Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Modifying the catalytic carboxylates of retaining [Beta]-glycosidases Lawson, Sherry L. 1997

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

Item Metadata

Download

Media
831-ubc_1997-196089.pdf [ 9.18MB ]
Metadata
JSON: 831-1.0059653.json
JSON-LD: 831-1.0059653-ld.json
RDF/XML (Pretty): 831-1.0059653-rdf.xml
RDF/JSON: 831-1.0059653-rdf.json
Turtle: 831-1.0059653-turtle.txt
N-Triples: 831-1.0059653-rdf-ntriples.txt
Original Record: 831-1.0059653-source.json
Full Text
831-1.0059653-fulltext.txt
Citation
831-1.0059653.ris

Full Text

MODIFYING THE CATALYTIC CARBOXYLATES OF RETAINING (3-GLYCOSIDASES by SHERRY L. L A W S O N B.Sc , University of Lethbridge, 1989 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES Department of Chemistry We accept this thesis as conforming to the required standard: THE UNIVERSITY OF BRITISH C O L U M B I A January 1997 © Sherry L. Lawson, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of ^<xU\ru The University of British Columbia Vancouver, Canada Date DE-6 (2/88) ABSTRACT n Glycosidases hydrolyze glycosidic bonds either with retention or inversion of anomeric configuration, the mechanism employed being dictated, in part, by the distance between the two key active site carboxylates. For retaining glycosidases, the average distance is 4.5-5.5 A , while for inverting glycosidases it is greater (9-9.5 A). In the retaining endo-P-o 1,4-xylanase from Bacillus circulans/subtilis, this critical distance (5.5 A) has been altered by replacing the active site nucleophile Glu78 with both a shortened and a lengthened analogue (aspartic acid and S-carboxymethyl cysteine, respectively). Shortening the nucleophilic side chain decreased k c a t / K m values at least 1600 fold for the aryl P-xylobiosides. In contrast, increasing the length (achieved by selective carboxymethylation of Cys78 of the Glu78Cys mutant) reduced these values by only 16 to 100 fold. These rate differences were not reflected in the degree of bond cleavage or proton donation at the glycosylation transition state, as demonstrated by similar p i g values (Br0nsted slopes). These results confirm the importance of precise positioning of the catalytic nucleophile at the active site of B. circulans/subtilis xylanase. The acid/base catalyst of B. circulans/subtilis xylanase, Glu 172, was substituted with a glutamine, a group with no significant capacity as a proton donor/acceptor. Removal of the carboxyl side chain eliminated activity with the natural substrate xylan, though some activity, which could be further rescued in the presence of the alternate nucleophile azide, was seen with activated substrates such as the nitrophenyl xylobiosides. In addition, Glul72 was also replaced with a shortened (aspartic acid) and lengthened (S-carboxymethyl cysteine) analogue. Both shortening and lengthening this carboxyl side chain had similar effects on iii xylan hydrolysis, with the k c a t / K m values being reduced -1000 fold relative to native xylanase. Modifying the length of the acid/base catalyst was less detrimental to the hydrolysis of aryl [3-xylobiosides. For these synthetic substrates, the k c a t / K m values were decreased only 3 to 24 fold. Again, no significant change was observed in the f3ig values, suggesting that these modifications have not seriously affected the degree of bond cleavage or proton donation at the glycosylation transition state. Thus, the precise placement of the acid/base catalyst is not as critical for the hydrolysis of aryl (3-xylobiosides. It has been suggested that a cysteine could fulfill the role of the active site nucleophile in retaining glycosidases (Hardy & Poteete (1991), Biochemistry 30, 9457). To test the validity of this proposal, a kinetic evaluation was conducted on the active site nucleophile cysteine mutants of two retaining 13-glycosidases. In the case of B. circulans xylanase, the cysteine mutant (Glu78Cys) was completely inactive, not even capable of undergoing the first step (glycosylation) of the double displacement mechanism. In contrast, the corresponding cysteine mutant (Glu358Cys) of Agrobacterium (3-glucosidase did complete the glycosylation step, but the rate constant for this step was reduced at least 2 x 106 fold relative to the native enzyme. The subsequent hydrolysis (deglycosylation) step was also severely affected by the replacement of Glu358 with a cysteine (the rate constant for this step was depressed 107 fold). Thus, Cys358 functions inefficiently in both the capacity of catalytic nucleophile and leaving group. On the basis of these results, it seems improbable that the role of the active site nucleophile in retaining glycosidases could successfully be filled by a cysteine residue. IV TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iv LIST OF TABLES viii LIST OF FIGURES ix LIST OF SCHEMES xiii ABBREVIATIONS AND SYMBOLS xv ACKNOWLEDGEMENTS xviii DEDICATION xix CHAPTER 1: GENERAL INTRODUCTION 1 1.1 Glycosidases and the reaction they catalyze 2 1.2 Classification of glycosidases 2 1.3 The catalytic mechanism of retaining glycosidases 4 1.4 Evidence for a double displacement mechanism in retaining glycosidases 6 1.4.1 Presence of a carboxylate nucleophile and acid/base catalyst 6 1.4.1.1 X-ray crystallographic studies 6 1.4.1.2 Labelling of the catalytic carboxylates using carbodiimides 8 1.4.1.3 The epoxy glycosides and conduritol epoxides 10 1.4.1.4 The 2-deoxy-2-fluoro glycosides 12 1.4.1.5 A strategy for identifying the acid/base catalyst 13 1.4.2 The nature of the glycosyl-enzyme intermediate 14 1.4.3 Oxocarbenium ion-like transition states 18 1.4.4 General acid catalysis 20 1.4.5 Noncovalent enzyme-substrate interactions .24 1.5 The proposed mechanisms for retaining and inverting glycosidases 26 1.6 Aims of this thesis 28 CHAPTER 2: MODIFYING THE ACTIVE SITE NUCLEOPHILE OF BACILLUS CIRCULANSISUBTILIS XYLANASE 31 2.1 Introduction \ 32 2.1.1 Xylan and xylan-degrading enzymes 32 2.1.2 Endo-p-1,4-xylanases 33 2.1.3 The endo-p-1,4-xylanases from Bacillus circulans and Bacillus subtilis 34 2.2 Specific aims of this study 38 2.3 Results and discussion 39 2.3.1 Preparation and preliminary characterization of native xylanase, Glu78Cys, and Glu78Asp 39 2.3.2 Kinetic analysis of the Glu78Cys mutant 41 2.3.3 Preparation and characterization of carboxymethylated Glu78Cys 44 2.3.3.1 Iodoacetate labelling of the Glu78Cys mutant under nondenaturing conditions 45 2.3.3.1.1 Search for the optimal labelling conditions 45 2.3.3.1.2 Active site titration using 2',4'-dinitrophenyl 2-deoxy-2-fluoro-p-xylobioside (2F-DNPX2) 48 2.3.3.1.3 Methyl methanethiolsulfonate labelling of Glu78Cys and subsequent IAA treatment 49 2.3.3.2 Iodoacetate labelling of the Glu78Cys mutant under partially denaturing conditions 51 2.3.3.3 Iodoacetate labelling of the Glu78Cys mutant under fully denaturing conditions 53 2.3.3.3.1 Preparation of IAA-Glu78Cys 53 2.3.3.3.2 Active site titration using 2F-DNPX2 54 2.3.3.3.3 Identification of the 2-fluoroxylobiosyl labelled active site peptide by ESMS 56 2.3.4 Kinetic evaluation of IAA-Glu78Cys, native xylanase, and Glu78Asp 61 2.3.4.1 Kinetic analysis of the 2F-DNPX2 inactivation of IAA-Glu78Cys 61 2.3.4.2 Reactivation of the inactivated IAA-Glu78Cys mutant 63 2.3.4.3 Determination of the stereochemical course of hydrolysis for IAA-Glu78Cys 65 2.3.4.4 Steady state kinetic studies using synthetic (3-xylobioside substrates 67 2.3.4.5 The pH dependence of k c a t / K m 71 2.4 Conclusion 73 CHAPTER 3: MODIFYING THE ACID/BASE CATALYST OF BACILLUS CIRCULANS/SUBTILIS XYLANASE 77 3.1 Introduction 78 3.2 Specific aims of this study 79 3.3 Results and discussion : 81 3.3.1 Preparation and preliminary characterization of native xylanase and the Glul72Cys, Glul72Asp, and Glul72Gln mutants 81 3.3.2 Preparation and characterization of the IAA-Glul72Cys mutant 84 3.3.2.1 Iodoacetate labelling of the Glul72Cys mutant under fully denaturing conditions 84 3.3.2.2 Identification of the iodoacetate labelled peptide in digested IAA-Glul72Cys 85 3.3.3 Kinetic evaluation of native xylanase and the Glul72 mutants 90 3.3.3.1 Steady state kinetic studies using the natural substrate xylan 90 3.3.3.2 Steady state kinetic studies using synthetic p -xylobioside substrates.... 92 3.3.3.3 The effects of exogenous nucleophiles on reaction rates and products .. 96 3.3.3.4 The Br0nsted relationships for native xylanase, IAA-Glul72Cys, and Glu 172 Asp 100 3.3.3.5 The pH dependence of k c a t / K m 104 3.3.4 The inactivation of Glul72Cys by di-P,P'-D-xylopyranosyl disulfide 109 vi 3.3.4.1 The effects of pH on the inactivation rate 110 3.3.4.2 Determination of the inactivation parameters kj and Kj 110 3.3.4.3 The proposed mechanism of inactivation 113 3.4 Conclusion 115 CHAPTER 4: INVESTIGATION OF THE GLU358CYS MUTANT OF AGROBACTERIUM p-GLUCOSIDASE 117 4.1 Introduction 118 4.2 Specific aim of this study 121 4.3 Results and discussion 122 4.3.1 Preparation and preliminary characterization of native (3-glucosidase and the Glu358Cys mutant 122 4.3.2 Kinetic evaluation of the Glu358Cys mutant 123 4.3.2.1 Determination of the kinetic parameters for the hydrolysis of 2',4'-dinitrophenyl (3-D-glucopyranoside 125 4.3.3 Analysis of glucosylated Glu358Cys 129 4.3.3.1 Treatment of the Glu358Cys and Glu358Gly mutants with excess 2,4-DNPG: ESMS analysis and thiol titrations 129 4.3.3.2 Identification of the glucosylated amino acid in 2,4-DNPG-treated Glu358Cys 132 4.3.3.3 Urea denaturation of glucosylated Glu358Cys 137 4.4 Conclusion 139 CHAPTER 5: MATERIALS AND METHODS 140 5.1 Materials 141 5.2 General Procedures 141 5.2.1 Enzyme assays and kinetic studies 141 5.2.2 Thiol titrations 142 5.2.3 Circular dichroism studies 142 5.2.4 Electrospray mass spectrometry studies 143 5.3 Investigation of native xylanase and mutants 144 5.3.1 Enzyme isolation and purification 144 5.3.2 Chemical modification of the Glu78Cys and Glul72Cys mutants 145 5.3.2.1 Methyl methanethiolsulfonate (MMTS) labelling of Glu78Cys 145 5.3.2.2 IAA labelling of Glu78Cys under nondenaturing conditions 145 5.3.2.3 IAA labelling of Glu78Cys under partially denaturing conditions 146 5.3.2.4 IAA labelling of Glu78Cys and Glul72Cys under fully denaturing conditions 146 5.3.3 Kinetic studies using synthetic substrates 147 5.3.3.1 Determination of extinction coefficients 147 5.3.3.2 A pre-steady state kinetic study of native xylanase with 2,5-DNPX2-.. 148 5.3.3.3 Steady state kinetic studies using synthetic (3-xylobioside substrates... 148 5.3.3.4 The effects of exogenous nucleophiles on reaction rates and products 150 5.3.3.5 The pH dependence of k c a t / K m 151 5.3.3.6 Determination of the stereochemical course of hydrolysis 152 5.3.4 Steady state kinetic studies using the natural substrate xylan 152 5.3.5 Inactivation studies using 2F-DNPX 2 153 5.3.5.1 Glu78Cys + 2F-DNPX 2 : contamination check 153 5.3.5.2 Active site titration of the IAA-labelled Glu78Cys mutant using 2F-DNPX 2 154 5.3.5.3 Determination of the k; and K, values for the inactivation of IAA-Glu78Cys by 2F-DNPX 2 154 5.3.5.4 Reactivation of 2F-DNPX 2 inactivated IAA-Glu78Cys 155 5.3.6 Inactivation studies using di-p\ fi'-D-xylopyranosyl disulfide 156 5.3.6.1 The effects of pH on inactivation rates 156 5.3.6.2 Determination of the inactivation parameters k, and Kj 157 5.3.7 ESMS analysis of digested 2F-DNPX 2 inactivated IAA-Glu78Cys 157 5.3.8 ESMS analysis of digested IAA-Glul72Cys and native xylanase 158 5.3.9 Molecular modelling of selected xylanase mutants 159 5.4 Investigation of the Glu358Cys mutant of (3-glucosidase 160 5.4.1 Enzyme isolation and purification 160 5.4.2 Glu358Cys + 2,4-DNPG: determination of kinetic parameters 161 5.4.3 Glu358Cys + 2F-DNPG: contamination check 162 5.4.4 Analysis of glucosylated Glu358Cys 163 5.4.4.1 Treatment of Glu358Cys and Glu358Gly with excess 2,4-DNPG: ESMS analysis and thiol titrations 163 5.4.4.2 ESMS analysis of digested glucosylated Glu358Cys 163 5.4.4.3 Urea denaturation of glucosylated Glu358Cys 165 REFERENCES 166 APPENDIX A: ENZYME KINETICS 174 APPENDIX B: INTRODUCTION TO B R 0 N S T E D RELATIONSHIPS 183 APPENDIX C: GRAPHICAL REPRESENTATION OF KINETIC DATA 186 APPENDIX D: SUPPLEMENTARY ESMS DATA 195 viii L I S T O F T A B L E S Table 2-1. The family G xylanases for which the catalytic residues have been identified .... 34 Table 2-2. The I A A labelling of Glu78Cys under different conditions 46 Table 2-3. Thiol titration and ESMS results for MMTS-treated enzyme 50 Table 2-4. The inactivation constants for IAA-Glu78Cys and native xylanase 61 Table 2-5. Kinetic parameters for aryl P-xylobiosides with native xylanase and mutants 69 Table 3-1. ESMS analysis of native xylanase and the Glul72 mutants 82 Table 3-2. Kinetic parameters for soluble xylan with native xylanase, IAA-Glul72Cys, and the Glul72Asp mutant 91 Table 3-3. Kinetic parameters for aryl P-xylobiosides with Glul72Cys, Glul72Gln, and native xylanase 94 Table 3-4. Kinetic parameters for aryl P-xylobiosides with IAA-Glul72Cys, Glul72Asp, and native xylanase 95 Table 3-5. The determined slopes and correlation coefficients for Figure 3-7 101 Table 3-6. The determined slopes and correlation coefficients for Figure 3-8 103 Table 3-7. The determined pK a values for native xylanase and the Glu 172 mutants 106 Table 3-8. Thiol titration and ESMS results for X S S X treated enzymes 114 Table 4-1. The pre-steady state kinetic parameters for the hydrolysis of 2,4-DNPG by Glu358Cys and native p-glucosidase 128 Table 4-2. Results of the modified Edman degradation of the glucosylated peptide 136 IX LIST OF FIGURES Figure 2-1. The three-dimensional structure of B. circulans xylanase, showing the (3-sheets and a-helix 36 Figure 2-2. The active site of B. circulans xylanase shown (a) in the absence of substrate and (b) for the Glul72Cys mutant complexed with xylotetraose 37 Figure 2-3. The circular dichroism spectra of native xylanase, Glu78Cys, and Glu78Asp . 40 Figure 2-4. Regain of activity upon IAA labelling of Glu78Cys and native xylanase (36 °C, pH 10, CAPSO) 47 Figure 2-5. Treatment of IAA-labelled Glu78Cys with 2F-DNPX 2 . The time-dependent inactivation of labelled Glu78Cys is shown in curve a, while the release of 2,4-DNP is shown in curve b 48 Figure 2-6. IAA treatment of MMTS-modified Glu78Cys and a Glu78Cys control 50 Figure 2-7. The effects of low urea concentrations on the IAA labelling of Glu78Cys at pH 10 and pH 9 52 Figure 2-8. The CD spectra of refolded IAA-treated Glu78Cys and native xylanase. The CD spectrum of untreated native xylanase was run as a control 54 Figure 2-9. The active site titration of IAA-Glu78Cys (prepared under fully denaturing conditions) 55 Figure 2-10. ESMS analysis of digested 2F-DNPX 2 inactivated IAA-Glu78Cys (prepared under fully denaturing conditions) 58 Figure 2-11. The MS/MS daughter ion spectrum of the 2-fluoroxylobiosyl labelled peptide59 Figure 2-12. The inactivation of IAA-Glu78Cys (prepared under fully denaturing conditions) by 2F-DNPX 2 62 Figure 2-13. The reactivation of inactivated IAA-Glu78Cys by xylobiose and buffer 64 Figure 2-14. Determination of the stereochemical course of hydrolysis of 2,5-DNPX 2 by IAA-Glu78Cys (prepared under fully denaturing conditions) 66 Figure 2-15. The Br0nsted plot (log k c a t / K m versus aglycone p K a ) for native xylanase, the Glu78Asp mutant, and the IAA-Glu78Cys mutant prepared under fully denaturing conditions 70 Figure 2-16. The pH dependence of k c a t / K m for native xylanase and IAA-Glu78Cys prepared under fully denaturing conditions 72 Figure 3-1. The CD spectra of native xylanase and the Glu 172 mutants 83 Figure 3-2. ESMS analysis of CNBr digested IAA-Glu 172Cys and native xylanase 88 Figure 3-3. The MS/MS spectrum of the IAA-labelled peptide in CNBr digested IAA-Glu 172Cys 89 Figure 3-4. The hydrolysis of soluble birchwood xylan by Glul72Asp, Glul72Cys, and refolded IAA-treated Glul72Cys 91 Figure 3-5. The effects of azide on xylanase-catalyzed hydrolysis rates for 2,5-DNPX2 and ONPX2. The enzymes examined were native xylanase, Glul72Asp, Glul72Gln, Glul72Cys, and refolded IAA-treated Glul72Cys 97 Figure 3-6. The effects of neutral nucleophiles on rates of native xylanase-catalyzed hydrolysis of 2,5-DNPX2. The nucleophiles tested were p-mercaptoethanol, DTT, and MeOH 100 Figure 3-7. The log k c a t versus p K a plot for native xylanase, Glul72Asp, and IAA-Glu 172Cys 101 Figure 3-8. The log k c a t / K m versus pK a plot for native xylanase, Glul72Asp, and IAA-Glu 172Cys 103 Figure 3-9. The pH dependence of k c a t / K m for native xylanase, Glul72Asp, Glul72Cys, and IAA-Glul72Cys 105 Figure 3-10. The active site of Bacillus circulans xylanase, showing the key residues and hydrogen bonding interactions 108 Figure 3-11. The effects of pH on the inactivation of Glul72Cys by X S S X I l l Figure 3-12. The inactivation of Glul72Cys by X S S X at pH 10 112 Figure 3-13. The activity loss of Glul72Cys and native xylanase upon treatment with XSSX (500 fold molar excess, pH 10, 40 °C) 115 Figure 4-1. The CD spectra of native (3-glucosidase and the Glu358Cys mutant 123 Figure 4-2. The hydrolysis of 2,4-DNPG by the Glu358Cys mutant 126 Figure 4-3. ESMS analysis of the Glu358Cys and Glu358Gly mutants after treatment with 2,4-DNPG (74 fold molar excess, 37 0 C, 24 hrs) 131 xi Figure 4-4. ESMS analysis of digested, 2,4-DNPG-treated Glu358Cys and its control 133 Figure 4-5. The mass spectrum of the 311 PTH derivatized amino acid released in cycle 4 of the modified Edman degradation procedure 137 Figure 4-6. The urea denaturation curves for the Glu358Cys mutant and the glucosylated mutant 138 Figure A - l . The plot of rate versus substrate concentration for a typical enzymatic reaction obeying Michaelis-Menten kinetics 168 Figure A-2. A typical Lineweaver-Burk plot for an enzymatic reaction 168 Figure A-3. The hypothetical Gibbs free energy diagram for a retaining glycosidase showing rate-limiting deglycosylation 172 Figure B - l . The Br0nsted plots for the hydrolysis of aryl (3-glucosides by C.fimi exoglycanase 177 Figure C - l . The Lineweaver-Burk plots for the hydrolysis of aryl P-xylobiosides by native xylanase 178 Figure C-2. The Lineweaver-Burk plots for the hydrolysis of aryl p-xylobiosides by the IAA-Glu78Cys mutant 179 Figure C-3. The Lineweaver-Burk plots for the hydrolysis of aryl P-xylobiosides by the Glu78Asp mutant 180 Figure C-4. The Lineweaver-Burk plots for the hydrolysis of aryl P-xylobiosides by the IAA-Glul72Cys mutant 181 Figure C-5. The Lineweaver-Burk plots for the hydrolysis of aryl P-xylobiosides by the Glul72Asp mutant 182 Figure C-6. The Lineweaver-Burk plots for the hydrolysis of aryl P-xylobiosides by the Glu 172Cys mutant 183 Figure C-7. The Lineweaver-Burk plots for the hydrolysis of aryl p-xylobiosides by the Glul72Gln mutant 184 Figure C-8. The Lineweaver-Burk plots for the hydrolysis of xylan by (a) native xylanase and (b) the IAA-Glul72Cys mutant 185 Figure C-9. The l /k 0 D S versus 1/[S] plot for the pre-steady state hydrolysis of 2,4-DNPG by the Glu358Cys mutant of Agrobacterium P-glucosidase 186 XII Figure D - l . ESMS analysis of Glu78Cys and native xylanase after treatment with 1200 fold molar excess I A A under nondenaturing conditions (27 hrs, 36 °C, pH 10, CAPSO) 187 Figure D-2. ESMS analysis of Glu78Cys and native xylanase after treatment with MMTS (100 fold molar excess, pH 8, 4 °C, 24 hrs) 188 Figure D-3. ESMS analysis of Glu78Cys and native xylanase after treatment with IAA (10 fold molar excess) under fully denaturing conditions (7.1 M urea, pH 7.5, 40 °C, 19 hrs) followed by refolding 189 Figure D-4. ESMS analysis of Glul72Cys and native xylanase after treatment with IAA (22 fold molar excess) under fully denaturing conditions (7.1 M urea, pH 7.5, 40 °C, 24 hrs) followed by refolding 190 Figure D-5. ESMS analysis of Glul72Cys, Glu78Cys, and native xylanase after treatment with X S S X (500 fold molar excess, pH 10, 40 °C, 24 hrs) 191 xiii LIST OF SCHEMES Scheme 1-1. Hydrolysis of glycosides by glycosidases 2 Scheme 1-2. The hydrolysis of a (3-glucoside catalyzed by retaining and inverting (3-glucosidases 3 Scheme 1-3. The proposed double displacement mechanism for a retaining glycosidase 5 Scheme 1-4. Bacterial cell wall hydrolysis catalyzed by lysozyme 7 Scheme 1-5. The proposed mechanism for the labelling of an active site carboxylic acid by carbodiimide 9 Scheme 1-6. The labelling of a glycosidase by an epoxy glycoside 10 Scheme 1-7. The reaction of conduritol C cis:epoxide with a P-galactosidase 12 Scheme 1-8. Identification of the active site nucleophile in Agrobacterium P-glucosidase using 2',4'-dinitrophenyl 2-deoxy-2-fluoro-P-D-glucopyranoside 13 Scheme 1-9. The reactivation of inactivated Agrobacterium P-glucosidase by P~D-glucosylbenzene 16 Scheme 1-10. A comparison of the transition state glucosyl oxocarbenium ion with its ground state glycoside 19 Scheme 1-11. The resonance structures of aldonolactones and aldonolactams 20 Scheme 1-12. The hydration of D-galacto-octenitol by coffee bean a-galactosidase and E. coli P-galactosidase 22 Scheme 1-13. Coffee bean a-galactosidase-catalyzed hydration of D-galactal 22 Scheme 1-14. The hydration of 2-acetamidoglucal by jack bean NAGase 23 Scheme 1-15. The substrate 2',4'-dinitrophenyl P-D-galactoside 26 Scheme 1-16. The proposed mechanisms for retaining and inverting glycosidases 27 Scheme 1-17. The replacement of a glutamic acid by a) aspartic acid, its shortened analogue and b) carboxymethylated cysteine, its lengthened analogue 29 Scheme 2-1. The structure of xylan, a P-l,4-linked xylose polymer 32 xiv Scheme 2-2. The side chains of glutamate and cysteine (shown in their expected ionization states atpH 6) 41 Scheme 2-3. The hydrolysis of 2',5'-dinitrophenyl (3-xylobioside by the Glu78Cys mutant. 42 Scheme 2-4. Carboxymethylation of Cys78 using iodoacetate (IAA) 44 Scheme 2-5. The modification of Cys78 by M M T S 49 Scheme 2-6. The proposed inactivation of IAA-Glu78Cys by 2F-DNPX 2 56 Scheme 2-7. The proposed reactivation of inactivated IAA-Glu78Cys by xylobiose 64 Scheme 3-1. The cleavage at methionine residues with CNBr under acidic conditions 86 Scheme 3-2. The proposed mechanism for xylanase-catalyzed hydrolysis of aryl f3-xylobiosides in the presence of sodium azide 98 Scheme 3-3. A comparison between Cys78, covalently modified with an oc-thioxylosyl moiety, and the normal covalent glycosyl-enzyme intermediate in native xylanase 110 Scheme 3-4. The proposed mechanism of inactivation for X S S X with the Glul72Cys mutant 113 Scheme 4-1. The conversion of cellobiose to glucose by p-glucosidase 118 Scheme 4-2. The side chains of glutamate and cysteine (shown in their expected ionization states at pH 7) 124 Scheme 4-3. The hydrolysis of 2',4'-dinitrophenyl P-D-glucopyranoside by the Glu358Cys mutant 125 Scheme 4-4. The proposed covalent glycosyl-enzyme intermediate formed upon treating Glu358Cys with 2,4-DNPG 130 Scheme 4-5. The Edman degradation procedure using the novel sequencing reagent 4-(3-pyridinylmethylaminocarboxypropyl) phenyl isothiocyanate (PITC 311).... 135 X V ABBREVIATIONS AND SYMBOLS 2,4-DNP 2,4-dinitrophenol 2.4- DNPG 2',4'-dinitrophenyl P-D-glucopyranoside 2.5- DNP 2,5-dinitrophenol 2,5-DNPX 2 2',5'-dinitrophenyl fj-xylobioside 2F-DNPG 2',4'-dinitrophenyl 2-deoxy-2-fluoro-(3-D-glucopyranoside 2F-DNPX 2 2',4'-dinitrophenyl 2-deoxy-2-fluoro-p-xylobioside 3,4-DNP 3,4-dinitrophenol 3,4-DNPX 2 3',4'-dinitrophenyl p-xylobioside Pig Br0nsted constant 8 chemical shift e extinction coefficient [0] mean residue ellipticity A Angstrom Ax absorbance at wavelength A, (where X is given in nm) AMPS O 3- [(1,1 -dimethy 1-2-hy droxyethyl)amino] -2-hydroxypropanesulfonic acid Asp78 aspartic acid at position 78 in the amino acid sequence B. circulans Bacillus circulans B. subtilis Bacillus subtilis B C X Bacillus circulans xylanase BIS-TRIS [/3w-(2-hydroxyethyl)-amino]?n,s'-(hydroxymethyl)methane B r A A bromoacetic acid BSA bovine serum albumin B S X Bacillus subtilis xylanase C A D collision assisted dissociation CAPSO 3-(cyclohexylamino)-2-hydroxy-l-propanesulfonic acid CD circular dichroism CNBr cyanogen bromide Cys78 cysteine at position 78 in the amino acid sequence D 2 0 deuterium oxide Da Dalton DTT 1,4-dithiothreitol E. coli Escherichia coli EC Enzyme Commission (classification number) of the International Union of Biochemistry EDTA ethylenediamine tetraacetic acid equiv equivalence ESMS electrospray mass spectrometry Glu78 glutamic acid at position 78 in the amino acid sequence Glu78Asp B. subtilis xylanase in which Glu78 has been replaced with an Asp Glu78Cys B. circulans xylanase in which Glu78 has been replaced with a Cys XVI Glul72Asp B. subtilis xylanase in which Glul72 has been replaced with an Asp Glul72Cys B. circulans xylanase in which Glu 172 has been replaced with a Cys Glul72Gln B. subtilis xylanase in which Glu 172 has been replaced with a Gin Glu358Cys Agrobacterium P^glucosidase in which Glu358 has been replaced with a Cys Glu358Gln Agrobacterium P-glucosidase in which Glu358 has been replaced with a Gin HEPES A/-(2-hydroxyethyl)piperazine-A/'-2-ethanesulfonic acid H E W L hen egg white lysozyme HPLC high performance liquid chromatography hr hour Hz Hertz IAA iodoacetate IAA-Glu78Cys Glu78Cys carboxymethylated at Cys78 IAA-Glu 172Cys Glul72Cys carboxymethylated at Cys 172 J coupling constant kDa kiloDalton MeOH methanol MES 2-(A/-morpholino)ethanesulfonic acid M M T S methyl methanethiolsulfonate MS " mass spectrometry MS/MS electrospray tandem mass spectrometry m/z mass/charge ratio NAGase p-A/-acetylhexosaminidase NaN 3 sodium azide N M R nuclear magnetic resonance ONP o-nitrophenol O N P X 2 o-nitrophenyl P-xylobioside . PhX 2 phenyl P-xylobioside PITC311 4-(3-pyridinylmethylaminocarboxypropyl) phenyl isothiocyanate PNP p-nitrophenol P N P X 2 p-nitrophenyl P-xylobioside ppm parts per million PTH phenylthiohydantoin Q l quadrupole one Rf retention factor SDS-PAGE sodium dodecyl sulfate-polyacrylamide gel electrophoresis UV/Vis ultraviolet/visible V reaction rate X S S X di-p,P'-D-xylopyranosyl disulfide Kinetic Constants k 2 glycosylation rate constant k 3 deglycosylation rate constant kcat catalytic rate constant (turnover number) kcat (app) apparent catalytic rate constant K d dissociation constant of an enzyme-substrate complex Ki dissociation constant for an enzyme-inhibitor complex ki rate constant for inactivation K m Michaelis constant of a substrate K m (app) apparent Michaelis constant of a substrate k-obs pseudo first-order rate constant kre,obs first-order rate constant for reactivation Abbreviations For Amino Acids Ala (A) alanine Arg (R) arginine Asn (N) asparagine Asp (D) aspartic acid Cys (C) cysteine Gin (Q) glutamine Glu (E) glutamic acid Gly (G) glycine His (H) histidine He (I) isoleucine Leu (L) leucine Lys (K) lysine Met (M) methionine Phe (F) phenylalanine Pro (P) proline Ser(S) serine Thr (T) threonine Trp (W) tryptophan Tyr(Y) tyrosine Val (V) valine xviii A C K N O W L E D G E M E N T S First, I would like to thank my supervisor, Dr. Steve Withers, for his guidance and support throughout the course of my doctoral studies. I also extend thanks to Dr. Shichang Miao, Dr. Dave Burgoyne, Dr. Curtis Braun, David Chow, and Shouming He for the ESMS work, Dr. Greg Connelly for the surface accessibility calculations, and Karen Rupitz for advice concerning enzyme kinetics and protein purification. Special thanks go to Dr. Warren Wakarchuk who kindly provided the xylanase mutants, Dr. Lothar Ziser who synthesized all the xylobioside substrates, and Manish Joshi who did all the molecular modelling and generated the active site figures of xylanase. I also wish to thank all the past and present members of the Withers group for the laughter, words of encouragement, insightful and stimulating conversations, and their friendship. To my friends and family, many thanks for the continual understanding, support, and love. Lastly, I thank my husband Rene who has been a source of encouragement, inspiration, and strength throughout the last four years. I could not have accomplished this without you Joe! xix *7o my frcw&tfo CHAPTER 1 GENERAL INTRODUCTION 1.1 GLYCOSIDASES AND THE REACTION THEY CATALYZE The degradation of polysaccharides, oligosaccharides, and other glycosides is accomplished through the concerted action of a large family of enzymes known as the glycosidases. These enzymes act by catalyzing the hydrolysis of the glycosidic linkage between two sugar residues, as shown in Scheme 1-1. The substrates of glycosidases comprise two parts, a glycone and an aglycone. The sugar glycone binds in the enzyme's catalytic site, while the aglycone binds in an adjacent subsite. The aglycone in most natural substrates is a sugar residue, but many glycosidases are capable of hydrolyzing substrates where the aglycone is an alkyl or aryl group. Glycosidases have considerable industrial and medical applications. For example, xylanases are already being commerically used in the pulp and paper industry (Paice et al., 1992), while amylases and 1,3-1,4-glucanases are utilized in the brewing industry (Olsen et al., 1991; Otey & Doane, 1984). Glycosidase inhibitors have potential therapeutic applications in the treatment of diabetes (Martin et al., 1991) and viral infections (Elbein et al., 1984) and in the study of lysosomal storage disorders (Atsumi et al., 1993). ROH Aglycone Scheme 1-1. Hydrolysis of glycosides by glycosidases. 1.2 CLASSIFICATION OF GLYCOSIDASES Glycosidases are divided into several subgroups on the basis of three criteria: glycone specificity, anomeric configuration of the substrate, and the stereochemical outcome of the reaction. 1. Glycone specificity. A glycosidase is classified according to the glycone, or sugar, against which it is most active. For example, a glucosidase is most active against glucosides (see Scheme 1-2), but may also be capable of hydrolyzing other sugars, such as mannosides or galactosides. 2. Anomeric configuration of the substrate. Glycosidases are further divided according to the anomeric configuration (a or (3) of the glycosidic bond to be cleaved. A (3-glucosidase, for example, catalyzes the hydrolysis of fj-glucosides and is generally inactive towards a-glucosides (refer to Scheme 1-2). HO HO OH O OH HO H O - ^ O HO P-Glucoside Retaining Inverting HO HO HO P-Glucose .OH HO| OH a-Glucose Scheme 1-2. The hydrolysis of a R-glucoside catalyzed by retaining and inverting (3-glucosidases (where R corresponds to a sugar residue, alkyl, or aryl group). 3. Stereochemical outcome of the reaction. A glycosidase is lastly classified as "retaining" or "inverting" on the basis of the relative anomeric configurations of the substrate and the initial glycone product. A glycosidase is termed "retaining" if the substrate and the initial product have the same anomeric configuration. In contrast, an "inverting" glycosidase releases a product with opposite anomeric configuration to the substrate. For a retaining fj-glucosidase, the initial product released is (3-glucose, while an inverting [3-glucosidase would produce a-glucose as illustrated in Scheme 1-2. 1.3 THE CATALYTIC MECHANISM OF RETAINING GLYCOSIDASES Retaining and inverting glycosidases employ two distinct mechanisms to achieve the hydrolysis • of glycosidic bonds. Retaining glycosidases are believed to act via a double displacement mechanism, illustrated in Scheme 1-3 for a p-l,4-xylanase. First proposed by Koshland (Koshland, 1953), a key feature of this mechanism is the covalent glycosyl-enzyme intermediate. The formation and hydrolysis of this covalent intermediate occurs via oxocarbenium ion-like transition states, with the assistance of two key active site carboxylates. One carboxylate acts as a nucleophile, attacking the anomeric carbon of the substrate to displace the aglycone (leaving group) and form the covalent intermediate (the glycosylation step). The other carboxylate plays the dual role of acid/base catalyst, in the first step (glycosylation) assisting in the formation of the intermediate by proton donation to the departing leaving group. In the subsequent hydrolysis (deglycosylation) step, this same residue functions as a base catalyst, deprotonating the attacking water. Noncovalent binding interactions between the enzyme and the substrate are believed to provide much of the enzymatic rate acceleration. In recent years an alternative ring-opening mechanism has been proposed to explain the catalytic action of retaining glycosidases (Franck, 1992; Post & Karplus, 1986). This 5 Scheme 1-3. The proposed double displacement mechanism for a retaining glycosidase (shown for fl-l,4-xylanases where R can be either one or more sugar residues or an alkyl or aryl group). 6 alternative mechanism, a distinguishing feature of which is the cleavage of the endocyclic C l - 0 bond, lacks serious experimental support (Sinnott, 1993). In contrast, the double displacement mechanism is supported by much experimental evidence (see the review by Sinnott, 1990). An overview of this evidence is presented in the following section. 1.4 EVIDENCE FOR A DOUBLE DISPLACEMENT MECHANISM IN RETAINING GLYCOSIDASES 1.4.1 PRESENCE OF A CARBOXYLATE NUCLEOPHILE AND ACID/BASE CATALYST The active site nucleophile and acid/base catalyst for many retaining glycosidases have been identified and invariably are either glutamic or aspartic acid residues. The identification of these catalytic carboxylates was accomplished using a combination of the following techniques: X-ray crystallographic studies, site-directed mutagenesis and kinetic analysis of the resulting mutant enzymes, labelling studies, and sequence alignment. 1.4.1.1 X- ray Crystallographic Studies Convincing evidence for the presence of a carboxylate nucleophile and acid/base catalyst has come from X-ray crystallographic studies done in conjunction with mutagenesis. Early evidence for the involvement of carboxylates in the reaction mechanism was derived from the X-ray crystal structure of hen egg white lysozyme, a retaining glycosidase which hydrolyzes the peptidoglycan component of bacterial cell walls (refer to Scheme 1-4). Examination of the tertiary structure revealed two active site carboxylates, Asp52 and Glu35, suitably positioned to function as the nucleophile (or charge stabilizer) and acid/base catalyst, 7 respectively (Phillips, 1967; for a listing of the amino acids and their corresponding three letter codes refer to Abbreviations and Symbols). Replacement of these residues with their corresponding amides (Asn52 and Gln35) resulted in essentially inactive mutant enzymes, consistent with these carboxylates being catalytically important (Malcolm et al., 1989). RO OH O AcNH HO OH O AcNH R O .OH AcNH HO Ac= — O II C — C H 3 R = OH O AcNH C H 3 O I II — C — C — N — peptide H H Scheme 1-4. Bacterial cell wall hydrolysis catalyzed by lysozyme (the glycosidic bond that is cleaved is indicated by the arrow). A number of crystal structures of glycosidases have since been determined (see Davies & Henrissat, 1995; McCarter & Withers, 1994) and these have supported the proposal of a carboxylate nucleophile and acid/base catalyst. Two examples are the X-ray crystal structures of the retaining glycosidases Streptomyces lividans xylanase A (Derewenda et al., 1994) and E. coli P-galactosidase (Jacobson et al., 1994). These glycosidases also possess active site carboxylates, in both these cases glutamic acids, appropriately located to act as the nucleophile and acid/base catalyst. In S. lividans xylanase A, mutating the active site glutamic acids, Glu 128 and Glu236, to glutamines completely abolished enzyme activity (Moreau et al., 1994). Site-directed mutagenesis done on the identified active site carboxylates, Glu537 and Glu461, of P-galactosidase also led to dramatic activity loss, thus 8 supporting the assignment of these residues as the nucleophile and acid/base catalyst, respectively (Yuan et al., 1994; Cupples et al., 1990). Direct evidence for the existence of catalytic carboxylates is obtained from the X-ray data of retaining glycosidases co-crystallized with either an inhibitor or substrate in the active site. The three-dimensional structure of a hybrid Bacillus l,3-l,4-(3-glucanase, with 3,4-epoxybutyl P-D-cellobioside (1.1) bound at the active site, showed a covalent bond between the side chain of Glu 105 and the inhibitor (refer to Scheme 1-6 for the proposed labelling mechanism). Site-directed mutagenesis of this conserved glutamic acid, in the homologous B. macerans l,3-l,4-(i-glucanase, resulted in an inactive enzyme confirming the catalytic importance of this carboxylate (Keitel et al., 1993). The B. macerans glucanase differs from the hybrid enzyme only in the amino acid residues at positions 1-16 near the amino terminus. 1.4.1.2 Labelling of the Catalytic Carboxylates using Carbodiimides In the absence of X-ray crystal data, the catalytic carboxylates of retaining glycosidases have been identified using different labelling reagents. One class of compounds used for this purpose is the carbodiimides. These carboxyl-specific reagents tend to be relatively nonspecific, labelling carboxyl groups both in the active site and on the enzyme's surface. Carbodiimides are proposed to react with active site carboxylic acids in the following manner (refer to Scheme 1-5). Following protonation by a suitably placed [1.1] carboxylic acid, the carbodiimide is attacked by a spatially close carboxylate anion to produce the intermediate 1.2 (Svensson et al., 1990). This intermediate can then undergo an intramolecular rearrangement, as shown in Scheme 1-5. Alternatively, an external amine may attack the carbonyl group of 1.2, thus converting the carboxylate group to an amide and releasing a urea derivative (Liu et al., 1991). " " " I " " "" I " " " I £o s o - o v o - o H ( - + R ' — N = C = N - R ^ R . _ N - C - N H - R ^ R . _ N - C _ N H _ R T I H + T C -O' o -cr II Rearrangement + R'-NH-C -* - R'-HN=C-NH-R I I [1.2] I I Scheme 1-5. The proposed mechanism for the labelling of an active site carboxylic acid by carbodiimide. The carbodiimide, l-(4-azonia-4,4-dimethyl-pentyl)-3-ethylcarbodiimide iodide (1.3), was used to label Glu87, an active site carboxylate residue in Schizophyllum commune xylanase A. A conserved residue in the family G glycosidases, Glu87 has been proposed to be the catalytic nucleophile (Bray & Clarke, 1994). This assignment is supported by site-10 directed mutagenesis work done on another family G xylanase from Bacillus pumilus. When Glu93, the homologous residue to Glu87 in the S. commune enzyme, was replaced with a serine or an aspartic acid, enzymatic activity was destroyed (Ko et al., 1992). This result is consistent with this conserved glutamic acid being catalytically important. C H 3 C H 3 - C H 2 - N = C = N - C H 2 - C H 2 - C H 2 - N - C H 3 l ~ C H 3 [1.3] 1.4.1.3 The Epoxy Glycosides and Conduritol Epoxides A second class of compounds commonly used to identify the catalytic carboxylates are the affinity labels epoxy glycosides and conduritol epoxides. Whereas the carbodiimides are relatively nonspecific, these reagents target the active site via their sugar moieties. The proposed labelling of a catalytic carboxylate by an epoxy glycoside is shown in Scheme 1-6. _ . . [ - . . _ _ _ L _ Scheme 1-6. The labelling of a glycosidase by an epoxy glycoside. Protonation of the epoxide by the acid catalyst, or some other proton donating residue, is followed by nucleophilic attack, resulting in the formation of a covalent adduct and enzyme inactivation. It should be noted that any suitably positioned nucleophile in the active site may make the attack, not just the catalytically essential carboxylates. 11 The catalytic carboxylates of several retaining glycosidases have been labelled using epoxy glycosides. Two such examples are Glu 105 and Asp52, the essential carboxylates of Bacillus amyloliquefaciens l,3-l,4-|3-glucanase and hen egg white lysozyme, respectively. Glul05 was labelled using 3,4-epoxybutyl P-D-cellobioside (1.1, H0j et al., 1992), whereas Asp52 was identified using an epoxyalkyl chitobioside derivative (Moult et al., 1973). Conduritol epoxides differ from epoxy glycosides in that the epoxide is endocyclic, incorporated within the sugar ring. These reagents are believed to label the catalytic carboxylates in much the same manner as the epoxy glycosides (refer to Scheme 1-7). In this case, a suitably positioned nucleophile attacks the epoxide at the center analogous to the substrate's anomeric center. For this reason, the conduritol epoxides were originally thought to be specific for the active site nucleophile. However, these reagents have occasionally labelled the acid/base catalyst as illustrated by the following example. Treatment of E. coli p-galactosidase with conduritol C ds-epoxide (1.4) resulted in the labelling of Glu461 and the incorrect assignment of this residue as the active site nucleophile (Herrchen & Legler, 1984). Subsequent labelling studies using a new class of inactivators, the 2-deoxy-2-fluoro glycosides, identified Glu537 as the actual nucleophile (Gebler et al., 1992b), this assignment being confirmed by mutagenesis work (Yuan et al., 1994). Glu461 is now proposed to be the acid/base catalyst, a role consistent with the mutagenesis results obtained by Cupples and colleagues (Cupples et al., 1990) and the recently solved X-ray crystal structure of [3-galactosidase (Jacobson et al., 1994). 12 0 ^ 0 0 ^ 0 I I Scheme 1-7. The reaction of conduritol C cis-epoxide (1.4) with a fi-galactosidase. The residues involved in the protonation and the nucleophilic attack may not necessarily be the catalytic carboxylates. 1.4.1.4 The 2-Deoxy-2-Fluoro Glycosides The active site nucleophiles in many retaining glycosidases have been successfully identified using 2-deoxy-2-fluoro glycosides (1.5, Scheme 1-8). These mechanism-based inactivators operate by trapping the covalent glycosyl-enzyme intermediate formed in the normal enzymic mechanism (Withers et al., 1988; refer to Scheme 1-3). The electronegative fluorine at the C-2 position of the glycone electronically destabilizes both oxocarbenium ion-like transition states. The replacement of the C-2 hydroxyl group with a fluorine also results in the loss of important transition state binding interactions, most likely hydrogen bonding (refer to section 1.4.5 for a more detailed description). These diminished binding interactions, when combined with the electronic destabilization effect, cause the rates of both the glycosylation and deglycosylation steps to be reduced. The presence of a good leaving group, 2,4-dinitrophenolate or fluoride, accelerates the glycosylation rate relative to the deglycosylation rate, resulting in the intermediate's accumulation and trapping. 13 _ _ _ [ _ _ _ Glu358 Scheme 1-8. Identification of the active site nucleophile in Agrobacterium P-glucosidase using 2',4'-dinitrophenyl 2-deoxy-2-fluoro-(l-D-glucopyranoside (1.5). These 2-deoxy-2-fluoro glycosides have not only been used to identify the active site nucleophile, Glu537, of E. coli P-galactosidase (Gebler et al., 1992b), but also the catalytic nucleophiles of several other retaining P-glycosidases. For example, treatment of Agrobacterium P-glucosidase with 2',4'-dinitrophenyl 2-deoxy-2-fluoro-P-D-glucopyranoside (1.5) led to the identification of Glu358 as the catalytic nucleophile (Withers et al., 1990; Scheme 1-8). Replacement of Glu358 with either an asparagine or glutamine totally abolished enzymatic activity, consistent with this residue being the active site nucleophile (Withers et al., 1992). Similarly, Glu233 and Glu78 have been identified as the active site nucleophiles of Cellulomonas fimi exoglycanase (Tull et al., 1991) and Bacillus subtilis xylanase (Miao et al., 1994), respectively, through labelling studies employing these 2-deoxy-2-fluoro glycosides. 1.4.1.5 A Strategy for Identifying the Acid/Base Catalyst No labelling reagent exists that specifically targets the acid/base catalyst of retaining glycosidases, although carbodiimides, epoxy glycosides, and conduritol epoxides have frequently labelled this catalytic residue. In the absence of a reliable chemical tagging method, MacLeod and coworkers (MacLeod et al., 1994) have developed a strategy for 14 identifying the acid/base catalyst which combines the techniques of sequence alignment, site-directed mutagenesis, and kinetic analysis. Highly conserved carboxylic acid residues, identified through the sequence alignments of related enzymes, are mutated to either an alanine or a glycine. A kinetic analysis of these mutants is then performed using substrates with different requirements for acid catalysis, in the absence or presence of exogenous nucleophiles such as sodium azide. This method has been used to identify Glu 127 and Glul70 as the acid/base catalysts of Cellulomonas fimi exoglycanase (MacLeod et al., 1994) and Agrobacterium p-glucosidase (Wang et al., 1995), respectively. The recently solved X -ray crystal structure of C. fimi exoglycanase shows Glu 127 suitably located in the active site to function as the acid/base catalyst, thus confirming this assignment (White et al., 1994). While the tertiary structure of Agrobacterium P-glucosidase is still undetermined, the X-ray structures of two other family 1 glycosidases, white clover P-glucosidase (Barrett et al., 1995) and Lactococcus lactis 6-phospho-p-galactosidase (Wiesmann et al., 1995), have been published. Inspection of these two structures shows the homologous residues to Glu 170 of Agrobacterium p-glucosidase appropriately positioned within the enzymes' active sites to fulfill the role of acid/base catalyst. 1.4.2 THE NATURE OF THE GLYCOSYL-ENZYME INTERMEDIATE A key element of Koshland's proposed double displacement mechanism is the covalent glycosyl-enzyme intermediate. Evidence for the covalent nature of this intermediate has been obtained from the cc-secondary deuterium kinetic isotope effects measured for several retaining glycosidases. With substrates (deuterated at the anomeric carbon) for which deglycosylation is rate determining, values of k H /k D = 1.21-1.25 and 1.10-1.12 have been 15 determined for E. coli f3-galactosidase (Sinnott & Souchard, 1973; Sinnott et al., 1978) and Agrobacterium (3-glucosidase (Kempton & Withers, 1992), respectively. These rate decreases, resulting from the substitution of a hydrogen with a deuterium at the anomeric center, indicate that significant sp3 to sp2 rehybridization occurs at this reaction center upon going to the transition state. The glycosyl-enzyme intermediate must, therefore, have considerably more sp3 character than the subsequent deglycosylation transition state. This result is consistent with the proposed mechanism in which a covalent glycosyl-enzyme intermediate is hydrolyzed via a transition state with significant oxocarbenium ion character. The strongest validation for the proposed covalent glycosyl-enzyme intermediate is the actual trapping of this species. This has been accomplished for several retaining glycosidases, one such enzyme being porcine pancreatic a-amylase. A covalent glycosyl-enzyme intermediate, formed upon treatment of this a-amylase with [1- 1 3C] maltotetraose (1.6) at -20 °C, was detected using 1 3 C N M R (Tao et al., 1989). The use of the subzero temperature and the cryosolvent DMSO sufficiently reduced the hydrolysis rate to "trap" this covalent intermediate and permit its detection using N M R techniques. 16 Trapping of the covalent glycosyl-enzyme intermediate has also been achieved through the use of the mechanism-based inactivators, the 2-deoxy-2-fluoro glycosides (refer to section 1.4.1.4 for a description of how these inactivators work). The glycosyl-enzyme intermediate formed by the inactivation of Agrobacterium p-glucosidase with 2-deoxy-2-fluoro-P-D-glycosides was extremely stable, having a half-life of greater than 500 hours at 30 °C (Withers et al., 1990; refer to Scheme 1-8). The longevity of this intermediate not only allowed the covalent nature and a-linked stereochemistry of the intermediate to be determined by 1 9 F N M R (Withers & Street, 1988), it also resulted in the identification of the active site nucleophile, Glu358 (Withers et al., 1990). This "trapped" glycosyl-enzyme intermediate was catalytically competent and could be turned over, in the presence of a suitable acceptor, to give fully active enzyme. Thus, treatment of inactivated P-glucosidase with the glycosyl acceptor, p-D-glucosylbenzene (1.7), resulted in enzyme reactivation and the release of the disaccharide, 2'-deoxy-2'-fluoro-P-D-cellobiosylbenzene (1.8, Scheme 1-9). The formation of this disaccharide is believed to occur via a transglycosylation reaction in which the C-4 hydroxyl of glucosylbenzene attacks the anomeric carbon of the glycosyl-enzyme intermediate (Withers et al., 1990). Glu358 Glu358 Scheme 1-9. The reactivation of inactivated Agrobacterium P-glucosidase by P-D-glucosylbenzene (1.7). 17 While the "trapped" glycosyl-enzyme intermediate of Agrobacterium P-glucosidase was detected using 1 9 F NMR, the "trapped" intermediate of another retaining P-glycosidase has been observed via X-ray crystallography. The X-ray crystal structure of Cellulomonas fimi exoglycanase, cocrystallized with 2',4'-dinitrophenyl 2-deoxy-2-fluoro-p-D-cellobioside, clearly showed a covalent bond between the C - l of the fluorocellobioside moiety and the side chain of Glu233, the active site nucleophile (White et al., 1996). As expected for a retaining p-glycosidase, the stereochemistry of this covalent linkage was a, as determined by 1 9 F N M R studies in addition to the structural data. Although there is strong evidence for the existence of a covalent glycosyl-enzyme intermediate for many retaining glycosidases, there is still debate over the nature of this intermediate in lysozyme. On the basis of the X-ray crystal structure of hen egg white lysozyme, it has been proposed that the negatively charged Asp52 acts as a charge stabilizer rather than a nucleophile, forming a long lived ion pair with the positively charged glycosyl oxocarbenium ion intermediate (Phillips, 1967). This ion pair would require a sufficiently long lifetime to permit the leaving group to diffuse out of the active site and the glycosyl acceptor to diffuse in and react. The lifetimes of glycosyl oxocarbenium ions are estimated to be 10"10 to 10"12 s in aqueous solution (Amyes & Jencks, 1989; Bennet & Sinnott, 1986), whereas covalent glycosyl-enzymes are more stable having estimated lifetimes of 1 to 100 ms at ambient temperatures (Weber & Fink, 1980). Given these estimated lifetimes, it seems improbable that the inherently reactive glycosyl oxocarbenium ion intermediate could exist in an ion pair, even within the stabilizing active site environment of an enzyme, without this ion pair collapsing to form the more stable covalent intermediate (Sinnott & Souchard, 1973). 18 1.4.3 OXOCARBENIUM ION-LIKE TRANSITION STATES In the proposed mechanism for retaining glycosidases, the formation and hydrolysis of the covalent glycosyl-enzyme intermediate occur via oxocarbenium ion-like transition states (refer to section 1.3, Scheme 1-3). Determination of the a-secondary deuterium kinetic isotope effects for several retaining glycosidases has provided evidence for the oxocarbenium ion-like character of both the glycosylation and deglycosylation transition states. With substrates (deuterated at the anomeric carbon) for which glycosylation is rate determining, values of k H / k D = 1.15-1.20, 1.06-1.12, and 1.06 have been obtained for E. coli (3-galactosidase (Sinnott & Withers, 1974; Sinnott et al., 1978), C.'fimi exoglycanase (Tull & Withers, 1994), and Agrobacterium P-glucosidase (Kempton & Withers, 1992), respectively. Similar k H / k D values were determined for these retaining glycosidases using deuterated substrates for which deglycosylation is rate limiting. For E. coli p-galactosidase (Sinnott & Souchard, 1973; Sinnott et al., 1978), C. fimi exoglycanase (Tull & Withers, 1994), and Agrobacterium P-glucosidase (Kempton & Withers, 1992), the corresponding k H / k D values of 1.21-1.25, 1.10-1.12, and 1.10-1.12 were obtained for deglycosylation. These large positive kinetic isotope effects, determined for both the glycosylation and deglycosylation transition states, indicate that significant sp3 to sp2 rehybridization occurs at the reaction center, the anomeric carbon, upon going to the corresponding transition state. Thus both transition states have substantial sp character, consistent with their proposed oxocarbenium ion-like character. Transition state analogue studies provide additional insight into the structure of the glycosylation and deglycosylation transition states. These transition states, in the proposed 19 double displacement mechanism, involve a glycosyl oxocarbenium ion, a species which possesses several unique features relative to its ground state glycoside. As illustrated in Scheme 1-10, the glycosyl oxocarbenium ion adopts a half-chair (or a 2 ' 5 B or B 2 , 5 boat) conformation, with the C - l , C-2, C-5, and 0-5 atoms being coplanar, and also contains a full positive charge which is shared between the C - l and 0-5 atoms (Sinnott, 1987). Since a transition state analogue mimics the structure of its target transition state, such a compound for a retaining glycosidase should incorporate several of the following features: a positive charge, a trigonal anomeric center, a half-chair (or boat) conformation, and the same configuration of hydroxyl groups as the ground state glycoside. 4 .OH .OH V ^ O R Glucosyl oxocarbenium ion Glucoside Scheme 1-10. A comparison of the transition state glucosyl oxocarbenium ion with its ground state glycoside. The aldonolactones (1.9) and aldonolactams (1.10), shown in Scheme 1-11, are putative transition state analogues of retaining glycosidases. These compounds resemble the glycosyl oxocarbenium ion in geometry and to some degree charge, since resonance places some positive charge on the ring oxygen and nitrogen atoms. Both compounds bind to retaining glycosidases 102 to 104 fold tighter than the corresponding aldohexoses (Legler, 1990). This enhanced binding of compounds 1.9 and 1.10, relative to the corresponding glycosides, is typical of transition state analogues since enzymes have evolved to selectively stabilize the transition state of a reaction relative to its ground state. 20 OH [1.9] OH [1.10] Scheme 1-11. The resonance structures of aldonolacton.es (1.9) and aldonolactams (1.10). Nojiritetrazole (1.11) and mannonojiritetrazole (1.12) are another class of transition state analogues of retaining glycosidases. Both compounds possess a half-chair conformation and a trigonal anomeric center, thus resembling the glycosyl oxocarbenium ion in geometry. As expected for transition state analogues, these tetrazole derivatives exhibited tighter binding (5 to 50 fold) toward glycosidases than their corresponding ground state glycosides. The binding efficiencies of compounds 1.11 and 1.12 are correlated with the glycone specificities of the glycosidases, with the "mannose" and "glucose" tetrazole derivatives binding tightest to mannosidases and glucosidases, respectively (Ermert et al., 1993). OH N=N OH N=N OH [1.11] [1.12] 1.4.4 GENERAL ACID CATALYSIS In the double displacement mechanism proposed for retaining glycosidases, the departure of the aglycone, or leaving group, may proceed with acid assistance. This general 21 acid catalysis is provided by an active site carboxylic acid located on the face of the sugar opposite to the nucleophile (refer to Scheme 1-3, section 1.3). In retaining p-glycosidases the acid catalyst is positioned above the sugar, while this catalytic residue is located below the sugar in retaining a-glycosidases. The acid catalysts of several retaining glycosidases have been identified using both X-ray crystallographic data and site-directed mutagenesis studies. As discussed in section 1.4.1, the carboxylic acids Glu35, Glu461, and Glul27 have been identified as the acid catalysts of hen egg white lysozyme (Phillips, 1967; Malcolm et al., 1989), E. coli p-galactosidase (Jacobson et al., 1994; Cupples et al., 1990), and C. fimi exoglycanase (MacLeod et al., 1994; White et al., 1994), respectively. Evidence supporting acid catalysis in glycoside hydrolysis is provided not only by the identification of the acid catalyst, but also from the glycosidase-catalyzed hydration of glycals and octenitol derivatives. Coffee bean a-galactosidase and E. coli P-galactosidase, both retaining glycosidases, catalyze the hydration of D-galacto-octenitol (1.13) to the galactooctulose derivatives illustrated in Scheme 1-12. The stereochemistry of the hydrated products was determined and revealed that protonation occurred from below the galactose ring for the a-galactosidase (Weiser et al., 1992), while it occurred from above the sugar for the p-galactosidase (Lehmann & Schlesselmann, 1983). The enzyme-catalyzed hydration of D-galactal (1.14) by coffee bean a-galactosidase was also examined (refer to Scheme 1-13). The stereochemistry of the hydration product, determined by  l H N M R spectroscopy, indicated that deuteration (protonation) occurred from below the sugar (Weiser et al., 1992). These results reveal the presence of a proton-donating residue, in both a retaining a- and P-galactosidase, positioned on the correct face of the sugar ring to function as the acid catalyst 22 in the normal enzymatic mechanism. OH OH HO O-OH [1.13] i X\CH 3 cx-Galactosidase H 20 H 20 p-Galactosidase OH OH HO-D GH 3 HO H OH D Scheme 1-12. The hydration of D-galacto-octenitol (1.13) by coffee bean a-galactosidase and E. coli fi-galactosidase. OD OD Scheme 1-13. Coffee bean cc-galactosidase-catalyzed hydration of D-galactal (1.14). The hydration of 2-acetamidoglucal (1.15, Scheme 1-14) by the retaining $-N-acetylhexosaminidase (denoted NAGase) from jack bean also supports the proposed acid catalysis. The compound A/-acetylglucosamine (1.16) was identified as the sole hydration product by HPLC analysis using authentic standards, indicating that proton donation came from the top face of the sugar (Lai & Withers, 1994). The location of this proton-donating residue makes it a likely candidate for the acid catalyst. 23 NAGase H 20 AcNH OH AcNH [1.15] [1.16] Scheme 1-14. The hydration of 2-acetamidoglucal (1.15) by jack bean NAGase. Although both jack bean NAGase and coffee bean a-galactosidase protonate glycals from the same face of the sugar as the proposed acid catalyst, this is not the norm. In general, both retaining (3- and a-glycosidases protonate glycals from the face of the sugar that is opposite to the acid catalyst (Legler, 1990). An example of this is the hydration of cellobial (1.17) by the exo- and endo-f3-l,4-cellulases from Irpex lacteus and Aspergillus niger, respectively. ' H N M R analysis of the hydration product revealed that both cellulases protonate (deuterate) the double bond of cellobial from below the face of the sugar ring, the face opposite the acid catalyst (Kanda et al., 1986). In these cases, the active site nucleophile is proposed to donate the proton rather than the acid catalyst (Kanda et al., 1986). While the above evidence clearly implicates the involvement of an acid catalyst in the mechanism of many retaining glycosidases, acid catalysis is not essential for the hydrolysis of all substrates. The (3-galactosidase from E. coli efficiently hydrolyzes (3-D-galactopyranosyl pyridinium salts (1.18), with rate enhancements of 108 to 10 1 3 fold over spontaneous hydrolysis (Jones et al., 1977; Sinnott & Withers, 1974). Since the pyridine leaving group OH [1.17] [1.18] 24 does not possess a lone pair of electrons, proton donation to this group from the acid catalyst is structurally impossible. Therefore, none of the observed rate acceleration can be attributed to acid catalysis. 1.4.5 NONCOVALENT ENZYME-SUBSTRATE INTERACTIONS While acid catalysis can contribute to the rate accelerations achieved by retaining glycosidases, the bulk of the rate enhancement is proposed to result from noncovalent binding interactions between the enzyme and the substrate. An important binding interaction is the hydrogen bonding between the hydroxyl groups on the sugar substrate and the enzyme. In addition to hydrogen bonding, there are also hydrophobic binding interactions between the sugar ring of the substrate and the side chains of aromatic amino acids. Upon going from the ground state to the transition state, these noncovalent binding interactions between the enzyme and the substrate are optimized, causing the enzyme to bind the transition state tighter than the ground state. This transition state stabilization translates into a decreased activation energy and subsequent rate acceleration. Evidence for the hydrogen bonding and hydrophobic binding interactions that occur between retaining glycosidases and oligosaccharides is obtained from X-ray crystallographic studies. Examination of the tertiary structure of Trichoderma reesei cellobiohydrolase I complexed with o-iodobenzyl-l-thio-(3-cellobioside (1.19) revealed hydrogen bonds between 0-3 and 0-4 of the first glucose unit and two active site residues, Glu217 and Asp214. In addition, a hydrophobic stacking interaction was observed between the (3-face of this glucose unit and the side chain of Trp376 (Divne et al., 1994). Similar noncovalent binding interactions were also seen in the X-ray crystal structures of hen egg white lysozyme 25 (Strynadka & James, 1991) and a hybrid Bacillus 1,3-1,4-p-glucanase (Keitel et al., 1993) co-crystallized with oligosaccharides. In the 1,3-1,4-glucanase crystal structure, hydrogen bonds were observed between the hydroxyl groups of the cellobioside unit (see structure 1.1) and the active site amino acids Glu63, His99, Tyr24, and Asn26 (Keitel et al., 1993). 2 1 OH [1.19] Kinetic analysis of retaining glycosidases, using deoxyglycosides as substrates, emphasizes the important contributions that these hydrogen bonding interactions make to the observed rate acceleration. In the deoxyglycosides, a hydroxyl group at a particular position on the normal substrate is replaced with a hydrogen. This substitution eliminates all significant hydrogen bonding interactions with the enzyme at this position, thereby resulting in reduced transition state stabilization and a rate decrease relative to the parent glycoside. For E. coli P-galactosidase, the k c a t / K m values for the 4-deoxy and 6-deoxy analogues of 2',4'-dinitrophenyl p-D-galactoside (see Scheme 1-15) were 500 to 1500 fold lower than that of the parent galactoside (McCarter et al., 1992). Less dramatic reductions in k c a t / K m values were observed for the deoxyglycosides with Agrobacterium P-glucosidase. A 3 to 17 fold decrease in k c a t / K m values were observed for the 3, 4, and 6-deoxyglucoside derivatives relative to the parent glucoside (Namchuk & Withers, 1995). From these results, it was determined that interactions with the C-4 and C-6 sugar hydroxyls contribute approximately 26 18 kJ/mol binding energy to transition state stabilization in E. coli p-galactosidase (McCarter et al., 1992), whereas the C-3, C-4, and C-6 sugar hydroxyls contribute 3 to 7 kJ/mol in Agrobacterium p-glucosidase (Namchuk & Withers, 1995). Interactions with the C-2 sugar hydroxyl play an even bigger role in transition state stabilization in retaining glycosidases, with binding energies of 34 kJ/mol and 18 kJ/mol being estimated for E. coli P-galactosidase (McCarter et al., 1992) and Agrobacterium P-glucosidase (Namchuk & Withers, 1995), respectively. The increased importance of the C-2 sugar hydroxyl, relative to the substrate's other hydroxyl groups, has been speculated to arise from a change in this hydroxyl's orientation which likely accompanies the substrate's conformational change upon going to the transition state. The repositioning of the C-2 hydroxyl group in the transition state could result in either new hydrogen bonding interactions or the strengthening of interactions previously present in the ground state (Namchuk & Withers, 1995). R 0 2 N R 1 Scheme 1-15. The substrate 2',4'-dinitrophenyl fi-D-galactoside (R is an OH group in the parent glycoside, whereas R is an H atom in the deoxy analogue). 1.5 THE PROPOSED MECHANISMS FOR RETAINING AND INVERTING GLYCOSIDASES Retaining and inverting glycosidases catalyze the hydrolysis of glycosidic bonds by two distinct mechanisms. Retaining glycosidases, which catalyze hydrolysis with net retention of anomeric configuration, employ a double displacement mechanism involving a 27 covalent glycosyl-enzyme intermediate as illustrated in Scheme 1-16 (refer to Scheme 1-3, section 1.3 for a more detailed description). By contrast, inverting glycosidases use a direct displacement reaction to achieve overall inversion of anomeric configuration (Koshland, 1953; Sinnott, 1990). While the two mechanisms are distinct, they possess several important Retaining Mechanism H O ^ O e d ^ O H O ^ O i H O ^ O , R _ ^ 0 ^ o 7 _ H O ^ ^ O H 4 , . , 5 A H O T H O K . H HO I 00^,0 0 ^ 0 Q o^,o • Inverting Mechanism Scheme 1-16. The proposed mechanisms for retaining and inverting glycosidases (shown for /5-1,4-xylanases where R can be either one or more xylose residues or an alkyl or aryl group. R' corresponds to one or more xylose residues). similarities since both proceed via oxocarbenium ion-like transition states and, in both cases, a pair of carboxylates function as the catalytic residues. In retaining glycosidases, one carboxylate functions as an acid/base catalyst and the other as a nucleophile, the average 28 o distance between the two essential carboxyl residues being 4.5-5.5 A. In inverting enzymes, one carboxylate functions as a general acid and the other as a general base and the distance between them is greater (9-9.5 A), accommodating a water molecule (Davies & Henrissat, 1995; McCarter & Withers, 1994; Wang et al., 1994; White et al., 1994). 1.6 AIMS OF THIS THESIS 1. What are the mechanistic consequences of modifying the distance between the two catalytic carboxylates? In glycosidases, the distance between the two essential carboxylates is critical in determining the catalytic mechanism followed. The main focus of this thesis is to explore the mechanistic consequences of altering this critical distance. The model enzyme system chosen for examining this question is the endo-(3-l,4-xylanase from Bacillus circulans/subtilis. For this retaining glycosidase, the catalytic carboxylates had previously been identified as two glutamic acids, Glu78 and Glu 172. The strategy that will be used to alter the distance between the two catalytic carboxylates consists of selectively replacing each glutamic acid with its shortened or o lengthened analogue. To increase the distance (>5.5 A), one of the glutamic acids will be replaced with its shortened analogue, aspartic acid (refer to Scheme 1-17). Substituting one of the glutamic acids with a lengthened analogue should have the opposite effect, a decrease in the distance between the two essential carboxylates (<5.5 A). This lengthened glutamic acid analogue will be produced using the combined techniques of site-directed mutagenesis and chemical modification. A cysteine, introduced at the position of interest, will be selectively carboxymethylated with iodoacetic acid, thus yielding a carboxylic side chain approximately 1.6 A longer than glutamic acid. 29 a) C H 2 ^ C H 2 CH? i i Glutamic acid Aspartic acid ° v O - || C H 2 C H 2 ^ S- l - C H 2 - C - Q - £ C H 2 9 H 2 C H 2 Glutamic acid Cysteine Carboxymethylated Cysteine Scheme 1-17. The replacement of a glutamic acid by a) aspartic acid, its shortened analogue and b) carboxymethylated cysteine, its lengthened analogue. 2. Can a cysteine function as the active site nucleophile in retaining glycosidases? The introduction of a cysteine at the active site nucleophile position fulfills two purposes. First, it provides the means by which to create a lengthened glutamic acid analogue, thereby allowing the effects of decreasing the critical distance between the two catalytic carboxylates to be examined. Second, the feasibility of a cysteine acting as the active site nucleophile in retaining glycosidases can be tested. This question, the second aim of this thesis, will be addressed by performing a kinetic analysis on the active site nucleophile cysteine mutants of two retaining P-glycosidases, B. circulans xylanase and Agrobacterium P-glucosidase. There are several enzymes in which the catalytic nucleophile has been identified as an active site cysteine residue. The cysteine proteases (Walsh, 1979), tyrosine phosphatases (Cho et al., 1992), thiolase (Thompson et al., 1989), and glyceraldehyde 3-phosphate dehydrogenase (Stryer, 1988; Walsh, 1979) are such examples. The possibility that cysteine could function as the catalytic nucleophile in retaining glycosidases was suggested by Hardy 30 and Poteete (1991). They found that replacing Asp20, the putative catalytic nucleophile of T4 lysozyme, with a cysteine resulted in nearly wild-type activity. At the time of this work, T4 lysozyme was believed to be a retaining glycosidase, but recently it has been discovered that this enzyme actually functions via an inverting mechanism. Consequently, Asp20 functions as a general base catalyst rather than as the catalytic nucleophile (Kuroki et al., 1995). 31 CHAPTER 2 MODIFYING THE ACTIVE SITE NUCLEOPHILE OF BACILLUS CIRCULANS/SUBTILIS XYLANASE 32 2.1 INTRODUCTION 2.1.1 XYLAN AND XYLAN-DEGRADING ENZYMES One of the major components of plant cell walls is xylan, a fj-l,4-linked polymer of D-xylose residues (refer to Scheme 2-1). Depending on the source, the xylose backbone may possess side chains of D-glucuronic acid, 4-O-methyl-D-glucuronic acid, L-arabinose, or acetyl groups (Coughlan & Hazlewood, 1993). Complete enzymatic degradation of this complex polysaccharide to its constituent monosaccharides is accomplished through the concerted action of endo-f3-l,4-xylanases, (3-1,4-xylosidases, a-L-arabinofuranosidases, a-D-glucuronidases, and various esterases (Coughlan & Hazlewood, 1993). There is considerable commercial interest in these xylan-degrading enzymes. They could potentially be used in the production of valuable fuel and chemical products from agricultural waste (Dekker, 1982). In addition, some poultry and pig feed is currently being supplemented with xylanases to improve the feed's digestibility (McCleary, 1992). Endoxylanases are also being utilized in the biobleaching of pulp for paper manufacture (Paice et al., 1992). Scheme 2-1. The structure of xylan, a fi-l,4-linked xylose polymer (where R can be a hydrogen, glucuronic acid, 4-O-methylglucuronic acid, L-arabinose, or an acetyl group). RO OR 33 2.1.2 ENDO-p-l,4-XYLANASES Endo-P-l,4-xylanases (P-l,4-D-xylan xylanohydrolase; EC 3.2.1.8) convert xylan to xylobiose and longer chain xylo-oligosaccharides by cleaving the p-l,4-glycosidic bonds between internal xylose residues. These enzymes are widely distributed in nature and have been found in plants, bacteria, fungi, and a variety of invertebrate animals such as snails and slugs (Dekker & Richards, 1976). Many bacterial and fungal xylanases have been isolated, cloned, and sequenced. As a consequence of this work, all sequenced xylanases to date have been assigned to two distinct groups, families F and G (Gilkes et al., 1991; Oku et al., 1993), also known as families 10 and 11 (Henrissat & Barrioch, 1993). These assignments were made by means of sequence alignment and comparison of the clustering of hydrophobic residues that are key in determining a protein's secondary structure. The family G (or family 11) xylanases, characterized by having low molecular weights and no cellulolytic activity, currently contain 26 members (Bairoch & Apweiler, 1996). For several family members, the catalytic residues have been identified using techniques such as X-ray crystallographic studies, site-directed mutagenesis, chemical modification studies, and sequence alignment. As illustrated in Table 2-1, the proposed catalytic residues are glutamic acids. Interestingly, these glutamic acids are conserved in all family G xylanases. 34 Table 2-1. The Family G Xylanases for Which the Catalytic Residues Have Been Identified Enzyme Proposed Catalytic Residues Identification Technique Reference Schizophyllum commune xylanase A Glu87/Glul84 chemical modification sequence alignment Bray & Clarke, 1990 Bray & Clarke, 1994 Bacillus circulans/subtilis xylanase Glu78/Glul72 X-ray crystallography mutagenesis chemical modification sequence alignment Campbell et al., 1993 Wakarchuk et al., 1994 Miao et al., 1994 Bacillus pumilus xylanase Glu93/Glul82 X-ray crystallography mutagenesis sequence alignment Katsube et al., 1990 Koet al., 1992 Trichoderma reesei X Y N I Glu75/Glul64 X-ray crystallography sequence alignment Torronen & Rouvinen, 1995 Trichoderma reesei XYNII Glu86/Glul77 X-ray crystallography sequence alignment Torronen et al., 1994 2.1.3 THE ENDO-f3-l,4-XYLANASES FROM BACILLUS CIRCULANS AND BACILLUS SUBTILIS The family G endo-(3-l,4-xylanases from Bacillus circulans and Bacillus subtilis (denoted B C X and B S X , respectively) are small (-20 kDa) glycosidases which share a high degree of homology. The xylanases, containing 185 amino acids, differ only in the residue at position 147. In B C X residue 147 is a threonine, while in B S X it is a serine. Despite this single amino acid substitution, the two xylanases are considered to be equivalent, having identical activities and stabilities (Wakarchuk et al., 1992). The genes encoding these xylanases have been cloned and expressed in E. coli (Paice et al., 1986; Yang et al., 1988; Sung et al., 1993). The purified recombinant xylanases are identical to the natural enzymes in their amino acid sequences, molecular weights, and kinetic behavior (Sung et al., 1993). 35 The xylanase from B. subtilis catalyzes the hydrolysis of both synthetic and natural substrates with net retention of anomeric configuration (Gebler et al., 1992a). As with other retaining [3-glycosidases, both BSX and BCX are proposed to employ the double displacement mechanism described in Chapter 1, section 1.3 (also refer to Scheme 1-3). This mechanism involves a covalent glycosyl-enzyme intermediate which is formed and hydrolyzed with the assistance of two active site carboxylates. One carboxylate functions as a nucleophile/leaving group, while the other carboxylate plays the dual role of acid/base catalyst. X-ray crystallographic data and mutagenesis results have identified Glu78 and Glu 172 as the catalytic residues in the xylanase from B. circulans/subtilis. X-ray crystal structures of both native BCX and a mutant enzyme, Glul72Cys, complexed with xylotetraose have been published and are shown in Figure 2-1 and 2-2 (Campbell et al., 1993; Wakarchuk et al., 1994). The three-dimensional structure of B. circulans xylanase consists of a single domain composed of three P-sheets (I, U, and IE) and one a-helix (see Figure 2-1). The active site cleft, located between sheets U and JJJ, spans the entire width of the protein. Within this cleft are located Glu78 and Glu 172, two glutamic acids that are conserved in all family G xylanases (refer to Figure 2-1 and 2-2). Separated by 5.5 A, these conserved glutamic acids are suitably positioned to function as the catalytic residues (Campbell et al., 1993). Site-directed mutagenesis results confirmed the catalytic importance of Glu78 and Glu 172. Mutation of the glutamic acids to either a glutamine or aspartic acid resulted in mutant enzymes that were essentially inactive against the natural substrate xylan (Wakarchuk et al., 1994). Glu78 and Glul72 have been assigned the specific roles of active site nucleophile and 36 acid/base catalyst, respectively. Glu78 was identified as the catalytic nucleophile by trapping the covalent glycosyl-enzyme intermediate using the mechanism-based inactivator, 2',4'-dinitrophenyl 2-deoxy-2-fluoro-p>-xylobioside (Miao et al., 1994; for a full explanation of how this 2-deoxy-2-fluoro glycoside works see section 1.4.1.4, Chapter 1). By default, Glu 172 must function as the acid/base catalyst. Three aspartic acids, Aspl 1, 83 and 106, are also observed in the active site (refer to Figure 2-2). Replacement of Asp83 and Asp 106 with asparagine had no effect on enzymatic activity, indicating that these carboxylates are not catalytically essential (Campbell et al., 1993). The corresponding mutation at Asp 11 resulted in a modest (75 fold) decrease in activity. While Asp 11 is not intimately involved in catalysis, it has been suggested that this carboxylate helps to maintain the structure of the active site (Campbell et al., 1993; Wakarchuk et al., 1994). Figure 2-1. The three-dimensional structure of B. circulans xylanase, showing the (3-sheets (in blue) and a-helix (in purple). The catalytic residues Glu78 and Glul72 are shown in red. The active site cleft is viewed from the side. 37 a) b) v i A l / i \ An , i cn m Figure 2-2. The active site of B. circulans xylanase shown (a) in the absence of substrate and (b) for the Glul72Cys mutant complexed with xylotetraose. Only key residues are displayed and are colored according to residue type, with Glu and Asp in red, Tyr in green, Trp in magenta, Arg in blue, Gin and Asn in cyan, and Cys in yellow. Xylotetraose is colored white and atoms CI and 01 of the reducing end of the oligosaccharide are indicated. Hydrogen bonds are shown as yellow dashed lines. 38 2.2 SPECIFIC AIMS OF THIS STUDY 1. Can a cysteine function as the catalytic nucleophile in B. circulans/subtilis xylanase? It has been suggested that a cysteine can fulfill the role of the catalytic nucleophile in retaining glycosidases (Hardy & Poteete, 1991; Coughlan, 1992). This proposal will be tested by conducting a kinetic analysis on the active site nucleophile cysteine mutant, denoted Glu78Cys, of B. circulans xylanase. 2. What are the mechanistic consequences of altering the distance between Glu78 and Glul72, the catalytic residues ofB. circulans/subtilis xylanase? As discussed in Chapter 1, section 1.5, the distance between the catalytic carboxylates of glycosidases is a primary determinant of the mechanism followed. In retaining glycosidases the average distance is 4.5-5.5 A, while in inverting enzymes it is greater (9-9.5 A) (Wang et al., 1994; McCarter & Withers, 1994; Davies & Henrissat, 1995; White et al., 1994). B. circulans/subtilis xylanase, a retaining glycosidase, contains two catalytic carboxylates, Glu78 and Glul72, which are separated by 5.5 A. One strategy for modifying the distance between Glu78 and Glu 172 is to selectively replace each glutamic acid with a shortened or lengthened analogue. The active site nucleophile, Glu78, is the chosen target for modification in this study. To increase the distance (>5.5 A) between the two catalytic residues, Glu78 will be replaced with aspartic acid (the mutant enzyme is denoted Glu78Asp). To decrease the distance (<5.5 A), Glu78 will be replaced with carboxymethylated Cys78, a lengthened carboxylic acid analogue (see Scheme 1-17, Chapter 1). This unnatural amino acid will be generated by treatment of the Glu78Cys mutant with iodoacetate. Several other investigators have used this combined approach of genetic 39 engineering with chemical modification to introduce unnatural amino acids into the active sites of other enzymes (Amano et al., 1994; Dhalla et al., 1994; Gloss & Kirsch, 1995; Lukac & Collier, 1988; Matsushima et al., 1994). This study describes the preparation and characterization of the carboxymethylated Glu78Cys mutant (denoted IAA-Glu78Cys). A detailed kinetic evaluation of this enzyme, native xylanase, and the Glu78Asp mutant will be performed to examine the mechanistic consequences of altering the length of the nucleophilic carboxyl side chain, thereby modifying the critical distance between the two catalytic carboxylates. 2.3 RESULTS AND DISCUSSION 2.3.1 PREPARATION AND PRELIMINARY CHARACTERIZATION OF NATIVE XYLANASE, GLU78CYS, AND GLU78ASP The mutagenesis, production, and purification of native xylanase and the Glu78 mutants were kindly performed by Dr. Warren Wakarchuk, National Research Council of Canada, Ottawa as previously described (Sung et al., 1993; Wakarchuk et al., 1994; refer to Chapter 5, section 5.3.1 for more details). Native xylanase from B. circulans/subtilis contains no cysteine residues. The introduction of a unique cysteine into the Glu78Cys mutant of B. circulans xylanase was confirmed by both thiol titrations and electrospray mass spectrometry (ESMS). The Cys mutant contained 1.07 (±0.03) free cysteines, as opposed to native xylanase which contained 0.06 (±0.06). The expected molecular mass decrease of 26 Da, due to replacement of a glutamic acid with a cysteine, was observed. For native B. circulans xylanase and Glu78Cys, 40 the molecular masses were 20 384 (±3) Da and 20 359 (±4) Da, respectively (the theoretical molecular masses are 20 384 Da and 20 358 Da). The replacement of a glutamic acid with an aspartic acid, in the Glu78Asp mutant of B. subtilis xylanase, was also confirmed by ESMS. The observed molecular mass of Glu78Asp, 20 358 (±4) Da, was within error of the theoretical mass of 20 356 Da (the theoretical mass of native B. subtilis xylanase is 20 370 Da). The secondary structure of Glu78Asp, and the Glu78Cys mutant, was examined using circular dichroism (refer to Figure 2-3). The close similarities between the spectra of the mutants and native xylanase suggest that the mutations have not significantly altered the conformation of the enzyme. _J I I l I i I i I 210 220 230 240 250 Wavelength (nm) Figure 2-3. The circular dichroism spectra of native xylanase ( ), Glu78Cys (— —), and Glu78Asp (- - -). 41 2.3.2 K I N E T I C A N A L Y S I S O F T H E G L U 7 8 C Y S M U T A N T Replacement of Glu78, the active site nucleophile, with a cysteine results in a carboxylate side chain being substituted with a shortened side chain containing a sulfhydryl group (-SH) (see Scheme 2-2). Relative to a carboxylate, a sulfhydryl group is a reactive nucleophile, but a poor leaving group. Therefore, one might expect the substituted side chain to still function effectively as a nucleophile if the sulfur can reach the anomeric center. However, it would serve as a relatively poor leaving group. C H 2 SH CHo CHo Glu78 Cys78 Scheme 2-2. The side chains of glutamate and cysteine (shown in their expected ionization states at pH 6). These alterations in the length, nucleophilicity, and leaving group ability of the side chain at the active site nucleophile position can potentially affect enzymatic activity in the following ways. One possibility is that the side chain, shortened by ~2 A, is now too far removed or sterically hindered to attack the anomeric carbon of the substrate. This would result in a "dead" enzyme, one incapable of hydrolyzing substrate. However, if the sulfhydryl containing side chain is of sufficient length to make this attack, one might expect the Glu78Cys mutant to hydrolyze substrate via the double displacement mechanism shown in Scheme 2-3. The initial formation of the covalent glycosyl-enzyme intermediate should be rapid due to the strong nucleophilicity of the sulfhydryl side chain. However, the breakdown of this intermediate to regenerate free enzyme, thereby continuing the catalytic cycle, should be impeded by the side chain's poor leaving group ability. Therefore, the rate of 42 glycosylation should be fast relative to the rate of deglycosylation, resulting in the accumulation of the intermediate and a phenolate burst. Scheme 2-3. The hydrolysis of 2',5'-dinitrophenyl fi-xylobioside (2.1) by the Glu78Cys mutant. The proposed reaction scheme is based on the double displacement mechanism illustrated in Chapter 1, Scheme 1-3 and assumes that the side chain of Cys78 is of sufficient length to attack the substrate's anomeric carbon. Incubation of Glu78Cys with 2',5'-dinitrophenyl (3-xylobioside (2,5-DNPX2, 2.1) did not result in a 2,5-dinitrophenolate (2.2) burst nor the accumulation of a thioglycoside intermediate. At all substrate concentrations tested (0.7 to 4.0 mM), a constant rate of 2,5-dinitrophenolate release was observed and the molecular mass of the enzyme was unchanged. The apparent k c a t and K m values for the Glu78Cys mutant, assayed with 2,5-DNPX2, were determined to be 1.1 x 10"5 s"1 (±1.1 x 10"6) and <0.7 mM, respectively. The corresponding 43 k c a t and K m values for native xylanase were 76 s"1 (±1.5) and 2.2 mM (±0.1), respectively. These results indicate that substituting a cysteine for a glutamic acid at the active site nucleophile position results in a dramatic decrease (~7 x 106 fold) in the k c a t value (refer to Appendix A for a brief discussion on the interpretation of k c a t , K m , and k c a t / K m ) . Since the mutant exhibited such low activity on 2,5-DNPX2, the most reactive substrate at our disposal, no other substrates were tested. The extremely low activity of the Glu78Cys mutant raised concerns that the activity might be due to contamination with native xylanase. An experiment using the mechanism based inactivator 2',4'-dinitrophenyl 2-deoxy-2-fluoro-(3-xylobioside (2F-DNPX 2) was performed to address this concern. Reaction of the mutant (95 nmol) with 2F-DNPX 2 (8.6 nmol) resulted in complete activity loss in 90 minutes. The observed activity cannot, therefore, be due to the Glu78Cys mutant since all activity was lost upon treatment with only 0.09 equivalents of inactivator. This was confirmed by ESMS analysis of the enzyme after inactivation, with identical molecular masses being observed for the treated and untreated mutant (20 358 ± 4 Da). The low activity observed for Glu78Cys is likely due to very low levels of contaminating native xylanase or another active mutant. Given the greater than three fold K m difference for native enzyme and the Glu78Cys mutant (2.2 and <0.7 mM, respectively), native xylanase is the least likely candidate. The results of the 2F-DNPX 2 experiment indicate that Glu78Cys is devoid of activity. This inactivity is not the consequence of a disrupted secondary structure, as indicated by circular dichroism (CD) analysis (refer to Figure 2-3), but is most likely due to the cysteine 44 side chain being too short or sterically hindered to attack the substrate's anomeric carbon. Therefore, a cysteine is incapable of functioning as the active site nucleophile in the retaining glycosidase, B. circulans xylanase. 2.3.3 PREPARATION AND CHARACTERIZATION OF CARBOXYMETHYLATED GLU78CYS As stated previously, a major goal of this study was to introduce a lengthened glutamic acid analogue into the active site nucleophile position of B. circulans/subtilis xylanase. It was believed this could be accomplished by alkylation of Cys78, in the Glu78Cys mutant, using iodoacetate (denoted IAA). In this reaction, illustrated in Scheme 2-4, the deprotonated sulfhydryl side chain of Cys78 makes an S N 2 nucleophilic attack on the methylene carbon of IAA to produce iodide and S-carboxymethyl Cys78. Conveniently Glu78Cys possesses only the cysteine at position 78, thus avoiding the problem of multiple cysteine labelling. IAA can also potentially alkylate functional groups other than thiols, but this can be minimized through the use of mild reaction conditions. 6 + C H 2 - C - 0 ' S" C H 2 S C H 2 i + Cys78 S-carboxymethyl Cys78 Scheme 2-4. Carboxymethylation of Cys78 using iodoacetate (IAA). 45 2.3.3.1 Iodoacetate Labelling of the Glu78Cys Mutant under Nondenaturing Conditions 2.3.3.1.1 Search for the Optimal Labelling Conditions The iodoacetate labelling of the Glu78Cys mutant was first attempted under nondenaturing conditions. A preliminary experiment showed very exciting results. Treatment of this "dead" mutant with IAA (300 fold molar excess, pH 9, 25 °C, 6 hours) caused a significant restoration of activity. This observed reactivation is presumably due to the carboxymethylation of Cys78, thus reintroducing a carboxylate into the active site nucleophile position of Glu78Cys. A search was then undertaken to find the optimal labelling conditions. Glu78Cys was incubated with different concentrations of IAA over a wide range of pH values and the activity monitored using the substrate 2,5-DNPX2. As shown in Table 2-2, increasing the reaction pH (from 7 to 10) and the IAA concentration (from 10 to 1200 fold molar excess) resulted in enhanced reactivation. Since a thiol must be deprotonated for alkylation to occur, these pH effects suggest that the sulfhydryl group of Cys78 is not fully deprotonated at pH values below 10. The largest observed activity regain occurred when Glu78Cys (0.1 mM) was treated with a 1200 fold molar excess of IAA (120 mM) for 27 hours (36 °C, pH 10, CAPSO). The reactivation of Glu78Cys, shown in Figure 2-4, is time-dependent and only occurs in the presence of IAA. Exposure of native enzyme to the identical conditions resulted in a decrease in activity. Since this activity loss occurs both in the presence and absence of IAA, it cannot be attributed to the nonselective labelling of residues other than Cys78. Rather, it is most likely due to enzyme instability at pH 10. 46 Table 2-2. The IAA Labelling of Glu78Cys under Different Conditions1 Labelling Reagent Molar Excess of Reagent pH Buffer Speciesb Time (hr)c Max. Activity (|imol 2,5-DNP/min/mg)d # of labelled amino acids0 IAA 325 7 BIS-TRIS 103 0.010 1-3 IAA 325 8 HEPES 102 0.012 1-5 IAA 325 9 AMPSO 102 0.072 1-10 IAA 325 9 f CAPSO 100 0.034 1-8 IAA 100 9 f CAPSO 100 0.015 1-5 IAA 10 9 CAPSO 101 0.002 1-2 IAA 1200 10 Borate 9 0.110g N D h I \.\ i:ofi • H i CAPSO 5-10 IAA 2000 10 CAPSO 27 0.157 N D h BrAA' 1200 10 CAPSO 28 0.118 4-10 IAA 325 11 pyrophos-phate 3 0.0006g N D h a All reactions were carried out at 36 °C using 120-140 mM buffer species. b For a listing of the full buffer names refer to Abbreviations and Symbols. c The incubation time at which maximum activity was observed. d DNP is the abbreviation for dinitrophenolate. e Determined by electrospray mass spectrometric analysis of the reaction mixtures. f The pH of the reaction mixture dropped during the reaction (at -100 hrs, the pH was ~6 and ~8 for 325 and 100 fold molar excess IAA, respectively). g The activity decreased rapidly when Glu78Cys was left at these conditions for longer than the indicated time. h Not determined (ND). 1 Bromoacetic acid (BrAA). 47 E *c E £ 'w z >. Q O CM 3 O O E 0) (0 tr 0.16 -0.12 -0.08 -0.04 -C M 0 > re O) E 1c E CL z Q in CN O E co re DC 400 800 1200 Time (min) 1600 Figure 2-4. Regain of activity upon IAA labelling of Glu78Cys and native xylanase (36 °C, pH 10, CAPSO). Glu78Cys was treated with 1200 (•) and 0 (B) fold molar excess IAA. Native xylanase was incubated with 1200 (O) and 0 (V) fold molar excess IAA also. The scales on the right and left y-axes are for native and Glu78Cys, respectively. ESMS analysis of the treated Glu78Cys mutant showed significant multiple labelling, with at least five different amino acids being alkylated by IAA (see Table 2-2 and Appendix D, Figure D- l ) . Similar multiple labelling was observed for IAA-treated native xylanase. In an effort to reduce the degree of multiple labelling, the alkylation reaction was repeated using bromoacetic acid (BrAA) instead of IAA. Since bromide has reduced leaving group ability relative to iodide, B r A A was expected to alkylate nucleophilic residues at a reduced rate, thereby increasing the reaction's selectivity. Unfortunately, no significant reduction in multiple labelling was observed when Glu78Cys was incubated with B r A A for 28 hours (see Table 2-2). 48 2.3.3.1.2 Active Site Titration using 2',4'-Dinitrophenyl 2-Deoxy-2-fluoro-pVxylobioside (2F-DNPX 9) [2.3] [2.4] In order to determine what fraction of the IAA-labelled GluVSCys1 mutant was active, an active site titration was performed using the mechanism based inactivator 2F-DNPX 2 (2.3). Incubation of labelled enzyme (0.15 mM) with 2F-DNPX 2 (1.6 mM) resulted in time-dependent inactivation, and a burst of released 2,4-dinitrophenolate (2,4-DNP, 2.4) was measured at 400 nm (refer to Figure 2-5). Since 1 mol of 2,4-DNP is released per mol of Time (min) Figure 2-5. Treatment of IAA-labelled Glu78Cys (0.15 mM) with 2F-DNPX2 (1.6 mM). The time-dependent inactivation of labelled Glu78Cys (•) is shown in curve a, while the release of 2,4-DNP (O) is shown in curve b. The left and right y-axes are for curves a and b, respectively. 1 IAA-labelled Glu78Cys was prepared using the optimal labelling conditions (1200 fold molar excess IAA, 36 °C, pH 10, CAPSO, 27 hrs). 49 "active" enzyme, the magnitude of the burst (observed AA400 = 0.112, expected AA400 = 1-6) indicates that only approximately 7% of the IAA-labelled enzyme was active and suggests that only partial carboxymethylation of Cys78 has occurred. 2.3.3.1.3 Methyl Methanethiolsulfonate Labelling of Glu78Cys and Subsequent IAA A thiol specific reagent, methyl methanethiolsulfonate (MMTS, 2.5) blocks the sulfhydryl side chain of cysteine with a thiomethyl group (molecular mass of 47 Da; refer to Scheme 2-5). In contrast to the IAA labelling, the modification of Cys78 by M M T S was selective and complete. ESMS analysis of Glu78Cys treated with M M T S (100 fold molar excess, pH 8, 4 °C, 24 hrs) revealed a single enzyme species with a molecular mass 46 Da greater than untreated Glu78Cys (refer to Table 2-3 and Appendix D, Figure D-2). Native xylanase, exposed to the same labelling conditions, showed no molecular mass increase relative to its control. This absence of labelling in native enzyme suggests that the modified amino acid in Glu78Cys is Cys78. Thiol titrations confirmed that Cys78 was indeed the residue being labelled, with MMTS-treated and untreated Glu78Cys having 0.12 and 1.07 free cysteines, respectively (see Table 2-3). Treatment C H 3 S 47 Da C H 2 SH O II + H 3 C - S - S - C H 3 CH 2 S + H - S - C H 3 o II II o II o Cys78 [2.5] CH3S- labelled Cys78 Scheme 2-5. The modification of Cys78 by MMTS (2.5). 2 For a description of how this value was determined, refer to section 5.3.5.2, Chapter 5. 50 Table 2-3. Thiol Titration and ESMS Results for MMTS-Treated Enzyme Sample Number of Free Cysteines/mol Enzyme Molecular Mass (g/mol) Glu78Cys 1.07 ±0.03 20 365 ± 2 Glu78Cys + MMTS 0.12 + 0.08 20 411 ± 2 Native 0.12 ±0.07 20 385 ± 2 Native + MMTS 0.05 ± 0.07 20 386 ± 4 E "c E S: Q in c\T "5 E =L CD flj DC 0.05 0.04 0.03 0.02 0.01 0 i 1 1 1 r ' I r 9-Q I ° i ° I i L 0 1000 2000 3000 4000 5000 Time (min) Figure 2-6. IAA treatment of MMTS-modified Glu78Cys (O) and a Glu78Cys control (•). Enzyme was treated with a 325 fold molar excess IAA (pH 9, AMPSO, 36 °C) and the activity monitored using 2,5-DNPX2-Reaction of MMTS-modified Glu78Cys with IAA resulted in insignificant reactivation relative to the Glu78Cys control (see Figure 2-6). Thus, blocking Cys78 with a thiomethyl group, thereby preventing the carboxymethylation of this residue by IAA, impedes enzyme reactivation. This finding strongly suggests that the carboxymethylation of Cys78 is 51 directly responsible for the observed reactivation of the Glu78Cys mutant upon treatment with IAA. 2.3.3.2 Iodoacetate Labelling of the Glu78Cys Mutant under Partially Denaturing Conditions In the absence of denaturants, the best IAA labelling conditions (1200 fold molar excess IAA, 36 °C, pH 10, CAPSO, 27 hrs) yielded a population of enzyme species of which only 7% was active. It was postulated that this low percentage of active species was due to Cys78 being relatively inaccessible to the anionic labelling reagent IAA. In an effort to improve the accessibility of Cys78, thereby increasing the degree of reactivation, the IAA reaction was repeated in the presence of low concentrations of urea (refer to Chapter 5, section 5.3.2.3 for the experimental details). The results, illustrated in Figure 2-7a, clearly show that the addition of urea (<3 M) served to increase the reaction rate, but did not increase the overall activity regain. Enzymatic activity was adversely affected when Glu78Cys was exposed to urea concentrations in excess of 1 M for more than 2.5 hours (at pH 10). It was hoped that lowering the reaction pH to 9 would lessen these adverse effects. As shown in Figure 2-7b, enzymatic activity was slightly more stable at pH 9, but again the overall activity regain was the same both in the presence and absence of urea. 52 a) 0 1000 2000 3000 Time (min) b) 0 1000 2000 3000 Time (min) Figure 2-7. The effects of low urea concentrations on the IAA labelling of Glu78Cys at (a) pH 10 and (b) pH 9. The urea concentrations tested were 3 (•), 2 (O), 1 (•), and 0 M (•). The reaction was performed at 36 °C using 1200 fold molar excess IAA. 53 2.3.3.3 Iodoacetate Labelling of the Glu78Cys Mutant under Fully Denaturing Conditions 2.3.3.3.1 Preparation of IAA-Glu78Cys The IAA labelling of Glu78Cys under nondenaturing conditions was nonselective, yielding approximately 7% active enzyme. While the addition of low concentrations of urea enhanced the reaction rate, this percentage of active species was unchanged. To improve the reaction's selectivity and increase the yield of active enzyme, the labelling was performed under fully denaturing conditions. The Glu78Cys mutant was incubated with IAA (10 fold molar excess) in the presence of 7.1 M urea (pH 7.5, 40 °C, 19 hrs). To refold the labelled enzyme, the reaction mixture was first diluted 20 fold with a 7.1 M urea solution (pH 7.5) and then dialyzed (refer to Chapter 5, section 5.3.2.4 for the complete experimental details). The IAA labelling of Glu78Cys under fully denaturing conditions proved to be a much better method to selectively and completely carboxymethylate Cys78. ESMS analysis of the refolded labelled mutant (denoted IAA-Glu78Cys; see Appendix D, Figure D-3) showed >90% conversion of Glu78Cys to a monolabelled species having a molecular mass of 20 420 (±4) Da (the mass of untreated Glu78Cys was 20 362 ± 3 Da, while the mass of a carboxymethyl group, -CH2COOH, is 59 Da). Native xylanase exposed to the same labelling conditions showed no sign of labelling by ESMS (refer to Figure D-3, Appendix D), thus suggesting that the amino acid carboxymethylated in the mutant is Cys78. Thiol titrations on the IAA-Glu78Cys mutant and a control sample (exposed to similar conditions, but in the absence of IAA) confirmed the selective carboxymethylation of Cys78, the samples having 0.12 and 0.92 free cysteines, respectively. The CD spectra of the labelled Cys mutant and 54 native xylanase, shown in Figure 2-8, are virtually superimposable indicating that correct refolding has occurred. 210 220 230 240 250 Wavelength (nm) Figure 2-8. The CD spectra of refolded IAA-treated Glu78Cys (- - -) and native xylanase (— —). The CD spectrum of untreated native xylanase ( ) was run as a control. 23.3.32 Active Site Titration using 2F-DNPX 2 An active site titration using 2F-DNPX2 (2.3) was performed to determine the amount of active IAA-Glu78Cys formed (refer to Chapter 5, section 5.3.5.2 for the experimental details). The results, illustrated in Figure 2-9, show the release of 2,4-dinitrophenolate (2.4) as a burst followed by a steady state. From the magnitude of the burst (observed AA400 = 0.48, expected AA400 = 0.53), it was determined that -90% of the enzyme was active. Since this activity is believed to be the result of carboxymethylation of Cys78, this value suggests 55 that -90% of the cysteine has been successfully alkylated. This corresponds well with the ESMS and thiol titration results described on the previous page which show the same degree of carboxymethylation. The steady state rate observed is most likely due to regeneration of free active enzyme that then undergoes a second round of inactivation. Assay of the IAA-Glu78Cys mutant after completion of the burst phase revealed an activity of less than 4% of the control. The molecular masses of the inactivated enzyme and a control were 20 684 (±4) Da and 20 420 (±3) Da, respectively. This molecular mass difference of 264 Da confirmed the covalent attachment of the 2-fluoroxylobiosyl label (molecular mass of 267 Da, structure 2.6 in Scheme 2-6). 0.6 -E c o o S 0.4 a> u c (0 Si o W Si < 0.2 -200 400 600 Time (min) 800 Figure 2-9. The active site titration of IAA-Glu78Cys (prepared under fully denaturing conditions). The time-dependent release of 2,4-dinitrophenolate was monitored at 400 nm. 56 2.3.3.3.3 Identification of the 2-Fluoroxylobiosyl Labelled Active Site Peptide by ESMS A 2-deoxy-2-fluoro glycoside, 2F-DNPX2 also inactivates native xylanase in a time-dependent fashion. Miao et al. (1994) demonstrated that this activity loss was due to the formation of a stabilized covalent glycosyl-enzyme intermediate between Glu78, the active site nucleophile, and the 2-fluoroxylobiosyl moiety of the inactivator. Since carboxymethylated Cys78 is a glutamic acid analogue, the inactivation of IAA-Glu78Cys should occur via the same process, the 2-fluoroxylobiosyl label now being covalently attached to S-carboxymethyl Cys78 (refer to Scheme 2-6). CH2 CH2 S-carboxymethyl Cys78 Scheme 2-6. The proposed inactivation of IAA-Glu78Cys by 2F-DNPX2. ESMS analysis of 2F-DNPX 2 inactivated IAA-Glu78Cys clearly demonstrated the covalent attachment of the 2-fluoroxylobiosyl label (molecular mass of 267 Da), thus lending support to the inactivation mechanism illustrated in Scheme 2-6. To confirm the identity of the amino acid to which the label was covalently linked, the inactivated mutant and a control (not exposed to 2F-DNPX2) were digested using pepsin and then subjected to ESMS analysis (refer to section 5.3.7, Chapter 5 for the complete experimental details). 57 The peptide mixtures were first separated by reverse-phase HPLC using the ESMS as a detector (shown in Figure 2-10a). The peptide bearing the 2-fluoroxylobiosyl label was located in this chromatogram by MS/MS analysis using a neutral loss experiment. This experiment involved exposing the charged peptides to low collision gas energies, thus inducing homolytic cleavage of the proposed ester linkage between the sugar moiety and the carboxyl side chain, while leaving the peptide linkages intact. The 2-fluoroxylobiosyl moiety departs as a neutral species and the parent peptide retains its charge. The two quadrupoles (Ql and Q3) are scanned in such a fashion that only charged peptides that differ by the mass of the sugar label are detected. When this experiment was conducted on the peptic digest and the spectrometer scanned for the mass loss m/z 133.5, corresponding to the loss of the sugar label from a doubly charged peptide, a single peak was observed (Figure 2-10b). This unique peak, having a m/z of 842.0 and a retention time of -20 minutes, was not detected in the control digest (Figure 2-10c). Since the doubly charged, labelled peptide has a m/z 842.0 (Figure 2-10d), the singly charged, unlabelled peptide must have a molecular mass of 1417 Da [(2 x 842) - 267]. A computer search for all peptides in the IAA-Glu78Cys mutant with a molecular mass of 1417 ± 2 Da yielded 13 possible peptides, three of which contained the expected carboxymethylated Cys78. 58 w c © > rr CO c © > a rr 100 75 50 25 Time (min) £5 >» 100-*co c 75 • 50 • <1) > 25 • CO rr 0 -Peptide 1 sCbea 12 r 15 T 18 Time (min) 0 JcSi^ejA. 12 15 Time (min) 18 t-21 0^  >. 100-1 '35 r- 75 • i— 0 ) Jl 50 • > 25 • CO n rr u 300 842.0 600 900 1200 1500 m/z Figure 2-10. ESMS analysis of digested 2F-DNPX2 inactivated IAA-Glu78Cys (prepared under fully denaturing conditions), (a) In normal MS mode, (b) in neutral loss mode, (c) digested IAA-Glu78Cys control in neutral loss mode, (d) mass spectrum of peptide I in panel b. 59 Sequence information on the 2-fluoroxylobiosyl labelled peptide was obtained by performing an MS/MS experiment at higher collision gas energies (as described in section 5.3.7, Chapter 5). The doubly charged, labelled peptide (m/z 842.0) was selectively introduced into a collision cell in the second quadrupole (Q2), where it was subjected to collision-induced fragmentation. The mass of the resulting daughter ions was detected in the third quadrupole (Q3). Illustrated in Figure 2-11 is the family of daughter ions produced. Y-G-W-lT-R b3: 408<-Figure 2-11. The MS/MS daughter ion spectrum of the 2-fluoroxylobiosyl labelled peptide (m/z 842.0 in the doubly charged state). 60 The fragmentation patterns of the 13 candidate peptides, having masses of 1417 ± 2 Da, were predicted using a computer program. Of these 13 candidates, only one peptide could yield the labelled daughter ions shown in Figure 2-11. This peptide had the sequence Y G W T R S P L I X Y (where X corresponds to a carboxymethylated cysteine; refer to Abbreviations and Symbols for the full names of the amino acids). The peak at m/z 1418 results from the loss of the 2-fluoroxylobiosyl label plus a proton from the doubly charged, labelled peptide (m/z 842.0), thus giving the singly charged, unlabelled peptide. The labelled peaks (blO, b9, b8, b6, b5, and b3) arise from the loss of neutral fragments from the C-terminus of the m/z 1418 peptide. The peak at m/z 1236 (blO) is attributed to the loss of the C-terminal tyrosine (Y, m/z 181) from the peak at m/z 1418. The other peaks (b9, b8, b6, b5, and b3) result from the respective losses of X Y , IXY, PLLXY, SPLLXY, and TRSPLLXY fragments from the C-terminus. A parallel "a" series of peaks are also observed, and these result from the additional loss of CO (28 Da) from the peaks in the "b" series. The 2-fluoroxylobiosyl labelled peptide in digested, inactivated IAA-Glu78Cys, identified as Y G W T R S P L I X Y (69-79), is identical to the previously identified active site peptide in digested, inactivated native xylanase (with the exception that X , carboxymethylated cysteine, is replaced with E, glutamic acid) (Miao et al., 1994). In native xylanase, the covalent attachment of the 2-fluoroxylobiosyl label to Glu78, through an ester linkage, was confirmed by performing an aminolysis experiment. Therefore by analogy, the sugar label should be attached to carboxymethylated Cys78 in 2F-DNPX 2 inactivated IAA-Glu78Cys. 61 2.3.4 KINETIC EVALUATION OF IAA-GLU78CYS, NATIVE XYLANASE, AND THE GLU78ASP MUTANT 2.3.4.1 Kinetic Analysis of the 2F-DNPX2 Inactivation of IAA-Glu78Cys The inactivation of IAA-Glu78Cys by 2F-DNPX2, first demonstrated in the active site titration experiment (section 2.3.3.3.2), was examined more closely by determining the inactivation rates at a series of inactivator concentrations (refer to section 5.3.5.3, Chapter 5 for the experimental details). From these results, shown in Figure 2-12, the values for k;, K,, and kj/K; were determined (refer to Table 2-4). The k; and K, values are very approximate since the relative insolubility of the inactivator precluded study at 2F-DNPX2 concentrations even approaching its Kj value. However, the k/Kj value, determined from the slope of the reciprocal plot in Figure 2-12b, is accurate (refer to Appendix A for a brief discussion on the interpretation of k; and Kj). The kinetic scheme shown below was followed, where E E + I-DNP « - EI -DNP E-I DNP corresponds to free enzyme, I to 2F-xylobiose, and DNP to 2,4-dinitrophenolate. Table 2-4. The Inactivation Constants for IAA-Glu78Cys and Native Xylanase Enzyme kj (min1) Kj (mM) ki/Kjtmin^mM-1) IAA-Glu78Cys 0.10 ±0.01 3.2 ±0 .3 0.031 Native3 2.2 ±0 .6 6.4 ±2.1 0.34 a The inactivation constants for native xylanase were taken from Miao et al. (1994). 62 Q I / I i I i I i L 0 0.8 1.6 2.4 1 / [2F-DNPXJ (mM 1 ) Figure 2-12. The inactivation of IAA-Glu78Cys (prepared under fully denaturing conditions) by 2F-DNPX2. (a) The plot of % activity versus time at the following inactivator concentrations: 0.46 mM (•), 0.68 mM (O), 1.08 mM (A), and 1.91 mM (T). (b) The double reciprocal plot of the first-order rate constants obtained from (a). 63 As shown in Figure 2-12a, the inactivation of IAA-Glu78Cys by 2F-DNPX2 occurred in a time-dependent fashion with the rate of enzyme death increasing as the inactivator concentration was increased. The determined kj/Kj value for IAA-Glu78Cys is reduced 11 fold relative to the corresponding value for native enzyme. Since the Kj values for the two enzymes are approximately the same, this decrease in kj/Kj is mainly the result of a depressed kj value. This suggests that replacing Glu78 with its lengthened analogue, S-carboxymethyl Cys78, does not significantly perturb the binding of 2F-DNPX2, but does decrease the rate of formation of the stabilized glycosyl-enzyme intermediate by -20 fold. 2.3.4.2 Reactivation of the Inactivated IAA-Glu78Cys Mutant Inactivated IAA-Glu78Cys, freed of excess 2F-DNPX 2 , was incubated with buffer or xylobiose (155 mM) at 40 °C and aliquots assayed for activity regain due to regeneration of free enzyme (refer to section 5.3.5.4, Chapter 5 for the experimental details). The return of activity,3 shown in Figure 2-13, occurred via a first-order process. Analysis of this data yielded the following reactivation rate constants, k r e o b s = 0.00150 ± 0.00001 min"1 and 0.00190 ± 0.00006 min"1, for buffer and 155 mM xylobiose, respectively. Thus, the addition of the sugar xylobiose (2.7) enhanced the rate of reactivation by a small factor. In the presence of xylobiose, the trapped intermediate is likely turned over via the transglycosylation process illustrated in Scheme 2-7. This observed reactivation, both by 3 The data shown in Figure 2-13 have been corrected for enzyme death due to denaturation using data from a non-inhibited control sample. 64 hydrolysis and transglycosylation, clearly demonstrates that the trapped intermediate is catalytically competent. c o CO > o CO CD rr 100 -0 600 1200 1800 2400 Time (min) Figure 2-13. The reactivation of inactivated IAA-Glu78Cys by 155 mM xylobiose (A) and buffer (•). The data shown have been corrected for enzyme death due to denaturation using data from a non-inhibited control sample. HH°0 H O y F 0 CHo CH 2 I HO tfcO^ [2.7] HO OH S-carboxymethyl Cys78 F HO CH 2 S • CH 2 S-carboxymethyl Cys78 HO HO HO OH Scheme 2-7. The proposed reactivation of inactivated IAA-Glu78Cys by xylobiose (2.7). The reactivation of 2F-DNPX2 inactivated native xylanase, incubated with just buffer, also occurred via a first-order process with a k r e o b s = 0.0021 min"1 (Miao et al., 1994). The 65 spontaneous reactivation rate constants for native xylanase and IAA-Glu78Cys only differ by a factor of 1.4. Thus, lengthening the carboxyl side chain at position 78 does not significantly alter the rate at which the trapped glycosyl-enzyme intermediate is hydrolyzed. As with IAA-Glu78Cys, the addition of sugars increased the reactivation rate for inactivated native xylanase (by up to a factor of 14 relative to spontaneous hydrolysis; Miao et al., 1994). 2.3.4.3 Determination of the Stereochemical Course of Hydrolysis for IAA-Glu78Cys The 2F-DNPX2 inactivation of IAA-Glu78Cys, and its subsequent reactivation, clearly demonstrated the existence of a covalent glycosyl-enzyme intermediate. This suggested that IAA-Glu78Cys (prepared under fully denaturing conditions) was operating via the double displacement (retaining) mechanism described in section 1.3, Chapter 1. To confirm that the modified cysteine mutant was indeed a retaining glycosidase, the enzyme-catalyzed hydrolysis of the substrate 2,5-DNPX2 was monitored using *H N M R spectroscopy (for the experimental details refer to section 5.3.3.6, Chapter 5). Figure 2-14a shows the anomeric proton region of 2,5-DNPX2 in buffer, prior to addition of enzyme. The doublet at 8 5.29 ppm (7 = 6.8 Hz) arises from the axial anomeric proton of the substrate, while the singlet at 8 4.48 ppm (marked x) is due to an impurity in the D2O. The broad peak at 8 4.60 to 4.74 ppm is due to HOD. Eight minutes after the addition of IAA-Glu78Cys (Figure 2-14b), the doublet at 8 5.29 ppm has almost disappeared and a new doublet (8 4.47 ppm, J = 7.8 Hz), due to the axial anomeric proton of |3-xylobiose, has appeared. Figure 2-14c, taken 20 minutes after enzyme addition, shows complete substrate hydrolysis and the emergence of a new doublet at 8 5.06 ppm (J = 3.6 Hz) which is increased further in intensity at 72 minutes (Figure 2-14d). This new doublet arises from the equatorial anomeric proton of a-xylobiose which is formed by mutarotation of the initially formed (3-xylobiose. These results confirm that IAA-Glu78Cys, like native xylanase, is a retaining glycosidase. —i 1 1 1 1 ' 1 5.2 5.0 +.8 4.6 (ppm) Figure 2-14. Determination of the stereochemical course of hydrolysis of 2.5-DNPX2 by IAA-Glu78Cys (prepared under fully denaturing conditions). 'H NMR spectra are for the anomeric proton region of the substrate before addition of enzyme (a) and 8, 20, and 72 minutes after enzyme addition (b, c, and d, respectively). 67 2.3.4.4 Steady State Kinetic Studies using Synthetic /5-Xylobioside Substrates The kinetic parameters k c a t , K m , and k c a t / K m were determined for a series of aryl (3-xylobiosides with IAA-Glu78Cys, native xylanase, and the Glu78Asp mutant (refer to Table 2-5; the data are presented as Lineweaver-Burk plots in Appendix C). Values of k c a t and k c a t / K m for IAA-Glu78Cys are based upon the assumption of 100% labelling, consistent with the active site titration performed (90 + 10% labelling, see section 2.3.3.3.2). For some substrates, as indicated in Table 2-5, the highest substrate concentration tested was below the listed K m value. For these substrates, the k c a t and K m values given are less accurate than the errors would suggest, but the k c a t / K m values, determined from the slope of the Lineweaver-Burk plot, are accurate (for a brief discussion on the interpretation of k c a t , K m , and k c a t / K m see Appendix A). IAA treatment of Glu78Cys, to produce a carboxymethylated cysteine at position 78, resulted in a remarkable recovery of activity. While Glu78Cys was unable to hydrolyze 2,5-DNPX2, IAA-Glu78Cys (prepared under fully denaturing conditions) turned over this substrate with a k c a t / K m value reduced only 16 fold relative to native xylanase. IAA-Glu78Cys was also active on the other aryl p-xylobiosides tested, with these k c a t / K m values being decreased 30-100 fold relative to native xylanase. These reductions in the k c a t / K m values are the cumulative result of both depressed k c a t values (-4-30 fold) and elevated K m values (-4-10 fold). Thus, inserting a lengthened carboxyl side chain into the active site nucleophile position only moderately perturbs substrate binding and catalysis. The IAA 68 treatment4 did not have a detrimental effect on native activity, with IAA-treated and untreated native xylanase having approximately the same k c a t / K m value for 2,5-DNPX2 (30 and 35 s"1 mM" 1 , respectively). The Glu78Asp mutant was significantly less active on the aryl [3-xylobiosides tested than either native xylanase or IAA-Glu78Cys. Its k c a t / K m values were reduced -1600-5000 fold relative to native xylanase, while -50-200 fold with respect to IAA-Glu78Cys. Since the K m values for the Asp mutant and native xylanase are very similar, these large reductions in k c a t / K m reflect depressed k c a t values. Replacing Glu78 with its shortened analogue, aspartic acid, appears to have no significant effect on substrate binding, but does have a dramatic influence on the rate of catalysis. For the three enzyme species, the Br0nsted relationships between log k c a t / K m for each substrate and the p K a of the aglycone are presented in Figure 2-15 (refer to Appendix B for a general introduction to Br0nsted relationships). A linear relationship was observed for all three enzymes, with the k c a t / K m values (which reflect the first irreversible step) increasing in step with the aglycone leaving group ability. Since the aglycone is only present in the glycosylation step of the double displacement (retaining) mechanism (refer to Scheme 1-3, Chapter 1), this correlation confirmed that glycosylation is the first irreversible step for native xylanase and the two mutants. 4 For the IAA labelling conditions, performed in the presence of 7.1 M urea, see section 5.3.2.4, Chapter 5. 69 Table 2-5. Kinetic Parameters for Aryl P-Xylobiosides with Native Xylanase and Mutants Substrate Kinetic Parameter Native IAA-Glu78Cys Glu78Asp 2,5-DNPX2a (pK a=5.15) b kcat ( s ') 76+ 1.5d 21 ± 1.9e 0.026 ± 0.0005 K m ( m M ) 2 . 2 ± 0 . 1d 10 ± 1.2e 2.4 ±0.1 k c a ^Cs - ' m M " 1 ) 0 35± l . l d 2.2 ± 0.03e 0.011 ±0.0002 3,4-DNPX2a (pK a = 5.36)b kCat ( s ) 8.3 ± 0.4d 1.6±0.2 e 0.005 ± 0.0002 Km(mM) 3.4 ± 0.3d 31 ± 6 . 1 e 9.2 ±0.8 k c a t / K m ( s - 1 m M - 1 ) c 2 . 7 ± 0 . 1d 0.042 ± 0.001e 0.0006 ± 2xl0" 5 ONPX 2 a (pK a = 7.22)b kcat ( s ) 9.6 ±0.1 1.3 ±0.04 0.005 ± 0.0002 K m ( m M ) 14 ±0.5 54 + 3.5 14 ±1.6 kcat/KmCs-'mM-1)0 0.66 ± 0.01 0.021 ±0.0003 0.0004 ± 9xl0" 6 PNPX 2 a (pK a = 7.18)b kcat( s ') 24 ±2.1 >0.3e,f 0.008 ± 0.0001e K m ( m M ) 49 ± 6.4 >25e>f 80 ± 1.5e k c a t / K m ( s - ' mM" 1) ' 0.43 ±0.01 0.013 ± 0 . 0 0 0 2 e f 9.0xl0" 5 ± l x l 0 " 6 e PhX 2 a (pK a = 9.99)b kcat ( s l ) 0.051 ±0.004 0.002 ± 8x10"5 e g K m ( m M ) 8.7 ± 1.3 31 ± 2 . 2e g k c a ^ K m ( s - 1 m M - 1 ) c 0.005 ± 0.0003 5.0xlO"5 ±9xlO" 7 e g d For a listing of the full substrate's name see Abbreviations and Symbols. b The pK a value given is for the aglycone. c The k<.JKm values were determined from the slope of the Lineweaver-Burk plot. d Data taken from Ziser et al. (1995). eThe highest substrate concentration tested was <Km. fNo saturation kinetics were observed for the substrate range examined (2-22 mM). 8 Enzyme-catalyzed hydrolysis was not detected. 70 PK a Figure 2-15. The Br0nsted plot (log kcat/Km versus aglycone pKa ) for native xylanase (•), the Glu78Asp mutant (A), and the IAA-Glu78Cys mutant (O) prepared under fully denaturing conditions. The calculated slope (Pig) and the correlation coefficient (p) for native xylanase, IAA-Glu78Cys, and Glu78Asp were determined to be p l g = -0.7 (±0.1), p = 0.97; p l g = -0.8 (±0.2), p = 0.94; and p i g = -0.6 (±0.3), p = 0.8, respectively. For native xylanase, the slope of the J3r0nsted plot (Pi g = -0.7) indicates very substantial glycosidic bond cleavage at the glycosylation transition state with very little proton donation. Interestingly, a very similar P i g value (-0.8) was obtained for IAA-Glu78Cys, with there being essentially no difference in the values within the error of their determination. While the data for the Glu78Asp mutant are more limited, a very similar trend was observed. This implies that even though there are substantial effects on rate due to changes in the nucleophile position, there is very little effect 71 on the reaction mechanism, at least with respect to charge development on the aglycone oxygen. 2.3.4.5 The pH Dependence of kcat/Km Values of k c a t / K m for the hydrolysis of ONPX2 by native xylanase and IAA-Glu78Cys were determined as a function of pH as described in Materials and Methods (Chapter 5, section 5.3.3.5).5 Previous work had indicated that the IAA-Glu78Cys mutant is relatively unstable over long periods at 40 °C, thus these studies were performed at 25 °C, a temperature at which the modified mutant is stable over the pH range 3.5 to 7.5. Presented in Figure 2-16 are the resulting plots of k c a t / K m versus pH. As can be seen, the pH profile for both native xylanase and the IAA-Glu78Cys mutant is a bell-shaped curve, suggesting that activity is dependent upon two ionizable catalytic groups. The ionization in the acidic limb (pK a 0 is most likely associated with the nucleophilic carboxyl group at position 78, since this residue must be deprotonated to be catalytically active. The ionization in the basic limb (pKa2) likely reflects that of the acid catalyst, Glu 172, which must be protonated in the glycosylation step to be catalytically active. 5 Due to the low activity of Glu78Asp on ONPX 2 (kc a t/Km = 0.0004 s"1 mM"1), the pH dependence of kca t/K, could not be examined for this mutant. 72 0.4 CO CO c JO >. 0.3 x 0) > CO 0.2 ^ 0.1 E o 1 1 1 /o 1 1 o 1 1 1 i i i i _ lo 1 — ' I 1 1 1 1 i i i PH 0.025 0.02 0.015 0.01 0.005 0 8 CO >. O 00 3 < E u Figure 2-16. The pH dependence of kcat/Kmfor native xylanase (•) and IAA-Glu78Cys (O) prepared under fully denaturing conditions. Lines shown represent fits to data for an enzyme with two ionizing groups. The scales on the right and left y-axes are for IAA-Glu78Cys and native xylanase, respectively. Analysis of the k c a t / K m versus pH data (as described in section 5.3.3.5, Chapter 5) yielded the following p K a values for native xylanase (pK a i = 4.6 ± 0.04, p K a 2 = 6.8 ± 0.03) and IAA-Glu78Cys (pK a l = 3.3 ± 0.08, p K a 2 = 6.5 + 0.07). These p K a values are attributed to the ionizable catalytic groups in the free enzyme (Fersht, 1985). The p K a values of 4.6 (pK a i) and 6.8 (pK a 2 ) , determined for native xylanase, have been shown by 1 3 C N M R studies to be due to Glu78 and Glul72, respectively (Mcintosh et al., 1996). Replacing Glu78 with an S-carboxymethyl cysteine results in very little change in p K a 2 , that of Glu 172, suggesting that no significant structural changes occur at the active site which affect the environment of 73 the putative acid/base catalyst. In contrast, a significant depression in p K a i , that of Glu/IAA-Cys78, is observed upon lengthening the nucleophilic carboxyl side chain. This decrease is most likely a consequence of an inductive effect from the thioether linkage. Indeed, a comparison of the p K a values of the analogous carboxylic acids C H 3 C H 2 C 0 2 H (pK a = 4.88) and C H 3 S C H 2 C 0 2 H (pK a = 3.72) reveals a p K a depression of approximately 1 pH unit (Brown et al., 1955). As illustrated in Figure 2-16, the pH optimum for IAA-Glu78Cys occurs at pH ~5, instead of at pH ~6 as for native enzyme. Upon going from pH 6 to 5, the k c a t / K m value determined for IAA-Glu78Cys with O N P X 2 (25 °C) increases by -1.5 fold. A similar enhancement in k c a t / K m was observed for 2,5-DNPX 2 , at 40 °C, upon adjusting the reaction pH from 6 to 5 (the k c a t / K m values were 2.2 and 3.2 s"1 mM" 1 , respectively). Therefore, the k c a t / K m values listed for IAA-Glu78Cys in Table 2-5, determined at pH 6 to allow for direct comparison with native xylanase, are slightly lower than what would be obtained at the mutant's optimum pH. 2.4 C O N C L U S I O N For the retaining endo-(3-l,4-xylanase from B. circulans, a cysteine is incapable of fulfilling the role of the active site nucleophile since substitution of a cysteine for Glu78, the catalytic nucleophile, resulted in an inactive mutant. Circular dichroism analysis indicated that this absence of activity was not due to a disrupted secondary structure. Since sulfhydryl 74 groups are typically good nucleophiles, a probable explanation is that the cysteine side chain is either too short or too sterically hindered to attack the anomeric carbon of the substrate. To probe this, the probable active site environment of the Glu78Cys mutant was examined by molecular modelling6 using the published X-ray crystal structure (Campbell et al., 1993; Wakarchuk et al., 1994). Replacement of Glu78 by a cysteine revealed that the sulfhydryl group is ~2 A further removed from the anomeric carbon of the scissile bond than the oxygens of the carboxylate group of Glu78 (the measured distances were 5.3 A and 3.4 A, respectively). This shortened sulfhydryl side chain, now situated in a hydrophobic environment, is relatively inaccessible to solvent. Surface accessibility calculations estimated the exposed surface area of the sulfhydryl group of Cys78 to be 2.6 A2 (the surface area of a fully exposed sulfhydryl group is >40 A2). As expected, the cysteine side chain appears to be too far removed, and located in too unfavorable an environment, to act as the nucleophile in the glycosylation step of the double displacement (retaining) mechanism. While completely inaccessible to the substrate 2,5-DNPX 2 , the side chain of Cys78 could be partially labelled using iodoacetate under nondenaturing conditions. The best IAA labelling conditions discovered (1200 fold molar excess IAA, 36 °C, pH 10, 27 hrs) yielded only 7% active enzyme and resulted in significant multiple labelling. In contrast, Cys78 could be selectively and completely modified using the neutral reagent methyl methanethiolsulfonate. Thus, the sulfhydryl side chain is fully accessible to small, neutral reagents, but only poorly accessible to small, anionic reagents. This preference for neutral compounds is in agreement with the predicted hydrophobic environment of the Cys78 side 6 The molecular modelling and surface accessibility calculations were kindly done by Manish Joshi and Dr. Greg Connelly, respectively. 75 chain. Such an environment would be highly unfavorable for both deprotonated Cys78 and the negatively charged IAA. To selectively and more completely carboxymethylate Cys78, the IAA labelling was performed under fully denaturing conditions (7.1 M urea, 10 fold molar excess IAA, pH 7.5). ESMS analysis and thiol titrations indicated that >90% of Cys78 had been selectively carboxymethylated, while the active site titration estimated -90% active enzyme. The IAA-modified Glu78Cys mutant (denoted IAA-Glu78Cys) was highly active with a range of aryl xylobioside substrates, catalyzing their hydrolysis via a retaining (double displacement) mechanism involving a covalent glycosyl-enzyme intermediate. Evidence for the double displacement (retaining) mechanism was obtained using a variety of techniques. The hydrolysis of 2,5-DNPX2 by IAA-Glu78Cys, monitored by *H N M R spectroscopy, clearly yielded an initial sugar product of retained stereochemistry. The existence of a covalent glycosyl-enzyme intermediate, a key feature of the retaining mechanism, was demonstrated in the 2F-DNPX2 inactivation studies. ESMS analysis of inactivated IAA-Glu78Cys revealed a mass increase consistent with the covalent attachment of a 2-fluoroxylobiosyl label. Evidence that the intermediate formed involved carboxymethylated Cys78 was obtained from the mass spectrometric identification of the glycosylated peptide as one including the modified Cys78. The catalytic competence of the intermediate trapped in this way was demonstrated by the reactivation of the purified trapped intermediate either by hydrolysis or transglycosylation. These results provide strong evidence that IAA-Glu78Cys functions via a double displacement mechanism in which carboxymethylated Cys78 acts as the catalytic nucleophile. 76 Comparison of the kinetic parameters for the IAA-Glu78Cys and Glu78Asp mutants with those for native xylanase allowed some insight into the effects of changing the position of the catalytic nucleophile. The effects of shortening the nucleophile on rate were significant with the k c a t / K m values, reflecting the glycosylation step, being reduced 1600-5000 fold relative to native enzyme. This rate reduction is consistent with that seen in other glycosidases modified in this same manner (Withers et al., 1992; Yuan et al., 1994). What is perhaps surprising is how small the effects of lengthening the nucleophile are, with k c a t / K m values being reduced only 16-100 fold relative to native xylanase. This indicates that the enzyme is flexible enough to accommodate the lengthened side chain, with the side chain presumably folding down to adopt a position in which the carboxylate can function effectively as a nucleophile. Indeed, molecular modelling studies in which a carboxymethyl group was built onto Cys78 confirmed that such space is available. The rate decreases observed probably reflect the steric and entropic costs of folding up the lengthened side chain. These rate differences do not appear to be reflected in the degree of bond cleavage or proton donation at the glycosylation transition state, as demonstrated by the similar p l g values. Thus, the relatively fixed environment of the active site dictates a very similar transition state structure in each case. That transition state structure, however, is less frequently attained when the carboxylate is imperfectly aligned. CHAPTER 3 MODIFYING THE ACID/BASE CATALYST OF BACILLUS CIRCULANS/SUBTILIS XYLANASE 78 3.1 I N T R O D U C T I O N The endo-p-l,4-xylanase from Bacillus circulans/subtilis catalyzes the hydrolysis of xylan, and P-xylobiosides, with net retention of anomeric configuration (Gebler et al., 1992a). The key catalytic residues in the proposed double displacement mechanism, illustrated in scheme 1-3 (Chapter 1), are two carboxylates, positioned 4.5-5.5 A apart in the active site (Sinnott, 1990; McCarter & Withers, 1994; Davies & Henrissat, 1995). One carboxylate functions as the catalytic nucleophile, attacking the anomeric carbon of the substrate to form the covalent glycosyl-enzyme intermediate. The other carboxylate plays the dual role of acid/base catalyst. In the glycosylation step, this residue acts as an acid catalyst, donating a proton to the departing leaving group. The carboxylate functions as a base catalyst in the subsequent deglycosylation step, deprotonating the attacking water. A member of glycosidase family G (Gilkes et al., 1991; Oku et al., 1993), B. circulans/subtilis xylanase possesses two conserved active site carboxylates, Glu78 and Glul72, which are mechanistically important (Wakarchuk et al., 1994; Campbell et al., 1993). Glu78 was previously identified as the active site nucleophile through labelling studies using the 2-deoxy-2-fluoro glycoside, 2',4'-dinitrophenyl 2-deoxy-2-fluoro-P-xylobioside (Miao et al., 1994). The detailed kinetic analysis of native xylanase and several Glu78 mutants described in Chapter 2 was consistent with this assignment. By default, Glu 172 has been assigned the role of acid/base catalyst. While the 2-deoxy-2-fluoro glycosides specifically target the active site nucleophile of retaining p-glycosidases, no equivalent labelling reagent exists to unequivocally identify the acid/base catalyst. To circumvent this problem, a method employing the techniques of 79 sequence alignment, site-directed mutagenesis, and kinetic analysis was recently developed to identify this catalytic residue (MacLeod et al., 1994; described in Chapter 1, section 1.4.1.5). Potential candidates for the acid/base catalyst are first identified by comparing the amino acid sequences of related enzymes. These candidates, conserved glutamic or aspartic acid residues, are then mutated to either a glycine or an alanine, amino acids bearing neutral side chains. The mutant enzymes are subsequently tested, in both the absence and presence of exogenous nucleophiles, with substrates possessing different requirements for acid catalysis. The above method was successfully applied to Cellumonas fimi exoglycanase (MacLeod et al., 1994) and Agrobacterium P-glucosidase (Wang et al., 1995), identifying Glu 127 and Glu 170 as the respective acid/base catalysts. These assignments were consistent with the recently published three-dimensional structures of C. fimi exoglycanase (White et al., 1994) and two P-glycosidases which are related to Agrobacterium P-glucosidase (Barrett et al., 1995; Wiesmann et al., 1995). 3.2 SPECIFIC A IMS OF THIS STUDY 1. What are the mechanistic consequences of removing the carboxyl side chain at position 172, the site of the putative acid/base catalyst of B. circulans/subtilis xylanase? Mutants have been constructed in which Glu 172 has been replaced with a cysteine or a glutamine (the mutants are denoted Glul72Cys and Glul72Gln, respectively). Kinetic studies done by Wakarchuk et al. (1994) indicated that these mutants, possessing no carboxyl side chain at position 172, were completely inactive on xylan, the natural substrate. In this study, the Glul72Cys and Glul72Gln mutants will be evaluated using synthetic p-xylobioside substrates, in both the presence and absence of the exogenous nucleophile azide. 80 These substrates, possessing different leaving groups,1 have varying requirements for acid catalysis. This kinetic analysis, modelled after the method developed by MacLeod et al. (1994; described in section 3.1), should provide definitive proof that Glul72 is the acid/base catalyst of B. circulans/subtilis xylanase. 2. What are the mechanistic consequences of altering the length of the carboxyl side chain at position 172, thereby modifying the distance (5.5 A) between the two catalytic carboxylates ofB. circulans/subtilis xylanase? To alter the length of the carboxyl side chain at the putative acid/base catalyst position, Glu 172 will be substituted with an aspartic acid and a S-carboxymethyl cysteine. A shortened glutamic acid analogue, aspartic acid will be introduced using site-directed mutagenesis (the mutant enzyme is denoted Glul72Asp). To produce S-carboxymethyl Cysl72, the lengthened analogue, the Glul72Cys mutant will be selectively carboxymethylated at position 172 using iodoacetate (the mutant enzyme is denoted IAA-Glul72Cys). This study describes the preparation and characterization of the IAA-Glul72Cys mutant. This enzyme, native xylanase, and the Glul72Asp mutant will be subjected to a detailed kinetic evaluation using both synthetic (3-xylobioside substrates and the natural substrate xylan. The active site nucleophile, Glu78, had previously been subjected to similar modifications and a kinetic analysis performed (the results are presented in Chapter 2, section 2.3.4). The mechanistic consequences of modifying the side chain length of the acid/base catalyst will be discussed in relation to the analogous work done on the catalytic nucleophile. 1 The pK a values of the leaving groups range from 5.15 to 9.99. 81 3. Investigating the inactivation of the Glul72Cys mutant by the disulfide reagent, di-R,p]'-D-xylopyranosyl disulfide (denoted XSSX). The effects of pH on the inactivation rate will be studied and the inactivation parameters kj and Kj determined. To discern the mechanism by which X S S X inactivates Glul72Cys, the inactivated mutant will be subjected to ESMS analysis and thiol titrations. Native xylanase and the active site nucleophile cysteine mutant, Glu78Cys, will also be treated with X S S X to determine the specificity of the disulfide. 3.3 RESULTS AND DISCUSSION 3.3.1 PREPARATION AND PRELIMINARY CHARACTERIZATION OF NATIVE XYLANASE AND THE GLU172CYS, GLU172ASP, AND GLU172GLN MUTANTS The mutagenesis, production, and purification of native xylanase and the Glu 172 mutants were kindly performed by Dr. Warren Wakarchuk, National Research Council of Canada, Ottawa, as previously described (Sung et al., 1993; Wakarchuk et al., 1994; refer to Chapter 5, section 5.3.1 for more details). The three Glul72 mutants were subjected to ESMS analysis. The results, summarized in Table 3-1, confirmed the replacement of a glutamic acid by a cysteine and an aspartic acid in the Glul72Cys and Glul72Asp mutants, respectively. The expected molecular mass reductions of 26 and 14 Da, relative to native xylanase, were observed for the Cys and Asp mutants, respectively. Thiol titrations provided additional evidence of the introduction of a unique cysteine in the Glul72Cys mutant. The Cys mutant contained 1.05 free cysteines, while native xylanase contained 0.00. 82 The secondary structure of the three Glu 172 mutants was examined using circular dichroism (CD). As shown in Figure 3-1, the CD spectra of the mutants and native xylanase are very similar. These results indicate that the mutations have had little effect on the enzyme' s conformation. Table 3-1. ESMS Analysis of Native Xylanase and the Glul72 Mutants Enzyme Observed Molecular Mass (Da) Observed Mass Reduction (Da)a Expected Mass Reduction (Da)a Native (BCX)b 20 394 ± 3 Glul72Cys (BCX) 20 367 ± 2 27 26 Native (BSX) (20 380)c Glul72Asp (BSX) 20 367 ± 2 13 14 Glul72Gln (BSX) 20 379 ± 2 1 1 a Relative to native BCX or BSX. b BCX and BSX correspond to Bacillus circulans xylanase and Bacillus subtilis xylanase, respectively. The xylanases from these two species are identical in activity and other properties, only differing in the amino acid residue at position 147 (in BCX it is a threonine while in BSX it is a serine). c Since native BSX was not available, the molecular mass given was calculated from the observed molecular mass of native BCX (20 394 Da) and the expected mass reduction (14 Da) upon replacing a threonine at position 147 with a serine. 83 a) 210 220 230 240 250 260 Wavelength (nm) b) 210 220 230 240 250 260 Wavelength (nm) Figure 3-1. The CD spectra of native xylanase and the Glul72 mutants, (a) A comparison of the CD spectra of native xylanase ( ), Glul72Asp (— —), and the Glul72Gln (- - -) mutant, (b) The CD spectra of Glul72Cys (- - -) and IAA-Glul72Cys ( ) versus native xylanase ( ). 84 3.3.2 PREPARATION AND CHARACTERIZATION OF THE IAA-GLU172CYS MUTANT 3.3.2.1 Iodoacetate Labelling of the Glul72Cys Mutant under Fully Denaturing Conditions Previous attempts to selectively carboxymethylate the nucleophile mutant, Glu78Cys, under nondenaturing conditions were unsuccessful, suggesting that Cys78 is relatively inaccessible to iodoacetate (refer to section 2.3.3.1, Chapter 2). Similar problems were anticipated with the Glul72Cys mutant since analysis of the modelled X-ray structures2 revealed that Cys78 and Cys 172 are solvent exposed to approximately the same degree. The calculated exposed surface area of the sulfhydryl group of Cys78 and Cysl72 is 2.6 and 0.2 A , respectively (a fully exposed sulfhydryl group has a surface area of >40 A 2 ) . Consequently, the iodoacetate labelling of Glul72Cys was attempted only under fully denaturing conditions using the procedure developed for the nucleophile cysteine mutant, as described in Materials and Methods (section 5.3.2.4, Chapter 5). The Glul72Cys mutant was incubated with IAA (22 fold molar excess) in the presence of 7.1 M urea (pH 7.5, 40 °C, 24 hrs). To refold the labelled enzyme, the reaction mixture was first diluted 20 fold with a 7.1 M urea solution (pH 7.5) and then dialyzed. The above labelling conditions were not only selective for Cys 172, but resulted in essentially complete carboxymethylation of this residue. ESMS analysis of the refolded labelled mutant (denoted IAA-Glul72Cys) showed >95% conversion of Glul72Cys to a single species having a molecular mass of 20 426 (±6) Da (refer to Appendix D, Figure D-4). 2 The molecular modelling of the Glu78Cys and Glul72Cys mutants was done using the published X-ray crystal structure of native xylanase (Campbell et al., 1993; Wakarchuk et al., 1994). 85 Since the molecular mass of untreated Glul72Cys was 20 368 (± 5) Da, this observed mass increase of 58 Da is consistent with the carboxymethylation of a single amino acid (the mass of a carboxymethyl group, -CH2COOH, is 59 Da). Evidence that Cys 172 is the amino acid being alkylated was obtained by both ESMS analysis and thiol titrations. Native xylanase exposed to the same labelling conditions, and possessing no cysteine at position 172, showed no sign of modification by ESMS (refer to Appendix D, Figure D-4). Thiol titrations revealed that IAA treatment results in the blocking of Cys 172, with the IAA-Glu 172Cys mutant having only 0.07 free cysteine. Iodoacetate labelling did not impede correct refolding, as indicated by Figure 3-lb, for the CD spectra of IAA-Glul72Cys and native xylanase are virtually identical. 3.3.2.2 Identification of the Iodoacetate Labelled Peptide in Digested IAA-Glul 72Cys To provide additional proof that Cys 172 was indeed the site of carboxymethylation in IAA-Glu 172Cys, the modified mutant and native xylanase were digested and the iodoacetate labelled peptide identified using ESMS (refer to section 5.3.8, Chapter 5 for the complete experimental details). The digestion was accomplished using cyanogen bromide (CNBr, 3.1), a reagent which cleaves predominantly at methionines (3.2), converting these residues into a mixture of C-terminal homoserine lactone (3.3) and homoserine (3.4) residues (Ambler, 1965; refer to Scheme 3-1). Since native xylanase possesses only two methionines, located at positions 158 and 169, treatment with CNBr should yield three peptides (peptide 1-158, 159-169, and 170-185). Similar treatment of the IAA-Glul72Cys mutant should produce the identical three peptides, the only difference being that the IAA-labelled peptide should have a greater molecular mass than the corresponding peptide in native xylanase. 86 H R O H R O R O I I II I I II | || . N - C H - C — . N - C H - C — —NH-CH-C-NH-CH-C— —NH-CH-CfO* — N H - C H - C / + / "I / / No / \ C H 2 0 H 2 C L H 2 C V 0 I ^ + Br" ^ \ / [3.2] C H 2 Br C H 2 ^ C H 2 Ls-^y M i l H 3 C - S f - C N + I' t 3 ' 1 ! + CH 3 SCN C H 3 N H 2 0, H + 0 II . 0 —NH-CH-C-OH H 2 0, H + —NH-CH-C^ R 0 I - / \ + I II CH 2 -CH 2 OH H 2 C X 0 + H 3N-CH-C-[3.4] C H 2 [3.3] Scheme 3-1. The cleavage at methionine residues with CNBr (3.1) under acidic conditions. The peptide mixtures resulting from the CNBr digestion were separated by reverse-phase HPLC using the ESMS as a detector (the results are shown in Figure 3-2). As expected, the HPLC traces of digested native xylanase and IAA-Glul72Cys were very similar, with each chromatogram containing three peaks corresponding to the three predicted peptides (Figure 3-2a and b). The singly charged parent peptide corresponding to HPLC peak 1 had a molecular mass of 1673 and 1705 Da in the native and IAA-Glul72Cys digests, respectively (Figure 3-2c and d). These masses are identical to those predicted for peptide 170-185 in native xylanase and IAA-Glul72Cys, with the observed 32 Da increase being as expected for the substitution of Glu 172 with a carboxymethylated cysteine. The other peaks shown in Figure 3-2c and d result from the fragmentation of the singly or doubly charged parent peptide. For both enzymes, HPLC peaks 2 and 3 correspond to peptides 159-169 and 87 1-158, respectively. Both peptides exhibited the same mass and fragmentation pattern in the native and IAA-Glul72Cys digests. These results indicate that IAA-Glul72Cys contains only one carboxymethylated amino acid, this residue being located in a peptide whose mass (m/z 1705, z = +1; m/z 853, z = +2) is identical to that predicted for peptide 170-185. The IAA-labelled peptide (m/z 853, z = +2), in the IAA-Glu 172Cys digest, was subsequently sequenced via an MS/MS experiment. The results, illustrated in Figure 3-3, are consistent with the predicted fragmentation pattern of peptide 170-185 having the sequence A T X G Y Q S S G S S N V T V W (where X corresponds to a carboxymethylated cysteine; refer to Abbreviation and Symbols for the full names of the amino acids). The peak at m/z 853 is due to the doubly charged, IAA-labelled peptide. The "b" series of peaks (bl5, bl4, bl3, bl2, b7,. b6, and b5) arise from the loss of neutral fragments from the C-terminus of the singly charged, labelled peptide (m/z 1705). The loss of the C-terminal tryptophan (W, m/z 204) from the labelled peptide (z = +1) yields the peak at m/z 1501 (bl5). The other peaks (bl4, bl3, bl2, b7, b6, and b5) result from the respective losses of V W , TVW, V T V W , SGSSNVTVW, SSGSSNVTVW, and QSSGSSNVTVW fragments from the C-terminus. The peak at m/z 741 (a7) is attributed to the additional loss of CO (28 Da) from b7, while the peak at m/z 1384 (M4-OH) could potentially result from the loss of OH from the C-terminal threonine (T) of bl4. CO c 0 c CD > DC 100 75 50 25 0 a i 12 I 16 20 Time (min) CO c CD CD > jo CD CE Time (min) CO c CD CD > JO CD DC 100 75 50 25 0 837 1 800 1673 I 1200 1600 m/z 2000 2400 c CD CD > re CD DC 100 n 75 50 25 .j 0 853 LbJL 1705 • •LJlJI, J , 800 1200 1600 2000 2400 m/z Figure 3-2. ESMS analysis of CNBr digested IAA-Glul72Cys and native xylanase. (a) The HPLC trace of digested native xylanase, (b) the HPLC trace of digested IAA-Glul72Cys, (c) the mass spectrum of HPLC peak 1 in panel a (native xylanase), (d) the mass spectrum of HPLC peak 1 in panel b (IAA-Glul72Cys). 89 100 75 • £ 50 25 0 M4-OH (+2) A-T-X-G-Y b5: 554 Q b6: 682 ^ b7: 769 <r S-G-S-S-N b12: 1 2 0 1 ^ -V b13:1301 < — b14: 1 4 0 2 ^ -fV-W b15: 1501 <-b14-OH b12 b13 ».n..J iiili.i-i^ -LLi.Mji b15 M4 JUL 1200 1400 Figure 3-3. The MS/MS spectrum of the IAA-labelled peptide in CNBr digested IAA-Glul72Cys (m/z 853 in the doubly charged state). 90 3.3.3 KINETIC EVALUATION OF NATIVE XYLANASE AND THE GLU172 MUTANTS 3.3.3.1 Steady State Kinetic Studies using the Natural Substrate Xylan Replacement of Glu 172, the putative acid/base catalyst, with a cysteine or a glutamine had a dramatic effect on activity. The Glul72Cys and Glul72Gln mutants were completely inactive on the natural substrate xylan (Wakarchuk et al., 1994). Substantial reactivation of the Glul72Cys mutant was observed upon IAA treatment, as illustrated in Figure 3-4. For IAA-Glul72Cys, native xylanase, and the Glul72Asp mutant, the kinetic parameters k c a t (app), K m (app), and k c a t / K m (app) were determined for xylan hydrolysis as described in section 5.3.4, Chapter 5 (the results are summarized in Table 3-2 and the data presented as Lineweaver-Burk plots in Appendix C). The highest substrate concentration tested with the IAA-Glul72Cys mutant was below the listed K m value. Consequently, the k c a t / K m value for this mutant was more accurately determined from the slope of the Lineweaver-Burk plot (refer to Appendix A for a brief discussion on the interpretation of k c a t , K m , and k c a t / K m ) . Altering the length of the carboxyl side chain at position 172 had a significant effect on xylan hydrolysis. The k c a t values for IAA-Glul72Cys and Glul72Asp were reduced 25 and 400 fold, respectively, relative to native xylanase. While these are substantial reductions, they are much less than that caused by complete removal of the carboxyl side chain. The side chain of IAA-Glul72Cys would appear to be better positioned to assist in xylan hydrolysis than the side chain of Glul72Asp. This presumably reflects the greater flexibility of the longer side chain, though clearly a price is paid for this in decreased binding as shown by the. 91 elevated K m value. As a result of this increased K m value, IAA-Glu 172Cys and Glul72Asp have approximately the same k c a t / K m value, this value being reduced -1000 fold with respect to the native enzyme. 800 <D CO CO CD 0) i -i— CD (0 ID E SU{ nzy CD CD C TO on E "O <D 1_ o E 600 400 200 i 1 1 1 1 1 r O O _ o — 1 1—0 1 §-50 100 150 200 Time (min) Figure 3-4. The hydrolysis of soluble birchwood xylan by Glul72Asp (•), Glul72Cys (•), and refolded IAA-treated Glul72Cys (1.1 mM (O) or 0 mM (V) IAA). For each enzyme, the concentration of xylan used was 10 mg/mL. Table 3-2. Kinetic Parameters for Soluble Xylan with Native Xylanase, IAA-Glul72Cys, and the Glul72Asp Mutant Enzyme K m (app)a (mg/mL) kcat(app)a (min1) kcat/Km (app)a (min1 mg'1 mL) Native 1.4 ±0 .2 6700 ±260 4800 ± 870 IAA-Glul72Cys 29 ± 8 270 ± 50 6.5 ± 0.4b Glul72Aspc 4.4 ±0.5 17 ± 1 3.9 ±0.7 a The (app) designations are used since the actual value observed depends upon the source and batch of xylan used, as well as the extent of reaction followed. These parameters were kept consistent in all studies described here. b The k c a t / K m value was determined from the slope of the Lineweaver-Burk plot. c Data taken from Wakarchuk et al. (1994). 92 3.3.3.2 Steady State Kinetic Studies using Synthetic j3 -Xylobioside Substrates The kinetic parameters k c a t , K m , and k c a t / K m were determined for a series of aryl (3-xylobiosides with the Glul72 mutants (the results are summarized in Table 3-3 and 3-4 and the data presented as Lineweaver-Burk plots in Appendix C). The values of k c a t and k c a t / K m determined for IAA-Glul72Cys (refer to Table 3-4) are based upon the assumption of 100% labelling, consistent with the ESMS and thiol titration results presented in section 3.3.2.1. Although Glul72Cys and Glul72Gln were completely inactive on xylan, these mutants were able to hydrolyze aryl xylobiosides with relatively good leaving groups which require little acid catalysis. Relative to native xylanase, the k c a t / K m values, which reflect the first irreversible step (most likely glycosylation), were decreased only 8 to 25 fold for substrates of aglycone pKa<5.5. Greater differences were observed for substrates of higher aglycone pK a , with no hydrolysis being detected for phenyl xylobioside (pK a = 9.99) with the Glul72Cys mutant. The inability of Glul72Cys and Glul72Gln to cleave substrates which require acid catalysis, such as xylan and PhX2, is as expected if Glu 172 functions as the acid catalyst. For both mutants, the K m values were depressed with respect to the native enzyme, indicating that the mutations have not adversely affected substrate binding. Modifying the length of the putative acid/base catalyst was less detrimental to the hydrolysis of aryl xylobiosides than to the hydrolysis of xylan. For these synthetic substrates, the k c a t / K m values for IAA-Glul72Cys and Glu 172Asp were reduced only 3 to 24 fold relative to native xylanase (refer to Table 3-4). The corresponding k c a t / K m values were decreased 1000 fold for xylan, a substrate requiring acid catalysis (see Table 3-2). The 93 precise placement of the carboxyl side chain at position 172 appears not to be critical for the hydrolysis of these aryl xylobiosides, with both mutants generally exhibiting the same k c a t / K m and k c a t values. For IAA-Glul72Cys and Glul72Asp, the k c a t / K m values determined with the substrates PNPX2 and P h X 2 were increased greater than 10 fold relative to the Glul72Cys mutant. Thus for synthetic substrates requiring protonic assistance, a mispositioned acid catalyst is better than none at all. Lengthening and shortening the side chain at position 172 did not significantly perturb substrate binding as evidenced by the similar K m values for IAA-Glu 172Cys, Glul72Asp, and native xylanase. 94 Table 3-3. Kinetic Parameters for Aryl p-Xylobiosides with Glul72Cys, Glul72Gln, and Native Xylanase Substrate Kinetic Parameter Native Glul72Gln Glul72Cys 2,5-DNPX2a ( p K a = 5 . 1 5 ) b kcat ( s ') 76 ± 1.5d 0.61 ± 0 . 0 2 1.5 ± 0 . 0 4 K m ( m M ) 2 . 2 ± 0 . 1 d 0.45 ± 0.05 0.29 ± 0.03 3 5 ± l . l d 1.5 + 0.1 4.2 ± 0 . 2 3,4-DNPX2a (pK a = 5.36) b kcat ( s ') 8 . 3 ± 0 . 4d e 0.18 ± 0 . 0 1 K m (mM) 3.4 + 0 . 3 d e 1.2 + 0.1 k A f s - ' m M 4 ) 1 2 . 7 ± 0 . 1 d e 0.17 ± 0 . 0 0 3 ONPX 2 a (pK a = 7.22) b kCat ( s ) 9.6 ± 0 . 1 0.62 ± 0.03 0.40 ± 0.004 K m ( m M ) 14 ± 0 . 5 8.3 ± 0 . 7 2.5 ± 0 . 1 0.66 ± 0 . 0 1 0.080 ± 0 . 0 0 1 0.16 ± 0 . 0 0 1 PNPX 2 a (pK a = 7.18) b k C at ( s ') 24 ± 2 . 1 e 0.018 ± 0 . 0 0 1 Km (mM) 49 ± 6 . 4 e 8.6 ± 1.2 kcat/KmCs-'mM"1)0 0.43 ± 0 . 0 1 e 0.002 ± 0.0001 PhX 2 a ( P K a = 9.99) b kcat ( s ') 0.051 ± 0 . 0 0 4 e f K m ( m M ) 8.7 ± 1 . 3 e f 0.005 ± 0.0003 e f a For a listing of the full substrate's name see Abbreviations and Symbols. bThe pK a value given is for the aglycone. c All the kc a t /K m values were determined from the slope of the Lineweaver-Burk plot. d Data taken from Ziser et al. (1995). eNot determined. f Enzyme-catalyzed hydrolysis was not detected. 95 Table 3-4. Kinetic Parameters for Aryl (3-Xylobiosides with IAA-Glul72Cys, Glul72Asp, and Native Xylanase Substrate Kinetic Parameter Native IAA-Glul72Cys Glul72Asp 2,5-DNPX2a (pK a=5.15) b k C a t ( s ' ) 76+ 1.5d 3.2 ±0 .1 4.1 ±0.4 K m ( m M ) 2 . 2 ± 0 . 1d 2.0 ±0.1 2.1 ±0.5 k ^ / K ^ s " 1 mM" 1) 0 35 ± l . l d 1.8 ±0.02 2.1 ±0.04 3,4-DNPX2a (pK a = 5.36)b k C a t ( s ' ) 8.3±0.4d 1.4 ±0.02 1.0 ±0.02 K m ( m M ) 3.4 ± 0.3d 3.6 ±0 .2 1.1 ±0 .1 kcat /KmCs- 'mM-'r 2 . 7 ± 0 . 1 d 0.46 ±0.01 0.90 ± 0.01 ONPX 2 a (pK a = 7.22)b kca t ( s ' ) 9.6 + 0.1 1.0 ±0.05 0.25 ± 0.005 K m ( m M ) 14 ±0 .5 36 ±3 .4 7.2 ±0 .4 kca . / K m ( s ' ' mM" 1) 0 0.66 ± 0.01 0.028 ±0.001 0.032 ± 0.001 PNPX 2 a (pK a = 7.18)b k c a t ( s ') 24 ±2.1 0.41 ±0.03 0.16±0.01 K m ( m M ) 49 ± 6.4 13 ± 1.6 6.4 ±0.8 0.43 ± 0.01 0.032 ± 0.0003 0.026 ± 0.0004 PhX 2 a (pK a = 9.99)b kca t ( s ' ) 0.051 ±0.004 0.017 ±0.001 0.002 ±0.0001 K m ( m M ) 8.7 ± 1.3 28 ± 2.3 5.8 ±0.6 k A i s - ' m M " 1 ) 1 0.005 ± 0.0003 0.0006 ±0.00001 0.0003 ± 0.00001 a For a listing of the full substrate's name see Abbreviations and Symbols. bThe pK a value given is for the aglycone. 0 Al l the k c a t /K m values were determined from the slope of the Lineweaver-Burk plot. d Data taken from Ziser et al. (1995). 96 3.3.3.3 The Effects of Exogenous Nucleophiles on Reaction Rates and Products The enzymatic cleavage of two synthetic substrates, ONPX2 and 2,5-DNPX2, by native xylanase and the Glu 172 mutants was performed in the presence of different concentrations of sodium azide (0-500 mM) (refer to section 5.3.3.4, Chapter 5 for the experimental details). As shown in Figure 3-5, the absence of an acid/base catalyst can, in part, be compensated for by the addition of the anionic nucleophile, azide, to the reaction buffer when activated substrates are studied. For the Glul72Gln and Glul72Cys mutants, the presence of sodium azide resulted in enhanced reaction rates (up to 8 fold) for 2,5-DNPX2 (refer to Figure 3-5a). No such rate increases were seen with native xylanase, Glul72Asp, or the IAA-Glu 172Cys mutant. Thin layer chromatographic analysis of the Glul72Gln and Glul72Cys reaction mixtures containing sodium azide revealed the formation of a new product (R f = 0.39), distinct from xylobiose (Rf = 0.08) and 2,5-DNPX 2 (Rf = 0.51), consistent with the formation of (3-xylobiosyl azide (structure 3.5, Scheme 3-2). Xylobiose and 2,5-dinitrophenol were the only reaction products of native xylanase, Glul72Asp, and IAA-Glu 172Cys reaction mixtures containing sodium azide. Analogous studies conducted on C. fimi exoglycanase (MacLeod et al., 1994), Agrobacterium P-glucosidase (Wang et al., 1995), and E. coli |3-galactosidase (Huber & Chivers, 1993) yielded similar results. For these enzymes, the rate enhancements were shown to be due to azide reacting with the glycosyl-enzyme intermediate more rapidly than does water, thus resulting in increased rates when deglycosylation is the rate-determining step (refer to Scheme 3-2 for the proposed reaction mechanism with xylanase). 97 a) > > b) > > 8 6 -4 -0 100 200 300 400 500 [NaN3] (mM) I 1 I 1 1 1 1 o o o o -0 -a I I H 1 1 1 1 1 1 100 200 300 400 [NaN3] (mM) 500 Figure 3-5. The effects of azide on xylanase-catalyzed hydrolysis rates for (a) 2,5-DNPX2 and (b) ONPX2. The enzymes examined were native xylanase (V), Glul72Asp (•), Glul72Gln (O), Glul72Cys (•), and refolded IAA-treated Glul72Cys (1.1 mM (A) or 0 mM (0) IAA). The symbols v and v0 correspond to the hydrolysis rates in the presence and absence of azide, respectively. 98 HO-HO- -O -O HO" HO HO [3.5] H O H O \ HO HOJ—^9 J Glycosylation HOJ . . . L . . . Glu78 ^ // V . . . L . . . Glu78 H,0-Deglycosylation HO HO O HO" HO . . . L . . . Glu78 HO OH Scheme 3-2. The proposed mechanism for xylanase-catalyzed hydrolysis of aryl /?-xylobiosides in the presence of sodium azide (X corresponds to a H or one or more NO 2 groups). Similar, but smaller, rate increases were seen for ONPX2 hydrolysis by Glul72Gln, but not for Glul72Cys (see Figure 3-5b). Analysis of the reaction mixtures for both mutants revealed the formation of a new product, presumably xylobiosyl azide. Again, no rate increases were seen with native enzyme, IAA-Glul72Cys, or Glul72Asp upon addition of azide, nor was any sugar product other than xylobiose formed. The above results provide information concerning the rate-determining step for the Glul72Cys and Glul72Gln mutants with the substrates 2,5-DNPX 2 and ONPX 2 . Deglycosylation appears to be the rate-limiting step for the Gin mutant with both substrates since rate enhancements were observed for both upon the addition of azide. For the Cys mutant, the azide data shown in Figure 3-5 suggest that the rate-determining step for 2,5-99 D N P X 2 and O N P X 2 are deglycosylation and glycosylation, respectively. The k c a t values for these two substrates with the Cys and Gin mutants are consistent with these assignments. For Glul72Gln, both substrates have the same k c a t value (0.6 s 1 , Table 3-3), as one would expect if deglycosylation is rate limiting. In contrast, these k c a t values differ by a factor of 3.7 for Glul72Cys, thus suggesting that the rate-determining step for this mutant is different for the two substrates. The absence of any azide-induced rate increase with native xylanase, Glul72Asp, and IAA-Glu 172Cys is presumably a consequence of charge screening by the anionic carboxyl group at position 172, inhibiting the access of azide. The existence of this charge screening in the three enzymes was confirmed by the appearance of only xylobiose as the reaction product and not xylobiosyl azide. The fact that azide effects are seen for Glul72Cys, with the substrate 2,5-DNPX 2 , strongly suggests that the cysteine side chain is not deprotonated under the assay conditions (pH 6). Further, the absence of azide effects with IAA-treated Glul72Cys provides additional evidence of the successful carboxymethylation of the cysteine residue, thus reinserting a carboxyl side chain at position 172. The hydrolysis of 2,5-DNPX 2 by native xylanase was also conducted in the presence of MeOH, DTT (1,4-dithiothreitol), and P-mercaptoethanol (for the experimental details see section 5.3.3.4, Chapter 5). Unlike sodium azide, these neutral nucleophiles should not be electrostatically repelled by the anionic carboxyl side chain at position 172. Thus, the nucleophile's attack on the glycosyl-enzyme intermediate should proceed without interference. As illustrated in Figure 3-6, the addition of these neutral nucleophiles to the 100 reaction mixture did not result in enhanced rates.3 Indeed, for DTT and (3-mercaptoethanol an inhibitory effect was observed at the higher nucleophile concentrations, presumably due to denaturation of the enzyme. These results suggest that glycosylation, rather than deglycosylation, is the rate-determining step for native xylanase with this substrate. 0 200 400 600 800 1000 1200 [nucleophile] (mM) Figure 3-6. The effects of neutral nucleophiles on rates of native xylanase-catalyzed hydrolysis of 2,5-DNPX2- The nucleophiles tested were R-mercaptoethanol (•), DTT (O), and MeOH (V). The symbols v and v 0 correspond to the hydrolysis rates in the presence and absence of the nucleophile, respectively. The enzyme-catalyzed rates were corrected for the cleavage of the substrate by the nucleophile alone. 3.3.3.4 The Br0nsted Relationships for Native Xylanase, IAA-Glul72Cys, and Glul72Asp No information regarding the identity of the rate-determining step for native xylanase, Glul72Asp, or IAA-Glu 172Cys could be gleaned from the azide data. For this it was When D T T and p-mercaptoethanol were tested at concentrations in excess of those shown in Figure 3-6, precipitate was observed in the assay cell. Due to this interference, the rates obtained at these higher nucleophile concentrations were unreliable and are not included in Figure 3-6. necessary to examine the Br0nsted relationships shown in Figure 3-7 (refer to Appendix B for a general introduction to Br0nsted relationships). The calculated slopes ((3ig) and correlation coefficients (p) for Figure 3-7, the log k c a t versus pK a plot, are summarized in Table 3-5. Figure 3-7. The log kcat versus pKa plot for native xylanase (O), Glul72Asp (•), and IAA-Glul72Cys (A). The data was taken from Table 3-4. Table 3-5. The Determined Slopes and Correlation Coefficients for Figure 3-7. Enzyme Slope (Plg) Correlation Coefficient (p) Native xylanase -0.5 (±0.2) 0.88 IAA-Glul72Cys -0.4 (±0.1) 0.96 Glul72Asp -0.7 (±0.1) 0.98 As illustrated in Figure 3-7, a linear relationship exists between log k c a t and aglycone pK a for the Glul72Asp and IAA-Glu 172Cys mutants. Since the aglycone only participates in 102 the glycosylation step of the double displacement (retaining) mechanism (see Scheme 1-3, Chapter 1), this correlation implies that glycosylation is the rate-determining step for these two mutant enzymes. While the data obtained for native xylanase are more scattered, the same trend is observed suggesting that glycosylation is also rate limiting for the native enzyme. Further evidence in support of glycosylation being the rate-determining step for native xylanase was obtained from two separate studies. The hydrolysis of 2,5-DNPX2 by native xylanase was examined using stopped flow techniques (refer to section 5.3.3.2, Chapter 5 for the experimental details). No pre-steady state burst of release of 2,5-dinitrophenolate was observed for native xylanase with either 0.5 or 1.9 mM 2,5-DNPX 2 . This result indicates that the glycosyl-enzyme intermediate is not accumulating and, thereby, suggests that glycosylation is the rate-determining step for this substrate. Additional evidence was obtained from the exogenous nucleophile studies presented in section 3.3.3.3. The addition of neutral nucleophiles to a reaction mixture containing native enzyme and 2,5-DNPX 2 did not result in the rate enhancements that would be expected if deglycosylation were rate limiting. These observations are again consistent with glycosylation being the rate-determining step for native xylanase with 2,5-DNPX 2 . Presented in Figure 3-8 are the Br0nsted relationships between log k c a t / K m and aglycone p K a for native xylanase, IAA-Glu 172Cys, and Glul72Asp. Good correlations of log k c a t / K m with aglycone leaving group ability are seen for these three enzymes with a range of aryl xylobioside substrates. Since k c a t / K m reflects the first irreversible step, these results 103 confirm that this step is indeed glycosylation. As shown in Table 3-6, very similar slopes ( P l g values) were obtained for the three enzymes. Thus, the placement of the acid catalyst does not appear to seriously affect the degree of proton donation at the glycosylation transition state. 4 5 6 7 8 9 10 11 pKa Figure 3-8. The log kcat/Km versus pKa plot for native xylanase (O), Glul72Asp (•), and IAA-Glul72Cys (A). The data was taken from Table 3-4. Table 3-6. The Determined Slopes and Correlation Coefficients for Figure 3-8. Enzyme Slope (Pig) Correlation Coefficient (p) Native xylanase -0.7 (±0.1) 0.97 IAA-Glul72Cys -0.7 (±0.1) 0.99 Glul72Asp -0.8 (±0.1) 0.99 104 3.3.3.5 The pH Dependence of kcat/Km The k c a t / K m values for the hydrolysis of O N P X 2 by native xylanase and the Glu 172 mutants were determined over a range of pH values as described in Chapter 5, section 5.3.3.5. The pH studies were performed at 25 °C since the Glul72 mutants are relatively unstable at 40 °C for extended periods of time. Even at this reduced temperature, activity loss was observed for Glul72Asp at the more acidic pH values, thus only k c a t / K m values for pH 4.5 to 9 were determined for this mutant. The resulting plots of k c a t / K m versus pH for native xylanase, Glul72Asp, and IAA-Glul72Cys are bell-shaped (see Figure 3-9). The shape of these pH profiles suggests that activity is dependent upon two ionizable catalytic residues. As previously discussed in Chapter 2 (section 2.3.4.5), the ionization in the acidic limb (pK a i) is most likely associated with the active site nucleophile (Glu78), while the ionization in the basic limb (pK a 2 ) is likely due to the carboxyl group at position 172 (the acid catalyst position). Analysis of the k c a t / K m versus pH data presented in Figure 3-9 (as described in section 5.3.3.5, Chapter 5) yielded values for pK a j and p K a 2 (refer to Table 3-7). These p K a values are attributed to the ionizable catalytic groups in the free enzyme (Fersht, 1985). For native xylanase, p K a values of 4.6 (pK a i) and 6.8 (pK a 2 ) were determined. 1 3 C N M R studies done by Mcintosh et al. (1996) confirmed that these p K a values were due to Glu78 (pK a i) and Glu 172 (pK a 2 ) , respectively. Upon substitution of Glu 172 by Asp, very little change is observed i n p K a i , that of Glu78, as might be expected since this replacement results in no 105 3 4 5 6 7 8 9 PH b) 3 4 5 6 7 8 9 PH Figure 3-9. The pH dependence of kcat/Km for (a) native xylanase (•) and Glul72Asp (A) and (b) Glul72Cys (•) and IAA-Glul72Cys (O). The lines shown for native, Glul72Asp, and IAA-Glul72Cys represent fits to data for enzymes with two ionizable groups. For Glul72Cys, the line shown represents a fit to data for an enzyme with a single ionizable group. For figure a (and b), the scales on the right and left y-axes are for Glul72Asp (IAA-Glul72Cys) and native xylanase (Glul72Cys), respectively. 106 Table 3-7. The Determined pKa Values for Native Xylanase and the Glul72 Mutants Enzyme PKai P K 3 2 Native xylanase 4.6 (±0.04) 6.8 (±0.03) Glul72Asp 4.2 (±0.13) 8.0 (±0.06) IAA-Glul72Cys 3.2 (±0.08) 6.7 (±0.07) Glul72Cys 4.0 (±0.23) significant change in local charge distribution. However, an increase of just over one unit in p K a 2 , that of Glu/Aspl72, is observed. Since electrostatic effects should diminish upon pulling this carboxyl group further away from the anionic Glu78, this change is opposite to that expected. The elevated p K a 2 value could result from either the loss of some hydrogen bonding interactions which ordinarily lower the p K a of Glu 172, presumably by stabilizing the carboxylate anion, or a closer positioning of Asp 172 to some other anionic or hydrophobic group. The X-ray crystal structure of native xylanase, shown in Figure 3-10, reveals hydrogen bonds between Glu 172 and two active site residues, Asn35 and Tyr80 (Campbell et al., 1993; Wakarchuk et al., 1994). Shortening Glul72 to an aspartic acid likely weakens or completely destroys these hydrogen bonds, thereby causing an elevation in p K a 2 . Lengthening the acid/base catalyst (IAA-Glul72Cys) had a reciprocal effect on the active site p K a values. The p K a of Glu78 (pK a0 is depressed just over one pH unit while that of Glu/IAA-Cysl72 stays approximately constant. The intrinsic p K a value of carboxymethylated cysteine is expected to be about one pH unit lower than that of glutamic 107 acid, based upon the p K a values for the analogous carboxylic acids CH3CH2CO2H (pK a = 4.88) and C H 3 S C H 2 C 0 2 H (pK a = 3.72) (Brown et al., 1955). Thus, in the absence of other effects a decrease of p K a 2 would have been expected. The absence of any change in p K a 2 suggests the presence of a countering effect which has increased p K a 2 by approximately one unit. Perhaps the simplest explanation for a decrease in p K a l , and an effective increase in p K a 2 , is that lengthening the side chain at position 172 has now allowed the two catalytic carboxyl groups to hydrogen bond to each other, most likely via an intervening water molecule. This would effectively lower pK a j , since the monoanion will be relatively stabilized, but raise p K a 2 . Molecular modelling of IAA-Glu 172Cys4 reveals that, in its fully extended conformation, the side chain carboxyl group of IAA-Cysl72 comes within 3.8 A of Glu78, a distance too great for direct hydrogen bonding. However, the intervention of a water molecule is possible, a candidate being that seen hydrogen bonded to Glu78 in the structure shown in Figure 3-10. Extension of the side chain at position 172 could allow the acid/base catalyst carboxyl group to also hydrogen bond to this water molecule, thereby completing the hydrogen bonding network. An alternative explanation for the p K a results is that lengthening the acid/base catalyst alters the conformation of the active site, placing Glu78 in an environment more capable of stabilizing a carboxylate anion while positioning IAA-Cysl72 in less favorable surroundings. Replacement of Glu 172 with a cysteine results in the abolition of the basic limb of the pH profile, as expected, along with a slight decrease in the p K a of Glu78. This result is 4 The molecular modelling was done by Manish Joshi as described in Materials and Methods (section 5.3.9, Chapter 5). completely consistent with the observation that Cys 172 provides no acid catalysis, as evidenced by the similar k c a t / K m values obtained for ONPX2 with Glul72Cys and Glul72Gln, a mutant incapable of providing acid catalysis (the values, shown in Table 3-3, are 0.16 and 0.08 s 1 m M 1 , respectively). It also agrees with the conclusion that Cys 172 remains protonated under the assay conditions (pH 6), as revealed by the azide effects observed with the substrate 2,5-DNPX 2 (discussed in section 3.3.3.3). Figure 3-10. The active site of Bacillus circulans xylanase, showing the key residues and hydrogen bonding interactions. The nucleophile (Glu78) and acid/base catalyst (Glul72) are located on opposite sides of the active site cleft. Hydrogen bonds are shown as dashed lines while the spheres represent water molecules. 109 3.3.4 THE INACTIVATION OF GLU172CYS BY DI-p,p-D-XYLOPYRANOSYL DISULFIDE The idea to test di-P,P'-D-xylopyranosyl disulfide (denoted X S S X , structure 3.6) as an inactivator of the Glul72Cys mutant developed from earlier work done on Glu78Cys, the active site nucleophile cysteine mutant. The goal of this original work was to covalently attach an a-thioxylosyl moiety to Cys78. It was postulated that the glycosylated Cys78 residue, bearing a resemblance to the normal covalent glycosyl-enzyme intermediate in native xylanase (refer to Scheme 3-3), could potentially be hydrolyzed via the normal enzymatic mechanism. One strategy to achieve the desired modification of Cys78 was to react this residue with the disulfide of 1-thio-a-D-xylose (structure 3.7). Unfortunately, there was no published synthetic route for this particular compound. However, a procedure for the P~ analogue (XSSX, structure 3.6) did exist (Stanek et al., 1965). Thus, it was decided to attempt the labelling of Cys78 using X S S X . ESMS analysis of Glu78Cys, treated with a 500 fold molar excess of X S S X (pH 10, 40 °C, 24 hrs), showed no signs of labelling (see Appendix D, Figure D-5). This result suggested that the a-linked analogue of X S S X is required for successful modification of Cys78 and prompted the testing of the P-linked disulfide with the Glul72Cys mutant. HO HO S — S [3.6] [3.7] 110 HO| H i H CH 2 Cys78 Glu78 Scheme 3-3. A comparison between Cys78, covalently modified with an a-thioxylosyl 3.3.4.1 The Effects of pH on the Inactivation Rate The Glul72Cys mutant was incubated with X S S X (500 fold molar excess) at pH 8, 9, and 10 (40 °C, 27 hrs) and the activity monitored (refer to section 5.3.6.1, Chapter 5 for the complete experimental details). As illustrated in Figure 3-11, essentially complete inactivation of Glul72Cys occurred in 27 hours at pH 9 and 10. In contrast, Glul72Cys incubated at pH 10 in the absence of X S S X suffered only a 20% activity loss over the same time period. Thus, the inactivation observed at pH 10 in the presence of X S S X can not be attributed solely to enzyme instability and is presumably due to derivatization. No inactivation was observed at pH 8 over the 27 hour period. 3.3.4.2 Determination of the Inactivation Parameters ki and Kj The inactivation of Glul72Cys by XSSX, at pH 10, was examined more closely by determining the inactivation rates at various concentrations of X S S X (refer to section 5.3.6.2, Chapter 5 for the experimental details). The results, illustrated in Figure 3-12a, clearly show that the rate of inactivation increased as the disulfide concentration was raised. Presented in Figure 3-12b and c are plots of the first-order rate constants (k0t,s) obtained from Figure 3-moiety, and the normal covalent glycosyl-enzyme intermediate in native xylanase. I l l 12a. Analysis of the k 0b s versus X S S X concentration plot (Figure 3-12b) yielded values of 0.021 (±0.002) min"1 and 170 (±20) mM for kj and Kj, respectively (refer to Appendix A for a brief discussion on the interpretation of kj and K;). These values are very approximate since only X S S X concentrations below the estimated Kj were examined. An accurate kj /Kj value of 1.5 x 10"4 (±3 x 10"6) min" mM" was determined from the slope of the reciprocal plot shown in Figure 3-12c. Clearly, X S S X is not a very potent inactivator of Glul72Cys since both a high X S S X concentration and an extended incubation period are required to achieve complete inactivation. 0 400 800 1200 1600 Time (min) Figure 3-11. The effects of pH on the inactivation of Glul72Cys (0.05 mM) by XSSX (25 mM, 500 fold molar excess). The experiment was performed at pH 8 (•), 9 (A), and 10 (V). As a control, enzyme was incubated in the absence of XSSX at pH 10(0). 112 Figure 3-12. The inactivation of Glu 172 Cys by XSSX at pH 10. (a) The plot of % activity versus time at the following inactivator concentrations: 0 mM (A), 1 mM (•), 5 mM (O), 25 mM (T), 50 mM (•), and 97 mM (0). (b) The plot of kobs versus XSSX concentration, (c) The double reciprocal plot of (b) (the data obtained at 1 mM are not included). The first-order rate constants (kobs) were obtained from (a) and corrected for enzyme inactivation due to denaturation using data from the 0 mM control sample. 113 3.3.4.3 The Proposed Mechanism of Inactivation The presence of a |3-xylosyl moiety in X S S X (structure 3.6) should facilitate binding in the active site of the Glul72Cys mutant. The deprotonated Cys 172, if suitably positioned, could then make a nucleophilic attack on the disulfide bond of X S S X to form the mixed disulfide (structure 3.8) shown in Scheme 3-4. The covalent attachment of a thioxylosyl moiety to Cys 172 would effectively prevent the productive binding and turnover of substrate, thus resulting in enzyme inactivation. Cys172 . . . . . . . C H 2 S [3.6, fKbngC_H°^°H OH"\JT Glu78 [3-8] K?oO^O_ Cys172 . . . . . . . C H 2 i S OH . . . L . . . Glu78 HO HOLO OH Scheme 3-4. The proposed mechanism of inactivation for XSSX with the Glul 72 Cys mutant. Evidence in support of the proposed mechanism was obtained from both ESMS analysis and thiol titrations. As shown in Table 3-8, treatment with X S S X resulted in essentially complete conversion of Glul72Cys to a single species having a molecular mass of 20 529 (±5) Da (see Figure D-5, Appendix D). The observed mass increase of 164 Da, relative to untreated Glul72Cys, is consistent with the covalent attachment of one thioxylosyl moiety (165 Da). Thiol titrations indicate that the sugar label is attached to Cysl72, with inactivated Glul72Cys possessing only 0.10 (±0.04) free cysteines (refer to Table 3-8). 114 Table 3-8. Thiol Titration and ESMS Results for XSSX Treated Enzymes Sample Number of Free Cysteines/mol Enzyme Molecular Mass (g/mol) Glul72Cys 0.99 ± 0.06 20 365 ± 6 Glul72Cys + XSSX a 0.10 ±0.04 20 529 ± 5 Native N D b 20 395 ± 4 Native + XSSX a N D b 20 394 ± 3 a All enzymes were treated with a 500 fold molar excess of XSSX at pH 10 (40 °C, 24 hrs). b Not determined (ND). Analysis of X S S X treated native xylanase provided additional evidence that the observed inactivation of Glul72Cys was due to the covalent modification of Cys 172. Native xylanase exposed to the same labelling conditions as Glul72Cys showed no sign of modification by ESMS analysis (refer to Table 3-8 and Figure D-5, Appendix D), nor was there any significant activity loss relative to the control (see Figure 3-13). The inability of X S S X to inactivate native enzyme can not be attributed to disrupted reagent binding since X -ray crystallographic analysis (refer to Figure 2-2a and b, Chapter 2) indicated that native xylanase and Glul72Cys share similar, if not identical, active site conformations. Rather, the absence of a cysteine at position 172 most likely makes native xylanase immune to inactivation by X S S X . 115 Time (min) Figure 3-13. The activity loss of Glu 172 Cys (O) and native xylanase (A) upon treatment with XSSX (500 fold molar excess, pH 10, 40 °C). As a control, Glul72Cys (•) and native xylanase (•) were also incubated at pH 10 in the absence of XSSX. 3.4 CONCLUSION The importance of a carboxyl group at position 172 in B. circulans/subtilis xylanase is clearly shown by the fact that replacement of this residue by a group with no significant capacity as a proton donor/acceptor (Gin or Cys), results in an enzyme which can not cleave xylan or phenyl p-xylobioside, two substrates requiring acid catalysis. However, this carboxyl side chain is not essential for the hydrolysis of substrates with good leaving groups (pKa<5.5) needing no protonic assistance. For the Glul72Cys and Glul72Gln mutants, the addition of sodium azide resulted in significant rate increases for the cleavage of 2,5-dinitrophenyl P-xylobioside and the formation of a xylobiosyl azide product. This indicates that deglycosylation is now the rate-determining step, the azide reacting preferentially with 116 the glycosyl-enzyme intermediate. These results are consistent with Glu 172 functioning as the acid/base catalyst in B. circulans/subtilis xylanase and emphasize the functional importance of the carboxyl group found at this position for the hydrolysis of unreactive substrates. The precise placement of the acid/base catalyst in B. circulans/subtilis xylanase is not critical for the hydrolysis of aryl P-xylobiosides since either shortening or lengthening this carboxyl side chain results in approximately the same modest decrease in k c a t / K m values (3 to 24 fold): This is in sharp contrast with the positional requirements of the catalytic nucleophile Glu78. Shortening the nucleophilic side chain (Glu78Asp) decreased k c a t / K m values at least 1600 fold for the aryl p-xylobiosides, whereas increasing the length (IAA-Glu78Cys) decreased these values by only 16 to 100 fold (discussed in Chapter 2, section 2.3.4.4). Thus as expected, the positional requirements for proton transfer are less demanding than those for carbon-oxygen bond formation. CHAPTER 4 INVESTIGATION OF THE GLU358CYS MUTANT OF AGROBACTERIUM p-GLUCOSIDASE 118 4.1 I N T R O D U C T I O N The conversion of cellobiose to glucose in natural systems is accomplished by the enzyme P-glucosidase ((5-D-glucoside glucohydrolase; EC 3.2.1.21; refer to Scheme 4-1). This enzyme is found in a variety of organisms such as plants, animals, fungi, and bacteria (Shewale, 1982; Dahlqvist, 1961). . O H HH°04^-0 OH OH ^-S /TV^-O P-Glucosidase H O T V < _ 0 O H H ° ^ L 0 H - H ° - - ^ ^ O H OH H20 OH Scheme 4-1. The conversion of cellobiose to glucose by P-glucosidase. One of the first bacterial P-glucosidases to be isolated and extensively characterized was the enzyme from Agrobacterium faecalis1 (Han & Srinivasan, 1969; Day & Withers, 1986). The gene encoding Agrobacterium P-glucosidase has been cloned and expressed in E. coli. The purified recombinant P-glucosidase is identical to the natural enzyme in amino acid sequence and kinetic behavior (Wakarchuk et al., 1986; Day & Withers, 1986; Kempton & Withers, 1992). The enzyme has a monomeric molecular mass of 51 171 Da and exists as a homodimer in its active form. In addition to cellobiose, Agrobacterium p-glucosidase catalyzes the hydrolysis of various p-glucosides including aryl p-glucosides, P-glucosyl fluoride, P-glucosyl azide, and p-glucosyl pyridinium salts (Day & Withers, 1986). The 1 In the rest of this chapter, the enzyme will be referred to as Agrobacterium P-glucosidase. The bacterium species was originally classified as Alcaligenes faecalis (Han & Srinivasan, 1969). 2 The monomeric molecular mass was determined from the amino acid sequence. 119 enzyme is also capable of hydrolyzing P-galactosides, (3-xylosides, and p-mannosides, but at a reduced rate relative to the fj-glucosides. A retaining glycosidase, Agrobacterium P-glucosidase catalyzes the hydrolysis of fj-glucosides with net retention of anomeric configuration (Day & Withers, 1986). This enzyme, like B. circulans/subtilis xylanase, is believed to employ the double displacement mechanism described in Chapter 1, section 1.3 (see Scheme 1-3). As stated previously, the key catalytic residues in this mechanism are two active site carboxylates. In the glycosylation step, one carboxylate functions as the active site nucleophile, attacking the anomeric carbon of the substrate to form the covalent glycosyl-enzyme intermediate. This same residue acts as a leaving group in the subsequent hydrolysis (deglycosylation) step. The other carboxylate acts as the acid/base catalyst. The formation and hydrolysis of the covalent glycosyl-enzyme intermediate are believed to occur via oxocarbenium ion-like transition states. Evidence in support of this proposal was obtained from a-secondary deuterium kinetic isotope effects. With substrates (deuterated at the anomeric carbon) for which glycosylation is rate determining, a k H /k D value of 1.06 was obtained. Larger k H / k D values (1.10-1.12) were determined when deuterated substrates were used for which deglycosylation is rate limiting (Kempton & Withers, 1992). These positive kinetic isotope effects indicate that both the glycosylation and deglycosylation transition states have substantial sp2 character, consistent with their proposed oxocarbenium ion-like character. The active site nucleophile of Agrobacterium P-glucosidase has been identified as Glu358. The identification of this catalytic carboxylate was accomplished by trapping the 120 covalent glycosyl-enzyme intermediate using the mechanism-based inactivator, 2',4'-dinitrophenyl 2-deoxy-2-fluoro-|3-D-glucopyranoside (4 . 1 , refer to Scheme 1-8, Chapter 1) (Withers et al., 1990). Results obtained from mutagenesis studies were consistent with this assignment. Replacement of Glu358 with residues incapable of functioning as nucleophiles, such as asparagine, glutamine, glycine, and alanine, resulted in dramatic decreases in activity (at least 104 fold) (Trimbur et al., 1992; Withers et al., 1992). In contrast, substituting Glu358 with its shortened analogue, aspartic acid, caused a 2500 fold reduction in the rate constant for the glycosylation step. However, the mutant still followed the double displacement mechanism (Withers et al., 1992). These results emphasize the importance of a correctly positioned carboxylate at position 358. Glu 170 has been identified as the acid/base catalyst of Agrobacterium (3-glucosidase. This assignment was made by performing a detailed kinetic analysis on the Glul70Gly mutant (Wang et al., 1995). Replacement of Glul70 with glycine, a group incapable of acting as a proton donor/acceptor, resulted in very little change in the k c a t / K m values for substrates not requiring protonic assistance. However, the k c a t / K m values for substrates needing acid catalysis were dramatically reduced (at least 7 x 103 fold). These results provide strong evidence for the role of Glu 170 as the acid/base catalyst. [4.1] 121 While the X-ray structure of Agrobacterium p-glucosidase is still undetermined, the three-dimensional structures of two related P-glycosidases, white clover P-glucosidase (Barrett et al., 1995) and Lactococcus lactis 6-phospho-P-galactosidase (Wiesmann et al., 1995), have been solved. Inspection of these two structures reveals that the homologous residues to Glu358 and Glu 170 are suitably positioned within the enzymes' active sites to fulfill the respective roles of catalytic nucleophile and acid/base catalyst. Thus, these results provide additional support for the assignment of Glu358 as the active site nucleophile and Glu 170 as the acid/base catalyst of Agrobacterium p-glucosidase. 4.2 SPECIFIC A I M OF THIS STUDY Can a cysteine function as the active site nucleophile in the retaining P-glucosidase from Agrobacterium? Hardy and Poteete (1991) proposed that cysteine could fulfill the role of the active site nucleophile in retaining glycosidases. This proposal was based on site-directed mutagenesis work done on T4 lysozyme, a glycosidase that was assumed to act via a retaining mechanism. Nearly wild-type activity was observed upon mutating Asp20, the putative catalytic nucleophile, to a cysteine. This result suggested that cysteine could function effectively as the catalytic nucleophile. It is important to note that recent work done by Kuroki et al. (1995) revealed that T4 lysozyme is actually an inverting glycosidase, not a retaining enzyme as originally believed. Consequently, Asp20 does not act as the catalytic nucleophile, but rather as a general base catalyst. The feasibility of a cysteine functioning as the catalytic nucleophile in Agrobacterium P-glucosidase will be tested by performing a kinetic analysis on Glu358Cys, the active site nucleophile cysteine mutant. Previously, a similar investigation had been conducted on the 122 active site nucleophile cysteine mutant, Glu78Cys, of B. circulans xylanase (refer to section 2.3.2, Chapter 2 for the results). The Glu78Cys mutant was totally inactive indicating that for B. circulans xylanase a cysteine is incapable of fulfilling the role of the catalytic nucleophile. 4.3 RESULTS AND DISCUSSION 4.3.1 PREPARATION AND PRELIMINARY CHARACTERIZATION OF NATIVE f3-GLUCOSIDASE AND THE GLU358CYS MUTANT The mutagenesis and production of the Glu358Cys mutant and native P-glucosidase were kindly performed by Dr. Don Trimbur, Department of Microbiology, University of British Columbia, as previously described (Trimbur et al., 1992). Both native P-glucosidase and Glu358Cys were isolated from E. coli cell paste, provided by Dr. Trimbur, using the protocol outlined in Withers et al. (1992) (refer to section 5.4.1, Chapter 5 for the full experimental details). Native p-glucosidase and the Glu358Cys mutant were subjected to electrospray mass spectrometric (ESMS) analysis and thiol titrations. For native enzyme and Glu358Cys, the molecular masses were determined to be 51 192 (±2) Da and 51 171 (±6) Da, respectively. The observed mass reduction of 21 (±8) Da, for the Glu358Cys mutant relative to the native enzyme, is as expected for the replacement of a glutamic acid by a cysteine (the theoretical mass reduction is 26 Da). The thiol titration results were also consistent with the introduction of a cysteine residue at position 358 in Glu358Cys. The Cys mutant contained 7.18 (±0.25) free cysteines, as opposed to the native enzyme which contained only 6.07 123 (±0.15). On the basis of the amino acid sequence, 6 cysteines were expected for native p-glucosidase. To ensure that native P-glucosidase and Glu358Cys possess approximately the same secondary structure, the circular dichroism (CD) spectra of the two enzymes were determined. As illustrated in Figure 4-1, the two enzymes exhibit very similar spectra. Thus, substituting Glu358 with a cysteine does not grossly alter the conformation of the enzyme. 200 210 220 230 240 250 Wavelength (nm) Figure 4-1. The CD spectra of native fi-glucosidase ( ) and the Glu358Cys mutant (- - -). 4.3.2 KINETIC EVALUATION OF THE GLU358CYS MUTANT The replacement of Glu358 with a cysteine (refer to Scheme 4-2), in the Glu358Cys mutant, changes the length, nucleophilicity, and leaving group ability of the side chain at the active site nucleophile position. Since a sulfhydryl group (-SH) is a reactive nucleophile, one 124 would expect the shortened cysteine side chain to still function effectively as a nucleophile, assuming it can reach the anomeric carbon of the substrate. However, this substituted side chain should serve as a relatively poor leaving group when compared to the carboxylate side chain of glutamate. }~2A CH 2 SH CH 2 CH 2 Glu358 Cys358 Scheme 4-2. The side chains of glutamate and cysteine (shown in their expected ionization states at pH 7). Enzymatic activity can potentially be affected in several ways as a result of the substitution of Glu358 with a cysteine. If the shortened sulfhydryl side chain is too far removed or sterically hindered to make the nucleophilic attack on the substrate, the Cys mutant would be inactive. This was indeed the case for the active site nucleophile cysteine mutant of B. circulans xylanase (refer to section 2.3.2, Chapter 2). Alternatively, if the cysteine side chain can reach the anomeric carbon of the substrate, one might expect the Cys mutant to hydrolyze substrate via the mechanism illustrated in Scheme 4-3. The strong nucleophilicity of the cysteine side chain should result in the rapid formation of the covalent glycosyl-enzyme intermediate. However, the subsequent hydrolysis of this intermediate to yield free enzyme should be hindered by the poor leaving group ability of the cysteine side chain. Thus, the glycosylation rate should be greater than the deglycosylation rate, thereby resulting in the accumulation of the intermediate and a phenolate burst. 125 [4.2] N 0 2 Glycosylation + Q 2N N 0 2 SH (k2) [4.3] i C H 2 CH 2 Cys358 Cys358 H 2Q Deglycosylation (k3) HO-HO OH S" CH 2 Cys358 Scheme 4-3. The hydrolysis of 2 ',4'-dinitrophenyl /3-D-glucopyranoside (4.2) by the Glu358Cys mutant. The proposed reaction scheme is based on the double displacement mechanism illustrated in Chapter 1, Scheme 1-3 and assumes that the side chain of Cys358 is of sufficient length to attack the anomeric carbon of the substrate. The rate constants ki and k$ are for the glycosylation and deglycosylation steps, respectively. .4.3.2.1 Determination of the Kinetic Parameters for the Hydrolysis of 2',4'-Dinitrophenyl (5 -D-glucopyranoside The Glu358Cys mutant (0.039 mM) was incubated with various concentrations (0.10 to 2.1 mM) of 2',4'-dinitrophenyl p-D-glucopyranoside (2,4-DNPG, structure 4.2, Scheme 4-3) and the release of 2,4-dinitrophenolate (2,4-DNP, structure 4.3) monitored at 400 nm (refer to section 5.4.2, Chapter 5 for the experimental details). Shown in Figure 4-2 are the absorbance versus time data3 obtained for the reaction of 0.26 mM 2,4-DNPG with 0.039 3 The data shown in Figure 4-2 have been corrected for the absorbance due to enzyme, 2,4-DNPG, and the spontaneous hydrolysis of 2,4-DNPG. 126 mM enzyme. As can be seen, the release of 2,4-DNP occurs as an initial burst followed by a steady state. From the magnitude of the burst (observed4 AA400 = 0.044 ± 0.0016, expected AA400 = 0.043), it was determined that 1 mol of 2,4-DNP is released per mol of enzyme. A full-sized 2,4-DNP burst was observed at all the substrate concentrations tested, with the exception of 0.10 mM 2,4-DNPG. At this reduced substrate concentration, only -0.6 equivalents of 2,4-DNP was released. 0.16 400 800 Time (min) 1200 Figure 4-2. The hydrolysis of 2,4-DNPG (0.26 mM) by the Glu358Cys mutant (0.039 mM). The data shown have been corrected for the absorbance due to enzyme, 2,4-DNPG, and the spontaneous hydrolysis of 2,4-DNPG. The fact that a full-sized burst of 2,4-DNP is released when Glu358Cys is treated with 0.26 mM 2,4-DNPG (refer to Figure 4-2) clearly indicates that the observed pre-steady state activity is associated with the Cys mutant and not the result of a contaminant. To determine 4 For a description of how this value was determined, refer to section 5.4.2, Chapter 5. The AA400 values given are for 1 mm pathlength cells. 127 whether the observed steady state activity is real or due to a contaminant, an inactivation study was performed using 2',4'-dinitrophenyl 2-deoxy-2-fluoro-(3-D-glucopyranoside (2F-DNPG, structure 4.1). Treatment of the mutant (0.3 mM) with 2F-DNPG (0.02 mM) resulted in a 70% reduction in the steady state rate in 26 hours5 (refer to section 5.4.3, Chapter 5 for the experimental details). Since only 0.07 equivalents of inactivator was used, a 7% reduction in the steady state rate is expected if this activity is solely due to Glu358Cys. Thus, the observation of a 70% decrease in activity indicates that 63% (70% - 7%) of the observed steady state activity for Glu358Cys is due to a contaminant (possibly either low levels of native (3-glucosidase or another active mutant). A parallel experiment was run in which native enzyme (7.2 x 10"8 mM) 6 was also incubated with 2F-DNPG (0.02 mM). Complete inactivation of native (3-glucosidase was observed in one minute. This suggests that the contaminant is most likely not native enzyme since a much longer incubation period (26 hours) was required to get the observed inactivation in the Glu358Cys mutant. Analysis of the absorbance versus time data obtained for 0.26 mM 2,4-DNPG (shown in Figure 4-2), and the other substrate concentrations tested, yielded values for the pre-steady state pseudo first-order rate constants (k0t,s) and the steady state rates. From these results, values for the pre-steady state kinetic parameters, and k 2 (refer to Table 4-1; for a graphical representation of the data see Appendix C), and the steady state kinetic parameters, K m and k c a t , were determined (see section 5.4.2, Chapter 5 for more details on the data analysis). 5 There was no significant effect on the pre-steady state activity of Glu358Cys upon treatment with 2F-DNPG (0.02 mM) for 26 hours. 6 This is the estimated concentration of native enzyme that would produce the steady state activity shown in Figure 4-2. 128 Table 4-1. The Pre-Steady State Kinetic Parameters for the Hydrolysis of 2,4-DNPG by Glu358Cys and Native fj-Glucosidase Enzyme Kd(mM) Glu358Cys 0.27 ± 0.068 5.5 x 10" 4 ±4 .5x 10"5 Native3 0.65 ± 0.022b 1300 ± 2 1 b a Data taken from Namchuk & Withers (1995). b The pre-steady state kinetic parameters were determined at 5 °C. As illustrated in Table 4-1, replacing the active site nucleophile, Glu358, with a cysteine had a dramatic effect on the glycosylation step of the double displacement mechanism (refer to Scheme 4-3; for a brief discussion on the interpretation of Kd, k 2 , K m , and k c a t see Appendix A). The rate constant for the glycosylation step (k2) is decreased at least 2 x 106 fold for Glu358Cys relative to the native enzyme. Since the cysteine side chain should function effectively as a nucleophile, this reduction in k 2 is most likely due to the shortened side chain being poorly positioned to make the attack on the substrate. The insertion of a cysteine at the active site nucleophile position also had a drastic impact on the deglycosylation step of the double displacement mechanism. The apparent k c a t and K m values for the Glu358Cys mutant were determined to be 3.5 x 10"5 s"1 (±1.8 x 10"6)7 and <0.10 mM, respectively. The results of the 2F-DNPG inactivation study indicate that 63% of the steady state activity (and k c a t value) is due to a contaminant, while 37% is attributed to the Glu358Cys mutant. Taking this into account, the corrected k c a t value for Glu358Cys is 1.3 x 10'5 s_1 (±6.7 x 10*7). For native (3-glucosidase, the corresponding k c a t At all substrate concentrations tested (0.10 to 2.1 mM), very similar steady state rates were obtained. Therefore, the k^ value given was determined from the average of the five obtained steady state rates. 129 and K m values were 130 s"1 (±3) and 0.022 mM (±0.0012), respectively (Namchuk & Withers, 1995). Like the Glu358Cys mutant, native (3-glucosidase produces a pre-steady state phenolate burst when incubated with the substrate 2,4-DNPG (Namchuk & Withers, 1995). This implies that deglycosylation is the rate-determining step for both enzymes with this particular substrate. In this case, the k c a t values given above reflect k 3 , the rate constant for the deglycosylation step (refer to Appendix A for a detailed explanation). Substituting a cysteine for a glutamic acid at the active site nucleophile position results in a 107 fold decrease in the k c a t (k3) value. Thus as expected, the cysteine side chain functions as a relatively poor leaving group when compared to the carboxyl side chain of glutamic acid. 4.3.3 A N A L Y S I S OF G L U C O S Y L A T E D GLU358CYS 4.3.3.1 Treatment of the Glu358Cys and Glu358Gly Mutants with Excess 2,4-DNPG: ESMS Analysis and Thiol Titrations The kinetic evaluation of the Glu358Cys mutant revealed that the cysteine side chain at position 358 is capable of functioning as the catalytic nucleophile in the glycosylation step of the reaction. Thus, the mutant should theoretically form the covalent glycosyl-enzyme intermediate shown in Scheme 4-4 (also refer to Scheme 4-3). The determined k c a t (k^) value of 1.3 x 10"5 s"1 (refer to section 4.3.2.1) shows that the hydrolysis of this intermediate is slow, while the observation of a 2,4-DNP burst when Glu358Cys is incubated with 0.26 mM 2,4-DNPG (refer to Figure 4-2) indicates that the intermediate is accumulating. This build-up (or "trapping") of the covalent glycosyl-enzyme intermediate should be detectable by ESMS. 130 HO- x U 163 Da I M U — 1 — HO| S CH 2 Cys358 Scheme 4-4. The proposed covalent glycosyl-enzyme intermediate formed upon treating Glu358Cys with 2,4-DNPG. Glu358Cys, and the inactive Glu358Gly mutant, were incubated with a 74 fold molar excess of 2,4-DNPG (37 °C, 24 hrs; refer to section 5.4.4.1, Chapter 5 for the complete experimental details). The treated mutants were subjected to ESMS analysis, the results of which are shown in Figure 4-3. As illustrated in Figure 4-3a, untreated Glu358Cys consists of two enzyme species. The peak at 51 176 Da is due to the unprocessed enzyme, still possessing its N-terminal methionine. The peak at 51 044 Da results from the processed enzyme which has lost its N-terminal methionine residue (-133 Da). Upon treatment with 2,4-DNPG, both of these enzyme species are completely converted to new species having a molecular mass 164-165 Da greater than the Glu358Cys control (see Figure 4-3b). This mass increase is as expected for the covalent attachment of a single glucosyl moiety (molecular mass -163 Da) and confirms the trapping of the covalent glycosyl-enzyme intermediate. Glu358Gly, exposed to the same treatment, showed no significant molecular mass increase Q relative to its control (refer to Figure 4-3c and d). The absence of any significant labelling in the Glu358Gly mutant suggests that the glucosyl moiety is attached to Cys358 in the treated The peaks at 51 127 and 50 998 Da are due to the unprocessed Glu358Gly mutant and the processed mutant which has lost its N-terminal methionine residue, respectively. The peak at 51 289 could result from a small amount of labelling of the unprocessed mutant. The active site residue Tyr298 is the likely target of glycosylation. Previous work done on another nucleophile mutant, Glu358Asp, demonstrated that this residue is capable of acting as a "substitute" active site nucleophile, attacking the substrate 2,4-DNPG to form a stable oc-D-glucopyranosyl tyrosine residue (Gebler et al., 1995). 131 Glu358Cys mutant. The thiol titration results were consistent with this proposal, with Glu358Cys possessing approximately one less free cysteine after treatment with 2,4-DNPG (treated and untreated Glu358Cys contained 5.64 (±0.06) and 7.11 (±0.02) free cysteines, respectively). _ , . . I . , • • ' I 1 I 1 ' W • ' 1 ' 1 ' 1 1 ' 1 ' • > • • • • ! 50600 50800 51000 51200 51400 50600 50800 51000 51200 51400 Molecular Mass (Da) Molecular Mass (Da) Figure 4-3. ESMS analysis of the Glu358Cys and Glu358Gly mutants after treatment with 2,4-DNPG (74 fold molar excess, 37 0 C, 24 hrs). Shown are the reconstructed mass spectra of (a) untreated Glu358Cys, (b) 2,4-DNPG-treated Glu358Cys, (c) untreated Glu358Gly, and (d) 2,4-DNPG-treated Glu358Gly. 132 4.3.3.2 Identification of the Glucosylated Amino Acid in 2,4-DNPG-Treated Glu358Cys To provide additional proof that Cys358 was indeed the site of glucosylation in 2,4-DNPG-treated Glu358Cys, the treated mutant and a control9 were digested using pepsin and the glucosylated peptide identified using HPLC-ESMS (refer to section 5.4.4.2, Chapter 5 for the complete experimental details). Shown in Figure 4-4a and b are the HPLC traces of the digested, treated mutant and its control. A comparative analysis of the mass spectra obtained for the two HPLC traces showed that the only significant differences between the treated Glu358Cys digest and its control were in HPLC peak areas 1 and 4. Examination of the mass spectra of HPLC peak area 1 (refer to Figure 4-4c and d 1 0) revealed the presence of a unique peak (m/z 1006.5) in the digest of treated Glu358Cys. Similarly, a comparison of the mass spectra of HPLC peak 4 (see Figure 4-4e and f) revealed a peak at m/z 844.5 unique to the digested control sample. These two unique peaks, differing in mass by 162 Da (z = +1; the mass of a glucosyl moiety is -163 Da), are attributed to the glucosylated peptide and its nonglucosylated counterpart. As expected, the glucosylated peptide (m/z 1006.5), being more polar, is retained less by the reverse-phase HPLC column than the corresponding nonglucosylated peptide (m/z 844.5) (the retention times for the two peptides are -21 to 24 min and -27 min, respectively). 9 These samples were prepared by incubating Glu358Cys with either a 0 or 74 fold molar excess of 2,4-DNPG (pH 7, 37 °C, 24 hrs). 1 0 For the Glu358Cys control digest, no peaks were detected at m/z > 900. 133 100] a in c I D c 50 <D > CL 24.3 31.1 37.9 Time (min) 44.6 100 Time (min) 100, (0 c CP c 50 > CD rr 664.0 371.0 466.5 i ! M l i 1006.5 ,l,lL.l. IL 400 600 800 m/z 1000 100, d 664.0 50 371.0 486.5 JiiJ ii li.l, I I , 400 600 m/z 800 100, 791.5 533.0 co c <D *-_C CO > 5 © rr 50 632.5 l i. I , I , 1 400 600 800 1000 1200 m/z 100 791.5 50 533.0 632.5 lliJllLl 844.5 400 600 800 m/z 1000 1200 Figure 4-4. ESMS analysis of digested, 2,4-DNPG-treated Glu358Cys and its control, (a) The HPLC trace of digested, treated Glu358Cys, (b) the HPLC trace of the digested Glu358Cys control, (c) the mass spectrum of HPLC peak area 1 in panel a (treated Glu358Cys), (d) the mass spectrum of HPLC peak area 1 in panel b (control), (e) the mass spectrum of HPLC peak 4 in panel a (treated Glu358Cys), (f) the mass spectrum of HPLC peak 4 in panel b (control). 134 A computer search was performed to find all peptides in the Glu358Cys mutant with a molecular mass of 844.5 ± 1 Da (the mass of the nonglucosylated peptide). This search yielded 17 possible peptides, 5 of which contained Cys358. To unambiguously identify the glucosylated peptide and confirm that the glucosyl moiety was attached to Cys358, the labelled peptide (m/z 1006.5) was isolated and sequenced11 using the modified Edman degradation procedure shown in Scheme 4-5 (for the experimental details refer to Chapter 5, section 5.4.4.2). The N-terminal amino acid of the peptide was labelled using the novel sequencing reagent 4-(3-pyridinylmethylaminocarboxypropyl) phenyl isothiocyanate (denoted PITC 311 1 2 , structure 4.4, Hess et al., 1995). Using mild acidic conditions, this derivatized amino acid was cleaved and spontaneously cyclized to form structure 4.5. Subsequent analysis using HPLC-ESMS resulted in the identification of the released 311 phenylthiohydantoin (PTH) derivatized amino acid (4.5). The isolated, glucosylated peptide was subjected to seven rounds of the modified Edman degradation procedure. The results, summarized in Table 4-2, identified the peptide sequence as Y I T X N G A (where X corresponds to an unknown amino acid). Examination of the mass spectrum obtained for cycle 4 (refer to Figure 4-5) revealed a unique peak at m/z 577.0 that was absent in cycle 3. While this mass does not correspond to that expected for the 311 PTH derivative of any known amino acid, it is the exact mass predicted for a derivatized, glucosylated cysteine (see structure 4.6). The other peaks in the spectrum result from either contaminating amino acid derivatives or other impurities which coelute with the m/z 577.0 species. 1 1 The amino acid sequencing was kindly performed by David Chow in the laboratory of Dr. Ruedi Aebersold at the Biomedical Research Centre, University of British Columbia. 1 2 311 refers to the molecular mass of the reagent. 135 N: O II H O R' O I II I II <\ / )-CH 2 NH-C-(CH 2 ) 3 —1 \ A — N = C = S + H 2 N-C-C-NH-CH-C-R [4.4] -CH 2 NH-C-(CH 2 ) 3 -Labelling S H O R' O II I II I II -NH-C-NH-C-C-NH-CH-C-N= O II CH 2 NH-C-(CH 2 ) 3 -[4.5] Release/cyclization •N o R' O I II H 2 N-CH-C-NH H R HPLC-ESMS analysis Scheme 4-5. The Edman degradation procedure using the novel sequencing reagent 4-(3-pyridinylmethylaminocarboxypropyl) phenyl isothiocyanate (PITC 311, structure 4.4). The 311 phenylthiohydantoin (PTH) derivatized amino acid is shown as structure 4.5. The side chain of the amino acid is represented by R and R'. The determined sequence of the glucosylated peptide in 2,4-DNPG-treated Glu358Cys, Y I T X N G A , aligns with residues 355-361 of the published sequence of Agrobacterium P-glucosidase (Wakarchuk et al., 1988). Therefore, the unknown amino acid (X), which exhibited a molecular mass consistent with a glucosylated cysteine (refer to Figure 4-5 and structure 4.6), corresponds to position 358, the active site nucleophile position. Thus, this result provides additional support for Cys358 being the site of glucosylation in 2,4-136 DNPG-treated Glu358Cys and is consistent with the proposal that Cys358 forms a stabilized covalent glycosyl-enzyme intermediate with this substrate. Table 4-2. Results of the Modified Edman Degradation of the Glucosylated Peptide Cycle m/z of 311 PTH Derivative Identified Amino Acida 1 475.0 Y b 2 425.2 I 3 413.2 T 4 577.0 X c 5 426.2 N 6 369.0 G 7 383.2 A a The identification was made by comparing the mass and chromatographic retention time of the released 311 PTH derivative with the values obtained for authentic 311 PTH standards. b Refer to Abbreviations and Symbols for the full names of the amino acids. 0 X corresponds to an unknown amino acid. 368 Da 209 Da A , r H [4.6] 137 100 CP 75 CO I 50 CD CC 25 577.0 255.0 285.2 413.2 399.2 369.0 426.2 300 400 500 m/z 600 700 Figure 4-5. The mass spectrum of the 311 PTH derivatized amino acid released in cycle 4 of the modified Edman degradation procedure. 4.3.3.3 Urea Denaturation of Glucosylated Glu358Cys Previous work had indicated that the presence of a covalently bound sugar in the active site of a glycosidase enhances its stability (Stevenson, 1994). In the case of native Agrobacterium P-glucosidase, the presence of a 2-fluoroglucosyl moiety covalently bound to Glu358 resulted in considerable stabilization (at least 3 kcal/mol) to the chemical denaturant urea (Stevenson, 1994). While exposure to 8 M urea (at 40 °C) caused essentially complete unfolding for native P-glucosidase, insignificant unfolding (<10%) was observed for the glycosylated enzyme.13 This increased stability is attributed to the hydrogen bonding interactions that occur between the bound sugar moiety and the active site residues. To test whether the presence of a sugar moiety covalently bound to Cys358 also confers stability, a urea denaturation study was conducted on Glu358Cys and the 3 The protein unfolding was monitored by CD. 138 glucosylated mutant.14 The two enzymes were incubated with various concentrations of urea (at 40 °C) and the extent of protein unfolding monitored by CD (refer to section 5.4.4.3, Chapter 5 for the complete experimental details). As illustrated in Figure 4-6, the glucosylated mutant is more resistant to urea denaturation. For Glu358Cys and glucosylated Glu358Cys, respective urea concentrations of ~5 and ~6 M are required to achieve 50% unfolding ([9]/[9]o= 0.5). Thus for both the native enzyme and Glu358Cys, the presence of a sugar moiety covalently bound to the side chain at the active site nucleophile position results in enhanced stability to urea denaturation. This suggests that in the glucosylated Glu358Cys mutant, the sugar is binding in the active site in a manner similar to that of the normal glycosyl-enzyme intermediate. 1.00 £ 0.80 "i r ® 0.60 ~ 0.40 0.20 0.00 A • • A T T 4 6 [urea] (M) 8 Figure 4-6. The urea denaturation curves for the Glu358Cys mutant (T) and the glucosylated mutant (A). The [0] values shown were determined at 222 nm after an incubation time of 1 hour (no further unfolding was observed at incubation times in excess of 1 hour). The symbol [6]o corresponds to the mean residue ellipticity atOM urea. 1 4 Glucosylated Glu358Cys was prepared by incubating the mutant with a 74 fold molar excess of 2,4-DNPG (pH 7.0, 37 °C, 24 hrs). 139 4.4 C O N C L U S I O N Can a cysteine residue function as the catalytic nucleophile in retaining glycosidases? This question was addressed by performing a kinetic analysis on the active site nucleophile cysteine mutants of two retaining p-glycosidases. In the case of B. circulans xylanase, the active site nucleophile cysteine mutant was totally inactive, not even undergoing the first step (glycosylation) of the double displacement mechanism to form a covalent thioglycoside intermediate. Thus, for this particular glycosidase the cysteine residue is unsuitable for the role of the catalytic nucleophile. In contrast, the corresponding cysteine mutant (Glu358Cys) of Agrobacterium p-glucosidase was able to complete the glycosylation step of the double displacement mechanism, thereby forming a covalent thioglycoside intermediate. However, the rate constant for the formation of this intermediate (k2) was reduced at least 2 x 106 fold relative to native p-glucosidase. Therefore, while Cys358 is capable of acting as the active site nucleophile in this retaining glycosidase, it does so rather ineffectively when compared to Glu358. The dramatic reduction in k 2 is most likely due to the shortened cysteine side chain being poorly positioned to make the nucleophilic attack on the substrate. As expected, Cys358 served as a very poor leaving group in the subsequent hydrolysis (deglycosylation) step. The rate constant for this step (k3) was decreased 107 fold for the Glu358Cys mutant relative to the native enzyme. Thus, Cys358 functions inefficiently in both the capacity of catalytic nucleophile and leaving group. Based on these results, it seems highly unlikely that a cysteine residue could effectively function as the active site nucleophile in a retaining glycosidase. CHAPTER 5 MATERIALS AND METHODS 141 5.1 MATERIALS The following compounds were prepared by members of Dr. Steve Withers' research group. Dr. Lothar Ziser synthesized 2',5'-dinitrophenyl p-xylobioside (2,5-DNPX2), 3',4'-dinitrophenyl P-xylobioside (3,4-DNPX2), o-nitrophenyl P-xylobioside (ONPX 2), p-nitrophenyl P-xylobioside (PNPX 2), phenyl P-xylobioside (PhX 2), and 2',4'-dinitrophenyl 2-deoxy-2-fluoro-P-xylobioside (2F-DNPX 2). Xylobiose was isolated and purified by Dr. Michael Korner, while Dr. Mark Namchuk prepared 2',4'-dinitrophenyl P-D-glucopyranoside (2,4-DNPG) and 2',4'-dinitrophenyl 2-deoxy-2-fluoro-p-D-glucopyranoside (2F-DNPG). Di -P,P'-D-xylopyranosyl disulfide (XSSX) was synthesized as reported previously (Stanek et al., 1965) and had a satisfactory ' H N M R spectrum, melting point, and elemental analysis. A l l chemicals and buffer materials were obtained from the Sigma or Aldrich Chemical Companies unless otherwise stated. The protease, pepsin, and the column chromatography resins were purchased from Boehringer Mannheim and Pharmacia Biotech, respectively. 5.2 GENERAL PROCEDURES 5.2.1 ENZYME ASSAYS AND KINETIC STUDIES A l l enzyme assays and kinetic studies, unless otherwise stated, were done with 1 cm pathlength micro black quartz cuvettes using either a Unicam 8700 or Pye Unicam PU 8800 UV/Vis spectrophotometer, each of which was equipped with a circulating water bath and 142 thermostatted cuvette holders. The xylan assays were performed using a single beam UV/Vis spectrophotometer equipped with an autocell (Pye Unicam PU 8600). 5.2.2 TH IOL T ITRATIONS Thiol titrations were performed with plastic cuvettes (1 cm pathlength) using a thermostatted double beam UV/Vis spectrophotometer (Pye Unicam PU8800). The number of free thiol groups was determined using the method of Ellman (1959) as follows. To a solution of 0.2 mM 5,5'-dithiobis(2-nitrobenzoic acid), 6 M guanidine hydrochloride, 20 mM HEPES, 1 mM E D T A (pH 7.4) warmed to 25 °C was added enzyme (2.5-5 uM). The absorbance at 412 nm, due to the released 2-nitro-5-thiobenzoate, was monitored over a 30 minute period. The concentration of free thiol groups was calculated from the net AA412, corrected for a buffer blank, using the extinction coefficient determined from a cysteine standard curve generated under the same conditions (£412 values typically ranged from 13 000 to 15 000 M ' 1 cm'1). Division of this obtained value by the protein concentration yielded the number of free thiol groups. 5.2.3 C I R C U L A R D ICHROISM STUDIES A l l circular dichroism (CD) spectra were obtained using a 1 mm pathlength thermostatted cylindrical quartz cell in a JASCO J-710 spectropolarimeter connected to a circulating water bath. For the xylanase samples, the spectra were recorded at 40 °C using enzyme (0.2 mg/mL) diluted in 10 mM MES, 50 mM NaCl (pH 6.0) buffer. For the p-glucosidase samples, the spectra were recorded at 37 °C using a solution of 0.1 mg/mL enzyme in 50 mM sodium phosphate buffer (pH 7.0). For both enzymes the CD spectra, 143 typically acquired from 260 to 200 nm, were corrected for the buffer background. The corrected data, normalized for protein concentration and the optical pathlength, are presented as the mean residue ellipticity [0] (deg cm 2 dmol"1) versus wavelength. 5.2.4 E L E C T R O S P R A Y MASS S P E C T R O M E T R Y STUDIES The electrospray mass spectrometry (ESMS) experiments were kindly performed by either Dr. Shichang Miao, Dr. Dave Burgoyne, David Chow, or Shouming He. A l l experiments were done using an HPLC-ESMS setup consisting of a microbore HPLC (Michrom U M A ) connected on-line to either a PE-Sciex API HI or API 300 MS as described by Hess et al. (1993). Intact xylanase samples (10-20 |j,g) were injected onto a microbore PLRP column (1 x 50 mm) and eluted with a gradient of 20-100% solvent B over 3 minutes followed by 100% solvent B for 7 minutes (solvent A : 0.05% trifluoroacetic acid, 2% acetonitrile in water; solvent B: 0.045% trifluoroacetic acid, 80% acetonitrile in water). A post-column flow splitter was used to introduce 15% of the HPLC eluant into the MS, via an ion spray ion source, while the remaining 85% was recovered. The mass spectrometer was operated in the single quadrupole mode with the quadrupole mass analyzer scanning over a m/z range of 400-2600 Da. Protein molecular weights were determined from this data using deconvolution software supplied by Sciex. Intact P-glucosidase samples (10-20 |Ltg) were analyzed using the above procedure with a modified elution gradient (20-100% solvent B over 10 minutes followed by 100% solvent B for 2 minutes). 5.3 INVESTIGATION OF NATIVE XYLANASE AND MUTANTS 144 5.3.1 ENZYME ISOLATION AND PURIFICATION The xylanases from Bacillus circulans and Bacillus subtilis used for this work were cloned and expressed in E. coli (Sung et al., 1993). The mutagenesis, detection of xylanase mutants, and the production and purification of the mutant enzymes and native xylanase were kindly performed by Dr. Warren Wakarchuk, National Research Council of Canada, Ottawa as previously described (Sung et a l , 1993; Wakarchuk et al., 1994)., The Glu78Cys mutant was isolated from E. coli cell paste, again provided by Dr. Wakarchuk, using a modification of the procedure described in Sung et al. (1993). The following purification steps were performed at 4 °C. The cell paste was ground with 2 times its weight in alumina and the resulting mixture diluted with 50 mM MES buffer (pH 6.5). The mixture was then centrifuged (23 000 g, 30 min), to pellet out the alumina and cellular debris, and the resulting supernatant loaded onto a SP Sepharose column (5 x 20 cm). After washing the column with 50 mM MES buffer (pH 6.5) to remove any unbound proteins, the bound proteins were eluted with a 0-1 M linear NaCl gradient (2 L). The column fractions were analyzed using SDS-PAGE and those fractions containing xylanase, indicated by the presence of a 20 kDa band, were pooled and concentrated using 10 kDa nominal cut-off centrifugal concentrators (Amicon Inc., Beverly, MA) . The concentrated pool (2.5 mL; approximately 30 mg) was centrifuged (15 000 g, 20 min) prior to being loaded onto a Sephacryl SI00 high resolution gel filtration column (1.6 x 80 cm). The column was eluted with 50 m M MES buffer (pH 6.5) and the column fractions containing xylanase were pooled and concentrated. SDS-PAGE analysis of the concentrated enzyme (at 1 mg/mL) revealed 145 the presence of only one band (at 20 kDa), indicating that the Glu78Cys mutant is >95% pure. The enzyme concentration was determined from A2go measurements using an e = 4.08 mg-1 mL c m 1 (Wakarchuk et al., 1994). 5.3.2 CHEMICAL MODIFICATION OF THE GLU78CYS AND GLU172CYS MUTANTS 5.3.2.1 Methyl Methanethiolsulfonate (MMTS) Labelling of Glu78Cys Enzyme (2 mg/mL; 0.1 mM) in 16 mM glycylglycine, 1 mM EDTA buffer (pH 8.0) was reacted with 10 m M M M T S (100 fold molar excess over the cysteine residue) for 24 hours at 4 °C. After removing an aliquot for ESMS analysis, excess reagent was removed by diluting the reaction mixture 20 fold with 20 mM MES, 50 mM NaCl buffer (pH 6.0) and then concentrating the sample down to its original volume using an Amicon 10 kDa centriprep (repeated three times). The protein concentration was then determined from A 2 so measurements. 5.3.2.2 Iodoacetate (IAA) Labelling of Glu78Cys under Nondenaturing Conditions The IAA labelling of the Glu78Cys mutant, under nondenaturing conditions, was tried using a wide range of conditions of pH (pH 7 to 11), buffer species, and IAA concentrations (10 to 2000 fold molar excess over the cysteine residue). The following procedure yielded the largest observed activity regain. Enzyme (2 mg/mL; 0.1 mM) was incubated, in the dark, with 120 mM IAA, 150 mM CAPSO, and 1 mM EDTA (pH 10) at 36 °C. A pH of approximately 10 was maintained through periodic additions of 1 M NaOH. Reactivation of 146 the enzyme was followed by assaying aliquots of the reaction mixture with 1.2 mM 2,5-D N P X 2 (pH 6.0, 40 °C) and following the release of 2,5-dinitrophenol at 440 nm (see section 5.3.3.3, paragraph one, for more details). Upon reaching maximal reactivation, excess reagent was removed by applying the reaction mixture to a Sephadex G-25 column. After the column was eluted with 50 mM MES buffer (pH 6.5), the appropriate fractions were pooled, concentrated (using 10 kDa concentrators), and the protein concentration determined. 5.3.2.3 IAA Labelling of Glu78Cys under Partially Denaturing Conditions A solution of enzyme (2 mg/mL; 0.1 mM), 120 mM IAA, 10% glycerol, 1 mM EDTA, and 150 m M CAPSO (pH 10) (or 150 mM AMPSO, pH 9) was incubated in the dark at 36 °C in the presence of different concentrations of urea (0, 1, 2, or 3 M). The pH of the solution was maintained through periodic additions of 1 M NaOH. At certain time intervals, an aliquot was removed and diluted five fold with 20 mM MES, 50 mM NaCl, 10% glycerol buffer (pH 6). After incubating the diluted sample at 25 °C for 24 hours (to permit refolding), the sample was assayed using 1 mM 2,5-DNPX 2 (pH 6.0, 40 °C) (see section 5.3.3.3, paragraph one, for more details). 5.3.2.4 IAA Labelling of Glu78Cys and Glul72Cys under Fully Denaturing Conditions The Glu78Cys mutant (1 mg/mL; 0.05 mM) in 7.1 M urea, 11% glycerol, 400 mM glycylglycine, 40 m M HEPES, and 1 mM EDTA (pH 7.5) was incubated at 60 °C for 20 minutes. The reaction mixture was then cooled to 40 °C, IAA added to give a final reagent concentration of 0.5 mM (10 fold molar excess of IAA over the cysteine residue), and the reaction mixture incubated in the dark for 19 hours. To stop the reaction and decrease the 147 enzyme concentration to 0.05 mg/mL (to minimize aggregation upon refolding), the reaction mixture was diluted 20 fold with the above 7.1 M urea solution. The labelled enzyme was then refolded by dialyzing the diluted enzyme mixture against 20 mM HEPES, 10% glycerol buffer (pH 7.0, room temperature, 48 hours). The dialyzed sample was filtered through a 0.2 |im filter, concentrated using a 10 kDa concentrator, and its protein concentration determined. The Glul72Cys mutant was labelled with IAA using the above protocol with the following modification. After adding a final reagent concentration of 1.1 mM IAA (22 fold molar excess over the cysteine residue), the reaction mixture was incubated in the dark for 24 hours. 5.3.3 K INET IC STUDIES USING SYNTHET IC SUBSTRATES 5.3.3.1 Determination of Extinction Coefficients The extinction coefficients for o-nitrophenol (ONP) and /?-nitrophenol (PNP) were determined at 40 °C in 20 m M MES, 50 mM NaCl buffer (pH 6.0) using matched quartz cells. The assay wavelength at which the extinction coefficient was measured, and subsequent kinetic studies done, was chosen as the point of maximal absorbance difference between the phenol and its corresponding xylobioside. This was determined by taking a wavelength scan of the two compounds under the assay conditions. Phenols and xylobiosides were dried in vacuo over P2O5, weighed, and dissolved in a known volume of water. Aliquots of the stock phenol and xylobioside solutions were added to the above buffer, warmed to 40 °C, and the absorbance measured (in duplicate) at the appropriate assay wavelength. From the obtained average absorbance value, the extinction coefficient (e) was 148 determined using Beer's law (equation 5.1), where A corresponds to the absorbance, b to the e = A (5.i) be pathlength (1 cm), and c to the concentration of phenol (mM). For ONP and PNP the determined extinction coefficients and assay wavelengths are Ae = 1.07 mM" 1 cm"1 (400 nm) and Ae = 1.66 mM" 1 cm"1 (400 nm) respectively (where Ae is the extinction coefficient difference between the phenol and its corresponding xylobioside). The extinction coefficients for 2,5-dinitrophenol (2,5-DNP), 2,4-dinitrophenol (2,4-DNP), and 3,4-dinitrophenol (3,4-DNP) were determined by Dr. Lothar Ziser (Ziser et al., 1995) using a similar procedure and are as follows: 2,5-DNP, 440 nm, Ae = 3.57 mM" 1 cm"1; 2,4-DNP, 400 nm, Ae = 10.80 mM" 1 cm"1; 3,4-DNP, 400 nm, Ae = 11.71 mM" 1 cm"1. 5.3.3.2 A Pre-Steady State Kinetic Study of Native Xylanase with 2,5-DNPX2 The pre-steady state kinetic study was performed using an Applied Photophysics M V 17 microvolume stopped flow spectrophotometer equipped with a temperature bath. A substrate solution (1 or 3.8 mM 2 ,5 -DNPX2) and enzyme solution (1 mg/mL; 0.05 mM) in 20 mM MES, 50 mM NaCl buffer (pH 6.0) were cooled to 5 °C. The reaction was started by driving together 50 \iL aliquots of the above two solutions and the release of 2,5-dinitrophenol monitored at 440 nm over a two second time period. 5.3.3.3 Steady State Kinetic Studies Using Synthetic (3-Xylobioside Substrates The initial rates of enzymatic hydrolysis for all the substrates, with the exception of PhX2, were determined using a continuous assay. An appropriate concentration of substrate 149 in 0.1% BSA, 20 mM MES, 50 mM NaCl buffer (pH 6.0) was warmed to 40 °C. Reaction was initiated by the addition of enzyme and substrate hydrolysis monitored by measuring the release of phenol at the appropriate wavelength. In order to ensure constant initial rates, enzyme concentrations and reaction times were chosen such that less than 10% of the total substrate was hydrolyzed. Xylanase-catalyzed hydrolysis rates for the substrate PhX 2 were determined using a stopped assay. Different concentrations of PhX 2 in 0.1% BSA, 20 mM. MES, 50 mM NaCl buffer (pH 6.0, 190 uX) were warmed to 40 °C and the reaction was initiated by the addition of a 10 | i L aliquot of enzyme. After an appropriate time, 0.6 mL of a 0.2 M N a 3 P 0 4 solution (pH 12.15) was added to stop the reaction. The absorbance of the released phenol at 288 nm was determined immediately and corrected for the spontaneous hydrolysis of PhX 2 and the background absorbance of the enzyme. The extinction coefficient for phenol under these assay conditions, Ae = 2.17 m M - 1 cm - 1 , was determined using the procedure described in section 5.3.3.1. Hydrolysis rates were generally determined at 6 to 8 different substrate concentrations ranging from 0.1 to 4 times the K m value, except when limited by substrate solubility. The observed rate of absorbance change (AA/min), during enzymatic hydrolysis, was converted to rate of phenol release (v) using equation 5.2, where Ae corresponds to the extinction v = AA/min H- Ae (5.2) coefficient difference between the phenol and its corresponding xylobioside (refer to section 5.3.3.1 for the Ae values for the various substituted phenols). 150 From the experimental rate versus substrate concentration data, values of K m and k c a t were determined directly by fitting the data to the Michaelis-Menten equation using the computer program GraFit (Leatherbarrow, 1992). Standard errors were calculated by the same fitting program. The data were also plotted according to the method of Lineweaver and Burk and the resulting double reciprocal plots are presented in Appendix C for visual convenience. 5.3.3.4 The Effects of Exogenous Nucleophiles on Reaction Rates and Products The reaction rates for the synthetic substrates, ONPX2 and 2,5-DNPX2, with native xylanase and the Glu 172 mutants were determined in the presence of sodium azide as follows. A solution containing various concentrations of sodium azide (0-500 mM), 20 mM O N P X 2 (or 2 mM 2,5-DNPX 2), 0.1% BSA, 20 mM MES, 50 mM NaCl (pH 6.0) was warmed to 40 °C. The reaction was initiated by the addition of enzyme and the release of product monitored at 440 nm or 400 nm for 2,5-DNPX2 and ONPX2, respectively. The possible cleavage of the substrates by azide (500 mM) alone was also monitored and proved to be negligible. Reaction product analysis was performed by thin layer chromatography on 60 F254 silica gel aluminum plates (Merck) run in 7:2:1 (v/v/v) ethyl acetate/methanol/water and developed with 10% H2SO4 in methanol. The reaction rates for 2,5-DNPX2 with native xylanase were also determined in the presence of neutral nucleophiles. A solution of 3 mM 2,5-DNPX 2 , 0.1% BSA, 20 mM MES, 50 mM NaCl (pH 6.0) was warmed to 40 °C. The reaction was initiated by the addition of 151 enzyme (or buffer) and various concentrations of the exogenous nucleophiles (0-1.2 M MeOH, 0-500 mM DTT, or 0-1 M p-mercaptoethanol). The reaction was monitored at 440 nm in both the presence and absence of enzyme. Cleavage of the substrate by the exogenous nucleophiles proved to be significant, thus requiring the enzyme-catalyzed rates to be corrected for this competing reaction. 5.3.3.5 The pH Dependence of kca/Km The k c a t / K m values for the hydrolysis of O N P X 2 at each pH value were determined as follows. A solution of O N P X 2 (0.5 mM (0.01-0.2 x K m ) for the IAA-Glu78Cys, IAA-Glu 172Cys, Glul72Asp, and Glul72Cys mutants and 0.3 m M (0.02 x K m ) for native xylanase), 0.1% BSA, and the appropriate buffer was warmed to 25 °C. The reaction was initiated by the addition of a 10 pL aliquot of enzyme (or buffer) and the release of o-nitrophenol monitored at 400 nm until the substrate was depleted. The pH of the reaction mixture was then determined and an aliquot assayed for activity at pH 6.0 to check enzyme stability. The buffers used were as follows: 20 mM succinic acid, 50 mM NaCl (pH 2.5-5); 20 mM MES, 50 mM NaCl (pH 5-7); 20 mM HEPES, 50 mM NaCl (pH 7-8.5); 20 mM AMPSO, 50 mM NaCl (pH 8.5-9.0). The absorbance versus time data were fit to a first-order rate equation using the program GraFit, thus yielding the pseudo first-order rate constants (k0bs) at each pH. At low substrate concentrations ([S] « K m ) , the reaction rate (v) is given by equation 5.3, where v = ^ [ E ] T [ S ] (5.3) 152 [E]T is the total enzyme concentration and k 0 b s = (k c a t /K M ) [E]x- The k c a t / K m values were thereby obtained by dividing the experimental k 0 b s values by the enzyme concentration. The p K a values were assigned by direct fitting of the k c a t / K m versus pH data to the appropriate pH dependent rate equation using GraFit. 5.3.3.6 Determination of the Stereochemical Course of Hydrolysis The hydrolysis of the synthetic substrate 2,5-DNPX 2 by the IAA-Glu78Cys mutant was monitored by J H N M R spectroscopy using a Bruker WH-400 spectrometer (the internal standard was the HOD signal). The substrate 2,5-DNPX 2 and the buffer, 10 m M sodium phosphate, 50 m M NaCl (pH 6.0), were freeze-dried from D 2 0 three times. The IAA-Glu78Cys mutant was repeatedly diluted with D 2 0 buffer and reconcentrated using a 10 kDa concentrator. For both the substrate and enzyme, the D 2 0 used in the final washing had an isotopic purity of 99.96%. To a 5 mm N M R tube was added 0.5 mL of 5.4 mM 2,5-DNPX 2, 10 mM sodium phosphate, and 50 m M NaCl (pH 6.0) in D 2 0 (99.96%). After recording the initial N M R spectrum of substrate and buffer, enzyme (6 [iL, 7.5 mg/mL) was added. The stereochemical course of the reaction was monitored at 25 °C by collecting spectra at time intervals over a 72 minute period. Based on chemical shifts and coupling constants it was possible to assign the stereochemistry at the anomeric center of the sugar product xylobiose. 5.3.4 S T E A D Y S T A T E K I N E T I C S T U D I E S U S I N G T H E N A T U R A L S U B S T R A T E X Y L A N The substrate used in these studies, soluble xylan, was prepared as follows. A 5% suspension of birchwood xylan was magnetically stirred at room temperature for 3 hours. 153 Upon removal of insoluble material via centrifugation, the supernatant was freeze-dried and stored in a dessicator. Xylanase-catalyzed hydrolysis rates for the soluble birchwood xylan were determined using a reducing sugar assay. Various concentrations of xylan in 0.1% BSA, 20 mM MES, 50 mM NaCl buffer (pH 6.0) were warmed to 40 °C and the reaction started by the addition of enzyme. At an appropriate time, a 50 (J,L aliquot was removed and diluted with 50 mM NaOH (1 mL) to stop the reaction. The concentration of reducing sugars produced was determined spectrophotometrically with the hydroxybenzoic acid hydrazide reagent (Lever, 1972), using xylose as a standard. Protein concentrations and reaction times were chosen so that each enzyme hydrolyzed xylan to the same extent, thereby making a relative comparison of kinetic parameters possible. Hydrolysis rates for xylan were determined at 6 to 7 different substrate concentrations ranging from 0.01 to 7 times the estimated K m value, where possible. From the experimental rate versus substrate concentration data, values of Km(app) and kca t(app) were calculated directly using the program GraFit, as described in section 5.3.3.3. For visual convenience, the data are presented as Lineweaver-Burk plots in Appendix C. 5.3.5 INACT IVAT ION STUDIES USING 2F-DNPX 2 5.3.5.1 Glu78Cys + 2F-DNPX2: Contamination Check To determine whether the observed activity for the Glu78Cys mutant was real or due to a contaminant, the following experiment was performed. The enzyme (9.7 mg/mL; 0.47 mM) was incubated at 40 °C with 0.043 mM 2F-DNPX 2 (0.09 equiv), 20 m M MES, and 50 154 mM NaCl (pH 6.0). At time intervals, residual activity was determined by assaying an aliquot of the inactivation mixture with 1 mM 2,5-DNPX 2 (pH 6.0, 40 °C) and following the release of 2,5-dinitrophenol at 440 nm (see section 5.3.3.3, paragraph one, for more details). 5.3.5.2 Active Site Titration of the lAA-Labelled Glu78Cys Mutant Using 2F-DNPX2 To a solution of 1.6 m M 2F-DNPX 2 , 20 mM MES, 50 mM NaCl (pH 6.0) warmed to 25 °C was added enzyme (final concentration of 1 mg/mL (0.05 mM) and 3 mg/mL (0.15 mM) for IAA-labelled Glu78Cys prepared under fully denaturing and nondenaturing conditions, respectively). The 2,4-dinitrophenol released, due to inactivation of the active enzyme species, was monitored at 400 nm. The obtained A400 versus time data were corrected for the absorbance due to enzyme, 2F-DNPX2, and the spontaneous hydrolysis of 2F-DNPX2. The corrected data were then fit to an equation describing a first-order reaction followed by a steady state, thus yielding a value for the y-intercept (net AA4oo)- From the net AA400 the concentration of 2,4-dinitrophenol released, and thereby the concentration of active enzyme, was determined (Ae = 10.80 m M - 1 cm - 1). The complete inactivation of IAA-labelled Glu78Cys was confirmed by assaying a 10 (iL aliquot of the reaction mixture with 1.2 mM 2,5-DNPX 2 (190 uL, pH 6.0, 40 °C). 5.3.5.3 Determination of the ki and Ki Values for the Inactivation of IAA-Glu78Cys by 2F-DNPX2 The kinetic parameters for the inactivation of IAA-Glu78Cys by 2F-DNPX2 were determined as follows. The IAA-Glu78Cys mutant (0.26 mg/mL; 0.013 mM) was incubated 155 at 40 °C with different concentrations of 2F-DNPX 2 (0, 0.46, 0.68, 1.08, and 1.91 mM) in 0.1% BSA, 20 m M MES, 50 mM NaCl buffer (pH 6.0). At time intervals, residual enzyme activity was determined by adding a 5 | i L aliquot of the inactivation mixture to a solution (190 |iL) of 1.2 m M 2,5-DNPX 2 , 0.1% BSA, 20 mM MES, and 50 mM NaCl (pH 6.0, 40 °C) and monitoring the release of 2,5-dinitrophenol at 440 nm. The residual enzyme activity versus time data were fit to a single exponential decay equation using the program GraFit, thus yielding pseudo first-order rate constants (k 0 b s ) at each inactivator concentration. The obtained k0t, s versus 2F-DNPX 2 (I) concentration data were subsequently fit to equation 5.4 kj[l] k o b s " K i + [l] ( 5 4 ) using GraFit, thus giving values for ki (the inactivation rate constant) and Kj (the equilibrium dissociation constant). Due to the low solubility of 2F-DNPX 2 , only inactivator concentrations below the estimated Kj value could be examined. Therefore, the values of k; and Kj are only estimates, but an accurate kj/Kj value was calculated from the slope of the reciprocal plot ( l / k o b s versus 1/[I]). 5.3.5.4 Reactivation of2F-DNPX2 Inactivated IAA-Glu78Cys Reactivation of inactivated IAA-Glu78Cys was investigated as follows. The IAA-Glu78Cys mutant (0.4 mg/mL; 0.02 mM) was incubated with 1.9 mM 2F-DNPX 2 , 20 mM MES, and 50 mM NaCl (pH 6.0) at 40 °C for 2 hours, at which time less than 5% activity remained. Excess inactivator was removed by concentrating the sample using a 10 kDa 156 concentrator, diluting 10 fold with 20 mM MES, 50 mM NaCl buffer (pH 6.0), and reconcentrating (repeated twice). Inactivated enzyme (0.10 mg/mL) in 0.1% BSA, 20 mM MES, 50 mM NaCl buffer (pH 6.0) was then incubated at 40 °C with either buffer or 155 mM xylobiose until full reactivation was observed. At time intervals, 5 (iL aliquots of the reactivation mixture were removed and assayed for activity using 1.2 mM 2,5-DNPX2 as described in section 5.3.5.3. The activity versus time data were corrected for enzyme inactivation due to denaturation using data from a non-inhibited control sample. The first-order rate constants for reactivation ( k r e o bs) were obtained from the slope of plots of ln(full rate minus observed rate) versus time. 5.3.6 INACTIVATION STUDIES USING DI-p\ p'-D-XYLOPYRANOSYL DISULFIDE 5.3.6.1 The Effects of pH on Inactivation Rates The effects of pH on the inactivation of Glul72Cys by di-(3,|3'-D-xylopyranosyl disulfide (XSSX) were studied as follows. The Glul72Cys mutant (1 mg/mL; 0.05 mM) was incubated at 40 °C with 25 mM X S S X (500 fold molar excess), 40 mM CAPSO, and 1 mM EDTA at pH 8, 9, and 10. At time intervals, a 4 |xL aliquot of the inactivation mixture was removed, added to a solution of 1 mM 2,5-DNPX 2 , 20 mM MES, 50 mM NaCl (200 u\L, pH 6.0, 40 °C), and the release of 2,5-dinitrophenol monitored at 440 nm. A control experiment was run in parallel in which enzyme, in the absence of inactivator, was incubated at pH 10. 157 5.3.6.2 Determination of the Inactivation Parameters ki and K{ The kinetic parameters for the inactivation of Glul72Cys by X S S X were determined using the following procedure. The Glul72Cys mutant (1 mg/mL; 0.05 mM) was incubated at 40 °C with various concentrations of X S S X (0, 1, 5, 25, 50, and 97 mM) in 40 mM CAPSO, 1 mM E D T A (pH 10). At time intervals, residual enzyme activity was determined by assaying a 4 p:L aliquot of the inactivation mixture with 1 mM 2,5-DNPX2 (pH 6.0, 40 °C) as outlined in section 5.3.6.1. The residual enzyme activity versus time data were analyzed as described in section 5.3.5.3 with the following modification. The obtained k 0b S values were corrected for enzyme inactivation due to denaturation using data from the 0 mM control sample. Since only inactivator concentrations below the estimated Kj were examined, the values of k, and Kj obtained are only approximate. An accurate k/Kj value was determined from the slope of the reciprocal plot ( l /k 0 D S versus 1/[I]). 5.3.7 ESMS ANALYS IS OF DIGESTED 2F-DNPX 2 INACT IVATED IAA-GLU78CYS The IAA-Glu78Cys mutant and 2F-DNPX 2 inactivated enzyme were digested as follows. Both enzymes were first heat-denatured by boiling for 2 minutes. The denatured enzymes (20 jiL, 4.6 mg/mL) were then incubated with 70 p i of pepsin (0.14 mg/mL) in 50 mM sodium phosphate buffer (pH 2) at room temperature for 12 hours. SDS-PAGE analysis of the proteolytic digests confirmed that both samples were completely digested. ESMS analysis of the proteolytic digests was performed by David Chow using a PE-Sciex API HI electrospray mass spectrometer. In each experiment, the digest was loaded onto 158 a Reliasil C]8 column (1 x 150 mm) and eluted with a gradient of 0-60% solvent B over 20 minutes followed by 100% solvent B for 2 minutes (refer to section 5.2.4 for the composition of solvents A and B). A neutral loss MS/MS experiment was performed in which the two quadrupoles, Q l and Q3, were offset by 133.5 Da, corresponding to the loss of the 2F-xylobiosyl label from a doubly charged peptide. Only those peptides that lost 133.5 Da in the m/z range of 300 to 1700 (0.5 Da step) were detected. A subsequent daughter ion MS/MS experiment of the single peptide which underwent neutral loss was carried out by selectively introducing the doubly charged m/z 842.5 peptide through Q l into the collision cell (Q2) and observing the daughter ions in Q3. Thus Q l was locked on m/z 842.5, while the scan range for Q3 was set from 50 to 1700 Da at 1 Da step sizes so that it would capture the entire mass range of the possible daughter ions as well as the parent ion. The collision gas thickness for the daughter ion MS/MS and neutral loss MS/MS experiment was set at approximately 438 (4.38 x 10 1 4 molecules/cm2) and 355 (3.55 x 10 1 4 molecules/cm2), respectively. 5.3.8 ESMS ANALYSIS OF DIGESTED IAA-GLU172CYS AND NATIVE XYLANASE The IAA-Glul72Cys mutant and native xylanase were digested with cyanogen bromide (CNBr) using the following procedure. A solution of enzyme (1 mg/mL; 0.05 mM) and 19 mM CNBr (190 fold molar excess over methionine residues) in 0.1 M HC1 was incubated in the dark, under a nitrogen atmosphere, at room temperature for 25 hours. After quenching the reaction by quick freezing (using liquid N 2 ) , the excess CNBr was removed by 159 freeze-drying (repeated twice). The freeze-dried, digested enzymes, dissolved in 6 M guanidine hydrochloride, 20 mM HEPES, and 1 mM EDTA (pH 7.4), were analyzed by Shouming He using a PE-Sciex API 300 electrospray mass spectrometer. The peptide mixtures were separated by reverse-phase HPLC, using the column and elution gradient described in section 5.3.7, prior to ESMS analysis. In the first ESMS experiment, the mass spectrometer was operated in the single quadrupole mode with the quadrupole mass analyzer scanning over a m/z range 300-2400 Da with a step size of 0.5 Da. A daughter ion MS/MS experiment was then conducted on the doubly charged m/z 853 peptide, in digested IAA-Glu 172Cys, by selectively introducing this peptide through Q l into the collision cell (Q2) and observing the daughter ions in Q3. Thus Q l was locked on m/z 853, while the scan range for Q3 was set from 50 to 1720 Da at 0.5 Da step sizes. The C A D (collision assisted dissociation) setting in the single quadrupole mode and under the MS/MS conditions was 0 and 3, respectively. 5.3.9 MOLECULAR MODELLING OF SELECTED XYLANASE MUTANTS Crystallographic coordinates of B. circulans xylanase and the Glul72Cys mutant, cocrystallized with xylotetraose, were supplied by Dr. Warren Wakarchuk (Wakarchuk et al., 1994; Campbell et al., 1993) and were visualized using Biosym Technologies Insight II graphics software. The molecular modelling of selected xylanase mutants was kindly done by Manish Joshi using the same program, while the surface accessibility calculations were carried out by Dr. Greg Connelly using the program SURFV (Sridharan et al., 1995). 160 5.4 INVESTIGATION OF THE GLU358CYS MUTANT OF 3-GLUCOSIDASE 5.4.1 ENZYME ISOLATION AND PURIFICATION Native Agrobacterium P-glucosidase and the Glu358 mutants used for this work were cloned and expressed in E. coli (Trimbur et al., 1992; Wakarchuk et a l , 1986). The mutagenesis, detection of p-glucosidase mutants, and the production of the mutant enzymes and native P-glucosidase were kindly performed by Dr. Don Trimbur, Department of Microbiology, University of British Columbia as previously described (Trimbur et al., 1992). The isolation of native Agrobacterium p-glucosidase and the Glu358Gly mutant from E. coli cell paste was done by Karen Rupitz and Dr. Qingping Wang as described previously (Withers et al., 1992; Kempton & Withers, 1992). The Glu358Cys mutant was isolated and purified from E. coli cell paste using a modification of the above protocol. The cell paste was ground with 2.5 times its weight in alumina followed by dilution with 25 mM sodium phosphate, 10 mM p-mercaptoethanol (pH 7.0). This step and all subsequent steps were performed at 4 °C. The mixture was then centrifuged (23 000 g, 15 min), to pellet out the alumina and cellular debris, and to the resulting supernatant was added 1.5% (w/v) streptomycin sulfate. After incubating for 4 to 5 hours, the precipitated nucleic acids were removed by centrifugation (25 000 g, 25 min) and the resulting supernatant loaded onto a D E A E Sephacel column (5 x 25 cm). The column was washed with 50 mM sodium phosphate buffer (pH 7.0) to remove any unbound proteins and then the bound proteins were eluted with a 4 L linear gradient of 0-1 M NaCl in the above buffer. The column fractions were analyzed using SDS-PAGE and those fractions containing p-glucosidase, indicated by 161 the presence of a 50 kDa band, were pooled and dialyzed against 50 mM sodium phosphate buffer (pH 7.0, 2 x 4 L, 48 hours). The dialyzed pooled fractions were then loaded onto a Q-Sepharose column (1.6 x 10.5 cm), the column washed with the above buffer, and eluted with a 0-0.5 M linear NaCl gradient (2.5 L). Fractions containing P-glucosidase (again determined via SDS-PAGE analysis) were pooled and concentrated using Amicon 30 kDa centripreps. The concentrated pool (2 mL; approximately 50 mg) was centrifuged (15 000 g, 20 min) prior to being loaded onto a Sephacryl S100 high resolution gel filtration column (1.6 x 80 cm). The column was eluted with 50 mM sodium phosphate buffer (pH 7.0) and the appropriate fractions pooled and concentrated. SDS-PAGE analysis of the concentrated enzyme (at 1 mg/mL) revealed the presence of only one band (at 50 kDa), thus the Glu358Cys mutant is >95% pure. The enzyme concentration was determined from A 2 8 0 measurements using an e = 2.184 mg"1 mL cm"1 (Kempton & Withers, 1992). 5.4.2 GLU358CYS + 2,4-DNPG: DETERMINATION OF KINETIC PARAMETERS The following kinetic experiment was performed using 1 mm pathlength quartz cells. Various concentrations of 2,4-DNPG in 50 m M sodium phosphate buffer (pH 6.8) were warmed to 37 °C. The reaction was initiated by the addition of enzyme (final concentration of 1.9 mg/mL; 0.039 mM) or buffer and substrate hydrolysis monitored by measuring the release of 2,4-dinitrophenol at 400 nm (Ae = 10.91 mM" 1 cm"1; Kempton & Withers, 1992). Reaction rates were measured at 6 different substrate concentrations ranging from 0.4 to 8 times the estimated value. 162 The A400 versus time data obtained were corrected for the absorbance due to enzyme, 2,4-DNPG, and the spontaneous hydrolysis of 2,4-DNPG. The corrected data were then fit to an equation describing a first-order reaction followed by a steady state, thus yielding the pre-steady state pseudo first-order rate constant (k0bS), the steady state rate, and the y-intercept (net AA400) at each substrate concentration. The pre-steady state kinetic parameters, K j and k 2[S] k o b s ~ K d + [S] ( 5 , 5 ) k 2, were determined from these k 0 b s values by direct fitting to equation 5.5. Direct fitting of the steady state rate versus substrate concentration data to the Michaelis-Menten equation yielded the steady state kinetic parameters, K m and k c a t . The program GraFit was used for the fitting of all data and the determination of standard error values. Refer to Appendix C for a graphical representation of the data. 5.4.3 GLU358CYS + 2F-DNPG: CONTAMINATION CHECK To determine whether the observed steady state activity for the Glu358Cys mutant was real or due to a contaminant the following experiment was performed. The enzyme (15 mg/mL; 0.3 mM) was incubated at 37 °C with either 0 mM or 0.02 m M 2F-DNPG (0.07 equiv), 50 m M sodium phosphate (pH 6.8). At time intervals, enzyme activity was determined by assaying a 30 u\L aliquot of the inactivation mixture with 0.3 mM 2,4-DNPG (170 u,L, pH 6.8, 37 °C) and following the release of 2,4-dinitrophenol at 400 nm. 163 5.4.4 A N A L Y S I S O F G L U C O S Y L A T E D G L U 3 5 8 C Y S 5.4.4.1 Treatment of Glu358Cys and Glu358Gly with Excess 2,4-DNPG: ESMS Analysis and Thiol Titrations The Glu358Cys mutant (3.6 mg/mL; 0.07 mM) was incubated with 5.2 mM 2,4-DNPG (74 fold molar excess), 1 mM EDTA, and 50 mM sodium phosphate buffer (pH 7.0) at 37 °C for 24 hours. An aliquot was removed for ESMS analysis (refer to section 5.2.4 for the experimental details) and the remaining reaction mixture was loaded onto a Sephadex G-25 column (to remove any unreacted 2,4-DNPG and the yellow color caused by the 2,4-dinitrophenol released during the incubation period). After eluting the column with 50 mM sodium phosphate buffer (pH 7.0), the appropriate fractions were pooled, and the protein concentration determined from A280 measurements. A thiol titration was then performed on the column-purified, colorless Glu358Cys mutant as described in section 5.2.2. A control experiment was run simultaneously in which the inactive Glu358Gly mutant was also treated with a 74 fold molar excess of 2,4-DNPG using the above conditions. The treated enzyme was subjected to ESMS analysis. 5.4.4.2 ESMS Analysis of Digested Glucosylated Glu358Cys Glucosylated Glu358Cys was prepared by incubating the mutant with a 74 fold molar excess of 2,4-DNPG (pH 7.0, 37 °C, 24 hours) as described in section 5.4.4.1. The glucosylated Glu358Cys and a control sample, exposed to the same conditions in the absence of 2,4-DNPG, were digested as follows. The enzymes (0.86 mg/mL) were incubated with 164 pepsin (0.016 mg/mL) in 50 mM sodium phosphate buffer (pH 2) at room temperature for 1 hour. The proteolytic digests were analyzed by Dr. Shichang Miao using a PE-Sciex API HI electrospray mass spectrometer. The peptide mixtures were first separated by reverse-phase HPLC using the column and elution gradient described in section 5:3.7. ESMS analysis was performed in the single quadrupole mode with the quadrupole mass analyzer scanning over a m/z range 300-2400 Da. A comparative analysis of the HPLC-ESMS data obtained for the glucosylated sample and its control resulted in the identification of the HPLC peak which contained the glucosylated peptide. The labelled peptide was subsequently collected via the post-column flow splitter using the mass spectrometer, operating in the single quadrupole mode, as a detector. This peptide was further purified on the same HPLC column by using a shallower gradient (0-20% gradient B over 50 minutes). The amino acid sequence of the isolated, glucosylated peptide was determined by David Chow using a modified Edman degradation procedure. The peptide was degraded, in a stepwise manner, using the novel Edman sequencing reagent 4-(3-pyridinylmethyl-aminocarboxypropyl) phenyl isothiocyanate (denoted PITC 311; Hess et al., 1995). The 311 phenylthiohydantoin derivatized amino acids were separated by reverse-phase HPLC, using a Reliasil column (1 x 50 mm) and a gradient of 0-100% solvent B in 8 minutes, and then analyzed using a PE-Sciex API HI electrospray mass spectrometer. The mass spectrometer was operated in the single quadrupole mode with the quadrupole mass analyzer scanning over a m/z range 300-850 Da. 165 5.4.4.3 Urea Denaturation of Glucosylated Glu358Cys Glucosylated Glu358Cys (prepared as described in section 5.4.4.1) and a control sample (0.12 mg/mL) were incubated at 40 °C with various concentrations of freshly prepared urea (0-8 M) in 35 mM HEPES buffer (pH 7.5). At time intervals (1, 4, and 8 hours), CD spectra were acquired from 260 to 200 nm. The data were corrected for the buffer background and analyzed as outlined in section 5.2.3. 166 REFERENCES Amano, T., Tozawa, K. , Yoshida, M . , & Murakami, H. (1994) FEBS Lett. 348, 93-98. Ambler, R. P. (1965) Biochem. J. 96, 32P. Amyes, T. L. , & Jencks, W. P. (1989) J. Am. Chem. Soc. Ill, 7888-7900. Atsumi, S., Nosaka, C , Iinuma, H. , & Umezawa, K. (1993) Arch. Biochem. Biophys. 304, 302-304. Bairoch, A. , & Apweiler, R. (1996) Nucleic Acids Res. 24, 21-25. Barrett, T., Suresh, C. G., Tolley, S. P., Dodson, E. J., & Hughes, M . A . (1995) Structure 3, 951-960. Bennet, A. , & Sinnott, M . L . (1986) J. Am. Chem. Soc. 108, 7287-7294. Bray, M . R., & Clarke, A. J. (1990) Biochem. J. 270, 91-96. Bray, M . R., & Clarke, A. J. (1994) Eur. J. Biochem. 219, 821-827. Brown, H. C , McDaniel, D. H. , & Hafliger, O. (1955) in Determination of Organic Structures by Physical Methods (Braude, E. A. , & Nachod, F. C , Eds.) pp 567-662, Academic Press, New York. Campbell, R., Rose, D., Wakarchuk, W., To, R., Sung, W., & Yaguchi, M . (1993) in Proceedings of the 2nd TRICEL symposium on Trichoderma reesei cellulases and other hydrolases (Suominen, P., & Reinikainen, T., Eds.) pp 63-72, Foundation for Biotechnical and Industrial Fermentation Research, Helsinki, Finland. Cho, H. , Krishnaraj, R., Kitas, E., Bannwarth, W., Walsh, C. T., & Anderson, K. S. (1992) / . Am. Chem. Soc. 114, 7296-7298. Coughlan, M . P. (1992) in Xylans and Xylanases (Visser, J., et al., Eds.) pp 111-139, Elsevier Science Publishers, Amsterdam. Coughlan, M . P., & Hazlewood, G. P. (1993) Biotechnol. Appl. Biochem. 17, 259-289. Cupples, C. G., Miller, J. H. , & Huber, R. E. (1990) J. Biol. Chem. 265, 5512-5518. 167 Dahlqvist, A . (1961) Biochim. Biophys. Acta 50, 55-61. Davies, G., & Henrissat, B. (1995) Structure 3, 853-859. Day, A . G., & Withers, S. G. (1986) Biochem. Cell Biol. 64, 914-922. Dekker, R. F. H . (1982) Biotechnol. Lett. 4, 411-416. Dekker, R. F. H. , & Richards, G. N . (1976) Adv. Carb. Chem. Biochem. 32, 317-352. Derewenda, U . , Swenson, L. , Green, R., Wei, Y. , Morosoli, R., Shareck, F., Kluepfel, D., & Derewenda, Z. S. (1994) J. Biol. Chem. 269, 20811-20814. Dhalla, A . M . , L i , B. , Alibhai, M . F., Yost, K. J., Hemmingsen, J. M . , Atkins, W. M . , Schineller, J., & Villafranca, J. J. (1994) Protein Sci. 3, 476-481. Divne, C., Stahlberg, J., Reinikainen, T., Ruohonen, L., Pettersson, G., Knowles, J. K. C , Teeri, T. T., & Jones, T. A . (1994) Science 265, 524-528. Elbein, A . D., Legler, G., Tlusty, A. , McDowell, W., & Schwarz, R. (1984) Arch. Biochem. Biophys. 235, 579-588. Ellman, G. L . (1959) Arch. Biochem. Biophys. 82, 70-77. Ermert, P., Vasella, A. , Weber, M . , Rupitz, K., & Withers, S. G. (1993) Carbohydr. Res. 250, 113-128. Fersht, A . (1985) Enzyme Structure and Mechanism, 2nd ed., W. H. Freeman and Company, New York. Franck, R. W. (1992) Bioorg. Chem. 20,77-88. Gebler, J., Gilkes, N . R., Claeyssens, M . , Wilson, D. B., Beguin, P., Wakarchuk, W., Kilburn, D. G., Miller, R. C. J., Warren, R. A . J., & Withers, S. G. (1992a) J. Biol. Chem. 267, 12559-12561. Gebler, J. C , Aebersold, R., & Withers, S. G. (1992b) J. Biol. Chem. 267, 11126-11130. Gebler, J. C , Trimbur, D. E., Warren, R. A . J., Aebersold, R., Namchuk, M . , & Withers, S. G. (1995) Biochemistry 34, 14547-14553. 168 Gilkes, N . R., Henrissat, B., Kilburn, D. G., Miller, R. C. J., & Warren, R. A . J. (1991) Microbiol. Rev. 55, 303-315. Gloss, L . M . , & Kirsch, J. F. (1995) Biochemistry 34, 3990-3998. Han, Y . W., & Srinivasan, V . R. (1969) J. Bacteriol. 100, 1355-1363. Hardy, L . W., & Poteete, A. R. (1991) Biochemistry 30, 9457-9463. Henrissat, B., & Barrioch, H. (1993) Biochem. J. 293, 781-788. Herrchen, M . , & Legler, G. (1984) Eur. J. Biochem. 138, 527-531. Hess, D., Covey, T. C , Winz, R., Brownsey, R. W., & Aebersold, R. (1993) Protein Sci. 2, 1342-1351. Hess, D., Nika, H. , Chow, D. T., Bures, E. J., Morrison, H. D., & Aebersold, R. (1995) Anal. Biochem. 224, 373-381. H0j, P. B., Condron, R., Traeger, J., McAuliffe, J., & Stone, B. (1992) J. Biol. Chem. 267, 25059-25066. Huber, R. E., & Chivers, P. T. (1993) Carbohydr. Res. 250, 9-18. Jacobson, R. H. , Zhang, X-J . , DuBose, R. F., & Matthews, B. W. (1994) Nature 369, 761-766. Jones, C. C , Sinnott, M . L. , & Souchard, I. J. L. (1977) J. Chem. Soc. Perkin Trans II, 1191-1198. Kanda, T., Brewer, C. F., Okada, G., & Hehre, E. J. (1986) Biochemistry 25, 1159-1165. Katsube, Y . , Hata, Y. , Yamaguchi, H , Moriyama, H. , Shinmyo, A. , & Okada, H. (1990) in Protein Engineering: Protein Design in Basic Research, Medicine and Industry (Ikehara, M . , Ed.) pp 91-96, Japan Scientific Societies Press, Tokyo. Keitel, T., Simon, O., Borriss, R., & Heinemann, U . (1993) Proc. Natl. Acad. Sci. U.S.A. 90, 5287-5291. Kempton, J. B. , & Withers, S. G. (1992) Biochemistry 31, 9961-9969. 169 Ko, E. P., Akatsuka, H. , Moriyama, H. , Shinmyo, A. , Hata, Y . , Katsube, Y . , Urabe, I., & Okada, H. (1992) Biochem. J. 288, 117-121. Koshland, D. E. (1953) Biol. Rev. 28, 416-436. Kuroki, R., Weaver, L . H. , & Matthews, B. W. (1995) Nature Struct. Biol. 2, 1007-1011. Lai, E., & Withers, S. G. (1994) Biochemistry 33, 14743-14749. Leatherbarrow, R. J. (1992) GraFit Version 3.0, Erithacus Software Ltd., Staines, U.K. Legler, G. (1990) Adv. Carb. Chem. Biochem. 48, 319-385. Lehmann, J., & Schlesselmann, P. (1983) Carbohydr. Res. 113, 93-99. Lever, M . (1972) Anal. Biochem. 47, 273-279. Liu, W., Madsen, N . B. , Braun, C., & Withers, S. G. (1991) Biochemistry 30, 1419-1424. Lukac, M . , & Collier, R. J. (1988) J. Biol. Chem. 263, 6146-6149. MacLeod, A . M . , Lindhorst, T., Withers, S. G., & Warren, R. A . J. (1994) Biochemistry 33, 6371-6376. Malcolm, B. A. , Rosenberg, S., Corey, M . J., Allen, J. S., de Baetselier, A. , & Kirsch, J. F. (1989) Proc. Natl. Acad. Sci. U.S.A. 86, 133-137. Martin, J. L. , Veluraja, K. , Ross, K., Johnson, L. N . , Fleet, G. W. J., Ramsden, N . G., Bruce, I., Orchard, M . G., Oikonomakos, N . G., Papageorgiou, A . C , Leonidas, D. D., & Tsitoura, H . S. (1991) Biochemistry 30, 10101-10116. Matsushima, Y. , Kim, D., Yoshimura, T., Kuramitsu, S., Kagamiyama, H. , Esaki, N . , & Soda, K. (1994) J. Biochem. 115, 108-112. McCarter, J., Adam, M . , & Withers, S. G. (1992) Biochem. J. 286, 721-727. McCarter, J. D., & Withers, S. G. (1994) Curr. Opin. Struct. Biol. 4, 885-892. McCleary, B. V . (1992) in Xylans and Xylanases (Visser, J., et al., Eds.) pp 161-169, Elsevier Science Publishers, Amsterdam. 170 Mcintosh, L. P., Hand, G., Johnson, P. E., Joshi, M . D., Korner, M . , Plesniak, L. A. , Ziser, L. , Wakarchuk, W. W., & Withers, S. G. (1996) Biochemistry 35, 9958-9966. Miao, S., Ziser, L. , Aebersold, R., & Withers, S. G. (1994) Biochemistry 33, 7027-7032. Moreau, A. , Roberge, M . , Manin, C., Shareck, F., Kluepfel, D., & Morosoli, R. (1994) Biochem. J. 502,291-295. Moult, J., Eshdat, Y. , & Sharon, N . (1973) J. Mol. Biol. 75, 1-4. Namchuk, M . N . , & Withers, S. G. (1995) Biochemistry 34, 16194-16202. Oku, T., Roy, C., Watson, D. C., Wakarchuk, W., Campbell, R., Yaguchi, M . , Jurasek, L., & Paice, M . G. (1993) FEBS Lett. 334, 296-300. Olsen, O., Borriss, R., Simon, O., & Thomsen, K. (1991) Mol. Gen. Genet. 225, 177-185. Otey, F. H. , & Doane, W. M . (1984) in Starch: Chemistry and Technology (Whistler, R. L., BeMiller, J. N . , & Paschall, E. F., Eds.) pp 389-416, Academic Press, Inc., New York. Paice, M . G., Bourbonnais, R., Desrochers, M . , Jurasek, L. , & Yaguchi, M . (1986) Arch. Microbiol. 144, 201-206. Paice, M . G., Gurnagul, N . , Page, D. H. , & Jurasek, L. (1992) Enzyme Microb. Technol. 14, 272-276. Phillips, D. C. (1967) Proc. Natl. Acad. Sci. U.S.A. 57, 484-495. Post, C. B., & Karplus, M . (1986) J. Am. Chem. Soc. 108, 1317-1319. Shewale, J. G. (1982) J. Biochem. 14, 435-443. Sinnott, M . L. (1987) in Enzyme Mechanisms (Page, M . I., & Williams, A. , Eds.) pp 259-297, Royal Society of Chemistry, London. Sinnott, M . L . (1990) Chem. Rev. 90, 1171-1202. Sinnott, M . L . (1993) Bioorg. Chem. 21, 34-40. Sinnott, M . L. , & Souchard, I. J. L . (1973) Biochem. J. 133, 89-98. 171 Sinnott, M . L. , & Withers, S. G. (1974) Biochem. J. 143, 751-762. Sinnott, M . L. , Withers, S. G., & Viratelle, O. M . (1978) Biochem. J. 175, 539-546. Sridharan, S., Nicholls, A. , & Sharp, K. A. (1995) J. Comp. Chem. 16, 1038-1044. Stanek, J., Sindlerova, M . , & Cerny, M . (1965) Collection Czechoslov. Chem. Commun. 30, 297-303. Stevenson, A . D. (1994) M.Sc. Thesis, Department of Chemistry, University of British Columbia, Vancouver, Canada. Stryer, L . (1988) Biochemistry, 3rd ed., W. H. Freeman and Company, New York. Strynadka, N . C. J., & James, M . N . G. (1991) J. Mol. Biol. 220, 401-424. Sung, W. L. , Luk, C. K. , Zahab, D. M . , & Wakarchuk, W. (1993) Protein Express. Purif. 4, 200-206. Svensson, B. , Clarke, A. J., Svendsen, I., & M0ller, H. (1990) Eur. J. Biochem. 188, 29-38. Tao, B. Y . , Reilly, P. J., & Robyt, J. F. (1989) Biochim. Biophys. Acta 995, 214-220. Thompson, S., Mayerl, F., Peoples, O. P., Masamune, S., Sinskey, A . J., & Walsh, C. T. (1989) Biochemistry 28, 5735-5742. Torronen, A. , Harkki, A. , & Rouvinen, J. (1994) EMBO J. 13, 2493-2501. Torronen, A. , & Rouvinen, J. (1995) Biochemistry 34, 847-856. Trimbur, D. E., Warren, R. A . J., & Withers, S. G. (1992) J. Biol. Chem. 267, 10248-10251. Tull, D., & Withers, S. G. (1994) Biochemistry 33, 6363-6370. Tull, D., Withers, S. G., Gilkes, N . R., Kilburn, D. G., Warren, R. A . J., & Aebersold, R. (1991) J. Biol. Chem. 266, 15621-15625. Wakarchuk, W., Methot, N . , Lanthier, P., Sung, W., Seligy, V. , Yaguchi, M . , To, R., Campbell, R., & Rose, D. (1992) in Xylans and Xylanases (Visser, J., et al., Eds.) pp 439-442, Elsevier Science Publishers, Amsterdam. 172 Wakarchuk, W. W., Campbell, R. L., Sung, W. L., Davoodi, J., & Yaguchi, M . (1994) Protein Sci. 3, 467-475. Wakarchuk, W. W., Greenberg, N . M . , Kilburn, D. G., Miller, R. C , & Warren, R. A. J. (1988) J. Bacteriol. 170, 301-307. Wakarchuk, W. W., Kilburn, D. G., Miller, R. C , & Warren, R. A. J. (1986) Mol. Gen. Genet. 205, 146-152. Walsh, C. T. (1979) Enzymatic Reaction Mechanisms, W. H. Freeman and Company, New York. Wang, Q., Graham, R. W., Trimbur, D., Warren, R. A . J., & Withers, S. G. (1994) J. Am. Chem. Soc. 116, 11594-11595. Wang, Q., Trimbur, D., Graham, R., Warren, R. A . J., & Withers, S. G. (1995) Biochemistry 34, 14554-14562. Weber, J. P., & Fink, A . L . (1980) J. Biol. Chem. 255, 9030-9032. Weiser, W., Lehmann, J., Matsui, H. , Brewer, C. F., & Hehre, E. J. (1992) Arch. Biochem. Biophys. 292, 493-498. White, A. , Tull, D., Johns, K. , Withers, S. G., & Rose, D. R. (1996) Nature Struct. Biol. 3, 149-154. White, A. , Withers, S. G., Gilkes, N . R., & Rose, D. R. (1994) Biochemistry 33, 12546-12552. Wiesmann, C , Beste, G., Hengstenberg, W., & Schulz, G. E. (1995) Structure 3, 961-968. Withers, S. G., Rupitz, K. , & Street, I. P. (1988) J. Biol. Chem. 263, 7929-7932. Withers, S. G., Rupitz, K., Trimbur, D., & Warren, R. A. J. (1992) Biochemistry 31, 9979-9985. Withers, S. G., & Street, I. P. (1988) J. Am. Chem. Soc. 110, 8551-8553. Withers, S. G., Warren, R. A. J., Street, I. P., Rupitz, K., Kempton, J. B., & Aebersold, R. (1990) J. Am. Chem. Soc. 112, 5887-5889. Yang, R. C. A. , MacKenzie, C. R., & Narang, S. A . (1988) Nucleic Acids Res. 16, 7187. Yuan, J., Martinez-Bilbao, M . , & Huber, R. E. (1994) Biochem. J. 299, 527-531. Ziser, L. , Setyawati, I., & Withers, S. G. (1995) Carbohydr. Res. 274, 137-153. 174 APPENDIX A: ENZYME KINETICS A - l . Fundamental Equations of Enzyme Kinetics The Michaelis-Menten treatment of enzyme kinetics is based on the following reaction scheme where E represents enzyme, S the substrate, and P the product. In the first E + S , ES - E + P step of the reaction, the enzyme and substrate combine in a rapid, reversible process to produce an enzyme-substrate complex ES. In the second step, the bound substrate is converted to product and subsequently released from the enzyme (Fersht, 1985). The rate equation derived from this scheme (known as the Michaelis-Menten equation) is shown below (equation 1). v is the rate of reaction (measured by substrate „ = k c a t [ E ] o [ S ] K m + [S] breakdown or product release), [E] 0 is the total enzyme concentration, [S] is the substrate concentration, k c a t is the catalytic constant, and K m is the Michaelis constant. The Michaelis constant K m is defined as the substrate concentration at which the reaction rate (v) is half-maximal (v = V m a x / 2 ) . It may be treated as an apparent dissociation constant of all the enzyme bound species and as such is expressed as equation 2. Therefore, K -Jaa ( 2 ) K m " Z [ E S ] ( ' 175 the K m value reflects the stability of the bound enzyme-substrate complex. Lower values of K m indicate tighter binding of the substrate to the enzyme. At low concentrations of substrate ([S] « K m ) , the Michaelis-Menten equation reduces to equation 3 and v is linearly dependent upon the substrate concentration (refer to „ _ k c a t [ E ] 0 [ S ] ( 3 ) K m Figure A - l ) . Since most of the enzyme is unbound, the total enzyme concentration [E] 0 can be approximated to the concentration of free enzyme [E]. Under these conditions, the Michaelis-Menten equation can now be expressed as equation 4. As illustrated by equation v = ^at[E][S] ( 4 ) K m 4, k c a t / K m is an apparent second-order rate constant for the reaction of free enzyme with free substrate. This kinetic parameter also is a measure of the overall catalytic efficiency of the enzyme. At high substrate concentrations ([S] » K m ) , v approaches a limiting value, V m a x , and the Michaelis-Menten equation can now be expressed as equation 5. Vmax = kcat [E] 0 ( 5 ) For the purpose of graphical representation, the enzymatic rate data are frequently displayed in the form of a Lineweaver-Burk plot. This technique converts the Michaelis-Menten equation into a linear plot, thereby highlighting data which deviate from the expected behavior. To do this conversion, both sides of the Michaelis-Menten equation are inverted, 176 v V + • K m max max [s] ( 6 ) thus giving equation 6. As illustrated in Figure A-2, plotting 1/v as a function of 1/[S] yields a straight line with a slope of K m / V m a x , a y-intercept of 1/V m a x , and an x-intercept of -1 /K m . Figure A-l. The plot of rate (v) versus substrate concentration for a typical enzymatic reaction obeying Michaelis-Menten kinetics (Fersht, 1985). / //\siope=/fM/l/m a x - ' A M / V 0 |/[S] Figure A-2. A typical Lineweaver-Burk plot for an enzymatic reaction (Fersht, 1985). 177 A-2. The Kinetic Parameters for the Double Displacement Mechanism The Michaelis-Menten kinetic model also is applicable to more complex reaction schemes in which additional intermediates occur on the reaction pathway. In this case, the kinetic parameters K m , k c a t , and k c a t / K m are a composite of the various rate constants. Here, we will apply the Michaelis-Menten kinetic model to the double displacement mechanism proposed for a retaining glycosidase (discussed in Chapter 1, section 1.3). The reaction scheme for this mechanism may be depicted as follows, where the formation of ES is ki k9 ko E + S ^ = ^ E S ^ — EP —J— E + P k- ^ r Pi H 2 0 referred to as the association step, the conversion of ES to EP as the glycosylation step, and the product release as the deglycosylation step. The dissociation constant (Kj) for the ES complex is expressed as K d " [ E S ] - k l ( ? ) Assuming a steady state concentration of ES and EP is reached during the reaction, then equations 8 and 9 apply. k 2 [ES] = k 3 [EP] (8 ) ^ = k i [E ] [S ] -k_ 1 [ES] -k 2 [ES] = 0 ( 9 ) The total concentration of enzyme [E] 0 is the sum of the concentration of free enzyme and all the enzyme-bound species: 178 [E] 0 = [E] + [ES] + [EP] (10) Substituting the expression for [EP] (see equation 8) into equation 10 and rearranging yields [E] = [ E ] o - [ E S ] - ^ [ E S ] (11) k-3 Replacing the above expression for [E] in equation 9 gives ki[E] 0[S] = (k_i + k 2 ) [ES] + ^ k i k 3 k i k 2 ^ V k3 k3 [ES][S] (12) Solving equation 12 for [ES] gives [ES] = ki[E] 0[S] , k i ( k 2 + k 3 ) r e - , k - l + k 2 + ; L S J (13) k 3 At steady state, the rate of product formation v p is given by equation 14. m dt = k 3 [ E P ] = k 2 [ E S ] (14) Substituting equation 13 into the above equation gives the following expression for vE k 2 k 3 k 2 + k 3 [E]0[S] V p k 3 k_! + k 2 k 2 + k 3 k i (15) + [S] Equation 15 follows the standard form of the Michaelis-Menten equation (see equation 1). Consequently, the kinetic parameters for the double displacement mechanism can be expressed as follows: kCat = - ^ (16) k 2 + k 3 179 _ k 3 ( k _ 1 + k 2 ) k i ( k 2 + k 3 j kcat/Km = 7 ^ 7 - ( 1 8 ) k _ i + k 2 The parameter k c a t is the rate constant for the rate-determining step of the reaction and is always associated with the highest free energy step in the reaction pathway. The kinetic parameter k c a t / K m is the pseudo second-order rate constant for the free enzyme and free substrate proceeding to the transition step of the first irreversible step (Fersht, 1985). Consider the case where the glycosylation rate constant k 2 is increased relative to the deglycosylation rate constant k 3 , thereby resulting in deglycosylation being the rate-determining step. If one also assumes a rapid, reversible association of enzyme and substrate (i.e. k _ i » k 2 ) , then the equations for the kinetic parameters k c a t and k c a t / K m reduce to k c a t = k 3 (19) k c a t / K m = ^ (20) k - i These kinetic parameters can be expressed as the following Eyring equations: k c a t = ^ e - A G | p / R T (21) h kT kca t /K m = - ^ e - A G T / R T ( 2 2 ) h where k is the Boltzmann constant, T is the temperature, h is the Planck constant, and R is the gas constant. The Eyring equations relate these rate constants to differences in energy levels among the various species on the reaction pathway. As shown by Figure A-3 and 180 equations 21 and 22, k c a t refers to the deglycosylation transition state with the enzyme-product complex as the initial reference point. The parameter k c a t / K m refers to the transition state of the glycosylation step with the free enzyme (E) and the free substrate (S) as the initial reference points. E + P Reaction coordinate Figure A-3. The hypothetical Gibbs free energy diagram for a retaining glycosidase showing rate-limiting deglycosylation (i.e kj«k2). Now consider the situation where glycosylation is the rate-limiting step and there is a rapid, reversible association of enzyme and substrate (i.e. k 3 » k 2 and k _ i » k 2 ) . . The equations for k c a t and k c a t / K m are now kcat = k 2 ( 2 3 ) k C a t / K m = ^ (24) k - i and the corresponding Eyring equations are kT k c a t ^ e - A G c / R T (25) h 181 k c a t / K m = ^ e - A G T / R T (26) h In this case, k c a t and k c a t / K m both yield information pertaining to the glycosylation transition state. However, the initial reference point for k c a t is the ES complex, whereas for k c a t / K m it is free enzyme and free substrate. As demonstrated by the above two scenarios, k c a t always refers to the rate-determining step while k c a t / K m corresponds to the first irreversible step in the reaction (glycosylation) with free enzyme (E) and free substrate (S) as the inital reference points. A-3. The Kinetics of Inactivation The inactivation of a glycosidase can be represented by the following reaction scheme: E + I . 1 • E I U - E-I The first step involves the reversible binding of the inactivator (I) and the free enzyme (E). The second step, which is rate limiting, involves an irreversible, bond forming step that produces an inactivated enzyme intermediate (E-I). The kinetic equation for this inactivation, shown below, is a variation of the Michaelis-Menten equation (see equation 1). v is the rate ki[E] [I] v= 1 1 JorL-. (27) K i + P] of inactivation, [E] 0 is the total enzyme concentration, and kj is the inactivation rate constant. Kj is the apparent dissociation constant of all the bound enzyme species and as such is expressed as: 182 K i = M J l ( 2 8 ) I[EI] If the concentration of inactivator is significantly greater than the enzyme concentration ( [ I ]»[E]) , then [I] remains essentially constant throughout the inactivation process. As a result, the kinetics are pseudo first-order with respect to the enzyme concentration and equation 27 can now be expressed as v = k 0 bs [E] 0 (29) where k*=iSi] ( 3 0 ) The value of k0t,s (the pseudo first-order rate constant of inactivation) at each inactivator concentration can be determined by fitting the residual enzyme activity versus time data to a single exponential decay equation (equation 31) using the program GraFit (Leatherbarrow, 1992) (t represents time, [E]j the initial enzyme concentration, and [E] the [E] = [E] i e-k 0 bs t ( 3 1 ) active enzyme concentration). Subsequent fitting of the obtained k 0 D S versus inactivator concentration data to equation 30 (again using GraFit) yields values for k; and K j . 183 APPENDIX B: INTRODUCTION TO B R 0 N S T E D RELATIONSHIPS B-l . General Concept Br0nsted relationships are based on the following general equation log k = p log K + A ( 32 ) where k is a rate constant, K an equilibrium constant, and A and P are constants. Since the equilibrium constant K a (also known as the ionization constant) can be expressed as -log K a = p K a ( 33 ) equation 32 can be written in the following form: logk = - p p K a + A (34) A plot of log k versus p K a should yield a straight line with a slope -P and a y-intercept A . The values of p typically range from 0 to -1, the magnitude of the value being an indication of the degree of charge development in the transition state. Br0nsted plots (log k versus pK a ) have been used to examine the dependence of the rate of an enzymatic reaction on the equilibrium constant for phenol ionization as the phenol substituent on the substrate is varied. Such studies have provided insight into the nature of the transition state relative to the structural variation. B-2. The Use of Br0nsted Relationships in Enzymology Br0nsted relationships can provide valuable insights into the relationship between substrate structure and enzymatic activity. Their applications, however, are limited by two important aspects of enzymatic reactions, substrate binding and catalysis. Since enzymes are 184 rather specific for their substrates, modifications of the substrate could potentially affect its binding to the enzyme's active site. The resulting binding effects may obscure the electronic effects of the substituents on the substrate. In addition, substrate modification could affect the orientation of the substrate relative to the enzyme's catalytic residues, thereby interfering with catalysis. Despite these limitations, Br0nsted relationships have provided valuable mechanistic information on several glycosidases (Sinnott, 1990; Kempton & Withers, 1992; Tull & Withers, 1994). The Br0nsted plots obtained for the hydrolysis of aryl P-glucosides by Cellulomonas fimi exoglycanase (Tull & Withers, 1994) are presented below. C. fimi exoglycanase, a retaining glycosidase, is quite reactive towards a wide range of aryl P-glucosides (see structure 1). As illustrated in Figure B - l , both log k c a t and log kCat/Km are strongly correlated with the pK a of the phenol leaving group for the glucoside substrates. This dependence indicates that the glycosylation step is both the rate-determining step and the first irreversible step in catalysis, since it is this step in which the C-O bond to the leaving group is cleaved. The large p i g value obtained (Pi g = -1) reflects a large degree of negative charge accumulation on the phenolate oxygen at the glycosylation transition state. This result suggests that there is almost complete C-0 bond cleavage and relatively little proton donation. As demonstrated by this example, the use of Br0nsted relationships in enzymology can provide useful mechanistic information regarding the identification of both the rate-determining step and the first irreversible step, as well as provide insight into the degree of bond cleavage and charge distribution at the transition state. 185 Figure B-l. The Br0nsted plots for the hydrolysis of aryl fi-glucosides by C. fimi exoglycanase (reproduced from Tull & Withers, 1994). 186 A P P E N D I X C : G R A P H I C A L R E P R E S E N T A T I O N O F K I N E T I C D A T A 0 1 eo o E f E c E 0.6 -0.4 -0.2 -a. z o o E O) E c E 0 1 0.8 0.6 0.4 0.2 0 0 2 4 6 8 1/ [2,5-DNPXJ (mM"1) ' I ' I ' l ' l ' l ' l ' I ' l ' 1 0 1 2 3 4 5 1/ [3,4-DNPXJ (mM1) | | i l i l i — l— i l i I i I • I/• I l l , 0.1 0.2 0.3 0.4 0.5 0.6 1/ [PNPXJ (mM'1) 0 0.1 0.2 0.3 0.4 1/ [PhXJ (mM"1) Figure C-l. The Lineweaver-Burk plots for the hydrolysis of aryl ^-xylobiosides by native xylanase (1 mg enzyme = 0.049 jimol; refer to Chapter 5, section 5.3.3.1 for the Ae values for the various phenols). The data for 2,5-DNPX2 and 3,4-DNPX2 were taken from Ziser et al. (1995). 187 Q. Z o o E -3-D) E c E -0.4 0 0.4 0.8 1.2 1.6 2 2.4 1/ [2,5-DNPXJ (mM1) 3 -2 -E 1 --0.04 0 0.04 0.08 0.12 0.16 0.2 1/ [ONPXJ (mM'1) 16 -Q • CO i : 12 -o E "8) E c E 8 -E 4 -^ 0 0 0.4 0.8 1.2 1.6 2 1/ [3,4-DNPXJ (mM1) 0.05 0.1 0.15 1/ [PNPXJ (mM1) 0.08 0.16 0.24 0.32 0.4 1/ [PhXJ (mM1) Figure C-2. The Lineweaver-Burk plots for the hydrolysis of aryl /3-xylobiosides by the IAA-Glu78Cys mutant (1 mg enzyme = 0.049 \imol; refer to Chapter 5, section 5.3.3.1 for the Ae values for the various phenols). 188 Figure C-3. The Lineweaver-Burk plots for the hydrolysis of aryl ^ -xylobiosides by the Glu78Asp mutant (1 mg enzyme = 0.049 fxmol; refer to Chapter 5, section 5.3.3.1 for the Ae values for the various phenols). 189 -1 0 1 2 3 4 5 -0.4 0 0.4 0.8 1.2 1.6 2 1/ [2,5-DNPXJ (mM1) 1/ [3,4-DNPXJ (mM1) 1/ [PhXJ (mM1) Figure C-4. The Lineweaver-Burk plots for the hydrolysis of aryl (3-xylobiosides by the IAA-Glul72Cys mutant (1 mg enzyme = 0.049 fJmol; refer to Chapter 5, section 5.3.3.1 for the Ae values for the various phenols). 190 -2 0 2 4 6 8 10 -2 0 2 4 6 8 10 1/ [2,5-DNPX2] (mM"1) 1/ [3,4-DNPXJ (mM1) -0.2 0 0.2 0.4 0.6 0.8 1 1/ [PhXJ (mM"1) Figure C-5. The Lineweaver-Burk plots for the hydrolysis of aryl p-xylobiosides by the Glul72Asp mutant (1 mg enzyme = 0.049 \imol; refer to Chapter 5, section 5.3.3.1 for the Ae values for the various phenols). 191 Figure C-6. The Lineweaver-Burk plots for the hydrolysis of aryl P-xylobiosides by the Glul72Cys mutant (1 mg enzyme = 0.049 pmol; refer to Chapter 5, section 5.3.3.1 for the Ae values for the various phenols). 192 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 1/[ONPX2](mM"1) Figure C-7. The Lineweaver-Burk plots for the hydrolysis of aryl ^ -xylobiosides by the Glul72Gln mutant (1 mg enzyme = 0.049 pjnol; refer to Chapter 5, section 5.3.3.1 for the Ae values for the various phenols). 193 Figure C-8. The Lineweaver-Burk plots for the hydrolysis of xylan by (a) native xylanase and (b) the lAA-Glu!72Cys mutant (1 mg enzyme = 0.049 \imol). 194 Figure C-9. The l/k0bs versus 1/[S] plot for the pre-steady state hydrolysis of 2,4-DNPG by the Glu358Cys mutant of Agrobacterium P-glucosidase. 195 A P P E N D I X D: S U P P L E M E N T A R Y ESMS D A T A 20200 20400 20600 20800 21000 Molecular Mass (Da) 20200 20400 20600 20800 21000 Molecular Mass (Da) Figure D-l. ESMS analysis of Glu78Cys and native xylanase after treatment with 1200fold molar excess IAA under nondenaturing conditions (27 hrs, 36 °C, pH 10, CAPSO). Shown are the reconstructed mass spectra of (a) untreated Glu78Cys, (b) IAA-treated Glu78Cys, (c) untreated native, and (d) IAA-treated native. 196 19600 20000 20400 20800 19800 20100 20400 20700 Molecular Mass (Da) Molecular Mass (Da) 19600 20000 20400 20800 19800 20100 20400 20700 Molecular Mass (Da) Molecular Mass (Da) Figure D-2. ESMS analysis of Glu78Cys and native xylanase after treatment with MMTS (100 fold molar excess,. pH 8, 4 °C, 24 hrs). Shown are the reconstructed mass spectra of (a) untreated Glu78Cys, (b) MMTS-treated Glu78Cys, (c) untreated native, and (d) MMTS-treated native. 197 Molecular Mass (Da) 100-1 cn 75-c Inte 50-CD > 25-re QJ 0-rx 20 420 20100 20300 20500 Molecular Mass (Da) 20700 20900 cn c 3 ^ 100 75 50 25 0 20 393 CD > CD rx 20100 T 20300 20500 Molecular Mass (Da) r — — i — 20700 1 f—' 20900 g 100-03 r-75-Inter 50-CD > 25-TO CD rx 0-20 392 20100 — r 20300 20500 - 20700 Molecular Mass (Da) 20900 Figure D-3. ESMS analysis of Glu78Cys and native xylanase after treatment with IAA (10 fold molar excess) under fully denaturing conditions (7.1 M urea, pH 7.5, 40 °C, 19 hrs) followed by refolding. Shown are the reconstructed mass spectra of (a) untreated Glu78Cys, (b) refolded, IAA-treated Glu78Cys, (c) untreated native, and (d) refolded, IAA-treated native. 198 20800 Molecular Mass (Da) °S 100 § 75 c S c 50 CO 1 25 a co 0 rr 20 426 20200 20400 20600 Molecular Mass (Da) °S 100 f 75 | 50 20 396 25 CD > CO © 0-T 20200 20400 20600 Molecular Mass (Da) 20800 £ 100 CO c o CD > CD rr 20 395 75-50-25 0 20200 20400 20600 Molecular Mass (Da) 20800 Figure D-4. ESMS analysis of Glul72Cys and native xylanase after treatment with IAA (22 fold molar excess) under fully denaturing conditions (7.1 M urea, pH 7.5, 40 °C, 24 hrs) followed by refolding. Shown are the reconstructed mass spectra of (a) untreated Glul72Cys, (b) refolded, IAA-treated Glul72Cys, (c) untreated native, and(d) refolded, IAA-treated native. 100 20 395 2 to c w c CD jo CD rr 50 c 100 20 394 50 20200 20400 20600 20800 Molecular Mass (Da) 20200 20400 20600 20800 Molecular Mass (Da) 100 CO c CD .1 jo CD rx 50 20 368 100 f 50 JSM 20 373 20200 20400 20600 20600 Molecular Mass pa) 20200 20400 20600 20800 Molecular Mass (Da) Figure D-5. ESMS analysis of Glul72Cys, Glu78Cys, and native xylanase after treatment with XSSX (500 fold molar excess, pH 10, 40 °C, 24 hrs). Shown are the reconstructed mass spectra of (a) untreated Glul72Cys, (b) XSSX-treated Glul72Cys, (c) untreated native, (d) XSSX-treated native, (e) untreated Glu78Cys, and (f) XSSX-treated Glu78Cys. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0059653/manifest

Comment

Related Items