Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A detailed mechanistic investigation of the exoglycanase from Cellulomonas fimi Tull, Dedreia L. 1995

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

Item Metadata

Download

Media
831-ubc_1995-060756.pdf [ 4.38MB ]
Metadata
JSON: 831-1.0059648.json
JSON-LD: 831-1.0059648-ld.json
RDF/XML (Pretty): 831-1.0059648-rdf.xml
RDF/JSON: 831-1.0059648-rdf.json
Turtle: 831-1.0059648-turtle.txt
N-Triples: 831-1.0059648-rdf-ntriples.txt
Original Record: 831-1.0059648-source.json
Full Text
831-1.0059648-fulltext.txt
Citation
831-1.0059648.ris

Full Text

A DETAILED MECHANISTIC INVESTIGATION OF THEEXOGLYCANASE FROM CELLULOMONAS FIMIByDEDREIA TULLB. Sc., McGill University, 1989M. Sc., University of British Columbia, 1991A THESIS SUBMITfED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDEPARTMENT OF CHEMISTRYWe accept this thesis as conformingto the required standard:THE UNWERSITY OF BRITISH COLUMBIAJuly, 1995© Dedreia L. Tull, 1995In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)___________________Department of - -The University of British ColumbiaVancouver, CanadaDate A’.L 4J9SDE-6 (2)88)IIABSTRACTThe exoglycanase from Cellulomonas fimi catalyses the hydrolysis of cellooligosaccharides to cellobiose as well as the hydrolysis of xylan and aryl 3-glycosides(Gilkes et al (1984) .1. Biol. Chem. 259, 10455). Its mechanism of action is thought toinvolve a double displacement reaction which is investigated here through detailed kineticstudies of the native enzyme and point mutants with a range of aryl f3-glycosides, andthrough inactivation studies with 2-deoxy-. and 2-deoxyfluoro-glycoside mechanism-basedinactivators and the affinity label, N-bromoacetyl cellobiosylamine.A pH study of the native enzyme revealed ionisations of PKa = 4.1 and 7.7 in thefree enzyme, likely corresponding to the catalytic nucleophile and the acid-base catalyst,respectively. The large secondary deuterium kinetic isotope effects measured on both stepsfor the glucosides and on the deglycosylation step for the cellobiosides reveal significantoxocarbonium ion character at the corresponding transition states, thus suggestingsubstantial C-O bond cleavage and little nucleophilic preassociation. By contrast, therelatively small secondary deuterium kinetic isotope effect and the small Broensted constantmeasured on the glycosylation step for the cellobiosides suggest that the cellobiosylationtransition state is less highly charged than the glucosylation transition state. These studiessuggest that the primary function of the distal glucosyl moiety of the cellobiosides is toincrease the rate of glycosylation, likely through improved acid catalysis and greaternucleophile preassociation, without affecting its rate of deglycosylation. The greater ratesof hydrolysis of the xylo-sugars, relative to those for the gluco-sugars, indicate that thesubstrate preference of C. fimi exoglycanase increases in the order glucosides <xylosides<cellobiosides <xylobiosides and that the C-5 hydroxymethyl group is slightly inhibitoryto catalysis. The role of the C-2 hydroxyl group was probed using 2,4-dinitrophenyl 2-deoxy-2-fluoro cellobioside (2F-.DNPC) and cellobial (a 2-deoxycellobiose analogue).Rates of hydrolysis of the 2-deoxyfluorocellobiosyl- and 2-deoxycellobiosyl-enzymes areiO and 106-fold lower respectively, than that for the cellobiosyl-enzyme, indicating that theifiC-2 hydroxyl group is necessary for catalysis and that it contributes a minimum of -9kcal/mole of stabilisation energy to the transition state.Electrospray ionisation mass spectrometry (ESI-MS) of the 2F-DNPC-inactivatedenzyme provided evidence for the covalent nature of the glycosyl-enzyme intermediatewhile 19F NMR analysis of this 2FCb-enzyme and the 2-deoxy-2-fluoro 4-O-(f-glucosyl)-j3-mannosyl fluoride (2F-GMF) -inactivated enzyme provided evidence for the - anomericstereochemistry of the intermediate.The catalytic nucleophile involved in C. fimi exoglycanase-catalysed hydrolysis ofthe cellobiosides was identified as Glu 233 by use of tandem MS techniques and 2F-DNPCand cellobial. Kinetic analysis of the Glu233Asp mutant revealed that pulling the catalyticnucleophile 1 A away from the reacting anomeric centre reduces the rates of glycosylationand deglycosylation —4 x 103-fold.ESI-MS analysis of N-bromoacetyl cellobiosylamine-inactivated C. fimiexoglycanase reveals that one mole of N-acetyl cellobiosylamine is incorporated per moleof enzyme. The labeled residue was identified as Glu 127 by use of a combination of MStechniques. This residue has recently been suggested to be the acid-base catalyst based onkinetic analysis of mutants (MacLeod et al (1994) Biochemistry 33, 6571). More detailedkinetic analysis of the Glul27Ala mutant revealed rate reductions of 200-300 fold on thedeglycosylation step while the rate reductions on the glycosylation step are dependent onthe leaving group ability of the phenolate. The larger Broensted constant seen with theGlul27Ala mutant compared to that for the native enzyme reflects greater negative chargeaccumulation on the leaving phenolate at the glycosylation transition state for theGlul27Ala mutant. These results are consistent with the role of Glu 127 as the acid-basecatalyst.These structural findings are completely consistent with the recently solved X-raycrystal structure of the catalytic domain of C.fimi exoglycanase (White et al (1994)Biochemistry 33, 12546).ivTABLE OF CONTENTSABSTRACT iiTABLE OF CONTENTS ivLIST OF TABLES xLIST OF FIGURES xiABBREVIATIONS AND DEFINITIONS xviGLOSSARY xxACKNOWLEDGEMENTS xixCHAPTER I 1GENERAL INTRODUCTION 11-1 Glycosidases 21-2 Classification Of Glycosidases 21-3 Cellulases 41-4 The Exoglycanase From Cellulomonas Fimi 61-5 Mechanism Of “Retaining” Glycosidases: Evidence For A DoubleDisplacement Mechanism 111-5-1 Presence of a carboxylate nucleophile 111-5-2 Nature of the glycosyl-enzyme intermediate 141-5-3 Oxocarbonium ion-like transition states 171-5-4 General acid assistance 221-5-5 Non-covalent enzyme-substrate interactions 251-5 Aim Of This Study 28CHAPTER II 30DETAILED KINETIC ANALYSIS OF THE NATIVE CELLULOMONASFIMI EXOGLYCANASE 302-1 Introduction To Different Kinetic Techniques 31V2-1-1 Kinetic scheme for a double displacement mechanism 312-1-2 Pre-steady state analysis of enzymatic reactions 322-1-3 Linear free energy relationships as mechanistic probes of enzymatic reaction 342-1-4 Secondary deuterium kinetic isotope effects as transition state probes 362-1-5 Fluorine and hydrogen substitutions as probes of enzymatic reactions 392-1-6 pH Dependence of enzymatic reactions 412-1-7 Background on kinetic analysis of C.flmi exoglycanase 422-2 Objectives Of This Project 472-2-1 Pre-steady state analysis of aryl f-cellobiosides and aryl 13 - glucosides. 472-2-2 Linear free energy relationships of aryl 13 -glycosides 482-2-3 a-Secondary deuterium kinetic isotope effects on aryl f3-glycosides. 492-2-4 Inactivation of C. fimi exoglycanase by 2-deoxy- and 2-deoxy-2-fluoro-sugars. 492-2-5 pH Study 512-3 Results 512-3-1 Steady state kinetics for xylo-substrates 512-3-2 Pre-steady state kinetic analysis of the hydrolysis of gluco-substrates 552-3-3 a-Secondary deuterium kinetic isotope effect measurements ongluco-substrates. 572-3-4 Inactivation of C.fimi exoglycanase 582-3-5 Reactivation of inactivated-C. fimi exoglycanase 582-3-6 19F NMR analysis of inactivated-C.fimi exoglycanase 592-3-7 pH Study of C.fimi exoglycanase 602-4 Discussion 642-4-1 pH Dependence of C. fimi exoglycanase 642-4-2 13-Glucanase activity of C. fimi exoglycanase 642-4-3 13-Glucosidase activity of C.JImi exoglycanase 662-4-4 Xylanase activity of C. fimi exoglycanase 702-4-5 Xylosidase activity of C. fimi exoglycanase 712-4-6 Inactivation-reactivation studies of C.fimi exoglycanase with 2F-DNPX2 722-4-7 Characterisation of the glycosyl-C. fimi exoglycanase intermediate 742-4-8 C.fimi exoglycanase-catalysed hydration of cellobial 762-4-9 Effect of substitutions at C-2 on cellobioside hydrolysis rates 792-5 Summary 81CHAPTER III 84DETAILED KINETIC ANALYSIS OF MUTANTS OF C. FIMIEXOGLYCANASE 843-1 Introduction 853-2 Objectives Of This Project 86vi3-3 Results For The Glu233Asp C. fimi Exoglycanase (Nucleophile)Mutant 883-3-1 Substrate reactivity 883-3-2 x-Secondary deuterium kinetic isotope effect 913-3-3 pH Study 913-4 Results For Glul27Ala C. Fimi Exoglycanase (Acid-Base Catalyst)Mutant 933-4-1 Substrate reactivity 933-4-2 Stopped-flow analysis 953-4-3 x-Secondary deuterium kinetic isotope effect 973-4-4 pH Study 973-5 Discussion Of The Glu233Asp C.Finii Exoglycanase (Nucleophile)Mutant 993-5-1 Proposed role of the catalytic nucleophile in C.fimi exoglycanase catalysis 993-5-2 pH Dependence of the Glu233Asp C.JImi exoglycanase mutant 993-5-3 Substrate specificity of the Glu233Asp mutant 1013-5-4 Rate determining steps for Glu233Asp mutant catalysis 1013-5-5 Effect of mispositioning the carboxylate group on the individual steps ofthe reaction. 1023-6 Discussion Of Glul27AIa C. Fimi Exoglycanase (Acid-Base Catalyst)Mutant 1053-6-1 Proposed role of the acid-base catalyst in C.flmi exoglycanase catalysis 1053-6-2 pH Dependence of Glul27Ala C.fimi exoglycanase mutant 1053-6-3 Effect of removal of the acid-base catalyst on f3 -glucanase activity 1063-6-4 Effect of removal of the acid-base catalyst on the individual steps in thereaction 1093-7 Summary 110CHAPTER IV 111LABELING STUDIES OF C. FIMI EXOGLYCANASE USING ACTIVESITE-DIRECTED IRREVERSIBLE INACTIVATORS 1114-1 Introduction 1124-1-1 Criteria for active site-directed irreversible inactivators 1124-1-2 Affinity labels 1134-1-3 Mechanism-based inactivators 1164-2 Background And Objectives Of This Project 1194-3 Results And Discussion 1214-3-1 Labeling studies of C.fimi exoglycanase using 2,4-dinitrophenyl 2- deoxy2-fluoro-13-cellobioside 121(a) Stoichiometry of inactivation 121(b) Strategyfor identification of the 2-deoxyfluorocellobiosyl-labeled residue 121(c) Identification of the residue modified by 2,4-dinitrophenyl 2-deoxy- 2-vilfluoro-J3-cellobioside 1244-3-2 Labeling studies of C.fimi exoglycanase using cellobial 127(a) Stoichiometry of inactivation 127(b) Identflcation of the residue modified by cellobial 1284-3-3 Labeling studies of C.flmi exoglycanase with N-bromoacetylcellobiosylamine 129(a) Protection against inactivation 129(b) Stoichiometry of inactivation 130(c) Strategyfor identification ofN-acetyl cellobiosylamine-labeled residue 132(d) IdentWcation ofpeptide modified by N-bromoacetyl cellobiosylamine 135(e) Identjfication of the labeled residue 1364-3-4 Comparison of the radiolabeled and the mass spectrometric approachesused to identify labeled residues in glycosidases 1404-4 Summary 142CHAPTER V 143MATERIALS AND METHODS 1435-1 Synthesis 1445-1-1 General materials and procedures 1445-1-2 General compounds 1455-1-3 {‘H}-aryl -cellobiosides 1475-1-4{2H}-aryl 13 -cellobiosides 1495-1-5 Inactivators 1525-2 Enzymology 1555-2-1 General materials and procedures 1555-2-2 Determination of steady state kinetic parameters 1565-2-3 Determination of pre-steady state kinetic parameters 1575-2-4 Secondary deuterium kinetic isotope effect measurements 1585-2-5 pH Study 1585-2-6 Inactivation kinetics 1605-2-7 Protection against inactivation 1615-2-8 Reactivation kinetics 1625-2-9 19F-NMR analysis of 2-deoxy-2-fluoro-glycosyl-CjImiexoglycanase 1625-2-10 Stoichiometry of inactivation 1635-2-11 Identification of the residue labeled by 2F-DNPC and cellobial 1635-2-12 Identification of the residue labeled by N-bromoacetyl cellobiosylamine 164REFERENCES 167APPENDIX A 174GRAPHICAL REPRESENTATION OF KINETIC DATA 174vmA-i Absorbance versus Time Plot Of The Hydrolysis of ONPC By TheGlu233Asp Mutant. i75A-2 Lineweaver-Burk Plots For The Hydrolysis Of Aryl 3-Xylosides By Native C. fimi Exoglycanase. 176A-3 Lineweaver-Burk Plots For The Hydrolysis Of Aryl -Xylobiosides By Native C. fimi Exoglycanase. 177A-4 Absorbance versus Time Plots for Pre-Steady State Analysis ofPNPC with Native C. fimi Exoglycanase i78A-5 Lineweaver-Burk Plots Of The Pre-Steady State Analysis OfAryl f-Cellobiosides With Native C. fimi Exoglycanase. 179A-6 Lineweaver-Burk Plots For The Hydrolysis Of Aryl 3-Glucosides And PNPX2 by The Glu233Asp Mutant Of C. fimiExoglycanase. 180A-7 Lineweaver-Burk Plots For The Hydrolysis Of Aryl-Cellobiosides By The Glu233Asp Mutant Of C. fimiExoglycanase. 181A-8 Lineweaver-Burk Plots For The Hydrolysis Of Aryl 3-Cellobiosides By The Glul27Ala Mutant Of C. fimiExoglycanase. 182A-9 Lineweaver-Burk Plots Of The Pre-Steady State Analysis OfAryl 13-Cellobiosides With The Glul27Ala Mutant Of C. fimiExoglycanase. 183A-b Inactivation-Reactivation Kinetics Of C. fimi Exoglycanase. 184APPENDIX B 188BASIC CONCEPTS OF ENZYME CATALYSIS 188xB-i Basic Enzyme Kinetics 189B-2 Interpretation Of kcat And kcat/Km 192B-3 Binding Energy And Enzyme Catalysis 195B-4 Inactivation Kinetics Of C. fimi Exoglycanase i98xLIST OF TABLESTable 1-1 Cellulases for which the X-ray crystal structures have been solved. 6Table 1-2 List of catalytic nucleophiles identified by mechanism-basedinactivators, 2-deoxy-2-fluoro-glycosides. 13Table 2-1 Size comparison of C-H, C-F and C-OH groups. 40Table 2-2 Michaelis-Menten parameters for the hydrolysis of aryl 3-xylosidesby C. fimi exoglycanase. 52Table 2-3 Michaelis-Menten parameters for the hydrolysis of aryl 3-xylobiosides by C.fimi exoglycanase. 52Table 2-4 Pre-steady state parameters for hydrolysis of aryl f3-cellobiosides byC. fimi exoglycanase. 55Table 2-5 Secondary deuterium kinetic isotope effects measured with C.fimiexoglycanase. 57Table 2-6 Rates of reactivation of inactivated C. fimi Exoglycanase in thepresence of a glycosyl-acceptor. 59Table 3-1 Michaelis-Menten parameters for the hydrolysis of aryl 3-cellobiosides by the Glu233Asp mutant. 89Table 3-2 Michaelis-Menten parameters for the hydrolysis of aryl f3-glucosidesby the Glu233Asp mutant. 89Table 3-3 Michaelis-Menten parameters for the hydrolysis of aryl J3-cellobiosidesby the Glul27Ala mutant. 93Table 3-4 Pre-steady state parameters for hydrolysis of aryl f3-cellobiosides bythe Glul27Ala mutant. 95xLIST OF FIGURESFigure 1-1Figure 1-2Figure 1-3Figure 1-4FigureFigureFigureFigureFigureFigureFigureFigure578910111216171819212324252727234Figure 1-5Figure 1-6Figure 1-7Figure 1-81-91-10Figure 1-11Figure 1-12Hydrolysis of glycosides by glycosidases.Examples of a and anomers of an aryl glucoside.Stereochemical classification of glycosidases.Schematic representation of the degradation of cellulose toglucose by the cellulase complex.Schematic representation of the structure of intact C.fimiexoglycanase.The X-ray crystal structure of the catalytic domain of C.fimiexoglycanase.Structure of xylan and cellulose.Double displacement mechanism proposed for C.fimiexoglycanase.The peptidoglycan substrate for lysozyme.Structure of 2-iodobenzyl- 1-thio-13-D-cellobiose.Labeling the catalytic nucleophile (Glu 358) of Agrobacterium 3-glucosidase with 2F-DNPG.Reactivation of 2F-DNPG-inactivated Agrobacterium3-glucosidase in the presence of 1-deoxy-f-glucosy1 benzene.Comparison of glycosyl cations with glycosides.Resonance structures for aldonolactones and aldonolactams.Nojirimycin, a transition state analogue for glycosidases.Structure of 1 -deoxy- 1 -fluoro-D-glucosyl fluoridesand 1, 1-deoxy- 1,1 -difluoro-D-glucosyl fluoride.Hydration of an octenitol derivative by E. coli 3-galactosidase.Hydration of 2-acetamidoglycal by 3-N-acetyl hexosaminidasesfrom Jack bean, bovine kidney and human placenta.Structure of f3-D-galactosyl pyridinium cation.Substrate analogue for lysozyme.Structure of deoxy-maltose derivatives.1-131-141-151-1615Figure 1-17Figure 1-18Figure 1-191-201-21xliFigure 2-1 Broensted plot relating values of for the hydrolysis of aryl-g1ucosides by Agrobacterium 3-glucosidase with theleaving group ability of the phenol. 36Figure 2-2 C.fimi exoglycanase-catalysed hydrolysis of aryl 3-ce1lobiosides. 42Figure 2-3 Broensted plots relating rates of C. fimi exoglycanase-catalysedhydrolysis of aiyl -glucosides with the leaving groupability of the phenols. 44Figure 2-4 Broensted plots relating rates of C. fimi exoglycanase-catalysedhydrolysis of aryl 13-cellobiosides with the leaving groupability of the phenols. 45Figure 2-5 Structures of aryl f3-xylobiosides and aryl [3-xylosides. 48Figure 2-6 Structure of 2,4-dinitrophenyl 2-deoxy-2-fluoro-f3-xylobioside(2F-DNPX). 49Figure 2-7 Structures of 2,4-dinitrophenyl 2-deoxy-2-fluoro-cellobioside(2F-DNPC) and 2-deoxy-2-fluoro 4-O-([3-glucosy1)-3-mannosy1fluoride (2F-GMF). 50Figure 2-8 Structure of cellobial. 50Figure 2-9 Broensted plots relating rates of C.flmi exoglycanase-catalysedhydrolysis of aryl f3-xylosides with the leaving group ability ofthe phenols. 53Figure 2-10 Broensted plots relating rates of C.fimi exoglycanase-catalysedhydrolysis of aryl 13-xylobiosides with leaving group ability ofthe phenols. 54Figure 2-11 Broensted plots relating pre-steady state rates of C.fimiexoglycanase-catalysed hydrolysis of aryl f3-cellobioside withleaving group ability of the phenols. 56Figure 2-12 ‘9F-NMR spectrum of C. Jlmi exoglycanase inactivated by2F-DNPC. 61Figure 2-13 ‘9F-NMR spectra of C.Jlmi exoglycanase inactivated by 2F-GMF. 62xmFigure 2-14 pH Dependence of the hydrolysis of 2,4-DNPC by C.Jlmiexoglycanase. 63Figure 2-15 Reaction coordinate diagram illustrating the stabilisationproduced by the distal glucosyl moiety of the cellobiosides. 68Figure 2-16 Mechanism for glycosidase-catalysed hydration of a glycal. 78Figure 3-1 Replacement of the catalytic nucleophile of C.fimi exoglycanase(Glu 233) by an aspartate. 87Figure 3-2 Replacement of the acid-base catalyst of C. fimi exoglycanase(Glu 127) by an alanine residue. 88Figure 3-3 Broensted plots relating rates of native enzyme andGlu233Asp mutant-catalysed hydrolysis of aryl -cellobiosideswith the leaving group ability of the phenols. 90Figure 3-4 pH Dependence of the hydrolysis of 2,4-DNPC by the nativeenzyme and the Glu233Asp mutant. 92Figure 3-5. Broensted plots relating rates of native enzyme and Glul27Alamutant catalysed hydrolysis of aryl -cellobiosides with the leavinggroup ability of the phenols. 94Figure 3-6 Broensted plots relating pre-steady state rate of Glu l27Ala mutantcatalysed hydrolysis of aryl 3-cellobiosides with the leaving groupability of the phenols. 96Figure 3-7 pH Dependence of the hydrolysis of cellobiosides by the nativeenzyme and the Glul27Ala mutant. 98Figure 3-8 Schematic diagram illustrating the hydrogen bonding networkaround Glu 233. 100Figure 3-9 Reaction coordinate diagram illustrating the effect of shorteningthe catalytic nucleophile on the glycosylation and deglycosylationtransition states. 104Figure 3-10 Linear free energy relationship correlating the glycosylation stepsfor the native and the Glul27Ala mutant. 108Figure 4-1 Examples of affinity labels. 114Figure 4-2 Mechanism for inactivation of glycosidases by glycosyl epoxides. 115Figure 4-3 Inactivation of a glycosidase by a conduritol epoxide and releaseof inositol upon hydroxylamine treatment. 118xivFigure 4-4 Electrospray ionisation mass spectra showing stoichiometry ofinactivation of C.fimi exoglycanase by 2F-DNPC. 122Figure 4-5 Scheme of the method used to identify the residue in C.fimiexoglycanase labeled by 2F-DNPC and cellobiaL 123Figure 4-6 Electrospray ionisation mass spectra of a peptic digestof C.Jimi exoglycanase inactivated by 2F-DNPC. 126Figure 4-7 Electrospray ionisation mass spectra showing stoichiometry ofinactivation of C. fimi exoglycanase by cellobial. 127Figure 4-8 Neutral loss tandem electrospray ionisation mass spectrum ofa peptic digest of C.fimi exoglycanase inactivated by cellobial. 129Figure 4-9 Structure of benzyl 4-O-((3-D-glucopyranosyl)- 1-thio--D-xylopyranoside. 130Figure 4-10 Mechanism of inactivation of C.fimi exoglycanase byN-bromoacetyl cellobiosylamine. 131Figure 4-11 Electrospray ionisation mass spectra showing stoichiometry ofinactivation of C. fimi exoglycanase by N-bromoacetylcellobiosylamine. 133Figure 4-12 Scheme of the method used to identify the residue in C.fimiexoglycanase labeled by N-bromoacetyl cellobiosylamine. 134Figure 4-13 Electrospray ionisation mass spectrum of a peptic digestof C. fimi exoglycanase inactivated by N-bromoacetylcellobiosylamine. 137Figure 4-14 Electrospray ionisation mass spectrum of the 311 PTH derivativereleased in cycleS during sequencing of the labeled peptide. 139Figure 4-15 Tandem electrospray ionisation mass spectrum of the 311 PTHderivative released in cycle 5 during sequencing of the labeledpeptide. 139Figure A-b-i Inactivation of C.fimi exoglycanase by cellobial. 184Figure A-i0-2 Inactivation of C.fimi exoglycanase by 2F-GMF. 185Figure A-10-3 Inactivation of C.fimi exoglycanase by N-bromoacetylcellobiosylamine. 186Figure A- 10-4 Reactivation of2F-DNPX-inactivated C. fimi Exoglycanase. 187Figure A-10-5 Reactivation of cellobial-inactivated C.fimi Exoglycanase. 187Figure B-i-i Plot of velocity versus substrate concentration for a typicalenzymatic reaction. 190Figure B-1-2 A typical Lineweaver-Burk plot for an enzymatic reaction. 191Figure B-2-1 Reaction coordinate diagram for an enzymatic reactioninvolving the interconversion of intermediates. 193Figure B-3-1 Reaction coordinate diagram for a typical reaction and thecorresponding uncatalysed reaction. 196Figure B-3-2 Reaction coordinate diagram illustrating complementarity ofthe enzyme to the ground state, and the transition state of thesubstrate. 197xvxviLIST OF ABBREVIATIONS AND DEFINITIONSGlucosides2,4-DNPG 2,4-Dinitrophenyl 3-D-g1ucoside2C1,4NPG 2-Chloro-4-nitrophenyl f3-D-glucoside3 ,4-DNPG 3,4-Dinitrophenyl 3-D-g1ucoside2,5-DNPG 2,5-Dinitrophenyl -D-g1ucosidePNPG 4-Nitrophenyl -D-g1ucosideCeliobiosides2,4-DNPC 2,4-Dinitrophenyl 13-cellobioside3 ,4-DNPC 3,4-Dinitrophenyl 13-cellobioside2,5-DNPC 2,5-Dinitrophenyl 13-cellobiosidePNPC 4-Nitrophenyl 13-cellobiosideONPC 2-Nitrophenyl f3-cellobioside3,5-DCIPC 3,5-Dichiorophenyl 3-ce11obioside4-CNPC 4-Cyanophenyl 3-ce11obioside4-BrPC 4-Bromophenyl 3-ce11obiosideXylosides2,3-DNPX 2,3-Dinitrophenyl f3-D-xylosidexvii2,5-DNPX 2,5-Dinitrophenyl 3-D-xy1oside3 ,4-DNPX 3,4-Diniirophenyl 3-D-xy1osidePNPX 4-Nitrophenyl j3-D-xyloside4-CNPX 4-Cyanophenyl -D-xy1osideXylobiosides2,5-DNPX 2,5-Dinitrophenyl -xy1obioside3 ,4-DNPX2 3,4-Dinitrophenyl f3-xylobiosidePNPX2 4-Nitrophenyl 3-xy1obiosideONPX2 2-Nitrophenyl f3-xylobiosideInactivators2F-DNPG 2,4-Dinitrophenyl 2-deoxy-2-fluoro-13-D-glucoside2F-DNPGa1 2,4-Dinitrophenyl 2-deoxy-2-fluoro-f3-D-galactoside2F-DNPC 2,4-Dinitrophenyl 2-deoxy-2-fluoro43-ce11obioside2F-GMF 2-Deoxy-2-fluoro 4-O-(3-g1ucosy1)-3-mannosy1 fluoride2F-DNPX 2,4-Dinitrophenyl 2-deoxy-.2-fluoro-3-xy1obioside2FCb- 2-Deoxy-2-fluoro-cellobiosyl-2FGM- 2-Deoxy-2-fluoro 4-O-(3-g1ucosy1)-3-mannosy1-2dCb- 2-Deoxy-cellobiosyl-2FX- 2-Deoxy-2-fluoro-xylobiosyl-xvmN-BrAc-CbNH2 N-Bromoacetyl 3-ce11obiosylamineBGTX Benzyl 4-O-(f-g1ucosyl)- 1 -thio-f-D-xy1osideAmino acidsAla AlanineArg ArginineAsn AsparagineAsp Aspartic AcidGlu Glutamic Acidlie IsoleucineLeu LeucineThr ThreonineVal ValineEnzymesC. fimi CellulomonasfimiE. coli Escherichia coliBSA Bovine serum albuminKinetic and physical constantsDissociation constant for a substrate-enzyme complexKm Michaelis constantK1 Dissociation constant for an inactivator-enzyme complexVm Maximal reaction velocitykcat Catalytic rate constantkObS Pseudo-first order rate constantk2 Glycosylation rate constantxixk3 Deglycosylation rate constantk1 First-order rate constant for inactivationig Broensted constantT Temperature (K)k Boltzmann constanth Planck’s constantOthersm. p. Melting pointTLC Thin layer chromatographyNMR Nuclear magnetic resonanceRP-HPLC Reverse-phase high performance liquid chromatographyMS Mass spectrometryESI-MS Electrospray ionisation mass spectromeiryESI-MS/MS Electrospray ionisation tandem mass spectrometryTIC Total ion chromatogramn/z Mass to charge ratio311 PITC 4-(3-Pyridinylmethyl-aminocarboxypropyl) PhenylisothiocyanatePTH PhenylthiohydantoinxxGLOSSARYCellulases: cellulolytic enzymes responsible for catalysing the conversion ofcellulose to glucose.Endoglucanases: enzymes that catalyse the hydrolysis of cellulose to cellooligosaccharides by cleaving relatively randomly along the cellulosechain.Exoglucanases: enzymes that catalyse the hydrolysis of cello-oligosaccharides andcellulose to cellobiose from either the non-reducing or reducingends.Exoglycanases: enzymes that catalyse the hydrolysis of oligosaccharides andpolysaccharidesXylanases: enzymes that catalyse the hydrolysis of xylan.xxiACKNOWLEDGEMENTSI wish to express my deepest appreciations to my supervisor, Dr. Stephen G.Withers, for his help and guidance throughout the time of this research. Thanks also to mycoworkers for their helpfull discussions, allowing me to use the compounds they painstakingly sythesised, and especially for their friendships. Special thanks to Ms. KarenRupitz and Dr. Qingping Wang for technical assistance, to Dr. Shichang Miao, Dr. DavidBurgoyne and Mr. David Chow for performing the mass spectrometric analysis as well asto Dr. Lawrence McIntosh for performing‘9F-NMR analysis of the protein samples. Iwould like to thank the NMR and elemental analysis staff of the Department of Chemistryfor all their help and advise. Thanks to Drs. R. Aebersold (Department of MolecularBiotechnology, University of Washington), A. Warren, D. Kilbum and N. Gilkes(Department of Microbiology, University of British Columbia) as well as Dr. David Roseand Mr. Andre White (Department of Medical Biophysics, University of Toronto) forcollaborating on this project.Thanks to the University of British Columbia and Natural Sciences and EngineeringResearch Council of Canada for financial support. Special thanks to the ProteinEngineering Network Centres of Excellence (PENCE) for funding such an exciting projectand providing me with opportunities to attend several enlightening and informativeconferences.Thanks to my family and all my friends for their support and encouragement.For VH, LT, HT, RT, NT, CT and AAM,and for “Women Who (dare to) Run WithThe Wolves” (C. P. Estés).1CHAPTER IGENERAL INTRODUCTION21-1 GlycosidasesGlycosidases are enzymes that catalyse the hydrolysis of the glycosidic linkagebetween two sugar residues (Figure 1-1). Interest in these enzymes has escalated in recentyears due to their potential applications in many areas of industry, for example cellulases inbiomass degradation and textiles, xylanases in the pulp and paper industry and of courseamylases in the food and beverage industries.Substrates for glycosidases are made up of two parts, a glycone portion and anaglycone portion (Figure 1-1). The aglycone portion in most natural substrates is anothersugar residue, however since most of these enzymes are less specific for the aglycone thanthe glycone, this sugar can often be replaced by an alkyl or aryl group. This allows for theuse of chromogenic substrates, making kinetic analysis of these enzymes more convenient.H20+ROHORGlycosidase OH\%%—/AglyconeGlyconeFigure 1-1 Hydrolysis of glycosides by glycosidases.1-2 Classification of GlycosidasesGlycosidases constitute a large class of enzymes which is subdivided into smallergroups based on three features.3(1) The glycone to which it is reactive; for example a galactofuranosidase or amannopyranosidase. However, it should be noted that although a glycosidase may bereactive against a variety of substrates, the enzyme is generally classified based on thesugar against which it is the most reactive.(2) The anomeric configuration (cL or ) of the glycosidic linkage that is to becleaved. For example, a f3-glucosidase will only catalyse the hydrolysis of [-glucosidesand not c-glucosides (Figure 1-2).(3) The relative anomeric configuration of the substrate versus the initial glyconeproduct that is, if the anomeric configurations of the substrate and the initial product are thesame then the enzyme is classified as a “retaining” glycosidase whereas if the anomericconfigurations are different, then the enzyme is termed “inverting” (Figure 1-3).13-GlucosideOHa-GlucosidexFigure 1-2 Examples of a and /3 anomers of an aryl glucoside.4ORaiming”OHFigure 1-31-3 CellulasesCellulose, the most abundant material in nature, is a carbohydrate polymer made upof glucose units connected by 3—(1-4) linkages. Conversion of cellulose to glucose isachieved by a cellulase complex which is composed of three types of glycosidases: endo1 ,4-f3-glucanases (endoglucanase, EC 3.2.1.4), exo- 1 ,4-13-glucanases (cellobiohydrolase,EC 3.2.1.91) and f3-glucosidases (cellobiase, EC 3.2.1.21). These enzymes are producedby a variety of organisms including microorganisms and plants. These enzymes can beindividually secreted by the organism, as is the case for most fungal cellulases. In somecases the excreted cellulases can be organized into supramolecular structures calledcellulosomes as is found in some bacterial cellulases. The cellulase complex is known tocatalyse the hydrolysis of cellulose synergistically, although the detailed mechanism of thissynergism is not well understood. The role of each individual glycosidase in the complex isthe following. The endoglucanase catalyses the random cleavage of the cellulose intooligosaccharides, the exoglucanase then converts the oligosaccharides into cellobiose,generally cleaving from the non-reducing end and finally the -glucosidase catalyses the“inverting”Stereochemical classfication ofglycosidases.5hydrolysis of cellobiose to glucose (Figure 1-4). Many cellulases have been isolated from avariety of organisms, cloned, sequenced and divided into families based on sequencesimilarity of their catalytic domains (Henrissat et al., 1989; Gilkes et al.; 1991; Henrissat &Bairoch, 1993). These families of glycanases also include a number of xylanases.Endoglucanase0-0-0-0 0-0-0-0I0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0I00000000 00000 00Figure 1-4 Schematic representation of the degradation of cellulose to glucose by thecellulase complex.Many cellulases are composed of at least two distinct domains; a catalytic domainand a cellulose-binding domain which are joined together by a linker region. The three-dimensional structures of the catalytic domains of a number of these cellulases haverecently been determined by X-ray crystallography and these are shown in Table 1-1. Otherstructural information on these proteins includes the ‘H-NMR structure of the cellulose-binding domain of Trichoderma reesei cellobiohydrolase II (Rouvinen et al., 1990) and that6of Cellulomonas fimi (Xu et al, 1995) as well as the tadpole-like structures of intactendoglucanase A and intact exoglycanase from Cellulomonasfimi as revealed by low angleX-ray studies (Pilz et al., 1990). Additional structural information has been derived fromlabeling studies of these enzymes with affinity labels and mechanism-based inactivators andthus has lead to the identification of catalytically important residues (discussed in ChapterIV).Table 1-1: Cellulases for which the X-ray crystal structures have been solved.Cellulase Organism ReferenceCellobiohydrolase II Trichoderma reesei (Rouvinen et al, 1990)Endoglucanase D Clostridium thermocellum (Juy et al., 1992)Endocellulase E2 Thermomonospora fusca (Spezio et al., 1993)Endoglucanase V Humicola insolens (Davies et aL, 1993)1,3-1,4-Glucanase Bacillus sp. (Keitel et al., 1993)1,3-Glucanase G II Barley (Varghese et al., 1994)1,3-1,4-Glucanase E II Barley (Varghese et al, 1994)Exoglycanase Cellulomonasfimi (White et al., 1994)Cellobiohydrolase I Trichoderma reesei (Divne et al., 1994)1-4 The Exoglycanase from Cellulomonas fimiThe enzyme used in this investigation is the exoglycanase from the soil bacterium,CellulomonasfImi. The gene encoding the exoglycanase (Cex) has been cloned, expressedin E. coli and subsequently sequenced (O’Neill et al., 1986). The C. fimi exoglycanase is a47 kDa protein comprising two domains which are separable by limited proteolysis into anactive catalytic domain of 35 kDa and a cellulose-binding domain (—12 kDa) which bindscellulose and chitin (Ong et aL, 1993) (Figure 1-5). These two domains are held together7by a linker region rich in proline and threonine residues. The 35 kDa catalytic domain iscapable of catalysing the hydrolysis of small substrates (e.g. aryl [3-cellobiosides), butcannot bind cellulose. C. fimi exoglycanase is a member of Family 10 (Henrissat &Bairoch, 1993) of 13—glycanases, a group which also contains 8 xylanases. The three-dimensional structure of the catalytic domain of this enzyme has recently been solved by Xray crystallography to 1.8 A resolution (Figure 1-6) (White et al, 1994). The catalyticdomain folds into an z43—barrel containing eight parallel 13-strands. There is an open cleftat the carboxyl terminal end and this is proposed to be the active site.e1lulosebinding domaincatalyticdomainFigure 1-5 Schematic representation of the structure of intact C.fimi exoglycanase.Previous investigations of the substrate specificity of C.fimi exoglycanase revealedit to be capable of hydrolysing cellulose (Figure 1-7, R = CH2OH), xylan (Figure 1-7, R =H) (Gilkes et al., 1984) and small substrates such as aryl [3-xylobiosides, aryl 13-cellobiosides and aryl 13-D-glucopyranosides (Tull et al., 1991). Analysis of thestereochemical outcome of the hydrolysis of glycosides by C. fimi exoglycanase using 1H-NMR revealed it to be a “retaining glycosidase (Withers et al., 1986).8Figure 1-6 The X-ray crystal structure of the catalytic domain of C.fimi exoglycanase(White et a!., 1994).9R=HorCH2OHFigure 1-7 Structure ofxylan and cellulose.Two mechanisms put forth to describe the action of ?IretainingH glycosidases are a doubledisplacement mechanism (Koshland, 1953) and a ring opening reaction (Capon, 1969; Post& Karplus, 1986). In recent years however, there has been an explosion of experimentalevidence in support of the double displacement mechanism (see review by Sinnott, 1990)while the latter mechanism still lacks serious experimental support. This doubledisplacement mechanism for C.fimi exoglycanase is illustrated in Figure 1-8. The first stepinvolves displacement of the aglycone by an appropriately positioned carboxylatenucleophile in the active site to form an x-glycosyl-enzyme intermediate. In a second step,the intermediate is hydrolysed to yield 13-cellobiose as the initial product and to regeneratethe free enzyme. Both formation and hydrolysis of the glycosyl-enzyme intermediate occurvia oxocarbonium ion-like transition states. Initial C-O bond cleavage is proposed to occurwith acid assistance from an appropriately located carboxylic acid residue while hydrolysisof the intermediate occurs with general base catalysis from the same residue. It has alsobeen suggested that most of the observed rate acceleration is expected to be derived fromthe non-covalent interactions between the enzyme and the substrate and its transition states.HROH10DEGLYCOSYLATION°N°1HHA6-0 HQ.HO H’0R11frRoH1k +H20H0 HO. IHOHc4HO1AHoc_o1GLYCOSYLATIONA-H AHFigure 1-8 Double displacement mechanism proposedfor C. fimi exoglycanase.111-5 Mechanism Of “Retaining” Glycosidases: Evidence For A DoubleDisplacement MechanismThe large quantity of experimental evidence in support of several features of adouble displacement mechanism for “retaining” glycosidases including C. fimiexoglycanase is discussed below.1-5-1 Presence of a carboxylate nucleophileSome of the most compeffing evidence for the presence of an enzymic nucleophileis derived from X-ray crystallographic studies. The earliest evidence came from X-raycrystallographic studies of hen egg white lysozyme which is responsible for catalysing thehydrolysis of the peptidoglycan component of bacterial cell walls (Figure 1-9).Figure 1-9 The peptidoglycan substratefor lysozyme.The three-dimensional structure of hen egg white lysozyme (HEWL) identifiedcarboxylate groups corresponding to Asp 52 and Glu 35 correctly positioned in the activesite (Grutter et al., 1983). Recent X-ray crystallographic studies on other “retaining”glycosidases including Trichoderma harzianum xylanase (Campbell et al., 1993), C. fimiexoglycanase (White et al, 1994) and E. coli -galactosidase (Jacobsen et al., 1994), havealso identified carboxylate residues appropriately located to function as the catalyticnucleophile. Even more direct evidence for the existence of a catalytic carboxylate isOHOH.OHR =CH3HCOO-Peptide12derived from glycosidases co-crystallized with either an inhibitor or a substrate in the activesite. For example the X-ray crystal structure of Bacillus circulans E172C xylanasexylotetraose complex revealed that Glu 78 is correctly oriented and within close enoughproximity of the substrate to act as the catalytic nucleophile (Campbell et al., 1993) whilethe structure of Trichoderina reesel cellobiohydrolase I with 2-iodobenzyl- 1-tMo--D-cellobioside (Figure 1-10) at the active site indicates that Glu 212 is likely the catalyticnucleophile based on its position and environment within the active site (Divne et al, 1994).In other cases the catalytic carboxylate groups have been identified by labelingexperiments using group-specific labels. For example, the carboxylate-modifying reagent,1-(4-azonia-4,4-dimethyl-pentyl)-3-ethylcarbodiimide iodide (EAC), has been used toidentify Glu 87 as a substrate-protected carboxylate group in the active site ofSchizophyllum commune xylanase A. This residue has been proposed to be the catalyticnucleophile in this enzyme (Bray & Clarke, 1994).Labeling experiments using the mechanism-based inactivators, 2-deoxy-2-fluoro-glycosides, have been used successfully to identify the catalytic nucleophile. Thesecompounds operate by trapping the glycosyl-enzyme intermediate, thus allowingidentification of the nucleophile by subsequent isolation and sequencing of the labeledpeptides. Indeed, the catalytic nucleophiles of several “retaining” glycosidases have beenidentified with these inactivators and these are listed in Table 1-2.OHOH0Figure 1-10 Structure of2-iodobenzyl-1 -thio-J3-D-cellobioside.13Table 1-2: List of catalytic nucleophiles identified by mechanism-based inactivators, 2-deoxy-2-fluoro-glycosides.Glycosidase Nucleophile Inactivator ReferenceAgrobacterium-Glucosidase Glu 358 2F-DNPG (Withers et al.,1990)C.JlmiExoglucanase Glu 233 2F-DNPG (Tull et al, 1991)C. thermocellumEndoglucanase C Glu 280 2F-DNPC (Wang et al., 1993)E. coli3-Galactosidase Glu 537 2F-DNPGa1 (Gebler et al., 1992)HumanGlucocerebrosidase Glu 340 2F-DNPG (Miao et al., 1994)B. subtilisxylanase Glu 78 2F-DNPX (Miao et aL, 1994)Labeling studies of this sort with C. fimi exoglycanase and E. coli f3-galactosidase haveidentified Glu 233 and Glu 537, respectively, as the catalytic nucleophiles in these enzymesand these findings were recently confirmed by the X-ray crystallographic structures whichrevealed these residues appropriately located within the active site to perform this role(White et al, 1994; Jacobsen et al, 1994).141-5-2 Nature of the glycosyl-enzyme intermediateThe double displacement mechanism initially described by Koshland (1953) for“retaining” glycosidases proposed the formation of a covalent glycosyl-enzymeintermediate. However, based on the X-ray structure of the HEWL active site, Phillips hasproposed that an ion-pair comprising Asp 52 and the oxocarbonium ion would besufficiently long-lived for the leaving group to diffuse away from the active site and allowthe glycosyl acceptor, water, to diffuse in and react (Blake et al., 1967). While additionalevidence in support of an ion-pair intermediate comes from a recent X-ray crystallographicstudy of HEWL complexed with a product, N-acetyl muramyl-N-acetyl glucosaminyl-Nacetyl muramic acid (Strynadka & James, 1991), there has been considerably moreevidence which strongly indicates formation of a covalent glycosyl-enzyme intermediate formany “retaining” glycosidases.Substantial evidence for the covalent nature of a glycosyl-enzyme intermediatecomes from secondary deuterium kinetic isotope effect measurements using substrates forwhich deglycosylation is rate determining. Values of kHIkJ = 1.20 - 1.25 have beenmeasured for E. coli (lac Z) -galactosidase (Sinnott, 1978), kJ/kD = 1.10 - 1.12 forAgrobacterium -glucosidase (Kempton & Withers, 1992) and kH/kD = 1.09 forBotrydiplodia theobromae -g1ucosidase (Umezerika, 1988). These positive kinetic isotopeeffects reflect significant sp3 to sp2 rehybridisation at the anomeric centre (C-i) in goingfrom the ground state glycosyl-enzyme intermediate to the deglycosylation transition stateand this can only occur if the intermediate has more sp3 character than the subsequenttransition state.Additional evidence for the covalency of the intermediate derives from inactivationstudies of Agrobacterium f3-glucosidase with 2-deoxy-2-fluoro-3-D-glycosides (Withers etal., 1987). These compounds inactivate 3-glucosidase because the presence of theelectronegative fluorine at C-2 destabilises both oxocarbonium ion-like transition states,thereby slowing down both the formation and the hydrolysis of the intermediate while the15presence of the good leaving group (fluoride or 2,4-dinitrophenolate) speeds up theglycosylation step thereby leading to the accumulation of the intermediate (discussed inChapter II) (Figure 1-1 1). This glucosyl-enzyme intermediate is stable, with t1,2 > 500hours at 37 °C in phosphate buffer, thus allowing the covalency and stereochemistry (a) tobe demonstrated by F-NMR (Withers & Street, 1988) as well as identification of thenucleophile as Glu 358 (Withers et al. 1990). Trapping of this covalent species shows thatsuch glycosyl-enzyme intermediates can exist and are stable.02N_- NO2Glu 358Figure 1-11 Labeling the catalytic nucleophile (Glu 358) ofAgrobacterium /3-glucosidase with 2F-DNPG.Furthermore, the catalytic competence of the intermediate is demonstrated by the fact that inthe presence of a suitable glycosyl acceptor such as 1-deoxy-3-D-glucosyl benzene,product is released and the free enzyme is reactivated. Analysis of this isolated reactivationproduct by mass spectrometry and both ‘H and 19F-NMR was consistent with the structureof the disaccharide, 2’-deoxy-2’-fluoro 1-deoxy-3-cellobiosyl benzene. Presumably, thisproduct is formed via a transglycosylation reaction involving attack on the anomeric carbonof the 2F-glucosyl-enzyme intermediate by the C-4 hydroxyl group of glucosyl benzenefollowed by subsequent displacement of the enzymic carboxylate group (Figure 1-12).OHH(OHNO2 HGlu 35816÷ Ho0f3Gin 358Glu 358Figure 1-12 Reactivation of2F-DNPG-inactivatedAgrobacterium f3-glucosidase in thepresence of 1 -deoxy-/3-glucosyl benzene.Finally, studies of spontaneous (non-enzymatic) hydrolysis of glycosides havesuggested that glycosyl cations are relatively unstable with estimated life times varyingfrom 10-10 to 10-12 seconds in aqueous solutions (Amyes & Jencks, 1989; Bennet &Sinnott, 1986) compared to life times of 1 - 100 milliseconds for glycosyl-enzymes atambient temperatures (Weber & Fink, 1980). It therefore seems unlikely that such anunstable species could exist, even within the stabilizing environment of an active site,without formation of a covalent bond (Sinnott & Souchard, 1973).OHHOFOH171-5-3 Oxocarbonium ion-like transition statesEvidence for oxocarbonium ion character at the transition states for formation andhydrolysis of the glycosyl-enzyme intermediate comes primarily from studies of transitionstate analogues, secondary deuterium kinetic isotope effects and linear free energyrelationships using fluorinated and deoxygenated substrates. In addition, studies with 1,1 -difluoro-substrates have provided a measure of the oxocarbonium ion character at theglycosylation transition state. The results from these studies are presented and discussedbelow.Transition state analogues are glycoside derivatives that sufficiently mimic theoxocarbonium ion-like transition state in structure that they bind to the glycosidasesignificantly tighter than does the substrate. Features distinguishing the glycosyl cationfrom the parent glycoside are the following. Both the C-i and 0-5 atoms of the glycosylcation share a full positive charge and the C-i, C-2, C-5 and 0-5 atoms are coplanar(Figure 1-13) (Sinnott, 1987). Hence, a transition state analogue must have one or both ofthese features.versusORGlucosyl cationGlucosideFigure 1-13 Comparison ofglycosyl cations with glycosides.Examples of classical transition state analogues include the aldonolactones andaldonolactams which resemble the glycosyl cation both in geometry and to some extent incharge (Figure 1-14). Generally, values of K for these lactones are 102 to i04-fold lowerHOH18than K values for the corresponding hexoses. For example, with Aspergillus wentii -glucosidase K (lactone) = 0.0095 mM versus K1 (hexose) 2.8 mM, with Jack bean ccmannosidase K1 (lactone) = 0.12 mM versus K (hexose) =22 mM and finally with HEWLK (lactone) = 0.000083 mM versus K (hexose) 0.01 mM (Legler, 1990). Comparablevalues of K1 for aldonolactams with glycosidases have been reported (Legler, 1990).HO>AldonolactoneHOç5AldonolactamFigure 1-14 Resonance structuresfor an aldonolactone and an aldonolactam.Some of the most potent inhibitors of glycosidases include the 5-amino-5-deoxy-aldose class of transition state analogues such as nojirimycin (Figure 1-15) and itsanalogues. Values of K for these 5-amino-5-deoxy-aldose inhibitors with theircorresponding glycosidase are found within the micromolar range, for example K1(nojirimycin) = 0.36 .tM for A. wentii 3-glucosidase, K1 (galactonojirimycin) = 0.045 I.LMfor E. coli 3-galactosidase and K (mannonojirimycin) 1.2 .tM for Jack bean (Xmannosidase (Legler, 1990). These compounds are thought to mimic the glycosyl cationbecause in their protonated form they are isoelectronic with the glycosyl cation. It is likely19that the dehydrated forms of 5-amino-5-deoxy-aldose inhibitors are also excellent inhibitorsof glycosidases, since in this form they are thought to mimic the glycosyl cation both incharge and in geometry.H OH-H20 HO-OH +H20 OHFigure 1-15 Nojirimycin, a transition state analoguefor glucosidases.Secondary deuterium kinetic isotope effect measurements on glycosidases usingspecific substrates for which the rate determining step is either glycosylation ordeglycosylation have provided substantial evidence for oxocarbonium ion-like transitionstates during glycoside hydrolysis. For example, kinetic isotope effects measured on theglycosylation transition state include values of kI-I/kD = 1.15 - 1.20 for E. coli (lac Z) fgalactosidase using 3-D-galactopyranosyl pyriclinium salts (Sinnott & Withers, 1974) andkH/kD = 1.05 - 1.07 for Agrobacterium 13-glucosidase using aryl 3-D-glucopyranosides(Kempton & Withers, 1992). Similarly, kinetic isotope effects have also been reported onthe deglycosylation transition state and these include values of kH/kD = 1.2 - 1.25 for E.coli -galactosidase (Sinnott, 1978) and kHJkD = 1.09 for Botrydiplodia theobromae -glucosidase (Umezerika, 1988).The presence of a secondary deuterium kinetic isotope effect of kWkD > 1 indicatesthat the isotopically-substituted carbon centre is converting from an sp3 hybridized groundstate to a transition state with significant sp2 character in the rate determining step. Ifhowever, a value of kH/kD = 1 was measured, then that would indicate either an SN2mechanism as there would be no change in hybridization or that another step was rateHO.OH20determining. Alternatively, values of kH/kD < 1 are indicative of a change in hybridizationfrom sp2 at the ground state to sp3 at the transition state. Thus, the kinetic isotope effectmeasurements presented above, kH/kD > 1, are entirely consistent with oxocarbonium ioncharacter at both transition states.Additional evidence for oxocarbonium ion-like transition states during glycosidasecatalysed hydrolysis of glycosides derives from a linear free energy relationship study of E.coil 3-galactosidase using deoxyfluoro-galactoside substrates substituted with fluorine atpositions 2, 3, 4 and 6 around the galactosyl ring and deoxy-galactoside substrates withhydrogen substituted at positions 4 and 6 (McCarter et al., 1992). The correlation observedbetween log(kca/Km) for the galactosidase-catalysed hydrolysis and log(k11) for thespontaneous hydrolysis with these galactosides suggested that both the enzymic transitionstate and that for spontaneous hydrolysis are similarly affected by the substitutions at eachposition on the galactose ring. Since the major effect of substitution of a hydroxyl group bya fluorine or a hydrogen atom on the rate of spontaneous hydrolysis is electronic in nature(Withers et al., 1989; Withers et aL, 1986), then the presence of this correlation withreaction constant of p = 0.8 provides direct evidence for the similar oxocarbonium ioncharacter in the enzymic transition state. The scatter observed in the log(kca/Km) versuslog(k) plot (McCarter et al., 1992) is due to binding effects that are important in theenzyme-catalysed hydrolysis reaction (discussed in Chapter II) but are not a component ofthe spontaneous hydrolysis reaction. Similar correlations between enzymic andspontaneous hydrolysis of deoxy-and deoxyfluoro-substrates have been observed in linearfree energy relationship studies of other related enzymes systems in which the transitionstate is believed to be oxocarbonium ion-like in nature. For example, with rabbit muscleglycogen phosphorylase b, an enzyme responsible for catalysing the reversiblephosphorolysis of glycogen to produce glucose-i-phosphate, a reasonable correlation witha reaction constant of p = 0.9 was observed between log(Vm) for the phosphorylasecatalysed reaction and log(khthO15)for the acid-catalysed hydrolysis reaction of deoxy21and deoxyfluoro-x-D-glucopyranosyl phosphates (Street et al., 1989). These results, likethose with E. coli 3-ga1actosidase, are consistent with oxocarbonium ion character at thetransition state for phosphorylase-catalysed hydrolysis of glucopyranosyl phosphates.Recently, Sinnott and coworkers (Konstantinidis & Sinnott, 1991; Srinivasan etal., 1993) have attempted to quantitate the oxocarbonium ion character at the first transitionstate of “retaining” glycosidase-catalysed hydrolysis of glycosides using 1-fluoro-D-glycosyl fluorides and 1, 1-difluoro-D-glycosyl fluorides (Figure 1-16). Values of kcat/Kmfor both the mono- and di-fluoride substrates were determined and used to calculate thedifferences in the free energies of activation (MG) for the glycosylation step.Figure 1-16 Structure of I -fluoro -D-glucosylfluorides and 1,1 -duoro-D-glucosylfluoride.Since introduction of an electronegative substituent such as fluorine at an already electron-deficient centre is expected to further destabilise that transition state then the values ofMG are estimates of the degree of destabilisation caused by this second fluorine atom atthe anomeric centre and thus provide some measure of the oxocarbonium ion character atthe first transition state during glycoside hydrolysis. Rates of hydrolysis for the appropriateglucosyl fluoride anomer were 5000-fold (AAG = 21.2 kJ mol1), 8000-fold (zXzG =23.2 kJ mol-1)and 200-fold (MGt = 14.5 U mo11)greater than those for the l-fluoro-Dglucosyl fluoride with yeast x-glucosidase, sweet almond 3—glucosidase B andAspergillus wentii 3-glucosidase A3, respectively (Konstantinidis & Sinnott, 1991).OH OHHOHF22Similarly, the rate of E. coli (lac Z) 3-galactosidase-cata1ysed hydrolysis of f-D-galactosy1fluoride was 400-fold (AAGt = 15.2 U mo!-1) greater than that for the correspondingdifluoride (Srinivasan et al, 1993). These results indicate that the presence of the secondfluorine atom at C-i destabiises the glycosylation transition state and thus is entirelyconsistent with the presence of oxocarbonium character at the glycosylation transition state.1-5-4 General acid assistanceKoshland proposed that departure of the aglycone may proceed with acidassistance. Evidence which supports acid catalysis includes identification of suitablypositioned carboxylic acid residues, Glu 35 and Glu 11, in the X-ray crystal structures ofHEWL (Anderson et al., 1981) and GEWL (Grutter et al, 1983) respectively, which couldfunction as acid catalysts. Similarly, the X-ray crystal structure of E. coli J3-galactosidaserevealed Glu 461 appropriately located within the active site to perform the role of acidcatalyst in this enzyme (Jacobsen et al, 1994).MacLeod and coworkers (1994) have developed a method to identify the acidcatalyst in glycosidases. This new method was applied to C. fimi exoglycanase and itinvolves identification of highly conserved carboxylic acid residues by means of sequencealignment, site-directed mutagenesis of these residues to alanine and glycine followed bykinetic analysis of these mutants using substrates for which protonation of the leavinggroup is required (e.g. 4-bromophenyl 13-cellobioside) and substrates requiring no acidassistance (e.g. 2,4-dinitrophenyl f3-cellobioside). Glu 127 was identified as the acidcatalyst in this enzyme by this means and this assignment has recently been confirmed byX-ray crystallographic analysis of the enzyme. Glu 127 was found to be appropriatelylocated to perform the role (White et al, 1994) (discussed in Chapter III).Additional supportive evidence for acid assistance in glycoside hydrolysis isderived from glycosidase-catalysed hydration of octenitol derivatives and glycals. E. coli 3-galactosidase and coffee bean tx-galactosidase catalyse the hydration of D-galacto-octenitol23to the galactooctulose derivatives (Figure 1-17) and these products were characterized by1H-NMR (Weiser et al., 1992). This study showed that based on the stereochemistry of thehydrated products, protonation occurred from the a-face of the galactose ring with the agalactosidase and the 3-face with the 3-galactosidase. These results are therefore consistentOHFigure 1-17 Hydration ofan octenitol derivative by E. coli J3-galactosidase.with the presence of a residue in these enzymes that could function as an acid catalyst in thenormal glycoside hydrolysis reaction.Another 1H-NMR stereochemical investigation, in this case of coffee bean agalactosidase with D-galactal in 1)20 revealed that deuteration (protonation) was from thea—face of the sugar ring and again this is consistent with the existence of an acid catalystthat could donate a proton in normal glycoside hydrolysis (Weiser et al, 1992). Similarresults have also been obtained with three 3-N-acetyl hexosaminidases from Jack bean,bovine kidney and human placenta using 2-acetamidoglucal. In all three cases, thehydration product identified by HPLC using authentic standards was N-acetyl-glucosaminewhile no observable amounts of N-acetyl-mannosamine were detected, indicating that theOHDHD24H‘OHTAcOHFigure 1-18 Hydration of2-acetamidoglycal by f3-N-acetyl hexosaminidasesfrom Jackbean, bovine kidney and human placenta.proton was indeed delivered (Figure 1-18) from the n-face of the sugar ring (Lai &Withers, 1994). It should be noted however that in general, protonation of glycals occursfrom the face of the sugar that is opposite to the acid catalyst (Sinnott, 1990). For example,1H-NMR analysis of the hydration of cellobial with Irpex lacteus exo-[-(1,4)-cellulase andAspergillus niger endo-3-(1,4)-cellulase in D20 revealed that deuteration (protonation)occurred from the a-face of the cellobiose ring (Kanda et al., 1986). In these cases it isbelieved that the proton is delivered by the catalytic nucleophile (Hehre et al., 1977; Kandaet al, 1986) and that reaction occurs via a concerted mechanism (Legler, 1990) (discussedin Chapter II).Although there is evidence which clearly indicates the involvement of a general acidcatalyst, there is also evidence suggesting that it is not essential for glycoside hydrolysis.For example, studies of E. coli f-galactosidase-catalysed hydrolysis of 13-D-galactopyranosyl pyridinium salts (Figure 1-19) revealed 108 to 103-fold rateenhancements for the enzymatic hydrolysis compared to spontaneous hydrolysis (Jones eta!., 1977). Since it is structurally impossible to protonate these compounds in a mannerH25which will assist aglycone departure, then none of this rate acceleration can be due to acidcatalysis.OHOHFigure 1-19 Structure of J3-D-galactosyl pyridinium cation.1-5-5 Non-covalent enzyme-substrate interactionsThe large majority of the rate enhancement observed in catalysis by most enzymesis believed to be derived from non-covalent interactions between the enzyme and thesubsirate as the transition state is reached. With glycosidases and other carbohydratebinding proteins, these interactions are likely to be predominantly hydrogen bonds formedbetween the hydroxyl groups of the sugar and the enzyme. The binding energy derivedfrom these interactions, when realized at the transition state, stabilizes that transition stateand leads to a decrease in the activation energy of the reaction, hence a rate acceleration isobserved.Evidence for the existence of hydrogen bonds between hydroxyl groups ofcarbohydrates and proteins is derived from a number of sources including X-raycrystallographic studies. Examples in which this has been seen have included lectins,proteins that recognize and bind to complex carbohydrates, with oligosaccharides bound atthe recognition sites. For example, the X-ray structure of Erythrina corallodendron lectinN-linked to a heptasaccharide and complexed with lactose revealed hydrogen bondsbetween the C-3 and C-4 hydroxyl groups of the galactose residue and Asp and Asnresidues in the binding site (Shaanan et al., 1991). Similarly, the X-ray structure of rat26serum lectin complexed to an oligomannose asparaginyl-oligosaccharide revealed hydrogenbonds between the C-3 and C-4 hydroxyl groups of terminal mannose and Asn and Gluresidues in the binding site (Weis et al., 1992). Indeed, hydrogen bonds have also beenidentified in glycosidases, for example hydrogen bonds between C-3 and C-6 hydroxylgroups of a f3—(1,4)-N-acetyl-D-glucosamine trisaccharide and the active site Asp and Trpresidues of lysozyme were identified in the X-ray crystal structure of the enzyme-sugarcomplex (Johnson et al, 1988).Studies of glycosidases using deoxy-glycosides as substrates have given someindication of how much the interaction of each individual hydroxyl group can contribute totransition state stabilization and thus rate acceleration. For example, with E. coli [3-galactosidase, values of kJKm for the hydrolysis of 4-deoxy and 6-deoxy analogues of 2,4dinitrophenyl 3-D-galactoside were 500- and 1000-fold lower than that for the parentgalactoside substrate (McCarter et a!, 1992). Interactions at the C-2 hydroxyl group mustbe more important since the rate of hydrolysis of the 2-deoxy-galactosyl-enzymeintermediate is 5 x 105-fold lower than that for the hydrolysis of the parent galactosylenzyme intermediate (Wentworth & Wolfenden, 1974). Thus, the C-2 hydroxyl groupcontributes at least 33.5 kJ/mole (8 kcal/mole) to the stabilization of the degalactosylationtransition state (McCarter et al, 1992). Similarly, the C-5 hydroxymethyl group inlysozyme substrates must also provide important non-covalent interactions at the transitionstate since the rate of hydrolysis of the 6-deoxy thsaccharide shown in Figure 1-20 was1300-fold lower than that for the corresponding C-5 hydroxymethyl substrate (Figure 1-20) (Ballardie et al., 1977).27R=HorCH2OHFigure 1-20 Substrate analogue for lysozyme.NO2With Aspergillus niger glucoamylase, the kca/Km values for 4’- and 6’-deoxy maltosederivatives (Figure 1-2 1) are 1000- and 700-fold lower than those for the parent substrate,clearly indicating that these hydroxyl groups are involved in stabilizing the transition state(Sierks et al., 1992).(a)R1,R2=H,OH(b)R,R=OH,HOHR2R1HOH OHFigure 1-21 Structure of deoxy-maltose derivatives.281-5 Aim of This StudyC. fimi exoglycanase has been classified as a retaining -glycosidase and thus adouble displacement mechanism is proposed. The aim of this study is to investigate themechanism of this enzyme. This will be accomplished by first performing a detailed kineticanalysis on the native enzyme using pre-steady state analysis, linear free energyrelationship studies, secondary deuterium kinetic isotope effect measurements, inactivationstudies and pH-dependence studies. Pre-steady state analysis of the hydrolysis of aryl -glycosides should provide evidence for a two step mechanism, identify the rate determiningstep and provide values for the rate constants for formation of the glycosyl-enzymeintermediate (k2) and values for the dissociation constant for the enzyme-substrate complex(Kd). Linear free energy relationship studies should identify the rate determining andprovide insights into the degree of negative charge accumulation on the phenolate oxygen atthe glycosylation transition state. Secondary deuterium kinetic isotope effect measurementsshould provide a measure of the oxocarbonium ion character at each transition state.Inactivation studies with 2-deoxyfluoro-sugars coupled with‘9F-NMR analysis shouldprovide substantial evidence for the covalent nature and stereochemistry of the glycosylenzyme intermediate. Furthermore, inactivation studies with both 2-deoxy- and 2-deoxyfluoro- glycosides should provide evidence for the role of the C-2 hydroxyl group ofthe sugar during exoglycanase-catalysed hydrolysis of glycosides. The pH-dependencestudy of the exoglycanase activity may provide insight into the ionisation state of active siteresidues.The mechanism of this exoglycanase will be further investigated by performing adetailed kinetic analysis similar to that performed on the native enzyme on point mutantsgenerated individually at the catalytic nucleophile (Glu 233) and the acid-base catalyst (Glu127). This is expected to provide further evidence in support of the proposed roles of theseresidues.29A second aim of this study is to use active site directed irreversible inactivators incombination with electrospray ionisation mass spectrometry to identify active site residuesthat are involved in either catalysis or binding.More detailed aims, specific to each project, are provided at the end of theintroduction of each chapter (Chapters 11,111 and IV).30CHAPTER IIDETAILED KINETIC ANALYSIS OF NATIVE CELLULOMONAS FIMIEXOGLYCANASE312-1 Introduction to Different Kinetic Techniques2-1-1 Kinetic scheme for a double displacement mechanism“RetainingT’glycosidases such as C. fimi exoglycanase are assumed to catalyse thehydrolysis of 3-glycosides via a double displacement mechanism (Koshland, 1953). Akinetic scheme describing this reaction is______k2 k3E + Sk 1E.S E-S’ j E + P2where k1 and k..1 are the rate constants for formation and dissociation of the Michaeliscomplex (ES), respectively, while k2 and k3 are the rate constants for formation andhydrolysis of the glycosyl-enzyme intermediate (E-S’), respectively. The dissociationconstant for the Michaelis-complex is expressed as[E] [SI[ES]thus,k1Kd=k1The Michaelis-Menten parameters, kcat, Km and kcatfKm can be defined in terms of the rateconstants ask23kcat= k2 + k332K(k..1+k2 k3m’\ k1 )\k2+3kcat — k12Km — k1 + k2Values for kcat, Km and kcat/Km are obtained by performing steady state kinetic analysis.2-1-2 Pre-steady state analysis of enzymatic reactionsAlthough steady state kinetic analysis of glycosidases can provide values for bothkcat and Km, such analysis provides very little information about the individual rateconstants. Alternatively, -steady state kinetic analysis of enzymatic reactions can providevalues for the rate constants of the individual steps in the mechanism as well as detecttransient intermediates along the reaction pathway. When performing pre-steady stateanalysis on “retaining” glycosidases, only the initial conversion of the Michaelis-complex(E.S) to the glycosyl-enzyme intermediate (E-S’) is observed. Since only one mole of theaglycone is released per mole of enzyme, then these studies require large amounts ofenzyme rather than catalytic quantities, as well as a large excess of substrate over enzyme inorder to maintain pseudo-first order kinetics. Pre-steady state kinetic analysis of “retaining”glycosidases can be used to specifically characterise the formation of the glycosyl-enzymeintermediate provided the following criteria are satisfied. First, the rate of formation of theMichaelis-complex (k1) must be greater than that for formation of the glycosyl-enzymeintermediate (k2), otherwise the reaction may be at least partially diffusion-controlled.Second, the rate of dissociation of the Michaelis-complex to free enzyme and substrate (Ici)must be greater than that for its conversion to the glycosyl-enzyme intermediate (k2) inorder to obtain a reliable estimate of the dissociation constant for the Michaelis-complex33(Kd). Last, if the rate of the formation of the glycosyl-enzyme intermediate (k2) is verymuch greater than that for its conversion to product and free enzyme (k3), then the reactionscheme can be simplified tok1E+S - E.S— E-S’k..11Ideally, if the rate of glycosylation (k2) is much greater than that for deglycosylation (k3)then formation of the glycosyl-enzyme intermediate would be expected to follow pseudo-first order kinetics over the entire observation period of the reaction. In general however,biphasic kinetics are observed for the release of the aglycone. During the pre-steady statephase of the reaction, initial aglycone release (the burst) follows pseudo-first order kinetics,but as the glycosyl-enzyme intermediate accumulates the rate of release of the aglyconebecomes dependent on the rate of conversion of the intermediate to product and freeenzyme. Eventually, a steady state is reached and the rate of aglycone release becomesconstant. These biphasic kinetics are then fitted to an equation describing a first orderprocess followed by a steady state phase to yield values for the pseudo-first order rateconstants (kobs) of the pre-steady state phase as well as values for the steady state rates.The pseudo-first order rate constants are fitted to the Michaelis-Menten equation,k2 [S]kobs= Kd + [S]thus yielding values for k2 and K1.342-1-3 Linear free energy relationships as mechanistic probes of enzymaticreaction(a) The concept: A series of modifications in the conditions of a reaction will nearlyalways result in a series of changes in either the rate and/or equilibrium of the reaction. Ifthe same modifications affect either the rate or equilibrium of a reaction in the same waythat they affect either the rate or equilibrium of a second reaction, then a linear free energyrelationship must exist between the two sets of effects. This relationship is expressedmathematically ask2, k1log= log ‘‘ x constantk2 k1where k represents either the rate or equilibrium constant of a reaction and k, that of thereaction in which there has been a modification in the reaction conditions (e.g. addition of asubstituent). The term “free energy” in this relationship is appropriate as the rate is areflection of the free energy of activation and the equilibrium constant reflects the standardfree energy change of the reaction.(b) Linearfree energy relationships in enzymology: Linear free energy relationshipsare valuable tools for the elucidation of the mechanism of enzymatic reactions since theycan provide valuable insights into the relationship between substrate structure andenzymatic activity. However, their applications are limited by two important aspects ofenzymatic reactions, substrate binding and catalysis. Enzymes are rather specific for theirsubstrates, thus modifications of the substrate could affect its binding to the enzyme activesite. The resulting binding effects may then obscure the electronic effects of the substituentson the substrate. Furthermore, substrate modification could affect the orientation of thesubstrate reacting centre relative to the enzymatic catalytic residues and this in turn could35interfere with catalysis. Nevertheless, in favourable circumstances linear free energyrelationships can be applied to enzymatic reactions. These have been applied extensively tothe study of glycosidases (Dale et al., 1986), acyl transferases (Lowe & Yuthauong, 1971)and proteases (Riddle & Jencks, 1971). Since this work focuses on a glycosidase, anexample from this group of enzymes has been selected in order to illustrate the type ofresults obtained and how they are interpreted to provide mechanistic insights.(c) Example of the application of linear free energy relationship: The “retainingglycosidase, Agrobacterium faecalis f3-glucosidase catalyses the hydrolysis of aryl 13-D-glucosides to glucose and the corresponding substituted phenol (Kempton & Withers,1992). Values of log for these substrates were plotted as functions of the phenolleaving group pKa in the form of a Broensted plot (Figure 2-1). The shape of the resultingplot is concave downward, providing evidence for a two step reaction mechanism and thusis consistent with the formation of a glucosyl-enzyme intermediate along the reactionpathway. The biphasic nature of this plot further indicates a change in the rate determiningstep as the phenol leaving group changes. The strong correlation observed between log k<for substrates with poor leaving groups (pKa> 8) and the phenol pKa suggests that theformation of the glucosyl-enzyme intermediate is the rate determining step for thesesubstrates as this is the step involving cleavage of the phenolate-sugar bond. The value ofthe reaction constant, 13g = -0.7, for this region of the plot is consistent with significantnegative charge development on the oxygen of the departing phenolate at the glucosylationtransition state. By contrast, the absence of any significant dependence of log kcat on thephenol pKa for the more reactive substrates (pKa < 8) indicates that the initial C-O bondcleavage is not rate determining, but instead the rate determining step is most likelyhydrolysis of the glucosyl-enzyme intermediate.36300 o oo2 o-Go1Oo0 • . i . .3.5 5.5 7.5 9.5pKaFigure 2-1 Broenstedplot relating values of kcaj for the hydrolysis of aryl /3-glucosidesby Agrobacterium J3-glucosidase with the leaving group ability of thephenols. (Reproducedfrom Kempton & Withers, 1992).The above example demonstrates that application of linear free energy relationshipsin enzymology can provide useful mechanistic information about the identity of ratedetermining steps, the existence of intermediates along the reaction path, as well as valuablestructural information about transition states.2-1-4 Secondary deuterium kinetic isotope effects as transition state probes(a) The concept of kinetic isotope effects : A kinetic isotope effect is the differencein the rate of a reaction resulting from an isotopic substitution in a reactant. It is given bythe ratio of the rate constants for the isotopically unsubstituted and substituted reactants, forexample kHIkD for deuterium effects. If the bond to the isotopically substituted atom is37formed or broken in the rate determining step, then a primary isotope effect can bemeasured for the reaction. Alternatively, if a bond other than that to the isotopicallysubstituted atom is formed or broken, then a secondary isotope effect can be measured forthe reaction. These secondary isotope effects are the result of rehybridization occurring atthe reacting centre in the rate determining step. Secondary isotope effects are classified asa, [3, y or (remote effects) based on the position of the isotopically substituted atomrelative to the reacting centre. If the isotope is bonded directly to the atom undergoingrehybridization, then an a-secondary isotope effect is measured whereas a 13-secondaryisotope effect is measured when the isotope is bonded to an atom adjacent to the reactingcentre.Although the isotopic substitution does not affect the electronic tructure of amolecule, the mass difference between the isotopes will be reflected in the frequency of thevibrating atoms according toI ‘. / \1/2\21tJ\jtwhere K, the force constant, of the bond does not change with isotopic substitution and p,the reduced mass of the vibrating system, is given bym12m1 + m2so that m1 and m2 are the masses of the two atoms forming the bond. If one mass is muchlarger than the other, for example a C-H bond, then is approximately equal to the smallermass. Since all molecules have an intrinsic non-zero minimum energy called the zero-pointenergy expressed asEo=4hV38then the zero-point energy can be related to the mass of the vibrating atoms by/ \112_l(K4itFor example, since deuterium has a greater mass than hydrogen, then the vibrations of theC-D bond contributes less to the zero-point energy of the molecule than those of the C-Hbond. Thus, substitution of deuterium for hydrogen results in reduction of the molecule’szero-point energy.(b) a-Secon&zry deuterium kinetic isotope effects: These effects arise from changesin the frequencies of the bending mode vibrations of the isotopic atoms upon reaching thetransition state. The difference in the zero-point energy between the isotopically labeledreactants upon reaching the transition state depends on the changes in the force constants(K) of the bending mode vibration for the isotopic bond. If the reacting centre undergoessp3 to sp2 rehybridization at the transition state of the rate determining step, then thefrequency of the C-H(D) bending mode vibration is lowered, decreasing the force constantof the vibration. Since the difference in zero-point energies between the C-H and C-D bondvibrations is proportional to (K4t)’/2, then the difference in zero-point energies at thetransition state will be less than that at the ground state. Thus, the activation energy for thehydrogen-substituted reaction is lower than that for the deuterium-substituted reaction,resulting in a greater reaction rate for the hydrogen-substituted reaction. For these reactiontypes, “normal” kinetic isotope effects of kJ.JIkD> 1.0 are measured. Alternatively, if thereacting centre undergoes sp2 to sp3 rehybridization at the transition state of the ratedetermining step, then the force constant for bond vibration increases, thus resulting in a39greater difference in zero-point energies at the transition state than that at the ground state.The activation energy for the hydrogen-substituted reaction is now greater than that for thedeuterium-substituted reaction. For these types of reactions, “inverse” kinetic isotopeeffects of kH/kD < 1.0 are measured.A “normal” a-secondary deuterium kinetic isotope effect can have a maximumvalue of kH/kD = 1.40. Generally however, values of kH/kD = 1.10-1.25 are consideredindicative of a significant degree of sp3 to sp2 rehybridisation at the transition state of therate determining step. An a-secondary deuterium kinetic isotope effect of kH/kD = 1.0suggests little or no change in hybridisation at the transition state.Enzymatic reactions are extremely sensitive to modifications in the electronicstructure of the substrates. However, as isotopic substitutions do not alter the electronicnature of the isotopically-labeled compounds, a-secondary deuterium kinetic isotope effectmeasurements can be used as powerful probes for elucidating the mechanism of enzymaticreactions. Unlike other substrate modifications, isotopic substitutions have no effect on thebinding of substrates at the active site of the enzyme. a-Secondary deuterium kineticisotope effects are particularly useful since they provide valuable information pertaining tothe structure of the transition state.2-1-5 Fluorine and hydrogen substitutions as probes of enzymatic reactionsAs previously discussed in Chapter I, non-covalent interactions between thesubstrate and the enzyme are proposed to account for most of the rate enhancement seen forenzymatic reactions. For glycosidases, these interactions are believed to be primarilyhydrogen bonds formed between the hydroxyl groups of the sugar and the residues withinthe active site of the enzyme. Replacement of these hydroxyl groups can therefore provideinsights into the enzymatic mechanism. Fluorine and hydrogen atoms are appropriatesubstitutions for the sugar hydroxyl groups since the bond lengths and van der Waals radiiof both the C-H and C-F bonds are less than those for C-OH, thus the fluorine and40hydrogen atoms do not introduce repulsive interactions within the enzyme active site (Table2-1) (Withers et a!., 1988). However, two major effects, electronic and binding, must beconsidered when investigating the effect of fluorine or hydrogen substitution on anenzymatic reaction.Table 2-1: Size comparison of C-H, C-F and C-OH groups (Withers et al., 1988).Group Bond length (A) Van der Waals Radius (A) Total (A)C-H 1.09 1.20 2.29C-F 1.39 1.35 2.74C-OH 1.43 2.10 3.53(a) Electronic effects: These effects will be important since differences in theelectronegativity of substituents will result in different responses of the oxocarbonium ion-like transition states. The relatively electronegative fluorine atom will destabilise theoxocarbonium ion-like transition state, thus slowing down the reaction. By contrast, therelatively electropositive hydrogen atom is expected to stabilise the oxocarbonium ion-liketransition state, thus speeding up the reaction. Indeed, such effects have been reportedpreviously for non-enzymatic, acid-catalysed hydrolysis of glycosides in which thereactions are known to proceed via oxocarbonium ion transition states (Capon, 1969;BeMiller, 1967; Withers et al., 1986; Withers et al., 1989). On the basis of electroniceffects, the rate of glycoside hydrolysis is expected to change in the order deoxy> hydroxy> deoxyfluoro.(b) Binding effects: These effects will be significant also since deletion of crucialhydrogen bonds will destabilise the transition state, thus slowing down the reaction.Binding effects upon the substitution of fluorine or hydrogen for a hydroxyl group wifi be41manifested in the following way. In a hydrogen bond, the functional group containing theinteracting hydrogen atom is referred to as the “donor” and the atom carrying the lone pairelectrons with which this hydrogen interacts is called the “acceptor” atom (for example,fluorine and oxygen). The hydroxyl group of the sugar can function both as a hydrogen-bond donor and acceptor while the fluorine atom of a deoxyfluoro-glycoside can onlyaccept a hydrogen bond. This hydrogen bond to the fluorine atom is likely weaker than thatto the oxygen of the hydroxyl group since the lone pair of electrons on the fluorine is moretightly held by the nucleus than those on the oxygen. The hydrogen of a deoxy-glycosidecannot participate in a significant hydrogen bond. Based on the hydrogen bonding potentialof fluorine, hydrogen and the hydroxyl group, the rates of glycoside hydrolysis areexpected to increase in the order hydroxy > deoxyfluoro > deoxy.In reality, the binding and electronic effects are superimposed on each other inenzymatic reactions and thus difficult to quantitate individually. However, minimumestimates of the amount of transition state stabilisation contributed by individual hydroxylgroups can be obtained from the comparison of rates with deoxy substrates with those forthe parent compound.2-1-6 pH Dependence of enzymatic reactionsAlthough enzymes contain many ionising groups, the most important ionisationsare those involved in substrate binding and catalysis, as well as those responsible formaintaining the enzyme in an active conformation. When interpreting pH profiles thefollowing assumptions are made (Fersht, 1985). The ionising groups act as perfectlytitrating acids and bases, only one active site ionisation state of the enzyme is capable ofconverting substrate to product, proton transfers of all ionising groups are faster than thecatalytic steps in the reaction, and the rate determining step of the reaction does not changewith pH. Plots of kcat/Km versus pH will yield pKa values for ionising groups within thefree enzyme and/or free substrate while plots of keat versus pH will yield pKa values for the42ionising groups within the enzyme-substrate complex whose decomposition is ratedetermining.2-1-7 Background on kinetic analysis of C. fimi exoglycanasePrevious studies of C.Jlmi exoglycanase (Tull, M. Sc. Thesis, 1991; Tull et al.,1991; Tull & Withers, 1994) revealed that it catalyses the hydrolysis of aryl 13-cellobiosidesto cellobiose and the corresponding substituted phenol (Figure 2-2).CellobioseogXExoglycan:e+Substituted phenolateFigure 2-2 C.fImi exoglycanase-catalysed hydrolysis of wy! J3-cellobiosides.Similarly, this enzyme converts aryl 3-D-glucosides to glucose and the correspondingsubstituted phenol. Values for k, Km and kcat/Km were determined for a series of aryl 3-cellobiosides as well as for a series of aryl 3-glucosides. Values of k/Km for thehydrolysis of the cellobiosides were much higher than those for the correspondingOHHOHOAryl 13-cellobiosideOHOH43glucoside, as might be expected given that the normal mode of cello-oligosaccharidecleavage is one in which cellobiose is released. To probe the relationship between substratestructure and enzymatic activity a linear free energy relationship study was undertaken inwhich values of log(kc) and log(kca/Km) for these two substrate series were plotted asfunctions of the phenol leaving group pKa in the form of Broensted plots (Figure 2-3 and2-4).For the glucosides, both log(kj) and log(kca/Km) were linearly correlated with thephenol pKa indicating that both the rate determining step and first irreversible step(Appendix B-2) in the reaction are the formation of the glucosyl-enzyme intermediate asthis is the step involving cleavage of the phenolate (Figure 2-3).For the cellobiosides, the absence of any significant dependence of log(kcat) UpOnthe phenol leaving group PKa across most of the range suggests that formation of thecellobiosyl-enzyme intermediate is not rate determining (Figure 2-4). Rather, hydrolysis ofthe cellobiosyl-enzyme intermediate is most probably rate limiting in this case. By contrast,the Broensted plot for log(k/Km) shows a modest linear dependence (I3 = -0.3) acrossthe entire pKa range (Figure 2-4). Since kcat/Km reflects the first irreversible step in thereaction, likely initial C-U bond cleavage, then this dependence on leaving group ability isexpected. Interestingly, a slight downward break at the higher pKa values seems to bepresent in the log(ka) plot suggesting a change to rate determining glycosylation.However, this is somewhat obscured due to the relatively weak dependence of log(kca) onthe pKa in this region. Such biphasic Broensted plots are not uncommon with glycosidasessince similar, though more pronounced, behaviour has been seen when probing substratereactivity of 3—glucosidases with aryl -D-glucosides (Dale et a!., 1986; Kempton &Withers, 1992). These results suggest that the rate determining step changes fromdeglycosylation to glycosylation with the poor leaving groups.II0.0-1.044Figure 2-3 Broenstedplots relating rates of C. fimi exoglycanase-catalysed hydrolysisofaryl /3-glucosides with the leaving group ability of the phenols. (Upper)Plot of log(ka,) versus PKa of the aglycone phenols; (lower) plot oflog(k/K) versus PKa of the aglycone phenols.3.02.01.0-2.0-3.03 5 7 9pKa3 5 7 9pKa.•.. •..3 5 7 9pKa452.01.5W1.0•C— O.50.0•2.52.0C—0.0pKaFigure 2-4 Broenstedplots relating rates of C. fimi exoglycanase-catalysed hydrolysisof aryl j3-cellobiosides with the leaving group ability of the phenols.(Upper) Plot of log(kcat) versus pKa of the aglycone phenols; (lower) plotof log(kg/K) versus PKa of the aglycone phenols.3 5 7 946These linear free energy relationship studies provide valuable insights into transitionstate structure. The large value of the Broensted constant (I3ig = -1) for the glucosides,which is similar to those found for several 3-glucosidases using similar glucosides (forexample, 13g = -0.7 for Agrobacterium 3-glucosidase (Kempton & Withers, 1992) andI31g — -1 for sweet almond f-glucosidase (Dale et al., 1986)), reflects a large degree ofnegative charge accumulation on the phenolate oxygen at the glucosylation transition state.This indicates that there is almost complete C-O bond cleavage and relatively little protondonation to the departing phenolate at this transition state. By contrast, the Broenstedconstant for the cellobiosides (I3ig = -0.3) is considerably less than that seen for theglucosides ({3ig = -1), thus reflecting some degree of negative charge build up on thephenolate oxygen at the cellobiosylation transition state, but not as much as that seen at theglucosylation transition state. There are two likely causes for this difference. One could bethat there is less C-O bond cleavage at the cellobiosylation transition state than at that forglucosylation. The other could be that general acid catalysis is more efficient in cellobiosidehydrolysis, resulting in more proton donation than was seen for the glucosides. It is noteasy to distinguish between these two possibilities.The inactivators, 2F-DNPC and 2F-DNPG, are known to inactivate thisexoglycanase by binding to the enzyme and forming covalent glycosyl-enzyme species(Tull et al., 1991; McCarter et al., 1993). The inactivated-enzymes are stable in buffer, butreactivate in the presence of a suitable glycosyl-acceptor (for example, cellobiose),presumably via a transglycosylation reaction. These results indicate that the covalent 2-deoxyfluoroglycosyl-enzymes are catalytically competent as they turn over to product, thussupplying further evidence for the proposed double displacement mechanism. Interestingly,inactivation of the enzyme by 2F-DNPC (k1/K = 6.12 x 10’ min1mM’) is times fasterthan by 2F-DNPG (k/K = 5.56 x i0 min’ mM’) while both inactivated forms of theexoglycanase reactivate at comparable rates (khydrolysjs = 8.5 x l0 min’ and khydrolysjs = 1.3x io min’ for the 2-deoxyfluorocellobiosyl- and 2-deoxyfluoroglucosyl-enzymes,47respectively). Interestingly, a somewhat similar situation is seen with the parentcompounds, 2,4-DNPG and 2,4-DNPC. The kcatfKm value for hydrolysis of 2,4-DNPC is—10 times greater than that for hydrolysis of 2,4-DNPG, while the k values forhydrolysis of both substrates are comparable. This likely indicates that the glycosylationstep is not rate determining for 2,4-DNPG, but rather another step, most likelydeglycosylation. Indeed this would be consistent with the finding that the log(k) value for2,4-DNPG falls slightly below the line defined by the other glucosides in the Broenstedplot. These results suggest that the distal glucosyl unit of the cellobiosides increases the rateof glycosylation relative to that for the glucosides, but not the rate of deglycosylation.2-2 Objectives of This ProjectThe aim of this project is the delineation of the detailed mechanism of CellulomonasjImi exoglycanase by means of several different kinetic techniques. This will provide thefirst detailed kinetic characterisation of a cellulase and will lay the basis for understandingthe consequences of mutations in other studies.2-2-1 Pre-steady state analysis of aryl 13-celiobiosides and aryl -glucosides.Stopped-flow kinetic analysis of the aryl 13-ceilobiosides may provide furtherevidence in support of the identity of the rate determining step previously deduced from thelinear free energy relationship study. This analysis should also provide values for the rateconstants for the formation of the cellobiosyl-enzyme intermediate (k2) as well as values forthe dissociation constant (KJ) of the E.S complex. Stopped-flow analysis of the aryl 3—glucosides, 2,4-DNPG and PNPG, may provide evidence concerning the “true” identity ofthe rate-determining step for 2,4-DNPG.482-2-2 Linear free energy relationships of aryl -g1ycosides(a) Cellobiosides: The presence of a linear free energy correlation between values ofk2 for the formation of the cellobiosyl-enzyme intermediate (derived from stopped-flowanalysis) and the pKa of the phenol leaving group may provide insights into the degree ofnegative charge development at the glycosylation transition state for cellobiosidehydrolysis.(b) Xylosides and xylobiosides. Determination of the Michaelis-Menten kineticparameters for C. fimi exoglycanase-catalysed hydrolysis of several aryl -xylosides andaryl 3-xylobiosides (Figure 2-5) may reveal further relationships between substratestructure and enzymatic activity. Furthermore, since gluco-sugars differ from xylo-sugarsby the presence of the C-5 hydroxymethyl group then a comparison of the results forxyloside and xylobioside hydrolysis with those obtained previously for the hydrolysis ofglucosides and cellobiosides may reveal any mechanistic function of the C-5hydroxymethyl group.xriio10HHOHFigure 2-5 Structures of aryl J3-xylobiosides and aryl J3-xylosides.492-2-3 a-Secondary deuterium kinetic isotope effects on aryl -glycosides.Values of kHIkD for C. fimi exoglycanase-catalysed hydrolysis of aryl -cellobiosides and aryl (3-glucosides are expected to yield information on the degree ofoxocarbonium ion character at the transition states, thus providing further insights into theidentity of rate determining steps, the degree of bond cleavage at the transition states, aswell as some indication of the extent of nucleophilic preassociation at the transition states.2-2-4 Inactivation of C. fimi exoglycanase by 2-deoxy- and 2-deoxy-2-flu o ro- sugars.(a) 2F-DNPX:Inactivation-reactivation studies with 2,4-dinitrophenyl 2-deoxy-2-fluoro-(3-xylobioside (Figure 2-6) can provide substantial evidence for the formation of acovalent glycosyl-enzyme intermediate during xylobioside hydrolysis by this enzyme.2F-DNPXFigure 2-6 Structure of 2 ,4-dinitrophenyl 2 -deoxy-2 -fluoro-f3-xylobioside (2F-DNPX2).(b) 2F-DNPC and 2F-GMF: 19F-NMR analysis of C.fimi exoglycanase inactivatedwith 2,4-dinitrophenyl 2-deoxy-2-fluoro-[3-cellobioside and 2-deoxy-2-fluoro 4-O-((3-glucosyl)-[3-mannosyl fluoride (Figure 2-7) can yield information pertaining to theanomeric stereochemistry of the glycosyl-enzyme intermediate.H0OHF NO250(c) Cellobial: As discussed in Chapter I, glycosidases are capable of catalysing thehydration of glycals to 2-deoxy-glycose products. Since glycals lack the hydroxyl group atthe C-2 position then investigation of C. fimi exoglycanase using cellobial (Figure 2-8) incombination with results from previous studies with 2,4-dinitrophenyl 2-deoxy-2-fluoro--cellobioside and 2,4-dinitrophenyl [-cellobioside can provide information about themechanistic role of the C-2 hydroxyl group during cellobioside hydrolysis.OH2F-DNPC2F-GMFFFigure 2-7 Structures of2,4-dinitrophenyl 2 -deoxy-2 -fluoro-cellobioside (2F-DNPC)and 2 -deoxy-2 -fluoro 4-O-(J3-glucosyl)-J3-mannosylfluoride (2F-GMF).H0OH0F NO2HOOHOHFOHHOOHOHOH0Figure 2-8 Structure of cellobial.512-2-5 pH StudyThe pH dependence of the Michaelis-Menten parameters for cellobioside hydrolysismay provide insights into the ionisation state of active site residues involved in substratebinding and/or enzyme catalysis.2-3 Results2-3-1 Steady state kinetics for xylo-substratesC. JImi exoglycanase catalyses the hydrolysis of aryl 3-D-xylosides to thecorresponding substituted phenol and xylose as well as catalysing the hydrolysis of aryl 3-xylobiosides to the corresponding substituted phenol and xylobiose. Values for V(maximum velocity) and Km were determined by fitting the initial rates of hydrolysis andthe substrate concentrations to the Michaelis-Menten equation using the program GraFit(Leatherbarrow, 1990). These results are illustrated as Lineweaver-Burk plots (AppendixA-i and A-2) for visual convenience. The Michaelis-Menten parameters for the xylosides(determined in collaboration with Ms. I. Setyawati) and xylobiosides, along with the pKvalues for the phenol leaving groups, are listed in Tables 2-2 and 2-3. Values of log(kca)and log(kca/Km) for the xylosides are plotted as functions of the aglycone pKa in the formof Broensted plots (Figure 2-9) and in both cases a strong dependence upon the PKa isobserved with Broensted constant, 13g = -0.8 (correlation coefficient = 0.99) and -0.9(correlation coefficient = 0.98) respectively. Equivalent plots of log(kca) and log(kca/Km)versus leaving group pKa for the xylobiosides are shown in Figure 2-10. Values oflog(k) appear invariant with P’<a over the narrow range of pKa of 5 -7. The log(kca/Km)Broensted plot appears scattered, however there are only four points thus interpretation isnot truly warranted.52Table 2-2: Michaelis-Menten Parameters for the Hydrolysis of Aryl f-Xylosides by C.JlmiExoglycanasePhenol substituent aPK Km (mM) kcat (s1) kcat/Km (mM1s)2,3-dinitro 4.89 5.82 ± 0.62 70±4 122,5-dinitro 5.15 5.85 ±0.89 108±7 183,4-dinitro 5.36 7.90± 2.32 22±4 2.84-nitro 7.18 20.0 ± 2.5 2.6 ± 0.2 0.134-cyano 8.49 11.4 ± 1.0 0.098 ± 0.004 0.0089aphenol PKa values were taken from Barlin & Perrin (1966), Kortum et al. (1961),Robinson et al., (1960), and Ba-Saif & Williams (1988).Table 2-3: Michaelis-Menten Parameters for the Hydrolysis of Aryl -Xylobiosides by C.fimi ExoglycanasePhenol substituent aK Km (mM) kcat (s1) kcatfKm (mM1s)2,5-dinitro 5.15 0.019±0.004 27±2 14433,4-dinitro 5.36 0.012±0.001 22±1 18404-nitro 7.18 0.015 ± 0.001 66±2 43862-nitro 7.22 0.078 ± 0.008 56±2 714aphenol PKa values were taken from Barlin & Perrin (1966), Kortum et al. (1961),Robinson et al., (1960), and Ba-Saif & Williams (1988).53Figure 2-9 Broensted plots relating rates of C. fimi exoglycanase-catalysed hydrolysisof aryl J3-xylosides with the leaving group ability of the phenols. (Upper)Plot of log(k) versus PKa of the aglycone phenols; (lower) plot oflog(kc1/Km) versus PKa of the aglycone phenols.3210-14 5 6 7 8 9pKaI4 5 6 7 8 9pKa543.02.52.0I.,C1.00.50.0 • I • I • • I5.0 5.5 6.0 6.5 7.0 7.5pKa4.54.08pKaFigure 2-10 Broensted plots relating rates of C.fimi exoglycanase-catalysed hydrolysisof aiyl J3-xylobiosides with leaving group ability of the phenols. (Upper)Plot of log(k) versus PKa of the aglycone phenols; (lower) plot oflog(k/K) versus PKa of the aglycone phenols.552-3-2 Pre-steady state kinetic analysis of the hydrolysis of gluco-substratesFive aryl f-cellobiosides were subjected to pre-steady state analysis using astopped-flow technique. Reaction rates were measured at five different substrateconcentrations ranging from 0.2 x Kd to K3 and fitted to an equation describing a first orderreaction followed by a steady state to yield values for the pseudo-first order rate constant(k0b). Values for k2 and Kd were determined from these k0b values and substrateconcentrations by direct fit to the Michaelis-Menten equation using the program GraFit(Leatherbarrow, 1990). These results are plotted as Lineweaver-Burk plots for visualconvenience (Appendix A-3). The kinetic parameters for these five cellobiosides arepresented in Table 2-4, along with the pKa values of the phenol leaving group as well as inthe form of plots of log(k2) and log(k2fK) versus aglycone pKa (Figure 2-11). Goodcorrelations are observed in both the log(k2) (correlation coefficient 0.93) and thelog(k2/K) (correlation coefficient = 0.91) plots, with slopes corresponding to Broenstedconstant, I1g = -0.3.Table 2-4: Pre-Steady State Parameters for Hydrolysis of Aryl -Cel1obiosides by C. fimiExoglycanase.Phenol substituent apKa k2 (s1) K(J (mM4) k2/Kj(s1mM)2,4-dinitro 3.96 1660 ± 236 10.0 ± 2.4 1662,5-dinitro 5.15 1160 ± 94 16.2 ± 2.0 723,4-dinitro 5.36 626±42 5.3 ± 0.7 1184-nitro 7.18 244 ± 67 12.4 ± 4.9 202-nitro 7.22 bND bj 13aphenol pKa values were taken from Barlin & Perrin (1966), Kortum et al. (1961),Robinson et al., (1960), and Ba-Saif & Williams (1988).bj-i= not determined560—0—3 4 5 6 7 8pKa3.43.02.62.22.62.21.81.41.0pKaFigure 2-11 Broensted plots relating pre-steady state rates of C.fimi exoglycanasecatalysed hydrolysis of aryl J3-cellobioside with leaving group ability ofphenols. (Upper) Plot of log(k2)versus PKa of the aglycone phenols;(lower) plot of log(k2/K) versus PKa of the aglycone phenols.3 4 5 6 7 857Two aryl f3-D-glucosides, 2,4-DNPG and PNPG, were also subjected to presteady state analysis. Each glucoside was investigated at two different substrateconcentrations. Like the cellobiosides, the exoglycanase concentration was selected toproduce a burst of phenolate with a total absorbance change of 0.06 A. A burst of 2,4-dinitrophenolate of 0.05 A was observed for 2,4-DNPG while no burst could be detectedfor PNPG.2-3-3 a-Secondary deuterium kinetic isotope effect measurements ongluco-substrates.x-Secondary deuterium kinetic isotope effects were measured for three aryl f3-cellobiosides and two aryl 3-D-glucosides and are listed in Table 2-5 along with the ratedetermining step for each substrate.Table 2-5: Secondary Deuterium Kinetic Isotope Effects Measured with C.flmiExoglycanaseSubstrate aRDS kcatHlkcatD2’,4’-dinitrophenyl glucoside DEGLY 1.12 ± 0.024’-nitrophenyl glucoside GLY 1.12 ± 0.022”,4”-dinitrophenyl cellobioside DEGLY 1.10 ± 0.024”-nitrophenyl cellobioside DEGLY 1.10 + 0.024”-bromophenyl cellobioside GLY 1.06 ± 0.02a)5rate determining step.582-3-4 Inactivation of C. fimi exoglycanaseInactivation kinetic parameters for cellobial inactivation of C. fimi exoglycanasewere determined by first calculating values of kobs by direct fit of the data to a first-orderfunction using the program GraFit (Leatherbarrow, 1990) (Appendix A-8). Values of K —500 mM (binding constant) and k1 2 min’ (inactivation rate constant) were detemiined bynon-linear regression analysis using the program GraFit (Leatherbarrow, 1990). However,these values must be taken only as estimates since the insolubility of cellobial preventedstudy of C. fimi exoglycanase inactivation at cellobial concentrations greater than 60 mM.A more reliable value ofk1/K = 2.9 x i0 ± 0.0001 mm4 mM’ was determined from theslope of the plot of k0b values versus cellobial concentration (Appendix A-8). Similaranalysis of 2F-GMF inactivation of C. fimi exoglycanase yielded kinetic parameters of K1160 mM, k1 — 0.069 min’ from the non-linear regression analysis while a more reliablevalue of k/K, = 4.4 x iO ± 0.00004 mM’ min1 was obtained from the slope of the kObSversus 2F-GMF concentration plot.2-3-5 Reactivation of inactivated-C. fimi exoglycanaseSamples of cellobial-inactivated and 2F-DNPX-inactivated C. fimi exoglycanasewere freed of excess inactivator by extensive dialysis, incubated in the presence a suitableglycosyl acceptor, and assayed for the return of enzymatic activity. Reactivation rates werethen calculated from the slopes of plots of ln(full rate - observed rate) versus time or directfit to a first order equation using the program GraFit (Leatherbarrow, 1990) (Appendix A-8and A-9) and these are listed in Table 2-6.59Table 2-6: Rates of Reactivation of Inactivated C.fimi Exoglycanase in the Presence of aGlycosyl-AcceptorGlycosyl-enzyme species kreact (buffer) (h1) kreaa(cellobiose) (h’)2-deoxycellobiosyl- 3.1 x 102± 0.003 7.9 x 102± 0.0082-deoxyfluoroxylobiosyl- 7.0 x i04± 0.0001 1.8 x i0 ± 0.00022-3-6 19F-NMR analysis of inactivated-C. fimi exoglycanaseSamples of C.fimi exoglycanase were inactivated with either 2F-DNPC or 2F-GMF and subjected to 19F-NMR analysis. The 19F-NMR spectrum of the 2FCb-exoglycanase sample is shown in Figure 2-12, the broad peak at -195.5 ppmcorresponding to 2FCb-exoglycanase. The signal at ö = -198.4 ppm corresponds to the C-2 fluorine of unreacted 2F-DNPC while that at = -198.1 ppm corresponds to the C-2fluorine of either the hydrolysis product or a transglycosylation product. The 19F-NMRspectrum of the 2FGM-exoglycanase sample (Figure 2-13) reveals resonances at = -149.9 and -224.1 ppm corresponding to the C-i and C-2 fluorines, respectively, ofunreacted 2F-GMF while that at = -121.4 corresponds to released fluoride. Theresonance at -205.4 ppm likely corresponds to the hydrolysis product, 2-deoxy-2-fluoro 4-O-(I3-glucosyl)-c-mannose. The peak for the corresponding f3-anomer would be expectedto be very small since very little of the 3-anomer is present at equilibrium and is likelyburied under the peak at -224.1 ppm. The resonance at = -205.6 ppm in the dialysedsample corresponds to the C-2 fluorine of the 2FGM-exoglycanase. For the 2FGM-enzymesample, since a high concentration of 2F-GMF (30 mM, K — 160 mM) was required toinactivate the enzyme and since a low concentration of enzyme (9 mg/ml, 0.2 mM) wasused, the signal for this 2FGM-exoglycanase is seen to be much smaller than those for theexcess, unreacted 2F-GMF. For the 2FCb-enzyme sample, since the concentration of the60enzyme used was 0.2 mM and that of 2F-DNPC was --1 mM, the signals for excess,unreacted 2F-DNPC and 2FCb-exoglycanase are seen to be comparable.2-3-7 pH Study of C. fimi exoglycanaseThe pH dependence of C. fimi exoglycanase was investigated using 2,4-DNPC assubstrate over a pH range of 4.5 to 9.4. Values for and K were determined and arepresented as plots of kcat versus pH in Figure 2-14. Values of are seen to beindependent of pH over this range, while kcat/Km is seen to be dependent upon twoionisations of pKa = 4.1 ± 0.1 and 7.7 ± 0.1. The higher pKa value is a reliable value, butinstability of the enzyme at low pH values precluded accurate determination of the lowerionisation constant. This value must therefore be taken simpiy as an estimate.61.195.5.198.198.4-195.0 •196.O .197.0 .198.0 .199.0ppmFigure 2-12 ‘9F-NMR spectrum of C.fImi exoglycanase inactivated by 2F-DNPC.[Enzyme] = 0.2 mM and [2F-DNPC] = 1.0 mM in 50 mM sodiumphosphate buffer, 10% D20.-121.462J49.9_J41-120 -140-160 .180 -200ppm.2049 •2.2 .205.5 405.8 4061 406.4ppmFigure 2-13 ‘9F-NMR spectra of C.flmi exoglycanase inactivated by 2F-GMF.[Enzyme] = 02 mM and [2F-GMF] =30 mM in 50 mM sodium phosphatebuffer, 10% D20. (Upper)‘9F-NMR spectrumfor the undialysed sampleand (lower)‘9F-NMR spectrumfor the undialysed sample overlapped withthatfor the dialysed sample over the same ppm range.-220.205.4E631614‘ 108620011•1•1 1• • • • ••••I • I • I4 5 6 7 8 9pH1501005004 5 6 7 8 9pHFigure 2-14 pH Dependence of the hydrolysis of2,4-DNPC by native C. fimiexoglycanase.642-4 Discussion2-4-1 pH Dependence of C. fimi exoglycanaseThe substrate employed for these studies was 2,4-DNPC, for which the ratedetermining step is hydrolysis of the cellobiosyl-enzyme intemiediate (vide infra). Valuesof kfK axe seen to be dependent upon ionisations of approximate PKa = 4.1 in the acidiclimb of the pH profile and PKa = 7.7 in the basic limb (Figure 2-14). This suggest that onlyone protonated state of the free enzyme is catalytically active. These pKa values are similarto those seen for free Agrobacterium 3-glucosidase (PKa < 5 d PKa = 7.0 - 7.2)(Kempton & Withers, 1992). Interestingly, binding of the substrate to the exoglycanaseshifts the pKa values of these groups considerably since no pH dependence is seen for theenzyme/substrate complex, as indicated by the pH independent kcat plot. Since the group inthe free enzyme of PKa = 7.7 must be protonated to be catalytically active then thisionisation likely reflects that of the general acid catalyst (Glu 127) which is required toprotonate the departing aglycone in the glycosylation step. The group in the free enzyme ofapproximate pKa = 4.1 which must be deprotonated to be catalytically active likelycorresponds to the catalytic nucleophile (Glu 233). Alternatively, this ionisation couldcorrespond to a group involved in accepting a proton from the catalytic nucleophile.2-4-2 f3-Glucanase activity of C. fimi exoglycanaseThe large bursts of released, substituted phenolates observed in the stopped-flowanalysis of the hydrolysis of the aryl 13-cellobiosides indicate that for these substrates a stepsubsequent to aglycone release, most likely hydrolysis of the cellobiosyl-enzymeintermediate, is the rate determining step. These results are entirely consistent with thosefrom a previous linear free energy relationship study of aryl 3-cellobiosides (Tull &Withers, 1994) which revealed that values of log(k) were essentially invariant with the65PKa of the phenol leaving group over the PKa range 4-8, thus suggesting thatdeglycosylation was most probably the rate determining step.Values of log(k2)and log(kJK) are seen to be dependent on the P’a of the phenolleaving group with a Broensted constant of fig = -0.3 in both cases (Figure 2-11). Theseresults suggest that both k2 and kz/Kd reflect cleavage of the bond to the phenolateaglycone, thus formation of the cellobiosyl-enzyme. The slopes of plots of log(k2) andlog(kc/Km) (Tull & Withers, 1994) versus pKa have exactly the same Broensted constant(I3ig = -0.3) as would be expected if both k2 and kcat/Km reflect the same step in thereaction. The stopped-flow results therefore provide further evidence in support of theproposal that the glycosylation step is the first irreversible step during cellobiosidehydrolysis. In the previous linear free energy relationship study, the log(k) versus pKaplot (Figure 2-4) revealed an apparent slight downward break at the higher pKa values,possibly indicating a change in the rate determining step from deglycosylation toglycosylation (Tull & Withers, 1994). The stopped-flow analysis provides further insightsinto this phenomenon. Since the value of the Broensted constant for the log(k2) plot is thesame as the approximate value of the Broensted constant seen for this somewhat obscuredpKadependent region (PKa —7 - 9) of the log(k) plot, this suggests that this region of thelog(k) likely reflects formation of the cellobiosyl-enzyme. Thus the stopped-flow results,in combination with the log(k) versus pKa plot from the linear free energy relationshipstudy, suggest that glycosylation is the rate determining step for aryl 3-cellobiosides withpoor leaving groups (phenol pKa > 7.5).The value of the Broensted constant (31g = -0.3) seen for both the log(k) andlog(k2IK) plots (Figure 2-11) reflects relatively little negative charge accumulation on thedeparting phenolate oxygen at the glycosylation transition state for hydrolysis of thesecellobiosides. As discussed in section 2-1-7, this Broensted constant (j3ig = 0.3) isconsiderably smaller than that previously seen for a series of aryl 3-D-glucosides(Iig 1).66The identical cc—secondary deuterium kinetic isotope effects (kH/kD = 1.10)measured for the more reactive cellobiosides, 2,4-DNPC and PNPC, indicate substantialoxocarbonium ion character at the deglycosylation transition state (Table 2-5). This value iscomparable to those reported for this step with other “retaining” -glycosidases (kHIkD =1.11 for Agrobacterium 3-glucosidase (Kempton & Withers, 1992) and kHJkD = 1.2-1.25for E. coli (lac z) f3-galactosidase (Sinnott, 1978)) thus providing further evidence for theinvolvement of a covalent glycosyl-enzyme intermediate during glycoside hydrolysis sincesuch secondary deuterium kinetic isotope effects could only be seen if the intermediate hadmore sp3 character than the subsequent transition state. This value of the kinetic isotopeeffect suggests that there is relatively little preassociation of the nucleophile (water) at thedeglycosylation transition state. By contrast, the smaller isotope effect (kH/kD = 1.06)measured for the cellobiosylation step using 4-BrPC indicates a transition state for thissubstrate with lesser oxocarbonium ion character. This value falls well within the range(kH/kD = 1.0 for E. coli (lac z) 3-galactosidase (Sinnott, 1978) and kH/kD = 1.10 forAgrobacterium 3-glucosidase (Kempton & Withers, 1992)) of isotope effects measured onthe glycosylation transition state with other “retaining” glycosidases. This small kineticisotope effect, taken in combination with the small Broensted constant (f3ig = -0.3) forcellobioside hydrolysis, is consistent with a relatively early cellobiosylation transition statewith either little bond cleavage or considerable protonic assistance coupled with substantialnucleophilic pre-association.2-4-3 f-GIucosidase activity of C. fimi exoglycanaseThe absence of a pre-steady state burst of 4-nitrophenolate for PNPG, underconditions where a full burst of 4-nitrophenolate was observed for PNPC, indicates thathydrolysis of the glucosyl-enzyme intermediate is not rate determining for PNPG. Instead,it is likely that formation of the glucosyl-enzyme intermediate is the rate determining stepfor this substrate. These stopped-flow results are entirely consistent with those from a67previous linear free energy relationship study which revealed that values of log(k) werelinearly dependent on the pKa of the phenol leaving group over the range pKa 4-8 and thussuggested that glucosylation is the rate determining step for these glucoside substrates. Bycontrast, a burst of 2,4 dinitrophenolate comparable in magnitude to that measured for 2,4-DNPC was observed for 2,4-DNPG. Like the cellobioside substrates, this result indicatesthat hydrolysis of the glucosyl-enzyme intermediate is most likely the rate determining stepfor 2,4-DNPG. This result is consistent with that from the linear free energy relationshipstudy which revealed that the log(kca) value for 2,4-DNPG falls below the line defmed bythe other substrates in the log(kca) Broensted plot (Figure 2-3), as would be expected if therate determining step for this glucoside was indeed deglycosylation. The deglycosylationrates for 2,4-DNPC and 2,4-DNPG are similar, indicating that, in strict contrast to thesituation with the glycosylation step, the presence of the distal glucose moiety does notassist the deglycosylation process. In fact, exactly the same situation has been reportedpreviously for the hydrolysis of 2FCb- and 2FGlu- (2-deoxyfluoroglucosyl) enzyme(discussed in Section 2-1-7, (Tull & Withers, 1994)).A possible reaction coordinate diagram which describes this situation is shown inFigure 2-15, the energy levels are arbitrarily chosen. As a first approximation, it isassumed that the transition states for glycosylation and deglycosylation are similar inenergy as expected for an evolved enzyme system and as shown by changes in the ratedetermining step with leaving group ability. A second assumption is that of uniformbinding (Albery & Knowles, 1976), in that the distal glucosyl unit stabilises both transitionstates and the glycosyl-enzyme intermediate to a similar extent. As a consequence of this,the activation energy required for hydrolysis of the glucosyl-enzyme is seen to becomparable to that for hydrolysis of the cellobiosyl-enzyme, and thus is consistent with thesimilar rates of deglycosylation observed with glucoside and cellobioside substrates.Furthermore, since values of Km for the glucosides having glycosylation as the ratedetermining step, thus Km — Kd, are comparable to Kd values for the cellobiosides, obtained68from stopped-flow analysis (K (PNPG) = 8.3 mM and Kd (PNPC) = 12.4 mM), it isassumed that the ground state binding energy is similar for the glucosides andcellobiosides. As a result of this, the activation energy for formation of the glucosylenzyme is seen to be greater than that for formation of the cellobiosyl-enzyme, thus theglycosylation rates for the glucosides are lower than those for the cellobiosides.GE+S -E.S ES EP JEPtES EPtFigure 2-15Reaction coordinateReaction coordinate diagram illustrating the stabilisation produced by thedistal glucosyl moiety of the cellobiosides (dashed line = glucosides, solidline = cellobiosides)The secondary deuterium kinetic isotope effect measured for PNPG (kHIICD = 1.12)reflects substantial sp3 to sp2 rehybridization, thus a large degree of oxocarbonium iontE+SESESEPE+P69character at the glucosylation transition state. This necessarily implies considerable bondcleavage and little preassociation of the enzymatic carboxylate nucleophile. The value ofthis isotope effect is slightly larger than those reported for the glucosylation transition statesin several other f3-glucosidases, values tending to range from 1.0 for E. coli (lac z) -galactosiclase (Sinnott, 1978) to 1.10 for Agrobacterium 3-glucosidase (Kempton &Withers, 1992). However it should be noted that interpretations are clouded somewhat bythe known dependence of the kinetic isotope value on the nature of the leaving groupinvolved in each case. For example, the kinetic isotope effect measured on theglucosylation transition state for Agrobacterium 3-glucosidase was kHIkD 1.05 usingMNPG while a value of kH/kD = 1.10 was measured using PNPG (Kempton & Withers,1992). This substantial degree of oxocarbonium ion character implied by the kinetic isotopeeffect measured for PNPG with C. fimi exoglycanase is however consistent with therelatively large Broensted constant(13g = -1) obtained from the previous linear free energyrelationship study of aryl f3—glucosides which indicated a large amount of negative chargedevelopment on the phenolate oxygen at the glucosylation transition state. The kineticisotope effect, taken in combination with the Broensted constant, reveals a glucosylationtransition state for PNPG that is relatively late, with almost complete C-O bond cleavage,little protonic assistance, and relatively little pre-association of the carboxylate nucleophile.By contrast, the smaller kinetic isotope effect and lesser negative charge development at thetransition state for formation of the cellobiosyl-enzyme suggest that this transition state isrelatively early. This likely arises from the increased protonic assistance, possibly coupledwith a greater degree of nucleophile preassociation by the catalytic nucleophile. The energyinput required for formation of this more highly organised transition state is likely derivedfrom the additional binding interactions with the distal glucosyl moiety.The kinetic isotope effect for 2,4-DNPG (kH/kD = 1.12) is also large, but falls wellwithin the range measured for other glycosidases on the deglycosylation step (kH/kD =1.09 for Botrydiplodia theobromae 3-glucosidase (Umezerika, 1988) to kH/kD = 1.2-1.2570for E. coli (lac z) 3-galactosidase (Sinnott, 1978)). Like the transition state for hydrolysisof the cellobiosyl-enzyme, the value of the isotope effect for 2,4-DNPG reflects a transitionstate for hydrolysis of the glucosyl-enzyme that involves substantial development ofpositive charge at the anomeric centre, extensive bond cleavage, and relatively littlepreassociation of the water.2-4-4 Xylanase activity of C. fimi exoglycanaseInitial substrate specificity studies showed that C.fimi exoglycanase was capable ofcatalysing the hydrolysis not only of cellulose but also xylan, with comparable efficiency(Gilkes et al., 1984). A reinvestigation using aryl f-xylobiosides revealed that kca/Kmvalues for xylobioside hydrolysis are in fact 30 -100 times higher than those for equivalentcellobiosides. This xylanase activity is not entirely surprising as C. fimi exoglycanase hasbeen classified, on the basis of sequence similarities of catalytic domains, as a member ofthe F family of f-g1ycanases which is made up largely of xylanases (Gillces et al., 1991;Henrissat & Bairoch, 1993). In addition, both C. JImi exoglycanase and xylanase Z fromClostridium therinocellum (whose cellulase activity is < 0.5% of the xylanase activity(Grepinet et al., 1988)) have been shown to be “retaining” glycosidases, thus implying thatthey share a common catalytic mechanism. Since cellobiosides differ from xylobiosides bythe presence of the C-5 hydroxymethyl group, it is clear that this group cannot play anessential role in catalysis in the way shown for other glycosidases (Sinnott, 1987; Kempton& Withers, 1992). Instead the presence of this group rather seems to be slightly inhibitory.The X-ray crystal structure of the free exoglycanase and that of the enzyme crystallisedwith a 2FCb moiety covalently bond at the active site provide insights into this situation(White et al., 1995). Comparison of these structures reveals that upon binding the 2FCbmoiety, Gln 87 is moved out of its original position and Trp 281 is rotated about its C-Cbond to accommodate the C-5 hydroxymethyl groups of the distal and proximal glucosylunits, respectively. These results suggest that the presence of the C-5 hydroxymethyl71groups introduce steric interactions within the active site. This relief of steric strain is notachieved without a “cost”, realised in less binding energy available for catalysis, thus lowerrates of hydrolysis are observed for the cellobiosides relative to those for the xylobiosides.Presumably, any favourable interactions (hydrogen bonds) present at 0-6 whencellobiosides are bound to the exoglycanase will be satisfied by bound water whenxylobiosides are present, as seen for the L-arabinose/D-galactose binding protein (Quiochoet al., 1989). Other cellulases that have xylanase activity include endoglucanases E and Hfrom Clostridium thermocelluin, however, in those cases the cellulase activity isconsiderably greater than that of the xylanase (Hall et aL, 1988; Yague et aL, 1990).Although only four xylobiosides were studied, the Broensted piots can still providevaluable insights into the rate limiting steps. Values of log(k) for hydrolysis of these aryl-xylobiosides are seen to be independent of the phenol pKa over the narrow range of PKa5-7 (Figure 2-10), indicating that hydrolysis of the xylobiosyl-enzyme is likely the ratedetermining step for these substrates. Since deglycosylation is also the rate determiningstep for the corresponding cellobiosides, these results suggest that the absence of the C-5hydroxymethyl group does not affect the relative rates of glycosylation anddeglycosylation. The log(ka/Km) versus pKa plot is scattered and, as there are only fourpoints, it is not possible to derive any meaningful information.2-4-5 Xylosidase activity of C. fimi exoglycanaseComparison of kcat/Km values for hydrolysis of xylosides (Table 2-2) andxylobiosides (Table 2-3) reveals that the exoglycanase catalyses the hydrolysis of thexylosides times less efficiently than the corresponding xylobiosides. These results areconsistent with previous findings that glucosides are hydrolysed less efficiently thancorresponding cellobiosides and with the fact that hydrolysis of oligosaccharide substratesresults in the formation of disaccharides, not monosaccharides. Values of kcat/Km forxyloside hydrolysis are comparable to those for the corresponding glucosides (Tull &72Withers, 1994) while values of k/Km for xylobiosides are 30 - 100 times higher thanthose for the corresponding cellobiosides, suggesting that it is the distal sugar whose C-5hydroxymethyl group is inhibitory.The strong correlations of both log(k) and 1og(k/Km) with the leaving grouppKa for the xyloside substrates (Figure 2-9) indicate that both the rate determining step andthe first irreversible step in catalysis are the formation of the xylosyl-enzyme. Comparisonof these results with those for the glucosides, xylobiosides, and cellobiosides indicates thatin general the rate determining step for C. Jlmi exoglycanase catalysed hydrolysis of themonosaccharide substrates is different from that for the corresponding disaccharidesubstrates. The large value of the Broensted constant (lig = -0.8) for the log(kat) plotreflects a large degree of negative charge build-up on the phenolate oxygen at the transitionstate for formation of the xylosyl-enzyme. This result is entirely consistent with the valueof the Broensted constant ([3ig = -0.9) for the log(k/K) plot. The similarity of theBroensted constant for hydrolysis of both the xyloside and the glucosides (I3ig = l) (TUII& Withers, 1994) indicates that at the transition states for formation of the xylosyl-enzymeand the glucosyl-enzyme, C-O bond cleavage of the phenolate has proceeded to a similarextent and that the degree of proton donation is similar.2-4-6 Inactivation-reactivation studies of C. fimi exoglycanase with 2FDNPX2As discussed in Chapter I, 2-deoxy-2-fluoro-3-D-glycosides have been used toinactivate several “retaining” glycosidases through trapping of covalent glycosyl-enzymeintermediates. The presence of the electronegative fluorine at C-2 destabilises theoxocarbonium ion-like transition states, thus slowing down both the glycosylation anddeglycosylation steps while the good leaving group (for example 2,4-dinitrophenolate)accelerates glycosylation resulting in accumulation of the glycosyl-enzyme. Previousstudies revealed that C. fimi exoglycanase is inactivated in the presence of 2F-DNPX73(Ziser et al., 1995 ). Formation of the 2-deoxyfluoroxylobiosyl-enzyme (2FX-enzyme)(k/K = 16 min1 mM1) is 26-fold more efficient than formation of the 2FCb-enzyme(k/K = 0.61 min’ mM1)(McCarter et a!., 1993; Tull & Withers, 1994) as indicated bythe relative kjJKj values. This rate ratio is comparable to that seen for the relative k’Kmvalues for the hydrolysis of xylobiosides and the corresponding cellobiosides and thus isconsistent with the view that the presence of the C-5 hydroxymethyl group is slightlyinhibitory to catalysis.Removal of excess 2F-DNPX by extensive dialysis does not immediately restoreenzymatic activity, consistent with the formation of a covalently bonded xylobiosyl-enzymespecies. Conversion of the 2FX-enzyme to free active enzyme, that is reactivation, canoccur by two separate routes; either by a hydrolysis reaction or by transglycosylation of thexylobiosyl moiety to a suitable acceptor such as cellobiose. These two routes have beendemonstrated previously for Agrobacterium -glucosidase (Withers & Street, 1988).Transglycosylation in this case is presumed to involve nucleophilic attack by the C-4hydroxyl group of cellobiose at the anomeric centre of the 2-fluoroxylobiosyl-enzyme.Reactivation of the inactivated-exoglycanase in this manner is consistent with the formationof a xylobiosyl-enzyme intermediate and its ability to be converted to product and freeenzyme.Reactivation of the 2-deoxyfluoroglycosyl-enzyme via the hydrolysis reactionprovides values for the deglycosylation rates of these 2-deoxyfluoroglycosyl-enzymespecies. Comparison of these deglycosylation rates with those for the parent glycosides canprovide insights into the deglycosylation transition state for xylobioside hydrolysis. Thedeglycosylation rate for hydrolysis of the parent xylobiosides is 30 - 100 times higher thanthat for the equivalent cellobiosides while the deglycosylation rate for hydrolysis of the2FX-enzyme = 7.0 x 10 h’) and the 2FCb-enzyme (kCL = 5.1 x i0 h’) (Tull &Withers, 1994) are comparable. These results suggest that the presence of the fluorine atomat the C-2 position of the sugar has a greater destabilising effect on the transition state for74hydrolysis of the 2FX-enzyme than on that for hydrolysis of the 2FCb-enzyme. Theseresults likely reflect the fact that highly positively charged transition states will be moresensitive to fluorine substitution than ones with less positive charge (discussed in section 2-1-5). This suggests that the degree of oxocarbonium ion character at the transition state forhydrolysis of the xylobiosyl-enzyme is greater than that at the transition state for hydrolysisof the cellobiosyl-enzyme.2-4-7 Characterisation of the glycosyl-C. fimi exoglycanase intermediate(a) Covalent nature: The downward break in the log(k) versus pKa Broenstedplot and the bursts of released phenolate seen in the stopped-flow study are consistent withthe formation of a glycosyl-enzyme intermediate during C. fimi exoglycanase-catalysedhydrolysis of the aryl 13-cellobiosides. Initial evidence for the covalent nature of thisintermediate derives from studies of the inactivation of this exoglycanase by 2F-DNPC(McCarter et al., 1993; Withers et al., 1993). The catalytic competence of this 2FCb-enzyme intermediate has been demonstrated previously by its conversion to product andactive enzyme via either a hydrolysis reaction or a transglycosylation reaction (McCarter etal., 1993; Tull & Withers, 1994). More recently, mass spectrometric analysis of thisinactivated-C. fimi exoglycanase provided direct evidence for the formation of a 2FCb-enzyme as well as identified the modified active site amino acid as a glutamate residue, thusan linkage is formed between the anomeric carbon of the sugar and the active siteresidue of C.fimi exoglycanase (discussed in detail in Chapter IV).(b) Stereochemistry: The stereochemistry of this ester linkage between the sugarand the exoglycanase was probed with the inactivators, 2F-DNPC and 2F-GMF, using19FNMR. The 19F-NMR spectrum of a sample of 2FCb-exoglycanase (Figure 2-12)reveals a resonance at 6= -198.4 ppm corresponding to the C-2 fluorine of unreacted 2F-DNPC (3- anomer). The broad signal resonating at 8 = -195.5 ppm is consistent with a75fluorine attached to a macromolecule and thus presumably corresponds to the C-2 fluorineof the 2FCb-enzyme. Similarly for Agrobacterium -g1ucosidase, the C-2 fluorine of the 2-deoxyfluoroglucosyl-enzyme is seen to resonate at ö = -197.3 ppm while the C-2 fluorineof the 2-deoxyfluoroglucosyl-[-fluoride inactivator resonated at = -203.4 ppm (Withers& Street, 1988). Although small differences (t6 = 2.9 ppm) in 19F chemical shiftsbetween a— and — anomers of 2-deoxyfluoroglycosyl esters are often seen (for example,the signal of the C-2 fluorine of the a-anomer of 1 ,3,4,6-tetra-O-acetyl-2-deoxy-2-fluoro-D-glucose is 1.5 ppm upfield of that of the 13-anomer (Csuk & Glanzer, 1988)) theanomeric configuration of the ester linkage of the intermediate cannot be reliably deducedbased only on this criteria.2F-GMF is seen to inactivate C. fImi exoglycanase in a time-dependent manner(Appendix A-8, Figure A-8-2). Thus, 19F-NMR analysis of 2F-GMF-inactivatedexoglycanase should provide further insights into the anomeric stereochemistry of theintermediate since the 19F chemical shift for 2F-mannose and its derivatives is moresensitive to the anomeric configuration than that for 2F-glucose and its derivatives, adifference of approximately 20 ppm being seen between a- and 13-anomers in all cases(Csuk & Glanzer, 1988). Samples of 2F-GMF-inactivated C. fimi exoglycanase anduntreated enzyme were subjected to mass spectrometric analysis. The molecular weight ofthe unlabeled catalytic domain of C. fimi exoglycanase is 34 815 ± 7 Da while that for thecatalytic domain treated with 2F-GMF is 35 155 ± 7 Da. The mass difference of 339 Dacorresponds well with the mass increase of 327 Da expected if the exoglycanase wasmodified by a single 2-deoxyfluoro-4-O-3-glucosylmannosyl- (2FGM-) moiety. Thus, themass spectrometric analysis is consistent with the formation of a covalent 2FGM-enzymespecies.‘9F-NIVIR analysis of this 2FGMF-inactivated exoglycanase sample (Figure 2-13)reveals signals at 6= -149.9 and -224.1 ppm corresponding to the C-i and C-2 fluorines,respectively, of excess 2F-GMF inactivator. This inactivated enzyme sample wasextensively dialysed against buffer to remove the sugars, and the dialysed sample76resubjected to‘9F-NMR analysis. The single broad peak seen at = -205.6 ppm in thespectrum of the dialysed sample (Figure 2-13), some 18.5 ppm downfield from the C-2fluorine of 2F-GMF, is consistent with the presence of 2FGM-enzyme. These 19F-NMRresults are consistent with those reported for Agrobacteriwn -glucosidase since the C-2fluorine of the 2-deoxy-2-fluoro-D--mannosyl fluoride inactivator is observed to resonateat 8 -224.4 ppm while that of the 2-deoxyfluoromannosyl-enzyme resonates 23.4 ppmdownfield at 8 = -201.0 ppm (Withers & Street, 1988). Since x-anomers of 2-deoxy-2-fluoromannosyl esters are known to generally resonate 16-20 ppm downfield from -anomers then the ‘9F signal at 6 = -205.6 ppm for the 2FGM-enzyme is completelyconsistent with the -anomeric configuration of the ester linkage. These results providefurther evidence in support of the formation of a covalent -glycosy1-enzyme intermediateduring C.fimi exoglycanase-catalysed hydrolysis of glycosides.2-4-8 C. fimi exoglycanase-catalysed hydration of cellobialHydration of glycals by retaining” glycosidases involves initial protonation at C-2of the glycal and attack at C-i by the catalytic nucleophile to form a covalent 2-deoxyglycosyl-enzyme intermediate similar to that formed during glycoside hydrolysis.Generally, the proton is delivered by the catalytic nucleophile as revealed by deuteriumlabeling experiments. However, in a few cases it is believed that the proton is donated bythe acid-base catalyst (discussed in detail in Chapter I). Subsequent hydrolysis of this 2-deoxyglycosyl-enzyme intermediate yields a 2-deoxy-glycose product and free enzyme.Legler (1990) has proposed a mechanism for the formation of the 2-deoxyglycosyl-enzymeintermediate involving the concerted addition of the carboxylic acid nucleophile acrossthe double bond via a cyclic 6-membered transition state (Figure 2-16) while Matsui andcoworkers (Matsui et al., 1993) have proposed an alternative mechanism involving anoxocarbonium ion. Hydrolysis of the 2-deoxyglycosyl-enzyme intermediate is believed tooccur via an oxocarbonium ion-like transition state (Matsui, 1993).77C. fimi exoglycanase-catalysed hydration of glycals was probed using cellobial.Interestingly, rather than hydration occuring, C.flmi exoglycanase is seen to be inactivatedin the presence of cellobial in a time-dependent manner (Appendix A-8). Massspectrometric analysis of the inactivated-exoglycanase is consistent with the formation of a2-deoxycellobiosyl- (2dCb-) enzyme (discussed in detail in Chapter IV), suggesting thatthe rate of formation of the intermediate is much greater than that of hydrolysis. This 2dCb-enzyme can be slowly converted to free enzyme and product via either a hydrolysis reactionor a transglycosylation reaction using a suitable glycosyl acceptor, such as cellobiose, in amanner similar to that seen for 2FG1u-, 2FCb- and2FX-exoglycanase, thus indicating thatthe 2dCb-enzyme intermediate is catalytically competent. These results are consistent withthose reported for E. coli f3-galactosidase since ‘H NMR experiments have revealed thatthis 3-galactosidase is capable of converting D-galactal to glyceryl 2-deoxy 3-D-galactosidein the presence of glycerol (Lehmann & Zieger, 1977). In contrast to C.fimi exoglycanase,studies with other glycosidases reveal that the corresponding 2-deoxyglycosyl-enzymeintermediates are not accumulated, but rather the glycals are readily converted to the 2-deoxy-glycose products (reviewed in Legler, 1990).The large value of the dissociation constant (K— 500 mM) for cellobial comparedto the dissociation constants (Kd) for the aryl -ceIlobiosides indicates that cellobial binds atleast 30 times less efficiently than the cellobioside substrates. These results are consistentwith those previously reported with other enzymes since glycals typically bind toglycosidases 10-100 times worse than do the corresponding glycosides (Legler, 1990).Given that glycals are somewhat similar to transition state analogues, in terms of theirplanar geometry78DEGLYCOSYLATI’kYCOSYLATIONOH0Ho0H•OHHIOHHH°THIOHOOHHOHFigure 2-16 Mechanism for glycosidase-catalysed hydration of a glycal.79at C-i, this poor binding is likely a consequence of the absence of the C-2 hydroxyl groupand thus the loss of important non-covalent interactions between the sugar and the enzyme.The X-ray crystal structure of the 2FCb-exoglycanase (White et al., 1995) provides someinsights into this situation. The side chains of Asn 126 and Glu 233 are seen to be withinhydrogen bonding distance of the fluorine atom at the C-2 position of the 2FCb moiety.This data suggest that the side chains of Asn 126 and Glu 233 are likely involved inhydrogen bonding interactions with the C-2 hydroxyl group of the parent cellobioside.Since the hydrogen atom at the C-2 position of cellobial cannot form hydrogen bondinginteractions with Asn 126 and Glu 233, the absence of these interactions likely contributesto the poor binding (K1— 500 mM) seen for this sugar.2-4-9 Effect of substitutions at C-2 on cellobioside hydrolysis rates(a) Fluorine substitution: Formation of the 2FCb-exoglycanase (k/K = 1.02 x 102s1 m1v11)is 104-fold less efficient than formation of the cellobiosyl-exoglycanase (k2/Kd =166 s1 mM1)for a dinitrophenyl cellobioside. Similarly, rate reductions of 3 x 102 and 1 xare seen on the glycosylation step for Agrobacterium 3-glucosidase (Street et al., 1992)and E. coli 3-galactosidase (McCarter et al., 1992) respectively, when the C-2 hydroxylgroup is replaced by a fluorine atom. These results are consistent with destabilization of theoxocarbonium ion-like transition state for glycosylation due to the presence of theelectronegative fluorine atom at C-2. As previously mentioned, the X-ray crystal structurefor 2FCb-exoglycanase reveals that the amine group of Asn 126 and the carboxylate groupof Glu 233 are seen to be within hydrogen-bonding distance of the C-2 position of theproximal glucosyl moiety of the 2FCb sugar, thus the fluorine atom can accept a hydrogenbond from the amine group. However, the hydroxyl group that is normally present at thisC-2 position of the sugar could either donate or accept a hydrogen bond with the aminegroup of Asn 126, as well as form a hydrogen bond with Glu 233. Although these80hydrogen bonds are to the intermediate, presumably they will be present and stronger at thetransition states. Thus, replacement of the C-2 hydroxyl group with a fluorine atom resultsin the loss of several important hydrogen bonds which serve to further destabilise theglycosylation transition state.The rate of hydrolysis of the 2FCb-exoglycanase is 107-fold lower than that of thecellobiosyl-enzyme. Similarly, rate reductions of 108-fold have been reported on thedeglycosylation step for both Agrobacterium 3-glucosidase (Street et al., 1992) and E. coli-galactosidase (McCarter et aL, 1992) when a fluorine atom is substituted for the normalC-2 hydroxyl group. Comparison of the rate reductions seen on the glycosylation step(104-fold) with that seen on the deglycosylation step (107-fold) upon replacing the C-2hydroxyl group with a fluorine atom indicates that the presence of the fluorine atom has agreater destabilising effect on the transition state for deglycosylation than on that forglycosylation. This interpretation is consistent with the previous suggestion of a greaterdegree of oxocarbonium ion character at the transition state for deglycosylation than at thatfor glycosylation, as suggested by the larger (x-secondary deuterium kinetic isotope effectmeasured on the deglycosylation step using 2,4-DNPC and PNPC compared to thatmeasured on the glycosylation step using 4-BrPC.(b) Hydrogen substitution: Hydrolysis of the 2dCb-exoglycanase is 106 times lessefficient than that of the cellobiosyl-exoglycanase. Similarly, the rate of hydrolysis of the 2-deoxygalactosyl-f3-galactosidase from E. coli is io times lower than that for thegalactosyl-enzyme (Wentworth & Wolfenden, 1974; Viratelle & Yon, 1980). Interestinglyhowever, for Aspergillus wentii 3-glucosidase, the rates of hydrolysis for both the 2-deoxyglucosyl- and glucosyl-enzymes were identical (Roeser & Legler, 1981).Substitution of a hydrogen atom for a hydroxyl group results in the loss of importanthydrogen bonds normally present between the C-2 hydroxy group of the sugar and theenzyme, thus leading to destabilisation of the deglycosylation transition state. Comparison81of the rate of hydrolysis of the 2dCb-enzyme with that of the Cb-enzyme (kcat for 2,4-DNPC) indicates that the non-covalent interactions between C-2 of the sugar and theenzyme contribute at least —9 kcal/molea. This represents only a minimum estimate of thestabiisation energy contributed by the C-2 hydroxyl group since the electronic effects onthe oxocarbonium ion-like transition state from the electropositive hydrogen at C-2 willpartially compensate for the loss of these binding interactions.2-5 SummaryC. fimi exoglycanase has been shown previously to hydrolyse its substrates withnet retention of anomeric configuration, thus a double displacement reaction has beenproposed to describe its mechanism of action.(Withers et al., 1986). This enzyme has beensubjected to a detailed kinetic investigation with a range of aryl f3-glycosides in order toprovide additional evidence in support of the proposed mechanism.Values of kcat are found to be invariant with pH whereas those of kcat/Km aredependent upon two ionisations of pKa = 4.1 in the acidic limb of the p11 profile and pKa =7.7 in the basic limb, presumably corresponding to the catalytic nucleophile (Glu 233) andthe acid catalyst (Glu 127), respectively. Secondary deuterium kinetic isotope effects on theglucosides revealed that the transition states for formation and hydrolysis of the glucosylenzyme are relatively late with almost complete bond cleavage and little preassociation ofaDetened usingkcat (2,4DNPC)=- RT lni(2dCb-enzyme)82the nucleophile. Further, the large negative charge accumulation seen on the phenolateoxygen as reflected by the value of the Broensted constant suggests that there is littleprotonic assistance at the transition state for formation of the glucosyl-enzyme.Similar analyses on the cellobiosides in combination with previous inactivationstudies with 2F-DNPC and 2F-DNPG reveal that the distal glucosyl moiety of thecellobiosides accelerates the rate of formation of the glycosyl-enzyme, but has nosignificant effect on its rate of hydrolysis. The smaller secondary deuterium kinetic isotopeeffect and lesser negative charge development on the phenolate oxygen for cellobiosylenzyme formation relative to that for glucosyl-enzyme formation suggest that the presenceof the second glucosyl moiety results in an earlier transition state for cellobiosyl-enzymeformation. This likely arises from increased acid catalysis, possibly coupled with a greaterdegree of nucleophilic preassociation by the catalytic nucleophile. Presumably the energyinput required for formation of this more highly organised cellobiosylation transition staterelative to the glucosylation transition state is derived from the additional bindinginteractions with the distal glucosyl moiety of the cellobiosides. By contrast, the transitionstate for hydrolysis of the cellobiosyl-enzyme is seen to be similar to that for hydrolysis ofthe glucosyl-enzyme, occurring relatively late with significant C-O bond cleavage andoxocarbonium ion character.Kinetic analysis with the xylo-substrates reveal that the substrate preference of thisenzyme increases in the order glucosides <xylosides <cellobiosides <xylobiosides. Theseresults, in combination with the X-ray crystal structure of 2FCb-exoglycanase (White etal., 1995), indicate that the distal C-5 hydroxylmethyl group is slightly inhibitory tocatalysis. The comparable Broensted constants for xyloside and glucoside hydrolysisreveals that the transition states for formation of the xylosyl-enzyme and the glucosylenzyme are similar.83Inactivation studies with 2F-DNPC and cellobial reveal that the C-2 hydroxyl groupof the sugar is necessary for catalysis. Inspection of the X-ray crystal structure of the2FCb-exoglycanase revealed that the C-2 hydroxyl group, normally present at thisposition, likely forms strong hydrogen bonds with the enzyme at the transition state, mostlikely with Asn 126 and Glu 233. Removal of this hydroxyl group is seen to decrease thebinding energy available for stabilisation of the deglycosylation transition state by at least 9kcal/mole.Mass spectrometric analysis of 2FCb-exoglycanase provided further evidence insupport of the covalent nature of the glycosyl-enzyme intermediate formed during C. Jlmiexoglycanase catalysed hydrolysis of glycosides. Additional evidence was derived from thevalues of the secondary deuterium kinetic isotope effects measured for glycosides havingdeglycosylation as the rate determining step (kH/kD = 1.12 for the glucoside and kH/kD =1.10 - 1.12 for the cellobiosides), since such effects could only be seen if the intermediatehas more sp3 character than the subsequent deglycosylation transition state. 19F-NMRanalysis of the 2FCb- and 2FGM-enzymes provided evidence for the a-stereochemistry ofthe anomeric linkage between the sugar and the enzyme.All the data presented are consistent with the proposed double displacementmechanism for this exoglycanase.84CHAPTER IIIDETAILED KINETIC ANALYSIS OF MUTANTS OF C. FIMIEXOGLYCANASE853-1 IntroductionSite directed mutagenesis can be used to systematically replace individual residueswithin a protein one at a time and thus provides a way of probing the role of specificresidues in binding and catalysis. These studies have been performed extensively onseveral glycosidases in order to identify possible candidates for the role of acid-basecatalyst (Sierks et al., 1990; Chauvaux et a!., 1992; Svensson & Sogaard, 1993; MacLeodet al., 1994; Juncosa et al., 1994; Wang et al., 1995; Damude et al., 1995) as well as forthe role of catalytic nucleophile (Malcolm et al., 1989; Holm et al., 1990; Grace et al.,1990; Py et al., 1991; Svensson & Sogaard, 1993; Juncosa, 1994). For example, severalconserved carboxylate residues within Clostridium thermocellum endoglucanase D wereindividually replaced with alanine residues and the mutants characterised kinetically(Chauvaux et al., 1992). The kcat value for the hydrolysis of PNPC by the Glu555Alamutant was reduced 4 x103-fold relative to the native enzyme while other properties (e.g.,Km and affmity for Ca2)remained basically unchanged. Based on the behaviour of thisGlu555Ala mutant relative to that for the other mutants, Glu 555 was proposed to be theacid-base catalyst in this enzyme. This assignment has been confirmed by the X-ray crystalstructure of the enzyme co-crystallised with the inhibitor, o-iodobenzyl-f3-D-thiocellobioside at the active site. This revealed that Glu 555 was appropriately positionedwithin the active site of the enzyme to perform this role (Juy et aL, 1992). Indeed a similarstrategy but with much more extensive kinetic analysis, has been used successfully toidentify the acid-base catalyst in C.fimi exoglycanase (MacLeod et al., 1994) andAgrobacterium f3-glucosidase (Wang et al., 1995) as Glu 127 and Glu 170 respectively. Inthe case of the exoglycanase, this assignment has been confirmed by X-ray crystallographicanalysis (White et al., 1994).Site directed mutagenesis of glycosidases can also be used to further investigate therole of putative catalytic nucleophiles and acid-base catalysts identified previously from86independent studies such as labeling experiments. For example, the catalytic nucleophile ofAgrobacterium f-glucosidase has been identified as Glu 358 by trapping of the enzyme as a2-deoxyfluoro-a-D-glucosyl-enzyme intermediate using the mechanism-based inactivator2,4-dinitrophenyl 2-deoxy-2-fluoro-f3-D-glucoside (2F-DNPG) (Withers et al., 1990).Replacement of Glu 358 by asparagine and glutamine residues using site directedmutagenesis generated mutant enzymes which were essentially completely inactivated(Withers et al., 1992). Since the catalytic nucleophile must stabilise the oxocarbonium ion-like transition states, form a covalent linkage with the substrate and in addition serve as agood leaving group, then the loss of enzymatic activity upon conversion of the carboxylategroup (Glu 358) to an amide group (Asn and Gln) is certainly consistent with the role ofGlu 358 as the catalytic nucleophile. By contrast, shortening the carboxylate side chain byreplacing Glu 358 by an aspartate residue resulted in a mutant for which the kcat value wasreduced 2.5 x 103-fold, corresponding to an increase in activation energy of —4.5kcal/mole. Thus withdrawing the negatively charged carboxylate group —1 A away from thereacting centre of the sugar destabilises the glycosylation transition state by at least 4.5kcal/mole. These results are completely consistent with the role of Glu 358 as the catalyticnucleophile. Furthermore, this kinetic analysis of Glu358Asp provided new insights intothe mechanistic consequences of mispositioning catalytic nucleophiles within glycosidases.3-2 Objectives Of This ProjectThe objective of this project is to probe the roles of the catalytic nucleophile and theacid-base catalyst of C. fimi exoglycanase through detailed kinetic analysis of mutantsgenerated at these positions in this enzyme by site directed mutagenesis.Catalytic nucleophile: Previously, Glu 233 was identified as the catalyticnucleophile in C.fimi exoglycanase from labeling studies using the mechanism-basedinactivator, 2F-DNPG (Tull et al., 1991). This assignment has been confirmed recently byX-ray crystallographic analysis of this enzyme (White et al., 1994). A mutant in which Glu87233 has been replaced with an aspartate residue has been prepared by Dr. AlasdairMacLeod, Department of Microbiology, University of British Columbia, allowing forfurther investigation of its role. A kinetic analysis of this mutant, similar to that previouslyperformed on the native enzyme, should provide insights into the mechanisticconsequences of pulling the catalytic nucleophile of C. Jimi exoglycanase away from thereacting centre of the substrate (Figure 3-1).ORGlcOR°N°H2 FGlu AspFigure 3-1 Replacement of the catalytic nucleophile of C.fimi exoglycanase (Glu 233)by an aspartate.Acid-base catalyst: Glu 127 has been identified previously as the acid-base catalystin C. fimi exoglycanase through a combination of site directed mutagenesis of conservedcarboxylate residues to alanine residues and kinetic analysis of the resulting mutants(MacLeod et al., 1994). In this investigation, the mechanistic consequences of removal ofthe acid-base catalyst of C.fimi, Glu 127, (Figure 3-2) are explored further throughdetailed kinetic analysis of the Glul27Ala mutant.88Glu AlaCH3Figure 3-2 Replacement of the acid-base catalyst of C.Jimi exoglycanase, Glu 127,by an alanine residue.3-3 Results For The Glu233Asp C. fimi Exoglycanase (Nucleophile)Mutant3-3-1 Substrate reactivityAryl 13-cellobiosides, aryl f3-D-glucosides and PNPX2were reacted with theGlu233Asp C.fimi exoglycanase mutant and the kinetic parameters determined. The kineticparameters for PNPX2are kcat = 0.011 ± 0.001 s_i, Km = 0.018 ± 0.001 mM and kcat/Km= 0.61 s_i mM4.Values for k, Km and kcatfKm for the cellobiosides and glucosides arelisted in Table 3-1 and 3-2 along with the pKa values of the phenol leaving group. TheMichaelis-Menten kinetic parameters for the cellobiosides are plotted as functions of thephenol pKa in the form of Broensted plots and these are shown in Figure 3-3 along withthose for the native enzyme. Values of log(kca) are independent of the pKa values over therange pKa = 4 - 9. In contrast, the log(k/Km) plot reveals a relatively weak dependence(I3ig = -0.3, correlation coefficient = 0.94) of log(k/Km) upon the pKa values across thefull range of substrates studied.HO.89Table 3-1: Michaelis-Menten Parameters for the Hydrolysis of Aryl 13-D-cellobiosides bythe Glu233Asp Mutant.Phenol substituent apKa Km (mM) kcae (s1) kcat/Km(s1mM)2,4 -dinitro 3.96 0.086 ± 0.006 0.0031 ± 0.0001 0.0363,4-dinitro 5.36 0.10 ± 0.004 0.0027 ± 0.0001 0.0274-nitro 7.18 0.66± 0.03 0.0038±0.0001 0.00582-nitro 7.22 0.62±0.02 0.0030±0.0001 0.00484-cyano 8.49 1.6 ± 0.4 0.0028 ± 0.0005 0.00184-bromo 9.34 1.2 ± 0.3 0.0025 ± 0.0002 0.0021aphenol PKa values were taken from Barlin & Pen-in (1966), Kortum et al. (1961),Robinson et al., (1960), and Ba-Saif & Williams (1988).Table 3-2: Michaelis-Menten Parameters for the Hydrolysis of Aryl 13-D-Glucosides by theGlu233Asp Mutant.Phenol substituent ap Km (mM) kcat (s1) kcat/Km(s1mM)2,4-dinitro 3.96 1.4 ± 0.1 0.0035 ± 0.0001 0.00262,5-dinitro 5.15 2.0 ± 0.1 0.0039 ± 0.0001 0.00203,4-dinitro 5.36 5.5 ± 2.0 0.0010 ± 0.0004 0.000182-chloro-4-nitro 5.45 20 ± 2 0.0018 ± 0.0001 0.000089aphenol pKa values were taken from Barlin & Perrin (1966); Kortum et al. (1961);Robinson et al (1960); and Ba-Saif & Williams (1988).9021• •• S.0•-10-20 o 0 o-3-4 -3 4 5 6 7 8 9 10pKaI37 8910pKaFigure 3-3 Broenstedplots relating rates of native enzyme- (solidpoints) andGlu233Asp mutant- (open points) catalysed hydrolysis of aryl /3-cellobiosides with the leaving group ability of the phenols.913-3-2 ct-Secondary deuterium kinetic isotope effectA secondary deuterium kinetic isotope effect on kc,t of kH/kD = 1.09 ± 0.02 wasmeasured for the Glu233Asp mutant using 2,4-DNPC, exactly as with the native enzyme.3-3-3 pH StudyThe pH dependence of Glu233Asp mutant was investigated using 2,4-DNPC as thesubstrate over the pH range 4.6- 8.4. Values for kcat and Km were determined at thedifferent pH values exactly as with the native enzyme and these are presented as plots ofkcat and kcat/Km versus pH in Figure 3-4. Interestingly values of both kcat and kcat/Km areseen to be independent of pH.EE9221-2.3.4I I I I I ‘ I ‘ I•-.sa.. i.• . IDD DD D D--i L_ i I4 5 6 7 8 9pH20-24 5 6 7pHFigure 3-4 pH Dependence of the hydrolysis of2,4-DNPC by the Glu233Asp mutant(open points) and native enzyme (solid points)8 9933-4 Results For Glul27Ala C. Fimi Exoglycanase (Acid-Base Catalyst)Mutant3-4-1 Substrate reactivityMichaelis-Menten kinetic parameters were determined for five aryl f3-cellobiosideswith the Glul27Ala mutant and these are listed, in addition to the two values determinedpreviously (MacLeod et al., 1994) in Table 3-3 along with the pKa values for thecorresponding phenolate leaving groups. Values of log(k) and log(kIKm) are plotted asfunctions of the pKa of the phenolate leaving group in the form of Broensted plots andthese are shown in Figure 3-5 along with those for the native enzyme. Values of log(kca)Table 3-3: Michaelis-Menten Parameters for the Hydrolysis of Aryl f3-Cellobiosides by theGlul27Ala MutantPhenol substituent bpKa Km (mM) keat (si) kcat/Km(s1mM)2,4dinitroa 3.96 0.0003 0.040 1333,4-dinitro 5.36 0.0008 ± 0.0002 0.072 ± 0.007 904-nitro 7.18 0.020 ± 0.002 0.033 ± 0.001 1.652-nitro 7.22 0.021 ± 0.003 0.021 ± 0.001 1.03,5-dichloro 8.19 0.15 ± 0.03 0.0027 ± 0.0005 0.0184-cyano 8.49 0.43 ± 0.01 0.015 ± 0.0001 0.0354bromoa 9.36 1.9 1.4x105 7.4x106a(MacLeod et al., 1994)bPhenol pKa values were taken from Barlin & Perrin (1966), Kortum et al. (1961),Robinson et al., (1960), and Ba-Saif & Williams (1988).Figure 3-5. Broenstedplots relating rates ofnative enzyme- (solid points) andGlul27Ala mutant- (open points) catalysed hydrolysis of aryl f3-cellobiosides with the leaving group ability of the phenols.940.5-1.5-3.5.55.• ••.%000003 4 5 6 7 8 9pKa1003.0 5.0 7.0 9.0 11.0pKa95are invariant with PKa over the range pKa 4-7 but become dependent on pKa at the higherpKa values as reflected by the downward break in the plot. The approximate Broenstedconstant for this PKa dependent region of the plot corresponds to — -1.2 (correlationcoefficient = 0.73). A similar trend is apparent in the log(kcat/Km) versus pKa Broenstedplot though the data are more scattered. The points fit reasonably well to a line of slope 13g-1.2 (correlation coefficient = 0.82) but with the more reactive substrates falling beneaththis line, as has been seen previously (Kempton & Withers, 1992).3-4-2 Stopped-flow analysisPre-steady state kinetic analyses of the Glul27Ala mutant were carried out for threecellobioside substrates, exactly as with the native enzyme. The results are presented inTable 3-4 and in the form of plots of log(k2) and log(k2IK) versus pKa (Figure 3-6) withslopes corresponding to Broensted constants of 13g — -0.8 (correlation coefficient = 0.95)and -0.7 (correlation coefficient = 0.87), respectively.Table 3-4: Pre-Steady State Parameters for Hydrolysis of Aryl -Cellobiosides by theGlul27Ala MutantPhenol substituent apK k2 (s1) KJ (mM) k2/KiJ(s1mM)2,4-dinitro 3.96 192±7 0.68 ± 0.06 2823,4-dinitro 5.36 3.31 ± 0.07 0.019 ± 0.002 1794-nitro 7.18 0.30 ± 0.01 0.17 ± 0.02 1.79aphl pKa values were taken from Barlin & Perrin (1966); Kortum et al. (1961);Robinson et al., (1960); and Ba-Saif & Williams (1988).0963210—13.0 4.0 5.0 6.0 7.0pKa3 4 5 6 7 8pKaFigure 3-6 Broensted plots relating pre-steady state rate of Glul27Ala mutant catalysedhydrolysis of aryl /3-cellobiosides with the leaving group ability of thephenols.973—4—3 x-Secondary deuterium kinetic isotope effectA secondary deuterium kinetic isotope effect on kcat of kH/kD = 1.08±0.2 wasmeasured for the Glul27Ala mutant using 2,4-DNPC, exactly as with the native enzyme.3-4-4 pH StudyTwo cellobioside substrates, 2,4-DNPC and PNPC, were used to investigate thepH dependence of Glul27Ala mutant catalysis over the pH range 4.6- 9.5. Values for kcatwere determined by measuring rates of hydrolysis at saturating concentrations of 2,4-DNPC (75 x Km). Values for kcat/Km were determined by following the time courses forthe hydrolysis of PNPC at concentrations much lower than Km (0.2 x Km), since at [SI <<Km, rates of hydrolysis are first order in substrate and the observed rate COflStafltS (‘Cobs)correspond to k/Km (discussed in Appendix B-i). These results are presented as plots ofiCcat and kcat/Km versus pH in Figure 3-7. Values of kcat/Km are seen to be dependent upononly one ionisation of pKa = 5.9 ± 0.1 in its acidic limb. Interestingly, values of keatappear to be dependent upon an ionisation of pKa — 5.0 ± 0.1 in the basic limb of the pH-profile. Although such pH-profiles are uncommon, this 1og(k) pH-profile isreproducible. This ionisation of pKa - 5.0 is only an estimate since instability ofGlul27Ala mutant at low pH values precludes a more accurate determination.98I • I I1 --iIp • U U— 0.-I—2 I I I i I i4 5 6 7 8 9pH2.5. 2p1.5Ii50.510-0.54 5 6 7 8 9 10pHpH Dependence of the hydrolysis of cellobiosides by the Glul27Ala mutant(open points) and native enzyme (solid points)Figure 3-7993-5 Discussion Of The GIu233Asp C.Fimi Exoglycanase (Nucleophile)Mutant3-5-1 Proposed role of the catalytic nucleophile in C. fimi exoglycanasecatalysisBased on the double displacement mechanism proposed for C.fimi exoglycanasecatalysed hydrolysis of glycosides (Withers et al., 1986) the role of the catalyticnucleophile involves the formation of a covalent linkage with the sugar in the glycosylenzyme intermediate in the first step of catalysis. In the second step, the catalyticnucleophile must serve as a good leaving group. Furthermore, the catalytic nucleophilemust also assist in the stabilisation of the oxocarbonium ion-like transition states throughelectrostatic interactions. The catalytic nucleophile presumably plays a minimal role inground state binding since interactions between the sugar and the catalytic nucleophile at theground state would likely be disrupted as the reaction proceeds to the transition state andwould therefore be inhibitory. This is consistent with the view that interactions realised atthe transition state rather than at the ground state, stabilise the transition state, thus loweringthe activation energy of the reaction (Fersht, 1985).3-5-2 pH Dependence of the GIu233Asp C. fimi exoglycanase mutantThe pH dependence of Glu233Asp was investigated using 2,4-DNPC, a substratefor which deglycosylation is the rate determining step. The similar pH profiles observed forkcat and kcat/Km indicate that binding of the substrate does not perturb the ionisations in thefree Glu233Asp mutant enzyme. Like that for the native enzyme, values of keat for theGlu233Asp mutant are seen to be independent of pH over the range studied. By contrast,the pH profile of kcat/Km for the native enzyme is quite different from that of theGlu233Asp mutant. Values of kcat/Km for the Glu233Asp mutant are seen to beindependent of pH while the kcat/Km pH profile for the native enzyme reveals two100ionisations of pKa = 4.1 and 7.7 likely coffesponding to the catalytic nucleophile (Glu 233)and the acid catalyst (Glu 127), respectively. These results suggest that the ionisations seenin the free native enzyme are perturbed to values outside the pH range studied. The X-raystructure of this exoglycanase provides some insights into this situation. For the nativeenzyme, Glu 233 is seen to be within hydrogen bonding distance of both His 205 and Asn169; and His 205 within hydrogen bonding distance of Asp 235 (White et al., 1994). Thishydrogen bonding net work is shown schematically in Figure 3-8. Since Asp 235, His 205and Glu 233 are highly conserved residues within this family of f3-glucanases, it has beensuggested that the hydrogen bonding network of Asp235-His205-G1u233 likely plays animportant role in maintaining the ionisation state of the nucleophile (Glu 233) (White et al.,1994). Modeling studies with the native enzyme X-ray crystal structure (without furtherstructural refinement) can provide some nreliminarv indication of the hydrogen bondinginteractions that are likely lost or altered by converting Glu 233 to an aspartate residue. Thisstudy reveals that both His 205 and Asn 169 are no longer within hydrogen bondingdistance of the catalytic nucleophile when Glu 233 is shortened by 1 A. The absence ofthese hydrogen bonding interactions in the Glu233Asp mutant would therefore be expectedto modify the environment of the active site, thus altering the ionisation states of both thecatalytic nucleophile and acid-base catalyst, as suggested by the pH-profiles.Asp 235Figure 3-8 Schematic diagram illustrating the hydrogen bonding net work aroundGlu 233.Asn 1691013-5-3 Substrate specificity of the Glu233Asp mutantComparison of the kcatfKm values for hydrolysis of the cellobiosides (Table 3-1)with those for the corresponding glucosides (Table 3-2) reveals that, like the nativeexoglycanase, the Glu233Asp mutant catalyses the hydrolysis of cellobiosides moreefficiently than glucosides. Likewise, the Glu233Asp mutant catalyses the hydrolysis ofPNPX2more efficiently than PNPC as was seen for the native enzyme. These resultsindicate that truncation of the catalytic nucleophile likely affects the mechanism forhydrolysis of the cellobiosides, the glucosides and the xylobiosides in a similar manner.3-5-4 Rate determining steps for GIu233Asp mutant catalysisThe Broensted plot of log(k/Km) versus pKa (Figure 3-3) for the Glu233Aspmutant-catalysed hydrolysis of aryl f-cellobiosides reveals a strong dependence of the rateon the leaving group ability of the phenolate aglycone across the PKa range studied, as wasseen with the native enzyme. This indicates that the first irreversible step in the reaction isstill the formation of the cellobiosyl-enzyme intermediate. The value of the Broenstedconstant, 13g = -0.3, for hydrolysis of the cellobiosides with the Glu233Asp mutant is thesame as that found previously for hydrolysis of these substrates with the native enzyme.Similarly for Agrobacterium f3-glucosidase, the Broensted constants for hydrolysis of aryl13-D-glucosides with the native enzyme(3ig = -0.7) and with the Glu358Asp catalyticnucleophile mutant (I3ig = -0.7) were found to be the same indicating that mispositioning ofthe carboxylate group does not affect the degree of negative charge development on thedeparting phenolate at the glycosylation transition state, thus the degree of C-O bondcleavage or proton donation (Withers et al., 1992). By contrast, the absence of a correlationbetween values of log(kca) and leaving group pKa over the entire pKa range studied (pKa =4-9) for the Glu233Asp mutant suggests that the rate determining step is not C-O bondcleavage of the phenolate but most likely hydrolysis of the cellobiosyl-enzyme intermediate.For the native enzyme, deglycosylation was seen to be the rate determining step only for102the cellobiosides with good leaving groups (pKa < --7.5) while glycosylation became therate determining step for those cellobiosides with poorer leaving groups (pKa> 7.5).Thus shortening the catalytic nucleophile side chain has changed the rate determining stepfor the poorer substrates from glycosylation to deglycosylation. Interestingly, exactly theopposite situation is found for Agrobacterium -glucosidase (Withers et al., 1992).Conversion of the glutamate catalytic nucleophile (Glu 358) of this 3-glucosidase to anaspartate residue resulted in a change in the rate determining step from deglycosylation toglycosylation for hydrolysis of all the aryl -D-glucosides studied.3-5-5 Effect of mispositioning the carboxylate group on the individualsteps of the reaction.Glycosylation step: Values of kcat/Km reflect the first irreversible step (AppendixB-2) in the reaction which is the glycosylation step for the cellobiosides and presumablyalso for the glucosides and the xylobioside studied. Estimates of the effect on theglycosylation step of mispositioning of the carboxylate group can therefore be obtainedfrom the relative kcailKm values for the Glu233Asp mutant and the native enzyme. Valuesof kcat/Km for hydrolysis of the cellobiosides (except 4-BrPC) by the Glu233Asp mutantare 3-4 x 103-fold lower than those for the native enzyme. These rate reductions are similarto those seen for Agrobacterium f-glucosidase (Withers et al., 1992) and E. coli -galactosidase (Yuan et al., 1994) (2.5 x 103-fold and 2 - 9 x 103-fold, respectively) uponreplacement of the glutamate catalytic nucleophiles with aspartate residues. Presuming thatthe majority of this effect is due to changes in the transition state and not ground stateinteractions (vide supra) then this rate reduction corresponds to an increase in theglycosylation transition state energy of —5 kcal/mole. Interestingly, the kcat/Km value forhydrolysis of 4-BrPC is reduced to a lesser extent (—9 x 102-fold, MG° —4 kcal/mole)than those for hydrolysis of the more reactive cellobiosides. This suggest that the movingthe carboxylate 1 A away from the reacting centre has a lesser destabilising effect on the103glycosylation transition state for hydrolysis of 4-BrPC than on those for hydrolysis of themore reactive cellobiosides. Presumably, this is because the glycosylation transition statefor hydrolysis of 4-BrPC is already more destabilised than those for hydrolysis of the morereactive cellobiosides with the native enzyme.Values of kcat/Km for hydrolysis of 2,4-DNPG and 3,4-DNPG are seen to bereduced (—‘3 x 103-fold) a similar amount to that seen for the equivalent cellobiosides.Similarly, both the kcatfKm values for PNPC and PNPX2hydrolysis are reduced —3 x 10-fold. These results indicate that mispositioning the carboxylate group destabilises theglycosylation transition states for hydrolysis of cellobiosides, glucosides and xylobiosidesto a similar extent.Deglycosylation step: The deglycosylation rate constant for cellobioside hydrolysis,obtained from kcat values for 2,4-DNPC, 3,4-DNPC, PNPC and ONPC, is also reduced 4x 103-fold, corresponding to an increase in activation energy of MG° = —5 kcal/mole,upon replacing the glutamate residue with an aspartate. Since shortening the side chain ofthe catalytic nucleophile would almost certainly result in a strained covalent cellobiosylenzyme intermediate, then the above value of MG represents a minimum estimate of theactual increase in the deglycosylation transition state energy. Since the increase in energy ofthe glycosylation transition state is —5 kcalfmole while that of the deglycosylation transitionstate is likely greater than 5 kcal/mole, this indicates that mispositioning the negativelycharged carboxylate group probably has a greater destabilising effect on the transition statefor deglycosylation than on that for glycosylation (Figure 3-9). This situation is mostclearly seen for 4-BrPC. The glycosylation rate for this substrate is reduced only —9 x 102fold while the deglycosylation rate is reduced at least 4 x 103-fold upon pulling thecarboxylate group —1 A away from the reacting anomeric carbon. This interpretation iscertainly consistent with the findings that for Glu233Asp mutant, deglycosylation is the ratedetermining step for the cellobiosides studied while for the native enzyme,104deglycosylation is niy rate determining for the more reactive cellobiosides (e.g. 2,4-DNPC) and glycosylation is rate determining for less reactive cellobiosides (e.g. 4-BrPC).GE + S - E.S ES EP EP E + PReaction coordinateE+PFigure 3-9 Reaction coordinate diagram illustrating the effect ofshortening the catalyticnucleophile on the glycosylation and deglycosylation transition states.The a-secondary deuterium kinetic isotope effects measured for the native enzymeand the Glu233Asp mutant provide further insights into this situation. The similar kineticisotope effects measured for 2,4-DNPC with the native enzyme (kH/kD = 1.10) and theGlu233Asp mutant (kH/kD = 1.09) indicate that there is a similar extent of oxocarboniumion development at the transition states for hydrolysis of the cellobiosyl-enzymeEPt4 -‘I IES I II S II$ II I$ I$ II I$ I II AGc’*(mut)• I I $I IAGgmut)E.SEP105intermediate in both cases, thus the extent of C-O bond cleavage of the carboxylate in thedeglycosylation step is not significantly affected by truncation of the catalytic nucleophile.3-6 Discussion Of Glul27AIa C. Fimi Exoglycanase (Acid-Base Catalyst)Mutant3-6-1 Proposed role of the acid-base catalyst in C. fimi exoglycanasecatalysisThe role of the acid-base catalyst, based on the proposed mechanism of action of C.fimi exoglycanase, involves donation of a proton (general acid catalysis) to the oxygen ofthe leaving aglycone to assist with C-O bond cleavage through stabilisation of the leavinggroup in the first step of the reaction. In the second step of catalysis, this same residue isproposed to abstract a proton (general base catalysis) from the attacking water, thusassisting with nucleophilic attack on the glycosyl-enzyme intermediate. Removal of thisresidue would therefore be expected to affect icth steps in catalysis.3-6-2 pH Dependence of Glul27AIa C. fimi exoglycanase mutantThe pH dependence of kcat/Km was investigated using PNPC as the substrate,while that of kcat was investigated with 2,4-DNPC, a substrate for which thedeglycosylation step is rate determining. Recall that values of kcat/Km for native enzymecatalysis are dependent upon two ionisations of pKa 4.1 and 7.7. In contrast, the kcat/Kmversus pH profile (Figure 3-7) of Glul27Ala catalysis reveals a single ionisation of pKa =5.9 in the acidic limb but no ionisations at the higher pH values. These results suggest thatthe group responsible for the ionisation of pKa = 7.7 in the native enzyme, which must bein its deprotonated state to be catalytically active, has been removed. Indeed similar resultshave been reported for TEM-1 13-lactamase (Delaire et al., 1991), recombinant proteintyrosine phosphatase from Yersinia enterocolitica (Zhang et al., 1994) and Agrobacterium(3-glucosidase (Wang et al., 1995) upon mutation of the acid-base catalysts. Insights into106this shift in pKa from 4.1 to 5.9 may be obtained from the X-ray crystal structure of C.fImi exoglycanase (White et al., 1994). Glu 127 is seen to interact with Asn 126, Gin 203,Trp 84 and a water molecule, which presumably are involved in maintaining the ionisationstate of Glu 127 in the free enzyme. Conversion of Glu 127 to an alanine residue wouldtherefore be expected to result in the loss of these hydrogen bonding interactions, as well asto alter the local charge density within the active site. Thus, the shift in pKa from 4.1 to 5.9is likely due to these changes occurring in the active site upon removal of the negativelycharged carboxylate group of Glu 127. Assignment of this ionisation of pKa 5.9 to aspecific group is not a thvial matter. Since this ionisation reflects a group that must bedeprotonated to be catalytically active, then it most likely corresponds to the catalyticnucleophile.Differences between the native enzyme and the Glul27Ala mutant are also seen inthe pH dependence of log(k). While no ionisations in the enzyme/substrate complex areseen for the native enzyme, an increase in k with decreasing pH, below pH— 6, is seenfor the mutant, corresponding to an ionisation of pKa — 5. Although such pH-profiles areuncommon, this log() pH-profile is reproducible. Since the rate determining step forthis substrate is deglycosylation, this ionisation is likely due to a group which interacts withthe catalytic nucleophile (Glu 233), increasing its leaving group ability.3-6-3 Effect of removal of the acid-base catalyst on 3-glucanaseactivityThe biphasic log(k) versus pKa Broensted plot (Figure 3-5) seen for Glul27Alacatalysis indicates that deglycosylation is likely the rate determining step for the morereactive cellobiosides (phenol leaving group pKa <7) whereas glycosylation is ratedetermining for the less reactive cellobiosides (phenol leaving group pKa > 7) as seen withthe native enzyme. The results from the stopped-flow analysis are consistent with thesefindings. For the three most reactive cellobiosides (2,4-DNPC, 3,4-DNPC and PNPC)107studied, the observed bursts of released, substituted phenolates indicate that a stepfollowing aglycone departure, likely deglycosylation, is rate determining for thesesubstrates.By contrast, the log(kcaj/Km) versus pKa Broensted plot (Figure 3-5) appearsscattered, thus it is difficult to interpret. Similarly for Agrobacterium f3-glucosidase, thelog(k/K) versus PKa Broensted plot for hydrolysis of f3-glucosides by the acid-basecatalyst mutant, Glul7OGly, is also seen to be scattered (Wang et al., 1995). Presumablythis scattering is due to the superimposition of binding effects onto the electronic factors(discussed in section 2-1-5). However values of log(kc/K) for the Glul27Ala mutantmay still be used to provide insights into transition state structure for hydrolysis of thesecellobiosides. If the scatter observed is truly due to binding effects and if these samebinding effects are present in the data on the native enzyme, then a plot of log(k/K) forthe Glul27Ala mutant versus log(kJK) for the native enzyme should be reasonably linearas these binding effects will be subtracted. The slope of such a plot will provide a measureof the relative charge development on the phenolate oxygen in the two cases. Indeed theversus log(kc/Km)najve plot (Figure 3-10) is linear with acorrelation coefficient of 0.9 and a slope of 4.0, indicating much greater chargedevelopment on the phenolate oxygen at the glycosylation transition state with the mutantenzyme. Since the Broensted constant for the log(ka/Km) versus pKa plot with the nativeenzyme corresponds to 13g = -0.3 (Tuil et al., 1994) then this slope of 4.0 would indicatethat the Glul27Ala mutant has a 13g -1.2. This value of the Broensted constant (I3ig =-1.2) is comparable to that for the pKa dependent region of the log(kca) versus pKa plot ,asexpected if they reflect the same step (glycosylation) in catalysis.Although only three substrates were subjected to pre-steady state kinetic analysis,the larger values of the Broensted constant seen for the log(k2) and log(k2IKAJ versus pKaplots (Figure 3-6) (I3ig = -0.8 and -0.7, respectively) with the mutant enzyme relative tothat seen with the native enzyme (13g = -0.3) are consistent with the larger Broensted108constant seen on log(kfKm) with the mutant enzyme. Since kcai/Km provides informationabout the first irreversible step in the enzymatic reaction (Appendix B-2), then these resultssuggest that glycosylation is still the first irreversible step with the Glul27Ala mutant as isthe case with the native enzyme. The larger value of the Broensted constant withGlu 1 27Ala (Iig = -1.2) relative to that for native exoglycanase (pig = -0.3) reflects a muchgreater degree of negative charge accumulation on the oxygen of the departing phenolate atthe glycosylation transition state for Glu l27Ala catalysis. These results suggest that there isless proton donation to the phenolate oxygen at the glycosylation transition state in theGlul27Ala mutant than in the native enzyme, exactly as would be predicted when the acidcatalyst has been removed. These results are therefore completely consistent with the roleof Glu 127 as the acid catalyst.‘-4IIFigure 3-10 Linearfree energy relationship correlating the glycosylation stepsfor thenative and the Glul27Ala mutant.0.0 0.5 1.0 1.5 2.0 2.5log(kcat/Km) (native)1093-6-4 Effect of removal of the acid-base catalyst on the individual steps inthe reactionGlycosylation step: The glycosylation rate constants (k2) (Table 3-4) from thestopped flow analysis for 2,4-DNPC and 3,4-DNPC hydrolysis are reduced 9- and 189-fold, respectively while that for PNPC hydrolysis is reduced 800-fold respectively, uponreplacement of the glutamate residue with an alanine residue. The glycosylation rateconstants for 3,5-DCIPC and 4-CNPC, obtained from kcat values, are reduced 3560- and600-fold, respectively with the Glul27Ala mutant. These results clearly indicate that theeffect of removal of the carboxylate group of Glu 127 on the glycosylation step isdependent on the leaving group ability of the aglycone. Thus, substrates with good leavinggroups such as 2,4-DNPC which require little acid catalytic assistance for departure are notgreatly affected while those substrates with poorer leaving groups such as 3,5-DCIPC aremore severely affected by removal of the carboxylate group. These results are in completeagreement with those previously reported on the glycosylation step of Glul27Ala catalysissince the value of k/Km for 2,4-DNPC was seen to be essentially unaffected by removalof the carboxylate group while that for 4-BrPC was reduced -3 x 105-fold relative to nativeenzyme catalysis. This behaviour of Glul27Ala is entirely consistent with the proposedrole of Glu 127 as the acid catalyst in C. fimi exoglycanase. Similar results have beenreported for Agrobacterium -glucosidase when Glu 170, the proposed acid-based catalyst,was replaced with a glycine residue (Wang et al., 1995).Deglycosylation step: The deglycosylation rate constant (k3) for the cellobiosylenzyme, obtained from keat values (Table 3-3), is seen to be reduced —200 to 600-fold withthe Glul27Ala mutant. Therefore removal of the general base catalyst increases thedeglycosylation transition state energy by 3-4 kcallmole. Interestingly however, thesecondary deuterium kinetic isotope effect measured for 2,4-DNPC with the Glul27Alamutant (kH/kD = 1.08) and with the native enzyme (kH,’kD = 1.10) are comparable,110indicating that there is a similar degree of oxocarbonium ion character at the twodeglycosylation transition states and thus likely a similar degree of C-O bond cleavage.3-7 SummaryAll the data presented on the Glu233Asp and Glul27Ala mutants are consistentwith the roles of Glu 233 and Glu 127 as the catalytic nucleophile and acid-base catalyst,respectively.Pulling the carboxylate group of the catalytic nucleophile 1 A away from thereacting anomeric centre of the sugar by conversion of Glu 233 to an aspartate residue isseen to reduce the rate of the glycosylation step 3-4 x 103-fold and that of thedeglycosylation step 4 x 103-fold relative to native enzyme catalysis. Interestingly, movingthe carboxylate group 1 A away is seen to have a greater destabilising effect on thetransition state for deglycosylation than on that for glycosylation. Presumably this isbecause the transition state for deglycosylation has greater oxocarbonium ion character thanthat for glycosylation.Removal of the carboxylate group of the acid-base catalyst is seen to reduce thedeglycosylation rate 200-300-fold while the pre-steady-state kinetic analysis reveals that theeffect on the glycosylation step is dependent upon the leaving group ability of the aglycone.111CHAPTER IVLABELING STUDIES OF C. FIMI EXOGLYCANASE USING ACTIVESITE-DIRECTED IRREVERSIBLE INACTIVATORS1124-1 IntroductionActive site-directed irreversible inactivators are structural analogues of the normalsubstrates, thus are capable of taking advantage of the binding specificity of the enzymefor its substrates. These inactivators bind within the active site forming a non-covalentenzyme-inactivator complex and are attacked by a nearby nucleophilic residue, resulting inthe formation of a covalent bond between the enzyme and the inactivator, thus inactivation.Active site-directed inactivators have been extensively used in enzymology. Forexample, in the absence of X-ray crystallographic data, irreversible inactivators can providestructural information, as the derivatised active site residue can be identified by using aradiolabeled analogue of the inactivator to identify a labeled, proteolytically derived peptide,which is then sequenced. The function of the labeled residue can be further probed bymutagenic studies.4-1-1 Criteria for active site-directed irreversible inactivatorsIn order to be characterised as an irreversible inactivator specifically targetted to theenzyme active site and capable of being used to identify active site residues, candidatecompounds must satisfy the following list of requirements (Silverman, 1988; Legler, 1990;Wold, 1978).(1) Inactivation should be time-dependent and follow pseudo-first order kinetics.(2) Saturation kinetics should be observed. That is, the rate of inactivation should beproportional to the concentration of the inactivator at low concentrations relative tothe dissociation constant, but independent at high concentrations when there is nofree enzyme in solution.(3) Protection against inactivation using a substrate or competitive inhibitor should beobserved. That is, the rate of inactivation should be lower in the presence of asubstrate or competitive inhibitor than in its absence, thus providing evidence forthe inactivator binding at the active site.113(4) Inactivation should be irreversible. Thus the inactivated enzyme should remaininactive after it is subjected to techniques such as dialysis or gel filtration whichremove excess and loosely bound inactivators.(5) The stoichiometry of the inactivation reaction should be one mole of inactivatorbound per one mole of active site since inactivation of the enzyme specificallyrequires blockage of the active site.(6) The inactivator should be reasonably stable against spontaneous decompositionunder the reaction conditions.(7) The inactivator must have a high non-covalent affinity for the active site, thusensuring that the active site is specifically labeled at low inactivator concentrations.(8) The enzyme-inactivator bond must be stable against denaturation, proteolysis andconditions used for isolation and sequencing of the labeled peptide, thus allowingidentification of the labeled amino acid residue.Two classes of active site-directed irreversible inactivators that satisfy these criteriaare affinity labels and mechanism-based inactivators.4-1-2 Affinity labelsThis class of active site-directed irreversible inactivators constitutes substrateanalogues containing a highly reactive functional group such as isocyanate, epoxide orbromoacetyl groups. However due to their high intrinsic reactivity, affinity labels can reactwith several different residues both within and outside the active site, resulting in nonspecific labeling of the enzyme. This problem can generally be circumvented by inactivatingthe enzyme at low concentrations of the inactivator, thereby taking advantage of the highspecificity of the active site for the inactivator. Alternatively, labeling experiments can beperformed in the absence and in the presence of a competitive inhibitor which protects theactive site from derivatisation by the affinity label. Labeled active site residues can then be114identified by eliminating from consideration residues labeled in the presence of thecompetitive inhibitor.(a) Glycosyl epoxides: Affinity labels have been extensively used to map the active siteof several glycosidases. Examples include the use of N-bromoacetyl glycosylamines(Figure 4-1) to inactivate E. coli 13-galactosidase (Naider et al., 1972), Aspergillus wentii13-glucosidase (Legler, 1977) and Agrobacterium 13-glucosidase (Black et at, 1993) as wellas f3-D-glucosyl isothiocyanate (Figure 4-1) which inactivates sweet almond f3-glucosidase(Shulman et al., 1976). However by far the most extensively used group of affmity labelsfor studying glycosidases are the glycosyl epoxide derivatives (Figure 4-2).11N%C. CH2Br=N-Bromoacetyl 3-D-glycosylamine 3-D-Glucosyl isothiocyanateFigure 4-1 Examples ofaffinity labels.Glycosyl epoxides inactivate glycosidases because when protonated either by theacid catalyst or another residue within close proximity, they are converted to reactivespecies which are susceptible to nucleophilic attack by neighbouring amino acid residues(Figure 4-2). Indeed these compounds have been successfully used to label and identifyactive site residues, the earliest study involving inactivation of hen egg white lysozymewith { 14C1-2,3-epoxy propyl chitobioside. Sequencing of the purified, radiolabeled,proteolytically derived peptide revealed Asp 52 as the modified residue (Eshdat et al.,OHOH1151974) while X-ray crystallographic analysis of the inactivated enzyme complex wasconsistentTHOO\)H- - -Figure 4-2 Mechanismfor inactivation ofglycosidases by glycosyl epoxides.with the role of this residue as the catalytic nucleophile (Moult et al., 1973). Other studiesinvolving the use of glycosyl epoxide inactivators have included those with severalcellulases. For example, both the Bacillus amyloliquefaciens l,3-1,4--D-glucan 4-glucanohydrolase and Bacillus macerans l,3-1,4-3-g1ucanase are inactivated with 3,4-epoxy butyl 13-cellobioside (Hoj et al., 1992; Keitel et al., 1993) while Trichoderma reeseiendoglucanase ifi is inactivated with 4’,5’-epoxy pentyl 3-cellobioside (Macarron et al.,1993). The labeled residues in the Bacillus amyloliquefaciens and Trichoderma reeseiOH116enzymes were identified as Glu 105 and Glu 329, respectively and proposed to be thecatalytic nucleophiles in these enzymes. In the Bacillus macerans enzyme, the labeledresidue was identified as Glu 105 by X-ray crystallographic analysis of the inactivatedenzyme complex and proposed to be a catalytically important residue based on mutagenesisstudies (Keitel et al., 1993).4-1-3 Mechanism-based inactivatorsAnother group of irreversible inactivators designed to specifically bind at the activesite of target enzymes is that of mechanism-based inactivators. These inactivators do notcontain highly reactive functional groups and thus are relatively unreactive structuralanalogues of the substrates. However, they are transformed into the inactivating species viaan enzyme-catalysed process. For an inactivator to be classified as mechanism-based, itmust comply with the following two additional requirements. First, a catalytic step must beshown to be involved in the inactivation process and second, inactivation of the enzymeshould occur prior to release of the activated species. This ensures the identity of thestructure of the inactivated-enzyme complex as well as ensures that the inactivation processoccurs at the active site (Silverman, 1988). As a consequence, mechanism-basedinactivators tend to be more specific for the enzyme active site than affinity labels due totheir inherently lower reactivity and requirement for catalytic activation.(a) Conduritol epoxides: One class of mechanism-based inactivators of glycosidasesthat has been widely used is that of the conduritol epoxides. These epoxides aretetrahydroxycyclohexanes containing an endocyclic epoxide, thus are structural analoguesof the glycone portion of glycosides (Figure 4-3). The electron-withdrawing effects of theneighbouring hydroxyl groups result in greatly reduced reactivity of the epoxide groupagainst acid catalysis and nucleophilic addition, thus rendering conduritol epoxidesrelatively inert to spontaneous hydration. These compounds are therefore good candidates117as mechanism-based inactivators since they will only react with proteins if bound in aposition that permits proton donation by an acidic group to the epoxide oxygen followed bynucleophilic attack at C-i and subsequent formation of a covalent bond (Figure 4-3). Inglycosidases, the residues appropriately located to perform these roles are the acid-catalystand the catalytic nucleophile, the residues normally involved in glycoside hydrolysis.Conduritol epoxide derivatives have been synthesized and shown to be inactivatorsof several glycosidases including Aspergillus wentii f-glucosidase and E. coli -galactosidase (Legler, 1990). In all cases studied, the modified residues were identified ascarboxylate groups and further confirmation of this was obtained by treating the inactivatedenzyme with hydroxylamine (Figure 4-3). The reaction product in all cases was identifiedas inositol produced by trans-opening of the epoxide which is consistent with the existenceof an ester linkage between a carboxylate group of the enzyme and the conduritol epoxideinactivator. The modified carboxylate residue was proposed to be the catalytic nucleophilebased upon the mechanism of enzyme inactivation by conduritol epoxides. For example,with Aspergillus wentii 3-glucosidase, an aspartate residue was labeled using 3H-conduritol cis-B epoxide (Bause & Legler, 1974), treatment of the inactivated enzyme withhydroxylamine produced (+)-chiro-inositol (Legler, 1968) and thus this aspartate residuewas assigned the role of the catalytic nucleophile (Bause & Legler, 1974). This assignmentwas later confirmed when the residue involved in the formation of the 2-deoxy-glucosyl-enzyme intermediate during p-nitrophenyl-2-deoxy-f-D-glucopyranoside hydrolysis wasidentified as the same asparate residue labeled by conduritol cis-B epoxide (Roeser &Legler, 1981). However it should be noted that in other cases, the enzymic residuemodified by conduritol epoxide derivatives is not always the catalytic nucleophile. Forexample, the residue labeled in E. coli -galactosidase by conduritol C epoxide wasidentified as Glu 461, and thus proposed to be the catalytic nucleophile (Herrchen &Legler, 1984). Kinetic analysis of the mutants generated at position 461 of this 1-galactosidase revealed that the kcat values for hydrolysis of 2-nitrophenyl--D-galactoside118were reduced 7500-, 1056-, 2083-, 18- and 3700-fold upon replacement of the glutamatewith aspartate, glycine, glutamine, histidine and lysine, respectively (Cupples et a!., 1990).ASurprisingly, the most conservative change, replacement of the glutamate with aspartate,resulted in the greatest rate reduction while less conservative changes such as thereplacement of the glutamate with histidine resulted in a mutant with significant enzymaticactivity. These results are inconsistent with those obtained from kinetic studies perfomiedon mutants generated at the catalytic nucleophile (Glu 358) of Agrobacterium -glucosidase(Withers et a!., 1992). In the case of the 3-glucosidase, values of keat were seen to beHi HOHHO.cçoI INH2OInositolFigure 4-3 Inactivation ofa glycosidase by a conduritol epoxide and release of inositolupon hydroxylamine treatment.119reduced 2500-fold with the aspartate mutant and greater than 105-fold with the glutaminemutant and thus are consistent with the role of Glu 358 as the catalytic nucleophile. Theseresults prompted a reinvestigation of the identity of the catalytic nucleophile by Gebler andcoworkers using the more specific mechanism-based inactivator, 2,4-dinitrophenyl-2-deoxy-2-fluoro-3-D-galactopyranoside, which functions via formation of a stabilised 2-deoxyfluoroglycosyl-enzyme intermediate (vide infra) (discussed in Chapter I). This studyidentified Glu 537 as the catalytic nucleophile (Gebler et al., 1992). Subsequent kineticstudies of the mutants made at position 537 were entirely consistent with this assignment(Yuan et al., 1994). It has been suggested by Gebler and coworkers (1992) that Glu 461 islikely the acid catalyst, an assignment more consistent with the mutagenesis results.Conduritol C epoxide is thought to have labeled Glu 461 instead of the catalytic nucleophilein E. coli f-galactosidase (Glu 537) because the absence of the C-5 hydroxymethyl groupwhich is present in normal galactoside substrates presumably allows alternative bindingmodes within the active site and thus allows the inactivator to react with other active siteresidues (Withers & Aebersold, 1995).4-2 Background And Objectives Of This ProjectThe aim of this project is to identify active site residues in C. fimi exoglycanase thatare important in either catalysis or substrate binding by using active site directed irreversibleinactivators and a combination of mass spectrometric techniques. This project has been acollaborative effort between our laboratory and that of Dr. Ruedi Aebersold where the massspectrometric analysis was performed.Previously, 2,4-dinitrophenyl 2-deoxy-2-fluoro-3-D-glycosides have been shownto be good mechanism-based inactivators of several “retaining” glycosidases, specificallyderivatising the catalytic nucleophile (discussed in Chapter I). They have therefore beenused to label and identify this key residue. Both 2,4-dinitrophenyl 2-deoxy-2-fluoro-3-D-glucoside and its cellobioside analogue are inactivators of C. fimi exoglycanase,120inactivating in a time dependent manner according to pseudo-first order kinetics (Tull &Withers, 1994; McCarter et al., 1993; Tull et al., 1991). The 2-deoxyfluoroglucosyl-inactivated exoglycanase is stable, with ti — 800 hours, thereby allowing identification ofthe catalytic nucleophile involved in glucoside hydrolysis as Glu 233 (Tull et al., 1991).Similarly, the 2-deoxyfluorocellobiosyl-exoglycanase intermediate is seen to be stable (t11300 hours) and thus should allow for the identification of the catalytic nucleophileinvolved in cellobioside hydrolysis.Hydration of glycals by glycosidases involves the same catalytic residues as thoseparticipating in normal glycoside hydrolysis (discussed in Chapters I & II). Recall that thereaction involves protonation of the glycal by either the acid catalyst or the catalyticnucleophile, attack at the C-i position of the glycal by the catalytic nucleophile, andformation of a 2-deoxyglycosyl-enzyme intermediate. In a second step the intermediate ishydrolysed, releasing a 2-deoxy-glycose product and free enzyme. If the rate of hydrolysisof the intermediate is lower than that of its formation, then the enzyme accumulates as theintermediate. Further, if hydrolysis is sufficiently slow, this may provide a means oftrapping the catalytic nucleophile. Identification of this 2dCb-amino acid residue cantherefore confirm the identity of the catalytic nucleophile in C.fimi exoglycanase.Previously, the affinity label, N-bromoacetyl cellobiosylamine, was shown toinactivate C.fimi exoglycanase in a time-dependent manner according to pseudo-first orderkinetics with inactivation kinetic parameters K, = 9.1 mM, k• = 0.083 min’ and k/K1 = 9.1x i0 min1M (Appendix A-8, Figure A-8-3, Black et al., 1993). Since N-bromoacetylglycosylamines can react with any suitable nucleophile within range due to their highintrinsic reactivity, then the cellobiosyl analogue of this inactivator has the potential toidentify residues important in either catalysis or substrate binding which have not beenidentified by other methods.Therefore, the objectives of this project are to use the mechanism-basedinactivators, 2,4-dinitrophenyl 2-deoxy-2-fluoro-3-cellobioside (2F-DNPC) and cellobial121to identify the catalytic nucleophile involved in C. fimi exoglycanase-catalyseci hydrolysisof cellobiosides as well as to use the affinity label, N-bromoacetyl cellobiosylamine, toidentify other active site residues.4-3 Results And Discussion4-3-1 Labeling studies of C. fimi exoglycanase using 2,4-dinitrophenyl 2-deoxy-2-fluoro- 13-D-cellobioside(a) Stoichiometry of inactivationA sample of C.fimi exoglycanase were inactivated with 2,4-dinitrophenyl 2-deoxy-2-fluoro--D-cellobioside and then subjected to electrospray ionisation mass spectrometricanalysis, along with a sample of unlabeled exoglycanase. Masses of 47,114 +1- 7 and47,441 +/- 7 Da were recorded and correspond to the molecular weights of unlabeledexoglycanase and 2-fluoro-cellobiosyl-exoglycanase, respectively (Figure 4-4). The massdifference between the 2-fluoro-cellobiosyl-exoglycanase sample and the unlabeled enzymecorresponds to 327 Da. Since the mass of the 2-deoxyfluoro-cellobiosyl moiety is 327 Da,then this result suggests that one mole of 2,4-dinitrophenyl 2-deoxy-2-fluoro-3-D-cellobioside inactivates one mole of exoglycanase.(b) Strategy for identification of the 2-deoxyfluorocellobiosyl-labeled residue.The approach that will be used to identify the 2-deoxyfluoro-cellobiosyl-modifledresidue is shown schematically in Figure 4-5. It involves labeling the exoglycanase with theinactivator, proteolytic digestion of the labeled enzyme, identification and purification of thelabeled peptides by high-performance liquid chromatography-electrospray ionisationtandem mass spectrometry (HPLC-ESI-MS/MS) and subsequent sequencing to identify thelabeled peptide.47,441Figure 4-4 Electrospray ionisation mass spectra showing stoichiometry ofinactivation of C. fimi exoglycanase by 2F-DNPC. Reconstructed massspectra ofunlabeled enzyme (upper) and labeled enzyme (lower).122.tISIABMolecular Weight100•75I£5025a‘46,600 47,000 47,400 47,800Molecular WeightRelativeIntensity- Purify &SequencePeptide123Figure 4-5 Scheme of the method used to identify the residue in C. fimi exoglycanaseInactivation of enzymeSugarLabeled-glycosidasePeptic digestionSugar\HPLC-ESMSTimeESMS/MS(neutral loss)labeled peptideTimelabeled by 2F-DNPC and cellobial.124(c) Identification of the residue modified by 2,4-dinitrophenyl 2-deoxy-2-fl uo ro- - cellobiosi deA sample of C. JImi exoglycanase was completely inactivated with 2F-DNPC andsubjected to peptic proteolysis, along with a sample of unlabeled enzyme. The resultingpeptide mixtures were individually separated by RP-HPLC using the electrosprayionisation mass spectrometer as a detector, scanning in the normal MS mode as describedin Chapter V by Dr. Shichang Miao in Dr. Ruedi Aebersold’s laboratory. A large numberof peaks is observed in the total ion chromatogram (TIC), consistent with the presence ofmany peptides in the mixture (Figure 4-6, upper spectrum). The 2FCb-labeled peptide inthe mixture was then identified in a second run using the tandem mass spectrometerscanned in the neutral loss mode (Figure 4-6, middle spectrum). In this mode the peptideions are subjected to limited fragmentation by collision with argon-nitrogen mixture in acollision cell. This results in the loss of a neutral sugar species from the peptide, leaving thepeptide with its original charge, but a decreased mass. In the case of the 2FCb-labeledsample the neutral species is expected to be the 2FCb-moiety since the ester linkagebetween the 2FCb-moiety and the peptide is one of the more labile linkages present andthus readily undergoes homolytic cleavage. Indeed, the collision conditions used weresufficient to break the ester bond but not generally the amide bonds of the peptidebackbone. This resulting peptide is identified when the two quadrupoles are scanned in alinked manner such that the peptide ions differing in mlz by the mass of the neutralspecies can pass through both quadrupoles and be detected. In some cases it may benecessary to scan for mlz differences of one half or one third the mass of the neutral speciesas the peptide may be doubly or triply charged. When the tandem mass spectrometer wasscanned in the neutral loss mode, searching for a mass loss corresponding to the 2FCb-moiety of m/z 327 Da, no signal was detected. However, when scanned for a mass loss125of m/z = 163.5 Da corresponding to the elimination of the 2FCb-moiety from the doublycharged peptide species, a single peak of m/z = 529 Da was observed in the TIC (Figure 4-6, lower spectrum). Similar tandem mass spectrometric analysis of a peptic digest ofunlabeled exoglycanase revealed no peaks in the TIC, suggesting that the signal at m/z =529 Da in the TIC of the labeled sample corresponded to the derivatised peptide.The identity of this peptide can be easily probed by calculation of its mass. Thepeptide observed of m/z= 529 Da for the doubly charged species (Figure 4-6, lowerspectrum) has molecular weight 1056 Da [(529 x 2) - 2H]. Since the mass of the label is327 Da the molecular weight of the unlabeled peptide must be 730 Da (1056 - 327 + 1H).Clues to the identity of this peptide were then obtained by searching the amino acidsequence of this exoglycanase (O,Neill et al., 1986) for all possible peptides of mass 730Da. Nine peptides of mass 730.4 ± 0.1 Da could be identified, one of which contains theactive site peptide (Val-Arg-fle-Thr-Glu-Leu) previously identified by the standard methodusing thtiated 2F-DNPG (Tull et al., 1991).The 2FCb-modified peptide was isolated from the peptic mixture by RP-HPLCusing a post column flow splitter to divert 85% of the eluent to a fraction collector while theremaining 15% was introduced into the mass spectrometer, scanning in the neutral lossmode. The purified 2FCb-peptide was sequenced using the Edman degradation. Thepeptide was identified as Val-Arg-Ile-Thr-Glu-Leu and the labeled residue as a glutamateresidue. Sequence alignment of this peptide with the known amino acid sequence of C. fimiexoglycanase revealed Glu 233 as the labeled residue. Thus, the catalytic nucleophileinvolved in cellobioside hydrolysis, Glu 233, is exactly the same as that involved inglucoside hydrolysis.126‘75E£ 50Ce 10 12 14 16 18 20mtha‘75so2508 10 12 14 16 18 20mm529Figure 4-6100-75J.E 50C)C)0400 600 800 1000 1200mlzElectrospray ionisation mass spectra ofa peptic digest of C.fimiexoglycanase inactivated by 2F-DNPC. (Upper) TIC of digest in nor,nalMS mode, (middle) TIC of digest in neutral loss mode and (lower) massspectrum ofpeptide in the middle panel.1274-3-2 Labeling studies of C. fimi exoglycanase using cellobial(a) Stoichiometry of inactivationSamples of cellobial-inactivated and untreated C. fimi exoglycanase were subjectedto electrospray ionisation mass spectrometric analysis. Molecular weights of 47,118 +1- 7Da and 47,423 +1- 7 Da corresponding to the unlabeled and labeled enzyme speciesrespectively, were recorded (Figure 4-7). This mass difference of 305 Da between the2dCb-exoglycanase sample and the unlabeled enzyme corresponds well with the mass of asingle 2dCb moiety (308 Da) and thus is consistent with a stoichiometry of inactivation ofone mole of 2dCb moiety per mole of enzyme.47,11846,600 47,000 47,400 47,SG0Molecular Weight100• 47,42311___46,600 47,150Molecular WeightFigure 4-7 Electrospray ionisation mass spectra showing the stoichiometry ofinactivation of C.fimi exoglycanase by cellobial. Reconstructed massspectra ofunlabeled enzyme (upper) and labeled enzyme (lower).128(b) Identification of the residue modified by cellobialFurther confirmation of Glu 233 as the catalytic nucleophile involved in cellobiosidehydrolysis could be obtained using cellobial since it is also expected to derivatise thecatalytic nucleophile and since the 2dCb-enzyme intermediate is stable with t1,2 23 hours.This approach has been used previously to identify the catalytic nucleophile in Aspergilluswentii 3-glucosidase. D-14C-glucal was seen to label the same aspartate residue inAspergillus wentli 3-glucosidase as that previously labeled by conduritol B epoxide andthus this aspartate residue has been assigned the role of the catalytic nucleophile (Legler etal., 1979). A sample of C. fimi exoglycanase was inactivated with cellobial, the labeledexoglycanase digested with pepsin and the digest subjected to neutral loss tandem massspectrometric analysis searching for a mass loss of 308 Da corresponding to the mass of a2dCb-moiety. The single peak observed in the neutral loss TIC at mlz = 1039 Dacorresponded to the singly charged labeled peptide (Figure 4-8), hence the labeled peptidehas molecular weight 1038 Da and the unlabeled peptide has molecular weight 730 Da(1038- 308 + 1H), likely corresponding to the Val-Arg-Ile-Thr-Glu-Leu peptide. Thissuggests that Glu 233 is the catalytic nucleophile involved in cellobial hydration and incellobioside hydrolysis.Indeed, the catalytic nucleophiles in other glycosidases have been identified usingthis new neutral loss tandem mass spectrometric method. For example, the labeled peptidesidentified in 2F-DPNC-inactivated Clostridium thermocellum endoglucanase C and 2F-DNPG-inactivated Agrobacterium 3-glucosidase by this mass spectrometric technique wereexactly the same as those identified by the standard radiolabeling technique (Tull et al.,1995). These results therefore further confirmed Glu 280 (Wang et al., 1993) and Glu 358(Withers et al., 1990) as the catalytic nucleophiles in the endoglucanase and the f3-glucosidase, respectively. In addition, this mass spectrometric approach was used toidentify, for the first time, Glu 340 and Glu 78 as the catalytic nucleophiles in human129glucocerebrosidase and Bacillus subtilis xylanase respectively, using the appropriate 2-deoxy-2-fluoro-glycosides (Miao et al 1994a; Miao et aL, 1994b).103810025______600 800 1000 1200 1400m/zFigure 4-8 Neutral loss tandem electrospray ionisation mass spectrum ofa peptic digestof C. fimi exoglycanase inactivated by cellobial.4-3-3 Labeling studies of C. fimi exoglycanase with N-bromoacetylcellobiosylamine(a) Protection against inactivationSamples of C.fimi exoglycanase were incubated in the presence of N-bromoacetylcellobiosylamine and either 0 or 9.7 mM of the competitive inhibitor, benzyl 4-O-((3-D-130glucopyranosyl)-1-thio-f-D-xy1opyranoside (Figure 4-9) (BGTX, K1 = 3.0 mM). Aliquotswere removed at different time intervals and assayed for exoglycanase activity usingsaturating concentrations of 2,4-DNPC (Appendix A-8, Figure A-8-3). Values of kobs mthe absence and presence of BGTX were 6.0 x 10-2 and 1.2 x 102 mirr1, respectively asexpected if BGTX and N-bromoacetyl cellobiosylamine are competing for the same site.These results suggest that inactivation of C. Jlmi exoglycanase by N-bromoacetylcellobiosylamine involves blockage of the active site and therefore likely involvesderivatisation of a residue present at the active site.Figure 4-9 Structure ofbenzyl 4-O-(/3-D-glucopyranosyl)-1 -thio-/3-D-xylopyranoside.(b) Stoichiometry of inactivationAs affmity labels, N-bromoacetyl glycosylamines are relatively reactive compoundscapable of modifying glycosidases (Figure 4-10) at more than one site and indeed suchresults have been reported. For example, electrospray ionisation mass spectrometricanalysis of Agrobacteriuin -glucosidase inactivated with N-bromoacetyl glucosylaminerevealed incorporation of up to three N-acetyl glucosylamine moieties into the enzymewhile radiolabeling studies of N-bromoacetyl galactosylamine inactivation of E. coli [-galactosidase revealed incorporation of two moles of N-acetyl galactosylamine per activesite (Yariv et al., 1971). However, in both cases the inactivation kinetics suggested asimple pseudo-first order process (Black et al., 1993).GTBX131T(NuTNu,CH2CII0Figure 4-10 Mechanism of inactivation of C.fimi exoglycanase by N-bromoacetylcellobiosylamine.The stoichiometry of inactivation of C. fimi exoglycanase by N-bromoacetylcellobiosylamine was investigated as described for 2F-DNPC inactivation. The molecularweight of the unlabeled exoglycanase sample was 47,118 +1- 7 Da while that for the labeledsample was 47,496 +1- 5 Da (Figure 4-11). This mass difference of 378 Da correspondswell with the mass increase of 382 Da expected if the exoglycanase was derivatised by asingle N-acetyl cellobiosylamine moiety. The very small peak observed at 47,873corresponds to an enzyme species in which two N-acetyl cellobiosylamine moieties havebound, however this could be suppressed by shortening the incubation time. Thisinactivation stoichiometry of one mole of N-acetyl cellobiosylamine incorporated per moleOHOHOHOH132of exoglycanase, together with the results from the protection study suggest that Nbromoacetyl cellobiosylamine is a good candidate affinity label to be used to identifyimportant active site residues of C. fimi exoglycanase.(c) Strategy for identification of N-acetyl cellobiosylamine-labeledresidueIdentification of the residue modified by N-bromoacetyl cellobiosylamine, asshown schematically in Figure 4-12, involved comparative LC-MS analysis of proteolyticdigests derived from labeled and unlabeled exoglycanase. The labeled peptide will beidentified by search the chromatogram of the labeled sample looking for a peptide ion that isabsent in the unlabeled sample and that differs from a specific peptide in the unlabeledsample by the mass of the inactivator.The labeled residue in the isolated peptide will thenidentified by first sequencing the peptide using a novel 311 PITC (phenyl isothiocyanatederivative) sequencing chemistry (Hess et al., 1995; Bures et al., 1995) and then by usingtandem mass spectromeiry to structurally characterise the released 311 PTH(phenylthiohydantoin) amino acid derivative. Sequence alignment of the peptide with theknown amino acid sequence of this exoglycanase will be used to detemñne the exactresidue labeled in the enzyme.1Ce133Figure 4-11 Electrospray ionisation mass spectra showing the stoichiometry ofinactivation of C. fimi exoglycanase by N-bromoacetyl cellobiosylamine.Reconstructed mass spectra of unlabeled enzyme (upper) and labeledenzyme (lower).1O(Moecuar WeightMolecular Weight134IUnlabeled enzyme Peptic digestion_____— \ \Labeled enzymeunlabeledRelativepetideintensityIdentification of labeled peptideby comparative HPLC-ESI-MSRelativeintensityIsolation of labeled peptide byHPLC-ESI-MS and chemicalpeptide sequencingA12rA456Labeled peptidesequenceIdentification of labeledresidue X by ESI-MS/MSFigure 4-12 Scheme of the method used to identify the residue in C.fimi exoglycanaseRelativet \ \HPLC-ESI-MSTime Timelabeled by N-bromoaceryl cellobiosylamine.135(d) Identification of peptide modified by N-bromoacetylcellobiosylamineA sample of the exoglycanase from C. Jimi was inactivated with N-bromoacetylcellobiosylamine and subjected to proteolytic digestion with pepsin, yielding a mixture ofpeptides. A sample of unlabeled exoglycanase was also subjected to peptic digestion usingsimilar conditions. The mixtures of peptides from the labeled and unlabeled samples wereindividually separated by RP-HPLC using the electrospray ionisation mass spectrometer asa detector. When the spectrometer is scanned in the normal MS mode, the total ionchromatogram (TIC) of the labeled and unlabeled exoglycanase peptic digests displayed alarge number of peaks which arise from every peptide in the mixture (Figure 4-13, upperspectrum). No significant, reproducible differences could be detected between thechromatograms of the two digests. The labeled peptide was identified by using this massinformation. It was expected that the masses of the relevant active site peptide in the labeledand unlabeled enzyme digests must differ by the mass of the N-acetyl cellobiosylaminemoiety (382 Da). Therefore, the masses of the peptides under each peak in the labeledsample were compared with the masses of the peptides from the corresponding region inthe TIC of the unlabeled exoglycanase sample. All the peptides present in the labeledsample were present in the unlabeled sample at comparable retention times, with theexception of a single peptide of mass 1028 Da (Figure 4-13, lower spectrum) which wasonly detected in the digest from the inhibited enzyme. The TIC of the unlabeled sample wasthen searched for a possible counterpart unlabeled peptide of mass 646 Da (Figure 4-13,middle spectrum) corresponding to the mass difference between this peptide of mass 1028Da and the N-acetyl cellobiosylamine label of mass 382 Da. Indeed such a peptide wasobserved in the unlabeled sample. This peptide of mass 646 Da was also present in the TICof the labeled sample (Figure 4-13, lower spectrum) at a retention time similar to that of the646 Da peptide in the unlabeled sample although its intensity was reduced compared to that136of the corresponding peptide in the unlabeled sample. These results suggested that thepeptide of mass 1028 Da was likely the modified peptide and that the appearance of thepeptide of mass 646 Da in the labeled sample was due to incomplete inactivation of theexoglycanase and/or cleavage of the N-acetyl cellobiosylamine from the labeled peptideduring proteolysis or chromatography. The labeled peptide was purified to homogeneityby RP-HPLC, monitored by ESI-MS. Candidate peptides were then identified byinspection of the amino acid sequence of the exoglycanase and searching for all possiblepeptides with this mass. Nine possible peptides of mass 646 +1- 1 were identified withinthe C.JImi exoglucanase sequence (O’Neill et al., 1986).(e) Identification of the labeled residueTo unambiguously assign the labeled peptide within the exoglycanase amino acidsequence and to characterise the labeled residue, the isolated peptide was sequenced usingthe novel 311 PITC (4-(3-pyridinylmethyl-aminocarboxypropyl) phenyl isothiocyanate,311 = molecular weight of reagent) sequencing reagent (Hess et al., 1995) by Mr. DavidChow in Dr. Reudi Aebersold’s laboratory. This automated sequencing involved attachingthe isolated peptide to a solid support and stepwise derivatisation of the N-terminal aminoacid by the 311 P1TC reagent. The 311 PTH- (phenylthiohydantoin-) derivatised residuesare sequentially separated by RP-HPLC and analysed by on-line electrospray ionisationmass spectrometry. Thus, the 311 PTh-residues are identified both by their mass and bythe chromatographic retention time.137IOC755025Time (mm)35 498.04.0o.0CnzFigure 4-13 Electrospray ionisation mass spectrum of a peptic digest of C.fimiexoglycanase inactivated by N-bromoaceiyl cellobiosylamine, TIC in thenormal MS mode (upper). Comparative ESI-MS ofa section (retention time= 24 -25 mm) of the TIC of the unlabeled (middle) and labeled (lower) C.fimi exoglycanase peptic digest.138In a first sequencing experiment, the amino acid derivatives in each step weredetected by ESI-MS, scanned over a range of 300 - 850 Da. The peptide sequence wasidentified as Asp-Val-Val-Asn-X-Ala. The residue X observed in cycle 5 denotes an ion ofm/z = 822 Da (Figure 4-14) which does not correspond to the modified 311 PTH derivativeof a known amino acid. However, this mass corresponded exactly with the expected massfor the 311 PTH of the N-acetyl cellobiosylamine ester of a glutamic acid residue shown inFigure 4-15. The other signals in the chromatogram represented background levels ofcontaminating amino acid derivatives or other impurities which coeluted with the in/z = 822Da species. To further characterise the derivative released in cycle 5, another aliquot of theisolated, labeled peptide was sequenced and the released residue with mass 822 Da wassubjected to collision-induced fragmentation in the mass spectrometer. Figure 4-15 showsthe generated daughter ions and an interpretation of the mass spectrum. Since some of thedaughter ions are known to be derived from the sequencing reagent and the structure of theinactivator was known, it was relatively simple to interpret the tandem mass spectrum.Sequence alignment of the identified peptide with the known amino acid sequenceof the exoglycanase indicates that the modified residue corresponds to Glu 127. Thisresidue has previously been identified as the acid-base catalyst in this exoglycanase throughmutation of conserved glutamate and aspartate residues and kinetic analysis of thesemutants (MacLeod et al., 1994). These results clearly illustrate the usefulness of Nbromoacetyl glycosylamine affinity labels in identifying important catalytic residues inglycosidases.139Figure 4-14VVVElectrospray ionisation mass spectrwn of the 311 PTH derivative released incycleS during sequencing of the labeled peptide.659497Figure 4-15 Tandem electrospray ionisation mass spectrum of the 311 PTH derivativereleased in cycles during sequencing of the labeledpeptide.ui/zWV109.025 ¶77.51lk0.100 200497.0539.0LMHr.822312.5180.0 121.5561.5423.0niikiiij LIIL L300 400 500 500 700 500TI’LIZ1404-3-4 Comparison of the radiolabeled and the mass spectrometricapproaches used to identify labeled residues in glycosidasesTraditionally, identification of the labeled residues in enzymes using irreversibleactive site-directed inactivators involves synthesis of a radiolabeled analogue of theinactivator which is then used to covalently derivatise the target enzyme. The radiolabeled,inactivated enzyme is then proteolysed, generating a mixture of peptides. The radiolabeledpeptide is isolated by HPLC and the purified peptide sequenced by the Edman degradation,thus revealing the identity of the modified peptide. Although this method has been usedsuccessfully, its application is limited by the complex syntheses frequently required for theintroduction of a radioisotope into the inactivator. In addition this method is timeconsuming, technically demanding and requires micromole amounts of the labeled enzyme.The mass spectrometric approach presented and discussed in the previous sectionscircumvents the limitations and difficulties associated with this standard radioactiveapproach. This new approach involves modification of the enzyme with a non-radioactiveactive site-directed inactivator, proteolytic digestion of the labeled enzyme, identificationand isolation of the purified labeled peptide by RP-HPLC electrospray ionisation massspectrometry and sequencing the purified peptide either by the standard Edmandegradation, by tandem mass spectrometry or by a novel sequencing approach whichcouples peptide sequencing chemistry to electrospray ionisation mass spectrometiy. Thelabeled peptide in the mixture is identified based on any one or a combination of thefollowing criteria. First, the presence of the label on the peptide can result in the peptidehaving a different retention time on the RP-HPLC column compared to the correspondingunlabeled peptide. The labeled peptide can then be identified by screening the HPLC-MSdata from the unlabeled and labeled samples looking for the disappearance of a specificpeptide ion within a selected time window upon inactivation of the enzyme (Withers &Aebersold, 1995). Second, labeling results in a mass increase of the peptide compared to141the unlabeled counterpart. This mass shift is then used, as in the case of N-bromoacetylcellobiosylamine-inactivated C. fimi exoglycanase, to identify the labeled peptide bysearching the HPLC-MS datasets looking for a new peptide ion in the labeled sample butabsent in the unlabeled sample that differs from a specific peptide ion in the unlabeledsample by the mass of the label. Finally, if a linkage within the labeled peptide, other thanthe amide bonds, is susceptible to collision-induced fragmentation (e.g. ester bond) andresults in the elimination of a neutral species of reproducible and predictable mass uponlimited fragmentation as was observed with the 2FCb- and 2dCb-C. fimi exoglycanasesamples, then the labeled peptide can be identified using the tandem mass spectrometer withthe second quadrupole offset from the first by the mass of the neutral species.A potential problem associated with the neutral loss tandem mass spectrometricapproach is cleavage of the amide bonds, releasing amino acids of the same m/z value asthat of the label which results in significant background signal. However, this problem maybe circumvented by performing a similar mass spectrometric analysis on a control sampleof unlabeled enzyme. For example, neutral loss tandem mass spectrometric analysis of a 2-deoxy-2-fluoroglucosyl Agrobacterium 3-glucosidase digest, searching for a mass loss ofm/z = 165 Da, revealed several peaks in the chromatogram (Tull et al., 1995). Similaranalysis of an unlabeled enzyme sample yielded an identical chromatogram with theexception of the absence of two strong signals. These two signals correspond to two 2-deoxy-2-fluoro-glucosyl-peptides that have lost the label while the other signals correspondto unlabeled peptides that underwent fragmentations resulting in loss of a neutral species ofmlz = 165 Da. In this case the most likely candidate for the neutral species lost in thecontrol was a phenylalanyl residue which has this identical m/z value.In contrast to the standard, radioactive approach, the mass spectrometric approachis rapid, non-radioisotopic, sensitive and only requires picomole amounts of labeledenzyme. The generality of the neutral loss approach has been demonstrated with differenttypes of glycosidase using mechanism-based inactivators (full et al., 1995; Miao et al.,1421994a; Miao et aL, 1994b). These mass spectromethc approaches offer an alternative to thestandard radioactive method.4-4 SummaryThe mechanism-based inactivators, 2F-DNPC and cellobial were used to identifythe catalytic nucleophile involved in C. fImi exoglycanase-catalysed hydrolysis ofcellobiosides as Glu 233. This residue has been identified previously as the catalyticnucleophile involved in glucoside hydrolysis. Identification of the labeled residue involvedinactivation of the enzyme with non-radioactive inactivator and proteolytic digestion of thelabeled enzyme. The 2FCb- and 2dCb-peptide were selectively identified by collision-induced loss of a neutral fragment of known mass (2FCb or 2dCb) from the labeledpeptide.The modified residue was identified by sequencing the isolated 2FCb-peptide bystandard Edman degradation.The N-acetyl cellobiosylarnine-labeled residue was identified as Glu 127, the acid-base catalyst. This involved labeling the enzyme with the non-radioactive inactivator, pepticdigestion of the labeled and unlabeled enzyme and ESI-MS analysis of the digests. The Nacetyl cellobiosylamine peptide was identified by comparative peptide mapping of thedigests. The labeled residue was identified using a new protein sequencing chemistrycoupled to ESI-MS and structurally characterised by tandem mass spectrometry.These results illustrate the usefulness of active site directed irreversible inactivatorsin identifying important catalytic residues in glycosidases. Further, these results show thatthese new mass spectrometric approaches present an alternative to the standard radioactiveapproach used to identify labeled residues within enzymes.143CHAPTER VMATERIALS AND METHODS1445-1 Synthesis5-1-1 General procedures and materialsMelting points (m.p.) were determined using a Laboratory Devices Mel-Temp Umelting point apparatus and are uncorrected.1H Nuclear magnetic resonance (NMR) spectroscopy was performed on either aVarian XL-300 spectrometer at 300 MHz or a Bruker WH-400 operating at 400 MHz.Chemical shifts given are on the 6 scale, using tetramethylsilane (TMS) as externalreference (6=0.00 ppm) for samples in CDC13 or 2,2-dimethyl-2-silapentane-5-sulphonate(8 =0.00 ppm) as an external reference for samples in D20.19F NMR spectra were recorded on a Bruker AC-200 spectrometer at a fieldstrength of 188 MHz. Chemical shifts are reported relative to CFC13 (6 = 0.00 ppm) andwere referenced against external trifluoroacetic acid (6 = -76.53 ppm). Signals upfield ofCFC13 are assigned negative values.Micro-analyses were performed by Mr. Peter Borda in the Micro-analyticalLaboratory, Department of Chemistry, University of British Columbia, Vancouver.Solvents and reagents used were either reagent, certified or spectral grade. Drysolvents were prepared as follows. Methanol was distilled from magnesium turnings in thepresence of iodine, acetyl chloride was refluxed over phosphorus pentachioride and thendistilled.Thin layer chromatography (TLC) was performed using aluminium-backed MerckKieselgel 60F-254 analytical plates. Generally, acetylated compounds were run in ethylacetate-hexanes mixtures (1:1) while deacetylated compounds were run in ethyl acetate-methanol-water (7:2:1). Compounds were visualized by employing UV light and/orcharring with 10% sulphuric acid in methanol. Preparative TLC was performed on a Model7924T chromatotron using 1 mm and 4 mm silica gel plates prepared from either silicaType H (10-40 microns) and calcium sulphate (1/2-hydrate) or silica gel 60 PF254145containing gypsun. Flash column chromatography was performed according to Clark-Stilland coworkers (1977) using Kieselgel 60 (230-400 mesh) silica gel.The following compounds were prepared and generously provided by co-workersin this laboratory. 2,4-Dinitrophenyl -D-glucopyranoside was prepared by Dr. MarkNamchuk while 2,4-dinitrophenyl- (1-2H} -13-D- and 4-nitrophenyl- { 1-2H}-3-D-glucopyranosides were synthesized by Ms. Julie Kempton. 3,4-Dinitrophenyl-, 2,5-dinitrophenyl-, 2-nitrophenyl- and 4-nitrophenyl--xylobiosides as well as 2,4-dinitrophenyl-2-deoxy-2-fluoro-3-xylobioside were synthesized by Dr. Lothar Ziser.Similarly, 3,4-dinitrophenyl-, 2,5-dinitrophenyl-, 2-nitrophenyl- and 4-nitrophenyl- 13-D-xylopyranosides were synthesized by Dr. Lothar Ziser. 4-Cyanophenyl 13-D-xylopyranoside was prepared in collaboration with Ms. Ika Setyawati. Benzyl 4-O-(13-D-glucopyranosyl)-1-thio-f3-D-xylopyranoside was prepared by Mr. Rajpal Chandi. 2,4-Dinitrophenyl-, 3 ,4-dinitrophenyl-, 3 ,5-dichlorophenyl-, 3-nitrophenyl-, and 4-cyanophenyl-f3-cellobioside as well as 2,4-dinitrophenyl-2-deoxy-2-fluoro f3-cellobiosidewere prepared for my M. Sc. thesis and their syntheses have been described (Tull &Withers, 1994).5-1-2 General compounds2, 3, 6, 2’, 3’, 4’, 6’-Hepta-O-acetyl-a-cellobiosyl bromide (1)Compound 1 was synthesized by the method of (Fischer & Zemplen, 1910).Cellobiose octa-O-acetate (10 g, 14.75 mmol) was dissolved in CHCI3 (50 mL), glacialacetic acid (350 mL) and 45% HBr-glacial acetic acid (15 mL) were added and the mixturestirred at 15 °C overnight. The reaction mixture was poured over ice-water and the productextracted with CHC13. The CHC13 layer was washed (3 times) with water, washed (2times) with saturated aqueous sodium bicarbonate, dried over MgSO4 and evapourated invacuo to leave an orange oil. The product (1) was crystallised from CHC13-diethyl ether146yielding white needle-like crystals (8.10 g, 11.59 mmol, 79%). M. p. 169-170 °C (lit. m.p. 180 °C) (Fischer & Zemplen, 1910).2, 3, 6, 2’, 3’, 4’, 6’-Hepta-O-aceiyl-a-{1-H}-cellobiosyl bromide (2)2, 3, 6, 2’, 3’, 4’, 6’-Hepta-O-acetyl-D-cellobiose (4.8 g, 7.55 mmol) dissolved inDMSO (16 mL) and acetic anhydride (10 mL) was stirred at room temperature overnight.The mixture was poured over water, centrifuged and the solvent decanted leaving acolourless syrup. The syrup was washed in this way 10 times and then dried overnight.The syrup was dissolved in TIIF (10 mL), sodium borodeuteride (190 mg) in D20 (1 mL)was added and the reaction allowed to proceed at room temperature for 2 hours. Thesolvent was evapourated in vacuo leaving a colourless syrup. 2, 3, 6, 2’, 3’, 4’, 6’-Hepta-O-acetyl-[1-2H]-cellobiose was purified from the contaminating protio-cellobioseoctaacetate which arose from incomplete oxidation, by flash column chromatography usingethyl acetate-hexanes (2:1). The purified deuterio-hemiacetal was dissolved in THF (5 mL),sodium borodeuteride (30 mg) in D20 (0.2 mL) and acetic anhydride added and thereaction mixture stirred at room temperature overnight. After concentration, the remainingsyrup was dissolved in CHC13,washed with water (2 x 50 mL), dried over MgSO4and thesolvent evapourated in vacuo. 1, 2, 3, 6, 2’, 3’, 4’, 6’-Octa-O-acetyl 3-[1-2H1-cellobiosewas crystallised from ethanol to yield a white solid (900 mg, 1.33 mmol, 18 %). M. p.197-199 °C. 1H NMR (CDC13): 5.45 (dd, 1 H, J23 10 Hz, J3,4 9 Hz, H-3), 5.15 (t, 1H, J,3’ 9 Hz, J3’’ 9 Hz, H-3’), 5.10 (t, 1 H, Jy,4 9 Hz, J4 9 Hz, H-4’), 5.00 (d, 1 H,J2,3 10 Hz, H-2), 4.95 (dd, 1 H, J1’2 8 Hz, J’,’ 9 Hz, H-2’), 4.55 (d, 1 H, Ji’,’ 8 HzH-i’), 4.50 (dd, 1 H, J56b 2 Hz, 6a6b 12 Hz, H-6b), 4.40 (dd, 1 H, 35’,6b’ 4 Hz, 6a’6b’12 Hz, H6b’), 4.15 (dd, 1 H, 5,6a 4 Hz, 36a,6b 12 Hz, H-6a), 4.05 (dd, 1 H, 5’6a’ 2Hz, J6a’,6b’ 12 Hz, H-6a’), 4.00 (ddd, 1 H, J45 9 Hz, 35,6a 4 Hz, 5,6b 2 Hz, H-5), 3.80(t, 1 H, J3,4 9 Hz, J4,5 9 Hz, H-4), 3.65 (ddd, 1 H, J4•5’ 9 Hz, 5’,6a 2 Hz, 56b 4 Hz,H-5’), 1.95-2.15 (8 s, 24 H, 8 (OAc)).147Deuterio-acetobromocellobiose (2) was then synthesised from deuterio-cellobioseoctaacetate by exactly the same method as the protio-compound (660 mg, 0.95 mmol,7 1%). M. p. 169-17 1 °C.5-1-3{1H}-aryl f3-celiobiosides4 “-Bromophenyl 13-cellobioside (4-BrPC)(3)Hepta-O-acetyl-c-ce1lobiosy1 bromide (1) (1.5 g, 2.21 mmol) dissolved in acetone(20 mL) was added to the 4-bromophenol dissolved in 1M NaOH (5 mL) and stirred atroom temperature for 48 hours. The solvent was then evapourated in vacuo and theremaining syrup diluted with water and extracted with CHC13. The organic phase waswashed (3 x 50 mL) with saturated sodium bicarbonate, dried over MgSO4,gravity filteredand evapourated in vacuo. The acetylated product was crystallised from ethyl acetate-ethanol as a white crystal (535 mg, 0.68 mmol, 31%). M. p. 231-233 °C 1H NMR(CDC13): 8 7.40 (d, 2 H, J 9 Hz, H-3”, H-5”), 6.85 (d, 2 H, J 9 Hz, H-2”, H-6”), 5.25(t, 1 H, J2,3 9 Hz, J34 9 Hz, H-3), 5.15-5.20 (m, 2 H, H-2 & 11-3’), 5.05 (t, 1 H, J3’4 9Hz, J4’5 9 Hz, H-4’), 5.00 (d, 1 H, J1,2 8 Hz, H-i), 4.95 (dd, 1 H, J12’ 8 Hz, J2,3’ 9Hz, H-2’), 4.50-4.55 (m, 2 H, H-i’ & H-6b), 4.35 (dd, 1 H, 5’,6b 4 Hz, 6a’,6b’ 12 Hz,H-6b’), 4.15 (dd, 1 H, J56a 5 Hz, 6a,6b 12 Hz, H-6a), 4.05 (dd, 1 H, 5’6a’ 2 Hz, 6a’,6b’12 Hz, H-6a’), 3.85 (t, 1 H, J3,4 9 Hz, J4,5 10 Hz, H-4), 3.75 (ddd, 1 H, J45 10 Hz,5,6a 5 Hz, 5,6b 2 Hz, H-5), 3.65 (ddd, 1 H, J4’,5’ 9 Hz, 5,6a’ 2 Hz, J5’,6b’ 4 Hz, 11-5’),1.95-2.15 (7 s, 21 H, 7(OAc)).Compound 3 was prepared from the acetylated compound (170 mg, 0.32 mmol)using the method of Sinnott & Souchard. The protected compound was dissolved inmethanol (15 ml), 1 M sodium methoxide (0.15 ml) was added and the reaction stirred atroom temperature for 2 hours. The excess base was neutralised with Amberlite IR-120(11+) cation exchange resin. The resin was removed by gravity filtration and washedseveral times with methanol, and the solvent evapourated in vacuo. Compound 3 was148recrystallised from methanol-diethyl ether as a white powder (13 mg, 0.026 mmol, 12%).M. p. 223-224 °C. ‘H NMR (CD3OD): 67.40 (d, J 8 Hz, H-3”, H-5”), 7.00 (d, J 8 Hz,H-2”, H-6”), 4.95 (d, Jj2 8 Hz, H-i), 3.20-4.90 (m, H(2-6)and H(1’-6’)). Anal. Calcd.forC18H25O1Br: C, 43.46; H, 5.23. Found: C,43.06; H, 5.09.2 “,S “-Dinitrophenyl /3-cellobioside (2,5-DNPC) (4)2,5-Dinitrophenyl 2, 3, 6, 2’, 3’, 4’, 6’-hepta-O-acety1--cellobioside wasprepared from hepta-O-acetyl-c-ce1lobiosyl bromide (760 mg, 0.95 mmol), silver oxide(760 mg) and 2,5-dinitrophenol (769 mg, 4.13 mmol) in CH3N (10 mL) stirred overDrierite at room temperature for 24 hours. The Drierite was removed by gravity filtration,wash several times with CHC13 and the solvent evapourated in vacuo .The remainingyellow syrup was dissolved in CHC13,washed with saturated sodium bicarbonate, driedover MgSO4and evapourated in vacuo, resulting in an pale yellow syrup. The acetylatedproduct was purified by flash column chromagraphy using ethyl acetate-hexanes (3 : 2) andwas crystallised from ethyl acetate-ethanol as a white solid (490 mg, 0.61 mmol, 64%). M.p. 208 - 209 °C. 1H N1\’IR (CDC13):6 8.23 (d, 1 H, J4”,6”2 Hz, H-6”), 8.05 (dd, 1 H,J3”,4 9 Hz, J4”,6 2 Hz, H-4”), 7.90 (d, 1 H, J3”4 9 Hz, H-3”), 5.23-5.33 (m, 3 H, H-i,H-2 & H-3), 5.18 (t, 1 H, J2’3 9Hz,J3’,49 Hz, H-3’), 5.08 (t, 1 H, J3’,4 9Hz,J4’59Hz, H-4’), 4.95 (dd, 1 H, J1’,2 8 Hz, J2’39 Hz, H-2’), 4.62 (dd, 1 H, 56b 2 Hz, 6a,6b12 Hz, H-6b), 4.55 (d, 1 H, J1’2 8 Hz, H-i’), 4.38 (dd, 1 H, 35’,6b’ 4 Hz,36a’,6b’ 12 Hz,H-6b’), 4.05-4.15 (m, 2 H, H-6a & H-6a’), 3.85-3.95 (m, 2 H, H-4 & H-5), 3.70 (ddd,1 H, J4’,5 9 Hz, 5’,6a’ 2 Hz, 15’,6b• 4, H-5’), 2.00-2.20 (7 s, 21 H, 7(OAc)).Compound 4 was prepared from this acetylated cellobioside using the method ofBallardie and coworkers. The protected compound (480 mg, 0.60 mmol) was suspended inmethanol (50 ml), cooled to 0 °C, acetyl chloride (0.5 ml) was added and the reactionstirred at 4°C overnight. The solvent was evapourated in vacuo and the resulting white soldwashed with anhydrous diethyl ether (10 x 5 ml). Compund 4 was crystallised from149ethanol as a white solid (77 mg, 0.15 mmol, 25%). M. p. 169-171 °C. ‘H NMR(CD3OD): 6 8.25 (d, J 2 Hz, H-6”), 8.10 (m, H-3” & H-4”), 5.40 (d, J12 8 Hz, H-i),3.25-3.95 (m, H(2-6) and H(1’-6’)). Anal. Calcd forC182405N.0. H: C, 41.78;H, 4.84; N, 5.42. Found: C, 41.94; H, 5.14; N, 5.17.5-1-4{2H}-aryi f-ceIIobiosides2, 4-.Dinitrophenyl J3-{1-2H}-cellobioside (D-2,4-DNPC) (5)2, 3, 6, 2’, 3’, 4’, 6’-Hepta-O-acetyl-D-{1-H}-cellobiose (300 mg, 0.51 mmol)and 1,4-diazabicyclo(2.2.2)octane (190 mg, 1.70 mmol) were stirred over molecular sieves(4 A) in DMF (20 mL) for three hours. Fluorodinitrobenzene (120 mg, 0.64 mmol) wasadded and the reaction was allowed to proceed at room temperature for 24 hours. Thesieves were removed by gravity filtration, washed with CHC13 and the combined extractsevapourated in vacuo. The remaining yellow solid was dissolved in CHC13 (20 mL),washed (3 x 20 mL) with saturated sodium bicarbonate and dried over anhydrous MgSO4.The solvent was evapourated in vacuo yielding 2”,4”-dinitrophenyl 2, 3, 6, 2’, 3’, 4’, 6’hepta-O-acetyl-{1-2H}-cellobioside which was crystallised from ethyl acetate-hexanes asa white solid (140 mg, 0.17 mmol, 33%). M. p. 214-215 °C. 1H NMR (CDC13)6: 8.70(d, 1 H, J3”5 3 Hz, H-3”), 8.43 (dd, 1 H, J3”,5 3 Hz, J”6”9 Hz, H-5”), 7.43 (d, 1 H,J5”,69 Hz, H-6”), 5.28 (t, 1 H, J23 8 Hz, J34 8 Hz, H-3), 5.15-5.23 (m, 2 H, H-2 & H-3’), 5.08 (t, 1 H, J3’,4 9 Hz, J4’,59 Hz, H-4’), 4.95 (dd, 1 H, 1’2’ 8 Hz, J2’,3 9 Hz, H2’), 4.55-4.65 (m, 2 H, H-i’ & H-6b), 4.38 (dd, 1 H, 35’6b’ 4 Hz, 6a’,6b’ 12 Hz, H-6b’),4.05-4.13 (m, 2 H, H-6a & H-6a’), 4.00 (dd, 1 H, J34 8 Hz, 145 10 Hz, H-4), 3.90(ddd, 1 H, J45 10 Hz, 15,6a 5 Hz, 35,6b 2 Hz, H-5), 3.70 (ddd, 1 H, J4’5 9 Hz, 15’,6a’ 2Hz, 5’,6b’ 4 Hz, H-5’), 2.00-2.15 (7 s, 21 H, 7(OAc)).Compound 5 was prepared from the acetylated cellobioside using the method ofBallardie and coworkers. The protected compound (130 mg, 0.16 mmol) was suspended inmethanol (15 ml), cooled to 0 °C, acetyl chloride (0.15 ml) was added and the reaction150stirred at 4°C overnight. The solvent was evapourated in vacuo and the resulting white soldwashed with anhydrous diethyl ether (10 x 5 ml). Compound 5 was crystallised frommethanol-diethyl ether as a white solid (22 mg, 0.027 mmol, 17%). M. p. 180 °C dec. 1HNMR (CD3OD) & 8.73 (d, 1 H, J3”,5” 3 Hz, H-3”), 8.45 (dd, 1 H, J3”,5” 3 Hz, J5”,6” 9Hz, H-5”), 7.63 (d, 1 H, J5”,6” 9 Hz, H-6”), 4.45 (d, 1 H, Ji’,2’ 8 Hz, H-i’), 3.20-3.95(m, H(2-6) & H(2’-6’). Anal. Calcd forC1823D05N.:C, 42.44; H, 4.72; N, 5.50.Found: C, 42.03; H, 4.82; N, 5.44.4-Nitrophenyl J3-{1-2H} -cellobioside (D-PNPC) (6)The acetylated compound, 4-nitrophenyl 2, 3, 6, 2’, 3’, 4’, 6’-hepta-O-acetyl--{1-2H}-cellobioside was prepared from hepta-O-acetyl-a-(1-H)-cellobiosyl bromide(290 mg, 0.42 mmol), silver oxide (290 mg) and 4-nitrophenol (290 mg, 2.09 mmol) inCH3N (5 mL) stirred over drierite at room temperature for 24 hours. The Drierite wasremoved by gravity filtration, wash several times with CHC13and the solvent evapouratedin vacuo .The remaining yellow syrup was dissolved in CHC13,washed with saturatedsodium bicarbonate, dried over MgSO4and evapourated in vacuo, resulting in an paleyellow syrup. This acetylated product was crystallised from ethyl acetate-ethanol as a whitesolid (150 mg, 0.20 mmol, 48%). M. p. 237-239 °C.1H NMR (CDC13)& 8.18 (d, 2 H, J9 Hz, H-3” & 5”), 7.05 (d, 2 H, J 9 Hz, H-2” & 6”), 5.28 (t, 1 H, J23 9 Hz, J3,4 9 Hz,H-3), 5.12-5.20 (m, 2 H, H-2 & H-3’), 5.05 (t, 1 H, J3’4 9 Hz, J4’,59 Hz, H-4”), 4.93(dd, 1 H, J1’,2’ 8Hz,J2’39 Hz, H-2’), 4.50-4.55 (m, 2 H, H-i’ & H-6b), 4.35 (dd, 1 H,5’,6b’ 4 Hz, 6a’6b’ 12 Hz, H6b’), 4.13 (dd, 1 H, J5,6a 5 Hz, J6a,6b 12 Hz, H-6a), 4.05(dd, 1 H, 5’6a’ 2 Hz, 6a’, 6b’ 12 Hz, H-6a’), 3.78-3.90 (m, 2 H, H-4 & H-5), 3.68 (ddd,1 H, J4’5 9 Hz, 5’,6a’ 2 Hz, 5’,6b’ 4 Hz, H-5’), 2.00-2.10 (7 s, 21 H, 7(OAc)).Compound 6 was prepared from the acetylated cellobioside using method ofBallardie and coworkers. The protected compound (143 mg, 0.19 mmol)) was suspendedin methanol (15 ml), cooled to 0 °C, acetyl chloride (0.15 ml) was added and the reaction151stirred at 4°C overnight. The solvent was evapourated in vacuo and the resulting white soldwashed with anhydrous diethyl ether (10 x 5 ml). Compound 6 was purified by flashcolumn chromatography (12 : 2: 1 ethyl acetate-MeOH-H20) and crystallised from waterdiethyl ether as a white solid (15 mg, 0.032 mmol, 17%). M.p. 254-255 °C. 1H NMR(CD3OD) 6: 8.23 (d, J 9 Hz, H-3” & H-5”), 7.23 (d, J 9 Hz, H-2” & H-6”), 4.45 (d, Ji’,2’8 Hz, H-i’), 3.20-3.95 (m, H(2-6) & H(2’-6’)). Anal. Calcd forC1824D03Ni.5H20.: C, 43.99; H, 5.70; N, 2.85. Found: C, 44.24; H, 5.65; N, 2.75.4-Bromophenyl /3-11-2H} -cellobioside (D-4-BrPC) (7)4-Bromophenyl 2, 3, 6, 2’, 3’, 4’, 6’-hepta-O-acetyl-f3-{i-2H}-cellobioside wasprepared from hepta-O-acetyl-a-cellobiosyl bromide (400 mg, 0.63 mmol), silver oxide(400 mg) and 4-bromophenol (400 mg, 2.32 mmol) in CH3N (10 mL) stirred overDrierite at room temperature for 24 hours. The Drierite was removed by gravity filtration,wash several times with CHC13 and the solvent evapourated in vacuo .The remaining syrupwas dissolved in CHC13,washed with saturated sodium bicarbonate, dried over MgSO4and evapourated in vacuo. The acetylated product was crystallised from ethyl acetate-ethanol as a white solid (260 mg, 0.33 mmol, 52%). M. p. 23 1-232 °C. 1H NMR(CDC13): 6 7.40 (d, 2 H, J 9 Hz, H-3”, H-5”), 6.85 (d, 2 H, J 9 Hz, H-2”, H-6”), 5.25(t, 1 H, J2,3 9 Hz, J34 9 Hz, H-3), 5.15-5.20 (m, 2 H, H-2 & H-3’), 5.05 (t, 1 H, J3’4 9Hz, J4’5 10 Hz, H-4’), 4.95 (dd, 1 H, J1’,2’ 8 Hz, J2’,3 9 Hz, H-2’), 4.50-4.55 (m, 2 H,H-i’ & H-6b), 4.35 (dd, 1 H, 5’6b’ 4 Hz, 6a’,6b’ 12 Hz, H-6b’), 4.15 (dd, 1 H, 56a 5Hz, 6a,6b 12 Hz, H-6a), 4.05 (dd, 1 H, 35’,6a’ 2 Hz, ‘6a’6b’ 12 Hz, H-6a’), 3.85 (t, 1 H,J3,4 9 Hz, J4,5 10 Hz, H-4), 3.75 (ddd, 1 H, J4,5 10 Hz, 5 Hz, 5,6b 2 Hz, H-5),3.65 (ddd, 1 H, J’5’ 10 Hz, 5’,6a’ 2 Hz, 35’6b 4 Hz, H-5’), 1.95-2.15 (7 s, 21 H,7(OAc)).Compound 7 was prepared from the acetylated compound (200 mg, 0.25 mmol)using the method of Sinnott & Souchard. The protected compound was dissolved in152methanol (20 ml), 1 M sodium methoxide (0.20 ml) was added and the reaction stirred atroom temperature for 2 hours. The excess base was neutralised with Amberlite IR-120(14+) cation exchange resin. The resin was removed by gravity filtration and washedseveral times with methanol, and the solvent evapourated in vacuo. Compound (7) waspurified by flash column chromatography (ii : 2 : 1 ethyl acetate-MeOH-H20) andcrystallised from water-cliethyl ether as a white solid (22 mg, 0.055 mmol, 18%) M.p.222-223 °C. 1H NMR (CD3OD) & 7.40 (d, J 9 Hz, H-3” & H-5”), 7.02 (d, J 9 Hz, H-2”& H-6”), 4.43 (d, J1’,2’ 8 Hz, H-i’), 3.20-3.95 (m, H(2-6) & H(2’-6’)). Anal. Calcd forC1824DOBr: C, 43.37; H, 5.22. Found: C, 42.96; H, 5.28.5-1-5 InactivatorsCellobial (8)3, 6, 2’, 3’, 4’, 6’-Hexa-O-acetyl cellobial was synthesised fromacetobromocellobiose (4 g, 5.73 mmol) dissolved in 90% glacial acetic acid (50 mL) wascooled to 12-15 °C. Chioroplatinic acid (one drop of —0.5% chioroplatinic acid in 50%acetic acid) and zinc (20 g) was added to the mixture and stirred at 12-15 °C for 2 hours.The reaction mixture was diluted with water (125 mL), the zinc filtered and washed and theproduct extracted with CHC13 (2 x 50 mL). The organic layer was washed with saturatedsodium bicarbonate, dried over MgSO4 and evapourated in vacuo resulting in an orangesyrup. The product crystallised from CHC13-hexanes (1:4) as clear crystals (1.8 g, 3.21mmol, 56%). M. p. 134-135 °C (Lit. m. p. 134-135°C, (Kanda et al., 1986))Compound 8 was prepared from the acetylated compound (350 mg, 0.62 mmol)using the method of Sinnott & Souchard. The protected compound was dissolved inmethanol (30 ml), 1 M sodium methoxide (0.30 ml) was added and the reaction stirred atroom temperature for 1 hours. The excess base was neutralised with Amberlite IR-120(H+) cation exchange resin. The resin was removed by gravity filtration and washedseveral times with methanol, and the solvent evapourated in vacuo. Compound 8 was153crystallised from hot ethanol as a white solid (130 mg, 0.42 mmol, 68%): M. p. 174-175°C. (lit. mp 175 -176 °C (Kanda et al., 1986)) 1H NMR.(CDOD) 6: 6.35 (dd, 1 H, J1,3 2Hz, Ji,2 6 Hz, H-i), 4.70 (dd, 1 H, J12 6 Hz, J23 2 Hz, H-2), 4.48 (d, 1 H, J1’,2’ 8 Hz,H-i’), 4.30 (dt, 1 H, 34,5 7 Hz, 35,6a 2 Hz, 35,6b 2 Hz, H-5), 3.80-4.00 (m, 4 H, H-3, H-3’, H-4’, H-5’ & H-6b), 3.73 (dd, 1 H, 345 7 Hz, J3,4 10 Hz, H-4), 3.20-3.63 (m, 4 H,H-2’, H-6a’ & H-6b’). Anal. Calcd forC12H2009:C, 46.75; H, 6.49. Found: C, 46.64;H, 6.38.2 -Deoxy-2-fluoro-4-O-{f3-D-glucopyranosyl}-13-D-mannosylfluoride (2F-GMF) (9)Acetylated cellobial (4 g, 7.14 mmol) was dissolved in acetonitrile (40 mL) and 2%F2 gas in neon bubbled into the reaction mixture for 10 minutes at room temperature. Thesolvent was evapourated in vacuo leaving an orange oil behind. The acetylated compound,2-deoxy-2-fluoro-3,6-di-O-acetyl-4-O- { 2’, 3’, 4’, 6’-tetra-O-acetyl-f3—D-glucopyranosyl }-mannosyl fluoride, was partially purified by flash column chromatography using diethylether-CHC13(1: 1) and on the chromatotron using diethyl ether-CHC13(3: 1). Theproduct was further purified from the 2-fluoro-cellobioside analogue on the chromatotronusing ethyl acetate-hexanes (45 : 55) and crystallised from hot ethanol as clear crystals (460mg, 0.78 mmol, 11%). M. p. 147-149 °C. 1H NMR (CDC13)6: 5.48 (dd, 1 H, J1,F1 51Hz, J1,F2 11 Hz, H-i), 5.20 (ddd, 1 H, J2,3 3 Hz, 334 8 Hz, J3J2 20 Hz, H-3), 5.15 (t,1 H, 2’,3’ 9 Hz, J3’4 9 Hz, H-3’), 5.03 (t, i H, J3’4’ 9 Hz, J4’,5 9 Hz, H-4’), 4.90 (dd, 1H, J1’,2’ 8 Hz, J2’,3’ 9 Hz, H-2’), 4.83 (dd, 1 H, 32,F1 12 Hz, 32,F2 48 Hz, H-2), 4.55 (d,1 H, J1’’ 8 Hz, H-i’), 4.50 (dd, 1 H, 35’,6b’ 5 Hz,36a’,6b’ 12 Hz, H-6b’), 4.30 (dd, 1 H,5,6b 6 Hz, 36a6b 12 Hz, H-6b), 4.18 (dd, 1 H, 5,6a 6 Hz,36a,6b 12 Hz, H-6a), 3.95-4.05 (m, 2 H, H-4 & H-6a’), 3.78 (q, 1 H, J45 5 Hz, 5,6a 6 Hz, 5,6b 6 Hz, H-5), 3.68(ddd, 1 H, J4’,5 9 Hz, 35’,6a’ 2 Hz, 35’,6b’ 5 Hz, H-5’). 19F NMR (CDC13)6: -145.5 (dt,Ji,Fi 51 Hz, Jzri 12 Hz, JF1,T2 13 Hz, F-i), -218.7 (m, F-2).154Compound 9 was prepared from the acetylated compound (110 mg, 0.18 mmol)using the method of Sinnott & Souchard. The protected compound was dissolved inmethanol (15 ml), 1 M sodium methoxide (0.15 ml) was added and the reaction stirred atroom temperature for 2 hours. The excess base was neutralised with Amberlite IR-120(H+) cation exchange resin. The resin was removed by gravity filtration and washedseveral times with methanol, and the solvent evapourated in vacuo.. Compound 9 waspurified by high performance liquid chromatography using a Waters 712 WISP HPLCequipped with a Waters 410 differential refractometer. The crude reaction mixture wasloaded onto a Dextropak (25 x 10 cm) column, eluted with water at a flow rate of 7 niL/mmand the fractions collected manually. Fractions containing pure compound 9 were pooledand lyophilised to yield a white solid (25 mg, 0.072 mmol, 40%): M. p. 170-172 °C. 1HNMR (CD3OD) ö: 5.58 (dd, 1 H, I,F1 48 Hz, J1,F2 14 Hz, H-i), 4.98 (d, 1 H, J2F2 50Hz, H-2), 4.45 (d, 1 H, J1’2 7 Hz, H-i’), 3.20-3.40 (m, H(3-6) & H(2’-6’)). WF NMR(D20) & -148.7 (dd, J1,F1 48 Hz, J1,F2 14 Hz, F-i), -222.8 (m, F-2). Anal. Calcd forC12H2009F:C, 41.62; H, 5.78. Found: C, 41.90; H, 5.86.Bromoacetic anhydride (10)Compound 10 was prepared from dicyclohexylcarbodiimide (1.2 g, 5.8 mmol) incold CCI1 (6 mL) was added slowly to bromoacetic acid (1 g, 7.2 mmol) dissolved in coldCCL1 (4 mL) and allowed to stir for 5 minutes at 0°C. The reaction mixture was allowed towarm up to room temperature and the dicyclohexylurea biproduct was filtered leavingproduct 10 in solution. Bromoacetic anhydride was crystallised from CCLI at -20 °C aswhite crystals (955 mg, 3.67 mmol, 63%). M. p. 39 - 41 °C (lit. m. p. 41-42 °C (Thomas,1977))155N-Bromoacetyl-f3-cellobiosylamine (N-BrAcCb-NH2)(11)Cellobiose (1.0 g, 2.79 mmol) was dissolved in water (10 mL), ammoniumbicarbonate added until the solution was saturated and stirred for 10 days with occasionalammonium bicarbonate resaturations. Evapouration of the solvent in vacuo yielded a whitegum. Partial purification by flash chromatography (2:2:1 ethyl acetate-methanol-water)gave a mixture of -D-cellobiosylamine and cellobiose. Bromoacetic anhydride (900 mg,3.46 mmol) was then added to a solution of this mixture in DMF (10 mL) and stirred atroom temperature for 3 hour. The solution was poured over ice-cold anhydrous ether andstirred for an hour. Purification by flash chromatography (5:2:1 ethyl acetate-methanol-water) and crystallization from methanol-diethyl ether yielded compound 11 as a whitesolid (582 mg, 45 %). M.p. 143-145 0; 1H NMR (CD3OD): 64.80 (d, 1 H, J12 8 Hz, H-1), 4.41 (d, 1 H, J12’ 8 Hz, H-i’), 3.87 (s, 2 H, CJj2BrCO), 3.25-3.90 (m, 11 H, H2,3,4,5,6a,6b,3’,4’,5’,6a’,6b’), 3.22 (t, 1 H, J2l’ 8, J2’,3 8 Hz, H-2’). Anal Calcd. forC14H2NO1Br: C, 36.36; H, 5.19; N, 3.03. Found: C, 36.35; H, 5.40; N, 2.95.5-2 Enzymology5-2-1 General materials and proceduresAll chemicals were of analytical reagent grade. 4-Nitrophenyl- and 2-nitrophenyl-3-cellobiosides, buffer materials and all other chemicals were obtained from either Sigma orAldrich Chemical Companies. Intact (molecular weight =47 kDa) and truncated (molecularweight = 35 kDa) native Cellulomonas fimi exoglycanase as well as the Glu233Asp andGlui27Ala mutants were provided by Drs. R. A. J. Warren, N. Gilkes, D. Kilburn, andDr. A. MacLeod, Department of Microbiology, University of British Columbia.All absorbance measurements were recorded on a Pye-Unicam 8700 UV/Visspectrometer equipped with a circulating water bath. All protein and peptide mass spectrawere recorded by either Dr. David Burgoyne or Dr. Shichang Miao in Dr. RuediAebersold’s laboratory, Biomedical Research Centre, University of British Columbia using156a PE-Sciex API 111 triple quadrupole mass spectrometer (PE-Sciex, Thornhill, Ontario)equipped with an ionspray ion source. Protein and peptide samples were separated byreverse phase high performance liquid chromatography (RP-HPLC) on an UltrafastMicroprotein Analyser (Michrom BioResources Inc., Auburn, CA) directly interfaced withthe mass spectrometer using solvent A: 0.05% trifluoroacetic acid, 2% acetonitrile in waterand solvent B: 0.045% trifluoroacetic acid, 80% acetonitrile in water. The system wasequipped with a post-column flow splitter to introduce 15% of the HPLC eluate into themass spectrometer while 85% was collected for further analysis (Hess et al., 1993).5-2-2 Determination of steady state kinetic parametersAll steady state kinetic studies were performed by recording changes in UV/Visabsorbance using a Pye-Unicam 8700 specirophotometer equipped with a circulating waterbath. Reactions were monitored at wavelengths where there was a convenient absorbancedifference between the initial glycoside and the phenol product as previously reported(Kempton & Withers, 1992) using the same extinction coefficients. Initial rates ofexoglycanase-catalysed hydrolysis of aryl 3-D-glycosides were determined by incubatingsolutions of the appropriate substrate concentrations in 50 mM sodium phosphate buffer,pH 7.0 and 1 mg/mL BSA at 37 °C in 1-cm cuvettes within the spectrophotometer untilthermally equilibrated. Reactions were initiated by the addition of enzyme, and release ofthe phenol product was monitored at the appropriate wavelength. In order to ensure linearkinetics and to obtain a sufficient absorbance change for accurate calulation of the rates, theconcentration of the enzyme added and the length of time that the reaction was monitoredwas selected such that less than 10% of the total substrate was converted to product.Initially, approximate values of Km and Vmax were calculated from three pointLineweaver-Burk plots where the initial rates of hydrolysis for three widely variedsubstrate concentrations were measured. More accurate values for Km and Vm were thendetermined by measuring the initial rates of hydrolysis for 5-8 different substrate157concentrations which generally ranged from 0.2 to 5 times Km except when limited bysubstrate insolubility. These results are illustrated as Lineweaver-Burk double reciprocalplots for visual convenience in Appendix A where the enzyme concentrations, thewavelengths and the molar extinction coefficients used are indicated in the legends.However, due to the inaccuracy introduced by the nonlinear error span of the doublereciprocal analysis, the values of Km and Vmar were determined actually by nonlinearregression analysis using the program GraFit (Leatherbarrow, 1990). Standard errors forthese constants were calculated by the same fitting program.5-2-3 Determination of pre-steady state kinetic parametersPre-steady state kinetic measurements were performed on an Applied PhotophysicsMV 17 microvolume stopped flow spectrophotometer equipped with a Grant constanttemperature bath. Reactions were monitored at the same wavelengths as in the steady statekinetic studies. The concentration of enzyme used in each case was chosen to yield a burstwith a total absorbance change of 0.06 A. Rates were determined by equilibrating solutionsof enzyme and of the appropriate concentrations of substrate in 50 mM sodium phosphatebuffer, pH 7.0 to 37 °C. The reactions were monitored by following the release of thephenol product at the appropriate wavelength. Reaction rates were measured at fivedifferent substrate concentrations ranging from 0.2 x Kd to -Kj whenever possible. Themeasurement at each substrate concentration was repeated 3-4 times, the traces averagedand fitted to an equation describing a first order reaction followed by a steady state. Thisyielded values of the pseudo-first order rate constant (kobs) and the steady state rate at eachsubstrate concentration. Values of KJ and k2 were determined from these kobs values bydirect fit to the equation1obs k2LSIKj + [Sj158using the program GraFit (Leatherbarrow, 1990). Standard errors for these constants werecalculated by the same fitting program. For visual convenience, these results are presentedas Lineweaver-Burk plots in Appendix A along with the enzyme concentrations,wavelength and molar extinction coefficients used.5-2-4 Secondary deuterium kinetic isotope effect measurementsIsotope effects were determined by comparison of the initial rates of hydrolysis ofhigh (4-7 times the Km value) concentrations of protio and deuterio substrates determinedspectrophotometrically. In most cases, quartz cells were filled with the appropriateconcentration of diluted enzyme and incubated at 37 °C, reaction being initiated by theaddition of a small volume (50 - 100 .tl) of (thermally equilibrated) substrate. Whensubstrate solubilities precluded this approach it was necessary to add the enzyme to the preequilibrated substrate. Rates of protio and deuterio substrate hydrolysis were determined inalternation until a total of 8 or 9 rates for each (protio and deuterio) substrate had beenmeasured. Average rates for the protio and deuterio substrates were then calculated and therate taken to give the isotope effect. Errors provided are the standard deviation of theaverage kinetic isotope effect.5-2-5 pH Study(a) Extinction coefficients of2,4 dinitrophenol2,4-Dinitrophenol was dried in vacuo overnight, weighed and dissolved in a knownvolume of water. Aliquots of stock 2,4-dinitrophenol were added to the appropriatebuffers, equilibrated to 37 °C and the absorbance recorded at 400 nm. The extinctioncoefficients of 2,4-dinitrophenol at the different pH values were determined from thesemeasurements using Beer’s law:159E=Abcwhere A corresponds to the absorbance, b to the cell pathlength (1 cm) and c to theconcentration of 2,4-dinitrophenol. In order to determine if the pH had changed due to theaddition of 2,4-dinitrophenol, the pH of the solution was measured after recording theabsorbance. The buffers used were 50 mM citrate (pH 4-6), 50 mM phosphate (pH 6-8),50 mM Tris (pH 8-9) and 50 mM carbonate-bicarbonate (pH 9-10). The extinctioncoefficients determined are 9.0, 9.2, 9.6, 9.9, 10.2, 10.6, 10.7 and 10.9 mM’ cm’ at pH4.53, 4.61, 4.78, 5.01, 5.16, 5.60, 6.25, and 6.8 respectively.(b) pH dependence ofenzyme kinetic parameters: The dependence of and kcaJKm on pHfor the native enzyme and the Glu233Asp mutant was determined as follows. Rates ofexoglycanase-catalysed hydrolysis at different pH values were determined by incubating 6different concentrations of 2,4-DNPC in the appropriate buffer containing 1 mg/mL BSAand 145 mM NaCl at 37 °C until thermally equilibrated. The reaction was initiated byaddition of enzyme, dinitrophenolate release was monitored at 400 nm and values for kand Km determined as previously described. The dependence of kcat on pH for Glul27Aspwas determined by thermally equilibrating saturating concentrations of 2,4-DNPC (75 xKm) in the appropriate buffer containing 1 mg/mL BSA and 145 mM NaC1 at 37 °C and thereaction initiated by enzyme addition. Values of kcat were then calculated from the initialrates of hydrolysis using the appropriate extinction coefficient for 2,4-dinitrophenol. ThepH dependence of kcat/Km for the Glu l27Ala mutant was determined by incubating PNPCat a fmal concentration of 0.2 x Km in the appropriate buffer and 1 mg/mL BSA and 145mM NaCl at 37 °C until thermally equilibrated. The reactions were initiated by the additionof enzyme and the release of 4-nitrophenolate was monitored by following the absorbanceat 400 nm until the substrate was depleted. The change in absorbance with time was fittedto a first order rate equation using the program GraFit (Leatherbarrow, 1990) which160yielded values for the pseudo-first order rate constant at each pH value. Since at lowsubstrate concentrations (S<<Km) the reaction rates are given by the equationv =Jat [El [SiKmthen kobs values correspond to kcatfKm. In order to ensure that the pH had not changedduring the reaction, the pH of each reaction mixture was measured after recording the ratesof hydrolysis.5-2-6 Inactivation kineticsThe kinetic parameters for the inactivation of C.fimi exoglycanase by cellobial and2F-GMF were determined as follows. C.Jlmi exoglycanase was incubated in 50 mMsodium phosphate buffer, pH 7.0 containing 1 mglml BSA at 37 °C in the presence ofvarying concentrations of the inactivator. Concentrations of cellobial used were 1.0, 7.1,10.0, 29.9, and 54.2 mM while concentrations of 2F-GMF used were 4.1, 8.2, 14.4,24.6, and 30.8 mM. Aliquots of these inactivation mixtures were removed at different timeintervals and diluted into reaction cells containing a large volume of substrate (2,4-DNPC)at saturating concentration (1 mM, 10 x Km). This stopped the inactivation both by dilutionof the inactivator and by competition for the excess substrate. The residual enzymaticactivity was determined from the rate of hydrolysis of the substrate, which is directlyproportional to the amount of active enzyme. The inactivation was monitored until 80-90%of the enzymatic activity was lost. Pseudo-first order rate constants (kobs) for inactivationwere calculated for each inactivator concentration by direct fit of each curve to a first-orderfunction, and then values of k and K were determined from these kobs values by direct fitto the equationkobs =K + [I]161using the program GraPh (Leatherbarrow, 1990) (Appendix A-8). However, due to thevery large values of K1 for both cellobial and 2F-GMF and the limited solubiity of theseinactivators, these values must be taken as only estimates of K and k. Nevertheless, sinceat very low inactivator concentrations ([II <<K1), values of kobs are given by the equationkobs= kLilthen accurate values ofk1/K for cellobial and 2FGMF inactivation can be calculated fromthe slopes of plots of kobs versus inactivator concentrations (Appendix A-8).5-2-7 Protection against inactivationProtection against exoglycanase inactivation by cellobial was investigated asfollows. Samples of exoglycanase (0.0 15 mg, 0.073 mg/mI) were incubated in 50 mMsodium phosphate buffer, pH 7.0 at 37 °C containing the cellobial (11.09 mlvi) and thecompetitive inhibitor, BGTX (Kd = 3.0 mM), at concentrations of 0 or --2.0 mM. Aliquotswere removed at different time intervals, diluted into reaction cells containing saturatingconcentrations of 2,4-DNPC (0.66mM —7 x Km) in 50 mM sodium phosphate buffer, pH7.0 at 37 °C, and the residual enzyme activity monitored by following the release of thedinitrophenolate at 400 nm as described above. Pseudo-first order rate constants forinactivation at each BGTX concentration were calculated and compared to determine thedegree of protection afforded by BGTX against inactivation.Protection against C. fimi exoglycanase inactivation by N-bromoacetylcellobiosylamine was investigated similarly using an inactivator concentration of 3.5 mMand BGTX concentration of 9.7 mM. Protection against 2F-GMF inactivation wasinvestigated using an inactivator concentration of 30.8 mM and BGTX concentration of 2mM.1625-2-8 Reactivation kineticsReactivation of cellobial-inactivated exoglycanase was investigated as follows. Asample of inactivated exoglycanase (0.25 mg, 5 mg/mi) were extensively dialysed at 4°Cagainst several changes of 50 mM sodium phosphate buffer pH 7.0 in order to remove theexcess inactivator. Aliquots of the dialysed inactivated enzyme (0.018 mg)were added tobuffer solutions containing BSA (1 mg/mL) and either buffer alone or 50 mM cellobiose.These solutions were incubated at 37 °C and monitored for return of enzymatic activity byperiodic removal of samples and assayed using 2,4-DNPC as described above.Reactivation rates were calculated by fitting these data to first-order curves by nonlinearregression analysis (Leatherbarrow, 1990) andthese are illustrated in Appendix 8.5-2-9 19F-NMR analysis of 2-deoxy-2-fluoro-glycosyl-C.fimiexoglycanaseSamples of C.JImi exoglycanase (5 mg, 9 mg/mi) were individually inactivatedwith either 2F-DNPC (1.0 mM) or 2FGMF (30 mM) in 50 mM sodium phosphate buffer(500 .t1) pH 7.0 containing 10% D20 at 37°C. The inactivated enzyme samples wereanalysed on a 470 MHz VARIAN for 19F NMR by Dr. Lawrence McIntosh in theDepartment of Chemistry, University of British Columbia, Vancouver. The 2F-GMF-inactivated exoglycanase sample was dialysed extensively against 50 mM sodiumphosphate buffer, pH 7.0 at 4 °C, reconcentrated to —9 mg/mi using a Miilipore UFC-10poiysuifone membrane concentrator and resubmitted for 19F-NMR analysis. The spectrumfor the 2F-DNPC-inactivated C.fimi exoglycanase sample was recorded using sweepwidthof 23460.4 Hz and repetition time of 2.2 seconds at a pulse angle of 90°C. The signals areaveraged over 6198 transient and referenced to external fluoride at = -121.39 ppm. Thespectrum for the 2F-GMF-inactivated enzyme sample was recorded using sweepwidth of53619.3 Hz and repetition time of 3.2 seconds at a pulse angle of 90 °C. The signals are163averaged over 14936 transients and referenced to internal fluoride at 6= -121.39 ppm. Thespectra were processed with a 2- 4 Hz linebroadening and baseline flattening.5-2-10 Stoichiometry of inactivationThe stoichiomeiry of C. fimi exoglycanase inactivation by the disaccharides 2F-DNPC, 2F-DNPX 2FGMF, cellobial and N-bromoacetyl cellobiosylamine wasdetermined by subjecting samples of inactivated exoglycanase (10 1 mg/mi) anduntreated enzyme (10 .tg, 1 mg/mI) to mass spectrometric analysis on a PE-Sciex API Illtriple quadrupole mass spectrometer. This involved introduction of the protein sample intothe mass spectrometer through a microbore PLRP column (1 x 50 mm) on the MichmmHPLC system using a gradient of 20-100% solvent B (solvent A: 0.05% trifluoroaceticacid, 2% acetonitrile in water and solvent B: 0.045% thfluoroacetic acid, 80% acetonitrilein water) over 10 minutes followed by 100% solvent B over 2 minutes. The massspectrometer, in the single quadrupole mode, was scanned over a nz range of 300-2400Da. Protein molecular weights were determined from this data using the deconvolutionsoftware supplied by Sciex.5-2-11 Identification of the residue labeled by 2F-DNPC and cellobialThe exoglycanase (100 jig, 9.8 mg/mi) was completely inactivated with either 2F-DNPC (1.0 mM) or cellobial (60 mM) in 50 mM sodium phosphate buffer, pH 7.0 at 37°C. Samples of labeled and unlabeled enzyme were digested using 1:100 pepsin (w/w,pepsin : exoglycanase) in 50 mlvi sodium phosphate buffer, pH 2.0 at room temperature.and the resulting digests subjected to mass spectromeiric analysis by Dr. Shichang Miao.This involved individually loading the digest (—0.05 .tg) onto a C18 Reliasil column (1 x150 mm) and eluting the separaed peptides directly into the mass spectrometer using agradient of 0-60% solvent B over 20 minutes followed by 100% solvent B for 2 minutes ata flow rate of 50 tl/minute. In this mass spectrometric experiment, the mass analyser in the164single quadrupole mode was scanned over m/z range of 300-2400 Da. The labeled peptidewas identified in a second mass spectrometric experiment in which the peptide ions weresubjected to collision-induced fragmentation in the second quadrupole. In this experimentthe mass spectrometer was scanned in the neutral loss mode searching for a mass losscorresponding to the loss of the label from a peptide ion in the singly or doubly chargedstate. The triple quadrupole mass analyser was scanned over a m/z range of 300-1200 Dawith collision gas (10% N2 diluted with Ar) thickness of 3.2-3.6 x iO molecules/cm2inthe second quadrupole.The 2-deoxyfluoro-cellobiosyl peptide was purified by RP-HPLC using the massspectrometer scanned in the neutral loss mode as the detector. This involved loadingaliquots of the labeled digest on to the Cl 8 Reliasil column and eluting with the gradientdescribed above. A post-column flow splitter was used to introduce 15% of the eluent intothe mass spectrometer while the remaining 85% was fractionated (Hess et al., 1993). Thefraction containing the labeled peptide was concentrated in vacuo and rechromatographedon the same column using a gradient of 0-20% solvent B over 50 minutes followed by100% solvent B over 2 minutes at a flow rate of 50 pi/min. The resultant labeled peptidewas sufficiently pure to be sequenced. This peptide was sequenced by solid phase Edmandegradation on a Milligen/BioSearch Model 6600 protein sequencer in the laboratory of Dr.Rued Aebersold at the Biomedical Research Centre, University of British Columbia,Vancouver. This involved the sequential derivatisation of the N-terminal amino acid tophenylthiohydantoins (PTH) followed by the separation and identification of the PThresidues by HPLC.5-2-12 Identification of the residue labeled by N-bromoacetylcellobiosylamineAn aliquot of C. fimi exoglycanase (1.0 mg, 8 mg/mi) was inactivated with Nbromoacetyl cellobiosylamine (30 mM) in 50 mlvi sodium phosphate buffer, pH 7.0 at 37165°C. Samples of labeled and unlabeled exoglycanase were completely digested using 1:100pepsin (w/w, pepsin: exoglycanase) in 50 mM sodium phosphate buffer pH 2.0 at roomtemperature. Mass spectromeiric analysis of these digests was performed by Dr. DavidBurgoyne in the laboratory of Dr. Ruedi Aebersold. This involved individually loading thedigest (10 .tg) onto a C18 column (Reliasil, 1 x 150 mm) and then eluting the peplidesdirectly into the mass spectrometer with a gradient of 0-20 % solvent B over 50 minutesfollowed by 100% solvent B for 2 minutes at a flow rate of 50 I.tl/min. The quadrupolemass analyser, in the single quadrupole mode, was scanned over a mlz range of 300 - 2400Da. The fraction containing the labeled peptide was identified from this mass information.The N-acetyl cellobiosylamine labeled peptide was purified by RP-HPLC on a C18Reliasil (1 x 150 mm) column using the mass spectrometer scanned in the singlequadrupole mode as a detector. The peptides were eluted with a gradient of 0-60% solventB over 20 minutes followed by 100% solvent B over 2 minutes at a flow rate of 50 p.1/mm.The post-column flow splitter was utilised to introduce 15% of the eluent into the massspectrometer while the remaining 85% was fractionated. The fraction containing the labeledpeptide was concentrated in vacuo and the labeled peptide further purified by reloading ontothe C18 Reliasil column and eluting with a gradient of 0-20% solvent B over 50 minutesfollowed by 100% solvent B over 2 minutes at a flow rate of 50 p.1/mm.The isolated, labeled peptide was sequenced in Dr. Ruedi Aebersold’s laboratory byMr. David Chow using a novel Edman sequencing reagent. This automated sequencinginvolved derivatising the N-terminal amino acid residues using 4-(3-pyridinylmethyl-aminocarboxypropyl) phenyl isothiocyanate (311 PITC, Hess et al., 1995), loading thePTH derivatives onto a Reliasil BDS C18 column (1 x 50 mm) and eluting the PThderivatives directly into the mass spectrometer with a gradient of 0-100% solvent B in 8minutes at a flow rate of 50 p.1/mm. In this experiment the mass analyser was scanned overm/z of 300 - 850 Da. This experiment identified the mass of the covalently modifiedresidue. The structure of the modified residue was further characterised in a second166sequencing experiment in which the ion corresponding to the modified amino acid wassubjected to collision-induced fragmentation. In this second experiment the massspectrometer was scanned over a mass range of 100-850 Da with collision gas thickness of6.0 x 1014 molecules/cm2in the second quadrupole.167REFERENCESAlbery, W. J., & Knowles, 3. R. (1976) Biochemistry 15, 5631.Amyes, T. L., & Jencks, W. P. (1989) J. Am. Chem. Soc. 111, 7888.Anderson, W. F., Grutter, M. G., Remington, S. J., Weaver, L. H., & Matthews, B. W.(1981) J. Mol. Biol. 147, 523.Ballardie, F., Capon, B., Derek, J., & Sutherland, G. (1973) J. Chem. Soc., Perkin.Trans. I, 2418.Ballardie, F. W., Capon, B., Cuthbert, M. W., & Deane, W. M. (1977) Bioorg. Chem.6, 483.Barlin, G. S., & Perrin, D. D. (1966) Q. Rev. Chem. Soc. 20, 75.Ba-Saif, S. A., & Williams, S. (1988) J. Org. Chem. 53, 2204.Bause, E., & Legler, G. (1974) Hoppe-Seyler’s Z. Physiol. Chem. 355, 438.BeMiller, J. N. (1967) Adv. Carb. Chem. Biochem. 22, 25Bennet, A., & Sinnott, M. L. (1986) J. Am. Chem. Soc. 108, 7287.Black, T., Kiss, L., Tull, D., & Withers, S. G. (1993) Carbohydr. Res. 250, 195.Blake, C. C. F., Johnson, L. N., Mair, G. A., North, A. C. T., Phillips, D. C., &Sarma, V. R. (1967) Proc. R. Soc. Lond. B 167, 365.Bray, M. R., & Clarke, A. J. (1994) Eur. J. Biochem. 219, 821.Bures, E. J., Nika, H., Chow, D. T., Morrison, H., Hess, D., & Aebersold, R. (1995)Anal. Biochem. 224, 364.Campbell, R., Rose, D., Wakarchuk, W., To, R., Sung, W., & Yaguchi, M., (1993) Acomparison of the structures of of the 20 kD xylanasesfrom Trichoderma harzianum andBacillus circulans, in Proceedings of the 2nd TRICEL symposium on Trichoderma reeseicellulases and other hydrolases, (Suominen, P., & Reinikainen, T., Ed.) p 63. Foundationfor Biotechnical and Industrial Fermentation Research, Helsinki, Finland.Capon, B. (1969) Chem. Rev. 69, 407.Chauvaux, S., Beguin, P., & Aubert, 3.-P. (1992) J. Biol. Chem. 267, 4472.Csuk, R., & Glanzer, B. I. (1988) Adv.Carbohydr. Chem. Biochem 46, 73.Cupples, C. G., Miller, 3. H., & Huber, R. E. (1990) J. Biol. Chem. 265, 5512.Dale, M. P., Kopfler, W. P., Chait, I., & Byers, L. D. (1986) Biochemistry 25, 2522.Damude, H. G., Withers, S. G., Kilburn, D. G., Miller, R. C. Jr., & Warren, R. A. 3.(1995) Biochemistry 34, 2220.168Davies, G. J., Dodson, G. G., Hubbard, R. E., Tolley, S. P., Dauter, Z., Wilson, K. S.,Hjort, C., Mikkelsen, J. M., Rasmussen, G., & Schulein, M. (1993) Nature, 365 362.Delaire, M., Lenfant, F., Labia, R., & Masson, K. (1991) Protein Eng., 4, 805.Divne, C., Stahlberg, J., Reinikainen, T., Ruohonen, L., Petterson, G., Knowles, I. K.C., Teen, T., & Jones, T. A. (1994) Science 265, 524.Eshdat, Y., Dunn, A., & Sharon, N. (1974) Proc. Nati. Acad. Sci. U. S. A. 71, 1658.Fersht, A., (1985) Enzyme Structure and Mechanism. 2nd ecL, Freeman, W. H. & Co.,New York.Fischer, E., & Zemplen, G. (1910) Ber. Dtsch. Chem. 43, 2536.Gebler, J. C., Aebersold, R., & Withers, S. G. (1992) J. Biol. Chem. 267, 11126.Gilkes, N. R., Henrissat, B., Kilbum, D. G., Miller, R. C. Jr., & Warren, R. A. J.(1991) Microbiol. Rev. 55, 303.Gilkes, N. R., Langford, M., Kilburn, D. G., Miller, R. C. Jr., & Warren, R. A. J.(1984) J. Biol. Chem 259, 10455.Gilkes, N. R., Warren, R. A. J., Miller, R. C. Jr., & Kilburn, D. G. (1988) J. Biol.Chem. 263, 10401.Grace, M. E., Graves, P. N., Smith, F. I., & Grabowski, G. A. (1990) J. Biol. Chem.265, 6827.Grepinet, 0., Chebrou, M.-C., & Beguin, P. (1988) J. Bacteriol. 170, 4576.Grutter, M. G., Weaver, L. H., & Matthews, B. W. (1983) Nature 303, 828.Hall, J., Hazlewood, G. P., Barker, P. J., & Gilbert, H. J. (1988) Gene 69, 29.Hehre, E. J., Genghof, D. S., Stemlicht, H., & Brewer, C. F. (1977) Biochemistry 16,1780.Henrissat, B., & Bairoch, A. (1993) Biochem. J. 293, 781.Henrissat, B., Claeyssens, M., Tomme, P., Lemsie, L., & Mornon, J.-P. (1989) Gene81, 83.Herrchen, M., & Legler, G. (1984) Eur. J. Biochem. 138, 527.Hess, D., Covey, T. C., Winz, R., Brownsey, R., & Aebersold, R. (1993) Protein Sci.2, 1342.Hess, D., Nilca, H., Chow, D. T., Bures, E. J., Morrison, H., & Aebersold, R. (1995)Anal. Biochem. 224, 373.Hoj, P. B., Condron, R., Traeger, J. C., McAuliffe, J. C., & Stone, B. A. (1992) J.Biol. Chem. 267, 25059.169Hoim, L., Koivula, A., Lehtovaara, P. M., Hemminki, A., & Knowles, K. C. (1990)Protein Eng. 3, 181.Huber, E. R., & Chivers, P. T. (1993) Carbohydr. Res. 250, 9.Jacobsen, R. H., Zhang, X.-J., DuBose, R. F., & Matthews, B. W. (1994) Nature 369,761.Johnson, L. N., Cheetham, J., McLaughlin, P. J., Acharya, K. R., Barford, D., &Phillips, D. C. (1988) Curr. Top. Microbiol. Immuno. 139, 81Jones, C. C., Sinnott, M. L., & Souchard, I. J. L. (1977) J. Chem. Soc. Perkin TransII, 1191.Juncosa, M., Pons, J., Dot, T., Querol, E., & Planas, A. (1994) J. Biol. Chem., 269,14530.Juy, M., Amit, A. G., Alzari, P. M., Poljak, R. J., Claeyssens, M., Beguin, P., &Aubert, J.-P. (1992) Nature 357, 89.Kanda, T., Brewer, C. F., Okada, G., & Hehre, E. J. (1986) Biochemistry 25, 1159.Keitel, T., Simon, 0., Borriss, R., & Heinemann, U. (1993) Proc. Nati. Acad. Sci.U.S.A. 90, 5287.Kempton, J. B., & Withers, S. G. (1992) Biochemistry 31, 9961.Konstantinidis, A., & Sinnott, M. L. (1991) Biochem. J. 279, 587.Kortum, G., Vogel, W., & Andrussow, K. (1961) Pure App!. Chem. I, 450.Koshland, D. E. (1953) Biol. Rev. 28, 416.Lai, E. C. K., & Withers, S. G. (1994) Biochemistry 33, 14743.Leatherbarrow, R. J. (1990) GraFit Version 2.0, Erithacus Software Ltd., Staines, U. K.Lehmann, J., & Zieger, B. (1977) Carbohydr. Res. 58, 73.Legler, G. (1968) Hoppe-Seyler’s Z. Physiol. Chem. 349, 767.Legler, G. (1977) Methods Enzymol. 46, 368.Legler, G. (1990) Adv. Carb. Chem. Biochem. 48, 319.Legler, G., Roeser, K.-R., & Illig, H.-K. (1979) Eur. J. Biochem. 101, 85.Lowe, 0., & Yuthauong, Y. (1971) Biochem. J. 124, 117.Macarron, R., van Beeumen, J., Henrissat, B., de la Mata, I., & Claeyssens, M. (1993)FEBSLett. 316, 137.MacLeod, A. M., Lindhorst, T., Withers, S. 0., & Warren, R. A. J. (1994)Biochemistry 33, 6571.170Malcolm, B. A., Rosenberg, S., Corey, M. J., Allen, I. S., DeBaetselier, A., & Kirsch,J. F. (1989) Proc. Nati. Acad. Sci. U. S. A. 86, 133.Matsui, H., Tanaka, Y., Brewer, C. F., Blanchard, J. S., & Hehre, E. J. (1993)Carbohydr. Res. 250, 45.McCarter, J., Adam, M., Braun, C., Namchuk, M., Tull, D., & Withers, S. G. (1993)Carbohydr. Res. 249, 77.McCarter, J., Adam, M., & Withers, S. 0. (1992) Biochem. J. 286, 721.Miao, 5., McCarter, J. D., Grace, M., Grabowski, G., Aebersold, R., & Withers, S. G.(1994) J. Biol. Chem., 269, 10975.Miao, S., Ziser, L., Aebersold, R., & Withers, S. G. (1994) Biochemistry 33, 7027.Moult, J., Eshdat, Y., & Sharon, N. (1973) J. Mol. Biol. 75, 1.Naider, F., Bohak, Z., & Yariv, J. (1972) Biochemistry 11, 3202.ONeill, G., Goh, S. H., Warren, R. A. J., Kilburn, D. G., & Miller, R. C. Jr. (1986)Gene 44, 325.Ong, E., Gilkes, N. R., Miller, R. C. Jr., Warren, R. A. J., & Kilbum, D. G. (1993)Biotechnology and Bioengineering 42,401.Pilz, I., Schwarz, E., Kilbum, D. G., Miller, R. C. Jr., Warren, R. A. J., & Gilkes, N.R. (1990) Biochem. J. 271, 277.Post, C., & Karplus, M. (1986) J. Am Chem. Soc. 108, 1317.Py, B., Bortoli-German, I., Haiech, J., Chippaux, M., & Barras, F. (1991) Protein Eng.4, 325.Quiocho, F. A., Wilson, D. K., & Vyas, N. K. (1989) Nature 340, 404.Riddle, B., & Jencks, W. P. (1971) J. Biol. Chem. 246, 3250.Ritchie, C. D. (1972) Acc. Chem. Res. 5, 348.Robinson, R. A., Davis, M. M., Paabo, M., & Bower, V. E. (1960) J. Res. Nazi. Bur.Stand. Sect. A64, 347.Roeser, K.-R., & Legler, G. (1981) Biochim. Biophys. Acta 657, 321.Rouvinen, J., Bergfors, T., Teen, T., Knowles, J. K. C., & Jones, T. A. (1990)Science 249, 380.Schowen, R. L. (1978) in Transition States ofBiochemical Processes (Gandour, R., &Schowen, R.L., Eds.) p 77, Plenum Press, New York.Shaanan, B., Lis, H., & Sharon, N. (1991) Science 254, 862.171Shulman, M. L., Shiyan, S. D., & Khorlin, A. Y. (1976) Biochim. Biophys. Acta 445,169.Sierks, M., Ford, C., Reilly, P., & Svensson, B. (1990) Protein Eng. 3, 193.Sierks, M. R., Bock, K., Refn, S., & Svensson, B. (1992) Biochemistry 31, 8972.Silverman, R. B. (1988) Chemistry and Enzymology 1, 3.Sinnott, M. L. (1978) FEBS Lett. 94, 1.Sinnott, M. L. (1987) in Enzyme Mechanism (Page, M. I., Williams, A., Eds.) p 259,Royal Society of Chemistry, London.Sinnott, M. L. (1990) Chem. Rev. 90 1171.Sinnott, M. L., & Souchard, I. J. L. (1973) Biochem. J. 133, 89.Sinnott, M. L., & Withers, S. G. (1974) Biochem. J. 143, 751.Spezio, M., Wilson, D. B., & Karplus, P. A. (1993) Biochemistry 32, 9906.Srinivasan, K., Konstantinidis, A., Sinnott, M. L., & Hall, G. (1993) Biochem. J. 291,15.Street, I. P., Kempton, J. B., & Withers, S. G. (1992) Biochemistry 31, 9970.Street, I. P., Rupitz, K., & Withers, S. G. (1989) Biochemistry 28, 1581.Strynadka, N. C. J., & James, M. N. G. (1991) J. Molec. Biol. 220, 401.Svensson, B., & Sogaard, M. (1993) J. Biotechnol. 29, 1.Thomas, E. W. (1977) Methods Enzymol. 46, 362.Tull, D., Miao, S., Withers, S. G., & Aebersold, R. (1995) Anal. Biochem. 224, 509.Tull, D., & Withers, S. G. (1994) Biochemistry 33, 6363.Tull, D., Withers, S. G., Gilkes, N. R., Kilburn, D. G., Warren, R. A. J., & Aebersold,R. (1991) J. Biol. Chem. 266, 15621.Tull, D. (1991) M. Sc. Thesis. University of British Columbia.Umezerika, G. M. (1988) Biochem. J. 254, 73.Varghese, J. N., Garrett, T. P. J., Colman, P. M., Chen, L., Hoj, P. B., & Fincher, G.B. (1994) Proc. Nat!. Acad. Sd. USA 91, 2785.Viratelle, 0. M., & Yon, J. M. (1980) Biochemistry 19, 4143.Wang, Q., Trimbur, D., Graham, R., Warren, R. A. J., & Withers, S. G. (1995)Biochemistry. In press.172Wang, Q., Tull, D., Meinke, A., Gilkes, N. R., Warren, R. A. J., Aebersold, R., &Withers, S. G. (1993) J. Biol. Chem. 268, 14096.Weber, J. P., & Fink, A. L. (1980) J. Biol. Chem. 255, 9030.Weis, W. I., Drickamer, K., & Hendrickson, W. A. (1992) Nature 360, 127.Weiser, W., Lehmann, J., Matsui, H., Brewer, C. F., & Hehre, E. J. (1992) Arch.Biochem. Biophys. 292, 493.Wentworth, D. F., & Wolfenden, R. (1974) Biochemistry 13, 4715.White, A., Tull, D., Withers, S. G., & Rose, D. R. (1995) In preparation.White, A., Withers, S. G., Gilkes, N. R., & Rose, D. R. (1994) Biochemistry 33, 12546Withers, S. G., & Aebersold, R. (1995) Protein Sci 4, 361.Withers, S. G., Dombroski, D., Berven, L. A., Kilburn, D. G., Miller, R. C. Jr.,Warren, R. A. 3., & Gilkes, N. R. (1986) Biochem. Biophys. Res. Commun. 139, 487.Withers, S. G., MacLennan, D. J., & Street, I. P. (1986) Carbohydr. Res. 154, 127.Withers, S. G., Percival, M. D., & Street, I. P. (1989) Carbohydr. Res. 187, 43.Withers, S. G., Rupitz, K., Trimbur, D., & Warren, R. A. J. (1992) Biochemistry 31,9979.Withers, S. G., & Street, I. P. (1988) J. Am. Chem. Soc. 110, 8551.Withers, S. 0., Street, I. P., Bird, P., & Dolphin, D. H. (1987) J. Am. Chem. Soc. 109,7530.Withers, S. G., Street, I. P., & Percival, M. D.(1988) in Fluorinated carbohydrates:Chemical and Biochemical Aspects (Taylor, N. F., Ed.) p 59, American Chemical Society,Washington, D. C.Withers, S. G., Tull, D., Gebler, 3., Braun, C., Aebersold, R., Wang, Q., Warren, R. A.J., Kilburn, D. G., & Gilkes, N. R. (1993) inMechanistic Studies on Cellulases.Trichodermareesei Cellulases and other Hydrolases (Suominen, P., & Reinikainen, T.,Eds.) p 117. Foundation for Biotechnical and Industrial Fermentation Research, Helsinki,Finland.Withers, S. G., Warren, R. A. J., Street, I. P., Rupitz, K., Kempton, J. B., &Aebersold, R. (1990) J. Am. Chem. Soc. 112, 5887.Wold, F. (1978) Methods Enzymol. XLVI, 3.Xu, G.-Y., Ong, E., Gilkes, N. R., Kilburn, D. G., Muhandiram, D. R., Harris-Brandts,M., Carver, J. P., Kay, L. E, & Harvey, T. S. (1995) Biochemistry 34, 6993.Yague, E., Beguin, P., & Aubert, J.-P. (1990) Gene 89 61.Yariv, J., Wilson, K. I., Hildersheim, J., & Blumberg, H. (1971) FEBS Lett. 15, 24.Yuan, J., Martinez-Bilbao, M., & Huber, R. E. (1994) Biochem. J. 299, 527.Zhang, Z.-Y., Wang, Y., & Dixon, J. E. (1994) Proc. Nat!. Acad. Sci. U. S. A. 91,1624.Ziser, L., Setyawati, I., & Withers, S. G. (1995) Carbohydr. Res., In press.173174APPENDIX AGRAPHICAL REPRESENTATION OF KINETIC DATA4O012001000800S600—40C2000A-i Absorbance versus Time Plots for Hydrolysis of ONPC with theGlu233Asp Mutant.175CCV.00.50.40.30.20.10.00—s•p .1a.a0A[ONPC] = 4.175 mM[ONPC] = 2.250 mM[ONPCJ = 1.285 mM[ONPCJ = 0.964 mM[ONPC] = 0.642 mM[ONPCJ = 0.32 1 mM[ONPC] = 0.161 mM50 100Time150.-E,___—”•2 •1 0 1 2 3 4 31! I8ONPC) (.M)6 7A-2 Lineweaver-Burk Plots for the Hydrolysis of Aryl -Xylosides byNative C. fimi Exoglycanase.176403632282420$1612840. I I I I I I I I I IE>7’.I-0.2 .0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.81/12,3 DNPXJ (mM’)rE] = 0.00035mg/mi=420nm;=5.4-4mM3cm’6050rE] = 0.00035mg/mi=400nm:= 11.05 rM cm3020100-0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25 0.3I44CNPX] (mM4)tEl = 0.11mg/mi=270nm;&=3,10mM cni1141210J8$6420.0.2 -0.1 0 01 0,2 0.3 0.4 0.5 0.6 0.7 0.81112,5 DNPX] (mM’)fE] = 0.00070mg/mi= 440 nm; & • 5.44 mM’ cm1go706050403020100CE403020100.0.2 .0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.61/[3,4 DNPX] (mM’1) 14PNPX] (mMj70605040fE] = 0.0069mg/mi= 400 nm; & = 7.28 mM’1 cm’1177A-3 Lineweaver-Burk Plots for the Hydrolysis of Aryl f3-Xylobiosides byNative C. fimi Exoglycanase.4003002001000-60 -40 -20 0 20 40 60 80 100 120 140 1601/[2,5-DNPX2] (mM’)[E) = 7.36 x iO mg/mi=440nm;=4.29 mMcm10 100 200IJfPNPX2] (mM’)fE] = 1.56 x 10 mg/mi= 400 nm: = 7.28 mM’ c&-500 50 W0150200250300350400143,4 DNPX2J (mM4)fE] = 7.36 x 1O mg/i)=40Onm;&= 11.09 mMcm’IILONPX2] (nM’)[E]= 1.16x 1O mg/mi.—400nm;&=2.17 mM’cni’I250200150100500806040200200e100500300 400 -10 0 10 20 30 40A-4 Absorbance versus Time Plots for Pre-Steady StateAnalysis ofPNPC with the Native Enzyme.178‘.2,,IIt551Sad-‘?____________________________________________Fit Li*t,‘-4”1.I.wbaaci aidLgct-?-’ Ii..33 FrIM i., 12.152LtI4 Paut,rgtPNPC) =73I.Na4 ,IFZt •ø i.i .t! LIflI. - - 1r1a,,.-W’...f - . 437f.I Fit t*.s‘—4”bl.Rs.rhaac. I.r4is,’—?’ Iv,IC i.t’ 3 3S” j:t41. rar;’’.PNPC) =6.1M:::::k 1i (f,M14$...s, •.e e.w e.. •.sI C4• •6S.—I..”Sisal. I1,.UatiI .th peua stat. t4flF1i.• frui_i;si.tLLM*4i(4L.-Ii‘.31.i._I1l1 •tat.I.I,,•. ta_b.Ibs.rIaa Svdise’’-’ Si9ij4 rrij4 3w t.5;5.IPNPC1 = 3.52eli,tiP.’(I.saJ44)1mtc.tTi., c-—’I 5I •. I 2 P t i.NPiCIiV5PdariaiCa-,__-“i•_-i r_i •JprVI fit LimesS 54 SIPS 5.121 5.165 I.l2.54E2. V hr.alas,gI •UiU:.— 1.4516-7•I fit LimitsiflI1t ix itiI siti. st•oø stat, £455f.ictig F)ILZP—P(Z’Z)+P(3aXP(4)i.I.i1gt..ic. Uagis’’-’ ic_I? Fri,14 3 1553.i5:14:fl Pv.t.,sV(ii7l[PNPC3 = 4....f __—s.lI! •_ V•1LbI Ta., (condi)•.st e.tzt •.ieS. 545liMit rntl 5t,a4 stat.Emit:.,. P(j)*tZ-Cs)sX,P(3)’54(4)5125,Ta.. (aesadLi0.0250.02. 0.0150.010.0050-01 0 0.1 02 0.3 0.4 0.5fgt.aLiit st,i stpt.(—P(..J.Z4F(3)s64P(4 MPNP (M’)= 0.28 mg/nil=400nm;&= 10.91 cm’0.0040.002-01 0 0.1 0.2 0.3 0.4 0.5 0.6 0.71112,5 DNPC] (mM1)[E] 0.71 mg/miA = 440 nm: E = 4.29 mM cm0.0040.004 0 0.04 0.08 0.12 0.16 0.2 0.24 0.28I4ONPC] (mM’)1IfE] = 0.28 mg/mIA400 nm; &= 11.05 mM’ cm0.0250.020.0150.010.0050-0.1 0 0.1 0.2 0.3 0.4 0.5L4PNPCJ (mM’)fE] = 0.40 mg/miA = 400 nm; = 7.28 mM cni1fE] = 1.30 mg/mIA=400nm;&=2.17mM’ cui0.0120.011790.0080.006A-5 Lineweaver-Burk Plots of the Pre-Steady State Analysis of Aryl f3-Cellobiosides with Native C. fimi Exoglycanase.0.020.0160.0120.0080.00400,0040.00200 0.4 0.8 1.2 1.6 2142,4.DNPCJ (mM”)0.010 0.4 0.8 1.2• 143,4 DNPC] (mM’)1.60.008000600.016001200081400012000.5j8000— 400020000-04 0 04 0.8 1.2 1.6 2 2.4 2.81! [3,441.DNPG] (mM)’fE] = 0.20 mg/mI= 400 nm. = 11.05 mM cm24002000a160012008004000-0.05 0 0.05 0.1 0.15 0.2 0.251/ L 2.CI-4-NPG] mM1.[El = 0.20 mg/mi= 400 rim, = 7.52mM1cr1’5004003002001000.50 0 50 100 150 200 250 300 350 4001/IPNPX2]E] = 0.083 mg/mIA=400nm,&=7.28M1c ’180A-6 Lineweaver-Burk Plots for the Hydrolysis of Aryl 3-Glucosides andPNPX2 by the Glu233Asp Mutant of C. fimi Exoglycanase.10001000800.5600;4002000$006004002000-0.8 .0.4 0 0.4 0.8 1.2 1.6 2I I (2,S$-DNPGJ (mM )4[E] = 0.14 mg/mi= 440 rim, & = 4.29 mM cr1’-1 -0.5 0 0.5 1 1.5 2 2.51i[B-DNPG] (mM’)[El = 0.31 mg/mi=400nm,&= 10.91 mM’ cn1I. 1.I.I.I7 I,/181A-7 Lineweaver-Burk Plots for the Hydrolysis of Aryl -Ceilobiosides bythe Glu233Asp Mutant of C. fimi Exoglycanase.16001400;12001 10006004002000140012001 1000800S600— 400200-2 -1 0 1 2 3 4 5 6 71/18-ONPC)(UIW)tE] = 0.12 mg/mI=400nm.E=2.17mM’cm1000‘gaoe 600400200-2 0 2 4ULB.PNPC] (mM4)fE] 0.305 mg/miA • 400 rim, = 7.28 mM’ cm1:/30002500f 2000150010005000-1 0 1 2 3 4 5 6M4-CNPCI (mM’)fE) = 0.12 mg/miA = 270 nm, = 3.10 mM cm1.1 0 1 2 31! 14-BrPCJ (mM’)fE]O.12mgIm1A — 288 nm, & • 0.68 mM’ cm800j6004002000.10 .5 0 5 10 15 20 25 301112A-DNPCJ (mM’)[E] = 0.16 mg/miA=400nm, = 10.91 mM1 ctn.10 0 10 20 30 40 50143A-DNPCJ (m.M’)fE] = 0.17 mg/mIA.400nm,àe= 11.05 mM1cni12000 0620001500I’:04 5182A-8 Lineweaver-Burk Plots for the Hydrolysis of Aryl -Cellobiosides by180016001400f 12001000. 8006004002000160012008004000the Glul27Ala Mutant of C. fimi Exoglycanase.-60 -40 -20 0 20 40 60 80 100 120 1401/[ONPC] (mM’).6 -4 -2 0 2 42113,5 DCIPC] (mM4)[El = 0.20 mg/mI= 280 nm; = 0.732 riM1 cm60005000f40003000. 200010000-2 -1 0 1 2 3 4 5 6 71113,4 DNPC] (mM’)fE] = 0.0010 mg/mi= 400 nm; & = 11.05 mM cm1.60 .40 .20 0 20 40 60 80 100 120 14014PNPC] (mM1)400350300I250200i0[E]= 0.040 mg/mi= 400 nm; & = 7.28 mM cm160140120E 100806040200144 CNPCJ (,,f4•1)[El = 0.20 mg/mI= 270 nm; & = 3.10 mM cm.2 -1 0 1 2 3 4fE] = 0(140 mg/mi= 400 nm; & = 217 mM1 cm2000I6A-9 Lineweaver-Burk Plots of the Pre-Steady State Analysis of Aryl -Cellobiosides with the Glul27Ala Mutant of C. fimi Exoglycanase.183‘liii I’I’I’ ‘i’i’6’<liIi-, I I i0.60 -50 .40 .30 .20 .10 0 10 20 30 40 50143,4 DNPCj[E] = 0.20 mg/mI).=400nrn;b.= 11.09 mM1cm0.60.40.20.0240.020.016I0.0120.0080.0040-2 -1 0 1 2 3 4 5 6142,4 DNPC] (mM’)EJ = 0.27 mg/mi= 360 nm .€ = 14.0 mM98__63I0-6 -4-2 0 2 4 6 8 101/[PNPC] (mM4)fE] = 0.39 mg/mIA.=400nm,7.28 mM1cm’184gEEA-1O Inactivation-Reactivation Kinetics of C. fimi Exoglycanase.300g2001000 50 100Time (mm)150 0.000.2 0.4 0.6 0.8 1.0 1.21f(cellobial) (1/mM)CC0 10 20 30Ecellobial] (mM)40 50 60 0 20 40 60 80 100 120Time (mm)Figure A-b-I Inactivation of C.fimi exoglycanase by cellobial. (A) Semi-logarithmic plotof residual activity versus time ((A), 1.0 mM; (•) 7.1 mM; (•) 10.0 mM; (1)29.9 mM;(0)542 mM), (B) double reciprocal plot ofpseudo-first order rate constantsfrom theupper plot; (C) plot of the pseudo-first order rare constants versus cellobial concentration;and (D) protection against cellobial inactivation in the presence ofGTBX ((s) 0 mM and(0)2mM).ECTime (mm)2FGMF (mM)Figure A-10-2 Inactivation of C.fimi exoglycanase by 2F-GMF. (A) Semi-logarithnic plotof residual activity versus time ((•), 4.1 mM; () 8.2 mM; (M), 14.4 mM; (0), 24.6 ,nM,(A), 30.82 mM); (B) double reciprocal plot ofpseudo-first order rate constants from theupper plot; (C) plotof the pseudo-first order rate constants versus 2F-GMF concentration;and (D) protection against 2F-GMF inactivation in the presence ofGTBX ((•) 0 mM and(0)2mM).185CI.g0 bC) 200 300 400 5(X) 600Time (mm)0.0 0.1 0.21/[2FGMF] (1/mM)0.3o00E10 20 30 40 0 200 300186ICo io 20 30 40 50 60 70 80 90Time (mm)0.2 0.6 1.01I[I] (1/mM)CC1.4Figure A-10-3 Inactivation of C.fimi exoglycanase by N-bromoaceryl cellobio.sylamine.(A) Semi-logarithmic plot of residual activity versus time ((I),0.8 mM; () 1.6 mM; (4),32 mM, (•), 4.0 mM; (0), 4.8 mM) (Black et al., 1993), (B) double reciprocal plot ofpseudo-first order rate constantsfrom the upper plot; and (C) protection againstNbromoacetyl cellobiosylamine inactivation in the presence ofGTBX ((4)0 mM and (A) 2mM).0 50 100Time (mm)150187Figure A-b-S Reactivation of cellobial-inactivated C.fimi Exoglycanase. Plots of rateversus time. Buffer only (0) and 50 mM cellobiose (ê.0 100 200 300 400 500 600.2.4-2.6-2.8.3.0>g.3.2.3.4-3.6Time (h)Figure A -10-4 Reactivation of2F-DNPX-inactivated C. fimi Exoglycanase. Plots ofln(full rate - observed rate) versus time. Buffer (A) only and 50 mM cellobiose (•).0.140.120.10.080.0200 20 40 60 80Time (h)188APPENDIX BBASIC CONCEPTS OF ENZYME CATALYSIS189B-i Basic Enzyme KineticsThe basic equation of enzyme kinetics is the Michaelis-Menten equation where v isthe velocity of the reaction measured either as the initial rate of formation of the products ordepletion of the substrates; [El is the total concentration of the enzyme; [S] is the substrateconcentration; kcat is the catalytic constant; and Km is the Michaelis constant.[E0] [SI kcatv=Km + [SIIn the Michaelis-Menten equation two assumptions are made; the enzymeconcentration is negligible compared to that of the substrate which is generally true sinceenzyme catalyse reactions with a high efficiency, and the velocity measured is the initial rateof product formation, thus there is no significant accumulation of product (or depletion ofsubstrates) hence, the reverse reaction can be ignored. Therefore, the change in substrateconcentration is generally linear with time.The Km is the substrate concentration when v = Vm12 Wm = maximum velocity). Itmay be treated as an apparent dissociation constant of all the bound enzyme species and assuch is expressed as:K— [E][S]m E[ES]The value of Km can be a measure of the enzyme affinity for the substrate, for example, alow Km means that the enzyme has a high affinity for the substrate.At low substrate concentration ([SI << Km) the Michaelis-Menten equationbecomes190At low [SI most of the enzyme is unbound such that the total enzyme concentration, whichis a sum of the concentration of the free and bound enzyme, can be approximated to theconcentration of the free enzyme, [E]. The Michaelis-Menten equation under theseconditions is expressed as[El [SI kcatv=Km[E0J [SIKmwhereas at saturating concentrations ([SI >> Km) the equation becomes:V = Vm keat [E0]Figure B-i-i Plot of velocity versus substrate concentrationfor a typical enzymaticreaction (Fersht, i985).191The kcailKm from the above equation is an apparent second-order rate constantwhich ielates the reaction rate to the concentration of the ft.ç enzyme and ft substrate.This kinetic parameter is also referred to as a specificity constant which is a measure of thecatalytic efficiency for the substrate.The Michaelis-Menten equation is often changed to a linear form which is useful forgraphical analysis of the data and detection of deviations from the expected values. Anexample of the Michaelis-Menten equation transformed is where both sides of theMichaelis-Menten equation have been inverted.1=1 + KmV Vm V[S]Plotting 1/v as a function of 1/[S] gives the Lineweaver-Burk plot (Figure B-1-2) where they-intercept is i/Vm, the x-intercept is- i/Km and the slope is Km/Vm.Figure B-1-2 A typical Lineweaver-Burkplorfor an enzymatic reaction (Fersht, 1985)192B -2 Interpretation Of kcat And kcat/KmThe rate constant k which equals Vm/[E]o is a reflection of the rate determiningstep and the rate constant, kcat/Km reflects the rate of the first irreversible step in thereaction (Schowen, 1978). In order to show this, first consider the general mechanismshown below, where the formation of ES is referrred to as the association step, theinterconversion of the ES and EP as the chemical step and the final step as product-release.k k2 k3E+S ES .. EP E+Pk1 k2The corresponding free energy diagram for this mechanism is shown in Figure B-2-1where the energy levels are arbitrarily chosen. It can be shown that the kinetic parameters,keat, Km and lccat/Km, for the reaction are give by23‘cat =k2+k3 (1)Kk3(1c +k2) +k12m (k+ (2)k123Km — k3(k1+ k2) + k1 (3)GESReaction coordinateE+PFigure B-2-1 Reaction coordinate diagramfor an enzymatic reaction involving theinterconversion of intermediates.193Now, consider the above reaction where there is a rapid, reversible association ipfollowed by a te determining chemical p. The kinetic relationships that describe thissituation are k..1 >> k2, k3 >> k..2 and k3 >> k2. When these conditions are applied toequations 1 and 3, the kinetic parameters are reduced tokcat = k2kcat k12Km k1which expressed in Eyring form become:kT [-(GES -kcat — eh=.I elc*/RThEPE+SEPE.S194— .!1 esGEh1fITKmh= e L\GT/RThwhere k is the Boltzmann constant and h is Planck’s constant. Thus, under theseconditions, and ktfKm both give information pertaining to the transition state of thechemical step. However, the initial reference point for is the ES complex and forkcajKm, it is free enzyme (E) and free substrate (S).If however, the restrictions on the reaction were rapid, reversible associationfollowed by an irreversible chemical and then te determining product release, thekinetic relationships would be k1 >> k2,k2>> k3 andk3>> k2. The kinetic constants thencan be reduced tokcat =1(3kcat k12Km k1and the Eyring equations arekT [(Gp* - GEp)]/RTkcat= — e=‘ ep1Th— = e Th1TKm h195In this case, k refers to the transition state for release of the product, with the enzyme-product complex as the initial reference point, and kfK refers to the transition state ofthe chemical step with E and S as the initial reference points.Thus, these examples, show that keat refers to the rate determining step whilekcat/Km corresponds to the first irreversible step in the reaction with E and S as thereference states.These general concepts can be extended to the hydrolysis of 13-glycosides by C.JImi exoglycanase where the chemical step corresponds to glycosylation, and productrelease to deglycosylation. Thus, the first example of rapid, reversible association followedby a rate determining chemical step is equivalent to the situation when glycosylation is therate determining step and the second example of rapid reversible association, an irreversiblechemical step and then rate determining product release corresponds to the situation whendeglycosylation is rate determining.B-3 Binding Energy And Enzyme CatalysisThe function of any catalyst, including enzymes, is to lower the activation energy ofa reaction, thus leading to rate acceleration. Enzymes are known to bind specifically to theirsubstrates and the binding energies involved may be quite large. However, since thestructure of the substrate changes as it is converted to product via a transition state, theenzyme can only be fiffly complementary to one form of the substrate. It will be shown thatit is catalytically advantageous for the enzyme to be complementary to the transition statestructure rather than to the ground state of the substrate.Consider a typical enzymatic reaction consisting of a binding step and a catalyticstep such as that shown below.Km kcatE+S ..‘ E.S196The energy diagram for this reaction is illustrated in Figure B-3-1 where k..1 >> k2, so thatKm = KJ and the kinetic parameters may be expressed as the following.GKm = e -t/RTkeat e= e GT*/RTKm hE+SFigure B-3-1 The reaction coordinate diagramfor a typical reaction (solid line) and thecorresponding uncatalysed reaction (dashed line).Now, consider the above situation when an extra amount of binding energy, AGR, hasbecome available, for example a hydrogen bond between the enzyme and the substrate. Ift’Gc*E+SE.SReaction coordinate197that extra binding energy is realised at the ground state (Figure B-3-2 (left plot)) rather thanthe transition state, then the ground state is stabilised and AGES increases by AGR, thusKm is decreased. The value of kcat is also reduced since AG (Michaelis complex (ES)proceeding to the transition state (ES*)) is increased by the ground state stabilisation, AGR.However, kJKm is unaffected as this is the rate constant for free enzyme (E) and freesubstrate (S) proceeding to the transition state (ES). This shows that if the enzyme iscomplementary to the ground state, then there is tighter substrate binding but slowerreaction rate, keat. Alternatively, if the extra binding energy, AGR, is realised at thetransition state (Figure B-3-2 (right plot)), then the transition state is stabilised and AGcand A&r will be lowered by AGR, thus, keat and kcat/Km will be increased. The value ofAGES is unchanged, and hence, Km remains the same. This shows that when the enzymeis complementary to the transition state rather than the ground state, then the rate of thereaction and the enzyme efficiency (kcatlKm) is increased.G ES G*Reaction coordinateFigure B-3-2 The reaction coordinate diagram illustrating complementarity of the enzymeto (left) the ground state, and (right) the transition state of the substrate.E÷S E÷SE. SReaction coordinateE.S198B-4 Inactivation Kinetics of C. fimi ExoglycanaseA schematic representation of the mechanism of the exoglycanase-catalysedhydrolysis of substrates is the following where k2 corresponds to glycosylation and k3 todeglycosylation.k1 k2 k3E + GX - E.GX E-G E + G-OH-1 HX H20For the inactivators used, k3 << k2 and k1 >> k2, thus the covalent glycosyl-enzymeintermediate accumulates. The kinetic equation for this inactivation, is the followingvariation of the Michaelis-Menten equation,k [E0J [GX]v=+ [GX]where k is the inactivation rate constant, and K is the apparent dissociation constant for allbound enzyme species.If [GX] >> [EJ, then [GXJ appears constant throughout the inactivation processand the kinetics are pseudo-first order with respect to the enzyme concentration. Thus, theMichaelis-Menten equation can be expressed asv = kobs [E0]k [GX]k0bK1 + [GX]The value of kobs can be calculated by directly fitting the data to a first-order function orfrom the slope of the natural logarithm of the residual enzymatic activity plotted as a199function of time. That is, since the rate of formation of HX is equal to the rate ofinactivation of the enzyme thend[HX] d[E01v=dt dtThe rate can then be expressed as- d[E0]= kobS [E0]dt- d[E0]= k0b dt[ ]Therefore,in [E0j obs tValues of k and K1 can then be calculated by direct fit of values of k(i,) to the equationk [GXJkobs= K + [GX]

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.831.1-0059648/manifest

Comment

Related Items