit trapping of a covalent 2-deoxy-2-fluorogalactosyl enzyme intermediate. The presence of electronegative fluorine adjacent to a developing positive charge at the anomeric centre would be expected to slow both the rates of galactosylation and degalactosylation while a good leaving group can accelerate only the galactosylation step. The substituent effects of these modified substrates may also provide insight into the electronic nature of the enzymic transition state. Comparison of the kinetic parameters 15 for the enzymic hydrolysis of these substrates with the spontaneous hydrolysis rates may provide evidence of similarity in electronic structure between the enzymic and non-enzymic transition states. In this way, the contribution of electronic factors to enzymic transition state structure may be investigated. 16 Chapter II RESULTS AND DISCUSSION 1. Syntheses of fluorinated analogs of 2'.4'-dinitrophenyl P-D-galactopvranoside All of the monofluorinated analogs of galactose have been synthesized by other workers previously.31'32*33'34 The intermediates synthesized in this work were characterized by comparison of their *H and 1 9 F nmr spectra and/or melting points, where possible, with literature values. The novel deoxyfluoro 2',4'-dinitrophenyl P-D-galactopyranosides were characterized by lH and 1 9 F nmr, melting point and elemental analysis. Excluding the characteristic resonances of the dinitrophenyl moiety, the signal farthest downfield in all cases is the anomeric proton. This appears as a doublet, except in the case of the 2-deoxyfluoro compound in which it also shows a 4 Hz coupling to fluorine. The magnitude of Ji,2 ( ~ 7.5 Hz) indicates a P-anomeric configuration in all cases. The magnitude of the ^ - i H coupling constants observed in the acetylated deoxyfluoro 2',4'-dinitrophenyl P-D-galactopyranosides indicates a trans-di&xial relationship between H-2 and H-3, suggesting the galactopyranose ring adopts the expected 4 C i conformation. 1.1 Fluorination at the 2 position The route employed for the preparation of 2',4'-dinitrophenyl 2-deoxy-2-fluoro-P-D-galactopyranoside 7 is shown in Figure 10. 17 Figure 10. Synthetic scheme for the preparation of2',4'-dinitrophenyl 2-deoxy-2-fluoro-p\D •galactopyranoside 7. l,3,4,6-Tetra-0-acetyl-2-deoxy-2-fluoro-a-D-galactopyranosyl bromide 4 was prepared in 92% yield from l,3,4,6-tetra-0-acetyl-2-deoxy-2-fluoro-a-D-galactopyranosyl fluoride (obtained from Dr. D. Dolphin, Department of Chemistry, U.B.C.) with 45% hydrobromic acid in acetic acid.35 The bromide 4 was converted to the alcohol 5 in 67% yield by reaction with silver carbonate in a mixture of acetone and water.36 The galactopyranoside 6 was prepared according to the procedure of van Boom et al. 3 7 as a 3:1 mixture of B and a anomers by reaction with l-fluoro-2,4-dinitrobenzene (FDNB) and 1,4 diazabicyclo[2.2.2]octane (DABCO) in N,N-dimemylformarnide. The p" anomer of 6 was selectively crystallized from ethyl acetate by addition of hexanes and isolated in 42% yield. Subsequent treatment of 6 with hydrogen chloride in methanol38 gave the deacetylated galactopyranoside 7 in 51% yield. 1.2 Fluorination at the 3 position 18 The synthetic route used for the preparation of the 3-deoxyfluoro analogs of galactose and 2',4' dinitrophenyl P-D-galactopyranoside is shown in Figure 11. Oxidation of 1,2:5,6 di-O-isopropylidene-a-D-glucofuranose39 by acetic anhydride in dimethyl sulphoxide (DMSO) has been reported to give the ulose hydrate 8.4 0 However, in our hands, the major product isolated from this reaction was the 3-thiomethoxy methyl ether 48 identified by comparison of its *H nmr spectrum with that reported in the literature.41 The desired product 8 was instead synthesized in 50% yield by oxidation of l,2:5,6-di-0-isopropylidene glucofuranose with phosphorus pentoxide in DMSO. 4 2 Preparation of the enol acetate 9 was carried out in 47% yield with a mixture of acetic anhydride and pyridine and was subsequently reduced stereoselectively using sodium borohydride in methanol to give the partially protected gulofuranose 10 in 20% yield.43 This compound was fluorinated with diemylaniinosulphur trifluoride (DAST) according to Kovac and Glaudemans44 to give the protected galactofuranose derivative 11 in 75% yield. The isopropylidene groups were removed by mild acid hydrolysis using Amberlite IR-120(H+) cation exchange resin in aqueous ethanol32 to give the free deoxyfluoro sugar 12 as a colourless oil in 73% yield. This was acetylated in 97% yield with acetic anhydride and pyridine to give a mixture of the tetra-O-acetates 13 which were selectively deacetylated at the 1 position with 2-aniinoethanol45 for 30 h at room temperature to give a mixture of the tri-O-acetates 14 in 58% yield. The galactopyranoside 15 was prepared in 47% yield by reaction of 14 with FDNB and DABCO 3 7 giving exclusively the P anomer. 48 19 Subsequent treatment of 15 with hydrogen chloride in methanol38 gave the deacetylated galactopyranoside 16 in 37% yield. 1. P2O5/DMSO 2 H 2 0 „ „ 0-iO CH 3 VH " U ~ C H 3 CH 3 g H 3C HO ^ J^CH, ho,, I AC20/ pyridiie H r /°"~\ NaBH4/CH3OH 0 " HO U0-VCK> A° 0 CH* AcO O C H „ 3 »3 10 9 DAST/CH 2C1 2 DMAP J 1 f 1. AmberliteIR-120(H+)/ l:2EtOH/H20 0 A c 2. AC2O/pyridine H 3C\ H 3 C W " F . ^ / V - - C H 3 ''O CH. 11 1. 2-aminoethanol/EtOAc 2. FDNB/DMF DAB CO t 3. HC1/ CH 3OH H O ^ - ° - ^ ^ ) - N 0 ' 16 °*N Figure 11. Synthetic scheme for the preparation of2',4'dinitrophenyl 3-deoxy-3-fluoro-p\D-galactopyranoside 16. 20 1.3 Fluorination at the 4 position The deacetylated galactopyranoside 21 was obtained from Dr. A. Becalski, Department of Chemistry, U.B.C. using the procedures indicated in Figure 12.4-Deoxy-4-fluoro-D-galactopyranose 19 was also synthesized separately in this study. HO HO OH Ph3CCl/ * pyridine H Q ^ H O ^ S HO OCH, HO OCPh, ?ASI7 OCPh, CH ^ NO-1. AC2O/pyridine 2. HBr/HOAc 3. Ag2C03/H20/ acetone 4. FDNB/DMF DAB CO 5. H C V C H 3 O H << H Amberlite m-120(H+)/H2O OH Figure 12 . Synthetic scheme for the preparation of 2',4' dinitrophenyl 4-deoxy-4-fluoro-P-D-galactopyranoside 21. Methyl a-D-glucopyranoside was treated with triphenylmethyl chloride in pyridine to give the trityl ether 17 in 77% yield This was fluorinated with DAST 4 6 to give the 4-deoxyfluoro galactopyranoside 18 in 21% yield and subsequently converted to the free deoxyfluoro sugar 19 in 89% yield by refluxing with Amberlite BR-120(H+) cation exchange resin in water for 3 days. 21 1.4 Fluorination at the 6 position The route employed for the preparation of the 6-deoxyfluoro analogs of galactose and 2',4' dinitrophenyl P-D-galactopyranoside is shown in Figure 13. 1. AmberliteIR-120(H+)/ 2. FDNB/DMF v DABCO 1 3. HCVCH3OH 27 Figure 13. Synthetic scheme for the preparation of 2',4' dinitrophenyl 6-deoxy-6-fluoro P-D-galactopyranoside 27. l,2:3,4-Di-0-isopropylidene-a-D-galactopyranose may be prepared in one step from D-galactose and acetone in the presence of sulphuric acid,39 It seemed likely that the 6-deoxyfluoro derivative 22 could be readily prepared by treatment of this compound with DAST in the usual manner. However, initial attempts, upon work-up, produced only a non-fluorinated product or mixtures of this product and the desired 6-deoxyfluoro derivative 22 in low (20 to 25%) yields. This non-fluorinated product was identified as the 6-c/t/orodeoxy analog, l,2:3,4-di-0-isopropylidene-6-chloro-6-deoxy-a-D-galactopyranose 49, by comparison of its *H nmr spectrum with that of an authentic 22 sample prepared by another route.83 Isolation of this compound was surprising, the only source of chlorine in the reaction being dilute hydrochloric acid used in the work-up. Normal work-up for fluorinations with DAST consists of quenching the reaction with excess methanol at -20°C, allowing the mixture to warm to room temperature and washing with cold, dilute hydrochloric acid to remove 2,4,6^ trimethylpyridine or a similiar organic base used in the reaction. When the reaction was quenched with excess methanol by stirring for several hours at room temperature and washed with 1 M aqueous acetic acid, the 6-deoxyfluoro compound 22, obtained in 40% yield, was the sole isolated product. Isolation of the chlorodeoxy compound 49 suggests interception of a putative DAST intermediate such as 50 by chloride ion during work-up. Relatively stable intermediates resulting from reaction of DAST or similar fluorinating reagents with sterically hindered alcohols have been reported in the literature. Tewson and Welch47 have observed a new 1 9 F nmr signal when DAST was reacted with the hindered alcohol l,2:5,6-di-0-isopropylidene-a-D-glucofuranose, which they attribute to a DAST intermediate. A fluorinated carbohydrate could not be isolated from this reaction but starting material was recovered upon treatment of the mixture with ethanol, while distillation from pyridine at 80°C yielded an elimination product Evelyn and Hall4 8 have reported that treatment of l,2:3,4-di-0-isopropylidene-a-D-galactopyranose with the so-called fluoramine reagent, N-(2-chloro-l,l,2-trifluoroemyl)-N,N-diethylamine gave none F 23 of the expected 6-deoxyfluoro product 22 but that aqueous work-up afforded the chlorofluoroacetate 51. 1 9 F nmr spectroscopic evidence was presented in support of the following scheme. F ( C H 2 C H 3 ) 2 N — C F 2 — C H C 1 F • ( C H J C H J ) ^ C C H C 1 F O R R O H rs I J l—C 1 \ O H 2 ° + C H C 1 F II M ( C H 2 C H 3 ) 2 N = C ^ R C T ^ C H C I F 0 R 51 H 3 C i o . A . R = ^ R F + ( C H 2 C H 3 ) 2 f T X H C 1 F 22 Figure 14. Proposed mechanism for reaction ofN-(2-chloro-l,12-trifluoroethyl)-Nfl-diethylamine with 12:3,4-di-O-isopropylidene-a-D-galactopyranose.48 It appears likely that steric hindrance of the 6 position by the axial isopropylidene group at C-4 contributes to the the relative stability of a DAST intermediate such as 50 and accounts for the formation of the 6-chlorodeoxy analog 49 when an aqueous hydrochloric acid work-up is employed in the synthesis of 22. 24 The protected deoxyfluoro sugar 22 was stirred with Amberlite IR-120(H+) cation exchange resin in aqueous ethanol to remove the isopropylidene groups,32 affording the free sugar 23 which crystallized exclusively as the a anomer in 75% yield. This compound was acetylated with acetic anhydride and pyridine in quantitative yield to give an anomeric mixture of per-O-acetates 24, then selectively deacetylated at the 1 position with 2-aminoethanol45 to give a mixture of the tri-O-acetates 25 in 56% yield. The galactopyranoside 26 was prepared by the reaction of 25 with FDNB and DAB CO 3 7 in N,N-dimethylformamide to give the P anomer in 40% yield. This compound was treated with hydrogen chloride in methanol38 to give the deacetylated galactopyranoside 27 in 70% yield. 2. Syntheses of deoxy analogs of 2'.4'-dinitrophenvl P-D-galactopyranoside Both the 6- and 4-deoxy analogs of 2',4'-dinitrophenyl P-D-galactopyranoside were prepared. The latter compound, 2',4'-dinitrophenyl 4-deoxy-P-D-xy/o-hexopyranoside 42 was obtained from M. Namchuk, Department of Chemistry, U.B.C. The complete synthesis of the 3-deoxy analog was not achieved. Its immediate precursor 2',4'-dinitrophenyl 2,4,6-tri-O-acetyl-3-deoxy-P-D-xy/0-hexopyranoside 39 was synthesized but attempts to isolate the deacetylated glycoside after treatment with hydrogen chloride in methanol38 were unsuccessful. 2.1 Deoxygenation at the 3 position 25 The preparation of 3-deoxy-ry/o-hexopyranosides by conventional methods, i.e. reductive dehalogenation49 of the protected halogeno derivative, involves long, multistep syntheses giving the final product in only modest yields. A recent improved synthesis50 by Lin, Kovac and Glaudemans is reported to give methyl 3-deoxy-P-D-xy/o-hexopyranoside 52 in 20% overall yield from methyl P-D-galactopyranoside in seven steps. For hexopyranosides, the synthetic problem is complicated by the propensity of 3-deoxy-D-xy/o-hexose to exist partially in its furanose form (19.5% at equilibrium at 31°C in water51). We proposed an alternative synthesis of methyl 3-deoxy-a-D-xy/o-hexopyranoside 35 prompted by two observations in the literature: 1) Hedgley et al. have reported that lithium aluminum hydride reduction of methyl 3,4-anhydro-a-D-galactopyranoside 28 afforded methyl 3-deoxy-a-D-xy/o-hexopyranoside 35 in 76% yield52; 2) according to Jones et al., the anhydro compound 28 could be prepared in 74% yield from methyl 3-chloro-3-deoxy-ct-D-gulopyranoside 53 in aqueous sodium hydroxide over 3 days at 40°C 53 Methyl 2,3,6-tri-0-benzoyl-4-chloro-4-deoxy-a-D-glucopyranoside 54 is available in 55% overall yield in two steps from methyl ct-D-galactopyranoside, by selective benzoylation at the 2-, 3- and 6-positions54 followed by treatment with sulphuryl chloride.55 The corresponding 4-O-toluenesulphonyl derivative 55, upon treatment with sodium methoxide in methanol at 0°C for several hours, is reported to give the anhydro compound 28 in 81% yield.56 It seemed plausible that treatment of the benzoylated 4-chloro-4-deoxy glucopyranoside 54 with base might give the anhydro compound 28 in one step, analogous to the base promoted epoxidation of the 3-chloro-3-deoxy gulopyranoside 53. Reduction of the anhydro compound 28 as described by Hedgley et al. 5 2 above would then give the desired 3-deoxy-ry/o-hexopyranoside 35. 26 Ri R-2 R3 R4 R5 R6 R? 29 H H OCH3 OH H a H 35 H H OCH3 H H H OH 52 H OCH 3 H H H H OH 53 H H OCH3 H a H OH 54 Bz H OCH3 OBz H a H 55 Bz H OCH3 OBz H OTs H Bz = benzoyl; Ts = p-toluenesulphonyl However, neither the aqueous conditions employed by Jones et al. 5 3 nor sodium methoxide in methanol gave satisfactory yields of the anhydro compound 28 from the debenzoylated 4-chloro-4-deoxy glucopyranoside 29 or the 4-chloro-4-deoxy glucopyranoside 54, respectively. Room temperature treatment of 4-chloro-4-deoxy glucoside 54 with sodium methoxide in methanol gave only the debenzoylated 4-chloro-4-deoxy glucopyranoside 29 in 87% yield. When 4-chloro-4-deoxy glucopyranoside 54 was heated with sodium methoxide in methanol at 45°C for 60 hours, a ~ 1:1:1 mixture of the anhydro compound 28, the debenzoylated 4-chloro-4-deoxy glucopyranoside 29 and 27 the 3,6-anhydrogalactopyranoside 30, separable by column chromatography, was produced (Figure 15). NaOCHs/ C H 3 O H Figure 15. Reaction of methyl 2,3,6-tri-0-benzoyl-4-chloro-4-deoxy-a-D-glucopyranoside 54 with sodium methoxide in methanol at45°C. When the 4-chloro-4-deoxy glucopyranoside 54 was refluxed with sodium methoxide in methanol for 18 h, the sole product was the 3,6-anhydro-a-D-galactopyranoside 30, identified by comparison of its lH nmr spectrum to literature values57, obtained in 77% yield. Clearly, it seemed the desired anhydro compound 28 could not be formed from the 4-chloro-4-deoxy glucopyranoside 54 under the conditions exarnined without irreversible formation of the 3,6-anhydrogalactoside 30 as well. Formation of a 3,6-anhydro-a-D-galactopyranosyl moiety has been reported58 upon treatment of 2,3,6^,,3'-penta-0-acetyl-4,r,6'-tri-0-toluenesulphonyl-D-sucrose 56 with re fluxing sodium ethoxide in ethanol (Figure 16). Presumably, the 3,4-epoxide ring is formed initially (step 1), followed by migration to the 2,3-position (step 2) which permits closure of the 3,6-anhydro ring (step 3). This behavior is well known59 in the 2,3-anhydro allopyranoside series. 28 TsO Figure 16. Proposed mechanism for formation of a 3,6-anhydro-a-D-galactopyranosyl structure from a 4-O-toluenesulphonyl glucopyranose moiety.58 The low reactivity of the 4-chloro-4-deoxy glucopyranoside 54 compared to the 3-chloro-3-deoxy gulopyranoside 53 and the 4-O-toluenesulphonyl glucopyranoside 55 is not difficult to rationalize. The hydroxyl of the guloside 53 is already well positioned to displace the axial chlorine and the glycopyranose ring does not have to undergo an unfavourable conformational change to position the hydroxyl for an optimal in-line attack. It would be expected, furthermore, that chloride is a much worse leaving group than the p-toluenesulphonate group in 55 (Table 3).60 29 Leaving group kre l trifluoromethanesulphonate 1.4 x 108 p-toluenesulphonate 3.7 x 104 methanesulphonate 3.0 x 104 I- 91 Br 14 ci- 1.0 F- 9X10-6 Table 3. Relative solvolysis rates of' 1-phenylethyl esters and halides.a& a Data from Noyce, D.S. and Virgilio, J.A. (1972) /. Org. Chem. 37, 2643 . b In 80% aqueous ethanol at 75°C. Iodide is generally assumed to be a better leaving group than chloride and indeed it was found that treatment of methyl 2,3,6-tri-0-benzoyl-4-deoxy-4-iodo-a-D-glucopyranoside 40 with sodium methoxide in 2:1 methanol/ N,N-dimethylformamide at room temperature for 4 days afforded the anhydro compound 28, identified by comparison of its lH nmr spectrum to uterature values61, in 66% yield (Figure 17). NaOCH3/ 1. BzCU pyridine 2:1 C H 3 O H / 2. CH3SO2CV pyridine DMF Figure 17. Synthetic scheme for the preparation of methyl 3,4 •anhydro-a-D-galactopyranoside 28. The 4-deoxy-4-iodo glucopyranoside 40 was prepared in 53% yield by iodination of methyl 2,3,6-tri-0-benzoyl-4-0-methanesulphonyl-a-D-galactopyranoside 57 with 30 sodium iodide in refluxing N,N-dimeftylformamide.62 The methanesulphonate 57 is available in 58% overall yield in two steps from methyl a-D-galactopyranoside by selective benzoylation at the 2-, 3- and 6-positions, followed by reaction with methanesulphonyl chloride in pyridine.55 Methyl 3-deoxy-a-D-xy/o-hexopyranoside 35 is thus available in five steps via this route in 15% overall yield from methyl a-D-galactopyranoside. An alternative strategy to block the irreversible formation of the 3,6-anhydro ring was also exarnined (Figure 18). [ Hr° H OCH, 1. BzCV pyridine l H 2. SO2CI2/ pyridine —> a BzO BzO 1. NaOCH3/CH3OH 0 B z 2. Ph3CCl/pyridine 0 — ° OCH, AcO ^.OAc 1. Amberlite IR-120(H+)/ 1:2 E1OH/H2O 2. AC2O/pyridine OH 36 O C H 3 CI2CHOCH3, ZnCl2 I*" OCPh3 LiAlH^THF 0 ^ 0 C P h 3 ^ - 0 H O ^ ^ j OCH 34 NaOCH 3 £H 3 OH reflux 9 H ^OCPh 3 OCH- OCH, 32 33 AcO O A^>OAc 1. Ag2CO3/H20/ I OAc I \ acetone L \ N - o • V ^ - o )—?—\ 2. FDNB/DMF A c ° I DAB CO CI 37 39 A c O ^ b -NO, O ^ Figure 18. Synthetic scheme for the preparation of 2',4'•dinitrophenyl 2,4,6-tri-O-acetyl-3-deoxy-P-D-xy\o-hexopyranoside 39. 3 1 The debenzoylated 4-chloro-4-deoxy glucopyranoside 29 was synthesized in 87% yield from the protected 4-chloro-4-deoxy glucopyranoside 54 as described previously, then treated with triphenylmethyl chloride in pyridine to give the 6-O-triphenylmethyl derivative 31 in 76% yield. This compound was refluxed with sodium methoxide in methanol for 15 h to afford a 1: 1.4 mixture of the 3,4-anhydro galactopyranoside 32 and the 2,3-anhydro gulopyranoside 33 in a combined yield of 80%. The two isomers were separated by column chromatography and the pure 3,4-anhydro galactopyranoside 32 was treated with lithium aluminum hydride in THF to give exclusively the 3-deoxy-xy/o-hexopyranoside 34 in 95% yield. The triphenylmethyl group was removed by mild acid hydrolysis with Amberlite IR-120(H+) cation exchange resin in aqueous ethanol32 and the resulting gum acetylated without purification to give the acetylated methyl 3-deoxy-xy/o-hexopyranoside 36 in 75% yield from 34. Primarily because of loss of material as the unwanted 2,3-anhydroguloside 33, this route is inefficient, with an overall yield of 36 from methyl cc-D-galactopyranoside of only 9%. Compound 36 was converted to the cc-hexopyranosyl chloride 37 with 1,1-dichloromethyl methyl ether and zinc chloride63 and hydrolyzed without purification to the alcohol 38 in 75% yield from 36.36 Compound 38 was converted to the glycopyranoside 39 by reaction with FDNB and DABCO in N,N-dimethylformamide37 in 23% yield. Attempted deacetylation of this compound with hydrogen chloride in methanol38 gave only hydrolyzed material, but it was anticipated that the deacetylated 2',4'-dinitrophenyl 3-deoxy-p*-D-.xy/<?-hexopyranoside would be extremely labile (see Section 6, p. 60). 2.2 Deoxygenation at the 6 position The route employed for the preparation of 2',4'-dinitrophenyl 6-deoxy-P-D-galactopyranoside 46 is shown in Figure 19. 32 1. NH 2NH 3 +OAc7 DMF Figure 19. Synthetic scheme for the preparation of2',4'-dinitrophenyl 6~deoxy-f5~D-galactopyranoside 46. D-Fucose (6-deoxy-D-galactose) was acerylated with acetic anhydride in pyridine to give a mixture of the tetra-O-acetates 43 in quantitative yield. This material was selectively deacetylated at the 1 position with hydrazine acetate in N.N-dimethylformamide64 in 69% yield and the resulting alcohol 44 converted to the galactopyranoside 45 in 69% yield with FDNB and DABCO in N,N-dimethylformarnide.37 Subsequent treatment with hydrogen chloride in methanol38 gave the deacetylated galactopyranoside 46 in 25% yield. 3. Inactivation of E. coli (lacZ) (3-galactosidase with 2-deoxv-2-fluoro galactopyranoside s The general mechanism proposed by Koshland1 for a "retaining" glycosidase is shown in Figure 2 (see Chapter 1, p. 3). Assuming only the chemical steps are kinetically significant, the mechanism may be depicted as follows. 33 k2 k3 E + S ES E-P » E + P k HX H 2 0 Figure 20. Kinetic model for a "retaining" glycosidase. E is the free enzyme, S the glycoside, HX the aglycone, ES the noncovalent Michaelis complex and E-P the covalent gtycosyl enzyme intermediate, ki and Ay are the first order rate constants for gfycosylation and deglycosylation, respectively. Assuming a steady state concentration of ES and E-P is reached during the reaction The total concentration of enzyme, E 0 is the sum of the concentrations of free enzyme and all enzyme-bound species then k2[ES]=k3[E-P] and d[ES] = ki[E][S] -k.i[ES] -k2[ES] dt = 0 [Eol = IE] + [ES] + [E-P] Therefore, by substituting for [E-P] and rearranging ki[Eo][S] = (k.i + k2)[ES] + ki[ES][S] +kik2fESirSl k3 = (k.i + k2)[ES] + (kik3/k3 + kik2/k3)[ES][S] Solving for [ES] [ES] = k.i + k2-r ki(k2 + k3) [S] Since at steady state the rate of production of P, vp is given by 34 d[P] =k3[E-P] " o T = k2[ES] then v P= k2k3_[Eo][S] k 2 + k? k3 k . i+k 2 + [S] k2 + k3 ki which follows the standard form of the Michaelis-Menten equation v= wfEnirsi Km+[S] Therefore the kinetic parameters for this mechanism are kcat= k2k3 k 2 + k3 K m = k3(k.i + k2) ki(k2 + k3) kcat/Km = kik2 k-l + k 2 It can be shown that kcat is the rate constant for the rate-determining step of the reaction and will always be associated with the highest free energy step in the pathway. kcat/Km will always be the pseudo second-order rate constant for the free enzyme and free substrate proceeding to the transition state of the first irreversible step.23 When the aglycone in Figure 20 is a highly activated leaving group (e.g. 2,4-dinitrophenolate), it would be expected that if the rate of glycosylation was increased sufficiently relative to deglycosylation then deglycosylation would become rate determining, i.e. k2 » k3. This has been shown by methanol competition experiments to be the case for the hydrolysis of p*GalDNP by E. coli (lacZ) (3-galactosidase.65 Assuming 35 further a rapid, reversible association of enzyme and substrate, i.e. k-i » k2,the kinetic parameters kcat and kcat/Km then become kcat = k3 kcat/Km = k i k 2 k-i The Eyring equation relates these rate constants to differences in energy levels among the various species on the reaction pathway. kcat = (kT/ h)exp{-(GEpt - GEP)/ RT} = (kT/h)exp{-AGEp*/RT} kcat/Km = (kT/h)exp{-(GES*-GE + s)/RT} = (kT/h)exp{-AGft/RT) The free energy diagram for this situation is depicted schematically in Figure 21. E + P Reaction coordinate Figure 21. Hypothetical Gibbsfree energy diagram for a "retaining" gfycosidase showing rate-limiting deglycosylation, i.e. k.3 « *2-36 Carrying rate-limiting deglycosylation to its extreme, if k3 is sufficiently small compared to k2 such that k3 approaches zero then the glycosyl enzyme intermediate will accumulate, inactivating the enzyme, and the kinetic model becomes k.i k2 E + I " EI v » E-P HX Figure 22. Kinetic model for inactivation via she accumulation of E-P, the glycosyl enzyme intermediate. E is the free enzyme, El the noncovalent enzyme-inhibitor complex and HX the aglycone. This model predicts a rapid time-dependent inactivation of the enzyme. If [I] is much greater than [EQ] and thus [I] is assumed to be essentially constant during the reaction, then pseudo first-order kinetics with respect to enzyme concentration will be observed. In Michaelis-Menten form, the equation for this process is vi = kifEnim (1) K i + [I] where ki is the rate constant for inactivation and Ki is an apparent dissociation constant for all forms of enzyme-bound inactivator. As [I] is assumed to be constant, Equation (1) can be rewritten as vj = kobslEo] where kobs =_MH_ (2) Ki+[I] The rate of inactivation, V J is equal to 37 -d[Eo] = kobs[E0] dt Thus ln[E0] =-kobst From inspection of Figure 21, it can be seen that ki is equal to the rate of glycosylation, k2. Thus, in Eyring form kj = k2 = (kT/ h)exp{-AGct/ RT) with Ki = k.i since the association of enzyme and inhibitor is assumed to be rapid and reversible, i.e. k-i » k 2 , then kj/Kj= kik2 = (kT/ h)exp{-AGT*/ RT} k-i Both kj and kj/Ki are associated with the energy barriers for glycosylation, but kj reflects the height of the energy barrier from the Michaelis complex to the transition state for glycosylation while kj/K, reflects the barrier height from free enzyme and free inactivator to the glycosylation transition state. Note that in this scheme, kj/Ki is analogous to kcat/Km in terms of the individual rate and equilibrium constants for each step. Recently it has been shown66 that incubation of E. coli (lacZ) JJ-galactosidase with 2-deoxy-2-fluoro-(3-D-galactopyranosyl fluoride at 37°C led to a rapid, time-dependent inactivation of the enzyme. Protection against inactivation by a competitive ligand, isopropylthio-P-D-galactopyranoside, indicated that inhibition was occuring at the enzyme active site. It was proposed that substitution of the highly electronegative fluorine at C(2) of the glycopyranose ring destabilizes the oxocarbonium ion-like transition states involved in both the glycosylation and deglycosylation steps, while acceleration of glycosylation by 38 deoxy-2-fluoro-glycosyl enzyme intermediate (Figure 5, p. 7). Similar behavior was observed when Agrobacterium pABG5 P-glucosidase was incubated with various aryl 2-deoxy-2-fluoro glucosides.67 Prompted by these results, E. coli (lacZ) (3-galactosidase was tested for similar time-dependent inactivation by 2',4'-dinitrophenyl 2-deoxy-2-fluoro-P-D-galactopyranoside at 25°C. Enzyme was incubated at 25°C in the presence of inactivator. Aliquots were removed at intervals and assayed for residual activity by dilution into a saturating concentration of jJPNPGal. This prevents further inactivation by diluting the inactivator with a large excess of a competing ligand. The observed pseudo first-order rate constants for the inactivation process, kobs. at each inhibitor concentration were determined by plotting the logarithm of the fraction of remaing active enzyme against time. Values for Kj and kj were calculated by fitting the rate constants so determined to Equation (2) using a weighted non-linear regression program. As expected, incubation of P-galactosidase with 2FPGalDNP resulted in inactivation of the enzyme according to pseudo first-order kinetics. The half-life of inactivation at saturating concentrations of inactivator was estimated to be 63 seconds. This result is consistent with the kinetic model discussed earlier in which deglycosylation is slow compared to glycosylation and the E-P complex rapidly accumulates (Figure 22). The results are shown in Figure 23. Values for ki and Kj were found to be 0.011 s'1 and 0.89 mM, respectively. The active site-directed nature of the inactivator was shown by protection against inactivation by PGalSiPr, a competitive ligand (Ki = 0.085 mM71) which binds reversibly at the active site (Figure 24). 39 A) B 0.022mM • 0.033 mM • 0.049 mM • 0.11 mM • 0.28 mM • 0.50 mM A 1.1 mM 0 -F ' 1 1 1 ' 1 ' 1 r 0 10 20 30 40 50 l/[2FpGalDNP] (1/mM) Figure 23. Inactivation ofE. coli (lacZ) B-galactosidase by 2FpGalDNP. A) A semilogarithmic plot of residual activity vs. time at 25°C, pH 7.00 in 50 mM sodium phosphate buffer containing 1 mM Mg2+, 0.1% BSA. Inhibitor concentrations are as indicated. B) Double reciprocal plot of the following values obtained for kobs ot their respective inhibitor concentrations: 0.016,0.024, 0.035, 0.071, 0.15, 025 and 0.35 min'K 40 • 0 mM • 0.5 mM 0 10 20 30 time (min) Figure 24. Protection against inactivation by fiGalSiPr. Inactivation at a single fixed concentration of2FpGalDNP (032 mM) with the indicated concentrations of BGalSiPr: 0 mM (k0bs = 0.13 min'1), 050 mM (k0bs =0.016 min'1). 4. Turnover of the E-P complex: hydrolysis and transglvcosvlation Inactivation of glycosidases by trapping of a 2-deoxy-2-fluoroglycosyl enzyme intermediate represents the first step in turnover of an exceedingly slow substrate. When freed from excess inhibitor, the inactivated glycosyl enzyme would be expected to slowly hydrolyze, releasing a 2-deoxy-2-fluoro sugar and the free enzyme. The return of activity with liberation of free enzyme could then be assayed with a normal substrate. Water is not the only molecule that can act as an acceptor for the 2-deoxy-2-fluorogalactosyl moiety released upon turnover of the galactosyl enzyme. The transglycosylase activity of E. coli (lacZ) p*-galactosidase is well known.68'69 When the natural substrate of the enzyme, lactose (a P(l-4)-linked (^ saccharide), is hydrolyzed, approximately 50% of the lactose consumed is converted by an intramolecular 41 trangalactolysis to allolactose, a P(l-6)-linked disaccharide.68 Allolactose is the natural inducer of the lac operon and its synthesis in this manner thus has a specific physiological role (Figure 25). + D'glucose + D-galactose Figure 25. Formation of allolactose by E. coli lac(Z) B-galactosidase via transglycosylation The efficiency of this reaction implies the presence of a second discrete "glucose" binding site adjacent to the active site which provides a high local concentration of a glycosyl acceptor enabling the transgalactolysis reaction to compete effectively with hydrolysis. Indeed, such a site might be expected since the natural substrate has a sugar "aglycone". Such multiple binding sites have been identified in hen egg white lysozyme by X-ray crystallography.70 However, unlike lysozyme or other glycohydrolases which have evolved to bind and hydrolyze oligosaccharides, the secondary binding site of E. coli lac(Z) P-galactosidase is apparently quite small since tri- or tetra-saccharides have no effect on the kinetic parameters of PGalONP hydrolysis by the enzyme,71 indicating that these compounds are unable to interact with the active site. 42 The enzyme catalyzes intermolecular transglycosylations to a wide variety of monosaccharides and aliphatic alcohols when these compounds are present during the enzyme-catalyzed hydrolysis of |3GalONP.69 It has been shown that compounds which could form transfer products were uncompetitive effectors of the enzyme, binding preferentially to the glycosyl enzyme rather than free enzyme. For example, Deschavanne et al.7 1 found that glucose binds relatively weakly to the free enzyme with a Ki greater than 600 mM while a dissociation constant (= k-s/ks in Figure 26 below) for glucose binding to the glycosyl enzyme has been determined by Huber et al. to be 17 mM. 6 9 A kinetic model for turnover of the isolated 2-deoxy-2-fluorogalactosyl enzyme in the presence of ligands which may act as transglycosylation acceptors is shown below. E + P E +P"S Figure 26. A kinetic model for turnover of the E-P complex via transglycosylation. EP and k3 are as defined previously, S is the ligand, E-PS is the glycosyl enzyme-ligand complex, ktrans is the rate constant for transglycosylation and Ktrans ~ (k-s + kp-ans)/ks. The rate of decomposition of E-P is given by -d[E-P3 = (k'3 + k'tnmS)[E-P] dt where k'trans =_kJraQS[S]L *nd k'3 = k3[H20] Ktrans + [S] Note that K^ans is equivalent to a true dissociation constant (= k_s/ks) for binding of S to the glycosyl enzyme if ktrans is slow compared to a rapid and reversible association of S E'P i H 2 0 S Ktrans E-PS • 43 and E-P, i.e. ktrans « k-i. When [S] is large compared to [E-P], the reaction will be pseudo first-order with respect to [E-P] and the observed rate constant, kobs, will be equal to k'3 + k'trans- The resulting time-dependent increase in activity can be related to the full activity of the sample by Inactivated jJ-galactosidase was produced by adding a fifty-fold excess of 2Fp*GalDNP to - 0.1 mg of enzyme in 1.6 mL buffer and incubating at 25°C for several hours to ensure complete inactivation. The inactivated galactosyl enzyme was freed of excess inactivator by centrifugation of the enzyme sample in a Millipore intrafree-PF filter unit fitted with a polysulphone membrane (30 000 molecular weight limit), diluted with buffer and assayed for activity at 25°C by diluting small aliquots into ^GalPNP at saturating concentration. The results are shown below (Figure 27). [E] = [EoKl-exp{-kobst}) A) 100 B buffer • 1 M MeOH • 1 M D-Glucose 0 100 200 time (h) 44 B) -2 Q buffer • 1 M MeOH time (h) Figure 27. Turnover of the 2-deoxy-2-fluoro galactosyl enzyme in the presence of various ligands. A) Plot of activity vs. time for reactivations in buffer, 1 M methanol and 1 M D-glucose, respectively B) Semilogarithmic plot of activity vs. time for reactivations in buffer and 1M methanol, respectively. All reactivations were performed in 50 mM sodium phosphate, 1 mM Mg2+,pH 7.00 at 25°C. Return of activity for the 2-deoxy-2-fluorogalactosyl enzyme in the presence of buffer alone or 1 M methanol followed a single exponential time course (Figure 27 B)). The half-life of the 2-deoxy-2-fluorogalactosyl enzyme in buffer at 25°C was estimated to be 69 hours. Control samples with non-inactivated enzyme to account for denaturation in buffer or 1 M methanol over the time course of the experiment showed negligible loss of activity. It can be seen from Figure 27 A) that the reactivation of the galactosyl enzyme in the presence of 1 M D-glucose does not follow simple pseudo first-order kinetics. The kobs value was estimated from the initial exponential increase in activity during the first 30 hours. Though it appears that some inactivation process is superimposed upon the reactivation curve, controls with non-inactivated enzyme in the presence of 1 M glucose showed only a very slight decrease in activity over ~ 75 hours. A speculative rationale that may account for the shape of the observed curve is that the 2-deoxy-2-fluorogalactosyl enzyme is more susceptible to denaturation in the presence of a high concentration of glucose than the free enzyme. acceptor kobs (s*1) relative rate half-life (h) H 2 0 2.8 x 10-6 1 69 1 M methanol 5.2 x 10"6 1.9 37 1 M D-glucose ~ 5 x 10-5 * -20 -3.5 Table 4. Turnover of the 2-deoxy-2-fluorogalactosyl enzyme in the presence of various ligands. * Calculated from initial rate of reactivation based on return of activity in first 30 h. From inspection of Figure 27 A) and Table 4 it is clear that both methanol and D-glucose can act as efficient transglycosylation acceptors, reactivating the enzyme at rates faster than in buffer alone. D-Glucose has been shown to bind relatively strongly to the galactosyl enzyme and to give transfer products when present during the enzyme-catalyzed hydrolysis of PGalONP.69 Sinnott and Viratelle found that the rate of solvolysis of PGalDNP at saturating concentration by E. coli lac(Z) P-galactosidase increased approximately two-fold in the presence of 1 M methanol.65 This rate acceleration in the presence of a better nucleophile than water was interpreted as evidence that degalactosylation of the enzyme was rate-determining for this substrate. These observations are wholly consistent with the reactivation data obtained in this study, which provides good evidence for a catalytically competent 2-deoxy-2-fluorogalactosyl enzyme intermediate that has not perturbed the normal mechanism other than by slowing the rates of glycosylation and deglycosylation. Although Ktrans or ktrans values for methanol or glucose were not determined in this study, accelerated transglycosylation rates with these compounds compared to hydrolysis 46 are evident. These rate increases probably arise from several factors: 1) greater nucleophilicity of the alcohol or sugar hydroxyl relative to water, 2) favourable binding of the acceptor by the "glucose" subsite to orient the nucleophile for attack and; 3) utilization of the binding energy from binding of the acceptor at the "glucose" subsite to stabilize the transition state for degalactosylation. The greater nucleophilicity of methanol, which is unlikely to bind strongly to the "glucose" site, probably accounts for the modest increase in turnover of the glycosyl enzyme compared to hydrolysis. Glucose itself, however, is more likely to utilize specific binding contacts at the secondary site to stabilize the transition state for the cleavage of the glycosyl-enzyme bond. Support for this notion is provided by comparison of k t^ values for the enzymic hydrolysis of |3GalPNP (163 s-1), for which glycosylation is though to be rate-limiting, with enzymic hydrolyses of lactose (66 s-1) and allolactose (102 s - 1).7 2 Though the leaving group ability (in terms of pKa) of glucose differs from that of 4-nitrophenol by many orders of magnitude, the rates of enzymic hydrolyses of the disaccharides are only 40 to 60% lower than PGalPNP. Note that kobs for the turnover of the 2-deoxy-2-fluoro galactosyl enzyme in buffer alone with water acting as a glycosyl acceptor represents a kcat value for 2F(3GalDNP since k o b s = k'3 = k3[H20] and kcat = k2ky(k2 + k'3) = 2.8 x 10"6 s-1. Recall that for inactivation of glycosidases by 2-deoxy-2-fluoro glycosides, k 2 is equivalent to the kj value, which was determined earlier. Comparing the kcat value of 2Fp,GalDNP above with the kcat of the unsubstituted 2',4'-dinitrophenyl galactopyranoside (1100 s*1; see Table 6, p. 56) provides a measure of the decrease in deglycosylation rate caused by placement of the fluorine atom at C(2) since deglycosylation is rate-limiting for both compounds. The k c a t of the 2-deoxy-2-fluoro galactopyranoside is thus reduced approximately 108 -fold. This vast decrease in deglycosylation rate probably arises from two factors. First, inductive de stabilization of an oxocarbonium ion-like transition state by the electron withdrawing effect of fluorine. Second, poorer binding of the 2-deoxy-2-fluorogalactosyl 47 enzyme intermediate at the transition state, possibly by disruption of specific non-covalent interactions (see Chapter 1, Section 3.4). A transition state with substantial positive charge would be expected to be inductively destabilized by the electron withdrawing effect of an adjacent fluorine. It has been, shown that the rates of spontaneous hydrolysis of 2',4'-dinitrophenyl glycopyranosides are independent of pH between pH 2 and 8 and that the hydrolyses proceed via oxocarbonium ion-like transition states.73 Fluorine substitution in the pyranose ring at C(2) results in a 20-fold decrease in the spontaneous hydrolysis rate of 2F|3GalDNP at 25°C relative to the parent compound (Table 7, p. 61). Clearly, the rate reduction arising from purely inductive destabilization of an electron-deficient transition state in the non-enzymic reaction cannot alone account for the 108-fold decrease in the enzymic rate. Some other factor, probably loss of specific binding involving the C(2) hydroxyl at the transition state must also contribute to the observed rate reduction. Note however that the transition states for the enzymic and spontaneous reactions may differ considerably in terms of charge development and hence the magnitude of the electronic effects of substitution on the different transition state structures may vary. Additional evidence that this rate reduction arising from substitution at the 2 position does not simply involve inductive destabilization of electron deficient transition states alone is provided by studies of the enzymic hydrolysis of 1-deoxyglycerol-l-yl-2-deoxy-P-/yxo-hexopyranoside 57 and its galacto analog 56 by Brockhaus et al. 7 4 (Table 5). It can be seen that deletion of the 2-hydroxyl reduces the rate of |3-galactosidase-catalyzed hydrolysis of 57 some lO^fold compared to 56. Note that on purely electronic grounds, the 2-deoxy compound 57 would be expected to be hydrolyzed faster than the galactopyranoside 56 with its C(2) hydroxyl intact. That this reduction is wholly due to poor ground state binding is unlikely by comparison of the Ki values for the corresponding hexitols 59 and 58. These results underline the importance of the C(2) hydroxyl to catalysis by the enzyme. 48 Substrate Relative rate Inhibitor Kj (mM) of enzymic hydrolysis (urrjolL^min-1) 0.0074 57 59 69.6 Table 5. Effect of deoxygenation at the 2 position on the hydrolysis of 1-deoxygtycerol-1-yl-P-D-galactopyranoside 56 by E. coli (lacZ) fi-galactosidase.* * Data from M. Brockhaus et al. (1979) Carb Res. 69,264 . 5. Dinitrophenolate burst and active site titration Accurate determinations of the concentration of enzyme in a sample often rely on spectrophotometric methods which depend on knowledge of the extinction coefficient of the enzyme at a particular wavelength and will often overestimate the enzyme concentration if the sample is not homogenous. However, with some enzymes, the concentration of active enzyme may be fetennined by measuring an initial burst of product. Such a burst will occur when the first step in a two-step process releases one product while the second step is slow, allowing an enzyme-bound intermediate to accumulate. One mole of enzyme will initially react with one mole of substrate to form one mole of the first product and one mole of the enzyme-bound intermediate. If this intermediate turns over very slowly, then the concentration of active enzyme may be determined from the amount of the first product released. This is, of course, the kinetic scheme for inactivation of p-galactosidase by 2FPGalDNP (see Figure 22). Such pre-steady state, burst phase kinetics have been observed in P-glycosidases previously. Mg2+-free E. coli (lacZ) P-galactosidase at 4°C showed a burst with PGalDNP 7 5 and a burst observed with inactivation of pABG5 P-glucosidase by 2FPGluDNP was used to quantitate the amount of enzyme present.55 In a similar experiment described below, the concentration of E. coli (lacZ) P-galactosidase in a sample was determined upon inactivation by 2FPGalDNP. The enzyme concentration in a dialyzed, homogenous sample was determined by absorbance at 280 nm (e = 2.1 mg"1 mL cm"1).76 A fifty-fold excess of 2FPGalDNP (0.41 mM) was added to the enzyme sample and the release of 2,4-dinitrophenol was followed at 400 nm for approximately 45 minutes at which time the change in absorbance showed a slight linear increase with time (Figure 28). The initial exponential rate of 2,4-dinitrophenol release (kobs = 0.19 min-1) is within 20% of the calculated rate of inactivation at this concentration of 2FPGalDNP (kobs = 0.23 min-1; see Figure 23 A), p. 40) confirming, within experimental error, that the inactivation and release of dinitrophenolate are associated with the same event The concentration of phenol released in the initial burst was determined by back-extrapolation of the linear portion of the curve. A plot of protein concentration (from absorbance at 280 nm) vs. 2,4-dinitrophenol concentration is linear (r 50 = 0.9990) with a slope of 0.9, indicating, within experimental error, approximately one mole of 2,4-dinitrophenol released per mole of enzyme (Figure 29). 0.07 B e © m 0.03 -\ 1 1 1 1 1 1 1 1 1 0 10 20 30 40 50 time (min) Figure 28. 2,4-Dinitrophenolate burst observed upon incubation ofE. coli (lacZ) fi-galactosidase with 0.41 mM 2FfiGalDNP in 50 mM sodium phosphate, 1 mM Mg2+,pH7.00at25°C. 3 Protein (nmol) Figure 29. Stoichiometry of 2,4-dinitrophenol released from fi-galactosidase incubated with 2F($GalDNP. 5 1 This near 1:1 stoichiometry is further evidence of the active site-directed nature of the inhibitor. Burst phase kinetics from 2Fj3GalDNP are readily observable at room temperature though the relatively slow rate of inactivation (and hence 2,4-dinitrophenolate release) and the significant rate of spontaneous hydrolysis of the galactoside are experimental drawbacks to the routine application of this method to active site titration of the enzyme. 6. Modified sugars as transition state probes Recall that non-covalent interactions between enzyme and substrate at the transition state are proposed to be primarily responsible for the rate acceleration of an enzyme-catalyzed reaction.23 In glycosyl transferases, these interactions are likely to be primarily due to hydrogen bonding between appropriate groups in the enzyme active site and the hydroxyl groups of the glycone. Hence deoxygenation (substitution of OH for H) at a particular position on the glycopyranose ring should result in deletion of any hydrogen bonding at that position between the glycone and the enzyme at the transition state. Substitution by fluorine, which can arguably accept a hydrogen bond but cannot possibly donate one, may still bind appreciably if the glycone normally accepts a hydrogen bond from the enzyme at that position. Deoxygenation or fluorination of the glycopyranose ring can also affect the rate of glycoside hydrolysis through purely electronic effects. Deoxygenation will tend to stabilize the developing oxocarbonium ion, increasing the rate of hydrolysis, while fluorination will have the reverse effect. Thus analysis of the kinetics of hydrolysis of a series of substrates with these modifications can provide information about both the electronic nature of the enzymic transition state and non-covalent interactions between the glycone and the enzyme active site (see Chapter 1, Section 4). 52 Hydrogen bonding energetics in water. Evaluation of the contribution of hydrogen bonding to the binding energy between substrate and enzyme is complicated by the fact that such binding involves an exchange of hydrogen bonds between substrate plus water and substrate plus enzyme, i.e. (3) E-H—OH2 + HO-H—A-S " [E-H—A-S] + HOH—OH2 where the enzyme donates a hydrogen bond to an acceptor, A on the substrate, S. The hydrogen bonding roles of the enzyme and substrate could, of course, be the reverse of that shown. There is no change in the number of hydrogen bonds upon binding (though bond geometries may differ considerably) but there is an increase in entropy associated with the release of an enzyme-bound water molecule into bulk water which gives rise to the favourable binding energy of the resulting hydrogen bond in the enzyme-bound complex.23 Deletion of the hydrogen bonding group from the substrate will probably not result in loss of the full hydrogen bonding energy because no hydrogen bond could exist between water and the substrate before binding, i.e. (4) E-H—OH2 + HOH + S " [E-H S] + HOH—OH2 Any free energy difference between the two systems represented by Equations (3) and (4) will be equal to the difference in hydrogen bond strengths between enzyme plus substrate and enzyme plus water. A hydrogen bond between a substrate and an enzyme that has evolved to specifically bind the substrate might well be expected to be stronger than that between a protein residue and a free water molecule in solution. While deoxygenation results in deletion of any hydrogen bond between the enzyme and substrate at that position, substitution of a hydroxyl by a fluorine atom results in two possible scenarios. In Equation (5), the enzyme donates a hydrogen bond which the 53 fluorine may accept, while in Equation (6) the enzyme accepts a hydrogen bond and therefore cannot interact favourably with the fluorine. (5) E-H + HO-H—F-S " [E-H—F-S] + HO-H— OH 2 E-A—H-OH + HO-H—F-S - [E-A F-S] + HO-H—OH2 ( 6 ) In Equation (6), as the fluorine on the substrate may still act as a hydrogen bond acceptor with water in solution, a net loss of hydrogen bonds results. In this case, the deoxy analog may bind more tightly than the deoxyfluoro substrate. Kinetic parameters of modified 2',4'-dinitrophenyl glycopyranosides. The series of deoxy and deoxyfluoro 2',4'-dinitrophenyl glycosides synthesized for this study were all substrates for E. coli (lacZ) P-galactosidase with the exception of 2FPGalDNP which undergoes degalactosylation sufficiently slowly to be an excellent inactivator (see Sections 3 and 4). Rates of hydrolysis were determined spectrophotometrically by measuring the rate of release of 2,4-dinitrophenol. The data so obtained were fitted to the Michaelis-Menten equation using a weighted least squares non-linear regression program and the kinetic constants kcat and K m thus determined (Table 6). The data are presented graphically according to the method of Lineweaver and Burk77 for convenient visual inspection only. These plots were not used to determine kcat and K m values due to the non-linear error span (Figures 30 to 36). The enzyme concentrations of the diluted stocks used for each determination are shown in the legends of Figures 30 to 36. Because of low solubility in buffer or high background absorbance, the kinetic parameters of two substrates (6FpGalDNP and 4FpGalDNP, respectively) were determined at substrate concentrations ranging from approximately one-tenth the reported K m to only slightly above the reported K m . Other substrates (4dPGalDNP and PGluDNP) bound sufficiently poorly that plots of rate vs. [S] 54 were linear over the substrate concentration range observed. In these cases, approximate values of kcat/Km were estimated from the Michaelis-Menten equation with [S] « K m v = kcatrEn1[Sl Km Kinetic parameters of the remaining substrates were generally detennined using substrate concentrations bracketing Km, ranging from one-seventh to seven times Km-55 2',4'-dinitro-phenyl glyco-pyranoside a Km (mM) kcat* (s-1) kcat/Km (s-imM-1) AAG 0 */ (kcal mol-1) PGalDNP 0.14 +/- 0.01 1100 7900 — 6d(JGalDNP 4.2 +/- 0.5 c 23 5.4 4.3 4d|3GalDNP — — 16 <* 3.7 3dpGalDNP n.d.e n.d* n.d.e 2dpGalDNP n.d.* n.d.e n.d.e 6FpGalDNP 0.61 +/- 0.06 5.5 9.0 4.0 4FpGalDNP 2.9 +/- 0.3 2.2 0.76 5.5 3F|3GalDNP 1.2+/- 0.1 0.63 0.53 5.7 2Fp*GalDNP Kj = 0.89 +/-0.02 ki = 0.011 kcat = 2.8 x 10"6 ki/Ki= 0.012 7.9 pGluDNP 0.13* Table 6. Kinetic parameters of2',4'-dinitrophenyl B-D-glycopyranosides with E. coli GacZ) B-galactosidase. a All measurements were made in 50 mM sodium phosphate buffer, 1 mM M g 2 + , pH 7.00 at 25°C. * kcat values are reported relative to (JGalPNP Qccat = 156 s"1) in accordance with Sinnott and Souchard13 and are + / - approximately 10%. c The maximal concentration of 6dPGalDNP in buffer is approximately 40% of the reported K m . d Saturation of enzyme not possible due to very poor binding of substrate, i.e. [S] « K m . e Not determined. /AAG 0 * = - RTln{0Ccat/Km)2/ Occat/Km)l} where 0£cat/Km)2 is & e value for the analog and O c^at/Km)! is the value for pGalDNP. 56 1000 1/[S] (1/mM) Figure 30. Hydrolysis of 2',4'-dinitrophenyl (5-D-galactopyranoside. Enzyme concentration -224 x 10~3 mg mL'1. 2000 s E c E 1000 -10 1/[S] (1/mM) Figure 31. Hydrolysis of 2',4'-dinitrophenyl 6-deoxy-R-D-galactopyranoside. Enzyme concentration = 1.79 mg mL'1. 57 8.00e-3 -4.24e-22 [S] (mM) Figure 32. Hydrolysis of 2',4'-dinitrophenyl 4-deoxy-fi-D-xylo-hexopyranoside. Enzyme concentration = 6.75 x lO"1 mg mLr1. 4000 _ 3000 -E E 2000 -i - 1000 -1/[S] (1/mM) Figure 33. Hydrolysis of 2',4'-dinitrophenyl 6-deoxy-6-fluoro-R-D-galactopyranoside. Enzyme concentration = 9.10 JC i(W mg mLr1. 58 0 I • 1 ' 1 ' 1 ' 1 1 — 0 2 4 6 8 10 1/[S] (1/mM) Figure 34. Hydrolysis of 2',4'-dinitrophenyl 4-deoxy-4-fluoro-fi-D-galactopyranoside. Enzyme concentration - 156mg mL'1. o H — i — i — • — i — i i i i • i i 0 2 4 6 8 10 12 1/[S] (1/mM) Figure 35. Hydrolysis of2',4'-dinitrophenyl3-deoxy-3-fluoro-fi-D-galactopyranoside. Enzyme concentration = 7.40 xlO"1 mg mLr1. [S] (mM) Figure 36. Hydrolysis of 2',4'-dinitrophenyl B-D-glucopyranoside. Enzyme concentration = 5.70 mg mL'1. Intrinsic electronic and other effects on glycoside hydrolysis. The double displacement mechanism proposed for the hydrolysis of aryl glycosides by glycosidases (see Figure 2, p. 3) involves transition states with considerable oxocarbonium ion character. On the basis of inductive effects alone, substitution of electronegative fluorine for a hydroxyl on the glycopyranose ring should destabilize these electron-deficient transition states while deoxygenation should have the reverse effect. The effect of such substitutions on the spontaneous, non-enzymic rates of hydrolysis, which have also been shown to proceed via oxocarbonium ion-like transition states73, were determined in order to interpret the effects on enzymic rates. The following spontaneous hydrolysis data were obtained from M. Namchuk, another worker in this laboratory (Table 7). 60 2\4'-dinitrophenyl glycopyranoside k(hvdrolvsis) x 10'6 at 25°C (s'1) PGalDNP 4.63 6dpGalDNP 29.3 4dpGalDNP 21.8 6FPGalDNP 1.72 4FpGalDNP 0.662 3FpGalDNP 1.68 2FPGalDNP 0.235 pGluDNP 0.870 Table 7. Spontaneous hydrolysis of a series of deoxy and deoxyfluoro 2',4'-dinitrophenyl B-D-glycopyranosides.a-b a Data obtained from M. Namchuk. b Hydrolyses were performed in 25 mM sodium phosphate buffer, 0.8 M KC1, pH 6.5 at various temperatures ranging from 15°C to 85°C., depending on the rate of hydrolysis of the glycoside. Measurements were made at three temperatures for each glycoside and the value of k^ hydrolysis) at 25°C determined by extrapolation or interpolation from the resulting Arrhenius plots. It is clear from the above data that the fluorinated derivatives are all hydrolyzed more slowly than the parent galactopyranoside, while the deoxy analogs are hydrolyzed faster. This is consistent with involvement of an oxocarbonium ion-like transition state which is inductively stabilized or destabilized depending on the electronic nature of the substituent. However, with the deoxyfluoro compounds, the magnitude of the rate constant does not simply reflect the distance of the fluorine atom from the anomeric centre since the observed order is 6-fluoro > 3-fluoro > 4-fluoro > 2-fluoro. This is the same order that was observed by Street et al. in the acid-catalyzed hydrolysis of a series of monofluorinated a-D-glucopyranosyl phosphates,78 which also hydrolyze with C-0 bond cleavage via oxocarbonium ion-like transition states between pH 2 and 5. It was proposed 61 by these authors that a combination of inductive and dipolar effects, electronic in nature, accounted for the observed order of hydrolysis rates, in particular the larger rate constant for the 3-deoxyfluoro derivative than for the 4-deoxyfluoro derivative. The reverse order was predicted for the deoxy analogs, i.e. 2-deoxy > 4-deoxy > 3-deoxy > 6-deoxy, and was in fact observed with the ce-D-glucopyranosyl phosphate derivatives.79 However, the factors determining the effect of substitution on the rates of hydrolysis of these and similar systems involving glycopyranosyl cation-like transition states have not yet been elucidated unambiguously. The same order is not observed for the deoxy galactopyranosides synthesized in this study. However, it has been shown80 that axial substituents which are better accomodated for steric reasons in the more planar half-chair conformation of the oxocarbonium ion transition state than the chair conformation of the ground state (Figure 37) will result in an increase in hydrolysis rate. Thus pGalDNP hydrolyzes some Figure 37. Chair to half-chair transition in the hydrolysis of a P-D-galactopyranoside via an oxocarbonium ion intermediate. five-fold faster than the corresponding gluco analog. This is sufficient to affect the order since the 4-deoxy derivative, which lacks an axial substituent at C(4), would be expected to hydrolyze more slowly relative to the other deoxy galactopyranosides. The predicted order HO 62 of hydrolysis rates will then be 2-deoxy > 3-deoxy > 6-deoxy > 4-deoxy > parent. This order is observed for the limited available deoxy analogs and the failure to isolate the 3-deoxy derivative is in itself an indication of its evident instability. Binding effects. Comparison of the data in Tables 6 and 7 reveals that the decreases in enzymic rate observed upon replacement of a hydroxyl by a fluorine in the deoxyfluoro galactopyranosides (103 to 108-fold) are significantly greater than the decreases in the rate of spontaneous hydrolysis of the same compounds (3 to 20-fold), indicating that some factor which only affects the enzymic rate contributes to the bulk of the observed rate reductions. As in the case of the 2-deoxyfluoro derivative (discussed in Section 4), this factor is probably related to the loss of specific transition state binding interactions. The importance of such binding interactions to catalysis is perhaps best probed by examining the data for the deoxy glycopyranosides. Purely on the basis of electronic effects, the deoxy substrates would be expected to undergo enzymic hydrolysis faster than the parent galactopyranoside, while in fact they are hydrolyzed more slowly by the enzyme. Recall that kcat/Km represents the change in free energy from free enzyme and free substrate to the transition state for the first irreversible step (Figure 21, p. 36). Thus the effect of modifications on a substrate that alter the stability of this transition state will be reflected in changes in kcat/Km. Neglecting for the moment other factors which may affect the structure of this transition state, comparison of kcat/Km for the parent compound with that for the modified substrates gives a measure of the energetic contribution to the stability of the transition state due to enzyme-substrate interactions at that position. These values are expressed as Gibbs free energies (AAG0*) in Table 6 (see Reference 23, p. 350). However, as noted above, the intrinsic electronic effects of the substituents affect transition state stability and also contribute to the value of AAG0*. With the deoxy analogs, these electronic effects will only serve to accelerate the reaction, thus the observed rate decreases are likely due solely to transition state binding interactions and the AAG0* values are minimum estimates of the loss of apparent binding energy resulting from deletion of the hydroxyl at that position. It has been proposed in a similar study with glycogen phosphorylase28, that such values are a measure of the difference in hydrogen bond strengths between enzyme plus substrate at the transition state (of the first irreversible step) and the unbound substrate plus water. The apparent binding energies are approximately 4 kcal mol"1 at the 4 and 6 positions. These values are consistent with estimates by Fersht23 that a strong hydrogen bond between one neutral and one charged group may contribute 3 kcal mol"1 or more to the overall binding free energy. Thus, assuming a deleted hydrogen bond is responsible for the bulk of the loss of binding energy, these results may suggest one strong, or several weak, hydrogen bonds at each of the 4- and 6-positions at the transition state. Unfortunately, no data is available for the 2- or 3-deoxy derivatives. Because the AAG0* values for the deoxyfluoro substrates cannot be simply interpreted in terms of hydrogen bonding, it is not possible to assign hydrogen bond polarities at each position. It is noteworthy that the gluco analog, PGluDNP, which differs from the corresponding galactopyranoside only in the orientation of its C(4) hydroxyl, is hydrolyzed by the enzyme some 1 (refold slower than PGalDNP. In contrast, the 4-deoxyfluoro analog, in which fluorine has been substituted for hydroxyl while retaining the galacto configuration, suffers a smaller reduction in rate even with a more unfavourable intrinsic electronic factor. It is likely that the great reduction in the rate of PGluDNP hydrolysis is due not simply to a loss of favourable hydrogen bonding interactions alone but to a severe steric clash in the active site as a hydroxyl group is thrust into a space normally occupied by a hydrogen atom. 64 Electronic effects. It is thought that the effect of substitution of hydrogen or fluorine for hydroxyl on the spontaneous hydrolysis rates of 2',4'-dinitrophenyl galactopyranosides is primarily electronic in nature.73,78'79 Fluorine substitution decreases the hydrolysis rate while deoxygenation leads to a marked increase. If the enzymic reaction proceeds through a transition state similar to the non-enzymic process and the electronic effect of the substitutions was the only factor affecting the free energy of the transition state, then a similar relationship between hydrolysis rate and substituent should also be seen with the enzymic reaction. This linear free energy relationship may be expressed as kA = CkBP where k A and ke are the rate constants for the two reactions, A and B, while C and p are constants. From the Eyring equation k A = (kT/h)exp{-AG*A/RT} and kB= (kT/h)exp{-AG*R/RT} where AG* is the activation energy for reaction of A or B. Thus plotting the logarithm of the rate constants for the two processes against each other provides information about the similarity of the transition states involved. The correlation coefficient (r) reflects the degree of certainty in the relationship, while the slope (P) of the line is a measure of the similarity of the two transition states in terms of charge development or some other parameter. However, we have already seen evidence for the importance of binding effects in catalysis by E. coli (lacZ) p-galactosidase. In addition to the electronic effects of deoxygenation and fluorine substitution on transition state stability, these modifications are also likely to perturb enzyme-substrate binding at the transition state. Thus the electronic effects on transition state stability will be superimposed upon perturbations of transition state energy due to altered binding interactions. A linear free energy relationship considering electronic factors alone may then be masked by binding effects. Street et al. 2 8 found a reasonable linear free energy correlation (r = 0.9) between logCVmax) of the enzyme-catalyzed hydrolysis of a series of deoxy and deoxyfluoro a-D-glucopyranosyl phosphates and log(k) of the acid-catalyzed hydrolysis of the same compounds. Since the acid-catalyzed reaction is known to proceed via an oxocarbonium ion-like transition state, these results provided good evidence that the enzymic process involves a similar transition state at the rate-limiting step of the reaction. A plot of log(Vmax/Km) vs. log(k) for the same compounds was essentially a scatter plot. This scatter was interpreted as being due to a greater contribution from binding interactions in this plot, since Vmax/Km represents the overall activation free energy from free enzyme and free substrate to the enzyme-bound transition state of the first irreversible step, while kcat represents the energy difference between an enzyme-bound intermediate and the enzyme-bound transition state of the rate-determining step (see Figure 21, p. 36). Thus in the latter case, some of the contribution of the binding energy is in effect cancelled. A similar treatment was applied to the data in this study and the results are shown in Figure 38. The plot of log(kCat/Km) log(khydrolysis) shows a correlation (r = 0.80), indicating that the electronic nature of the enzymic and non-enzymic transition states is similar. Since the spontaneous process has been shown to proceed via oxocarbonium ion-like transition states73, these data are consistent with an enzymic mechanism involving a similar transition state at the galactosylation step. Recall that kcat/Km is associated with the energy barrier between free enzyme and free substrate and the transition state of the first irreversible step (presumably galactosylation). 66 W V / v 1 2^ - 6 -j 1 1 • 1 • -7 -6 -5 -4 k>g(k hydrolysis) Figure 38. Linear free energy relationships between kinetic parameters for the B-galactosidase-catalyzed reaction and the spontaneous hydrolysis of a series of deoxy and deoxyfluoro 2',A'-dinitrophenyl glycopyranosides. A) Plot of logikcatlKjn) vs. the logarithm of the spontaneous hydrolysis rate constant; B) plot ofbg(kcat) vs. the logarithm of the spontaneous hydrolysis rate constant. 67 That a correlation is observed between log(kcat/Km) and log(knydroiysis) in Figure 38 A), even with interferences from binding effects, might be interpreted as suggesting that binding effects are relatively unimportant to stabilization of this transition state. However, this is unlikely in light of the vast decreases in rate observed even with the deoxy substrates. The very fact that the correlation is not perfect suggests the presence of transition state binding interactions not present in the non-enzymic reaction. Note that scatter due to binding effects will occur if the magnitude and polarity of the contributions to transition state stabilization due to binding at each position are significantly different from each other. If, for example, hydrogen bonding interactions at each position contributed approximately the same stabilization to the transition state then a linear free energy correlation reflecting electronic effects might still be observed among the deoxy substrates. However, the ability of the deoxyfluoro substrates to accept hydrogen bonds from appropriately positioned donor residues and consequent possible rate increases for these substrates may introduce considerable scatter in a correlation reflecting electronic effects alone. Nevertheless, an efficient enzyme would be expected to utilize potential hydrogen bonding interactions at all positions as fully as possible in order to stabilize the transition state. Specific changes in the geometry of the substrate as the transition state is approached, however, impose certain restrictions on the relative importance of binding interactions at a given position. Thus, in the chair to half-chair transition of a developing oxocarbonium ion (Figure 37), interactions at C(2), where a considerable amount of restructuring is occuring at the transition state, would be expected to be more important. However, in a highly efficient enzyme, these differences would likely be minimized because contributions at other positions would be utilized to their fullest potential. The plot of log(kcat) vs. log(knydroiysis)m Figure 38 B) shows a similar correlation (r = 0.78), if the point for the 2-deoxyfluoro deriviative is excluded. Recall that kcat is associated with the activation energy for the rate deterrnining step (degalactosylation for the parent galactopyranoside and the turnover of the 2-deoxy-2-fluorogalactosyl enzyme). The 68 essential difference between the two plots arises from the very slow turnover of the 2-deoxy-2-fluorogalactosyl enzyme, reflected by a very large decrease in kcat. This may be accounted for by the loss of strong transition state binding interactions with the C(2) hydroxyl which are more important at the transition state for degalactosylation than at the transition state for galactosylation, or perhaps more likely, that the transition state for degalactosylation has more oxocarbonium ion character and is inductively destabilized to a greater extent by the adjacent fluorine atom-Excluding the data for the 2-deoxyfluoro derivative from both plots, the slopes are 0.60 for the log(kcat/Km) plot in Figure 38 A) and 0.73 for the log(kcat) plot in Figure 38 B). This suggests that both the transition state associated with galactosylation and the transition state associated with degalactosylation have significantly less oxocarbonium ion character than the transition state for spontaneous hydrolysis. This is consistent with participation of an enzymic nucleophile which can pre-asssociate with a developing oxocarbonium ion-like species. A greater degree of oxocarbonium ion character at the transition state for degalactosylation as suggested above may be accounted for by preassociation of the nucleophile to a greater extent at the galactosylation transition state. This is reasonable since the nucleophile in the galactosylation step, a potentially charged carboxylate, can readily interact with the developing positive charge at the anomeric centre. In the degalactosylation transition state, the nucleophile is a neutral water and hence is likely to preassociate to a lesser extent with a developing oxocarbonium ion-like species. This interpretation is supported by the a-deuterium kinetic isotope effect data of Sinnott and Souchard13 as follows. Aryl galactosides with activated leaving groups (e.g. 2,4-dinitrophenolate and 3,5-dinitrophenolate), where degalactosylation is rate limiting (as shown by methanol competition experiments65), exhibited larger kinetic isotope effects than galactosides with less activated leaving groups (e.g. 4-nitrophenolate), where galactosylation is thought to be the rate determining step.81 69 7. Conclusions 2FPGalDNP was found to be an effective mechanism-based inactivator of E. coli (lacZ) p-galactosidase. This compound thus joins a class of 2-deoxy-2-fluoro glycosides which inactivate "retaining" glycosidases by the accumulation of a stable glycosyl-enzyme intermediate. The active site-directed nature of 2FPGalDNP was shown by protection against time-dependent inactivation by a competitive ligand and the near 1:1 stoichiometry of dinitrophenolate release with enzyme inactivation. Furthermore, when freed from excess inactivator the 2-deoxy-2-fluorogalactosyl enzyme intermediate turned over slowly in buffer and exhibited enhanced rates of reactivation in the presence of the acceptors methanol or glucose, providing strong evidence that the intermediate is catalytically competent A series of deoxy and deoxyfluoro analogs of PGalDNP was prepared for this study. All of the compounds, with the exception of 2FPGalDNP which undergoes degalactosylation sufficiently slowly to be an excellent inactivator, were substrates for E. coli (lacZ) P-galactosidase although the rates of enzymic hydrolysis were two to four orders of magnitude slower than have been measured for pGalDNP. These large rate reductions are thought to result primarily from the loss of important transition state binding interactions due to the substitution of a hydrogen or a fluorine for a hydroxyl at a given position. These results strongly suggest that much of the catalytic power of the enzyme is derived from non-covalent interactions between the enzyme active site and the galactopyranose ring of the substrate. A linear free energy relationship was shown to exist between the logarithm of kcat/Km for the enzyme-catalyzed reaction and the logarithm of the first order rate constant for the spontaneous hydrolysis of the same series of deoxy and deoxyfluoro glycopyranosides, khydrolysis- Since the spontaneous process has considerable oxocarbonium ion character at the transition state, this suggests that the enzymic transition 70 state associated with kcat/Km (presumably representing the galactosylation step) is also electron deficient and is affected by the electronic effects of the substituents in a manner similar to that observed for the spontaneous process. A similar correlation is observed in a plot of the logarithm of kcat vs. the logarithm of khydrolysis. if the point for the 2-deoxyfluoro derivative is excluded (kcat represents the degalactosylation step for the parent compound, the turnover of the 2-deoxy-2-fluorogalactosyl enzyme and presumably represents degalactosylation for the other glycosides as well). While it is possible that galactosylation has become rate determining for some substrates, this seems unlikely since both the fastest and slowest substrates (the parent compound and the 2-deoxyfluoro derivative, respectively) exhibit rate-limiting degalactosylation. The very small value of kcat for the 2-deoxyfluoro derivative is significant, and when interpreted in light of the linear free energy relationships discussed above, may suggest that some binding or electronic factor associated with substition of fluorine at C(2) destabilizes the transition state for degalactosylation to a greater extent than the transition state for galactosylation. 71 Chapter in MATERIALS AND METHODS 1. Synthesis 1.1 General methods and materials Melting points (m.p.) were determined on a Laboratory Devices Mel-temp U melting point apparatus, and are uncorrected. *H-nuclear magnetic resonance (nmr) spectra were recorded on 300 MHz Varian XL-300 or Bruker 400 MHz WH-400 instruments. Chemical shifts are given in the delta (5) scale. With samples in C D C I 3 or C D 3 O D , tetramethylsilane (TMS) was used as an internal reference (5 = 0.00 ppm). Samples in D 2 O are referenced to external 2,2-dimethyl-2-silapentane-5-sulphonate (8 = 0.00 ppm). 19F nmr spectra were recorded on a 200 MHz Bruker AC-200 instrument and are proton decoupled. Chemical shifts are given in the delta (8) scale with reference to C F C I 3 (8 = 0.00 ppm). Signals upfield of this resonance are assigned a positive value. External trifluoroacetic acid (8 = 76.53 ppm) was used as reference. Micro-analyses were performed by Mr. P. Borda, Microanalytical laboratory, University of British Columbia, Vancouver, B.C. Solvents and reagents used were either reagent grade, certified or spectral grade. Dry solvents or reagents where indicated were prepared as follows: Methanol was distilled from magnesium methoxide prepared in situ by reaction of methanol with magnesium turnings. Acetonitrile, dichloromethane and pyridine were distilled from calcium hydride. Tetrahydrofuran (THF) was distilled from sodium benzophenone ketyl. Dimethyl sulphoxide (DMSO) was distilled from sodium hydroxide under reduced pressure. N,N-dimethyl formamide (DMF) was pre-dried over calcium sulphate then distilled from 4 A molecular sieves. Acetyl chloride was dried by refluxing several hours over phosphorus pentachloride, followed by distillation. Thin-layer chromatography (tic) separations were performed using Merck Kieselgel 60 F-254 analytical plates. Compounds were visualized (where possible) under UV light or by charring with 10% sulphuric acid in methanol. Column chromatography was performed according to the method of Clarke-Still et al.82, u s m g silica gel columns of Kieselgel 60 (180-230 mesh). The following solvent systems were commonly employed: Solvent A: 1:1 ethyl acetate/ hexanes; Solvent B: 5% methanol/ ethyl acetate; Solvent C: 7:2:1 ethyl acetate/ ethanol/ water. Several compounds or precursors in this work were obtained from others: From Dr. D. Dolphin; 3)4,6-tri-0-acetyl-2-deoxy-2-fluoro-a-D-galactopyranosyl fluoride. From Dr. A. Becalski; 2',4'-dinitrophenyl 4-deoxy-4-fluoro-p-D-galactopyranoside. From M. Namchuk; 2',4'-dinitrophenyl P-D-glucopyranoside and 2',4'-dinitrophenyl 4-deoxy-p-D-xy/o-hexopyranoside. From I. Street; methyl 2,3,6-tri-0-benzoyl-4-deoxy-4-chloro-a-D-glucopyranoside and methyl 2,3,6-tri-0-benzoyl-4-0-methanesulphonyl-a-D-galactopyranoside. 1.2 General procedures 1.2.1 Fluorination To a 3-necked round bottom flask fitted with a pressure equalized addition funnel was added the suitably protected carbohydrate derivative and the appropriate base dissolved in dry dichloromethane. The apparatus was flushed with dry nitrogen and cooled to - 20°C (carbon tetrachloride/ dry ice bath). Diethylaminosulphur trifluoride (DAST, Aldrich Chemical Co.) was then slowly added over a period of 15 min. After the addition was complete, the mixture was allowed to warm slowly to ambient temperature. Upon completion, the reaction mixture was cooled to -20°C and quenched with addition of excess methanol. Solvent was removed in vacuo , the resulting gum was dissolved in dichloromethane, washed with a saturated sodium bicarbonate solution, dried (MgS04) and purified by column chromatography. 1.2.2 Acetvlation Acetic anhydride was added dropwise to an ice-cooled solution of pyridine and the carbohydrate derivative. The mixture was stirred until the reaction was complete by tic (Solvent A) then cooled (0°C), stirred with excess methanol for 1 h and the solvent evaporated. Removal of pyridine was facilitated by several co-evaporations with toluene. The resulting oil was then washed successively with 5% hydrochloric acid, saturated sodium bicarbonate and water and dried (MgSC»4) or purified by column chromatography. 1.2.3 Selective deacetylation at the 1 position a) with 2-aminoethano^ The appropriate peracetylated carbohydrate derivative was dissolved in ethyl acetate, 2-aminoethanol was added and the resulting mixture stirred at ambient temperature. Upon completion of reaction, as indicated by tic (Solvent A), solvent was removed in vacuo, the residual oil was dissolved in chloroform, washed once with water, dried (MgS04) and purified by column chromatography. b) with hydrazine acetate^ A suspension of the appropriate peracetylated carbohydrate derivative and hydrazine acetate was heated in DMF at 50°C for ~ 3 minutes to dissolve the starting material. The mixture was then stirred ~ 20 minutes at ambient temperature. Upon completion of reaction, as indicated by tic (Solvent A), the mixture was diluted with ethyl acetate, washed with a saturated brine solution and dried (MgS04). DMF was removed by several co-evaporations with toluene and the residual oil was purified by column chromatography. 1.2.4 2'.4'-Dinitrophenvl glycopyranoside formation^ The appropriate 1-deacetylated glycopyranose derivative and 1,4 diazabicyclo[2.2.2]octane (DABCO) were dissolved separately in DMF and stirred over 4 A molecular sieves for 2 to 3 h. The DABCO solution was added to the carbohydrate derivative under nitrogen and l-fluoro-2,4-dinitrobenzene (FDNB, dried over P2O5 for 3 h) was added. The mixture was stirred for 2 h at ambient temperature, filtered to remove sieves and concentrated in vacuo to an orange syrup. This was dissolved in dichloromethane, washed with saturated sodium bicarbonate solution and water and dried (MgS04). Removal of solvent gave a brownish yellow gum which yielded a solid upon trituration with ethanol. Subsequent recrystallization from ethanol or ethyl acetate/ hexanes gave white or faintly yellow crystals of the acetylated glycopyranoside. 1.2.5 Deacetvlation with HC17 methanol The acetylated glycopyranoside was suspended in dry methanol (14 mg/mL), cooled to 0°C under nitrogen and sufficient freshly distilled acetyl chloride added to give a final HC1 concentration of 3-5% (w/v). The reaction mixture was stirred ~ 18 to 36 h at 4°C until the reaction was complete by tic (Solvent B) then the solvent was evaporated in vacuo, and the product was washed several times with anhydrous diethyl ether to remove 75 residual acid. The resulting gum was dissolved in dry methanol or acetone, then triturated with diethyl ether or diethyl ether plus hexanes to induce crystallization. 1.3 2'.4'-Dinitrophenyl glvcopyranosides 2',4'-Dinitrophenyl 2,3,4,6-tetra-O-acetyl-B-D-galactopyranoside (2) 1,2,3,4,6-Penta-O-acetyl-D-galactopyranose ( 1 . 0 0 g, 2.60 mmol) was selectively deacetylated at the 1 position with 2-aminoethanol (0.23 mL, 3.8 mmol) according to the general procedure by stirring in ethyl acetate (75 mL) at room temperature for 60 h. Column chromatography (Solvent A) afforded an anomeric mixture of tetra-O-acetates 1 as a faintly yellow foam (0.82 g, 2.4 mmol, 60%). This material ( 0 . 5 4 g, 1.6 mmol) was converted to the galactopyranoside 2 according to the general procedure by reaction with DAB CO (0.60 g, 5.5 mmol) and FDNB (0.34 g, 1.8 mmol). The resulting solid was recrystallized from ethanol to afford a white crystalline powder 2 (0.50 g, 0.97 mmol, 63%). M.p. 174-175°C (lit.38 m.p. 174-176°C); LH nmr (300 MHz, C D C I 3 ) : 8 8.72 (d, 1 H, J3',5' 3 Hz, H(3')), 8.45 (dd, 1 H, J5',6' 10, J3S51 3 Hz, H(5')), 7.50 (d, 1 H, Js',61 10 Hz, H(6')), 5.60 (dd, 1 H, J i , 2 8, J 2 3 11 Hz, H(2)), 5.50 (d, 1 H, 1 3 , 4 3 .5 Hz, H(4)), 5.23 (d, 1 H, Ji ,2 8 Hz, H(l)), 5.14 (dd, 1 H, J 2 , 3 11 . J3,4 3 .5 Hz, H(3)), 4 . 3 -4.1 (m, 3 H, H(5), H(6, 6')), 2 . 2 0 , 2.15, 2.10, 2.05 (4 x s, 12 H, 4 x OAc). Anal. calc. for C 2 0 H 2 2 O 1 4 N 2 ; C, 46.70; H, 4.31; N, 5 .45%; Found: C, 46.77; H, 4.40; N, 5 .33%. 2',4'-Dinitrophenyl B-D-galactopyranoside (3) The tetra-O-acetate 2 (300 mg, 0.585 mmol) was deacetylated according to the general deacetylation procedure for 18 h. The product crystallized upon trituration with diethyl ether and was recrystallized twice from acetone/ diethyl ether to give white powdery crystals of 3 (64 mg, 0.185 mmol, 32%). M.p. 134-1350C (dec) (lit.38 m.p. 161-163<>C); lH nmr (400 MHz, C D 3 O D ) : 8 8.70 (d, 1 H, Jyj 3 Hz, H(3')), 8.48 (dd, 1 H, Js-,6 110, J3 - .5 - 3 Hz, H(5')), 7.65 (d, 1 H, Js-,6' 1 0 Hz, H(6')), 5 . 2 2 (d, 1 H, J i ,2 8 Hz, H(l)), 4 . 2 - 3 . 2 (m, 6 H, H(2-6,6')). Anal. calc. for C 1 2 H 1 4 O 1 0 N 2 + O . 5 C H 3 O H + H 2 O ; C, 39.48; H, 4 .77 ; N, 7 .37%; Found: C, 39.15; H, 4 . 3 9 ; N, 7.05%. 3,4,6-Tri-0-acetyl-2-deoxy-2-fluoro-a-D-galactopyranosyl bromide (4) 2-Deoxy-2-fluoro-a-D-galactopyranosyl fluoride (1 .0 g, 3.2 mmol) was dissolved in 45% w/v hydrogen bromide in glacial acetic acid ( 1 0 mL), 5 drops of acetic anhydride was added and the reaction mixture was stirred at room temperature for 4 h. The mixture was dissolved in dichloromethane, washed with saturated sodium bicarbonate solution and water, and dried (MgS04). Evaporation of solvent gave 4 as a white semi-solid mass which was stored under hexanes at 0°C ( 1 . 1 0 g, 2 . 9 7 mmol, 92%). lH nmr (300 MHz, C D C I 3 ) : 8 6.62 (d, 1 H, Ji,2 4 Hz, H(l)), 5 . 5 5 (t, 1 H, J4,F 3 , J3,4 3 Hz, H(4)), 5.49 (dt, 1 H, J 3 > 4 3, J 2 , 3 10 , J 3 . F 1 0 Hz, H(3)), 4 . 7 7 (ddd, 1 H, J i , 2 4 , J 2 > 3 10, J 2 , F 51 Hz, H(2)), 4.53 (t, 1 H, J 5 , 6 ' 6 .5 , J 5 , 6 7 Hz, H(5)), 4.16 (AB multiplet, 2 H, J 6 , 6 ' 11 .5 , J 5 , 6 ' 6 .5 , J 5 , 6 7 Hz, H(6,6')), 2.16, 2.07, 2.06 (3 x s, 9 H, 3 x OAc). 1 9 F nmr ( C D C I 3 ) : 8 195.52. 2 ',4 '-Dinitrophenyl 3,4,6-tri-0-acetyl-2-deoxy-2-fluoro-B-D-galactopyranoside (6) The bromide 4 (0.58 g, 1.57 mmol) was dissolved in acetone (5 mL) containing 4 drops of water and freshly prepared silver carbonate (0.37 g, 1.34 mmol) at 0°C in the dark. The reaction mixture was stirred for 1 h at 0°C then 24 h at room temperature. The suspension was filtered, the precipitated solids washed well with acetone and the filtrate evaporated to give a yellow oil. This was purified by column chromatography (Solvent A) to give an anomeric mixture of 5 (0.32 g, 1.04 mmol, 67%), as a colourless oil. *9p nmr ( C D C I 3 ) : 8 207.38, 207.86. This material (0.30 g, 0.97 mmol) was convened to the galactopyranoside 6 according to the general procedure by reaction with DABCO (0.38 g, 3.30 mmol) and FDNB (0.22 g, 1.18 mmol). *H nmr indicated the resulting yellow syrup was a 3:1 mixture of P and a anomers. This syrup was dissolved in ethyl acetate and hexanes were added to selectively crystallize the P anomer. This material was recrystallized from ethyl acetate/ hexanes to afford white needles of 6 (0.19 g, 0.40 mmol, 42%). M.p. 160-161°C; *H nmr (300 MHz, C D C I 3 ) : 8 8.80 (d, 1 H, Jy# 3 Hz, H(3')), 8.45 (dd, 1 H, J5 . .6- 10, J 3 - , 5 . 3 Hz, H(5,))> 7.42 (d, 1 H, J 5 ' ) 6 ' 10 Hz, H(6')), 5.54 (t, 1 H, J 4,F 3 , J 3 > 4 3 Hz, H(4)), 5.39 (dd, 1 H, Ji,2 7.5, Ji,p 4 Hz, H(l)), 5.25 (dt, 1 H, 13,4 3, J 2 , 3 10, J3 .F 10 Hz, H(3)), 4.90 (ddd, 1 H , Ji,2 7.5, J2,3 9, J2.F 52 Hz, H(2)), 4.18 (m, 3 H, H(5), H(6,6*)), 2.20, 2.12, 2 . 1 0 (3 x s, 9 H, 3 x OAc). 19p nmr ( C D C I 3 ) : 8 206.74. Anal. calc. for C 1 8 H 1 9 O 1 2 N 2 F ; C, 45.58; H, 4.04; N, 5.91%; Found: C, 45.71; H, 4.10; N, 5.85%. 2',4'-Dinitrophenyl2-deoxy-2-fluoro-B-D-galactopyranoside (7) The tri-O-acetate 6 (80 mg, 0.169 mmol) was deacetylated according to the general deacetylation procedure for 33 h. The product was recrystallized from methanol/ diethyl ether/ hexanes to give white crystals of T (30 mg, 0.086 mol, 51%). M.p. 174-175°C; *H nmr (400 MHz, C D 3 O D ) : 8 8.73 (d, 1 H, J 3 ' J 5 - 3 Hz, H(3')), 8.48 (dd, 1 H, J 3 ' , 5 ' 3 , Jyfi 10 Hz, H(5')), 7.67 (d, 1 H, J5'& 10 Hz, H(6')), 5.50 (dd, 1 H, Ji,2 7.5, Ji,F 4 Hz, H(l)), 4.65 (ddd, Ji,2 7.5, J2.3 9, J2,F 52 Hz, H(2)), 4.0-3.7 (m, 5 H, H(3-6,6')). 19F nmr ( D 2 O ) : 8 216.32. Anal. calc. for C 1 2 H 1 3 O 9 N 2 F ; C, 41.39; H, 3.76; N, 8.04%; Found: C, 41.41; H, 3.80; N, 7.98% 78 12:5,6-Di-04sopropylidene-a-D-nbo-hexofuranos-3-ulose hydrate (8) l,2:5,6-Di-0-isopropylidene-a-D-glucofuranose39 (io g, 38.4 mmol) was dissolved in DMSO (100 mL) then phosphorus pentoxide (10 g, 70.4 mmol) was added slowly with ice-bath cooling. After stirring 60 h at room temperature, the mixture was diluted with dichloromethane (100 mL) and washed successively with water, saturated sodium bicarbonate solution and water, then dried (MgS04). The resulting syrup was chromatographed (Solvent A) to afford faintly yellow crystals of 8 (5.0 g, 18.1 mmol, 50%). M.p. 98-99°C (lit.40 m . p . 109-111°C); lH nmr (300 MHz, C D C I 3 ) : 8 6.15 (d, 1 H, Ji,2 4 Hz, H(l)), 4.42-4.3 (m, 3 H, H(2), H(4), H(5)), 4.10-4.00 (m, 2 H, H(6,6')), I. 47 (s, 3 H, CH3) , 1.45 (s, 3 H, CH3), 1.34 (s, 6 H, 2 x CH3). 3-O-Acetyl-l 2:5,6-di-0-isopropylidene-a-D-ery\hio-hex-3-enofuranose (9) To a solution of the ulose hydrate 8 (5.0 g, 18.1 mmol) in dry pyridine (50 mL) was added acetic anhydride (25 mL, 210 mmol). The flask was sealed and heated to 75°C for 16 h. After evaporation of the pyridine in vacuo, the resulting syrup was dissolved in dichloromethane, washed with cold 0.5 N hydrochloric acid, saturated sodium bicarbonate solution and water, charcoal-treated and dried (MgS04). After evaporation of solvent the gummy residue was crystallized from methanol/ water to give white crystals of 9 (2.53 g, 8.43 mmol, 47%). M.p. 60°C (lit.42 m.p. 620Q; lH nmr (300 MHz, C D C I 3 ) : 8 6.06 (d, 1 H, J i , 2 5.5 Hz, H(l)), 5.41 (d, 1 H, J i , 2 5.5 Hz, H(2)), 4.71 (t, 1 H, J5,6 7, J5,6' 7 Hz, H(5)), 4.09 (m, 2 H, H(6, 6')), 2.21 (s, 3 H, OAc), 1.55, 1.48, 1.45, 1.38 (4 x s,12 H, 4 x CH3). 79 12:5,6-Di-0-isopropylidene-a-D-gulofuranose (10) To an ice-cooled solution of enol acetate 9 (2.50 g, 8.33 mmol) in dry methanol (60 mL) was added quickly sodium borohydride (1.25 g, 33.0 mol). The inixture was stirred at 0°C for 30 min then at room temperature for 30 min before quenching with water. Solvent was evaporated in vacuo then the residue was dissolved in chloroform, washed once with water and dried (MgS04). Column chromatography (Solvent A) afforded 10 as a white solid (0.44 g, 1.69 mmol, 20%). M.p. 103°C (lit.43 m.p. 105-106°C); lH nmr (300 MHz, C D C I 3 ) : 8 5.79 (d, 1 H, J\2 3.5 Hz, H(l)), 4.67 (dd, 1 H, J]2 3.5, J2.3 6 Hz, H(2)), 4.48 (ddd, 1 H, J2,3 6, J3,4 6, J3 .0H 7 Hz, H(3)), 4.24 (m, 2 H, H(4), H(5)), 3.80 (AB multiplet, 2 H, J5,6 6, Js,6' 7, J6,6' 9 Hz, H(6,6')), 2.68 (d, 1 H, J3,OH 7 Hz, 3-OH), 1.63, 1.47, 1.43, 1.39 (4 x s, 12 H, 4 x CH3). 12:5,6-Di-0-isopropyliaene-3-deoxy-3-fluoro-a-D-galactofuranose (11) The partially protected gulofuranose 10 (0.44 g, 1.69 mmol) was treated with DAST (0.60 g, 3.72 mmol) and 4-(dimemylamino)pyridine (0.41 g, 3.38 mmol) according to the general fluorination procedure and stirred at room temperature for 18 h. Column chromatography (10% ethyl acetate/ hexanes) gave white crystals of 11 (0.33 g, 1.26 mmol, 75%). M.p. 46-47»C; *H nmr (400 MHz, C D C I 3 ) : 8 5.90 (d, 1 H, 1\2 4 Hz, H(l)), 4.82 (ddd, 1 H, J2,3 0.5, J3,4 3, J3 .F 52 Hz, H(3)), 4.72 (ddd, 1 H, J2,3 0.5, Jl,2 4, J2,F 15 Hz, H(2)), 4.34 (m, 1 H, H(5)), 4.08 (m, 2 H, H(4), H(6)), 3.80 (dd, 1 80 H, J5,6' 6.5, J6,6' 8 Hz, H(6')), 1.52, 1.43, 1.35, 1.33 (4 x s, 12 H, CH3). 1$F nmr (CDCI3): 6 187.79. 3-Deoxy-3-fluoro-D-galactopyranose (12) The protected galactofuranose derivative 11 (0.33 g, 1.26 mmol) was suspended in 1:2 ethanol/ water and stirred with Amberlite IR-120(H+) cation exchange resin for 6 h at 60°C. The resin was removed by filtration and the residual gum purified by column chromatography (Solvent C) to afford an anomeric mixture of the free deoxyfluoro sugar 12 (0.17 g, 0.93 mmol, 73%) as a colourless oil. 1 9 F nmr (D2O): 8 199.14, 203.25. 12,4,6-Tetra-0-acetyl-3-deoxy-3-fluoro-D-galactopyranose (13) The deoxyfluoro sugar 12 (0.16 g, 0.90 mmol) was dissolved in a cold (-20°C) stirred solution of dry pyridine (10 mL) to which acetic anhydride (4.0 mL, 42 mmol) was added dropwise according to the general acetylation procedure. The mixture was stirred for 16 h at room temperature then worked up in the normal manner to afford a colourless oil of 13 (0.30 g, 0.86 mmol, 97%), as an anomeric mixture. *H nmr (300 MHz, CDCI3): 5 6.41 (t, 1 H, Jia,2a 4, Jla.F 5 Hz, H(la)), 5.64 (d, 1 H, Jip,2p 8 Hz, H(1P)), 5.63 (m, 2 H, H(4aB)), 5.39 (m, 2 H, H(2ap)), 4.86 (ddd, 1 H, J2a,3a 10, J3a,4a 4, J3a,F 48 Hz, H(3a)), 4.70 (ddd, 1 H, J2p,3P 10, J3p,F 48 Hz, H(3p)), 4.20-3.95 (m, 6 H, H(5aP), H(6,6'ap)), 2.18-2.05 (8 x s, 32 H, 8 x OAc). 1 9 F nmr (CDCI3): 8 200.75 (P), 204.30 (a). 81 2\4'-Dinitrophenyl2A,6-tri-0-acetyl-3-deoxy-3-fluoro-B-D-galactop (15) The tetra-O-acetate 13 (0.30 g, 0.86 mmol) was selectively deacetylated at the 1 position with 2-aminoethanol (0.90 mL, 1.46 mmol) according to the general procedure by stirring in ethyl acetate ( 1 0 mL) for 30 h at room temperature. Column chromatography (Solvent A) afforded an anomeric mixture of tri-O-acetates 14 as a colourless syrup (0.15 g, 0.49 mmol, 58%). 19F nmr ( C D C I 3 ) : 8 2 0 0 . 4 5 , 205.33. This material (0.15 g, 0.49 mmol) was converted to the galactopyranoside 15 according to the general procedure by reaction with DAB CO (0.19 g, 1.65 mmol) and FDNB (0.11 g, 0.59 mmol). The resulting solid was recrystallized from ethanol to give a white crystalline powder (109 mg, 0.23 mmol, 47%) . M.p. 187-188°C; lH nmr (300 MHz, C D C I 3 ) : 8 8.70 (d, 1 H, J3 ' ,5 ' 3 Hz, H(3')), 8.42 (dd, 1 H, J3\5' 3, Js\6' 1° Hz, H(5')), 7.50 (d, 1 H, l5\& 1 0 Hz, H(6')), 5.68 (m, 2 H, H(2), H(4)), 5.18 (d, 1 H, Ji,2 7.5 Hz, H(l)), 4.77 (ddd, 1 H, J2,3 9, J3,4 5, J3 .F 48 Hz, H(3)), 4.25 (m, 2 H, H(6, 6'), 4.10 (m, 1 H, H(5)), 2.25, 2 . 2 0 , 2.10 (3 x s, 9 H, 3 x OAc). I 9 F nmr ( C D C I 3 ) : 8 201.00. Anal. calc. for C 1 8 H 1 9 O 1 2 N 2 F ; C, 45.58; H, 4.04; N, 5.91%; Found: C, 45.48; H, 4.03; N, 5.97%. 2',4'-Dinitrophenyl 3-deoxy-3-fluoro-B-D-galactopyranoside (16) The tri-O-acetate 15 ( 1 0 0 mg, 0.21 mmol) was deacetylated according to the general deacetylation procedure for 19 h. The product was recrystallized from methanol/ diethyl ether/ hexanes to give white crystals of 16 (27 mg, 0.078 mmol, 37%) . M.p. 162-163°C; lH nmr (400 MHz, C D 3 O D ) : 8 8.73 (d, 1 H, J3 ' ,5 ' 3 Hz, H(3')), 8.45 (dd, 1 H, J3\5' 3, J5\6' 10 Hz, H(5')), 7.63 (d, 1 H, J 5 ' t 6 ' 1 0 Hz, H(6')), 5.25 (d, 1 H, J i , 2 7.5 Hz, H(l)), 4.48 (ddd, J2.3 3, J3,4 9, J3 ,F 48 Hz, H(3)), 4.2-4.1 (m, 2 H, H(2), H(4)), 3.85-3.65 (m, 3 H, H(5), H(6,6')). 1 9F nmr ( D 2 O ) : 8 199.51. Anal. calc. for 82 C 1 2 H 1 3 O 9 N 2 F + O . 5 H 2 O ; C, 40.34; H, 3 .95; N, 7.84%; Found: C, 40.24; H, 4.04; N, 7.64%. Methyl 6-O-triphenylmethyl-a-D-glucopyranoside (17) To a suspension of methyl ct-D-glucopyranoside (3.0 g, 15.5 mmol) in dry pyridine (15 mL) was added triphenylmethyl chloride (4.9 g, 17 mmol). The mixture was stirred at 40°C for 20 h. Pyridine was removed by several co-evaporations with toluene and the residue was dissolved in chloroform, washed with water and dried (MgSC»4). The resulting solid was recrystallized from ethanol and dried thoroughly (under vacuum at 75°C) to give colourless needles of 17 (5.21 g, 11.9 mmol, 77%). M.p. 150-152°C; *H nmr (300 MHz, C D C I 3 ) : 8 7.60-7.20 (m, 15 H, 3 x phenyl), 4.79 (d, 1 H, Ji,2 4.5 Hz, H(l)), 3.70-3.40 (m, 6 H, H(2-6,6')), 3.45 (s, 3 H, OMe). Methyl 6-0-triphenylmethyl-4-deoxy-4-fluoro-a-D-galactopyranoside (18) A suspension of the protected glucopyranoside 17 (2.00 g, 4.57 mmol) was treated with DAST (3.3 g, 21 mmol) in 25 mL dry dichloromethane according to the general fluorination procedure. After 3 days at room temperature, the reaction was quenched and the product purified by column chromatography (ethyl acetate) to afford colourless needles of 18 (0.42 g, 0.96 mmol, 21%). M.p. 158-159°C; *H nmr (300 MHz, C D C I 3 ) : 8 7.60-7.20 (m, 15 H, 3 x phenyl), 4.93 (dd, 1 H, 13 ,4 2.5, J4,F 49 Hz, H(4)), 4.81 (d, 1 H, Ji,2 4.5 Hz, H(l)), 3.90-3.30 (m, 5 H, H(2), H(3), H(5), H(6,6')), 3.42 (s, 3 H, OMe). l^F nmr ( C D C I 3 ) : 8 221.66. 4-Deoxy-4-fluoro-D-galactopyranose (19) 83 The protected deoxyfluoro galactopyranoside 18 (0.40 g, 0.91 mmol) was stirred with Amberlite IR-1200H+) cation exchange resin (4 mL) and water (40 mL) at reflux for 3 days. The mixture was filtered hot to remove resin and the bulk of the triphenylmethanol and the precipitated solids were washed with water. The filtrate was washed with dichloromethane and the aqueous portion evaporated to afford a gum which was purified by column chromatography (Solvent C) to give an anomeric mixture of the free deoxyfluoro sugar 19 as a colourless oil (0.15 g, 0.80 mmol, 89%). * 9 F nmr (D2O): 8 217.48, 220.04. 2',4'-Dinitrophenyl4-deoxy-4-fluoro-P-D-galactopyranoside (21) The above compound was obtained from Dr. A. Becalski, Department of Chemistry, U.B.C. Its tri-O-acetate 20, had m.p. 243-245°C; *H nmr (400 MHz, C D C I 3 ) : 8 8.68 (d, 1 H, .Jyty 3 Hz, H(3')), 8.40 (dd, 1 H, Js',6' 10. T3',5' 3 Hz, H(5')), 7.52 (d, 1 H, J5-.6 1 10 Hz, H(6')), 5.63 (dd, 1 H, J2,3 10, Jl,2 8 Hz, H(2)), 5.23 (d, 1 H, Ji,2 8 Hz, H(l)), 5.05 (ddd, 1 H, 13,4 2.5, J2,3 10, J3,F 26.5 Hz, H(3)), 4.95 (dd, 1 H, 13,4 2.5, J4.F 49 Hz, H(4)), 4.36 (AB multiplet, 1 H, J5.6 6, Js,6' 7, J6,6' 12 Hz, H(6,6'), 4.06 (dt, 1 H, J5,6 6, J5)6' 7, J5.F 25 Hz, H(5)), 2.14, 2.12, 2.11 (3 x s, 9 H, 3 x OAc). 1 9 F nmr (CDCI3): 8 216.80. Anal. calc. for C18H19O12N2F; C, 45.57; H, 4.04; N, 5.90%; Found: C, 45.68; H, 3.99; N, 5.94%. The deacetylated galactoside 21 had m.p. 127°C (dec); ! H nmr (400 MHz, C D 3 O D ) : 8 8.73 (d, 1 H, J3',5' 3 Hz, H(3')), 8.48 (dd, 1 H, J5',6' 10, J3\5' 3 Hz, H(5')), 7.66 (d, 1 H, J5',6' 10 Hz, H(6')), 5.28 (d, 1 H, Ji,2 7.5 Hz, H(l)), 4.82 (dd, 1 H, 13,4 3, J4,F 50 Hz, H(4)), 4.00-3.10 (m, 5 H, H(2), H(3), H(5), H(6,6')). 1 9 F nmr (D2O): 8 217.62. Anal. calc. for C 1 2 H 1 3 O 9 N 2 F ; C, 41.39; H, 3.76; N, 8.04%; Found: C, 41.21; H, 3.93; N, 7.94%. 12:3,4-Di-0-isopropylidene-6-deoxy-6-fluoro-a-D-galactopyranose (22) 84 Thoroughly dried l,2:3,4-di-0-isopropyUdene-a-D-galactopyranose39 (2.20 g, 8.45 mmol) was treated with DAST (2.40 mL, 18.2 mmol) and 2,4,6-trimethylpyridine (2.20 mL, 16.7 mmol) in dry dichloromethane (30 mL) according to the general fluorination procedure and stirred at room temperature for 24 h. The reaction was quenched by stirring with excess methanol at room temperature for 2 h and washed with 1 M aqueous acetic acid. Column chromatography (10% ethyl acetate/ hexanes) afforded 22 as a colourless oil (0.87 g, 3.30 mmol, 40%). *H nmr (300 MHz, CDCI3): 8 5.55 (d, 1 H, J i > 2 5 Hz, H(l)), 4.63 (dd, 1 H, J2,3 3, J 3 , 4 8 Hz, H(3)), 4.56 (dm, 2 H, J6/6',F ~ 47 Hz, H(6,6')), 4.35 (dd, 1 H, J2.3 3, J 1,2 5 Hz, H(2)), 4.27 (dd, 1 H, J 3 , 4 8, J4,5 4 Hz, H(4)), 4.08 (m, 1 H, H(5)), 1.55 (s, 3 H, CH3), 1.47 (s, 3 H, CH3), 1.35 (s, 6 H, 2 xCH3). 1 9 F nmr (CDCI3): 8 231.67. 12:3,4 -Di-0 -isopropylidene-6-chloro-6-deoxy-a-D -galactopyranose (49) A solution of l,2:3,4-di-0-isopropylidene-ot-D-galactopyranose39 (0.70 g, 2.69 mmol) and triphenylphosphine (0.70 g, 2.67 mmol) in carbon tetrachloride (5 mL) was refluxed for 12 h. The mixture was cooled, the precipitated solids removed by filtration and the resulting clear solution was evaporated to a syrup which was purified by column chromatography (10% ethyl acetate/ hexanes) to give 49 (0.31 g, 1.11 mmol, 41%). *H nmr (400 MHz, CDC1 3 ) : 8 5.54 (d, 1 H, J U 4 Hz, H(l)), 4.65 (dd, 1 H, J 2 , 3 3, J 3 > 4 8 Hz, H(3)), 4.35 (m, 2 H, H(2), H(4)), 3.95 (m, 1 H, H(5)), 3.65 (AB multiplet, 2 H, J 5 ) 6 8, J5.6' 6, J6.6' 12 Hz, H(6,6'), 1.55, 1.45, 1.37, 1.33 (4 x s, 12 H, 4 x CH3). 6-Deoxy-6-fluoro-a-D-galactopyranose (23) 85 The protected deoxyfluoro sugar 22 (0.71 g, 2.70 mmol) was dissolved in 1:2 ethanol/ water (100 mL) and stirred at 65°C with Amberlite IR-120(H+) cation exchange resin (8 mL). The reaction was complete by tic (Solvent C) after 15 h. The resin was removed by filtration and the gummy residue was crystallized from ethanol/ diethyl ether to afford white crystals of the free sugar 23 (0.37 g, 2.0 mmol, 75%), which crystallized exclusively as the a anomer. M.p. 159-160°C (lit.34 m.p. 160°C); lH nmr (300 MHz, D20): 5 5.25 (d, 1 H, Jia,2ct 4 Hz, H(l)), 4.80-3.40 (m, 6 H, H(2-6,6')). 1 9 F nmr ( D 2 O ) : 8 229.63. 12 $,4-Tetra-0-acetyl-6-deoxy-6-fluoro-D-galactopyranose (24) The deoxyfluoro sugar 23 (0.33 g, 1.80 mmol) was treated with acetic anhydride (0.85 mL, 9.0 mmol) in pyridine (20 mL) according to the general acetylation procedure and purified by column chromatography (30: 70 ethyl acetate/ hexanes) to afford 24 (0.64 g, 1.80 mmol, 100%) as a colourless oil. *H nmr (400 MHz, CDCI3): 8 6.35 (d, 1 H, Jla,2cc 4.5 Hz, H(la)), 5.77 (d, 1 H, Jip,2p 8 Hz, H(lp)), 5.65-5.10 (m, 8 H, H(2-5aP)), 4.70-4.30 (m, 4 H, H(6,6*ap)), 4.25 (m, 1 H, H(5a)), 4.12 (m, 1 H, H(5p)), 2.09-2.01 (8 x s, 32 H, 8 x OAc). 1 9 F nmr (CDCI3): 8 232.37 (p), 233.69 (a). 2\4'-Dinitrophenyl2£,4-tfi-0-acetyl-6-deoxy-6-fliwro^ (26) The tetra-O-acetate 24 (0.65 g, 1.9 mmol) was selectively deacetylated at the 1 position with 2-arninoethanol (0.17 mL, 2.9 mmol) according to the general procedure by stirring in ethyl acetate (50 mL) for 60 h at room temperature. Column chromatography (Solvent A) gave an anomeric mixture of tri-O-acetates 25 as a colourless syrup (0.32 g, 1.1 mmol, 56%). 1 9 F nmr (CDCI3): 8 231.54, 231.66. This material (0.32 g, 1.0 mmol) was converted to the galactopyranoside 26 according to the general procedure by reaction 86 with DAB CO (0.40 g, 3.5 mmol) and FDNB (0.23 g, 1.3 mmol). The resulting solid was recrystallized from ethyl acetate/ hexanes to give white crystals (196 mg, 0.41 mmol, 40%). M.p. 181-1820Q *H nmr (300 MHz, C D C I 3 ) : 8 8.70 (d, 1 H, J3',5' 3 Hz, H(3')), 8.45 (dd, 1 H, Js',6' 10, Jy,5' 3 Hz, H(5')). 7.55 (d, 1 H, J5',6' 10 Hz, H(6')), 5.63 (dd, 1 H, J2.3 10, Jl,2 8 Hz, H(2)), 5.52 (d, 1 H, J3,4 4 Hz, H(4)), 5.30 (d, 1 H, Jl,2 8 Hz, H(l)), 5.15 (dd, 1 H, J2.3 1 ° , J3,4 4 Hz, H(3)), 4.55 (dm, 2 H, J6/6',F 48 Hz, H(6,6')), 4.22 (m, 1 H, H(5)), 2.20, 2.15, 2.05 (3 x s, 9 H, 3 x OAc). i 9 F nmr (CDCI3) : 8 230.08. Anal. calc. for C 1 8 H 1 9 N 2 F ; C, 45.58; H, 4.04; N, 5.91%; Found: C, 45.38; H, 4.16; N, 5.75%. 2',4'-Dinitrophenyl 6-deoxy-6-fluoro-B-D-galactopyranoside (27) The tri-O-acetate 26 (100 mg, 0.210 mmol) was deacetylated according to the general deacetylation procedure for 18 h. The product was recrystallized from acetone/ diethyl ether to give white powdery crystals of 27 (51 mg, 0.15 mmol, 70%). M.p. 180-18 PC; lH nmr (400 MHz, C D 3 O D ) : 8 8.73 (d, 1 H, J3',5' 3 Hz, H(3')), 8.50 (dd, 1 H, ly,5' 3, J5\6' 10 Hz, H(5*)), 7.65 (d, 1 H, J5<t6< 10 Hz, H(6')), 5.25 (d, 1 H, 3\,2 7.5 Hz, H(l)), 4.60 (dm, 2 H, J6/6',F 48 Hz, H(6,6')), 4.12 (m, 1 H, H(5)), 3.95-3.85 (m, 2 H, H(2), H(4)), 3.63 (dd, 1 H, J2.3 10, J3.4 4 Hz, H(3)). l9F nmr (D2O): 8 230.37. Anal. calc. for C 1 2 H 1 3 O 9 N 2 F + H 2 O ; C, 39.35; H, 4.13; N, 7.65%; Found: C, 39.15; H, 4.22; N, 7.30%. Methyl 4-chloro-4-deoxy-6-0-triphenylmethyl-a-D-glucopyranoside (31) Methyl 2,3,6-tri-0-benzoyl-4-chloro-4-deoxy-a-D-glucopyranoside 545 4'5 5 (3.0 g, 5.76 mmol) was debenzoylated in 0.13 M sodium methoxide/ methanol (10 mL) for 20 h at room temperature, neutralized with Amberlite IR-120(H+) cation exchange resin and 87 filtered to afford the crude debenzoylated compound 29 as a white solid. This was dissolved in dry pyridine (12 mL) and triphenylmethyl chloride (1.66 g, 1.95 mmol) added. The mixture was stirred at 40°C for 65 h then pyridine was removed by successive co-evaporations with toluene and the residue was dissolved in dichloromethane, washed with water and dried (MgSO*). Column chromatography (2.5% methanol/ ethyl acetate) gave 31 as a white foam (1.73 g, 3.81 mmol, 66% from 54). *H nmr (300 MHz, CDC13): 8 7.65-7.20 (m, 15 H, 3 x phenyl), 4.90 (d, 1 H, Ji , 2 4 Hz, H(l)), 3.95-3.80 (m, 3 H, H(3-5)), 3.65 (dt, 1 H, Ji , 2 4, J 2 > 3 7.5, J 2 > 0H 7.5 Hz, H(2)), 3.50 (m, 4 H, H(6), OMe), 3.29 (dd, 1 H, J 5 , 6 ' 6, J6.61 9 Hz, H(6')), 2.70 (br s, 1 H, 3-OH), 2.25 (d, 1 H, J2,OH 7.5 Hz, 2-OH). Methyl 3,4-anhydro-6-0-triphenylmethyl-a-D-galactopyranoside (32) and methyl 2,3-anhydro-6-O-triphenylmethyl-a-D-gulopyranoside (33) The methyl glucopyranoside 31 (0.10 g, 0.22 mmol) was dissolved in dry methanol (5 mL) containing sodium methoxide (6 mg, 0.26 mmol) and heated at reflux for 15 h. Solvent was evaporated in vacuo and the residue was chromatographed (Solvent A) to afford 32 (30 mg, 0.067 mmol, 33%, Rf 0.48) and 33 (43 mg, 0.096 mmol, 47%, Rf 0.38), both as white foams. lH nmr (400 MHz, C D C I 3) 32: 8 7.50-7.20 (m, 15 H, 3 x phenyl), 4.60 (d, 1 H, Ji,2 5 Hz, H(l)), 3.98 (t, 1 H, Js,6 5, J5,6' 5 Hz, H(5)), 3.78 (dd, 1 H, Ji,2 5, J2,OH 10 Hz, H(2)), 3.42 (s, 3 H, OMe), 3.35 (m, 3 H, H(6,6'), H(4)), 3.22 (d, 1 H, 13,4 4.5 Hz, H(3)), 2.48 (d, 1 H, 2-OH). lH nmr (400 MHz, C D C I 3 ) 33: 8 7.50-7.20 (m, 15 H, 3 x phenyl), 5.00 (d, 1 H, Ji,2 4 Hz, H(l)), 4.10-4.00 (m, 2 H, H(4), H(5)), 3.52 (s, 3H, OMe), 3.43-3.22 (m, 4 H, H(2), H(3), H(6,6')), 2.40 (d, 1 H, 4-OH). Methyl 6-0-triphenylmethyl-3-deoxy-a-D-xy\o-hexopyranoside (34) 88 The 3,4-anhydro galactopyranoside 32 (0.41 g, 0.98 mmol) was dissolved in dry THF (25 mL), cooled to 0°C and lithium aluminum hydride (0.15 g, 4.0 mmol) added in one portion. The mixture was stirred under nitrogen at 0°C for 15 min then at room temperature for 5 h. The reaction was quenched with water, extracted with diethyl ether and dried (MgSC»4). Evaporation of solvent in vacuo afforded a chromatographically homogenous (Solvent A) white foam 34 (0.39 g, 0.93 mmol, 95%). *H nmr (400 MHz, C D C I 3 ) : 8 7.50-7.20 (m, 15 H, 3 x phenyl), 4.72 (d, 1 H, J i > 2 4 Hz, H(l)), 4.10-3.95 (m, 2 H, H(2), H(4)), 3.79 (t, 1 H, J5,6 5, Js,6' 5 Hz, H(5)), 3.45 (s, 3 H, OMe), 3.40-3.00 (m, 2 H, H(6,6')), 2.08 (dt, 1 H, J 3 e,4 4, J2,3e 4, J3e,3a 12.5 Hz, H(3e)), 1.75 (dt, 1 H, J3 a,4 3, J2,3a 12, J3e,3a 12.5 Hz, H(3a)). Methyl 2,4,6-tri-0-acetyl~3-deoxy-a-D-xy\o-hexopyranoside (36) The tritylated methyl glycopyranoside 34 (0.36 g, 0.86 mmol) was stirred in 1:2 ethanol/ water (25 mL) at room temperature for 6 h with Amberlite IR-120(H+) cation exchange resin (10 mL). The mixture was filtered to remove the resin and precipitated triphenylmethanol and the solvent evaporated in vacuo to give methyl 3-deoxy-a-D-xy/o-hexopyranoside 35 as a yellow gum which was acetylated according to the general procedure without purification and chromatographed (Solvent A) to afford 36 as a colourless oil (0.19 g, 0.65 mmol, 75% from 34). lH nmr (400 MHz, CDCI3) : 8 5.13 (m, 1 H, H(4)), 5.08 (ddd, 1 H, J i , 2 4, J 2 ,3 e 5, J2,3a 12 Hz, H(2)), 4.90 (d, 1 H, J i , 2 4 Hz, H(l)), 4.15-4.05 (m, 3 H, H(5), H(6,6')), 3.45 (s, 3 H, OMe), 2.20-2.10 (m, 2 H, H(3e,3a)), 2.15, 2.10, 2.08 (3 x s, 9 H, 3 x OAc). 2,4,6-Tri-0-acetyl-3-deoxy-D-xy\o-hexopyranose (38) 89 The acetylated methyl glycopyranoside 36 (0.19 g, 0.65 mmol) was dissolved in 1,1-dichloromethyl methyl ether (3 mL) and stirred with a catalytic amount of freshly fused zinc chloride under nitrogen at 65°C. After 3 h, the excess ether was removed in vacuo and the residue was dissolved in dichloromethane, washed with saturated sodium bicarbonate solution and dried (MgSC«4). Evaporation of solvent afforded 2,4,6-tri-O-acetyl-3-deoxy-a-D-xy/o-hexopyranosyl chloride 37 as a pale yellow oil which was used without further purification. J H nmr (400 MHz, CDCI3): 6 6.32 (d, 1 H, Ji,2 4 Hz, H(l)), 5.25-5.15 (m, 2 H, H(2), H(4)), 4.35 (m, 1 H, H(5)), 4.17 (AB multiplet, 2 H, J 5 i 6 5.5, J 5 > 6 - 7, J6.61 11 Hz, H(6,6')), 2.15, 2.09, 2.07 (3 x s, 9 H, 3 x OAc). To the above oil, dissolved in acetone (3 mL) and 2 drops water, was added freshly prepared silver carbonate (0.15 g, 0.55 mmol) at 0°C in the dark. The mixture was stirred for 24 h at room temperature then the suspension was filtered, and the precipitated solids washed thoroughly with acetone. Column chromatography (Solvent A) gave 38 as a syrupy mixture of anomers (0.14 g, 0.48 mmol, 75% from 36). *H nmr (400 MHz, CDCI3): 8 5.43 (d, 1 H, J i a ,2a3 Hz, H(la)), 5.18 (m, 2 H, H(4a(5)), 5.15-4.95 (m, 2 H, H(lp), H(2a)), 4.90 (m, 1 H, H(2p)), 4.68 (br d, 1 H, (5-OH), 4.35 (m, 1 H, H(5a)), 4.20-4.00 (m, 4 H, H(6,6'ap)), 3.92 (m, 1 H, H(5J3)), 2.32 (dm, 1 H, J3e,3a I 3 Hz, H(3ep)), 2.20-2.00 (m, 19 H, 6 x OAc, H(3ea)), 2.00-1.80 (m, 2 H, H(3aap)). 2',4'-Dinitrophenyl 2,4,6-tri-0-acetyl-3-deoxy-pr-D-xy\o-hexopyranoside (39) The tri-O-acetate 38 (0.14 g, 0.48 mmol) was converted to the glycopyranoside 39 according to the general procedure by reaction with DABCO (0.18 g, 1.6 mmol) and FDNB (0.11 g, 0.57 mmol). The resulting solid was chromatographed (Solvent A) and recrystallized from ethyl acetate/ hexanes to give yellowish crystals (49 mg, 0.11 mmol, 23%). M.p. 116-1190C; !H nmr (400 MHz, CDC13): 8 8.65 (d, 1 H, J 3' > 5' 3 Hz, H(3')), 8.35 (dd, 1 H, J5\6' 10, J3\5' 3 Hz, H(5')), 7.42 (d, 1 H, J 5 ' i 6- 10 Hz, H(6')), 5.30-5.20 90 (m, 2 H, H(l), H(2)), 5.15 (m, 1 H, H(4)), 4.20-3.95 (m, 3 H, H(5), H(6,6')), 2.43 (dt, 1 H, J2,3e 4, J3e,4 4, J 3 e , 3 a 15 Hz, H(3e)), 2.09, 2.08, 1.98 (3 x s, 9 H, 3 x OAc), 1.84 (ddd, 1 H, J3a,4 3, J 2 , 3 a 10, J3a,3e 15 Hz, H(3a)). Anal. calc. for C18H20O12N2; C, 47.38; H, 4.42; N, 6.14%; Found: C, 47.38; H, 4.79; N, 5.57%. Methyl 2J,6-tri-0-benzoyl-4-deoxy-4-iodo-a-D-glucopyranoside (40) Sodium iodide (0.38 g, 2.50 mmol) was added to a solution of methyl 2,3,6-tri-O-benzoyl-4-O-methanesulphonyl-a-D-galactopyranoside 5755 (0.50 g, 0.85 mol) in dry DMF (10 mL). The mixture was refluxed for 3 h then washed with water, charcoal-treated and dried (MgSCU). Column chromatography (20% ethyl acetate/ hexanes) afforded 40 as a white foam (0.28 g, 0.45 mmol, 53%). ! H nmr (300 MHz, CDCI3): 8.30-7.31 (m, 15 H, 3 x OBz), 6.08 (t, 1 H, J 2 , 3 10, J3>4 10 Hz, H(3)), 5.22 (m, 2 H, HQ), H(2)), 4.75 (AB multiples 2 H, J 5 ( 6 4, J 5 i 6 - 6, ^  13 Hz, H(6,6')), 4.30 (t, 1 H, J 3 > 4 10, J 4 ) 5 1° Hz, H(4)), 3.50 (s, 3 H, OMe). *H nmr showed minor resonances, presumably arising from the presence of the galacto isomer, but the desired gluco compound was favoured by a ratio of greater than 10 to 1. Methyl 3,4-anhydro-a-D-galactopyranoside (28) A solution of the iodo glucopyranoside 40 (48 mg, 0.078 mmol) in 2:1 methanol/ DMF (3.5 mL) containing sodium methoxide (2.3 mg, 0.10 mmol) was stirred at room temperature for 4 days. Tic (9:1 dichloromethane/ methanol) showed ~ 75% methyl 4-deoxy-4-iodo-a-D-glucopyranoside and ~ 25% 28 after 18 h. The reaction was worked up when methyl 3,6-anhydro-a-D-galactopyranoside 30 began to appear after 90 to 100 h at room temperature. Solvent was evaporated and the residue was chromatographed (Solvent B) to afford 28 as a gummy solid (9 mg, 0.051 mmol, 66%). lH nmr (300 91 MHz, D20): 8 4.70 (d, 1 H, J 1 ) 2 4 Hz, H(l)), 4.11 (m, 1 H, H(5)), 3.85 (m, 2 H, H(2), H(6)), 3.72 (dd, 1 H, J5,6' 7, J6,6' 12 Hz, H(6')), 3.45-3.40 (m, 4 H, H(4), OMe), 3.31 (d, 1 H, J3,4 4.5 Hz, H(3)). 2',4'-Dinitrophenyl 4-deoxy-B-D-xy\o-hexopyranoside (42) The above compound was obtained from M. Namchuk, Department of Chemistry, U.B.C. Its tri-O-acetate 41 had m.p. 169-170°C; lH nmr (300 MHz, CDCI3): 8 8.70 (d, 1 H, J3\5' 3 Hz, H(3')), 8.40 (dd, 1 H, J3',5' 3, J5\6' 10 Hz, H(5')), 7.49 (d, 1 H, J5',6 10 Hz, H(6')), 5.30-5.05 (m, 3 H, H(l-3)), 4.20 (m, 2 H, H(6,6')), 4.05 (m, 1 H, H(5)), 2.25 (ddd, J4e,5 3, J3,4e 6, J4e,4a 13 Hz, H(4e)), 2.15, 2.11, 2.10 (3 x s, 9 H, 3 x OAc), 1.80 (q, 1 H, J4a,5 13, J3,4a 13, J4e,4a 13 Hz, H(4a)). Anal. calc. for Ci8H 2 u Oi 2 N 2 ; C, 47.37; H, 4.42; N, 6.14%; Found: C, 47.53; H, 4.58; N, 6.07%. The deacetylated glycoside 42 had m.p. 76-79°C; J H nmr (300 MHz, C3D6O): 8 8.70 (d, 1 H, J3.5' 3 Hz, H(3')), 8.50 (dd, 1 H, J 3', 5' 3, h',6' 10 Hz, H(5')), 7.45 (d, 1 H, J5'> 6- 10 Hz, H(6')), 5.30 (d, 1 H, Ji, 2 8 Hz, H(l)), 4.10-3.00 (m, 5 H, H(2), H(3), H(5), H(6,6')), 2.00 (ddd, 1 H, J 4 e,5 3, J3,4e 6, J4e,4a 13 Hz, H(4e)), 1.55 (q, 1 H, J 3 , 4 a 13, J 4 a > 5 I 3, J 4 a > 4 e 13 Hz, H(4a)). Anal. calc. for Ci 2Hi 40 9N 2+2H 20; C, 39.35; H, 4.95; N, 7.65%; Found: C, 39.68; H, 4.78; N, 5.88%. 2J,4,-Tri-0-aceryl-6-deoxy-D-galactopyranose (44) D-Fucose, 6-deoxy-D-galactose (0.655 g, 3.99 mmol) was acetylated with acetic anhydride (20 mL, 210 mmol) in pyridine (50 mL) for 18 h according to the general procedure to afford an anomeric mixture of tetra-O-acetates 43 (1.32 g, 3.99 mmol, 100%) as a pale yellow oil. This material (1.30 g, 3.91 mmol) was selectively deacetylated at the 1 position with hydrazine acetate (0.44 g, 4.80 mmol) according to the general procedure to 92 afford, after chromatography (Solvent A), a white solid 44 (0.79 g, 2.72 mmol, 69%). M.p. 112-1130C; lH nmr (300 MHz, CDC13): 8 5.49 (t, 1 H, J i a i 2 a 4, J i a > 0 H ~4 Hz, H(la)), 5.40 (dd, 1 H, J2 a,3a H , J3a,4a4 Hz, H(3a)), 5.32 (m, 1 H, H(4a )), 5.28 (m, 1 H, H(4B)), 5.18 (dd, 1 H, J 2 a , 3 a 11, J 3 a > 4 a 4 Hz, H(2a)), 5.08 (m, 1 H, H(2B), H(3B)), 4.65 (dd, 1 H, Jip.2p 8, Jip.oH 10 Hz, H(lB)), 4.42 (q, 1 H, J 5 a,6a 6 Hz, H(5a)), 3.85 (q, 1 H, J5p,6B 6 Hz, H(5B)), 3.50 (d, 1 H, J i p , 0 H 10 Hz, p-OH), 2.85 (br s, 1 H, a-OH), 2.18, 2.12, 2.02 (6 x s, 18 H, 6 x OAc), 1.23 (d, 3 H, J 5 a i 6 a 6 Hz, H(6a)), 1.15 (d, 3 H, J5p)6p 6 Hz, H(6P)). 2 ',4 '-Dinitrophenyl 2 3,,4-tri-0-acetyl-6-deoxy-B-D-galactopyranoside (45) The tri-O-acetate 44 (0.79 g, 2.72 mmol) was converted to the galactopyranoside 45 according to the general procedure by reaction with DABCO (1.04 g, 9.25 mmol) and FDNB (0.61 g, 3.27 mmol). The resulting solid was recrystallized from ethyl acetate/ hexanes to afford white crystals (0.86 g, 1.89 mmol, 69%). M.p. 166-167°C (lit.38 m.p. 169-172°C); ] H nmr (300 MHz, CDCI3): 8 8.70 (d, 1 H, Jj, 5' 3 Hz, H(3')), 8.43 (dd, 1 H, J3-.51 3, J5-.6' 10 Hz, H(5')), 7.49 (d, 1 H, J 5-, 6- 10 Hz, H(6')), 5.58 (dd, 1 H, J 1 ) 2 7.5, J 2 > 3 9 Hz, H(2)), 5.35 (m, 1 H, H(4)), 5.22 (d, 1 H, Ji.2 7.5 Hz, H(l)), 5.12 (dd, 1 H, J 2 > 3 9, J 3 i 4 4.5 Hz, H(3)), 4.10 (m, 1 H, H(5)), 2.22, 2.15, 2.05 (3 x s, 9 H, 3 x OAc), 1.32 (d, 3 H, J 5 , 6 6 Hz, H(6)). Anal. calc. for Ci8H 2 0 Oi 2 N 2 ; C, 47.37; H, 4.42; N, 6.14%; Found: C, 47.31; H, 4.52; N, 6.00%. 2\4 '-Dinitrophenyl 6-deoxy-B-D-galactopyranoside (46) The tri-O-acetate 45 (175 mg, 0.38 mmol) was deacetylated according to the general deacetylation procedure for 20 h. The product was recrystallized from methanol/ diethyl ether/ hexanes to give faintly yellow crystals of 46 (31 mg, 0.094 mmol, 25%). 93 M.p. 115-118°C (dec) Git.38 m.p. 135-136<>C); *H nmr (400 MHz, CD3OD): 8 8.72 (d, 1 H, J 3 . t 5 . 3 Hz, H(3')), 8.47 (dd, 1 H, J 3', 5' 3, J5.,6.10 Hz, H(5')), 7.63 (d, 1 H, J 5', 6' 10 Hz, H(6')), 5.19 (d, 1 H, Ji, 2 8 Hz, H(l)), 3.93 (q, 1 H, J 5 > 6 6 Hz, H(5)), 3.82 (dd, 1 H, Ji , 2 8, J 2 , 3 9 Hz, H(2)), 3.70 (d, 1 H, J 3 > 4 3 Hz, H(4)), 3.59 (dd, 1 H, J 2 f 3 9, J 3 > 4 3 Hz, H(3)), 1.30 (d, 3 H, J 5 > 6 6 Hz, H(6)). Anal. calc. for Ci 2Hi 40 9N 2+2H 20; C, 39.35; H, 4.95; N, 7.65%; Found: C, 38.96; H, 5.06; N, 5.41%. 2. Enzymology 2.1 General procedures Escherichia coli (lacZ) B-galactosidase (B-D-galactoside galactohydrolase; EC 3.2.1.23) was purchased as a suspension in 1.7 M ammonium sulphate from the Sigma Chemical Co. All absorbance measurements were made on a Pye Unicam PU-8800 UV-Visible spectrophotometer equipped with a circulating water bath. Matched quartz cells were used throughout except for the determination of Ki and ki for 2FBGalDNP, for which plastic cells were used. All cells used in this work had a path length of 1 cm. 2.2 Extinction coefficients Enzymatic hydrolyses of 2',4'-dinitrophenyl glycopyranosides were followed at 400 nm, the wavelength of maximal absorbance (Xmax) of 2,4-dinitrophenol. The intact glycosides do not absorb at this wavelength so the rate of change in absorbance (AA400/ min) is proportional to the rate of 2,4-dinitrophenol release. The extinction coefficient (e) of 2,4-dinitrophenol at 400 nm at 25°C in phosphate buffer (pH 7.00) was determined in the following manner. A stock solution of accurately known (by weight) phenol concentration was prepared. This stock solution was diluted with buffer to give absorbance values of 0.2 to 1.0. Accurate absorbance values at 400 nm were recorded, and the extinction coefficient determined from Beer's Law: z-A/be where A is absorbance, b is cell path length and c is the molar concentration of the solution. The extinction coefficient (10-3 e) of 2,4-dinitrophenol at 400 nm was deterrriined to be 11.1 M' 1 cm"1 at 25°C in 50 mM sodium phosphate buffer at pH 7.00. 2',4'-Dinitrophenyl glycopyranoside concentrations were determined from their absorbance at 255 nm (10"3 e = 10.7 M _ 1 cm*1 for BGluDNP at 37°C in 50 mM sodium phosphate buffer at pH 6.8). The extinction coefficient of the glucoside was determined in the following manner by J. Kempton and M. Namchuk, other workers in this laboratory. A solution of PGluDNP of known absorbance at 255 nm was incubated with pABG5 P-glucosidase until hydrolysis was complete. Since the initial concentration of glucoside is equal to the concentration of phenol released, Egiucoside could be calculated from the final absorbance at 400 nm. Extinction coefficients for the deoxy and deoxyfluoro glycosides were assumed to be the same as that of the unsubstituted glucoside. 2.3 Determination of Km and krat for hydrolyses of 2'.4'-dinitrophenvl p-D  glvcopvranosides hv E. coli (lacZ) B-palactosidase All kinetic measurements were performed at 25°C in 50 mM sodium phosphate buffer (pH 7.00) containing 1 mM Mg 2 + (added as MgCl2- 6H2O) and 0.1% bovine serum albumin (BSA). Enzyme concentrations and reaction times were chosen so that less than 10% of the total substrate was hydrolyzed to ensure linear kinetics. The enzyme dilution used for each substrate was assayed for activity with BGalPNP at saturating concentration 95 (~ 0.3 mM). All kcat values are standardized with respect to a value for pGalPNP of 156 s*1, in accordance with Sinnott and Souchard.13 Rates of enzymic hydrolysis at each substrate concentration were determined as follows. Substrate stock, BS A and buffer were mixed in a quartz spectrophotometer cell to give the desired substrate concentration. The cell was thermally equilibrated at 25°C, and an appropriate dilution of P-galactosidase (10 |iL) was added to the cell, which was mixed several times by inversion. The rate of hydrolysis was determined spectrophotometrically at 400 nm. Values of K m and kcat were determined by measuring rates at six to eight substrate concentrations bracketing K m and ranging from, if possible, 7 times to 0.15 times K m . The rate of spontaneous glycoside hydrolysis in each cell was subtracted from the observed rate. A slightly different procedure was employed for the deoxy substrates because of their high rates of spontaneous hydrolysis. Thermally pre-equilibrated buffer was added to a weighed quantity of the deoxy galactoside, the stock concentration was determined by absorbance at 255 nm and appropriate volumes added to thermally pre-equilibrated cells containing buffer and BSA. Enzyme was added immediately and the rates determined as above. The data are presented as double reciprocal plots according to the method of Lineweaver and Burk77, but this method was not used for calculating kinetic parameters due to the inaccuracy introduced by the non-linear error span. The values of K m and kcat were calculated from a weighted non-linear regression program, Curvefitter, written for an Apple He computer by I. Street. Some of the glycosides were sufficiently insoluble in buffer (or bound so poorly) that the concentrations of saturated solutions were far below K m . In these instances, approximate values for kcat/Km were determined from plots of v vs. [S] using the Michaelis-Menten equation with [S] « K m v = kcat[E0][S3 Km where [EQ] is the total enzyme concentration and [S] is the concentration of substrate. 2.4 Determination of K; and kj for 2'.4'-dinitrophenvl 2-deoxv-2-fluoro-B-D-galactopvranoside bv time-dependent inactivation Solutions containing several concentrations of 2FBGalDNP and 0.1% BSA were incubated at 25°C. B-Galactosidase was added to each inactivation mixture, and aliquots were removed at intervals until the enzyme was ~ 90% inactivated. Residual activity at each time interval was determined by diluting aliquots (10 |iL) of the inactivation mixture into 1.6 mL of BGalPNP at saturating concentration (~ 0.3 mM), and measuring the rate of release of 4-nitrophenolate, which is directly proportional to the amount of remaining active enzyme. The rate constant at each inactivator concentration, kobs, w a s determined by plotting the logarithm of the fraction of remaining active enzyme against time. Kj and kj were found by fitting the rate constants so determined to kobs= frm Ki+[I] using the Curvefitter weighted regression program, described previously. 2.5 Determination of B-galactosidase concentration bv absorbance at 280 nm and 2.4-dinitrophenolate burst Several concentrations of freshly dialyzed B-galactosidase in 50 mM phosphate buffer (pH 7.00) containing 1 mM Mg 2 + were incubated in quartz cells at 25°C in the spectrophotometer. Absorbances at 280 nm were recorded and the concentration of protein 97 determined from Beer's Law (E280 = 2.1 cm*1 mg*1 mL76). A solution of 2FPGalDNP to give a final inhibitor concentration of ~ 0.4 mM (approximately one-half Kj and fifty times the estimated enzyme concentration) was added to the cell, which was mixed several times by inversion. The change in absorbance at 400 nm over ~ 45 minutes was recorded. After corrections had been made for the added volume, the initial absorbance and the spontaneous hydrolysis of the galactoside during the course of the experiment, the enzyme concentration could be calculated according to concentration of = AA400 x Sample Volume (^ L) enzyme (mol L'1) 1.11 x 104 M' 1 cnr1 since the moles of enzyme present is equal to the moles of 2,4-dinitrophenolate released. 2.6 Reactivation of covaiemiy inactivated p-gaiactpsidase Inactivated enzyme (Section 2.5) was centrifuged (3 x 30 min, 5500 rpm) in a Millipore Ultrafree-PF filter unit fitted with a polysulphone membrane (30 000 molecular weight limit) using a Sorvall RC-5B centrifuge with an SS-34 rotor to free it of excess, non-covalently bound inhibitor. The sample was reduced to a volume of approximately 50 \iL, washed with approximately 2 mL of buffer after each centrifugation and finally diluted with 2.5 mL of buffer. The end point activity was calculated from the ratio of the absorbances at 280 nm before and after centrifugation multiplied by the initial activity (prior to inactivation). Samples of the inactivated enzyme were then incubated in 50 mM phosphate buffer (pH 7.00) containing 1 mM Mg 2 + at 25°C in the presence of any desired transglycosylation acceptors. Reactivation was monitored by the removal of aliquots at 98 appropriate time intervals and assaying for activity by addition to 2.5 mL of PGalPNP at saturating concentration ( ~ 0.3 mM). The observed rate constant, k0bs. for each reactivation process was determined by fitting the data to a single exponential function using an Archimedes 410/1 computer with an Applied Photophysics kinetic workstation. The return of activity can be related to the full activity of the enzyme sample by [E] = [E0](l-exp{-kobst}) Control samples with non-inactivated enzyme were assayed for activity with PGalPNP to account for denaturation during the time course of the experiment with the incubation conditions used. 99 REFERENCES 1 Koshland, D.E. (1953) Biol, Rev. 28, 416. 2 Wallenfels, K., & Weil, R. (1972) in The Enzymes; Academic Press: New York, Vol. 7, p. 617. 3 Wallenfels, K., & Malhotra, O.P. (1961) Adv. Carb. Chem. 16, 239. 4 Stokes, T.M., & Wilson, LB. (1972) Biochemistry 11, 1061. 5 Blake, C.C.F., Mair, G.A., North, A.C.T., Phillips, D.C., & Sarma, V.R. (1967) Proc. R. Soc. London Ser. B 167, 365. 6 Chiang, Y., & Kresge, A J. (1985) J. Amer. Chem. Soc. 109, 7530. 7 Herrchen, M., & Legler, G. (1984) Eur. J. Biochem. 138, 527. 8 Bader, D.E., Ring, M., & Huber, R.E. (1988) Biochem. Biophys. Res. Commun. 153, 301. 9 Sinnott, M.L. (1990) Chem. Rev. 90, 1171. 1 0 Withers, S.G., & Street, LP. (1988) J. Amer. Chem. Soc. 110, 8551. U Withers, S.G., Warren, R.A.J., Street, LP., Rupitz, K., Kempton, J.B., & Aebersold, R. (1990) / . Amer. Chem. Soc. 112, 5887. 1 2 Sinnott, M.L., & Withers, S.G. (1974) Biochem. J. 143, 751. 1 3 Sinnott, M.L., & Souchard, I.J.L. (1973) Biochem. J. 133, 89. 1 4 Blake, C.C.F., Johnson, L.N., Mair, G.A., North, A.T.C., Phillips, D.C., & Sarma, V.R. (1967) Proc. R. Soc. London Ser. B 167, 378. 1 5 Tenu, J.-P., Viratelle, O.M., & Yon, J. (1972) Eur. J. Biochem. 26, 112. 1 6 Sinnott, M.L., & Withers, S.G. (1978) Biochem. J. 175, 539. 1 7 Ring, M„ Bader, D.E., & Huber, R.E. (1988) Biochem. Biophys. Res. Commun. 152, 1050. 1 8 Lehmann, J., & Schlesselmann, P. (1983) Carb. Res. 113, 93. 1 9 Edwards, R.A., Cupples, C.G., & Huber, RJ£. (1990) Biochem. Biophys. Res. Commun. 171, 33. 100 20 Loeffler, R.S.T., Sinnott, M.L., Sykes, B.D., & Withers, S.G. (1979) Biochem. J. 177, 145. 2 1 Haldane, J.B.S. (1930) in Enzymes; Longmans, Green & Co., p. 182. 2 2 Pauling, L. (1946) Chem. Engng. News 24, 1375. 2 3 Fersht, A.R. (1985) Enzyme Structure and Mechanism, 2n d ed.; W.H. Freeman & Co.: San Francisco. 2 4 Quicho, F., Wilson, D., & Vyas, N. (1989) Nature 340, 404. 2 5 Fersht, A.R. (1987) Biochemistry 26, 8031. 2 6 Withers, S.G., Street, LP., & Percival, M.D. (1988) in Fluorinated Carbohydrates: Chemical and Biochemical Aspects; American Chemical Society: Washington, D.C., p. 59. 2 7 Street, LP., Armstrong, C , & Withers, S.G. (1986) Biochemistry 25, 6021. 2 8 Street, LP., Rupitz, K., & Withers, S.G. (1989) Biochemstry 28, 1581. 2 9 Murray-Rust, P., Stallings, W.G., Monti, C.T., Preston, R., & Glusker, J. (1983) / . Amer. Chem. Soc. 105, 3206. 30 Karpides, A., & Miller, C. (1984) / . Amer. Chem. Soc. 106, 1494. 51 Adamson, J., & Marcus, D. (1972) Carb. Res. 22, 257. 3 2 Brimacombe, J., Foster, A., Westwood, J., Hems, A., & Hall, L. (1970) Can. J. Chem. 48, 3946. 3 3 Marcus, D., & Westwood, J. (1971) Carb. Res. 17, 269. 3 4 Taylor, N.F., & Kent, P.W. (1958) / . Chem. Soc. London, 872. 3 5 Haynes, L.J., & Newth, F.J. (1955) Adv. Carb. Chem. 10, 213. 36 McCloskey, CM., & Coleman, G.H. (1955) Org. Syn. Coll. Vol. 3, 439. 37 van Boom, J.H., Koeners, H.J., de Koke, A.J., & Romers, C. (1980) Rec. Trav. Chim. Pays-Bos 99, 355. 38 Ballardie, F., Capon, B., Sutherland, J.D. Cocker, D., & Sinnott, M.L. (1973) J. Chem. Soc. Perkin Trans. I, 2418. 39 Schmidt, O. (1963) in Meth. Carb. Chem., Vol. U [83]. 40 Sowa, W., & Thomas, G. (1966) Can. J. Chem. 44, 836. 101 4 1 Jewell, J.S..& Szarek, W.H. (1969) Tetr. Lett. 1, 43. 4 2 Onodera, K., Hirano, S., & Kashimura, N. (1965) J. Amer. Chem. Soc. 87, 4651. 4 3 zu Reckendorf, W. (1963) in Meth. Carb. Chem., Vol H [17]. 4 4 Kovac, P., & Glaudemans, C. (1983) Carb. Res. 123, 326. 4 5 Grynkiewicz, G., Fokt, I., Szeja, W., & Fitak, H. (1989) / . Chem. Res. (S), 152. 4 6 Card, P.J., & Reddy, G.S. (1983) / . Org. Chem. 48, 4737. 4 7 Tewson, T.J., & Welch, M.J. (1978) / . Org. Chem. 43, 1090. 4 8 Evelyn, L., & Hall, L. (1976) Carb. Res. 47, 285. 4 9 Maradufu, A., & Perlin, A.S. (1974) Carb. Res. 32, 261. 5 0 Lin, Tsu-Hsing, Kovac, P., & Glaudemans, C. (1989) Carb. Res. 188, 228. 51 Angyal, S.J., & Pickles, V.A. (1972) Aust. J. Chem. 25, 1711. 52Hedgley, E.J., Overend, W.G., & Rennie, R.A.C. (1963) J. Chem. Soc, 4701. 5 3 Jones, J.K.N., Perry, M.B., & Turner, J.C. (1960) Can. J. Chem. 38, 1122. 5 4 Williams, J.M. & Richardson, A.C. (1967) Tetr. 23, 1369. 5 5 Street, LP. (1988) Ph.D. Thesis, University of B.C. 56 Buchanan, J.G. (1958) J. Chem. Soc, 2511. 5 7 Izumi, K. (1973) Carb. Res. 27, 278. 58 Lemieux, R.U., & Barette, J.P. (1958) / . Amer. Chem. Soc. 80, 2422. 5 9 Foster, A., Stacey, M., & Vardheim, S. (1958) Acta Chem. Scand. 12, 1819. 6 0 Noyce, D.S., & Virgilio, J.A. (1972) / . Org. Chem. 37, 2643. 6 1 Buchanan, J.G., Fletcher, R., Parry, K., & Thomas, W. (1969) / . Chem. Soc. (B), 377. 6 2 Lopes, D., & Taylor, N. (1979) Carb. Res. 73, 125. 63 Gross, H., & Farkas, I. (1960) Chem. Ber. 93, 95. ^Excoffier, G., Gagnaire, D., & Utille, J.-P. (1975) Carb. Res. 39, 368. 102 6 5 Sinnott, M.L., & Viratelle, O.M. (1973) Biochem. J. 133, 81. 66 Withers, S.G., Rupitz, K., & Street, LP. (1988) / . Biol Chem. 263, 7929. 6 7 Kempton, J.B. (1990) M.Sc. Thesis, University of B.C. 68 Huber, KB., Kurz, G., & Wallenfels, K. (1976) Biochemistry 15, 1994. 69 Huber, R.E., Gaunt, M.T., & Hurlburt, K.L. (1984) Arch. Biochem. Biophys. 234, 151. 70lmoto, T., Johnson, L.N., North, A.C.T., Phillips, D.C., & Rupley, J.A. (1972) in The Enzymes; Academic Press: New York, Vol. 7, p. 665. 71 Deschavanne, P.J., Viratelle, O.M., & Yon, J. (1978) / . Biol. Chem. 253, 833. 7 2 Huber, R.E., & Gaunt, M.J. (1983) Arch. Biochem. Biophys. 220, 263. 73 Cocker, D., & Sinnott, M.L. (1975) / . Chem. Soc. Perkin Trans. II, 1391. 7 4 Brockhaus, M., Dettinger, H.M., Kurz, G., Lehmann, J., & Wallenfels, K. (1979) Carb. Res. 69, 264. 7 5 Withers, S.G. (1977) Ph.D. Thesis, University of Bristol. 76Tenu, J.-P., Viratelle, O.M., Gamier, J., & Yon, J. (1971) Eur. J. Biochem 20, 363. 7 7 Lineweaver, H., & Burk, D. (1934) J. Amer. Chem. Soc. 56, 658. 78 Withers, S.G., MacLennan, D.J., & Street, LP. (1986) Carb. Res. 154, 127. 79 Withers, S.G., Percival, M.D., & Street, LP. (1989) Carb. Res. 187, 43. 80 Capon, B. (1969) Chem. Rev. 69, 407. 81 Rosenberg, S., & Kirsch, J.F. (1981) Biochemistry 20, 3189. 8 2 Clark-Still, W., Kahn, M., & Mitra, M. (1977) / . Org. Chem. 43, 2923. 83 Lee, J.B., & Nolan, TJ. (1966) Can. J. Chem. 44, 1331. 103 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0059772/manifest

Comment

Related Items