Mode of action of exoglucanases from the cellulolytic fungus trichoderma reesi

MISSING IMAGE

Material Information

Title:
Mode of action of exoglucanases from the cellulolytic fungus trichoderma reesi
Physical Description:
Book
Language:
English
Creator:
Chirico, William Joseph, 1956-
Publication Date:

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
aleph - 20498153
oclc - 13866298
System ID:
AA00026359:00001


This item is only available as the following downloads:


Full Text















MODE OF ACTION OF EXOGLUCANASES FROM THE CELLULOLYTIC FUNGUS Trichoderma reesei:
ACTIVITY ON REDUCING END-LABELED CELLOOLIGOSACCHARIDES
AND TOPOGRAPHY OF ACTIVE SITES OF 8-GLUCOSIDASE, CELLOBIOHYDROLASE I(D), AND CELLOBIOHYDROLASE II






BY






WILLIAM JOSEPH CHIRICO













A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA

1984





















This dissertation is dedicated to my parents,

Dominick and Catherine Chirico,

and to my Aunt and Uncle, Inez and Frank Montalbano.














ACKNOWLEDGMENTS

I wish to express my sincere gratitude to Dr. Ross D. Brown, Jr., for his guidance and support during the course of this work. I would also like to thank the other members of my supervisory committee, Drs. Charles M. Allen, Jr., Kim J. Angelides, Vicent Chau, Michael Roberts, and Jesse F. Gregory, for their suggestions and support. I wish to thank Drs. Phil Laipis, Mike Kilberg, and Daniel Purich for their support.

I am also grateful to Charles duMee, John Denny, Larry Weissbach, Mark Eller, and Sal Pietromonaco for friendship and many valuable discussions.

I am extremely grateful for the excellent technical assistance of Cynthia Fazenbaker and Vicki Andersen who shared the duties of scraping TLC plates. I would also like to thank Jim Parkes for incorporating J.D. Allen's depolymerase computer model in my computer file. I wish to thank Cindy Zimmerman for typing the dissertation and Nancy Shaskey for drawing the figures.

Finally, I would like to thank Crystal Willis, without whose support, encouragement, and love this dissertation could not have been written.






iii
















TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ..........................................iii

LIST OF TABLES ..........................................vii

LIST OF FIGURES ............................................

LIST OF ABBREVIATIONS AND SYMBOLS ......................xv

ABSTRACT ................................. ............. xvi

CHAPTER

1 INTRODUCTION .............................. .......

Structure of Cellulose ............................. 3
Cellulases ........................................4
Subsite Mapping of Enzymes.........................12
Assessment................................ ....... 15

2 EXPERIMENTAL PROCEDURES..........................17

Materials ........................................ 17
Enzymes .....................................17
Substrates .................... ..............17
Chemicals ......... ...................... 18
Chromatographic Supplies.................... 18
Methods .......................................... 19
Carbohydrate Determination.................. 19
Protein Determination ...................... 20
High Performance Liquid Chromatography
(HPLC) ............... ....................20
Preparation of Cellooligosaccharides ........21 Oxidation of Cellooligosaccharides..........21
Lactonization of Cellooligosaccharide
Aldonic Acids ...............................22
Reduction of Aldonic Lactones with
Sodium Borohydride...........................23
Reduction of Lactones of
Cellooligosaccharide Aldonic Acids
with Sodium Boro[3H]hydride ....... .......24
Specificity of Radiolabeling...............25
Measurement of Ra ioactivity...............26
Preparation of [ 4C-U]cellooligosaccharides.26


iv









Thin-layer Chromatography (TLC).............27
Enzymic Assays .............................. 31
Kinetic Analysis of Hydrolysis Data .........34 Determination of Kinetic Constants..........34 Evaluation of Subsite Affinities............35

3 PRqPARATION OF [1-3H]CELLOOLIGOSACCHARIDES AND
[ C-U]CELLOOLIGOSACCHARIDES .....................36

Introduction ..................................... 36
Results and Discussion .......................... 38
Preparation of Cellooligosaccharides........38 Oxidation of Cellooligosaccharides..........43
Lactonization of Cellooligosaccharide
Aldonic Acids............................... 46
Reduction of Lactones of
Cellooligosacchar de Aldonic Acids
with Sodium Boro[ H]hydride...............49
Purity and Stability of
[1- H]Cellooligosaccharides............... 55
Specificity of Radiolabeling................54
Prep ration of
[ 4C-U]Cellooligosaccharides.............. 54
Conclusions .....................................57

4 SEPARATION OF [1-3H]CELLOOLIGOSACCHARIDES
BY THIN-LAYER CHROMATOGRAPHY... ................ 58

Introduction..................................... 58
Results and Discussion ........................60
Separation of [1- H]Cellooligosaccharides...60 Extraction of [1-5H]Cellooligosaccharides...69
Conclusions...... ............... ................ 71

5 MODE OF ACTION OF -GLUCOSIDASE.................73

Introduction......................... .......... 73
Results and Discussion .................... 75
Hydrolysis of [1-3H]Cellooligosaccharides...75
Hydroly s of [1-5H]Cellooligosaccharides
and [ C-U]Cellooligosaccharides.........113
Inhibition of Methylumbelliferylglucoside
(MUG) Hydrolysis by
Cellooligosaccharides. ............ .....119
Subsite Mapping of 8-Glucosidase...........135
Conclusions ....................................144

6 MODE OF ACTION OF CELLOBIOHYDROLASE I(D) ........146

Introduction* ** ...........* .....o........... 146
Results and Discussion..................... 148
Hydro ysis of
[1- H]Cellooligosaccharides..............148


V









Hydro ysis of
[1 -H]Cellooligosaccharides and
[ 4C-U]Cellooligosaccharides.............184
Subsite Mapping of
Cellobiohydrolase I(D) ....................196
Conclusions ............................. ....... 205

7 MODE OF ACTION OF CELLOBIOHYDROLASE II..........206

Introduction ....................................206
Results and Discussion........................... 208
Hydro ysis of
[1- H]Cellooligosaccharides ...........208
Hydro ysis of
[ -H]Cellooligosaccharides and
[ 4C-U]Cellooligosaccharides.............240
Subsite Mapping of Cellobiohydrolase II....249
Conclusions.....................................258

8 GENERAL DISCUSSION.............. ..... ...........260

Hydrolysis of [1-3H]Cellooligosaccharides.......260 Synergism of Cellulolytic Enzymes...............263
Individual Roles of Cellulolytic Enzymes.......270 Methods of Subsite Mapping...................... 274
Future Experiments ......................... 281

9 SUMMARY...................

APPENDICES

A SUBSITE MAPPING OF ENZYMES ................286

B SAMPLE CALCULATIONS FOR BOND CLEAVAGE
FREQUENCIES AND INITIAL RATES OF
HYDJOLYSIS OF
[1- H]CELLOOLIGOSACCHARIDES .................304

C KINETIC EQUATIONS... .......................09

D SUBSITE AFFINITIES FOR B-GLUCOSIDASE ........310

E KINETIC CONSTANTS FOR HYDROLYSIS OF
B-LINKED DISACCHARIDES OF GLUCOSE BY
B-GLUCOSIDASE........... 9..... ......... .....311

F KINETIC CONSTANTS FOR HYDROLYSIS OF
[1- H]CELLOOLIGOSACCHARIDES BY
B-GLUCOSIDASE, CBH I(D), AND CBH II......... 312

BIBLIOGRAPHY ................................. ............ 31

BIOGRAPHICAL SKETCH.......................... ............323


vi














LIST OF TABLES

Table Page 3-1 Oxidation of Cellooligosaccharides...............44
3-2 Lactonization of Calcium Salts of Cellooligosaccharide Aldonic Acids............48

3-3 Characteristic Infrared Bands of Calcium Salts of Cellooligosaccharide Aldonic
Acids and Cellooligosaccharides Lactones.........50 3-4 Reduction of Cellgoligosaccharide Lactones with Sodium Boro[H]hydride .....................51

3-5 Preparation of [14C-U]Cellooligosaccharides......56 4-1 Resolution of Cellooligosaccharides With Different Adsorbents, Impregnants and
Solvents .........................................61

4-2 Values of the Migration Parameter, RF............68

4-3 Ext action of a Mixture of [1- H]Cellooligosaccharides From Thin-layer
Chromatographic Plates..........................70

4-4 Ext action of Individual
[1-SH]Cellooligosaccharides From Thin-layer
Chromatographic Plates..........................72

5-1 Kinetic Constants for
[1- H]Cellooligosaccharide Hydrolysis by
8-Glucosidase of T. reesei ....................... 83
5-2 Hydrolysis of [1-3H]Cellopentaose and
[' C-U]Cellopentaose by 8-Glucosidase ............114

5-3 Initial Rates for 8-Gluco~ dase Activity on
[1-3H]Cellopentaose and [ 'C-U]Cellopentaose....116 5-4 Hydrolysis of [1-3H]Cellohexaose and
[14C-U]Cellohexaose by 8 -Glucosidase .............117

5-5 Initial Rates for 8 -Glucolidase Activity on
[1-9H] Cellohexaose and ['4C-U]Cellohexaose ..... 118

vii









5-6 Inhibition of 8-Glucosidase by
Cellooligosaccharides .......................... 134
6-1 Kinetic Constants for
[1- H]Cellooligosaccharide Hydrolysis by
Cellobiohydrolase I(D) ........................... 183

6-2 H yrolysis of [1-3H]Cellopentaose and
[' C-U]Cellopentaose by Cellobiohydrolase
I(D) ................ ......................... 185
6-3 Initial Rates of Cellobiohydrolase I(D)
A ivity on [1- H]Cellopentaose and
[ C-U]Cellopentaose ...................... ...... 189
6-4 Hydrolysis of [1-3H]Cellohexaose and
[ C-U]Cellohexaose by Cellobiohydrolase I(D)...191

6-5 Initial Rates of Hydrolysis of
Cel]obiohydrolase I(D) Ajqivity on
[1- H]Cellohexaose and [ 'C-U]Cellohexaose ......195
6-6 Subsite Affinities of CBH I(D) Calculated
Using Values of V ~K and Bond Cleavage
Frequencies from T- H Cellooligosaccharides.....197

7-1 Kinetic Constants for
[1- H]Cellooligosaccharide Hydrolysis by
Cellobiohydrolase II .............. ............. 239

7-2 Hy rolysis of [1-3H]Cellopentaose and
[' C-U]Cellopentaose by Cellobiohydrolase II .....242
7-3 Initial Rates o4 Cellobiohydrolase II
A ivity on [1- H]Cellopentaose and
[ C-U]Cellopentaose ............................243

7-4 H yrolysis of [1-3H]Cellohexaose and
[ 4C-U]Cellohexaose by Cellobiohydrolase II .....245
7-5 Initial Rates o4 Cellobiohydrolase II
Agivity on [1- H]Cellohexaose and
[' C-U]Cellohexaose ....................... ...... 250

7-6 Subsite Affinities of CBH II Calculated
Using Values of V ~K and Bond Cleavage
Frequencies from [-2HlCellooligosaccharides....252

B-1 Distribution of [1-3H]Cellooligosaccharides
from Hydrolysis of [1-5H]Cellotetraose by
CBH II(D).......................................305


viii









B-2 Product Ratios for [1-3H]Cellotetraose
Hydrolysis by CBH I(D)..........................306

B-3 Concentration of [1-3H Cellooligosaccharides
from Hydrolysis of [1- H]Cellotetraose by
CBH I(D)...................... .... .... .......... 308














































ix














LIST OF FIGURES

Figure Page

3-1 Synthesis of [1-5H]cellooligosaccharides.........40

3-2 Elution pattern from Bio-Gel P-2 column chromatography of cellooligosaccharides
generated from acid hydrolysis of cellulose......42

4-1 Thin-layer chromatographic pattern of cellooligosaccharides ............................65

4-2 Thig-layer chromatographic pattern of [1- H]cellooligosaccharides ......................67

5-1 Time course hydrolysis of [1-3H]cellobiose by 8-glucosidase................ .............. .. 78

5-2 Bond c eavage frequency plot for hydrolysis of [1-H]cellobiose by 0-glucosidase.............80

5-3 Lineweaver-Burk a d Eadie-Hofstee plots for hydrolysis of [1- H]cellobiose by
8-glucosidase................................... 82

5-4 Time course hydrolysis of [1-3H]cellotriose by B-glucosidase.................................85

5-5 Bond c eavage frequency plot for hydrolysis
of [1- H]cellotriose by 8-glucosidase............88

5-6 Lineweaver-Burk aqd Eadie-Hofstee plots for
hydrolysis of [1- H]cellotriose by
S-glucosidase................................... 90
5-7 Time course hydrolysis of [1-3H]cellotetraose
by 8-glucosidase ................................. 92
5-8 Bond c eavage frequency plot for hydrolysis
of [1- H]cellotetraose by -glucosidase..........95
5-9 Lineweaver-Burk and Eadie-Hofstee plots for
hydrolysis of [1-3H]cellotetraose by
8 -glucosidase ................................... 97



X









5-10 Time course hydrolysis of [1-3H]cellopentaose
by 8-glucosidase ............................... 99

5-11 Bond cleavage frequency plot for hydrolysis
of [1-9H]cellopentaose by 6-glucosidase.........101

5-12 Lineweaver-Burk agd Eadie-Hofstee plots for
hydrolysis of [1- H]cellopentaose by
a-glucosidase...................................104

5-13 Time course hydrolysis of [1-3H]cellohexaose
by 8-glucosidase .......................... .....106

5-14 Bond c eavage frequency plot for hydrolysis
of [1--H]cellohexaose by s-glucosidase..........108

5-15 Lineweaver-Burk agd Eadie-Hofstee plots for
hydrolysis of [1- H]cellohexaose by
8-glucosidase...................................111

5-16 Inhibition of s-glucosidase by cellobiose
using 4-methylumbelliferyl-s -D-glucopyranoside
as the substrate................................121

5-17 Inhibition of 8-glucosidase by cellotriose
using 4-methylumbelliferyl- -D-glucopyranoside
as the substrate ............................123

5-18 Inhibition of 8-glucosidase by cellotetraose
using 4-methylumbelliferyl-a-D-glucopyranoside
as the substrate................................125

5-19 Inhibition of B-glucosidase by cellopentaose
using 4-methylumbelliferyl-8 -D-glucopyranoside
as the substrate.................................127

5-20 Inhibition of a-glucosidase by cellohexaose
using 4-methylumbelliferyl-8--D-glucopyranoside
as the substrate ......... ..... .......... ..... 129

5-21 Dixon replots of inhibition of -glucosidase
by cellobiose, cellotriose and cellotetraose....131

5-22 Dixon replots of inhibition of -glucosidase
by cellopentaose and cellohexaose ..............133

5-23 Subsite map for -glucosidase constructed
using values of V x/Km for
cellooligosacchar es...........................138

5-24 Subsite map for -glucosidase constructed
using values of inhibition constants for
cellooligosaccharides................ ............141


xi









6-1 Time course hydrolysis of [1-3H]cellotriose
by CBH I(D) ....................................151
6-2 Bond c eavage frequency plot for hydrolysis
of [1- H]cellotriose by CBH I(D)................ 153
6-3 Effect of [1-3H]cellotriose concentration on
product ratios..................................155
6-4 Ead e-Hofstee plot for hydrolysis of
[1-H]cellotriose hydrolysis by CBH I(D)........157
6-5 Time course hydrolysis of [1-3H]cellotetraose
by CBH I(D) ....................................161
6-6 Bond c eavage frequency plot for hydrolysis
of [1- H]cellotetraose by CBH I(D)..............163
6-7 Lineweaver-Burk agd Eadie-Hofstee plots for
hydrolysis of [1- H]cellotetraose by CBHI(D)....165
6-8 Time course hydrolysis of [1-3H]cellopentaose
by CBH I(D) ...................................167
6-9 Bond c eavage frequency plot for hydrolysis
of [1-iH]cellopentaose by CBH I(D)..............170
6-10 Lineweaver-Burk agd Eadie-Hofstee plots for
hydrolysis of [1- H]cellopentaose by CBH I(D)...172
6-11 Time course hydrolysis of [1-3H]cellohexaose
by CBH I(D) .....................................174
6-12 Bond c eavage frequency plot for hydrolysis
of [1-'H]cellohexaose by CBH I(D)...............176
6-13 Lineweaver-Burk agd Eadie-Hofstee plots for
hydrolysis of [1- H]cellohexaose by CBH I(D)....179
6-14 Boni cleavage frequencies of
[1- H]cellooligosaccharide hydrolysis by
CBH I(D) .... ........................... 0 0 181
6-15 Possible types of repetitive attack of
Cj I(D) with [1-3H]cellopentaose and
[ C-U]cellopentaose .............. ........... ..188
6-16 Possible types of repetitive attack of
CB I(D) with [1-3H]cellohexaose and
[4C-U]cellohexaose ............................ 193



xii








6-17 Subsite map for CBH I(D) constructed using
values gf V, /K and bond cleavage frequencies
for [(1- H]celoo igosaccharides.................200

6-18 Subsite map for CBH I(D) constructed using
values of K., Vm,_ and bond cleavage
frequencies for x- H]cellooligosaccharides .....203

7-1 Time course hydrolysis of [1-3H]cellotriose
by CBH II.......................................210

7-2 Bond c eavage frequency plot for hydrolysis
of [1- H]cellotriose by CBH II ...................212

7-3 Lineweaver-Burk agd Eadie-Hofstee plots for
hydrolysis of [1- H]cellotriose by CBH II........214

7-4 Time course hydrolysis of [1-3H]cellotetraose
by CBH II...................................... 217

7-5 Bond cleavage frequency plot for hydrolysis
of [1- H]cellotetraose by CBH II................ 219
7-6 Lineweaver-Burk agd Eadie-Hofstee plots for
hydrolysis of [1- H]cellotetroase by CBH II.....221

7-7 Time course hydrolysis of [1-3H]cellopentaose
by CBH II.......................................223

7-8 Bond c eavage frequency plot for hydrolysis
of [1-H]cellopentaose by CBH II .................226

7-9 Lineweaver-Burk aqd Eadie-Hofstee plots for
hydrolysis of [1- H]cellopentaose by CBH II.....228

7-10 Time course hydrolysis of [1-3H]cellohexaose
by CBH II...................................... 230
7-11 Bond cleavage frequency plot for hydrolysis
of [1- H]cellohexaose by CBH II ................. 232

7-12 Lineweaver-Burk agd Eadie-Hofstee plots for
hydrolysis of [1- H]cellohexaose by CBH II......235
7-13 Bon cleavage frequencies of
[1- H]cellooligosaccharide hydrolysis by
CBH II.................*..................27

7-14 Possible types of repetitive attack of
CBH II with [1-3H]cellohexaose and
[14C-U]cellohexaose ................... .........247




xiii









7-15 Subsite map for CBH II constructed using
values gf V a/K and bond cleavage frequencies
for [1- H]ceiTooTigosaccharides.................254

7-16 Subsite map for CBH II constructed using
values of K., V and bond cleavage
frequencies for H]cellooligosaccharides.....257

8-1 Structure of cellulose I and cellulose II.......266

8-2 Orientation of glycosidic bonds in cellulose ....269

A-i Binding modes of cellotriose on a
hypothetical four-subsite enzyme.................288

A-2 Combination of Vmax K values and bond cleavage
frequencies for [1- H cellotetraose and
[1-H]cellotriose used to calculate subsite
affinities......................................301


































xiv














LIST OF ABBREVIATIONS AND SYMBOLS

A Substrate CBH Cellobiohydrolase CM-Cellulose Carboxymethylcellulose GC Cellooligosaccharide of chain length i HPLC High performance liquid chromatography kcat Turnover number k+2 Hydrolytic rate coefficient Ki Inhibition constant Kint Microscopic dissociaton constant for a binding mode in which the entire binding region is occupied Km Michaelis constant K Dissociation constant for substrate MOG 4-Methylumbelliferyl-8-Dglucopyranoside
MeUmb 8(Glcj)n 4-Methylumbelliferyl-8-D-glycosides from cellobiose (n=2) to cellohexaose (n=6)
TLC Thin-layer chromatography V Measured velocity Vmax Maximum velocity























xv














Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

MODE OF ACTION OF EXOGLUCANASES FROM THE CELLULOLYTIC FUNGUS Trichoderma reesei:
ACTIVITY ON REDUCING END-LABELED CELLOOLIGOSACCHARIDES
AND TOPOGRAPHY OF ACTIVE SITES OF 3-GLUCOSIDASE, CELLOBIOHYDROLASE I(D), AND CELLOBIOHYDROLASE II BY

WILLIAM JOSEPH CHIRICO

December, 1984

Chairman: Ross D. Brown, Jr., Ph.D. Major Department: Biochemistry and Molecular Biology

The activity of purified 8-glucosidase, cellobiohydrolase I(D) and cellobiohydrolase II from Trichoderma reesei was studied using [1-3H1cellooligosaccharides. Reducing end-labeled cellooligosaccharides of high specific radioactivity were synthesized by reducing lactones of cellooligosaccharide aldonic acids with sodium boro[3H]hydride. A thin-layer chromatographic method for separating [1-3H]cellooligosaccharides with high resolution and quantitatively extracting them from silica gel was developed.

The 8-glucosidase was shown to bind (1-3H]cellooligosaccharides in one productive mode and sequentially remove glucosyl residues from the nonreducing end. The B-glucosidase exhibited little tendency to repetitively attack [1-3H]cellooligosaccharides. Values of Vmax for hydrolysis



xvi









of [1-3H]cellooligosaccharides remained constant as chain length increased, whereas corresponding values of K, decreased. Values of inhibition constants determined for cellooligosaccharide inhibition of 4-methylumbelliferyl-8-Dglucopyranoside hydrolysis were similar to corresponding values of Km. The active site of 8-glucosidase was shown to comprise primarily three subsites, of which subsite 1 contributes the greatest proportion of binding energy.

Cellobiohydrolase I(D) was shown to bind [1-3H]cellooligosaccharides in more than one productive mode. Bond cleavage frequency analysis indicated that cellobiohydrolase I(D) preferentially hydrolyzes glycosidic bonds at the reducing end. Product distribution resulting from activity on [14C-U]cellooliogosaccharides and [1-3H]cellooligosaccharides suggests that cellobiohydrolase I(D) repetitively attacks cellooligosaccharides from the reducing end. Values of Vmax/Km for [1-5H]cellooligosaccharide hydrolysis increase with increasing chain length. Subsite mapping procedures revealed that the active site of cellobiohydrolase I(D) comprises seven subsites with catalytic groups located between subsites 5 and 6.

Cellobiohydrolase II was also shown to bind

[1-3H]cellooligosaccharides in more than one productive mode; however, glycosidic bonds at the nonreducing end are preferentially hydrolyzed. Product distribution resulting from hydrolysis of [14C-U]cellooligosaccharides and [1-3H]cellooligosaccharides suggests that cellobiohydrolase


xvii









II repetitively attacks cellooligosaccharides from the nonreducing end. Values of Vmax/Km for [1-3H]cellooligosaccharide hydrolysis increase with increasing chain length. Subsite mapping procedures show that the substrate binding region of cellobiohydrolase II comprises seven subsites with catalytic groups located between subsites 3 and 4.

The results are discussed in terms of the individual

roles of exoglucanases during the saccharification of cellulose. A mechanism by which cellobiohydrolase I(D) and cellobiohydrolase II may synergistically degrade crystalline cellulose is proposed.






























xviii














CHAPTER 1
INTRODUCTION

The utilization of cellulose, the world's most abundant renewable carbon source, for fuel, food and chemical feedstocks has received increasing attention as depletion of nonrenewable resources continues. The amount of photosynthetically fixed carbon dioxide which is converted to cellulose has been estimated by Ghose (1) at 1011 metric tons per year. Glucose, a product of cellulose hydrolysis, can serve as a substrate for fermentations to produce single-cell protein, ethanol and methane (2). Cellulose can be converted to soluble products by acid or enzymic hydrolysis. Although technology for acid hydrolysis exists, degradation of products, interaction of acid with noncellulosic substances in natural sources of cellulose, corrosion of equipment and high capital costs argue for development of alternative methods of conversion (3). Enzymic hydrolysis of cellulose has the advantage of efficiency and specificity; however, it is the most expensive process. Lignin and hemicelluloses, which are constituents of natural sources of cellulose, protect the substrate from enzymic hydrolysis and limit conversion to 50 percent (2). Removal of lignin and hemicelluloses from native cellulose or addition of hemicellulases is necessary to obtain higher yields of glucose (2).



1





2



An excellent source of cellulolytic enzymes is the

fungus Trichoderma reesei QM 9414, which possesses the complete array of enzymes required for the conversion of cellulose to glucose (4). The component enzymes of the cellulase system act synergistically, in order to convert native, crystalline cellulose to soluble products (5). Since the product of one enzyme may serve as a substrate for another, purification and characterization of each component is a prerequisite for a clear description of the enzymic hydrolysis of cellulose. Although the extracellular, depolymerizing enzymes (6,7) and 8-glucosidase (8) of T. reesei have been purified and partially characterized, little quantitative information is available regarding the affinity of cellulolytic enzymes for cellulose (or cellooligosaccharides). In addition, the mechanism by which cellulolytic enzymes act synergistically and the topography of their active sites remain unknown. Since cellooligosaccharides are symmetric, the precise glycosidic bonds cleaved by cellulolytic enzymes have not been determined. Furthermore, the direction from which exo-glucanases remove glucosyl or cellobiosyl residues from cellulosic substrates has not been determined.

In this report, the synthesis of reducing end-labeled cellooligosaccharides and a quantitative method for separating them using thin-layer chromatography are described. The activity of cellobiohydrolase I(D) [CBH I(D)], cellobiohydrolase II (CBH II) and 8-glucosidase of T. reesei on





3



these asymmetrically labeled substrates provides information regarding substrate affinities, action patterns, topography of active sites and the direction from which substrates are degraded.

Structure of Cellulose

Cellulose is a linear polymer of B-1,4-D-glucosyl residues. Four crystalline forms of cellulose (I-IV) have been recognized on the basis of infrared spectra and X-ray diffraction studies (9). The chains in each structure have approximately the same backbone conformation, with a unit of cellobiose repeating every 10.3 A (10). The distinguishing feature of the structures is the arrangement of adjacent chains. Cellulose I (native cellulose) is found in plant cell walls. When native cellulose is dissolved and recrystallized, it is converted irreversibly to cellulose II, a form of cellulose more stable than cellulose I. Cellulose III and IV can be produced from cellulose I or II by treatment with liquid ammonia and hot glycerol, respectively (9).

The most crystalline, native cellulose is found in the cell walls of the alga Valonia ventricosa. The X-ray diffraction patterns of cellulose fibers isolated from cell walls of Valonia ventricosa indicate that intramolecular hydrogen bonds are located on both sides of the glycosidic linkage: one hydrogen bond between 02'-H---06 and another between 03-H---05' (9). The 06-H hydroxyl is bonded to 03 of an adjacent chain. Thus, native cellulose is seen as a series of hydrogen-bonded sheets of parallel chains which





4


are held together by van der Waal forces between hydrophobic faces of sugar rings. In contrast, X-ray patterns of cellulose II indicate that chains are arranged in an antiparallel fashion. Cellulose II also has intramolecular hydrogen bonds between 02'-H---06 and another between 03-H---05'. The orientation of -CH20H groups allows formation of intermolecular hydrogen bonds between 06-H---03, 06-H---02 and 02-H---02'. The additional intermolecular hydrogen bonding accounts for the increased stability of cellulose II.

Recently, Atalla and VanderHart (11) compared the

solid-state carbon-13 nuclear magnetic resonance spectra of various native celluloses. The spectra were consistent with the presence of two distinct forms of cellulose, I and Ig, in each sample. They estimated that the relative amounts of I. as a fraction of the total are Acetobacter xylinium = Valonia ventricosa > cotton = ramie > regenerated cellulose I. Thus, Is is dominant in bacterial and algal cellulose, whereas Ig is dominant in celluloses from higher plants. Carbon-13 NMR spectra of cellulose II closely resembled that of Ig.

Cellulases

The cellulase system comprises 1,4-8-D-glucan-4-glucanohydrolases (EC 3.2.1.4), 1,4-8-D-glucan cellobiohydrolases (EC 3.2.1.91) and 8-glucosidases (EC 3.2.1.21), which together act sequentially and cooperatively to convert native, crystalline cellulose to oligosaccharides and glucose. Endoglucanases randomly attack internal glycosidic





5



bonds of cellulose producing chain ends and soluble oligosaccharides. Cellobiohydrolases cleave cellobiosyl residues from ends of cellulose chains. Cellobiose, an inhibitor of the depolymerizing enzymes, and oligosaccharides are converted to glucose by the 8-glucosidase. Component enzymes have been purified and characterized from true cellulolytic organisms, such as Trichoderma viride (5,12-16), Trichoderma reesei (6,8,17), Trichoderma koningii (18,19) and Irpex lacteus (20).

Endo-1,4,-8-D-glucanases catalyze the hydrolysis of

internal glycosidic bonds of cellulose providing chain ends from which cellobiosyl residues are removed by exo-cellobiohydrolases (5). The most frequently used technique for evaluating endoglucanase activity is measurement of the reduction of viscosity of a carboxymethylcellulose (CMcellulose) solution (21). Endoglucanases are differentiated from exo-cellobiohydrolases according to the anomeric configuration and identity of products from hydrolysis of polymeric substrates and cellooligosaccharides (22). Endoglucanases proceed with retention of configuration and form a variety of products, whereas exo-glucanases proceed with inversion of configuration and cleave almost exclusively cellobiosyl residues from the ends of substrates (22).

Gritzali (6) reported the purification of a single

endoglucanase, which has a molecular weight of 45,200 and a neutral carbohydrate content of 14.2 percent (by weight), from extracellular filtrates of T. reesei. The





6



endoglucanase accounts for 15 percent (by weight) of the extracellular protein (5). Purification of endoglucanases from T. reesei has also been reported by Hakansson et al.

(17), Ladisch et al. (23) and duMee (24). Bhikhabhai and Pettersson (25) recently reported the partial sequence of an endoglucanase from T. reesei. Cyanogen bromide peptides of the endoglucanase showed significant homology with the amino acid sequence of CBH I. Comparison of carbon-13 NMR spectra of three endoglucanases isolated by duMee indicates that oligosaccharides attached to endoglucanases are similar if not identical to those of CBH I(D) and CBH II of T. reesei

(24).

The rate of hydrolysis of cellooligosaccharides by the endoglucanase from T. reesei increases from cellotetraose to cellopentaose; however, this trend does not extend to cellohexaose (6). Cellotetraose is hydrolyzed to glucose, cellobiose and cellotriose. Cellotriose is resistant to further degradation and its large concentration relative to glucose is indicative of transglycosylation by the endoglucanase. A variety of products are formed from cellopentaose and cellohexaose. The endoglucanase readily catalyzes hydrolysis of CM-cellulose and phosphoric acid-swollen cellulose. Microcrystalline cellulose is resistant to attack, whereas filter paper discs are degraded to short fibers.

The release of cellobiose from ends of cellulose chains is due primarily to the hydrolytic action of 1,4-a-D-glucan cellobiohydrolases. Other names for this enzyme include





7


exo-glucanase and exo-cellulase (20). The activity of cellobiohydrolases is determined by measuring the reducing sugar released from a suspension of swollen or crystalline cellulose. Since each enzyme in the cellulase system produces reducing sugar, highly purified preparations are required for enzymic characterization. Cellobiohydrolases exhibit little activity toward CM-cellulose as measured by change in viscosity (5).

Two immunologically distinct cellobiohydrolases [CBH

I(D) and CBH II] have been purified from extracellular filtrates of T. reesei grown on cellulose or incubated with sophorose (6). Mandels et al. (26) discovered that sophorose is a potent inducer of cellulase in T. viride QM 6a. Cellobiohydrolase I(D) has a molecular weight of 53,200

(6), contains 5.9 percent (by weight) of neutral carbohydrate (24) and is isoelectric at pH 3.95 (7), whereas the corresponding values for CBH II are 54,700 (6), 18.1 percent

(24) and pH 4.9 (6), respectively. Cellobiohydrolase I(D) and CBH II constitute approximately 60 and 25 percent (by weight), respectively, of the extracellular protein produced by this fungus (5). Recently, duMee (24) reported that CBH I(D) contains 6 or 7 oligosaccharide chains linked through mannose to the protein, whereas CBH II contains approximately 25 oligosaccharide chains also linked through mannose to the protein. Complete amino acid sequence analysis of CBH I revealed that most of the neutral carbohydrate is located in a short region near the carboxy terminus (27).





8



The sites of attachment of three glucosamine residues were also determined. Fagerstam and Pettersson (7) reported that no homology exists within the first 20 amino acid residues of CBH I(D) and CBH II, although both enyzmes have a pyroglutamyl residue at the N-terminus of the polypeptide chain. The molecular cloning and characterization of the gene encoding CBH I of T. reesei have been reported by Shoemaker et al. (28). The nucleotide sequence shows that the gene contains two introns and that a signal peptide is located at the N-terminus.

Enzymic properties of cellobiohydrolases purified from extracellular filtrates of T. reesei QM 9414 on insoluble and soluble substrates have been reported (6). Cellobiohydrolase II cleaves cellooligosaccharides approximately 10fold and polymeric substrates 2-fold more rapidly than CBH I(D) (6). Cellobiohydrolase I(D) and CBH II are optimally active at pH 5.2-5.6 and 4.9, respectively (6).

Phosphoric acid-swollen cellulose is hydrolyzed more rapidly than microcrystalline cellulose by CBH I(D) or CBH II (6). Cellobiohydrolase II produces exclusively cellobiose from phosphoric acid-swollen cellulose, whereas CBH I(D) yields some glucose in addition to cellobiose. Carboxymethylcellulose is mostly resistant to attack by either enzyme. The two cellobiohyrolases show strong synergistic activity in the hydrolysis of crystalline cellulose, thereby demonstrating their essential role in the degradation of cellulose (7).





9


The action patterns of CBH I(D) and CBH II have been investigated using high performance liquid chromatographic (HPLC) analysis of products formed during enzymic hydrolysis of cellooligosaccharides (6). Cellobiohydrolase I(D) catalyzes the hydrolysis of cellotriose forming glucose and cellobiose, whereas CBH II has little activity toward cellotriose. As the chain length of oligosaccharide substrate increases, CBH I(D) produces cellobiose and increasing proportions of glucose and cellotriose. In contrast, CBH II yields exclusively cellobiose from cellotetraose and a mixture of cellobiose and cellotriose from cellopentaose and cellohexaose. Each cellobiohydrolase yields cellobiose from hydrolysis of cellohexaose without concomitant formation of cellotetraose, indicating a sequential cleavage of cellobiosyl residues. Cellohexaose is hydrolyzed 2-fold more rapidly than cellotetraose or cellopentaose by CBH I(D) or CBH II.

Hsu et al. (29), using a cellobiohydrolase isolated

from T. reesei QM 9414, reported that Km's for cellotriose and cellotetraose are 0.2 and 0.08 mM, respectively. Cellotetraose was hydrolyzed 27-fold faster than cellotriose. Since low sensitivity of HPLC analysis of cellooligosaccharides precludes initial rate studies, Hsu et al. (29) estimated kinetic constants from analysis of the entire time course of the reaction.

Recently, van Tilbeurgh et al. (30) reported the action patterns of CBH I(D) on a series of 4-methylumbelliferyl





10



glycosides derived from cellooligosaccharides. Cellobiohydrolase I(D) preferentially cleaves the glycosyl bond at the reducing end of the modified cellooligosaccharides. Whether the modification influences the action pattern of CBH I(D) is not known.

Cellobiohydrolases are also characterized by their affinities for inhibitors and substrates. Inhibition of depolymerizing enzymes by cellobiose is an important regulatory feature of the cellulase system. Inhibition constants of 2.1 and 0.2 mM for glucose and cellobiose, respectively, have been reported using a cellobiohydrolase from T. reesei

(29). Using 4-methylumbelliferylcellobioside as the substrate, Brown and Greenberg (31) determined that glucose, cellobiose, cellotriose, cellotetraose, cellopentaose and cellohexaose inhibit and bind to one site on CBH I(D) with inhibition constants of 0.476 M, 90.4 iM, 28.4 iM, 14.1 iM, 14.6 UM and 7.6 PM, respectively. Similar information was not obtained for CBH II, because this enzyme exhibits little activity toward 4-methylumbelliferylcellobioside.

The 8-glucosidases catalyze the hydrolysis of aryl-8-Dglucopyranosides and 8-linked glucose oligomers as well as transfer of glucosyl residues (32). Although the hydrolysis of dimers, trimers and tetramers of glucose is catalyzed by glucosidases and exo-glucanases, glucosidases are often reported to act more rapidly on short oligomers, whereas exo-glucanases prefer longer oligomers (22). Glucosidases proceed with retention of configuration, whereas





11


exo-glucanases proceed with inversion (22). The role of $glucosidase within the complement of cellulase enzymes is to cleave cellobiose, an inhibitor of the depolymerizing enzymes, to glucose (5,33).

Glucosidases have been purified from extracellular

filtrates of several cellulolytic fungi, such as T. viride (16,34,35), T. koningii (19) and T. reesei (8). The Bglucosidase purified from T. reesei (8) has a molecular weight of 74,600, contains 0.7 percent (by weight) carbohydrate and is isoelectric at pH 8.5. The 8-glucosidase constitutes about 0.4 percent (by weight) of the extracellular protein produced by this fungus. Transglucosylation reactions and hydrolysis of cellobiose, 2-nitrophenylB-D-glucopyranoside and 4-methylumbelliferyl-8-D-gluconpyranoside are catalyzed by the 6-glucosidase. The enzyme has an unusually high affinity for its product, D-glucose, and is strongly inhibited by D-glucono-1,5-lactone, nojirimycin and 1-deoxynojirimycin.

Most cellulolytic fungi have similar cellulase systems containing one to several 8-glucosidases, endo-a-glucanases and exo-a-glucanases that act synergistically to hydrolyze insoluble cellulose (36). Synergism resulting from the concerted action of endoglucanases and exoglucanases has been explained by the production by endoglucanases of free polymer chain ends which serve as substrates for exoglucanases. Recently, strong exo-exo synergism has been reported for CBH I(D) and CBH II from T. reesei acting on crystalline





12


cellulose (7). Fagerstam and Pettersson (7) offered two possible explanations for exo-exo synergism: (1) the synergism may reflect a structural asymmetry in the substrate or (2) a change in the mode of action resulting from the formation of a complex between the two enzymes. However, no evidence for complex formation has been reported.

Subsite Mapping of Enzymes
Studies of the action pattern of polysaccharide hydrolases using their natural polymeric substrates led to the concept that glycosyl residues distant from the point of catalysis interact with the enzyme (37). Thus, the substrate binding region of the enzyme is considered to be an array of tandem subsites, each of which interacts independently with a specific glycosyl residue and with a certain affinity. The location of catalytic groups between two of the subsites, the number of subsites, the subsite affinity and the change in the unitary binding free energy at a subsite are characteristic features of polysaccharide hydrolases. Subsite models have been described to account for the enzymic properties of proteinases (38), nucleases (39) and carbohydrases (40-43).

The literature offers three methods to determine

experimentally the number of subsites comprising the binding region, the energetics of interaction of each subsite with a glycosyl residue and the hydrolytic rate coefficients (37,43-45). A synopsis of equations most relevant to each





13



subsite-mapping procedure is found in Appendix A. Presented below is a brief desription of each method.

Hiromi and coworkers (44,46) applied the subsite model to exo-enzymes by using the first-order rate constant (Vmax/Km) and Km for each of a series of oligosaccharides. A substrate can be bound to an exoenzyme in a variety of binding modes; however, only one mode is productive. Since hydrolytic rate coefficients for each oligosaccharide are assumed equal, Vmax/Km values are directly proportional to the association constant of the single productive complex. Michaelis constants are assumed to approximate association constants. Thus, the ratios of Vmax/Km values for n-mer and n+1-mer substrates permit the estimation of the n+1-th subsite affinity. Using this method, the substrate binding region of glucoamylase from Rhizopus sp. was shown to consist of seven subsites, of which subsite II contributes the greatest proportion of the binding energy. Recently, Koyama et al. (47) applied Hiromi's method to glucoamylase from Aspergillus saito. This enzyme was also shown to consist of seven subsites, of which subsite II is the major contributor to the binding energy.

Although the method of Hiromi (46) is suitable for

mapping exo-enzymes having one productive complex, it is not applicable to enzymes, such as endoenzymes, having more than one productive complex. Thoma and coworkers (43,48) have proposed a method to evaluate subsite affinities of endoenzymes from product distribution ratios of end-labeled





14


oligosaccharides. Enzymic processes are separated into microscopic and macroscopic events. A microscopic coefficient characterizes one particular binding mode of a substrate. Each binding mode has a microscopic association constant, and each productive complex has a microscopic hydrolytic rate coefficient. Although microscopic constants are difficult to measure, they are related to measurable macroscopic parameters, which are a function of all binding modes of the substrate. The experimentally accessible parameters that are useful in subsite mapping are Km, Vmax, first order rate constant and bond cleavage frequencies measured as a function of chain length. Ratios of bond cleavage frequencies can be used to measure the size of the binding region, to locate the catalytic groups between two of the subsites and to calculate the apparent binding energies for some of the subsites. Allen and Thoma (49) using bond cleavage frequency analysis reported that the substrate binding region of Bacillus amyloliquefaciens aamylase is composed of 10 subsites with catalytic groups located between subsites VI and VII.

Suganuma et al. (45) have proposed a method for evaluating subsite affinities that retains the advantages of Hiromi's and Thoma's methods. Suganuma et al. (45) reported that the substrate binding region of Taka-amylase A consists of 9 subsites with catalytic groups located between subsites

4 and 5. The subsite binding region of an endo-1,4-0xylanase from Cryptococcus albidus was examined using the





15



method of Suganuma et al. (45). The enzyme was shown to consist of 4 subsites with the catalytic groups located in the center (42).

Assessment

The mechanisms by which cellulolytic enzymes attack

cellulose remain poorly understood, because the substrate, cellulose, is insoluble and heterogeneous. To define the mode of attack of cellulolytic enzymes, soluble substrates, such as cellooligosaccharides, should be used in enzymic studies. Although previous studies of cellulolytic enzymes using cellooligosaccharides as substrates provide information regarding rates of hydrolysis, quantitative estimates of the affinity of the enzymes for these substrates were not possible due to low sensitivity of the assays. In addition, the symmetric nature of cellooligosaccharides precludes identification of the precise glycosidic bonds cleaved by the enzymes.

In this report, the synthesis of reducing end-labeled cellooligosaccharides and a thin-layer chromatographic method for separating and quantitating these substrates are described. Reducing end-labeled cellooligosaccharides were used to identify the precise glycosidic bond(s) cleaved by 0-glucosidase, CBH I(D) and CBH II of T. reesei. These substrates were also used in initial rate studies to determine the affinity and maximal rates of hydrolysis by cellulolytic enzymes. Cellooligosaccharides uniformly labeled with carbon-14 were used in combination with reducing





16



end-labeled cellooligosaccharides to determine the direction from which cellulolytic enzymes attack. A subsite map for each enzyme was constructed from bond cleavage frequencies, Km and Vmax for the series of cellooligosaccharides. The results are discussed in terms of the role of each enzyme in the conversion of cellulose to glucose. A mechanism by which CBH I(D) and CBH II may exert exo-exo synergism is proposed.













CHAPTER 2
EXPERIMENTAL PROCEDURES

Materials

Enzymes

Cellobiohydrolase I(D) and CBH II, from extracellular culture filtrates of T. reesei QM 9414, were purified by Gritzali (6). Extracellular 8-glucosidase was purified from T. reesei as described previously (8).

Statzyme-Glucose 16 was purchased from Worthington Biochemical Corporation, Freehold, NJ. Substrates

Cellobiose was obtained from Schwartz/Mann, Orangeburg, NY.

Sophorose was obtained from Adams Chemical Co., Round Lake, IL.

Laminaribiose was prepared by hydrolysis of laminarin and separated on a column of Bio-Gel P-2 by B. Greenberg in this laboratory.

The 4-methylumbelliferyl-a8-D-glucopyranoside (lot 8650263), 2-nitrophenyl-6-D-glucopyranoside (lot 88C-5039), pnitrophenyl-8-D-galactopyranoside (lot 58C-5063), 2.-nitrophenyl-8-D-xylopyranoside (lot 117C-0369), -nitrophenyl-aD-glucopyranoside (lot 59B-0100), j-nitrophenyl-l-thio-8-Dglucopyranoside and gentiobiose (lot 35C-0173) were obtained from Sigma Chemical Company, St. Louis, MO.


17





18


Chemicals

Sodium boro[3H]hydride (18.1 Ci/mmol, lot 25 and 12.1 Ci/mmol, lot 26) was obtained from Amersham, Arlington Heights, IL. *The [14C-U]cellulose (Nicotiana Tobacum L., 34.5 VCi/mg, lot 1396-151) and Aquasol were purchased from New England Nuclear, Boston, MA. Cellulose (SIGMACELL Type 100, lot 129C-0079) was obtained from Sigma Chemical Co., St. Louis, MO. Dextrose was obtained from the National Bureau of Standards, Washington, DC. Bromine (lot 22460) and Celite 545 were obtained from J.T. Baker Chemical Co., Phillipsburg, NJ. Calcium cellobionate (lot 101207) was obtained from ICN Pharmaceuticals, Inc., Cleveland, OH. Lactobionic acid (calcium salt) was obtained from Nutritional Biochemicals Corp., Cleveland, OH. The 2propanol (certified ACS), acetonitrile (HPLC grade), ethyl acetate (certified ACS), methanol (HPLC grade), dimethyl sulfoxide (certified ACS), glucono-1,5-lactone (lot 793699) and ~-anisaldehyde were obtained from Fisher Scientific, Fair Lawn, NJ.

All other chemicals were reagent grade. Chromatographic Supplies
The LK5 (5 x 20 cm), LK6 (5 x 20 cm) and LK5D (20 x 20 cm, 250 um) (lot 002431) Silica Gel TLC plates, Whatman 3 MM paper and Whatman Partisil PXS 10/25 PAC column (bonded cyano-amino type, polar phase) were obtained from Whatman Chemical Separations Division, Clifton, NJ.





19



Silica Gel IB2 Baker-flex (20 x 20 cm) TLC plates were purchased from J.T. Baker Chemical Co., Phillipsburg, NJ.

Cellulose (20 x 20 cm) (lot 6214) TLC plates were obtained from Eastman Kodak Company, Rochester, NY.

Polyamide Layer Sheets (15 x 15 cm, 50 um) were obtained from Accurate Chemical and Scientific Corp., Hicksville, NY.

Amberlite IRA-45 (-OH) Anion Exchange Resin was obtained from Sigma Chemical Co., St. Louis, MO.

Amberlite IR-120 (H+) Cation Exchange Resin was obtained from Fisher Scientific, Fair Lawn, NJ.

Amberlite MB-3 Mixed Cation-Anion Exchange Resin was obtained from Mallinckrodt, Inc., Paris, KY.

Bio-Gel P-2 (200-400 mesh) and Bio-Gel P-2 (-400) mesh were purchased from BioRad Laboratories, Richmond, CA.

Millipore filters (0.22 u) were purchased from Millipore Corp., Bedford, MA.

Molecular sieves (3 oA) were obtained from Davison Chemical, Baltimore, MD.

Methods

Carbohydrate Determination

Total neutral carbohydrate was determined by the

phenol-sulfuric acid method of Dubois et al. (50). Alternatively, carbohydrate was determined by HPLC using differential refractometry. Reducing sugar was determined using the method of Nelson (51) and Somogyi (52). Glucose was used as a standard.





20


Protein Determination

Protein content was determined according to the method of Lowry et al. (53) as described by Bailey (54). Bovine serum albumin was used as a standard. Protein concentration of CBH I(D) was determined from the extinction coefficient, 13.8 (24).

High Performance Liquid Chromatography (HPLC)

Instrumentation. Cellooligosaccharides were separated and purity assessed using a Waters Association Model ALC 202/401 Liquid Chromatograph supplemented with a Model 6000 Solvent Delivery System (Waters Associates, Inc., Milford, MA). Column effluents were monitored using a differential refractometer and a Spectra-Physics Autolab System I integrator (Spectra-Physics, Santa Clara, CA) computed relative area of chromatographic peaks.

Reverse phase chromatography. Cellooligosaccharides were separated by reverse phase chromatography using a Whatman Partisil PXS 10/25 PAC column with an acetonitrile:water (75:25; v:v) solvent system (55). Cellooligosaccharides were eluted isocratically from the column at a flow rate of 1.5 ml/min at room temperature.

Bio-Gel P-2 chromatography. Cellooligosaccharides were also separated by gel permeation chromatography using a Bio-Gel P-2 (-400 mesh) column (0.75 x 240 cm). Cellooligosaccharides were eluted with water at a flow rate of 0.7 ml/min at 600C.





21


Preparation of Cellooligosaccharides

Cellulose powder (15 g) was dissolved at 40C in 200 ml of concentrated HCl (-300C) (56). The slurry was stirred for 20 min at 40C and then for 2 h at room temperature. After the hydrolysate was added to 600 ml of water (4oC), insoluble cellulose was removed by centrifugation (Sorvall RC-5B Centrifuge, Du Pont, Co., Newtown, CT) and the pellet was washed with 2 aliquots (100 ml) of water. The supernatants were combined, filtered (Whatman No. 1 paper) and neutralized using anion exchange resin (Amberlite IR-45 [-OH]). Cellooligosaccharides were concentrated in a rotary evaporator (Buchler Instruments, Fort Lee, NJ) and then fractionated on a Bio-Gel P-2 (200-400 mesh) column (4.4 x 115 cm). Approximately 200 g of cellulose powder yielded at least 500 mg of cellotriose through cellohexaose. Carbohydrates in column effluents were monitored using the method of Dubois et al. (50).

Oxidation of Cellooligosaccharides

Bromine oxidation. The method of Diehl et al. (57) was modified as described below and used to oxidize purified cellotriose through cellohexaose. Cadmium carbonate (0.45 g, 2.7 mmoles) was added to a solution of purified cellotriose (0.449 g, 0.89 mmoles) in water (19 ml). Bromine (0.028 ml, 1.1 mmoles) was added and the suspension was stirred in a stoppered flask in the dark for 24 h at room temperature. After the mixture was filtered through Celite, silver carbonate (0.32 g, 1.2 mmoles) was added to the





22


filtrate and the suspension was stirred for 20 min. The reaction mixture was filtered through Celite and the filtrate treated with an excess of hydrogen sulfide. The precipitate was removed by filtration through Celite and the filtrate freed of residual hydrogen sulfide by aeration. The filtrate was neutralized with calcium carbonate (58), concentrated to approximately 3 ml and filtered through a Millipore filter. The calcium salt of cellotrionic acid was separated from cellotriose and by-products of the reaction using Bio-Gel P-2 chromatography. Fractions containing calcium cellotrionate were combined, concentrated to a syrup and the product precipitated and washed with 2-methoxyethanol. Alternatively, isolation of calcium salts of cellooligosaccharide aldonic acids may be omitted and aldonic acids may be lactonized directly.

Oxidation with Adams' catalyst. Cellobiose (1 g, 1.5 mmoles) was oxidized in the presence of Adams' platinum oxide catalyst (0.5 g) and gaseous oxygen in aqueous solution (10 ml) at pH 3.7 as described by Conchie et al. (59) for oxidation of monosaccharides. Lactonization of Cellooligosaccharides Aldonic Acids

Calcium cellotrionate (0.357 g) in 2 ml of water (40C) was passed through a column containing 3 ml of cation exchange resin (Amberlite IR-120 [+H]) and the column washed with 30 ml of cold water (60). The effluent was concentrated to a thin syrup at 500C under reduced pressure in a rotary evaporator and dried by successive evaporation in 6





23



ml of a mixture of dry 2-methoxyethanol and toluene (2:1) followed by evaporation to a syrup. The drying procedure was repeated three times. During the synthesis of cellotetranolactone, cellopentanolactone and cellohexanolactone precipitation of product before removal of water was prevented by using dimethyl sulfoxide in place of 2-methoxyethanol. After 5 ml of 2-methoxyethanol was added, cellotrionolactone precipitated from the syrup and the resulting suspension was centrifuged. Filtration was avoided, because precipitated lactones were hygroscopic at this stage of the synthesis. The pellet was washed with 3 aliquots (5 ml) of 2-methoxyethanol and then dried under reduced pressure at room temperature.

Reduction of Aldonic Lactones with Sodium Borohydride

Several procedures for reducing aldonic lactones to aldoses were evaluated using glucono-1,5-lactone, lactobionolactone or cellobionolactone as the starting compound. The procedures included reduction of lactones in the presence of cation exchange resin, boric acid, carbon dioxide or sodium acid oxalate as described by Frush and Isbell (61). Alternatively, 1.0 N H2SO4 was added to reaction mixtures to maintain weakly acidic conditions during reductions as described by Wolfrom and Wood (62).





24


Reduction of Lactones of Cellooligosaccharides with Sodium Boro[3H]hydride

Reduction in the presence of carbon dioxide. Under an efficient hood, 10 ml of water was cooled to 40C and a stream of carbon dioxide was bubbled through the solution. After the hydrogen ion concentration of the solution was adjusted to pH 6-7 (pH paper) with 1.0 N NaOH, 100 mg of cellotrionolactone was added and the solution was stirred. Immediately after cellotrionolactone was dissolved, 250 ul (5 mCi) of sodium boro[3H]hydride (12.1-18.1 Ci/mmole) was added dropwise during 15 minutes. After an additional 15 minutes, the hydrogen ion concentration was adjusted to pH

2.5 with 1.0 N sulfuric acid and the solution stirred for 15 minutes. The solution was passed through a column containing 3 ml of cation exchange resin (Amberlite IR120 [+H]) and the effluent was concentrated to a syrup under reduced pressure in a rotary evaporator. The syrup was evaporated three times from 5 ml of water followed by evaporation of three 5 ml aliquots of methanol to remove borate esters. The resulting powder was dissolved in 1 ml of water and the [1-H]cellootriose was separated from unreacted lactone and by-products of the reaction using Bio-Gel P-2 HPLC. To prevent radiochemical decomposition, column effluents containing radioactive cellooligosaccharide were adjusted immediately to 10 percent ethanol and stored at -700c. Unreacted lactone was converted to the corresponding calcium





25


salt and reused for future reductions. Chemical and radiochemical purity of [1-3H]cellooligosaccharides were determined using thin-layer chromatography, reverse phase HPLC and Bio-Gel P-2 HPLC. Specific radioactivity was determined by measuring radioactivity by scintillation counting and carbohydrate by Bio-Gel P-2 HPLC, reverse phase HPLC or the method of Dubois et al. (50).

Reduction in the presence of sodium acid oxalate.

Glucono-1,5-lactone (100 mg, 0.56 mmoles) was added to 20 ml of water (40C) containing 7.3 mg (0.56 mmoles) of sodium acid oxalate. Under an efficient hood, 250 l (5 mCi) of sodium boro[3H]hydride (16.4 Ci/mmole) and then sodium borohydride (4.2 mg, 1.1 mmoles) were added dropwise to the solution. After 0.5 h, the hydrogen ion concentration was adjusted to pH 2.5 with 1.0 N H2SO4. The solution was stirred for 15 min and then neutralized with mixed-bed ion exchange resin (Amberlite MB-3). After removing ion exchange resin by filtration, the filtrate was concentrated to a syrup using a rotary evaporator. The syrup was dissolved in 75 ml of water and concentrated to remove tritiated water. Volatile and nonvolatile radioactivity was measured by scintillation counting.

Specificity of Radiolabeling

To determine the location of tritium in [3H]cellooligosaccharides, [3H]cellotriose (0.28 uCi) was diluted with cold cellotriose to a total of 4.1 mg (8.1 moles) and oxidized with an excess of bromine (20 ul, 390 moless.






26


After the solution was stirred at room temperature for 24 h, water was removed using a rotary evaporator. Radioactivity in water and oxidized products was measured by scintillation counting.

Measurement of Radioactivity

Radioactive samples were dissolved in 5 ml of Aquasol and counted using a Beckman LS-9000 liquid scintillation counter (Beckman Instruments, Inc., Palo Alto, CA). All radioactive measurements were automatically corrected for quenching. Aquasol was acidified (0.4 percent) with glacial acetic acid to retard chemiluminescence. Preparation of [14C-U]Cellooligosaccharides

Batch method. The [14C-U]cellulose (750 uCi, 34.5

uCi/mg) was added to concentrated HC1 (3 ml), which had been stored at -300C. After the slurry was stirred at room temperature for 3 h, 3 ml of cold water (40C) was added and the solution was neutralized with sodium bicarbonate. The reaction mixture was centrifuged and the supernatant was filtered through a Millipore filter. The [14C-U]cellooligosaccharides were separated on a Bio-Gel P-2 (200-400 mesh) column (4.4 x 115 cm). Isolated [14C-U]cellooligosaccharides were rechromatographed on the same column or using Bio-Gel P-2 HPLC until each [14C-U]cellooligosaccharide was greater than 99 percent pure as judged by scintillation counting. Specific radioactivity of [14C-U]cellooligosaccharides was assumed equal to that of the





27


[14C-U]cellulose. Purified [14C-U]cellooligosaccharides were stored in aqueous solution at -300C.

Column method. The following procedure was used to

increase the yield of [14C-U]cellooligosaccharides of longer chain length (4-6) from acid hydrolysis of [14C-U]cellulose. During a 3 h period, 90 ml of concentrated HC1 (40C) were passed through a column (0.6 x 7 cm) containing a mixture of [14C-U]cellulose (250 uCi, 34.5 uCi/mg) and fine glass beads. The effluent was collected in a flask containing 50 ml saturated sodium bicarbonate solution. The hydrogen ion concentration of the solution in the collection flask was maintained at approximately pH 7 with the addition of sodium bicarbonate. The effluent was desalted and the [14C-U]cellooligosaccharides were separated on a Bio-Gel P-2 (200-400 mesh) column (4.4 x 115 cm). Fractions containing soluble cellooligosaccharides of chain length greater than two were combined and rechromatographed through the same column. Radioactivity in column effluents was measured by liquid scintillation counting. Thin-Layer Chromatography (TLC)

Separation of cellooligosaccharides. Aqueous samples (5-15 ug) of cellooligosaccharides in 1-2 U1 aliquots were applied to TLC plates. To a developing tank (7 x 27 x 25 cm) lined with two sheets of Whatman 3 MM paper (25 x 30 cm), 50-100 ml of the appropriate solvent was added; the paper was saturated with the same solvent and the tank equilibrated for 0.5 h. Plates were developed twice with





28



solvent A (ethyl acetate:water:isopropanol; 2:1:2; v:v:v)

(63), solvent B (ethyl acetate:water:isopropanol; 40:25:27; v:v:v), solvent C (ethyl acetate:water:n-propanol; 40:30:34; v:v:v), solvent D (ethyl acetate:methanol:acetic acid:water; 2:1:1:1; v:v:v:v), solvent E (acetonitrile:water; 70:30; v:v), solvent F (ethyl acetate:dimethyl sulfoxide:water: isopropanol; 60:40:20:20; v:v:v:v), solvent H (ethyl acetate:water:methanol; 1:2:1; v:v:v) or I (ethyl acetate:water:methanol; 2:1:1; v:v:v). Plates were developed until the solvent front migrated to 0.5 cm from the top of the plate and then dried at 1100C for 5 min between developments. Carbohydrates were stained with potassium dichromate reagent or silver nitrate reagent

(64). In some instances, before applying samples silica gel plates were immersed in 0.03 M sodium borate, sodium metabisulfite, sodium acetate or 0.2 M sodium phosphate for 5 min and then dried for 1 h at 1100C (65).

Separation of [1-3H]cellooligosaccharides. Aqueous samples (20-30 ul) of [1-3H]cellooligosaccharides were applied to preadsorbent zones of Whatman LK5D silica gel plates. After samples were applied, the plate was dried in an efficient hood with cool, dry air for 45 min. To remove traces of water, the preadsorbent zone was saturated with several drops of acetone and the plates were dried for an additional 15 min. On humid days, plates were dried at room temperature under reduced pressure for an additional 30 min. After the developing tank was lined with two sheets of





29


Whatman 3 MM paper, 30 ml of solvent G (ethyl acetate: water:methanol; 40:15:20; v:v:v) were added to the tank. Whatman 3 MM paper was saturated with solvent G and the tank equilibrated for 0.5 h. Plates (maximum of 2 per tank) were developed until the solvent front migrated to 0.5 cm from the top of the plate (1.5 h). The plates were removed from the tank and dried in an efficient hood with dry, cool air .or under reduced pressure at room temperature for 0.5 h. While the plates were drying, solvent in the tank was replenished. The plates usually were developed 3 times, after which they were dried at 1100C for 5 min. However, when initial rates of enzymic hydrolysis of [1-3H]cellohexaose were to be determined, plates were developed a total of 4 times to assure baseline separation between [1-5H]cellopentaose and [1-5H]cellohexaose. After channels containing radioactive cellooligosaccharides were covered with Whatman

3 MM paper, channels containing cellooligosaccharide standards were stained with g-anisaldehyde reagent (64).

Extraction of a mixture of [1-3H]cellooligosaccharides from thin-layer chromatographic plates. A mixture containing equal amounts of radioactivity of [1-3H]glucose through [1-3H]cellohexaose was applied to Whatman LK5 TLC plates. The plates were dried under an efficient hood for I h with dry, cool air or in an oven at 1100C for 15 min. Areas containing radioactive samples were scraped with a single-edged razor blade onto weighing paper and then transferred into 7 ml scintillation vials. The following





30


procedures for extracting [1-3H]cellooligosaccharides from scrapings were evaluated.

Method A: Water (0.5 ml) was added to vials and then

vials were placed in boiling water for 15

min.

Method B: Dimethyl sulfoxide:water (1:1; v:v) (0.5 ml)

was added to vials and then vials were placed

in boiling water for 15 min.

Method C: Water (0.5 ml) was added to vials and then

vials were sonicated for 0.5 h.
Method D: Aquasol (5 ml) and 20 Pi of glacial acetic

acid were added directly to vials containing

scrapings.

Aquasol (5 ml) and 20 ul of glacial acetic acid were added to each sample and the vials were shaken vigorously. Radioactivity was measured using a scintillation counter to 2 percent sigma error or for a maximum of 2 min.

Extraction of individual [1-3H]cellooligosaccharides.

Samples (2.5-10 Pl) of each [1-3H]cellooligosaccharide containing equal amounts of radioactivity were applied to Whatman LK5D TLC plates and then the plates were dried at 1100C for 0.5 h. Areas containing individual [1-3H]cellooligosaccharides were scraped with a single-edged razor blade onto weighing paper and then transferred to a 7 ml scintillation vial. For Method A, water (0.5 ml) was added to the vial and the vial was then vortexed for 30 s. For Method B, water (1.5 ml) was added to the vial and then the





31


vial was vortexed for 30 s. For Method C, water (0.5 ml) was added to the vial and then the vial was sonicated for 15 min. After samples were treated as described above, Aquasol (5 ml) and 20 ul of glacial acetic acid were added to the slurry and the vials were shaken vigorously. A gel formed when samples treated according to Method B were shaken. Samples were counted in a scintillation counter to 2 percent sigma error or a maximum of 2 min. Enzymic Assays

8-Glucosidase. Aryl-a-D-glycosidase activity was

determined by measuring the release of 2-nitrophenol from .nitrophenyl-D-glycosides after incubation with purified 8glucosidase (66). Samples were incubated at 400C for 20 or 30 min in 2.5 ml of 50 mM sodium acetate buffer, pH 5.0, which contained 3 mM sodium azide and 2.4 mM p-nitrophenylD-glycosides. The reactions were terminated by immersing the tubes containing the assay mixture in a boiling water bath for 5 min. After the addition of 1.0 ml of 7.5 percent (w/v) potassium phosphate (K2HP04), the absorbance of jnitrophenol at 400 nm was measured. Specific activity was expressed as Umoles of 2-nitrophenol released/min/mg of protein. The absorbance was linear from 5 to 60 ug of 2nitrophenol.

Activity of the 8-glucosidase on 8-linked glucose
disaccharides was determined by measuring the production of glucose from sophorose, laminaribiose and gentiobiose. Samples were incubated at 400C for 30 min in 0.5 ml of





32


sodium acetate buffer, pH 5.0, which contained 3 mM sodium azide and various concentrations of substrate. The reaction was stopped by immersing tubes containing the assay mixture in a boiling water bath for 5 min. Glucose was measured by a modification of the Statzyme Glucose Reagent method as described by Worthington Biochemical Corporation (67). The absorbance of NADH at 340 nm was measured after the addition of 100 ul of assay mixture to 1.0 ml of Statzyme Glucose Reagent. The absorbance measured was linear from 5 to 50 ug of D-glucose. Specific activity was expressed as moles of disaccharide hydrolyzed/min/mg of protein.

Initial rates of hydrolysis of 4-methylumbelliferyl-SD-glucopyranoside catalyzed by 8-glucosidase were measured by continuously monitoring the formation of 4-methylumbelliferone. Samples were incubated at 400C in 1.0 ml of 50 mM sodium acetate buffer, pH 5.0, which contained 3 mM sodium azide and various concentrations of 4-methylumbelliferyl-BD-glucopyranoside. The absorbance due to 4-methylumbelliferone at 346 nm was measured continuously as described by Rosenthal and Saifer (68). The absorbance response was linear in the range of 0.05 to 0.50 mole of 4-methylumbelliferone. Specific activity was expressed as Pmoles of 4-methylumbelliferone released/min/mg of protein.

Hydrolysis of labeled cellooligosaccharides. Activities of 8-glucosidase, CBH I(D) and CBH II on [1-3H]cellooligosaccharides were determined by measuring the change in the distribution of radioactive products and substrates





33



during the time course of reactions. The a-glucosidase, CBH I(D) or CBH II were incubated at 400C in 100 ul of 5 mM sodium acetate buffer, pH 5.0, which contained 0.3 mM sodium azide and 0.1-0.5 pCi of [1-3H]cellooligosaccharide. When activities of exoglucanases on [1-3H]cellooligosaccharides and [14C-U]cellooligosaccharides were determined, 0.19-0.22 uCi of [14C-U]cellooligosaccharide were included in the assay mixture. Total concentration of cellooligosaccharide in assay mixtures was adjusted with corresponding unlabeled cellooligosaccharide. The concentration of enzyme in an assay was chosen to give approximately 50 percent degradation of substrate in 30 min. Assay mixtures were contained in 2 ml conical, glass centrifuge tubes. The reaction was stopped at various times by mixing a 10 Ul sample of the assay mixture with 10 4l of 2 N HC1. The volume of sample removed from the assay mixture was chosen to give sufficient counts to detect 0.5 percent hydrolysis at a single bond of a labeled cellooligosaccharide. Labeled cellooligosaccharides in samples removed from assay mixtures were separated using thin-layer chromatography as described above. Labeled cellooligosaccharides were extracted routinely from silica gel scrapings by adding 1.5 ml of 2 percent glacial acetic acid and vortexing for 15 s. Aquasol (5 ml) was added and the mixture was shaken vigorously until a gel formed. Radioactivity was measured in a scintillation counter to 2 percent sigma error or a maximum of 5 min.





34


Kinetic Analysis of Hydrolysis Data

Allen (69) showed that bond cleavage frequencies are

best evaluated by plotting product ratios (radioactivity of each product divided by total radioactivity of the sample) vs. extent of reaction (total radioactivity of the products divided by total radioactivity of the sample). Bond cleavage frequencies were determined from the slope of product ratio vs. extent of reaction plots. Subtraction of sample background is unnecessary, because zero-time sample background affects only the intercept of the plot.

Initial rates of hydrolysis of [1-3H]cellooligosaccharides were estimated from the slope of a line through the linear, initial region of time course plots. Time course plots were constructed by plotting the amount of [1-3H]cellooligosaccharides (moles) vs. time (min). The concentration of a [1-3H]cellooligosaccharide was determined by multiplying the product ratio by the initial concentration of substrate. Specific activities were expressed as moles of [1-3H]cellooligosaccharide hydrolyzed/min/mg of protein. Sample calculations for determining bond cleavage frequencies and initial rates of hydrolysis are provided in Appendix B.

Determination of Kinetic Constants

Values of Km, Vmax and Vmax/Km for activity of exoglucanases were estimated from initial rates of hydrolysis using HYPER Fortran program described by Cleland (70). The HYPER program was also used to determine apparent values of





35


K, and Vmax for competitive inhibition plots. The activity of CBH I(D) on [1-3H]cellotriose was analyzed using Program Two/One kindly provided by W.W. Cleland. Kinetic equations used in the analyses are listed in Appendix C. Evaluation of Subsite Affinities

Subsite affinities of the 8-glucosidase were determined in part from the values of Vmax/Km obtained for cleavage of [1-3H]cellotriose through [1-3H]cellohexaose as described by Hiromi et al. (44). Subsite affinities of the 8-glucosidase were also determined using competitive inhibition constants for glucose through cellohexaose as described by Roeser and Legler (71). Subsite affinities of CBH I(D) and CBH II were determined from bond cleavage frequencies, Km and Vmax for [1-3H]cellotriose through [1-3H]cellohexaose using the method of Allen and Thoma (48) and the method of Suganuma et al. (45). Dr. J. Allen kindly provided a copy of the depolymerase computer model and sample input/output data. A brief description of each method is provided in Appendix A.















C4APTER 3
PREPARATION OF [1- H]CELLOOLIGOSACCHARIDES
AND [14C-U]CELLOOLIGOSACCHARIDES Introduction

Cellooligosaccharides have been used as substrates to investigate the mode of action and to estimate steady state kinetic parameters of cellulolytic enzymes (5,29). However, some glycosidic bonds of cellooligosaccharides that are susceptible to attack cannot be identified precisely, because identical products are formed from hydrolysis at different glycosidic bonds. Furthermore, low sensitivity of high performance liquid chromatographic analysis (HPLC) of cellooligosaccharides limits the substrate concentration at which initial rates of enzymic reactions can be measured. To overcome these limitations, cellooligosaccharides which are asymmetrically and radioactively labeled are required.

Although the enzymic methods for synthesizing oligosaccharides radioactively labeled at either the reducing or nonreducing end have been reported (42,72-74), similar methods have not been developed for synthesizing end-labeled cellooligosaccharides. Carbohydrates can be labeled chemically with tritium at carbon one of the reducing end glucosyl residue by catalytic exchange of unlabeled compounds in the presence of tritium gas and PdO/BaSO4 catalyst (75). Alternatively, carbon one-labeled aldoses can be prepared by


36





37



the reduction of aldonolactones with sodium boro[5H]hydride

(76) or lithium boro[3H]hydride (77) in weakly acidic solutions. The formation of alditols is prevented under these conditions, because the hydroxyl group at carbon one in its hemiacetal form is not readily displaced by hydride (78). The reduction of aldonolactones with sodium amalgam in tritiated water resulted in lower yields of products than reduction with lithium boro[3H]hydride (77). A practical synthesis of cellooligosaccharides based on the Koenigs-Knor type of reaction has recently been reported by Takeo et al.

(79) and may be adapted to the synthesis of nonreducing endlabeled cellooligosaccharides.

Although catalytic exchange is a straightforward, rapid method and affords products with high specific radioactivities, the necessity of using large amounts of tritium gas (10 Ci per reaction) makes this procedure potentially more hazardous and less economical than reduction of lactones with sodium boro[3H]hydride. The synthesis of lactones of cellooligosaccharides is a prerequisite for their subsequent reduction with sodium boro(3H]hydride. However, these lactones may serve as inhibitors in mechanistic studies of cellulolytic enzymes. The synthesis of reducing end-labeled cellooligosaccharides using sodium boro[3H]hydride and the preparation of [14C-U]cellooligosaccharides from [14C-U]cellulose are described in this report.





38



Results and Discussion

A flowchart of the synthesis of [1-3H]cellooligosaccharides is shown in Fig. 3-1. Purified cellooligosaccharides, which were prepared from acid hydrolysis of cellulose, were oxidized specifically at carbon 1 of the reducing glucosyl residue with mild bromine oxidation. The resulting aldonic acids of cellooligosaccharides were converted to calcium salts with calcium carbonate and then purified using Bio-Gel P-2 chromatography. Calcium salts of cellooligosaccharide aldonic acids facilitated purification of oxidized products and subsequent lactonization. Cellooligosaccharide lactones were formed from the corresponding calcium salts by first removing calcium with anion exchange resin and then shifting the equilibrium from acid to lactone by removing water. After sodium boro[5H]hydride was used to reduce cellooligosaccharide lactones specifically at carbon 1, [1-3H]cellooligosaccharides were purified using Bio-Gel P-2 chromatography.

Preparation of Cellooligosaccharides

Cellooligosaccharides used in synthesis of [1-3H]cellooligosaccharides and in enzymic assays as substrates or inhibitors were generated from cellulose by acid hydrolysis. Soluble cellooligosaccharides were separated on a Bio-Gel P-2 column as shown in Fig. 3-2. Cellotriose through cellohexaose were rechromatographed on the same column until they were >99 percent pure as determined by reverse-phase
























Figure 3-1 Synthesis of [1-3H]cellooligosaccharides

In this scheme, purified cellooligosaccharides
are oxidized at carbon 1 of the reducing end
glycosyl residue with bromine. The resulting
cellooligosaccharide aldonic acids are converted to calcium salts, which are then
separated from unoxidized cellooligosaccharides
using Bio-Gel P-2 chromatography. After calcium is removed from purified calcium salts of
cellooligosaccharide aldonic acids using cation
exchange resin, lactones of cellooligosaccharides are formed by evaporating water from a
solution containing cellooligosaccharide
aldonic acids and 2-methoxyethanol (or dimethylsulfoxide). Reduction of lactones of
celloqligosaccharide aldonic acids with sodium
borq[ H]hydride in the presence of CO2 yields
[1- H]cellooligosaccharides.






40







CH20H CH20H 0 CH20H 0 HO 0 HO'\ OH OH OH
n
1. Br2, CdCO3, H20
2. AgCO3
3. H2S



HO H H COH CH20H

O 0



; CaCO3



CH20H CH20H 1 CH20H
HO 0O --J OH OH 0 OH i0


2-methoxyethanol, toluene



CH20H
HO CHOH O CH20H CH20H HO 0H H 0=0 + Ca(OH)2

n


1. NaB3H4 S 2. H20

CH20H
HO C20 CH20H CH20H

OH H HO H
n











0


0 -d co O )



0 O C



I C .0 *q 0a r c ) q-0 .J0 L




4.) 0 f4OC 0= 0 0 rq..l 0 LO O .00









-) 4-3r C



00 0 -4 4.) : CO CH ta O I I O








4- *-IV) OrO0 .O 4-o











"- i O 0 0 r 0 c 0m r. O0 0 00















O W 0 0 O 0 0 -C 0








.0 o *4 =-4 O4 a s" 0
*4 '5 r 0 .C













C-4O 0 0 C 4 4 r-C4 *a L,

















CU Uo **o.HO0 O O O DC


















1-4 4) 40 m0 amMO WOl 000 4nI
*< *Har att





42







0
0
O




0 co










0






O 000









0 0 0 0
r') 0W


(NOIIOVI_-! / 70 ) 3Soon70





43


HPLC. Cellooligosaccharides of chain length greater than 6 precipitated from solution.

Preparation of cellooligosaccharides using Bio-Gel P-2 chromatography avoids the disadvantages of slow flow rates and necessity of repacking columns associated with the charcoal-Celite column method described by Miller et al. (56). Bio-Gel P-2 chromatography has also been used to prepare maltooligosaccharides as described by John et al. (80). Hsu et al. (29) prepared cellotriose and cellotetraose using a Bio-Gel P-2 (-400 mesh) column (1.5 x 200 cm). Oxidation of Cellooligosaccharides
The results of oxidizing purified cellobiose through

cellohexaose with bromine using a modification of the method of Diehl et al. (57) are presented in Table 3-1. Although on a weight basis, oxidation yielded greater than 100 percent of crude calcium cellooligosaccharide aldonic acids, reducing sugar assay indicated that yields ranged from 79-95 percent. The general decrease in yield as chain length increases may have resulted from the corresponding decrease in concentration of groups available for oxidation.

Diehl et al. (57) isolated cellobiono-1,5-lactone in 54 percent yield (weight) from oxidation of cellobiose, whereas cellobiose was oxidized in 93 percent yield using the method described in this report. At different steps in the synthesis, Diehl et al. (57) filtered the reaction mixture through a bed of decolorizing carbon overlaid with Celite. When the method of Diehl et al. (57) was modified by using






44



Table 3-1

Oxidation of Cellooligosaccharides


Cellooligosaccharides were oxidized with bromine using a modification of a method described by Diehl et al. (57). Oxidized cellooligosaccharides were converted to their calcium salts with calcium carbonate.


Weight of Weight of
Oligo- Cellooligosaccharide Crude Calcium saccharide (g) Salt (g) % Yielda


Cellobiose 0.50 0.56 93 Cellotriose 0.45 0.49 95 Cellotetraose 0.50 0.54 93 Cellopentaose 0.50 0.52 90 Cellohexaose 0.36 0.44 79

a Yields were determined by measuring loss of reducing
sugar.






45



only Celite as a filtering aid, cellobiose was oxidized in 93 percent yield. The affinity of cellooligosaccharides for charcoal has been used as the basis for separating cellooligosaccharides on charcoal-Celite columns (56).

Cellobiose was also oxidized using Adam's platinum

oxide catalyst as described by Conchie et al. (59). However, the resulting product was reduced with sodium borohydride in 12 percent yield. Since higher yields of reducible products were obtained from oxidation with bromine, oxidation in the presence of Adam's catalyst was not used to oxidize cellooligosaccharides.

Unoxidized cellooligosaccharides significantly reduce the specific radioactivity of [1-3H]cellooligosaccharides, if they are not removed before cellooligosaccharide lactones are reacted with sodium boro[3H]hydride. Since oxidized cellooligosaccharides contained 5-21 percent of contaminating cellooligosaccharides, oxidized cellooligosaccharides were converted to their calcium salts and separated using Bio-Gel P-2 chromatography. The separation of calcium salts of cellooligosaccharide aldonic acids from cellooligosaccharides was more reproducible than that of the corresponding aldonic acids. Although cellooligosaccharide aldonic acids and calcium salts eluted before cellooligosaccharides, cellooligosaccharide aldonic acids eluted as 2 peaks. The relative size of the peaks may represent the equilibrium between the aldonic acid and aldonolactone. Occasionally, cellooligosaccharide aldonic acids appeared to adhere to the





46



column matrix necessitating repacking the column. Thus, oxidized cellooligosaccharides were purified from cellooligosaccharides as their calcium salts. Furthermore, cellooligosaccharide lactones obtained from calcium salts of cellooligosaccharide aldonic acids were reduced with sodium borohydride in higher yields than cellooligosaccharide lactones obtained directly from aldonic acids as described below.

Lactonization of Cellooligosaccharide Aldonic Acids

The method of Diehl et al. (57) affords cellobiono-1,5lactone directly from oxidation of cellobiose. In preliminary experiments, cellobiono-1,5-lactone prepared using the method of Diehl et al. (57) was reduced with sodium borohydride in <25 percent yield. However, commercially available glucono-1,5-lactone was reduced in >80 percent yield suggesting that cellobiono-1,5-lactone synthesized using the method of Diehl et al. (57) may contain significant amounts of cellobionic acid.

Isbell and Frush (60) reported the lactonization of

calcium lactobionate in 87 percent yield. Cellobiono-1,5lactone prepared from calcium cellobionate using the method of Isbell and Frush (60) was reduced with sodium borohydride in >80 percent yield. Thus, cellooligosaccharide aldonic acids were converted to corresponding calcium salts and then lactonized using the method of Isbell and Frush (60).

The results of converting purified calcium salts of

cellooligosaccharide aldonic acids to corresponding lactones





47


are presented in Table 3-2. The yields of cellooligosaccharide lactones are comparable to those obtained for glucono1,5-lactone and lactobionolactone reported by Isbell and Frush (60). The low yield of cellobiono-1,5-lactone may be due to its higher solubility in 2-methoxyethanol. Samples of cellobiono-1,5-lactone and cellopentano-1,5-lactone were reduced with sodium borohydride to cellobiose and cellopentaose, respectively, in >50 percent yield.

When calcium salts of cellotetranoic acid, cellopentanoic acid or cellohexanoic acid were lactonized in the presence of 2-methoxyethanol, the resulting lactones were reduced with sodium boro[3H]hydride in <0.1 percent yield of radioactivity. Furthermore, products of the lactonization precipitated from 2-methoxyethanol before water was completely removed suggesting that aldonic acids formed instead of aldonolactones. Isolated products were slightly soluble in water. However, when 2-methoxyethanol was substituted with dimethyl sulfoxide, lactones, which were very soluble in water and were reduced with sodium borohydride in >50 percent yield of carbohydrate, were obtained.

Calcium salts of cellooligosaccharides served as an aid to purify oxidized cellooligosaccharides from contaminating cellooligosaccharides using Bio-Gel P-2 chromatography. Precipitation of these calcium salts with 2-methoxyethanol may be omitted and lactones may be obtained directly after removal of calcium by anion exchange resin.





48




Table 3-2

Lactonization of Calcium Salts of Cellooligosaccharide Aldonic Acids


Calcium salts of cellooligosaccharide aldonic acids were converted to their corresponding lactones, after removal of calcium by anion exchange resin and removal of water using a rotary evaporator as described in Experimental Procedures.


Weight of
Weight of CellooligoCalcium saccharide
Calcium Salt Salt (g) Lactone (g) % Yielda


Calcium Cellobionate 1.0 0.54 60 Calcium Cellotrionate 0.536 0.33 98 Calcium Cellotetranate 0.017 0.015 94 Calcium Cellopentanate 0.016 0.013 90 Calcium Cellohexanate 0.020 0.017 89

a Yields are based on the measured weights of lactones and
calcium salts assuming anhydrous molecular weights.





49


The identification of synthesized products as calcium salts of cellooligosaccharide aldonic acids and corresponding 1,5-lactones is supported by characteristic infrared bands for C-0 asymmetric stretch of C02 carboxylate and C=O stretch of 1,5-lactone groups listed in Table 3-3. Whereas 1,5-lactones have characteristic C=O stretch at 1760-1725 cm-1, 1,4-lactones have characteristic C=O stretch at 1790-1765 cm-1 (81). The formation of 1,4-lactones is unlikely, because oxygen of carbon 4 contributes to the glycosidic bond of the next residue. Reduction of Lactones of Cellooligosaccharide Aldonic Acids with Sodium Boro[5H]hydride

Procedures for reduction of aldonolactones in the presence of cation exchange resin, boric acid, carbon dioxide, sodium acid oxalate or dilute H2SO4 were evaluated on the basis of yields and stability of pH of the solution during the reaction. Although lactobionolactone or cellobiono-1,5lactone prepared by the method of Diehl et al. (57) was reduced in the presence of each of the above buffering systems to the corresponding sugar in 15-25 percent yield, carbon dioxide and sodium acid oxalate provided more stable pH's. The results of reduction of cellooligosaccharide lactones in the presence of carbon dioxide or sodium acid oxalate with sodium boro[3H]hydride are described below.

The results of reducing cellooligosaccharide lactones in the presence of carbon dioxide with sodium





50



Table 3-3
Major Infrared Bands of Calcium Salts of Cellooligosaccharide Aldonic Acids
and Cellooligosaccharides Lactones



Cellooligosaccharide Wavenumber (cm-1)


C-0 Asymmetric Stretch
of CO" Carboxylatea
Calcium Cellobionate 1557 Calcium Cellotrionate 1565 Calcium Cellotetranate 1585 Calcium Cellopentanate 1590 Calcium Cellohexanate 1598

C=O Stretch of 1,5-Lactonea
Cellobiono-1,5-lactone 1719 Cellotriono-1,5-lactone 1720 Cellotetrano-1,5-lactone 1722 Cellopentano-1,5-lactone 1728 Cellohexano-1,5-lactone 1736

a Assignments were based on characteristic infrared bands
shown by various carboxylate and 1,5-lactone groups
reported by Tipson and Parker (81).






51




O1-e

0 t* N Gr I 0

C *H
w *M co 4 ~
O -P-w
a) 0 ao z w ,

* C 4 -4 aO>O 4 D
bO ~ ~ OQ*rC
OLO 0 0 O Od o v N O 0 0 n 4 C~r i 0 "-1 a 0
O 0 4 O--- 0 L O \ O Q) U4- a ) 4H 0 0 0
94 Z. t -H 0 C- L T O0 414 :4 )1 -Cj W m o0 a






-4 t o O
0 e 0
Sa t a
a) V- A' -- 0- .,94 1




>0O S ,-4-) CO ,0 F4 a





4-) CO Z "- a,40, C o -,..4 I a ,.0 l 0 f" o OO O
C tn a -- o .








4-) > 0 .., 0 -H 0 Q I -I 0 0 h O S* 0*4 CL 0o 0 O 4 0, = O CC :3












C13 0d -H 4- .-n 0V- V-! I4 I4o O t 0<)L4 L* 0





004 4)0 > a C E'- o O O- O1 (O1C a) 1 :1 O- -H '-l or 4- 00










o ,- x 0,- 4,. 1- .,, 0 -4 0 2 CO e 4 O 0 O





o a 4 O O O O O O O

0 O", L t0 0 -) 3c 4) c 10 tO




00 Co C I O O I as O Ob~ C0O OOO CO COO k
-H O Om C c m azO cO OO 0 0 to m








r-4 OD -H ro >H 400 a) ro cO O) O) O O O O 00 4 a 0 0 T o




C~ O 0T(~
4-4 ca Vq ca AA o tk 0 0









0 0 c ci0 "- 1
0 S OQ -P 4 ) C: Q m 0 ca :it o

"-4 a co
0 >1a~ aa et no C co ca 4 4 r -H tk o ) 1) ) 4-


co C Z C 0 r_ 0 C 9 C) 4- C c 0Z L4 F" Ica 0 ca -I c 4- CO r-CO x ca

4)- > 0 0 U \ 0 ul 0 U- 0 n 0 U) 0 V 9. L 00 0 1 4) 0 r-4 r-4 0 ca co6CL r- ;-,r a r- r
o o a) C13 a)
() o Ld 0 0r~r~ ~ UU





52


boro[3H]hydride at pH 6-7 are shown in Table 3-4. Highly radioactive cellooligosaccharides (2707-3750 Ci/mole) were obtained in sufficient quantities for enzymic analysis of cellulases. The specific radioactivities of [1-5H]glucose and [1-3H]cellobiose are lower than those of [1-3H]cellotriose through [1-3H]cellohexaose, because after the addition of sodium boro[3H]hydride more lactone was converted to sugar with sodium borohydride. Two factors may have contributed to the low yield of [1-3H]cellotetraose:
(1) 10-fold less carbohydrate was present initially, although on a weight basis the concentration was similar to the other reductions and (2) the pH was lower, approximately 6. Preliminary experiments indicated that using 100 mg of lactone significantly improves yields of incorporation of tritium. The yields of [1-3H]cellopentaose and [1-3H]cellohexaose are higher, because the pH of the reaction mixtures was approximately 7. The initial concentration of sodium hydroxide was 0.017 M during the synthesis of [1-3H]cellopentaose and [1-3H]cellohexaose. This seemed to stabilize the pH during the addition of sodium boro[3H]hydride which is stored in 0.1-0.3 M sodium hydroxide.
The initial pH of the reaction mixture strongly influences the reduction of aldonolactones with sodium boro[3H]hydride. At alkaline pH's, lactones hydrolyze to aldonic acids (60), whereas at acidic pH's sodium boro[3H]hydride hydrolyzes. In addition, lactones are more readily converted to alditols at alkaline pH's. Thus, the





53


pH must be sufficiently high to stabilize sodium boro[3H]hydride, but sufficiently low to prevent hydrolysis of lactones and formation of alditols.
Specific radioactivities of [1-3H]cellotetraose through [1-3H]cellohexaose are comparable to those obtained for [1-3H]glucose (3900 mCi/mmol) and [1-3H]lactose (3000 mCi/mmol) using PdO/BaSO4 catalyst as described by Evans et al. (75). Specific radioactivities of [1-3H]cellooligosaccharides are 10-fold higher than those reported by Biely et al. (82) for [1-3H]xylooligosaccharides. Biely et al.

(82) used PdO/BaSO4 catalyst and tritium gas to label xylooligosaccharides.

Reduction of glucono-1,5-lactone in the presence of sodium acid oxalate with sodium boro[3H]hydride produced [3H]glucose in 0.65 percent yield. Since the initial pH of the solution was 3.8, sodium boro[3H]hydride may have hydrolyzed before reacting with glucono-1,5-lactone. Although Frush and Isbell (61) reduced aldonolactones to aldoses in >95 percent yield with excess sodium borohydride, the initial pH of the reaction is not suited for reductions with sodium boro[3H]hydride.
Purity and Stability of [1-3H]Cellooligosaccharides
The [1-3H]cellooligosaccharides were purified on a Bio-Gel P-2 column (4.4 x 115 cm) and using Bio-Gel P-2 HPLC. The [1-3H]cellooligosaccharides were shown to be >99 percent pure using thin-layer chromatography, reverse-phase HPLC and Bio-Gel P-2 HPLC. Alditols of





54


cellooligosaccharides were not detected by refractive index or scintillation counting. Reducing end-labeled cellooligosaccharides of specific radioactivity >250 Ci/mole were stored in 10 percent ethanol at -700C. Reducing endlabeled cellohexaose (3109 Ci/mole) is extremely labile and produced several decomposition products (approximately 80 percent of total radioactivity) after 24 h in 10 percent ethanol at 40C. Reducing end-labeled cellooligosaccharides of specific radioactivity <250 Ci/mole were stored in aqueous solution at -300C.

Specificity of Radiolabeling

To confirm that tritium was incorporated at carbon 1 of [3H]cellooligosaccharides, 0.28 uCi of [3H]cellotriose was oxidized with an excess of bromine for 24 h. After removal of tritiated water using a rotary evaporator, 92 percent of the initial radioactivity was accounted for, of which 99.1 percent was in tritiated water and 0.9 percent was in residue of oxidized products. Thus, 99.1 percent of recovered tritium was located at carbon 1.

Although catalytic tritiation yields sugars specifically labeled at carbon 1 (>98 percent) (75), reduction of lactones with lithium boro[3H]hydride in some cases yields significant quantities of labeled alditols (77). Preparation of [14C-U]Cellooligosaccharides

Batch method. The [14C-U]cellooligosaccharides, which were used in conjunction with [1-3H]cellooligosaccharides to investigate the action patterns of cellulolytic enzymes,





55


were generated from [14C-U]cellulose during hydrolysis with concentrated HC1. The distribution of [14C-U]cellooligosaccharides resulting from acid hydrolysis of [14C-U]cellulose using the batch method and subsequent separation using Bio-Gel P-2 chromatography is shown in Table 3-5. Hydrolysis yielded 39 percent soluble cellooligosaccharides ranging from [14C-U]glucose to [14C-U]celloheptaose. Preliminary experiments, in which the hydrolysis time was varied from 1 to 21.5 h, indicated that the greatest proportion of [14C-U]cellopentaose and (14C-U]cellohexaose was obtained after 3 h of hydrolysis. Labeled [14C-U]cellopentaose and [14C-U]cellohexaose are the most useful for double-label experiments with corresponding [1-3H]cellooligosaccharides. Labeled [14C-U]cellotriose through [14C-U]cellohexaose were purified using Bio-Gel P-2 HPLC until each eluted as a single symmetric peak of radioactivity. Purified [14C-U]cellooligosaccharides were stored safely in aqueous solution at -300C.

Column method. To promote formation of [14C-U]cellooligosaccharides of chain length 4-6, [14C-U]cellulose was hydrolyzed by pouring concentrated HCl through a column containing a mixture of glass beads and [14C-U]cellulose. The eluant was collected in a flask containing a saturated solution of sodium bicarbonate. Hydrolysis yielded 24 percent soluble cellooligosaccharides, of which predominantly [14C-Ujcellopentaose (36 percent) and [14C-U]cellotriose (12 percent) were obtained. The column method did not provide





56



Table 3-5

Preparation of [14C-U]Cellooligosaccharides

[14C-U]Cellulose (750 uCi) was Ydrolyzed with 3 ml of concentrated HC1 for 3 h. Soluble [ C-U]cellooligosaccharides were separated on a Bio-Gel P-2 column (4.4x115 cm).



Percent of
[14C-U]Cello- Radioactivity Recovered oligosaccharide (uCi) Cellooligosaccharides

Glucose 100 34 Cellobiose 52 18 Cellotriose 40 14 Cellotetraose 52 18 Cellopentaose 22 7.6 Cellohexaose 18 6.0 Celloheptaose 11 3.6 295 (% of initial
[ C-Ulcellulose
hydrolyzed)





57


[14C-U]cellooligosaccharides of chain length 4-6 in high yield.

Conclusions
Several [3H]cellooligosaccharides have been synthesized by sodium boro[3H]hydride reduction of cellooligosaccharide lactones. The [3H]cellooligosaccharides were shown to be specifically labeled at carbon 1 of the reducing end glucosyl residue. The specific radioactivities of [1-3H]cellooligosaccharides are sufficiently high to measure initial rates of hydrolysis of cellulolytic enzymes at micromolar substrate concentrations. Purified [14C-U]cellooligosaccharides have also been prepared and can be used in conjunction with [1-3H]cellooligosaccharides to study the action patterns of cellulolytic enzymes.















CHAPTER 4
SEPARATION OF [1- H]CELLOOLIGOSACCHARIDES
BY THIN-LAYER CHROMATOGRAPHY Introduction

Soluble cellooligosaccharides have been used to investigate kinetics and action patterns of cellulolytic enzymes (5,14,29) and have been separated using high performance liquid chromatography (HPLC)(55,83-85), paper chromatography

(PC)(20), and thin-layer chromatography (TLC)(19,63,86). Sensitivity afforded by HPLC detection methods limits substrate concentrations at which cellulolytic enzymes may be assayed to values greater than 1 mM. Furthermore, long elution times for cellooligosaccharides makes HPLC impractical for processing large numbers of samples. Previously reported TLC techniques provide limited resolution and quantitation of cellooligosaccharides has not been demonstrated.

The synthesis of [1-3H]cellooligosaccharides of high specific radioactivity affords a detection sensitivity not previously possible with cellooligosaccharides. A TLC technique which is quantitative, reliable and provides excellent resolution of cellooligosaccharides is required to take advantage of this sensitivity in analyzing enzymic digests of low concentrations of cellooligosaccharides.




58





59


A compilation of methods for separating cellooligosaccharides using HPLC has been reported recently (87). Ladisch and Tsao (83) separated glucose through cellohexaose using an Aminex 50W-X4 (Ca2+) column; however, elution times exceeded 25 min. Although Vratny et al. (84) reported the separation of glucose through cellohexaose, glucose through cellotriose were resolved poorly. The method of Chen and McGinnis (85) requires that cellooligosaccharides be converted to oximes before separation on silica gel columns containing polar-bonded alkylamino groups. Excellent resolution was reported by Gum and Brown (55) using a Whatman PXS-1025 PAC column; however, elution times exceeded 25 min for baseline separation.

Kanda et al. (20) separated cellooligosaccharides on Whatman No. 1 paper using 1-butanol:pyridine:water (6:4:3, v/v) solvent. Although glucose through cellotriose were well separated, cellotetraose through cellohexaose comigrated even after 100 h of development.

Cellooligosaccharides have been separated using thinlayer chromatography on Kieselgel G or Kieselguhr G adsorbents with ethyl acetate:water:isopropanol solvents. Saif-ur-Rahman et al. (63) separated glucose through cellohexaose on Kieselgel G plates (20x60 cm) using ethyl acetate:water:isopropanol (2:1:2, v/v) solvent. The disadvantages of this method are the commercial unavailability of such plates and the limited resolution of cellopentaose and cellohexaose. Brown and Andersson (86) separated





60


cellooligosaccharides on Kieselguhr G plates (20x20 cm) buffered with 0.02 M sodium acetate using 65 percent aqueous isopropanol:ethyl acetate (1:1). Although RF values for individual sugars indicated good separation, quality of resolution of mixtures was not reported. The 8-glucosidase from T. koningii was assayed using cellooligosaccharides as substrates; the products then were separated on Kieselgel G using 2 ascents of ethyl acetate:isopropanol:water (18:13:9; v:v:v)(19). The quality of the separation was not reported. None of the above methods reviewed have been shown to permit quantitative recovery of the compounds.

Here is reported a TLC method for the separation with high resolution and the quantitative extraction of [1-3H]cellooligosaccharides from silica gel. The silica gel plates employed are readily available commercially and separation of carbohydrates is similar to that on Kieselgel G or Kieselguhr G (65) without the requirements for impregnation with salts. This method facilitated the study of kinetics and action patterns of cellulolytic enzymes.

Results and Discussion
Separation of [1-3H]Cellooligosaccharides

Silica gel, cellulose and polyamide adsorbents were evaluated for their ability to separate cellooligosaccharides with a variety of solvents and impregnants (Table 4-1). Glucose through cellopentaose were separated well on silica gel with solvent A; however, cellopentaose and cellohexaose were poorly separated. Although Saif-ur-Rahman et





61

Table 4-1

Resolution of Cellooligosaccharides With
Different Adsorbents, Impregnants and Solvents

DevelopAdsorbent Impregnanta Solvent ments Resolution
IB2 none A 2 G1-G5 fair;G6 poor
Cellulose none H 2 G1-G2 poor;G2-G5 fair;G5-G6 poor
IB2 none H 2 G1-G6 poor Polyamide none A 2 G1-G6 poor IB2 none E 1 G1-G3 fair;G4-G6 poor
IB2 none F 1 G1-G6 fair (DMSO good solvent)
IB2 borate A 2 G1-G3 fair;G4-G6 poor
IB2 borate D 2 G1-G4 good;G5-G6 poor
IB2 metabisulfite A 2 G1-G3 excellent; G4-G6 poor
IB2 metabisulfite D 2 G1-G4 good;G5-G6 poor
IB2 acetate A 2 G1-G4 fair;G5-G6 poor
IB2 acetate D 2 G1-G5 good;G5-G6 fair
LK5 phosphate A 3 G1-G4 good;G5-G6 poor
LK5 none B 3 GI-G5 good;G5-G6 poor
LK5D none B 6 G1-G5 good;G5-G6 fair
LK5 none C 3 G1-G5 good;G5-G6 poor
LK5/LK5D none G 3 G1-G6 excellent K5 none I 3 G1-G4 excellent; GS-G6 poor
K6 none I 3 G1-G4 excellent; G5-G6 poor
a Impregnants were 0.03 M sodium borate, sodium metabib sulfite, sodium acetate or 0.2 M sodium phosphate. Solvent compositions were A (ethyl acetate:water:isopropanol; 2:1:2), B (ethyl acetate:water:isopropanol; 40:25: 27), C (ethyl acetate:water:n-propanol; 40:30:34), D (ethyl acetate:methanol: acetic acid:water; 2:1:1:1), E
(acetonitrile:water; 70:30), F (ethyl acetate:dimethyl sulfoxide:water:isopropanol; 60:40:20:20), G (ethyl acetate:water:methanol; 40:15:20) or H (ethyl acetate: water:methanol; 1:2:1), or I (ethyl acetate:water: methanol; 2:1:1).





62



al. (63) reported good separation of cellooligosaccharides using solvent A, we observed negligible migration of cellohexaose. Separation of cellooligosaccharides on cellulose TLC plates was fair; however, migration of cellohexaose was negligible. Furthermore, the cellulose adsorbent easily flaked, thereby eliminating this adsorbant for use with radioactive cellooligosaccharides. Polyamide adsorbent poorly separated cellooligosaccharides.

Separation of sugars on silica gel is improved by impregnating silica gel with inorganic salts, such as bisulfite, boric acid, mono- and dibasic phosphate and sodium acetate (65). Although impregnating silica gel with salts typically improved resolution of glucose through cellopentaose, cellopentaose and cellohexaose were poorly resolved. Sodium acetate and sodium metabisulfite increased resolution more than did boric acid and sodium phosphate. The use of impregnants, especially boric acid, lowered RF values for cellooligosaccharides.

Isopropanol, n-propanol, methanol and dimethyl

sulfoxide each characteristically influenced the migration of cellooligosaccharides. Isopropanol and n-propanol typically separated glucose through cellotetraose with excellent resolution, whereas longer cellooligosaccharides were poorly separated. Dimethyl sulfoxide served as an excellent solvent for cellooligosaccharides, yielding high RF values with decreasing resolution of longer cellooligosaccharides. Methanol also yielded high RF values; however,





63


resolution between adjacent cellooligosaccharides was similar regardless of chain length. Concentrations of methanol higher than those in solvent G increased the spreading of bands. Lower concentrations of methanol increased the sharpness of bands, but decreased the difference between RF values.

Although Baker-flex IB2 silica gel plates were used to evaluate most solvents and impregnants, Whatman K5 and K6 series plates provided higher resolution of cellooligosaccharides. The LK5D series is especially suited for separation of [1-3H]cellooligosaccharides, because (1) plates contain a preadsorbent zone which increases resolution, (2) preadsorbent zone allows addition of up to 50 ul of sample, (3) prechanneled plates prevent contamination of samples and (4) 150 A pore size provides excellent separation of polar compounds.

The best separation of cellooligosaccharides was obtained using Whatman LK5D silica gel plates with three ascents of solvent G and is shown in Fig. 4-1. The ability of preadsorbent zone to focus cellooligosaccharides is reflected in the sharpness of the bands. Similarly, [1-3H]glucose and through [1-3H]cellohexaose were separated with high resolution as shown by the distribution of radioactivity in Fig. 4-2. The RF values are listed in Table 4-2.

Preliminary experiments indicated that approximately 510 percent of [1-3H]cellooligosaccharides adsorbed























Figure 4-1 Thin-layer chromatographic pattern of cellooligosaccharides

Cellooligosaccharides were separated on Whatman
LK5D TLC plates using 3 ascents of solvent G
(ethyl acetate:water:methanol; 40:15:20;
v:v:v). The TLC plate was stained for carbohydrate using j-anisaldehyde reagent as
described in Experimental Procedures. Lanes, from left to right, contain 1 ug of glucose,
cellobiose, cellotriose, cellotetraose, cellopentaose, cellohexaose and a mixture of glucose-cellohexaose (1 ug each).






65 I ..........






















Figure 4-2 Thin-layer chromatographic pattern of
[1-3H]cellooligosaccharides

Reducing end-labeled cellooligosaccharides were
separated on Whatman LK5D TLC plates using 3
ascents of solvent G (ethyl acetate:water: methanol; 40:15:20; v:v:v). The lane was
div ded into 2 mm zones from which
(1--H]cellooligosaccharides were eluted and
quantitated by scintillation counting as
described in Experimental Procedures. The mixture applied contained 8300 dpm of each
[1-'H]cellooligosaccharide. The peaks represent, from left to right [1-3H]cellohexaose,
[1-3H]cellopentaose, [1- H]cellotetraose,
[1-3H]cellotriose, [1- H]cellobiose and
[1-3H]glucose. The arrow represents the end of
the preadsorbent zone.


















3.6 3.0




2.4



0 x 1.8 a








0.6




0
0 4 8 12 16 20

DISTANCE FROM BOTTOM (CM)





68





Table 4-2

Values of the Migration Parameter, RF
[1-3H]Cellooligosaccharides were separated using
Whatman LK5D TLC plates with 3 developments in solvent G (ethyl acetate:water:methanol; 40:15:20; v:v:v). [1-3H]Cellooligosaccharide RFa


Glucose 0.59 Cellobiose 0.50 Cellotriose 0.39 Cellotetraose 0.30 Cellopentaose 0.20 Cellohexaose 0.13 a RF values were based on distance migrated by
cellooligosaccharides and solvent from top of preadsorbent zone.





69


irreversibly, when plates were dried at 1100C after application [1-3H]cellooligosaccharides. To avoid irreversible adsorption at the origin or at intermediate positions of migration, plates must be dried at room temperature instead of at 1100C. Extraction of [1-3H]Cellooligosaccharides

Four methods of extracting a mixture containing equal amounts of radioactivity of [1-3H]glucose through [1-3H]cellohexaose from Whatman LK5 TLC plates were evaluated and the results are shown in Table 4-3. When plates were dried at room temperature before extraction, cellooligosaccharides were recovered 10-14 percent more efficiently than when plates were dried at 1100C. Extracting cellooligosaccharides before adding Aquasol improved recoveries by 42-45 percent with plates dried at room temperature and 32-37 percent with plates dried at 1100C. Percent recovery of cellooligosaccharides was approximately equal for each extraction procedure, when plates were dried at room temperature. However, extraction using dimethyl sulfoxide:water solvent yielded slightly lower recoveries than other extraction procedures when plates were dried at 1100C.
Low recoveries of radioactivity from silica gel

suspended in Aquasol may result from physical quenching elicited by silica gel. Aqueous solvents may serve to desorb cellooligosaccharides and disperse silica gel, thereby promoting contact of [1-3H]cellooligosaccharides with






70
















a)
00 ,0 2 (U -C -4 V
n Ld cc a ) cq (N N\J



GrB0 (> 0 0 0 0 0, :4 0 .C 0 0 0 0
ca > co Ol O O O1: S0E O cO t O
S O CO 0 a) C 0 cu 0 Oa

0 -4 P 0 4 +1 +1 1 'i 4

0 -* '- O O IO O OU o C -4 Q> 0 0 0 0 0.




>o- O > 0 0 0 0 C 0L) 0 .4 --) c 0 Li L 0 m 1 0 0 0 H P 3 PMw <94-P a -H -H H -H

Q > 4- E -4 -P 00 X o 0 0 0

So -4d a0 w U -H w (a 0 C 0 0 0 0 V
0 4 3 o .4 -- 0 0 EA 9 L:- w m W t -q C -4 ,,- 0 a 0 > *0 0 V

m00 0 00- 4 L C- 0 L > N

0 rO CO 4-nl >\ +- ( O o 3 o a o OP "- 0 4-,--) I 0 4 0 6 4 (1 0 ( U 00

acc 0 t C a) 4 Q>
bD 0 4L4 ia2nCO 0V >: > n '.0 To a) ca ca L 0 ON : -0 I0 ,0 0 o 4 0lM L q 0 a 0 "


M-,- C -rM o4- 0 0 0 0-4
- 0 (U-(DHo a0k


X--04 *) Li) C 0 0 0 0 (i K40 -)0(U OWj00-H t- 0 00( Ci( *U 00m O O O O

r 000 a( OOL. 44 4 4a 44 0O

90 a aOO co O- O O- O("t1 0


0U.( 4---'*: 0 > > CO V 02 O C -C 0-0 O O O O (
0 4-) 0 O 4- Lic


O E0 i ** a)c.r 4 0 > c F4 C O( r-4 4-) 04 0 V(
*H O -4-w0CL 0 1 4( 0 *Lt



o )0 o 80 c- 0 0
X000 (C OO

w c*4- .HM-4H C >0
-c:C r n r- 3 x ,(Ul-. 0o 0 -H 0 L C000 0 *N4fC'j ( 0 0 n Oh -4 0l.. a) > L QU
1I-00(-= 0 VC x E-- 4-4M w0CO Q
-B >3 O M m 0 Q (U





71


scintillation cocktail. Scott (88) eluted glucose, mannose, galactose, arabinose, xylose and sucrose from silica gel with 95 percent recovery.

To assure equal recoveries of [1-3H]cellooligosaccharides from silica gel, methods for extracting individual [1-3H]cellooligosaccharides from Whatman LK5D plates were evaluated. The results are shown in Table 4-4. Although recoveries of [1-3H]cellooligosaccharides using methods A, B and C were similar, the range of recoveries for method B was less than those of methods A and C. Thus, method B, by which [1-3H]cellooligosaccharides are counted with an efficiency of 883, was chosen as the extraction procedure. The efficiency of extraction varies slightly from plate to plate and with length of heating at 1100C. To efficiently stain cellooligosaccharide standards, plates were heated at 1100C after the final development.

Conclusions

A thin-layer chromatographic method for quantitating

and separating [1-3H]cellooligosaccharides with high resolution on silica gel has been developed. Although TLC methods for separating cellooligosaccharides have been reported (19,63,86), resolution and quantitation of cellooligosaccharides comparable to that reported here have not been demonstrated. The separation of [1-3H]cellooligosaccharides on Whatman LK5D silica gel plates with 3 ascents of ethyl acetate:water:methanol (40:15:20; v:v:v) will facilitate the analysis of products formed from the action of cellulolytic enzymes on radioactive cellooligosaccharides.






72







V i IN n rc N rc N
0 D -H H "+ 4 .. 4.

0 4.4 0 0 O O

*0 0 0 0
o > o 0 0 0 0 0 0

0i) 0 0 0 0 0 0
ro I D 0 D N N rN q3





0 0. .- 0 O O O 0 O 0 8 0 0 0 I CO O ) O O O Q0 -4 c 4 4 4 4 4 4 S C -. c 0 0 0 0 0 0 0



0 CO > 4 CO 0 0 0 0 0 0 ,C 1 ---t m Nr N C0 OO 1- > 0 0 0 0 0
0 0 no to C N 'T 0 n 0') to c c c co
Q Eco a -H



OO ) 0 >
-e r 0 0 -4 cO 0-4CO -H0 0 CO4 4 4 0H 44 44 4 O 0 o > ~ a) o\ M Cto 0 0 co 4-) 4-) a) a) x a) am m )







0 0-00> r O O O O O O-Sa c O O 0 >
0 0-O (C1 0 0 0 0 0 0

O I V0 r-4 W 0 w 0 0 0 0 0 0
-4C V o *-q 0 W

-r O0 0 41 -4 4 4 4 4 o 0 OC ar-- .-4 co 0 0 0 0 0 0
rQ.O -O > t4 0 0 0 0 0 0 4o CO C- O4 o1- 0 n- n 0 0 -T0




0 x 0 N 0 N U0 O O 0 W
000C --4 OOOO





SO 44 4' 4 4 4 4 CC o a0 O 0 a





4 W O 4-- 0 0 0 0 0 0 S', 0 -0 n 0 Or O m I' 0 0 mO E4 4o 4C 44 4c 40
-H O C4Or- 0 0 0 0 0 0


-a rC-4 4 KLa I'D V, N n



0 0nco0 n 4-) 0 0 0 0 0 0 4:z 1 co 0 0 0 0 0 0 4- CO L N \10 N n 0 L 00 O 44 A 44 -4 4 CO D4-4C- 3- O0 0 0 O O0 C.. c C 0 Q0C 0 0 0 0 0 0 4 ) O 0. O 0 r- n 0 x c*MMO -< C N co wO- 90 0 a M IC' O


4-) t O O O O to CO
C O-0 I CO CO
4OO4-4 C o 0 0 C r--I O -4-CO Co CO 0 0. V CO C a) o 0 9-4 4- 3 CO 2 cuCOV 0 r-10 r_ 4 C X C 4- -H (1) O a) Cq.t 4) a CO c)C to > C CO ..0 4.) 4-) Q ..c
CCO cccO n0 0 0 0 0 0 0
C-PO) a) L- -4 4 1- -4
Q) > 0 CD0C-) 0 u U u














CHAPTER 5
MODE OF ACTION OF 8-GLUCOSIDASE Introduction

The cellulase system of T. reesei is comprised of endoglucanases, cellobiohydrolase I (D), cellobiohydrolase II and 8-glucosidases. Cellulolytic enzymes act sequentially and cooperatively to degrade crystalline cellulose to glucose. The role of the 8-glucosidase in saccharification of cellulose is to degrade cellobiose, an inhibitor of the depolymerizing enzymes, and cellooligosaccharides to glucose (5,21,36). Although extracellular 8-glucosidases have been purified and partially characterized from Trichoderma reesei

(8), Trichoderma viride (16,34,35), Trichoderma koningii

(19), Irpex lacteus (20) and Botryodiplodia theobromae (89), little information is available regarding the direction from which 8-glucosidases remove glucosyl residues, mode of attack (repetitive or multiple attack), kinetic constants governing the binding of cellooligosaccharides and topography of active sites.

A variety of aglycones linked through 8-glycosidic

bonds are hydrolyzed by 8-glucosidases (32). The nature of the aglycone influences rates of hydrolysis. Cellulolytic 8-glucosidases are specific for the C-4 hydroxyl group and do not hydrolyze 8-galactosides (90). In general, as cellooligosaccharide chain length increases, Km decreases


73





74



indicating that cellooligosaccharides are good substrates for B-glucosidases (90). Rates of hydrolysis of cellooligosaccharides correspondingly increase with increasing chain length. The dependence of rates of hydrolysis and K, on chain length suggested to Shewale (90) that in saccharification of cellulose, a-glucosidases are engaged more in hydrolysis of cellooligosaccharides than in hydrolysis of cellobiose. Hydrolysis of reduced cellooligosaccharides by 8-glucosidases indicates that glycosyl residues are sequentially removed from the nonreducing end; however, the effects of modifying the reducing glucosyl residue on enzymic activity have not been investigated (91).

Although the arrangement of subsites in the active site of 0-glucosidases is poorly understood, subsite affinities have been evaluated for glucoamylase from Rhizopus delemar

(44) and Aspergillus saito (47). The subsite maps, which were constructed using the method of Hiromi et al. (44), show that the active site of each enzyme consists of 7 subsites. Subsite II (numbered from the nonreducing end) contributes the greatest proportion of binding energy, whereas the affinity of subsite I is negligible. Subsite affinities decrease in order from subsites III through VII. Since only one productive complex is possible with glucoamylases, the catalytic groups are positioned between subsites I and II.

To determine the mode of action, topography of the active site and to more clearly define the role of the





75


8-glucosidase in the cellulase system of T. reesei, the activity of the 6-glucosidases on [1-3H]cellooligosaccharides is examined in this report. The use of [1-3H]cellooligosaccharides permits measurement of initial rates of hydrolysis at micromolar substrate concentrations. In addition, these asymmetrically labeled cellooligosaccharides permit identification of glycosidic bonds susceptible to enzymic hydrolysis. The 8-gluco'sidase was shown to remove glucosyl residues from the nonreducing end of cellooligosaccharides in a multi-chain mode of attack. Maximum velocities were determined to be independent of chain length, whereas Km's decrease as chain length increases. Cellooligosaccharides were also used as inhibitors of 8glucosidase-catalyzed hydrolysis of methylumbelliferyl-8-Dglucopyranoside (MUG). Inhibition constants of cellooligosaccharides were shown to be similar to the corresponding Michaelis constants. Two approaches were used to evaluate subsite affinities of the 8-glucosidase and revealed that the active site consists primarily of 3 subsites. The results are discussed in terms of the role of the 8-glucosidase in saccharification of cellulose.

Results and Discussion
Hydrolysis of [1-5H]Cellooligosaccharides

To determine initial rates and to identify products formed from 8-glucosidase-catalyzed hydrolysis of [1-3H]cellooligosaccharides, samples were removed from assay mixtures at various times and products separated using





76


TLC. Michaelis parameters were obtained by measuring initial rates of hydrolysis at various concentrations of [1-3H]cellooligosaccharides.

The activity of 8-glucosidase on [1-3H]cellobiose is

shown in Fig. 5-1. Reducing end-labeled cellobiose was hydrolyzed with a corresponding increase of [1-3H]glucose. Since only reducing end-labeled products were monitored in the assay, glucose at the nonreducing end was not observed. Initial bond cleavage frequencies were determined from the slopes of product ratios versus extent of reaction plots. Deviation from linearity of the slope indicates a change in mode of action. Bond cleavage frequency analysis of 8-glucosidase catalyzed hydrolysis of [1-3H]cellobiose indicated that only one bond was cleaved since the slope of [1-5H]glucose product ratio versus extent of reaction curve is equal to 1.0 (Fig. 5-2). Recovery of [1-3H]cellobiose and [1-3H]glucose from TLC plates was constant for each time point and at each concentration of substrate assayed indicating that transfer reactions are unlikely to occur from 0.25 to 10.0 mM [1-3H]cellobiose.

Linear Lineweaver-Burk and Eadie-Hofstee plots were

obtained using initial rates of hydrolysis from 0.25 to 10.0 mM [1-3H]cellobiose (Fig. 5-3). Values of Km and Vmax were determined to be 88040 uM and 17.70.2 umol/min/mg, respectively (Table 5-1).

Reducing end-labeled cellobiose and [1-3H]glucose were formed during 8-glucosidase hydrolysis of [1-3H]cellotriose (Fig. 5-4). The initial appearance of only [1-3H]cellobiose























Figure 5-1 Time course hydrolysis of [1-3H]cellobiose by
8-glucosidase

The 8-glucosidase (1.05x10-4 mg) was incubated
in 100 ul of 5 mM sodium acetate buffer, pH
5.0, containing 250 uM [1- H]cellobiose.
Samples were removed from the reaction mixture after various intervals and analyzed using TLC
as described in Experimenta Procedures. The
early, linear region of [1- H]cellobiose degradation curve yielded the initial velocit. The curves represent the distribution of [1- H]glucose (o) and [l- H]cellobiose (*).





78









I I I I
25





20











0
I 10
0
10 o






5






0 10 20 30 40 50 60 TIME (MIN)
























Figure 5-2 Bond cleavage frequency plot for hydrolysis of
[1-3H]cellobiose by a-glucosidase

The 8-glucosidase (1.05x10-4 mg) was incubated
in 100 ul of 5 mM sodium acetate buffer, pH
5.0, containing 250 uM [1- H]cellobiose.
Samples were removed from the reaction mixture and analyzed using TLC as described in Experimental Procedures. The initial slope of each
line is the bond cleavage frequency of the
substrate bond yielding the product (Gi). The
curves represent the product ratio of
[1-3H]glucose (o) and substrate ratio of
[1-5H]cellobiose (*).

















1.0





0.8





\J 0.6
+




, 0.4





0.2





0 I I I I
0 0.1 0.2 0.3 0.4 0.5

GI / (GI +G2)























Figure 5-3 Lineweaver-Burk and Eadie-Hofstee plots for
hydrolysis of [1-3H]cellobiose by a-glucosidase

Initial rates for hydrolysis of [1-3H]cellobiose (0.25-10 mM) by B-glucosidase were used
to construct Lineweaver-Burk (A) and EadieHofstee (B) plots.





82





A
0.25
z

a 0.20 a


r 0.15


o 0.10 o0.0



0 I 2 3 4

[3H G2]-1 ( mM '





B
16


- 12


8


4



0 4 8 12 16

V* [H G2]"' C(MOL MIN-'* MG PROTEIN"* mM')




Full Text

PAGE 1

MODE OF ACTION OF EXOGLUCANASES FROM THE CELLULOLYTIC FUNGUS Trichoderma reesei ; ACTIVITY ON REDUCING END-LABELED CELLOOLIGOSACCHARIDES AND TOPOGRAPHY OF ACTIVE SITES OF S-GLUCOSIDASE, CELLOBIOHYDROLASE 1(D), AND CELLOBIOHYDROLASE II BY WILLIAM JOSEPH CHIRICO DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1984

PAGE 2

4 This dissertation is dedicated to my parents, Dominick and Catherine Chirico, and to my Aunt and Uncle, Inez and Frank Montalbano.

PAGE 3

ACKNOWLEDGMENTS I wish to express my sincere gratitude to Dr. Ross D. Brown, Jr., for his guidance and support during the course of this work. I would also like to thank the other members of my supervisory committee, Drs. Charles M. Allen, Jr., Kim J. Angelides, Vicent Chau, Michael Roberts, and Jesse F. Gregory, for their suggestions and support. I wish to thank Drs. Phil Laipis, Mike Kilberg, and Daniel Purich for their support. I am also grateful to Charles duMee, John Denny, Larry Weissbach, Mark Eller, and Sal Pietromonaco for friendship and many valuable discussions. I am extremely grateful for the excellent technical assistance of Cynthia Fazenbaker and Vicki Andersen who shared the duties of scraping TLC plates. I would also like to thank Jim Parkes for incorporating J.D. Allen's depolymerase computer model in my computer file. I wish to thank Cindy Zimmerman for typing the dissertation and Nancy Shaskey for drawing the figures. Finally, I would like to thank Crystal Willis, without whose support, encouragement, and love this dissertation could not have been written. lii

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGMEfJTS iii LIST OF TABLES vii LIST OF FIGURES x LIST OF ABBREVIATIONS AND SYMBOLS XV ABSTRACT xvi CHAPTER j 1 INTRODUCTION 1 Structure of Cellulose 3 Cellulases 4 Subsite Mapping of Enzymes 12 Assessment I5 2 EXPERIMENTAL PROCEDURES 17 Materials 1 7 Enzymes I7 Substrates 1 7 Chemicals 18 Chromatographic Supplies 18 Methods 19 Carbohydrate Determination I9 Protein Determination 20 High Performance Liquid Chromatography (HPLC) 20 Preparation of Cellooligosaccharides 21 Oxidation of Cellooligosaccharides 21 Lactonization of Cellooligosaccharide Aldonic Acids 22 Reduction of Aldonic Lactones with Sodium Eorohydride 23 Reduction of Lactones of Cellooligosaccharide Aldonic Acids with Sodium Boro[3H]hydride 24 Specificity of Radiolabeling 25 Measurement of Radioactivity 26 Preparation of ['''^C-Ujcellooligosaccharides.26

PAGE 5

Thin-layer Chromatography (TLC) 27 Enzymic Assays 31 Kinetic Analysis of Hydrolysis Data 34 Determination of Kinetic Constants 34 Evaluation of Subslte Affinities 35 3 PREPARATION OF [1 -^HjCELLOOLIGOSACCHARIDES AND [ ^ ^C-U] CELLOOLIGOSACCHARIDES 56 Introduction 36 Results and Discussion 38 Preparation of Cellooligosaccharides 38 Oxidation of Cellooligosaccharides 43 Lactonization of Cellooligosaccharide Aldonic Acids 46 Reduction of Lactones of Cellooligosaccharide Aldonic Acids with Sodium Boro[^H]hydride 49 Purity and Stability of [1-^H] Cellooligosaccharides 53 Specificity of Radiolabeling 54 Preparation of ["^+0-1;] Cellooligosaccharides 54 Conclusions 57 4 SEPARATION OF [ 1 -^H] CELLOOLIGOSACCHARIDES BY THIN-LAYER CHROMATOGRAPHY 53 Introduction 58 Results and Discussion 60 Separation of [1-^H]Cellooligosaccharides. .60 Extraction of [1-^H]Cellooligosaccharides. .69 Conclusions 71 5 MODE OF ACTION OF 6-GLUCOSIDASE 73 Introduction 73 Results and Discussion^ 75 Hydrolysis of [1-^H]Cellooligosaccharides. .75 Hydrolysis of [1-^H]Cellooligosaccharides and [ '^C-U]Cellooligosaccharides 113 Inhibition of Methylumbellif erylglucoside (MUG) Hydrolysis by Cellooligosaccharides II9 Subsite Mapping of 6-Glucosidase 135 Conclusions 1 44 6 MODE OF ACTION OF CELLOBIOHYDROLASE 1(D) I46 Introduction 145 Results and Discussion 148 Hydrolysis of [1 -^H] Cellooligosaccharides 143 V J

PAGE 6

Hydrolysis of [1-^H]Cellooligosaccharides and [ '^C-U]Cellooligosaccharides 184 Subsite Mapping of Cellobiohydrolase 1(D) I96 Conclusions 205 7 MODE OF ACTION OF CELLOBIOHYDROLASE II 206 Introduction 206 Results and Discussion 208 Hydrolysis of [1 -^HjCellooligosaccharides 208 Hydrolysis of [1--^H]Cellooligosaccharides and [ ^C-U]Cellooligosaccharides 240 Subsite Mapping of Cellobiohydrolase II.... 249 Conclusions 258 8 GENERAL DISCUSSION 260 Hydrolysis of [1-%]Cellooligosaccharides 260 Synergism of Cellulolytic Enzymes 263 Individual Roles of Cellulolytic Enzymes 270 Methods of Subsite Mapping 274 Future Experiments 281 9 SUMMARY 283 APPENDICES •. A SUBSITE MAPPING OF ENZYMES 286 B SAMPLE CALCULATIONS FOR BOND CLEAVAGE FREQUENCIES AND INITIAL RATES OF HYDROLYSIS OF [ 1 -^H ] CELLOOLIGOSACCHARIDES 3O4 C KINETIC EQUATIONS 309' D SUBSITE AFFINITIES FOR 8-GLUCOSIDASE 310 E KINETIC CONSTANTS FOR HYDROLYSIS OF 6-LINKED DISACCHARIDES OF GLUCOSE BY 6-GLUCOSIDASE 311 F KINETIC CONSTANTS FOR HYDROLYSIS OF [1-^H] CELLOOLIGOSACCHARIDES BY 3-GLUCOSIDASE, CBH 1(D), AND CBH II 31 2 BIBLIOGRAPHY BIOGRAPHICAL SKETCH ^02 vi

PAGE 7

LIST OF TABLES Table Page 3-1 Oxidation of Cellooligosaccharides 44 3-2 Lactonization of Calcium Salts of Cellooligosaccharide Aldonic Acids 48 3-3 Characteristic Infrared Bands of Calcium Salts of Cellooligosaccharide Aldonic Acids and Cellooligosaccharides Lactones 50 3-4 Reduction of Cellooligosaccharide Lactones with Sodium Boro[-^H]hydride 51 35 Preparation of ['''^C-UjCellooligosaccharides 56 41 Resolution of Cellooligosaccharides With Different Adsorbents, Impregnants and Solvents 61 4-2 Values of the Migration Parameter, Rp 68 4-3 Extcaction of a Mixture of [1-^H]Cellooligosaccharides From Thin-layer Chromatographic Plates 70 44 Extcaction of Individual [1-^H]Cellooligosaccharides From Thin-layer Chromatographic Plates 72 51 Kinetic Constants for [1 -^HjCellooligosaccharide Hydrolysis by 8 -Glucosidase of T. reesei 83 5-2 Hydrolysis of [1 -^HjCellopentaose and [^4c_U]Cellopentaose by g -Glucosidase 114 5-3 Initial Rates for 8 -Glucosidase Activity on [1-^H]Cellopentaose and [""^C-UjCellopentaose. .116 5-4 Hydrolysis of [ 1 -^HjCellohexaose and [ '4c-U]Cellohexaose by 8 -Glucosidase 117 5-5 Initial Rates for 8 -Glucosidase Activity on [l-^'H] Cellohexaose and [^^C-U]Cellohexaose 118 vii i

PAGE 8

56 Inhibition of 8-Glucosidase byCell ooligosaccha rides 134 61 Kinetic Constants for [1-^H]Cellooligosaccharide Hydrolysis by Cellobiohydrolase 1(D) 183 6-2 Hydrolysis of [1--^H]Cellopentaose and [ '4c-U]Cellopentaose by Cellobiohydrolase 1(D) 185 6-3 Initial Rates of Cellobiohydrolase 1(D) Activity on [ 1 -^HjCellopentaose and [^^C-UjCellopentaose 189 6-4 Hydrolysis of [1 -^HjCellohexaose and [ '4c-U]Cellohexaose by Cellobiohydrolase 1(D)... 191 6-5 Initial Rates of Hydrolysis of Cellobiohydrolase 1(D) Activity on [1-^H]Cellohexaose and [^"^C-ujCellohexaose 195 6-6 Subsite Affinities of CBH 1(D) Calculated Using Values of V /K and Bond Cleavage Frequencies from n -^H JCellooligosaccharides. 197 7-1 Kinetic Constants for [1 — ^HjCellooligosaccharide Hydrolysis by Cellobiohydrolase II 239 7-2 Hydrolysis of [ 1 -^H] Cellopentaose and [ ^C-U]Cellopentaose by Cellobiohydrolase II.... 242 7-3 Initial Rates of Cellobiohydrolase II Activity on [ 1 -^HjCellopentaose and [ '^C-U]Cellcpentaose 243 7-4 Hydrolysis of [1-^H]Cellohexaose and [ '^C-U]Cellohexaose by Cellobiohydrolase II 245 7-5 Initial Rates of Cellobiohydrolase II Activity on [1 -^HjCellohexaose and [ '4C-U]Cellohexaose 250 7-6 Subsite Affinities of CBH II Calculated Using Values of V_ ZKj and Bond Cleavage Frequencies from IT-^H JCellooligosaccharides 252 B-1 Distribution of [ 1 -^HjCellooligosaccharides from Hydrolysis of [ 1 -^HjCellotetraose by CBH 11(D) \ 305 viii

PAGE 9

Product Ratios for [1 -^HjCellotetraose Hydrolysis by CBH 1(D) Concentration of [1 -^H iCellooligosaccharides from Hydrolysis of [1 --^HjCellotetraose by CBH 1(D) ix

PAGE 10

LIST OF FIGURES Figure Page 3-1 Synthesis of [1 -^Hjcellooligosaccharides 40 32 Elution pattern from Bio-Gel P-2 column chromatography of cellooligosaccharides generated from acid hydrolysis of cellulose 42 41 Thin-layer chromatographic pattern of cellooligosaccharides 65 42 Thin-layer chromatographic pattern of [1 --^H] cellooligosaccharides 67 51 Time course hydrolysis of [1 --^Hjcellobiose by 3 -glucosidase 78 5-2 Bond cleavage frequency plot for hydrolysis of [1-^H]cellobiose by g -glucosidase 80 5-3 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [1-^H]cellobiose by 3 -glucosidase 82 5-4 Time course hydrolysis of [ 1 --^Hjcellotriose by 6 -glucosidase 85 5-5 Bond cleavage frequency plot for hydrolysis of [1-^H]cellotriose by g -glucosidase 88 5-6 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [1-%]cellotriose by 8 -glucosidase 90 5-7 Time course hydrolysis of [1 -^Hjcellotetraose by 0 -glucosidase 92 5-8 Bond cleavage frequency plot for hydrolysis of [1-^H]cellotetraose by g -glucosidase 95 5-9 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [ 1 -^Hjcellotetraose by 8 -glucosidase 97 X

PAGE 11

5-10 Time course hydrolysis of [1-^H]cellopentaose by 3-glucosidase 99 5-11 Bond cleavage frequency plot for hydrolysis of [1--^H]cellopentaose by g-glucosldase 101 5-12 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [1-^H]cellopentaose by B-glucosidase 104 5-13 Time course hydrolysis of [1-^H]cellohexaose by S -glucosidase 106 5-14 Bond cleavage frequency plot for hydrolysis of [1 — 'Hjcellohexaose by 8 -glucosidase 108 5-15 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [ 1 -^Hjcellohexaose by 3-glucosidase 111 5-16 Inhibition of 6 -glucosidase by cellobiose using 4-methylumbellif eryl-s -^-glucopyranoside as the substrate 121 5-17 Inhibition of s -glucosidase by cellotriose using 4-methylumbelliferyl-6 -^-glucopyranoside as the substrate 123 5-18 Inhibition of 8 -glucosidase by cellotetraose using 4-methylumbelliferyl-8 -D_-glucopyranoside as the substrate 125 5-19 Inhibition of 8 -glucosidase by cellopentaose using 4-methylumbelliferyl-B -£-glucopyranoside as the substrate 127 5-20 Inhibition of 3 -glucosidase by cellohexaose using 4-methylumbelliferyl-g -^-glucopyranoside as the substrate 129 5-21 Dixon replots of inhibition of 3 -glucosidase by cellobiose, cellotriose and cellotetraose .... 131 5-22 Dixon replots of inhibition of g -glucosidase by cellopentaose and cellohexaose 133 5-23 Subsite map for b -glucosidase constructed using values of V /Kjjj for cellooligosaccharides I38 5-24 Subsite map for 8 -glucosidase constructed using values of inhibition constants for cellooligosaccharides I4I xi

PAGE 12

6-1 Time course hydrolysis of [1 -^Hjcellotriose by CBH 1(D) ..151 6-2 Bond cleavage frequency plot for hydrolysis of [1-^H]cellotriose by CBH 1(D) 153 6-3 Effect of [1 --^Hjcellotriose concentration on product ratios 155 6-4 Eadie-Hof stee plot for hydrolysis of [1-^H]cellotriose hydrolysis by CBH 1(D) 157 6-5 Time course hydrolysis of [1--^H]cellotetraose by CBH 1(D) 161 6-6 Bond cleavage frequency plot for hydrolysis of [1-^H]cellotetraose by CBH 1(D) 163 6-7 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [1-^H]cellotetraose by CBHI(D) 165 6-8 Time course hydrolysis of [1 --^Hjcellopentaose by CBH 1(D) 167 6-9 Bond cleavage frequency plot for hydrolysis of [1-^H]cellopentaose by CBH 1(D) 170 6-10 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [1-^H]cellopentaose by CBH 1(D). ..172 6-11 Time course hydrolysis of [1-^H]cellohexaose by CBH 1(D) 174 6-12 Bond cleavage frequency plot for hydrolysis of [1--5H]cellohexaose by CBH 1(D) 176 6-13 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [ 1 -%]cellohexaose by CBH 1(D) 179 6-14 Bond cleavage frequencies of [1 --^Hjcellooligosaccharide hydrolysis by CBH 1(D) 181 6-15 Possible types of repetitive attack of CBH 1(D) with [1-^H]cellopentaose and [ '^C-U] cell open taose 188 6-16 Possible types of repetitive attack of CBH 1(D) with [l-^Hjcellohexaose and [^'^C-Ujcellohexaose .193 xil i

PAGE 13

6-17 Subsite map for CBH 1(D) constructed using values of V /K_, and bond cleavage frequencies for [ 1 --^H] eel iooTigosacchar ides 200 618 Subsite map for CBH 1(D) constructed using values of K^, V^^y and bond cleavage frequencies for tiHjcellooligosaccharides 203 71 Time course hydrolysis of [1 --^Hjcellotriose by CBH II 210 7-2 Bond cleavage frequency plot for hydrolysis of [1-'H]cellotriose by CBH II 212 7-3 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [1 -^Hjcellotriose by CBH II 214 7-4 Time course hydrolysis of [1--^H]cellotetraose by CBH II 217 7-5 Bond cleavage frequency plot for hydrolysis of [1-^H]cellotetraose by CBH II 219 7-6 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [1-^H]cellotetroase by CBH II 221 7-7 Time course hydrolysis of [ 1 --^Hjcellopentaose by CBH II 223 7-8 Bond cleavage frequency plot for hydrolysis of [1-^H]cellopentaose by CBH II 226 7-9 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [1-^H]cellopentaose by CBH II 228 7-10 Time course hydrolysis of [ 1 --^Hjcellohexaose by CBH II 230 7-11 Bond cleavage frequency plot for hydrolysis of [1— ^Hjcellohexaose by CBH II 232 7-12 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [1-^H]cellohexaose by CBH II 235 7-13 Bond cleavage frequencies of [1-^H]cellooligosaccharide hydrolysis by CBH II 237 7-14 Possible types of repetitive attack of CBH II with [1-^H]cellohexaose and [ ''+C-U]cellohexaose 247 xiii

PAGE 14

7-15 Subsite map for CBH II constructed using values of Vjjj /K_ and bond cleavage frequencies for [ eel loo iigosacchar ides 254 716 Subsite map for CBH II constructed using values of K^, V^j and bond cleavage frequencies for 1 1 -^Hjcellooligosaccharides 257 81 Structure of cellulose I and cellulose II 266 8-2 Orientation of glycosidic bonds in cellulose .... 269 A-1 Binding modes of cellotriose on a hypothetical four-subsite enzyme 288 A-2 Combination of V^j^j^/K values and bond cleavage frequencies for [1 --^HJcellotetraose and [1 — ^Hjcellotriose used to calculate subsite affinities 301 xiv

PAGE 15

LIST OF ABBREVIATIONS AND SYMBOLS A Substrate CBH Cellobiohydrolase CM-Cellulose Carboxy methyl cellulose G-j Cellooligosaccharide of chain length i HPLC High performance liquid chromatography ^cat Turnover number ^+2 Hydrolytic rate coefficient Kj^ Inhibition constant ^int Microscopic dissociaton constant for a binding mode in which the entire binding region is occupied Kjj Michaelis constant Kg Dissociation constant for substrate MUG 4-Methylumbelliferyl-3-D_glucopyranoside MeUmb S (Glc£_)n 4-Methylumbelliferyl-S-_D-glycosides from cellobiose (n=2) to cellohexaose (n=6) TLC Thin-layer chromatography V Measured velocity Vjjjgjj Maximum velocity XV

PAGE 16

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MODE OF ACTION OF EXOGLUCANASES FROM THE CELLULOLYTIC FUNGUS Trichoderma reesei : ACTIVITY ON REDUCING END-LABELED CELLOOLIGOSACCHARIDES AND TOPOGRAPHY OF ACTIVE SITES OF s -GLUCOSIDASE CELLOBIOHYDROLASE 1(D), AND CELLOBIOHYDROLASE II BY WILLIAM JOSEPH CHIRICO December, 1984 Chairman: Ross D. Brown, Jr., Ph.D. Major Department: Biochemistry and Molecular Biology The activity of purified 6 -glucosidase, cellobiohydrolase 1(D) and cellobiohydrolase II from Trichoderma reesei was studied using [1--^H]cellooligosaccharides. Reducing end-labeled cellooligosaccharides of high specific radioactivity were synthesized by reducing lactones of cellooligosaccharide aldonic acids with sodium boro[^H]hydride. A thin-layer chromatographic method for separating [1--^H]cellooligosaccharides with high resolution and quantitatively extracting them from silica gel was developed. The 0 -glucosidase was shown to bind [1-%]cellooligosaccharides in one productive mode and sequentially remove glucosyi residues from the nonreducing end. The 6 -glucosidase exhibited little tendency to repetitively attack [1--^H]cellooligosaccharides. Values of Y^^^ for hydrolysis xvi

PAGE 17

of [1--^H]cellooligosaccharides remained constant as chain length increased, whereas corresponding values of Kjjj decreased. Values of inhibition constants determined for cellooligosaccharide inhibition of 4-methylumbelliferyl-8-Dglucopyranoside hydrolysis were similar to corresponding values of K^. The active site of S-giucosidase was shown to comprise primarily three subsites, of which subsite 1 contributes the greatest proportion of binding energy. Cellobiohydrolase 1(D) was shown to bind [1-^H]cellooligosaccharides in more than one productive mode. Bond cleavage frequency analysis indicated that cellobiohydrolase 1(D) preferentially hydrolyzes glycosidic bonds at the reducing end. Product distribution resulting from activity on ['''^C-Ujcellooliogosaccharides and [1-^K]cellooligosaccharides suggests that cellobiohydrolase 1(D) repetitively attacks cellooligosaccharides from the reducing end. Values of V^^^^/K^j for [1 -^Hjcellooligosaccharide hydrolysis increase with increasing chain length. Subsite mapping procedures revealed that the active site of cellobiohydrolase 1(D) comprises seven subsites with catalytic groups located between subsites 5 and 6. Cellobiohydrolase II was also shown to bind [1-%]cellooligosaccharides in more than one productive mode; however, glycosidic bonds at the nonreducing end are preferentially hydrolyzed. Product distribution resulting from hydrolysis of [^^C-U]cellooligosaccharides and [l-^H]cellooligosaccharides suggests that cellobiohydrolase xvii

PAGE 18

II repetitively attacks cellooligosaccharides from the nonreducing end. Values of ^^Q^^^m [ 1 --^H] cellooligosaccharide hydrolysis increase with increasing chain length. Subsite mapping procedures show that the substrate binding region of cellobiohydrolase II comprises seven subsites with catalytic groups located between subsites 3 and 4. The results are discussed in terms of the individual roles of exoglucanases during the saccharif ication of cellulose. A mechanism by which cellobiohydrolase 1(D) and cellobiohydrolase II may synergistically degrade crystalline cellulose is proposed. xviii

PAGE 19

CHAPTER 1 INTRODUCTION The utilization of cellulose, the world's most abundant renewable carbon source, for fuel, food and chemical feedstocks has received increasing attention as depletion of nonrenewable resources continues. The amount of photosynthetically fixed carbon dioxide which is converted to cellulose has been estimated by Ghose (1) at lO^'' metric tons per year. Glucose, a product of cellulose hydrolysis, can serve as a substrate for fermentations to produce single-cell protein, ethanol and methane (2). Cellulose can be converted to soluble products by acid or enzymic hydrolysis. Although technology for acid hydrolysis exists, degradation of products, interaction of acid with noncellulosic substances in natural sources of cellulose, corrosion of equipment and high capital costs argue for development of alternative methods of conversion (3). Enzymic hydrolysis of cellulose has the advantage of efficiency and specificity; however, it is the most expensive process. Lignin and hemicelluloses, which are constituents of natural sources of cellulose, protect the substrate from enzymic hydrolysis and limit conversion to 50 percent (2). Removal of lignin and hemicelluloses from native cellulose or addition of hemicellulases is necessary to obtain higher yields of glucose (2). 1

PAGE 20

An excellent source of cellulolytic enzymes is the fungus Trichoderma reesei QM 9414, which possesses the complete array of enzymes required for the conversion of cellulose to glucose (4). The component enzymes of the cellulase system act synergistically, in order to convert native, crystalline cellulose to soluble products (5). Since the product of one enzyme may serve as a substrate for another, purification and characterization of each component is a prerequisite for a clear description of the enzymic hydrolysis of cellulose. Although the extracellular, depolymerizing enzymes (6,7) and 8-glucosidase (8) of T. reesei have been purified and partially characterized, little quantitative information is available regarding the affinity of cellulolytic enzymes for cellulose (or cellooligosaccharides). In addition, the mechanism by which cellulolytic enzymes act synergistically and the topography of their active sites remain unknown. Since cellooligosaccharides are symmetric, the precise glycosidic bonds cleaved by cellulolytic enzymes have not been determined. Furthermore, the direction from which exo-glucanases remove glucosyl or cellobiosyl residues from cellulosic substrates has not been determined. In this report, the synthesis of reducing end-labeled cellooligosaccharides and a quantitative method for separating them using thin-layer chromatography are described. The activity of cellobiohydrolase 1(D) [CBH 1(D)], cellobiohydrolase II (CBH II) and s-glucosidase of ^. reesei on

PAGE 21

3 these asymmetrically labeled substrates provides information regarding substrate affinities, action patterns, topography of active sites and the direction from which substrates are degraded. Structure of Cellulose Cellulose is a linear polymer of S-1 4-£-glucosyl residues. Four crystalline forms of cellulose (I-IV) have been recognized on the basis of infrared spectra and X-ray diffraction studies (9). The chains in each structure have approximately the same backbone conformation, with a unit of cellobiose repeating every 10.3 A (10). The distinguishing feature of the structures is the arrangement of adjacent chains. Cellulose I (native cellulose) is found in plant cell walls. Vhen native cellulose is dissolved and recrystallized, it is converted irreversibly to cellulose II, a form of cellulose more stable than cellulose I. Cellulose III and IV can be produced from cellulose I or II by treatment with liquid ammonia and hot glycerol, respectively (9). The most crystalline, native cellulose is found in the cell walls of the alga Valonia ventricosa The X-ray diffraction patterns of cellulose fibers isolated from cell walls of Valonia ventricosa indicate that intramolecular hydrogen bonds are located on both sides of the glycosidic linkage: one hydrogen bond between 02 '-H 06 and another between 03-H— 05' (9). The 06-H hydroxy 1 is bonded to 03 of an adjacent chain. Thus, native cellulose is seen as a series of hydrogen-bonded sheets of parallel chains which

PAGE 22

4 are held together by van der Waal forces between hydrophobic faces of sugar rings. In contrast. X-ray patterns of cellulose II indicate that chains are arranged in an antiparallel fashion. Cellulose II also has intramolecular hydrogen bonds between 02 '-H 06 and another between 03-H 05'. The orientation of -CH2OH groups allows formation of intermolecular hydrogen bonds between 06-H 03, 06-H 02 and 02-H 02'. The additional intermolecular hydrogen bonding accounts for the increased stability of cellulose II. Recently, Atalla and VanderHart (11) compared the solid-state carbon-13 nuclear magnetic resonance spectra of various native celluloses. The spectra were consistent with the presence of two distinct forms of cellulose, I^ and Ig, in each sample. They estimated that the relative amounts of Iq as a fraction of the total are Acetobacter xyllnium = Valonia ventricosa > cotton = ramie > regenerated cellulose I. Thus, la is dominant in bacterial and algal cellulose, whereas 1 3 is dominant in celluloses from higher plants. Carbon-13 NMR spectra of cellulose II closely resembled that of Ig. Cellulases The cellulase system comprises 1 4-S-_D-glucan-4-glucanohydrolases (EC 3.2.1.4), 1 4-S-_D-glucan cellobiohydrolases (EC 3.2.1.91) and S-glucosidases (EC 3.2.1.21), which together act sequentially and cooperatively to convert native, crystalline cellulose to oligosaccharides and glucose. Endoglucanases randomly attack internal glycosidic

PAGE 23

5 bonds of cellulose producing chain ends and soluble oligosaccharides. Cellobiohydrolases cleave cellobiosyl residues from ends of cellulose chains. Cellobiose, an inhibitor of the depolymerizing enzymes, and oligosaccharides are converted to glucose by the 6-glucosidase. Component enzymes have been purified and characterized from true cellulolytic organisms, such as Trichoderma viride (5,12-16), Trichoderma reesei (6,8,17), Trichoderma koningii (18,19) and Irpex lacteus (20). Endo-1 ,4,-B-^-glucanases catalyze the hydrolysis of internal glycosidic bonds of cellulose providing chain ends from which cellobiosyl residues are removed by exo-cellobiohydrolases (5). The most frequently used technique for evaluating endoglucanase activity is measurement of the reduction of viscosity of a carboxymethylcellulose (CMcellulose) solution (21). Endoglucanases are differentiated from exo-cellobiohydrolases according to the anomeric configuration and identity of products from hydrolysis of polymeric substrates and cellooligosaccharides (22). Endoglucanases proceed with retention of configuration and form a variety of products, whereas exo-glucanases proceed with inversion of configuration and cleave almost exclusively cellobiosyl residues from the ends of substrates (22). Gritzali (6) reported the purification of a single endoglucanase, which has a molecular weight of 45,200 and a neutral carbohydrate content of 14.2 percent (by weight), from extracellular filtrates of T. reesei. The

PAGE 24

6 endoglucanase accounts for 15 percent (by weight) of the extracellular protein (5). Purification of endoglucanases from T. reesei has also been reported by Hakansson et al. (17), Ladisch et al. (23) and duMee (24). Bhikhabhai and Pettersson (25) recently reported the partial sequence of an endoglucanase from T. reesei Cyanogen bromide peptides of the endoglucanase showed significant homology with the amino acid sequence of CBH I. Comparison of carbon-13 NMR spectra of three endoglucanases isolated by duMee indicates that oligosaccharides attached to endoglucanases are similar if not identical to those of CBH 1(D) and CBH II of T. reesei (24). The rate of hydrolysis of cellooligosaccharides by the endoglucanase from T. reesei increases from cellotetraose to cellopentaose ; however, this trend does not extend to cellohexaose (6). Cellotetraose is hydrolyzed to glucose, cellobiose and cellotriose. Cellotriose is resistant to further degradation and its large concentration relative to glucose is indicative of transglycosylation by the endoglucanase. A variety of products are formed from cellopentaose and cellohexaose. The endoglucanase readily catalyzes hydrolysis of CM-cellulose and phosphoric acid-swollen cellulose. Microcrystalline cellulose is resistant to attack, whereas filter paper discs are degraded to short fibers. The release of cellobiose from ends of cellulose chains is due primarily to the hydrolytic action of 1 4-S-^-glucan cellobiohydrolases. Other names for this enzyme include

PAGE 25

7 exo-glucanase and exo-cellulase (20). The activity of cellobiohydrolases is determined by measuring the reducing sugar released from a suspension of swollen or crystalline cellulose. Since each enzyme in the cellulase system produces reducing sugar, highly purified preparations are required for enzymic characterization. Cellobiohydrolases exhibit little activity toward CM-cellulose as measured by change in viscosity (5). Two immunologically distinct cellobiohydrolases [CBH 1(D) and CBH II] have been purified from extracellular filtrates of T. reesei grown on cellulose or incubated with sophorose (6). Mandels et al. (26) discovered that sophorose is a potent inducer of cellulase in T. viride QM 6a. Cellobiohydrolase 1(D) has a molecular weight of 53,200 (6), contains 5.9 percent (by weight) of neutral carbohydrate (24) and is isoelectric at pH 3.95 (7), whereas the corresponding values for CBH II are 54,700 (6), 18.1 percent (24) and pH 4.9 (6), respectively. Cellobiohydrolase 1(D) and CBH II constitute approximately 60 and 25 percent (by weight), respectively, of the extracellular protein produced by this fungus (5). Recently, duMee (24) reported that CBH 1(D) contains 6 or 7 oligosaccharide chains linked through mannose to the protein, whereas CBH II contains approximately 25 oligosaccharide chains also linked through mannose to the protein. Complete amino acid sequence analysis of CBH I revealed that most of the neutral carbohydrate is located in a short region near the carboxy terminus (27).

PAGE 26

8 The sites of attachment of three glucosamine residues were also determined. Fagerstam and Pettersson (7) reported that no homology exists within the first 20 amino acid residues of CBH 1(D) and CBH II, although both enyzmes have a pyroglutamyl residue at the N-terrainus of the polypeptide chain. The molecular cloning and characterization of the gene encoding CBH I of T. reesei have been reported by Shoemaker et al. (28). The nucleotide sequence shows that the gene contains two introns and that a signal peptide is located at the N-terminus. Enzymic properties of cellobiohydrolases purified from extracellular filtrates of T. reesei QM 9414 on insoluble and soluble substrates have been reported (6). Cellobiohydrolase II cleaves cellooligosaccharides approximately 10fold and polymeric substrates 2-fold more rapidly than CBH 1(D) (6). Cellobiohydrolase 1(D) and CBH II are optimally active at pH 5.2-5.6 and 4.9, respectively (6). Phosphoric acid-swollen cellulose is hydrolyzed more rapidly than microcrystalline cellulose by CBH 1(D) or CBH II (6). Cellobiohydrolase II produces exclusively cellobiose from phosphoric acid-swollen cellulose, whereas CBH 1(D) yields some glucose in addition to cellobiose. Carboxymethylcellulose is mostly resistant to attack by either enzyme. The two cellobiohyrolases show strong synergistic activity in the hydrolysis of crystalline cellulose, thereby demonstrating their essential role in the degradation of cellulose (7).

PAGE 27

9 The action patterns of CBH 1(D) and CBH II have been investigated using high performance liquid chromatographic (HPLC) analysis of products formed during enzymic hydrolysis of cellooligosaccharides (6). Cellobiohydrolase 1(D) catalyzes the hydrolysis of cellotriose forming glucose and cellobiose, whereas CBH II has little activity toward cellotriose. As the chain length of oligosaccharide substrate increases, CBH 1(D) produces cellobiose and increasing proportions of glucose and cellotriose. In contrast, CBH II yields exclusively cellobiose from cellotetraose and a mixture of cellobiose and cellotriose from cellopentaose and cellohexaose. Each cellobiohydrolase yields cellobiose from hydrolysis of cellohexaose without concomitant formation of cellotetraose, indicating a sequential cleavage of cellobiosyl residues. Cellohexaose is hydrolyzed 2-fold more rapidly than cellotetraose or cellopentaose by CBH 1(D) or CBH II. Hsu et al. (29), using a cellobiohydrolase isolated from T. reesei QM 9414, reported that Km's for cellotriose and cellotetraose are 0,2 and 0.08 mM, respectively. Cellotetraose was hydrolyzed 27-fold faster than cellotriose. Since low sensitivity of HPLC analysis of cellooligosaccharides precludes initial rate studies, Hsu et al. (29) estimated kinetic constants from analysis of the entire time course of the reaction. Recently, van Tilbeurgh et al. (30) reported the action patterns of CBH 1(D) on a series of 4-methylumbellif eryl

PAGE 28

10 glycosides derived from cellooligosaccharides. Cellobiohydrolase 1(D) preferentially cleaves the glycosyl bond at the reducing end of the modified cellooligosaccharides. Whether the modification influences the action pattern of CBH 1(D) is not known. Cellobiohydrolases are also characterized by their affinities for inhibitors and substrates. Inhibition of depolymerizing enzymes by cellobiose is an important regulatory feature of the cellulase system. Inhibition constants of 2.1 and 0.2 mM for glucose and cellobiose, respectively, have been reported using a cellobiohydrolase from T. reesei (29). Using 4-methylumbelliferylcellobioside as the substrate. Brown and Greenberg (31) determined that glucose, cellobiose, cellotriose, ceilotetraose cellopentaose and cellohexaose inhibit and bind to one site on CBH 1(D) with inhibition constants of 0.476 M, 90.4 uM, 28.4 i^M, 14.1 pM, 14.6 m and 7.6 UM, respectively. Similar information was not obtained for CBH II, because this enzyme exhibits little activity toward 4-methylumbelliferylcellobioslde. The B-glucosidases catalyze the hydrolysis of aryl-6-Dglucopyranosides and 6-linked glucose oligomers as well as transfer of glucosyl residues (32). Although the hydrolysis of dimers, trimers and tetramers of glucose is catalyzed by glucosidases and exo-glucanases, glucosidases are often reported to act more rapidly on short oligomers, whereas exo-glucanases prefer longer oligomers (22). Glucosidases proceed with retention of configuration, whereas

PAGE 29

11 exo-glucanases proceed with inversion (22). The role of 6glucosidase within the complement of cellulase enzymes is to cleave cellobiose, an inhibitor of the depolymerizing enzymes, to glucose (5,33). Glucosidases have been purified from extracellular filtrates of several cellulolytic fungi, such as T. viride (16,34,35), T. koningii (19) and T. reesei (8). The 6glucosidase purified from T. reesei (8) has a molecular weight of 74,600, contains 0.7 percent (by weight) carbohydrate and is isoelectric at pH 8.5. The B-glucosidase constitutes about 0.4 percent (by weight) of the extracellular protein produced by this fungus. Transglucosylation reactions and hydrolysis of cellobiose, _£_-nitrophenylg_£_glucopyranoside and 4-methylumbellif eryl6-_D-gluconpyranoside are catalyzed by the 6-glucosidase The enzyme has an unusually high affinity for its product, _D-glucose, and is strongly inhibited by ^-glucono-1 5-lactone nojirimycin and 1 -deoxyno jirimycin. Most cellulolytic fungi have similar cellulase systems containing one to several 8-glucosidases, endo8-glucanases and exo-6-glucanases that act synergistically to hydrolyze insoluble cellulose (36). Synergism resulting from the concerted action of endoglucanases and exoglucanases has been explained by the production by endoglucanases of free polymer chain ends which serve as substrates for exoglucanases. Recently, strong exo-exo synergism has been reported for CBH 1(D) and CBK II from T. reesei acting on crystalline

PAGE 30

12 cellulose (7). Fagerstam and Pettersson (7) offered two possible explanations for exo-exo synergism: (1) the synergism may reflect a structural asymmetry in the substrate or (2) a change in the mode of action resulting from the formation of a complex between the two enzymes. However, no evidence for complex formation has been reported. Subsite Mapping of Enzymes Studies of the action pattern of polysaccharide hydrolases using their natural polymeric substrates led to the concept that glycosyl residues distant from the point of catalysis interact with the enzyme (37). Thus, the substrate binding region of the enzyme is considered to be an array of tandem subsites, each of which interacts independently with a specific glycosyl residue and with a certain affinity. The location of catalytic groups between two of the subsites, the number of subsites, the subsite affinity and the change in the unitary binding free energy at a subsite are characteristic features of polysaccharide hydrolases. Subsite models have been described to account for the enzymic properties of proteinases (38), nucleases (39) and carbohydrases (40-43). The literature offers three methods to determine experimentally the number of subsites comprising the binding region, the energetics of interaction of each subsite with a glycosyl residue and the hydrolytic rate coefficients (37,43-45). A synopsis of equations most relevant to each

PAGE 31

13 subsite-raapping procedure is found in Appendix A. Presented below is a brief desription of each method. Hiromi and coworkers (44,46) applied the subsite model to exo-enzymes by using the first-order rate constant ( Vjjjgjj/Knj) and Km for each of a series of oligosaccharides. A substrate can be bound to an exoenzyme in a variety of binding modes; however, only one mode is productive. Since hydrolytic rate coefficients for each oligosaccharide are assumed equal, "^^g^y^/^^ values are directly proportional to the association constant of the single productive complex. Michaelis constants are assumed to approximate association constants. Thus, the ratios of Vjjj^^/K^, values for n-mer and n+1-mer substrates permit the estimation of the n+1-th subsite affinity. Using this method, the substrate binding region of glucoamylase from Rhizopus sp. was shown to consist of seven subsites, of which subsite II contributes the greatest proportion of the binding energy. Recently, Koyama et al. (47) applied Hiromi 's method to glucoamylase from Aspergillus salto This enzyme was also shown to consist of seven subsites, of which subsite II is the major contributor to the binding energy. Although the method of Hiromi (46) is suitable for mapping exo-enzymes having one productive complex, it is not applicable to enzymes, such as endoenzymes, having more than one productive complex. Thoma and coworkers (43,48) have proposed a method to evaluate subsite affinities of endoenzymes from product distribution ratios of end-labeled

PAGE 32

14 oligosaccharides. Enzymic processes are separated into microscopic and macroscopic events. A microscopic coefficient characterizes one particular binding mode of a substrate. Each binding mode has a microscopic association constant, and each productive complex has a microscopic hydrolytic rate coefficient. Although microscopic constants are difficult to measure, they are related to measurable macroscopic parameters, which are a function of all binding modes of the substrate. The experimentally accessible parameters that are useful in subsite mapping are Km, Vmax, first order rate constant and bond cleavage frequencies measured as a function of chain length. Ratios of bond cleavage frequencies can be used to measure the size of the binding region, to locate the catalytic groups between two of the subsites and to calculate the apparent binding energies for some of the subsites. Allen and Thoma (49) using bond cleavage frequency analysis reported that the substrate binding region of Bacillus amyloliquef aciens aamylase is composed of 10 subsites with catalytic groups located between subsites VI and VII. Suganuma et al. (45) have proposed a method for evaluating subsite affinities that retains the advantages of Hiromi's and Thoma's methods. Suganuma et al. (45) reported that the substrate binding region of Taka-amylase A consists of 9 subsites with catalytic groups located between subsites 4 and 5. The subsite binding region of an endo-1,4-0xylanase from Cryptococcus albidus was examined using the

PAGE 33

15 method of Suganuma et al. (45). The enzyme was shown to consist of 4 subsites with the catalytic groups located in the center (42). Assessment The mechanisms by which cellulolytic enzymes attack cellulose remain poorly understood, because the substrate, cellulose, is insoluble and heterogeneous. To define the mode of attack of cellulolytic enzymes, soluble substrates, such as cellooligosaccharides, should be used in enzymic studies. Although previous studies of cellulolytic enzymes using cellooligosaccharides as substrates provide information regarding rates of hydrolysis, quantitative estimates of the affinity of the enzymes for these substrates were not possible due to low sensitivity of the assays. In addition, the symmetric nature of cellooligosaccharides precludes identification of the precise glycosidic bonds cleaved by the enzymes. In this report, the synthesis of reducing end-labeled cellooligosaccharides and a thin-layer chromatographic method for separating and quantitating these substrates are described. Reducing end-labeled cellooligosaccharides were used to identify the precise glycosidic bond(s) cleaved by 8-glucosidase, CBH 1(D) and CBH II of T. reesei These substrates were also used in initial rate studies to determine the affinity and maximal rates of hydrolysis by cellulolytic enzymes. Cellooligosaccharides uniformly labeled with carbon-14 were used in combination with reducing

PAGE 34

16 end-labeled cellooligosaccharides to determine the direction from which cellulolytic enzymes attack. A subsite map for each enzyme was constructed from bond cleavage frequencies, Km and Vmax for the series of cellooligosaccharides. The results are discussed in terms of the role of each enzyme in the conversion of cellulose to glucose. A mechanism by which CBH 1(D) and CBH II may exert exo-exo synergism is proposed.

PAGE 35

CHAPTER 2 EXPERIMENTAL PROCEDURES Materials Enzymes Cellobiohydrolase 1(D) and CBH II, from extracellular culture filtrates of T. reesei QM 9414, were purified by Gritzali (5). Extracellular 3-glucosidase was purified from T. reesei as described previously (8). Statzyme-Glucose 16 was purchased from Worthington Biochemical Corporation, Freehold, NJ. Substrates Cellobiose was obtained from Schwartz/Mann, Orangeburg, NY. Sophorose was obtained from Adams Chemical Co,, Round Lake, XL, Laminaribiose was prepared by hydrolysis of laminarin and separated on a column of Bio-Gel P-2 by B. Greenberg in this laboratory. The 4-methylumbelliferyl-0-^-glucopyranoside (lot 8650263), £.-nitrophenyl-e-_D-glucopyranoside (lot 88C-5039), d.nltrophenyl-8-D_-galactopyranoside (lot 58C-5063), £.-nitrophenyl-8-D_-xylopyranoside (lot 117C-0369), _£-nitrophenylaD_-glucopyranoside (lot 59B-0100), _2.-nitrophenyl-1-thio-8-^glucopyranoside and gentiobiose (lot 35C-0173) were obtained from Sigma Chemical Company, St. Louis, MO. 17

PAGE 36

18 Chemicals Sodium boro[^H]hydride (18.1 Ci/mmol, lot 25 and 12.1 Ci/mmol, lot 26) was obtained from Amersham, Arlington Heights, XL. The [''^C-Ujcellulose ( Nicotiana Tobacum L. 34.5 uCi/mg, lot 1396-151) and Aquasol were purchased from New England Nuclear, Boston, MA. Cellulose (SIGMACELL Type ICQ, lot 129C-0079) was obtained from Sigma Chemical Co., St. Louis, MO. Dextrose was obtained from the National Bureau of Standards, Washington, DC. Bromine (lot 22460) and Celite 545 were obtained from J.T. Baker Chemical Co., Phillipsburg, NJ. Calcium cellobionate (lot 101207) was obtained from ICN Pharmaceuticals, Inc., Cleveland, OH. Lactobionic acid (calcium salt) was obtained from Nutritional Biochemicals Corp., Cleveland, OH. The 2propanol (certified ACS), acetonitrile (HPLC grade), ethyl acetate (certified ACS), methanol (HPLC grade), dimethyl sulfoxide (certified ACS), glucono-1 5-lactone (lot 793699) and _£^anisaldehyde were obtained from Fisher Scientific, Fair Lawn, NJ. All other chemicals were reagent grade. Chromatographic Supplies The LK5 (5 x 20 cm), LK6 (5 x 20 cm) and LK5D (20 x 20 cm, 250 um) (lot 002431) Silica Gel TLC plates, Whatman 3 MM paper and Whatman Partisil PXS 10/25 PAC column (bonded cyano-amino type, polar phase) were obtained from Whatman Chemical Separations Division, Clifton, NJ.

PAGE 37

19 Silica Gel IB2 Baker-flex (20 x 20 cm) TLC plates were purchased from J.T. Baker Chemical Co., Phillipsburg, NJ. Cellulose (20 x 20 cm) (lot 6214) TLC plates were obtained from Eastman Kodak Company, Rochester, NY. Polyamide Layer Sheets (15 x 15 cm, 50 jjm) were obtained from Accurate Chemical and Scientific Corp., Hicksville, NY. Amberlite IRA-45 (-0H) Anion Exchange Resin was obtained from Sigma Chemical Co., St. Louis, MO. Amberlite IR-120 (H+) Cation Exchange Resin was obtained from Fisher Scientific, Fair Lawn, NJ. Amberlite MB-3 Mixed Cation-Anion Exchange Resin was obtained from Mallinckrodt Inc., Paris, KY. Bio-Gel P-2 (200-400 mesh) and Bio-Gel P-2 (-400) mesh were purchased from BioRad Laboratories, Richmond, CA. Millipore filters (0.22 v) were purchased from Millipore Corp., Bedford, MA. Molecular sieves (3 A) were obtained from Davison Chemical, Baltimore, MD. Methods Carbohydrate Determination Total neutral carbohydrate was determined by the phenol-sulfuric acid method of Dubois et al. (50). Alternatively, carbohydrate was determined by HPLC using differential ref ractometry. Reducing sugar was determined using the method of Nelson (51) and Somogyi (52). Glucose was used as a standard.

PAGE 38

20 Protein Determination Protein content was determined according to the method of Lowry et al. (53) as described by Bailey (54). Bovine serum albumin was used as a standard. Protein concentration of CBH 1(D) was determined from the extinction coefficient, 13.8 (24). High Performance Liquid Chromatography (HPLC) Instrumentation Cellooligosaccharides were separated and purity assessed using a Waters Association Model ALC 202/401 Liquid Chromatograph supplemented with a Model 6000 Solvent Delivery System (Waters Associates, Inc., Milford, MA). Column effluents were monitored using a differential ref ractometer and a Spectra-Physics Autolab System I integrator (Spectra-Physics, Santa Clara, CA) computed relative area of chromatographic peaks. Reverse phase chromatography Cellooligosaccharides were separated by reverse phase chromatography using a Whatman Partisil PXS 10/25 PAC column with an acetonitrile:water (75:25; v:v) solvent system (55). Cellooligosaccharides were eluted isocratically from the column at a flow rate of 1.5 ml/min at room temperature. Bio-Gel P-2 chromatography Cellooligosaccharides were also separated by gel permeation chromatography using a Bio-Gel P-2 (-400 mesh) column (0.75 x 240 cm). Cellooligosaccharides were eluted with water at a flow rate of 0.7 ml/min at 60c.

PAGE 39

21 Preparation of Celloollgosaccharldes Cellulose powder (15 g) was dissolved at 4C in 200 ml of concentrated HCl (-30C) (56). The slurry was stirred for 20 min at 4C and then for 2 h at room temperature. After the hydrolysate was added to 600 ml of water (4^C), insoluble cellulose was removed by centrifugation (Sorvall RC-5B Centrifuge, Du Pont, Co., Newtown, CT) and the pellet was washed with 2 aliquots (100 ml) of water. The supernatants were combined, filtered (Whatman No. 1 paper) and neutralized using anion exchange resin (Amberlite IR-45 [-0H]). Cellooligosaccharides were concentrated in a rotary evaporator (Buchler Instruments, Fort Lee, NJ) and then fractionated on a Bio-Gel P-2 (200-400 mesh) column (4.4 x 115 cm). Approximately 200 g of cellulose powder yielded at least 500 mg of cellotriose through cellohexaose Carbohydrates in column effluents were monitored using the method of Dubois et al. (50). Oxidation of Cellooligosaccharides Bromine oxidation The method of Diehl et al. (57) was modified as described below and used to oxidize purified cellotriose through cellohexaose. Cadmium carbonate (0.45 g, 2.7 mmoles) was added to a solution of purified cellotriose (0.449 g, 0.89 mmoles) in water (19 ml). Bromine (0.028 ml, 1.1 mmoles) was added and the suspension was stirred in a stoppered flask in the dark for 24 h at room temperature. After the mixture was filtered through Celite, silver carbonate (0.32 g, 1.2 mmoles) was added to the

PAGE 40

22 filtrate and the suspension was stirred for 20 min. The reaction mixture was filtered through Celite and the filtrate treated with an excess of hydrogen sulfide. The precipitate was removed by filtration through Celite and the filtrate freed of residual hydrogen sulfide by aeration. The filtrate was neutralized with calcium carbonate (58), concentrated to approximately 3 ml and filtered through a Millipore filter. The calcium salt of cellotrionic acid was separated from cellotriose and by-products of the reaction using Bio-Gel P-2 chromatography. Fractions containing calcium cellotrionate were combined, concentrated to a syrup and the product precipitated and washed with 2-methoxyethanol. Alternatively, isolation of calcium salts of cellooligosaccharide aldonic acids may be omitted and aldonic acids may be lactonized directly. Oxidation with Adams' catalyst Cellobiose (1 g, 1.5 mmoles) was oxidized in the presence of Adams' platinum oxide catalyst (0.5 g) and gaseous oxygen in aqueous solution (10 ml) at pH 3.7 as described by Conchie et al. (59) for oxidation of monosaccharides. Lactonization of Cellooligosaccharides Aldonic Acids Calcium cellotrionate (0.357 g) in 2 ml of water (4C) was passed through a column containing 3 ml of cation exchange resin (Amberlite IR-120 [+H]) and the column washed with 30 ml of cold water (60). The effluent was concentrated to a thin syrup at 50C under reduced pressure in a rotary evaporator and dried by successive evaporation in 6

PAGE 41

23 ml of a mixture of dry 2-methoxyethanol and toluene (2:1) followed by evaporation to a syrup. The drying procedure was repeated three times. During the synthesis of cellotetranolactone cellopentanolactone and cellohexanolactone precipitation of product before removal of water was prevented by using dimethyl sulfoxide in place of 2-methoxyethanol. After 5 ml of 2-methoxyethanol was added, cellotrionolactone precipitated from the syrup and the resulting suspension was centrifuged. Filtration was avoided, because precipitated lactones were hygroscopic at this stage of the synthesis. The pellet was washed with 3 aliquots (5 ml) of 2-methoxyethanol and then dried under reduced pressure at room temperature. Reduction of Aldonic Lactones with Sodium Borohydride Several procedures for reducing aldonic lactones to aldoses were evaluated using glucono-1 5-lactone lactobionolactone or cellobionolactone as the starting compound. The procedures included reduction of lactones in the presence of cation exchange resin, boric acid, carbon dioxide or sodium acid oxalate as described by Frush and Isbell (61). Alternatively, 1.0 N H2SO4 was added to reaction mixtures to maintain weakly acidic conditions during reductions as described by Wolfrom and Wood (62).

PAGE 42

24 Reduction of Lactones of Cellooligosaccharides with Sodium Boro[-^H]hydride Reduction in the presence of carbon dioxide Under an efficient hood, 10 ml of water was cooled to 4C and a stream of carbon dioxide was bubbled through the solution. After the hydrogen ion concentration of the solution was adjusted to pH 6-7 (pH paper) with 1.0 N NaOH, 100 mg of cellotrionolactone was added and the solution was stirred. Immediately after cellotrionolactone was dissolved, 250 ul (5 mCi) of sodium boro[^H]hydride (12.1-18.1 Ci/mmole) was added dropwise during 15 minutes. After an additional 15 minutes, the hydrogen ion concentration was adjusted to pH 2.5 with 1.0 N sulfuric acid and the solution stirred for 15 minutes. The solution was passed through a column containing 3 ml of cation exchange resin (Amberlite IR120 [+H]) and the effluent was concentrated to a syrup under reduced pressure in a rotary evaporator. The syrup was evaporated three times from 5 ml of water followed by evaporation of three 5 ml aliquots of methanol to remove borate esters. The resulting powder was dissolved in 1 ml of water and the [1-^H]cellootriose was separated from unreacted lactone and by-products of the reaction using Bio-Gel P-2 HPLC. To prevent radiochemical decomposition, column effluents containing radioactive cellooligosaccharide were adjusted immediately to 10 percent athanol and stored at -700c, Unreacted lactone was converted to the corresponding calcium

PAGE 43

25 salt and reused for future reductions. Chemical and radiochemical purity of [1--^H]cellooligosaccharides were determined using thin-layer chromatography, reverse phase HPLC and Bio-Gel P-2 HPLC. Specific radioactivity was determined by measuring radioactivity by scintillation counting and carbohydrate by Bio-Gel P-2 HPLC, reverse phase HPLC or the method of Dubois et al. (50). Reduction in the presence of sodium acid oxalate Glucono-1 5-lactone (100 mg, 0.56 mmoles) was added to 20 ml of water (4C) containing 7.3 mg (0.56 mmoles) of sodium acid oxalate. Under an efficient hood, 250 yl (5 mCi) of sodium boro[^H]hydride (16.4 Ci/mmole) and then sodium borohydride (4.2 mg, 1.1 mmoles) were added dropwise to the solution. After 0.5 h, the hydrogen ion concentration was adjusted to pH 2.5 with 1.0 N H2SO4. The solution was stirred for 15 min and then neutralized with mixed-bed ion exchange resin (Amberlite MB-3). After removing ion exchange resin by filtration, the filtrate was concentrated to a syrup using a rotary evaporator. The syrup was dissolved in 75 ml of water and concentrated to remove tritiated water. Volatile and nonvolatile radioactivity was measured by scintillation counting. Specificity of Radiolabeling To determine the location of tritium in [^Hjcellooligosaccharides, [^Hjcellotriose (0.28 yCi) was diluted with cold cellotriose to a total of 4.I mg (8.1 umoles) and oxidized with an excess of bromine (20 ul, 390 ymoles).

PAGE 44

26 After the solution was stirred at room temperature for 24 h, water was removed using a rotary evaporator. Radioactivity in water and oxidized products was measured by scintillation counting. Measurement of Radioactivity Radioactive samples were dissolved in 5 ml of Aquasol and counted using a Beckman LS-9000 liquid scintillation counter (Beckman Instruments, Inc., Palo Alto, CA). All radioactive measurements were automatically corrected for quenching. Aquasol was acidified (0.4 percent) with glacial acetic acid to retard chemiluminescence. Preparation of [ ''^C-U]Cellooligosaccharides Batch method The ['''^C-Ujcellulose (750 uCi, 34.5 uCi/mg) was added to concentrated HCl (3 ml), which had been stored at -30C. After the slurry was stirred at room temperature for 3 h, 3 ml of cold water (4C) was added and the solution was neutralized with sodium bicarbonate. The reaction mixture was centrifuged and the supernatant was filtered through a Millipore filter. The [ '''^C-U]cellooligosaccharides were separated on a Bio-Gel P-2 (200-400 mesh) column (4.4 x 115 cm). Isolated [ ''^C-Ujcellooligosaccharides were rechromatographed on the same column or using Bio-Gel P-2 HPLC until each [ 'I ^C-Ujcellooligosaccharide was greater than 99 percent pure as judged by scintillation counting. Specific radioactivity of [''4c_u]cellooligosaccharides was assumed equal to that of the

PAGE 45

27 [''4c-U]cellulose. Purified ['•^C-Ujcellooligosaccharides were stored in aqueous solution at -30C. Column method The following procedure was used to increase the yield of ['''^C-Ujcellooligosaccharides of longer chain length (4-6) from acid hydrolysis of '^C-U] cellulose. During a 3 h period, 90 ml of concentrated HCl (4C) were passed through a column (0.6 x 7 cm) containing a mixture of [''4c-U]cellulose (250 uCi, 34.5 yCi/mg) and fine glass beads. The effluent was collected in a flask containing 50 ml saturated sodium bicarbonate solution. The hydrogen ion concentration of the solution in the collection flask was maintained at approximately pH 7 with the addition of sodium bicarbonate. The effluent was desalted and the [''4c_u]cellooligosaccharides were separated on a Bio-Gel P-2 (200-400 mesh) column (4.4 x 115 cm). Fractions containing soluble cellooligosaccharides of chain length greater than two were combined and rechromatographed through the same column. Radioactivity in column effluents was measured by liquid scintillation counting. Thin-Layer Chromatography (TLC) Separation of cellooligosaccharides Aqueous samples (5-15 yg) of cellooligosaccharides in 1-2 yl aliquots were applied to TLC plates. To a developing tank (7 x 27 x 25 cm) lined with two sheets of Whatman 3 MM paper (25 x 30 cm), 50-100 ml of the appropriate solvent was added; "che paper was saturated with the same solvent and the tank equilibrated for 0.5 h. Plates were developed twice with

PAGE 46

28 solvent A (ethyl acetate rwater : isopropanol ; 2:1:2; v:v:v) (63) solvent B (ethyl acetate : water : isopropanol ; 40:25:27; v:v:v), solvent C (ethyl acetate :water :n-propanol ; 40:30:34; v:v:v), solvent D (ethyl acetate : methanol :acetic acid:water; 2:1:1:1; v:v:v:v), solvent E (acetonitrile :water ; 70:30; v:v), solvent F (ethyl acetate : dimethyl sulfoxide :water : isopropanol; 60:40:20:20; v:v:v:v), solvent H (ethyl acetate:water:methanol; 1:2:1; v:v:v) or I (ethyl acetate :water: methanol; 2:1:1; v:v:v). Plates were developed until the solvent front migrated to 0.5 cm from the top of the plate and then dried at 110C for 5 min between developments. Carbohydrates were stained with potassium dichromate reagent or silver nitrate reagent (64) In some instances, before applying samples silica gel plates were immersed in 0.03 M sodium borate, sodium metabisulfite, sodium acetate or 0.2 M sodium phosphate for 5 min and then dried for 1 h at 110C (65). Separation of [1-^H]cellooligosaccharides Aqueous samples (20-30 yl) of [1-^H]cellooligosaccharides were applied to preadsorbent zones of Whatman LK5D silica gel plates. After samples were applied, the plate was dried in an efficient hood with cool, dry air for 45 min. To remove traces of water, the preadsorbent zone was saturated with several drops of acetone and the plates were dried for an additional 15 min. On humid days, plates were dried at room temperature under reduced pressure for an additional 30 min. After the developing tank was lined with two sheets of

PAGE 47

29 Whatman 3 MM paper, 30 ml of solvent G (ethyl acetate: water: methanol; 40:15:20; v:v:v) were added to the tank. Whatman 3 MM paper was saturated with solvent G and the tank equilibrated for 0.5 h. Plates (maximum of 2 per tank) were developed until the solvent front migrated to 0.5 cm from the top of the plate (1.5 h). The plates were removed from the tank and dried in an efficient hood with dry, cool air or under reduced pressure at room temperature for 0,5 h. While the plates were drying, solvent in the tank was replenished. The plates usually were developed 3 times, after which they were dried at 110C for 5 min. However, when initial rates of enzymic hydrolysis of [1-^H]cellohexaose were to be determined, plates were developed a total of 4 times to assure baseline separation between [ 1 -^Hjcellopentaose and [1-^H]cellohexaose. After channels containing radioactive cellooligosaccharides were covered with Whatman 3 MM paper, channels containing cellooligosaccharide standards were stained with ^-anisaldehyde reagent (64). Extraction of a mixture of [1-^H]cellooligosaccharides from thin-layer chromatographic plates A mixture containing equal amounts of radioactivity of [ 1 -^Hjglucose through [l-^H]cellohexaose was applied to Whatman LK5 TLC plates. The plates were dried under an efficient hood for 1 h with dry, cool air or in an oven at 110C for 15 min. Areas containing radioactive samples were scraped with a single-edged razor blade onto weighing paper and then transferred into 7 ml scintillation vials. The following

PAGE 48

30 procedures for extracting [1--^H]cellooligosaccharides from scrapings were evaluated. Method A: Water (0.5 ml) was added to vials and then vials were placed in boiling water for 15 min. Method B: Dimethyl sulfoxide : water (1:1; v:v) (0.5 ml) was added to vials and then vials were placed in boiling water for 15 min. Method C: Water (0.5 ml) was added to vials and then vials were sonicated for 0.5 h. Method D: Aquasol (5 ml) and 20 ul of glacial acetic acid were added directly to vials containing scrapings. Aquasol (5 ml) and 20 ul of glacial acetic acid were added to each sample and the vials were shaken vigorously. Radioactivity was measured using a scintillation counter to 2 percent sigma error or for a maximum of 2 min. Extraction of individual [1-^H]cellooligosaccharides Samples (2.5-10 yl) of each [1--^H]cellooligosaccharide containing equal amounts of radioactivity were applied to Whatman LK5D TLC plates and then the plates were dried at 110C for 0.5 h. Areas containing individual [1-^H]cellooligosaccharides were scraped with a single-edged razor blade onto weighing paper and then transferred to a 7 ml scintillation vial. For Method A, water (0.5 ml) was added to the vial and the vial was then vortexed for 30 s. For Method B, water (1.5 ml) was added to the vial and then the

PAGE 49

31 vial was vortexed for 30 s. For Method C, water (0.5 ml) was added to the vial and then the vial was sonicated for 15 min. After samples were treated as described above, Aquasol (5 ml) and 20 ul of glacial acetic acid were added to the slurry and the vials were shaken vigorously. A gel formed when samples treated according to Method B were shaken. Samples were counted in a scintillation counter to 2 percent Sigma error or a maximum of 2 min. Enzymic Assays g-Glucosidase Aryl-8-D^-glycosidase activity was determined by measuring the release of _£;-nitrophenol from _2.nitrophenyl-_D-glycosides after incubation with purified eglucosidase (66). Samples were incubated at 40C for 20 or 30 min in 2.5 ml of 50 mM sodium acetate buffer, pH 5.0, which contained 3 mM sodium azide and 2.4 mM £_-nitrophenyl_D-glycosides. The reactions were terminated by immersing the tubes containing the assay mixture in a boiling water bath for 5 min. After the addition of 1.0 ml of 7.5 percent (w/v) potassium phosphate (K2HPO4), the absorbance of nitrophenol at 400 nm was measured. Specific activity was expressed as ymoles of ^-nitrophenol released/min/mg of protein. The absorbance was linear from 5 to 60 yg of £nitrophenol. Activity of the g-glucosidase on 0-linked glucose disaccharides was determined by measuring the production of glucose from sophorose, laminaribiose and gentiobiose. Samples were incubated at 40C for 30 min in 0.5 ml of

PAGE 50

32 sodium acetate buffer, pH 5.0, which contained 3 mM sodium azide and various concentrations of substrate. The reaction was stopped by immersing tubes containing the assay mixture in a boiling water bath for 5 min. Glucose was measured by a modification of the Statzyme Glucose Reagent method as described by Worthington Biochemical Corporation (67). The absorbance of NADH at 340 nm was measured after the addition of 100 ul of assay mixture to 1.0 ml of Statzyme Glucose Reagent. The absorbance measured was linear from 5 to 50 yg of _D-glucose. Specific activity was expressed as umoles of disaccharide hydrolyzed/min/mg of protein. Initial rates of hydrolysis of 4-methylumbelliferyl-S^-glucopyranoside catalyzed by S-glucosidase were measured by continuously monitoring the formation of 4-methylumbelliferone. Samples were incubated at 40C in 1.0 ml of 50 mM sodium acetate buffer, pH 5.0, which contained 3 mM sodium azide and various concentrations of 4-raethylumbelliferyl-3D[;-glucopyranoside. The absorbance due to 4-methylumbelliferone at 346 nm was measured continuously as described by Rosenthal and Saifer (68). The absorbance response was linear in the range of 0.05 to 0.50 umole of 4-methylumbellif erone Specific activity was expressed as umoles of 4-methylumbelliferone released/min/mg of protein. Hydrolysis of labeled cellooligosaccharides Activities of 8-glucosidase, CBH 1(D) and CBH II on [1-^Hjcellooligosaccharides were determined by measuring the change in the distribution of radioactive products and substrates

PAGE 51

33 during the time course of reactions. The 6-glucosidase CBH 1(D) or CBH II were incubated at 40C in 100 ul of 5 mM sodium acetate buffer, pH 5.0, which contained 0.3 mM sodium azide and 0.1-0.3 yCi of [1-%]cellooligosaccharide. When activities of exoglucanases on [ 1 --^Hjcellooligosaccharides and [^"^C-Ujcellooligosaccharides were determined, 0.19-0.22 jiCi of [^^C-Ujcellooligosaccharide were included in the assay mixture. Total concentration of cellooligosaccharide in assay mixtures was adjusted with corresponding unlabeled cellooligosaccharide. The concentration of enzyme in an assay was chosen to give approximately 50 percent degradation of substrate in 30 min. Assay mixtures were contained in 2 ml conical, glass centrifuge tubes. The reaction was stopped at various times by mixing a 10 ul sample of the assay mixture with 10 yl of 2 N HCl. The volume of sample removed from the assay mixture was chosen to give sufficient counts to detect 0.5 percent hydrolysis at a single bond of a labeled cellooligosaccharide. Labeled cellooligosaccharides in samples removed from assay mixtures were separated using thin-layer chromatography as described above. Labeled cellooligosaccharides were extracted routinely from silica gel scrapings by adding 1.5 ml of 2 percent glacial acetic acid and vortexing for 15 s. Aquasol (5 ml) was added and the mixture was shaken vigorously until a gel formed. Radioactivity was measured in a scintillation counter to 2 percent sigma error or a maximum of 5 min.

PAGE 52

34 Kinetic Analysis of Hydrolysis Data Allen (69) showed that bond cleavage frequencies are best evaluated by plotting product ratios (radioactivity of each product divided by total radioactivity of the sample) vs. extent of reaction (total radioactivity of the products divided by total radioactivity of the sample). Bond cleavage frequencies were determined from the slope of product ratio vs. extent of reaction plots. Subtraction of sample background is unnecessary, because zero-time sample background affects only the intercept of the plot. Initial rates of hydrolysis of [1 -^H]cellooligosaccharides were estimated from the slope of a line through the linear, initial region of time course plots. Time course plots were constructed by plotting the amount of [1-^H]cellooligosaccharides (moles) vs. time (min). The concentration of a [1--^H]cellooligosaccharide was determined by multiplying the product ratio by the initial concentration of substrate. Specific activities were expressed as ymoles of [1-^H]cellooligosaccharide hydrolyzed/min/mg of protein. Sample calculations for determining bond cleavage frequencies and initial rates of hydrolysis are provided in Appendix B. Determination of Kinetic Constants Values of K^,, V^^^ and ^^g^^/K^ for activity of exoglucanases were estimated from initial rates of hydrolysis using HYPER Fortran program described by Cleland (70). The HYPER program was also used to determine apparent values of

PAGE 53

35 Kjjj and 'V^g.^ for competitive inhibition plots. The activity of CBH 1(D) on [1 --^Hjcellotriose was analyzed using Program Two/One kindly provided by W.W. Cleland. Kinetic equations used in the analyses are listed in Appendix C. Evaluation of Subsite Affinities Subsite affinities of the 8-glucosidase were determined in part from the values of Vjj,g^jj/Kjj, obtained for cleavage of [1-^H]cellotriose through [1-^H]cellohexaose as described by Hiromi et al. (44). Subsite affinities of the S-glucosidase were also determined using competitive inhibition constants for glucose through cellohexaose as described by Roeser and Legler (71). Subsite affinities of CBH 1(D) and CBH II were determined from bond cleavage frequencies, Kjjj and Vjjjg^ for [1-^H]cellotriose through [1-^H]cellohexaose using the method of Allen and Thoma (48) and the method of Suganuma et al. (45). Dr. J. Allen kindly provided a copy of the depolymerase computer model and sample input/output data. A brief description of each method is provided in Appendix A.

PAGE 54

CHAPTER 3 PREPARATION OF [ 1 -^H ] CELLOOLIGOSACCHARIDES AND [^'^C-U] CELLOOLIGOSACCHARIDES Introduction Cellooligosaccharides have been used as substrates to investigate the mode of action and to estimate steady state kinetic parameters of cellulolytic enzymes (5,29), However, some glycosidic bonds of cellooligosaccharides that are susceptible to attack cannot be identified precisely, because identical products are formed from hydrolysis at different glycosidic bonds. Furthermore, low sensitivity of high performance liquid chromatographic analysis (HPLC) of cellooligosaccharides limits the substrate concentration at which initial rates of enzymic reactions can be measured. To overcome these limitations, cellooligosaccharides which are asymmetrically and radioactively labeled are required. Although the enzymic methods for synthesizing oligosaccharides radioactively labeled at either the reducing or nonreducing end have been reported (42,72-74), similar methods have not been developed for synthesizing end-labeled cellooligosaccharides. Carbohydrates can be labeled chemically with tritium at carbon one of the reducing end glucosyl residue by catalytic exchange of unlabeled compounds in the presence of tritium gas and PdO/BaSO^ catalyst (75). Alternatively, carbon one-labeled aldoses can be prepared by

PAGE 55

37 the reduction of aldonolactones with sodium boro[^H]hydride (76) or lithium boro[^H]hydride (77) in weakly acidic solutions. The formation of alditols is prevented under these conditions, because the hydroxyl group at carbon one in its hemiacetal form is not readily displaced by hydride (78). The reduction of aldonolactones with sodium amalgam in tritiated water resulted in lower yields of products than reduction with lithium boro[^H]hydride (77). A practical synthesis of cellooligosaccharides based on the Koenigs-Knor type of reaction has recently been reported by Takeo et al. (79) and may be adapted to the synthesis of nonreducing endlabeled cellooligosaccharides. Although catalytic exchange is a straightforward, rapid method and affords products with high specific radioactivities, the necessity of using large amounts of tritium gas (10 Ci per reaction) makes this procedure potentially more hazardous and less economical than reduction of lactones with sodium boro[^H]hydride. The synthesis of lactones of cellooligosaccharides is a prerequisite for their subsequent reduction with sodium boro[-^H]hydride. However, these lactones may serve as inhibitors in mechanistic studies of cellulolytic enzymes. The synthesis of reducing end-labeled cellooligosaccharides using sodium boro[^H]hydride and the preparation of [''^C-U]cellooligosaccharides from [ C-U]cellulose are described in this report.

PAGE 56

38 Results and Discussion A flowchart of the synthesis of [ 1 -^Hjcellooligosaccharides is shown in Fig. 3-1. Purified cellooligosaccharides, which were prepared from acid hydrolysis of cellulose, were oxidized specifically at carbon 1 of the reducing glucosyl residue with mild bromine oxidation. The resulting aldonic acids of cellooligosaccharides were converted to calcium salts with calcium carbonate and then purified using Bio-Gel P-2 chromatography. Calcium salts of cellooligosaccharide aldonic acids facilitated purification of oxidized products and subsequent lactonization. Cellooligosaccharide lactones were formed from the corresponding calcium salts by first removing calcium with anion exchange resin and then shifting the equilibrium from acid to lactone by removing water. After sodium boro[^H]hydride was used to reduce cellooligosaccharide lactones specifically at carbon 1, [1-^H]cellooligosaccharides were purified using Bio-Gel P-2 chromatography. Preparation of Cellooligosaccharides Cellooligosaccharides used in synthesis of [1-^H]cellooligosaccharides and in enzymic assays as substrates or inhibitors were generated from cellulose by acid hydrolysis. Soluble cellooligosaccharides were separated on a Bio-Gel P-2 column as shown in Fig. 3-2. Cellotriose through cellohexaose were rechromatographed on the same column until they were >99 percent pure as determined by reverse-phase

PAGE 57

Figure 3-1 Synthesis of [ 1 -^Hjcellooligosaccharides In this scheme, purified cellooligosaccharides are oxidized at carbon 1 of the reducing end glycosyl residue with bromine. The resulting cellooligosaccharide aldonic acids are converted to calcium salts, which are then separated from unoxidized cellooligosaccharides using Bio-Gel P-2 chromatography. After calcium is removed from purified calcium salts of cellooligosaccharide aldonic acids using cation exchange resin, lactones of cellooligosaccharides are formed by evaporating water from a solution containing cellooligosaccharide aldonic acids and 2-methoxye thanol (or dimethylsulf oxide ) Reduction of lactones of cellooligosaccharide aldonic acids with sodium boro[^H]hydride in the presence of CO2 yields [1-^H] cellooligosaccharides.

PAGE 58

40 n 1. Br2, CdC03, H2O 2. AgC03 V 3. H2S CHoOH HOOH CH2OH 0 HO OH CH2OH ^'^-^^OH HO 5CaC03 HO HO CH2OH OH CH2OH OH CH2OH Hq\^.^^^^^Q JsCa'^t J5HCO3 0 2-methoxyethanol toluene CHoOH HO HO CH2OH "V-i--— 0 CH2OH hX >o OH J5Ca{0H)2 i 1. NaB^H^ 2. H2O CH,GH HO 0 OH CH20H "W^^ 0 HoV----l^>\/^ OH CH2OH OH

PAGE 59

0 "O 0 (U ~ C 0 ^ CO 0 CO -P OJ (1) c c -J di OJ 0 CQ 0 CO f-H r—l 4J ITS 0 -P CD C ^— •H 0 ^ o 0 ^ O •H 0 r-l C3 X3 rH Q CO Q) 3 d) u CO Co 1— J U CD r rn <--t r— T S B /-\ 1 0300 -M O rH CQ • 'H 0 t3 0 CD a 0 0 "rH CO O >4 t-i -C ^ 0 <1) C -P 0 CO •P "H d) rH X O 'D CO S rH (U •H 3c d) d) Jm C O +J T3 0 0 S CO CO -H rH P tl 0 • rH •H CO d} dJ O O I* >i CO 0 o jC u 0 T3 0 *H CJ ^ CM CU 3 C 1 'O P £-1 3 rH CO 3 CO *H 00 f-H 0 • CO Q) 3 -P d) CO 'H W 0 (U
PAGE 60

c NOiiovdd / lOYr:) 3soonno

PAGE 61

43 HPLC. Cellooligosaccharides of chain length greater than 6 precipitated from solution. Preparation of cellooligosaccharides using Bio-Gel P-2 chromatography avoids the disadvantages of slow flow rates and necessity of repacking columns associated with the charcoal-Celite column method described by Miller et al. (56). Bio-Gel P-2 chromatography has also been used to prepare maltooligosaccharides as described by John et al. (80). Hsu et al. (29) prepared cellotriose and cellotetraose using a Bio-Gel P-2 (-400 mesh) column (1.5 x 200 cm). Oxidation of Cellooligosaccharides The results of oxidizing purified cellobiose through cellohexaose with bromine using a modification of the method of Diehl et al. (57) are presented in Table 3-1. Although on a weight basis, oxidation yielded greater than 100 percent of crude calcium cellooligosaccharide aldonic acids, reducing sugar assay indicated that yields ranged from 79-95 percent. The general decrease in yield as chain length increases may have resulted from the corresponding decrease in concentration of groups available for oxidation. Diehl et al. (57) isolated cellobiono-1 5-lactone in 54 percent yield (weight) from oxidation of cellobiose, whereas cellobiose was oxidized in 93 percent yield using the method described in this report. At different steps in the synthesis, Diehl et al. (57) filtered the reaction mixture through a bed of decolorizing carbon overlaid with Celite. When the method of Diehl et al. (57) was modified by using

PAGE 62

44 Table 3-1 Oxidation of Cellooligosaccharides Cellooligosaccharides were oxidized with bromine using a modification of a method described by Diehl et al. (57). Oxidized cellooligosaccharides were converted to their calcium salts with calcium carbonate. Weight of Weight of OligoCellooligosaccharide Crude Calcium saccharide (g) ,^ Salt (g) % Yield^ Cellobiose 0.50 Cellotriose 0.45 Cellotetraose 0.50 Cellopentaose 0.50 Cellohexaose 0.36 0.56 93 0.49 95 0.54 93 0.52 90 0.44 79 ^ Yields were determined by measuring loss of reducing sugar

PAGE 63

45 only Celite as a filtering aid, cellobiose was oxidized in 93 percent yield. The affinity of cellooligosaccharides for charcoal has been used as the basis for separating cellooligosaccharides on charcoal-Celite columns (56). Cellobiose was also oxidized using Adam's platinum oxide catalyst as described by Conchie et al. (59). However, the resulting product was reduced with sodium borohydride in 12 percent yield. Since higher yields of reducible products were obtained from oxidation with bromine, oxidation in the presence of Adam's catalyst was not used to oxidize cellooligosaccharides. Unoxidized cellooligosaccharides significantly reduce the specific radioactivity of [1--^H]cellooligosaccharides, if they are not removed before cellooligosaccharide lactones are reacted with sodium boro[^H]hydride. Since oxidized cellooligosaccharides contained 5-21 percent of contaminating cellooligosaccharides, oxidized cellooligosaccharides were converted to their calcium salts and separated using Bio-Gel P-2 chromatography. The separation of calcium salts of cellooligosaccharide aldonic acids from cellooligosaccharides was more reproducible than that of the corresponding aldonic acids. Although cellooligosaccharide aldonic acids and calcium salts eluted before cellooligosaccharides, cellooligosaccharide aldonic acids eluted as 2 peaks. The relative size of the peaks may represent the equilibrium between the aldonic acid and aldonolactone Occasionally, cellooligosaccharide aldonic acids appeared to adhere to the

PAGE 64

46 column matrix necessitating repacking the column. Thus, oxidized cellooligosaccharides were purified from cellooligosaccharides as their calcium salts. Furthermore, cellooligosaccharide lactones obtained from calcium salts of cellooligosaccharide aldonic acids were reduced with sodium borohydride in higher yields than cellooligosaccharide lactones obtained directly from aldonic acids as described below. Lactonization of Cellooligosaccharide Aldonic Acids The method of Diehl et al. (57) affords cellobiono-1 ,5lactone directly from oxidation of cellobiose. In preliminary experiments, cellobiono-1 5-lactone prepared using the method of Diehl et al. (57) was reduced with sodium borohydride in <25 percent yield. However, commercially available glucono-1 5-lactone was reduced in >80 percent yield suggesting that cellobiono-1 5-lactone synthesized using the method of Diehl et al. (57) may contain significant amounts of cellobionic acid. Isbell and Frush (60) reported the lactonization of calcium lactobionate in 87 percent yield. Cellobiono-1 5lactone prepared from calcium cellobionate using the method of Isbell and Frush (60) was reduced with sodium borohydride in >80 percent yield. Thus, cellooligosaccharide aldonic acids were converted to corresponding calcium salts and then lactonized using the method of Isbell and Frush (60). The results of converting purified calcium salts of cellooligosaccharide aldonic acids to corresponding lactones

PAGE 65

47 are presented in Table 3-2. The yields of cellooligosaccharide lactones are comparable to those obtained for glucono1,5-lactone and lactobionolactone reported by Isbell and Frush (60). The low yield of cellobiono-1 5-lactone may be due to its higher solubility in 2-methoxyethanol. Samples of cellobiono-1 5-lactone and cellopentano-1 5-lactone were reduced with sodium borohydride to cellobiose and cellopen• taose, respectively, in >50 percent yield. When calcium salts of ce llotetranoic acid, cellopentanoic acid or cellohexanoic acid were lactonized in the presence of 2-methoxyethanol, the resulting lactones were reduced with sodium boro[^H]hydride in <0.1 percent yield of radioactivity. Furthermore, products of the lactonization precipitated from 2-methoxyethanol before water was completely removed suggesting that aldonic acids formed instead of aldonolactones. Isolated products were slightly soluble in water. However, when 2-methoxyethanol was substituted with dimethyl sulfoxide, lactones, which were very soluble in water and were reduced with sodium borohydride in >50 percent yield of carbohydrate, were obtained. Calcium salts of cellooligosaccharides served as an aid to purify oxidized cellooligosaccharides from contaminating cellooligosaccharides using Bio-Gel P-2 chromatography. Precipitation of these calcium salts with 2-methoxyethanol may be omitted and lactones may be obtained directly after removal of calcium by anion exchange resin.

PAGE 66

48 Table 3-2 Lactonization of Calcium Salts of Cellooligosaccharide Aldonic Acids Calcium salts of cellooligosaccharide aldonic acids were converted to their corresponding lactones, after removal of calcium by anion exchange resin and removal of water using a rotary evaporator as described in Experimental Procedures. Weight of Weight of CellooligoCalcium saccharide Calcium Salt Salt (g) Lactone (g) % Yield^ Calcium Cel lobionate 1 .0 0.54 60 Calcium Cellotrionate 0.36 0.33 98 Calcium Cellotetranate 0.017 0.015 94 Calcium Cellopentanate 0.016 0.013 90 Calcium Cellohexanate 0.020 0.017 89 ^ Yields are based on the measured weights of lactones and calcium salts assuming anhydrous molecular weights.

PAGE 67

49 The identification of synthesized products as calcium salts of cellooligosaccharide aldonic acids and corresponding 1,5-lactones is supported by characteristic infrared bands for C-0 asymmetric stretch of carboxylate and C=0 stretch of 1,5-lactone groups listed in Table 3-3. Whereas 1,5-lactones have characteristic C=0 stretch at I76O-I725 cm" 1,4-lactones have characteristic C=0 stretch at 1790-1765 cm"'' (81). The formation of 1,4-lactones is unlikely, because oxygen of carbon 4 contributes to the glycosidic bond of the next residue. Reduction of Lactones of Cellooligosaccharide Aldonic Acids with Sodium Boro[-^H]hydride Procedures for reduction of aldonolactones in the presence of cation exchange resin, boric acid, carbon dioxide, sodium acid oxalate or dilute H2S0^ were evaluated on the basis of yields and stability of pH of the solution during the reaction. Although lactobionolactone or cellobiono-1 5lactone prepared by the method of Diehl et al. (57) was reduced in the presence of each of the above buffering systems to the corresponding sugar in 15-25 percent yield, carbon dioxide and sodium acid oxalate provided more stable pH's. The results of reduction of cellooligosaccharide lactones in the presence of carbon dioxide or sodium acid oxalate with sodium boro[^H]hydride are described below. The results of reducing cellooligosaccharide lactones in the presence of carbon dioxide with sodium

PAGE 68

50 Table 3-3 Major Infrared Bands of Calcium Salts of Cellooligosaccharide Aldonic Acids and Cellooligosaccharides Lactones Cellooligosaccharide Wavenumber (cm~^) C-0 Asymmetric Stretch of C05 Carboxylate^ Calcium Cellobionate Calcium Cellotrionate Calcium Cellotetranate Calcium Cellopentanate Calcium Cellohexanate 1557 1565 1585 1590 1598 C=0 Stretch of 1,5-Lactone^ Cellobiono-1 ,5-lactone Cellotriono-1 ,5-lactone Cellotetrano-1 ,5-lactone Cellopentano-1 ,5-lactone Cellohexano-1 ,5-lactone 1719 1720 1722 1728 1736 Assignments were based on characteristic infrared bands shown by various carboxylate and 1,5-lactone groups reported by Tipson and Parker (81).

PAGE 69

51 I r— I CO Eh T3 •H £4 -a >, r— I ac I— J o o CQ B 3 •H TJ O CO s: -p CO 0) c o o CO J 0) T3 •H Li CO JZ o o CO CO o t>0 o o o o c o •H o 3 0) c o •H a 3 C -H (U TD ClO O O CO (1 -a >. • m CD Lj 0) -p CO CU -p CO C\>H O L, 0 CJ -P H o o -p CO c •H CO 3 T3 0) •H tH •H Li 3 a (U IS CO CU e CO (U TJ T3 Li C -H CO -P CO CO 3 O 3 C •H +J c o o -a o •H Li CU a L CD o o CD CO O 60 CU > r-H LPl i-H o o CO x: T5 -H CU 0) I CD O •H f Li t3u_i T3 CO X CU CU C3. CO C -H O 4J S -P CO a o o CO T3 t< -H (U T3 C •H 73 CD CU -P 73 C T} •H CD CO B CO 0) 73 Li CO JC o o CO CO Li o -p CO Li o Q, • CO >, > x: CD a, CD >^ L4 Li bO CD O o o CQ CO o s CU eiO 73 C -H O Li •H 73 p 1— I CO x: r-t L|l 1 0 -P X o o C Li o o O X) -p -p O CD L< S o CO Li x: 60 O c •H CM CO I a Oi 73 rH Q) a o s , O I -P O -H 73 -H > >-H 73 -H 0) CO -P •H cc: o >> CD -p o > . a xi'-v 3r— 1^ •H n: o 73K^ a c— CO o Li o OQ o CU C -P O'-N x; -p bO CjO o •H CD (U J • • J3 in o T— o XI o in • • 0 0 0 0 0 0 0 in 0 0 0 00 CO in 0 • • • • • • 0 0 0 0 in in in o o o o o o o CD CO (U (U i 0) 1 CD CD c c 1 c 0 c 0 C 1 C 0 1 0 0 0 c 0 c 0 0 0 p 0 -p C -P CO +J CO -P C -P 0 c 0 0 0 Li 0 -P 0 CO 0 1 CO 0 CO •H CD -P CO C CO X CO •H 1-H Li >-H (U f-t CD rH CD >-H c C 1 XI 1 -P 1 -P 1 a 1 x: 1 0 0 in 0 in 0 in 0 in 0 in 0 m -p 0 • r-H i-H ~ rH ^ 1— ( i-H •> 0 3 TCO 1— 1 CU CU (U
PAGE 70

52 boro[-^H]hydride at pH 6-7 are shown in Table 3-4. Highlyradioactive cellooligosaccharides (2707-3750 Ci/mole) were obtained in sufficient quantities for enzymic analysis of cellulases. The specific radioactivities of [1--^H] glucose and [1--^H]cellobiose are lower than those of [1--^H]cellotriose through [1-^H]cellohexaose, because after the addition of sodium boro[^H] hydride more lactone was converted to sugar with sodium borohydride. Two factors may have contributed to the low yield of [1 -^Hjcellotetraose : (1) 10-fold less carbohydrate was present initially, although on a weight basis the concentration was similar to the other reductions and (2) the pH was lower, approximately 6. Preliminary experiments indicated that using 100 rag of lactone significantly improves yields of incorporation of tritium. The yields of [1-^H]cellopentaose and [1-^H]cellohexaose are higher, because the pH of the reaction mixtures was approximately 7. The initial concentration of sodium hydroxide was 0.017 M during the synthesis of [1--^H]cellopentaose and [1 -^Hjcellohexaose. This seemed to stabilize the pH during the addition of sodium boro[-^H]hydride which is stored in 0.1-0.3 M sodium hydroxide. The initial pH of the reaction mixture strongly influences the reduction of aldonolactones with sodium boro[^Hjhydride. At alkaline pH's, lactones hydrolyze to aldonic acids (60), whereas at acidic pH's sodium boro[^H]hydride hydrolyzes. In addition, lactones are more readily converted to alditols at alkaline pH's. Thus, the

PAGE 71

53 pH must be sufficiently high to stabilize sodium boro[-^H]hydride but sufficiently low to prevent hydrolysis of lactones and formation of alditols. Specific radioactivities of [1-^H]cellotetraose through [1 -^Hjcellohexaose are comparable to those obtained for [1-^H]glucose (3900 mCi/mmol) and [ 1 -^H] lactose (3000 mCi/mmol) using PdO/BaSO^ catalyst as described by Evans et al. (75). Specific radioactivities of [1-^H]cellooligosaccharides are 10-fold higher than those reported by Biely et al. (82) for [1-^H]xylooligosaccharides. Biely et al. (82) used PdO/BaSO^ catalyst and tritium gas to label xylooligosaccharides. Reduction of glucono-1 5-lactone in the presence of sodium acid oxalate with sodium boro[^H]hydride produced [^Hjglucose in 0.65 percent yield. Since the initial pH of the solution was 3.8, sodium boro[-^H]hydride may have hydrolyzed before reacting with glucono-1 5-lactone Although Frush and Isbell (61) reduced aldonolactones to aldoses in >95 percent yield with excess sodium borohydride, the initial pH of the reaction is not suited for reductions with sodium boro[^H]hydride. Purity and Stability of [1-^H]Cellooligosaccharides The [1--^H]cellooligosaccharides were purified on a Bio-Gel P-2 column (4.4 x 115 cm) and using Bio-Gel P-2 HPLC. The [1-^H]cellooligosaccharides were shown to be >99 percent pure using thin-layer chromatography, reverse-phase HPLC and Bio-Gel P-2 HPLC. Alditols of

PAGE 72

54 cellooligosaccharides were not detected by refractive index or scintillation counting. Reducing end-labeled cellooligosaccharides of specific radioactivity >250 Ci/mole were stored in 10 percent ethanol at -70C. Reducing endlabeled cellohexaose (3109 Ci/mole) is extremely labile and produced several decomposition products (approximately 80 percent of total radioactivity) after 24 h in 10 percent ethanol at 4C. Reducing end-labeled cellooligosaccharides of specific radioactivity <250 Ci/mole were stored in aqueous solution at -30'-'C. Specificity of Radiolabeling To confirm that tritium was incorporated at carbon 1 of [^H]cellooligosaccharides, 0.28 uCi of [^Hjcellotriose was oxidized with an excess of bromine for 24 h. After removal of tritiated water using a rotary evaporator, 92 percent of the initial radioactivity was accounted for, of which 99.1 percent was in tritiated water and 0.9 percent was in residue of oxidized products. Thus, 99.1 percent of recovered tritium was located at carbon 1. Although catalytic tritiation yields sugars specifically labeled at carbon 1 (>98 percent) (75), reduction of lactones with lithium boro [^Hjhydride in some cases yields significant quantities of labeled alditols (77). Preparation of [ ''^C-U]Cel loollgosaccharides Batch method. The [l^c-Ujcellooligosaccharides, which were used in conjunction with [1-^H]cellooligosaccharides to investigate the action patterns of cellulolytic enzymes,

PAGE 73

55 were generated from [ '*C-U]cellulose during hydrolysis with concentrated HCl. The distribution of [ '''^C-Ujcellooligosaccharides resulting from acid hydrolysis of ['' '^C-U] cellulose using the batch method and subsequent separation using Bio-Gel P-2 chromatography is shown in Table 3-5. Hydrolysis yielded 39 percent soluble cellooligosaccharides ranging from ['''*C-U]glucose to [''^C-Ujcelloheptaose. Preliminary experiments, in which the hydrolysis time was varied from 1 to 21.5 h, indicated that the greatest proportion of ["^^C-Ujcellopentaose and [''^C-Ujcellohexaose was obtained after 3 h of hydrolysis. Labeled [''^C-U]cellopentaose and [''^C-Ujcellohexaose are the most useful for double-label experiments with corresponding [1-^H]cellooligosaccharides. Labeled [ '''*C-U]cellotriose through [^^C-Ujcellohexaose were purified using Bio-Gel P-2 HPLC until each eluted as a single symmetric peak of radioactivity. Purified [^'*C-U]cellooligosaccharides were stored safely in aqueous solution at -30C. Column method To promote formation of [^"^C-Ujcellooligosaccharides of chain length 4-6, [^"^C-Ujcellulose was hydrolyzed by pouring concentrated HCl through a column containing a mixture of glass beads and ['''*C-U]cellulose. The eluant was collected in a flask containing a saturated solution of sodium bicarbonate. Hydrolysis yielded 24 percent soluble cellooligosaccharides, of which predominantly [''^C-Ujcellopentaose (36 percent) and [''^C-Ujcellotriose (12 percent) were obtained. The column method did not provide

PAGE 74

56 Table 3-5 Preparation of [ ^C-U]Cellooligosaccharides [''^C-UjCellulose (750 uCi) was concentrated HCl for 3 h. Soluble [ charides were separated on a Bio-Gel P-2 column (4.4x115 cm) hvdrolyzed with 3 ml of ^ '^C-U]cellooligosac[''4c-U]Cellooligosaccharide Radioactivity (uCi) Percent of Recovered Ce 1 1 00 ligo sac char ides Glucose 100 Cellobiose 52 Cellotriose 40 Cellotetraose 52 Cellopentaose 22 Ce llohexaose 18 Celloheptaose 11 295 mi of initial [ '^C-U]cellulose hydrolyzed) 34 18 14 18 7.6 6.0 3.6

PAGE 75

57 [^'+C-U]cellooligosaccharides of chain length 4-6 in high yield. Conclusions Several [-^Hjcellooligosaccharides have been synthesized by sodium boro[^H]hydride reduction of cellooligosaccharide lactones. The [^Hjcellooligosaccharides were shown to be specifically labeled at carbon 1 of the reducing end glucosyl residue. The specific radioactivities of [1-^H]cellooligosaccharides are sufficiently high to measure initial rates of hydrolysis of cellulolytic enzymes at micromolar substrate concentrations. Purified [''^C-Ujcellooligosaccharides have also been prepared and can be used in conjunction with [1-^H]cellooligosaccharides to study the action patterns of cellulolytic enzymes.

PAGE 76

CHAPTER 4 SEPARATION OF [1-^H]CELL00LIG0SACCHARIDES BY THIN-LAYER CHROMATOGRAPHY Introduction Soluble cellooligosaccharides have been used to investigate kinetics and action patterns of cellulolytic enzymes (5,14,29) and have been separated using high performance liquid chromatography (HPLC) (55 83-85 ) paper chromatography (PC)(20), and thin-layer chromatography (TLC) (19,63,86) Sensitivity afforded by HPLC detection methods limits substrate concentrations at which cellulolytic enzymes may be assayed to values greater than 1 mM. Furthermore, long elution times for cellooligosaccharides makes HPLC impractical for processing large numbers of samples. Previously reported TLC techniques provide limited resolution and quantitation of cellooligosaccharides has not been demonstrated. The synthesis of [ 1 -^H] cellooligosaccharides of high specific radioactivity affords a detection sensitivity not previously possible with cellooligosaccharides. A TLC technique which is quantitative, reliable and provides excellent resolution of cellooligosaccharides is required to take advantage of this sensitivity in analyzing enzymic digests of low concentrations of cellooligosaccharides. 58

PAGE 77

59 A compilation of methods for separating cellooligosaccharides using HPLC has been reported recently (87). Ladisch and Tsao (83) separated glucose through cellohexaose using an Aminex 50W-X4 (Ca^"^) column; however, elution times exceeded 25 min. Although Vratny et al. (84) reported the separation of glucose through cellohexaose, glucose through cellotriose were resolved poorly. The method of Chen and McGinnis (85) requires that cellooligosaccharides be converted to oximes before separation on silica gel columns containing polar-bonded alkylamino groups. Excellent resolution was reported by Gum and Brown (55) using a Whatman PXS-1025 PAC column; however, elution times exceeded 25 min for baseline separation. Kanda et al. (20) separated cellooligosaccharides on Whatman No. 1 paper using 1 -butanol .-pyridine :water (6;4:3, v/v) solvent. Although glucose through cellotriose were well separated, cellotetraose through cellohexaose comigrated even after 100 h of development. Cellooligosaccharides have been separated using thinlayer chromatography on Kieselgel G or Kieselguhr G adsorbents with ethyl acetate :water : isopropanol solvents. Saif-ur-Rahman et al (63) separated glucose through cellohexaose on Kieselgel G plates (20x60 cm) using ethyl acetate:water:isopropanol (2:1:2, v/v) solvent. The disadvantages of this method are the commercial unavailability of such plates and the limited resolution of cellopentaose and cellohexaose. Brown and Andersson (86) separated

PAGE 78

60 cellooligosaccharides on Kieselguhr G plates (20x20 cm) buffered with 0.02 M sodium acetate using 65 percent aqueous isopropanol :ethyl acetate (1:1). Although Rp values for individual sugars indicated good separation, quality of resolution of mixtures was not reported. The S-glucosidase from _T. koningii was assayed using cellooligosaccharides as substrates; the products then were separated on Kieselgel G using 2 ascents of ethyl acetate : isopropanol : water (18:13:9; v:v:v)(19). The quality of the separation was not reported. None of the above methods reviewed have been shown to permit quantitative recovery of the compounds. Here is reported a TLC method for the separation with high resolution and the quantitative extraction of [1-^H]cellooligosaccharides from silica gel. The silica gel plates employed are readily available commercially and separation of carbohydrates is similar to that on Kieselgel G or Kieselguhr G (65) without the requirements for impregnation with salts. This method facilitated the study of kinetics and action patterns of cellulolytic enzymes. Results and Discussion Separation of [1-^H]Cellooligosaccharides Silica gel, cellulose and polyamide adsorbents were evaluated for their ability to separate cellooligosaccharides with a variety of solvents and impregnants (Table 4-1). Glucose through cellopentaose were separated well on silica gel with solvent A; however, cellopentaose and cellohexaose were poorly separated. Although Saif-ur-Rahman et

PAGE 79

61 Table 4-1 Resolution of Cellooligosaccharides With Different Adsorbents, Impregnants and Solvents Deve lopAdsorbent Impregnant Solvent" ments Resolution IB2 none A 2 G1-G5 fair;G6 poor Cellulose none H 2 G1-G2 poor;G2-G5 fair;G5-G6 poor IB2 none H 2 G1-G6 poor Polyaraide none A 2 G1-G6 poor IB2 none E 1 G1-G3 fair;G4-G6 IB2 poor none F 1 G1-G6 fair (DMSO IB2 good solvent) borate A 2 G1 -G3 fair;G4-G6 IB2 poor borate D 2 G1-G4 good;G5-G6 poor IB2 metabisulfite A 2 G1-G3 excellent; IB2 G4-G6 poor metabisulf ite D 2 G1-G4 good;G5-G6 IB2 poor acetate A 2 G1-G4 fair;G5-G6 IB2 poor acetate D 2 G1-G5 good;G5-G6 fair LK5 phosphate A 3 G1-G4 good;G5-G6 LK5 poor none B 3 G1-G5 good;G5-G6 LK5D poor none B 6 G1-G5 good;G5-G6 LK5 fair none C 3 G1-G5 good;G5-G6 LK5/LK5D none G 3 poor G1-G6 excellent K5 none I 3 G1-G4 excellent; K6 G5-G6 poor none I 3 G1-G4 excellent; G5-G6 poor Impregnants were 0.03 M sodium borate, sodium metabi^ sulfite, sodium acetate or 0.2 M sodium phosphate. Solvent compositions were A (ethyl acetate : water : isopropanol; 2:1:2), B (ethyl acetate : water : isopropanol ; 40:25: 27), C (ethyl acetate : water :n-propanol ; 40:30:34), D (ethyl acetate :methanol : acetic acid:water; 2:1:1:1), E (acetonitrile:water; 70:30), F (ethyl acetate : dimethyl sulfoxide:water:isopropanol; 60:40:20:20), G (ethyl acetate:water:methanol; 40:15:20) or H (ethyl acetate: water :methanol; 1:2:1), or I (ethyl acetate : water : methanol; 2:1:1).

PAGE 80

62 al. (63) reported good separation of cellooligosaccharides using solvent A, we observed negligible migration of cellohexaose. Separation of cellooligosaccharides on cellulose TLC plates was fair; however, migration of cellohexaose was negligible. Furthermore, the cellulose adsorbent easilyflaked, thereby eliminating this adsorbant for use with radioactive cellooligosaccharides. Polyamide adsorbent poorly separated cellooligosaccharides. Separation of sugars on silica gel is improved by impregnating silica gel with inorganic salts, such as bisulfite, boric acid, monoand dibasic phosphate and sodium acetate (65). Although impregnating silica gel with salts typically improved resolution of glucose through cellopentaose, cellopentaose and cellohexaose were poorly resolved. Sodium acetate and sodium metabisulf ite increased resolution more than did boric acid and sodium phosphate. The use of impregnants, especially boric acid, lowered Rp values for cellooligosaccharides. Isopropanol, n-propanol, methanol and dimethyl sulfoxide each characteristically influenced the migration of cellooligosaccharides. Isopropanol and n-propanol typically separated glucose through cellotetraose with excellent resolution, whereas longer cellooligosaccharides were poorly separated. Dimethyl sulfoxide served as an excellent solvent for cellooligosaccharides, yielding high Rp values with decreasing resolution of longer cellooligosaccharides. Methanol also yielded high Rp values; however,

PAGE 81

63 resolution between adjacent cellooligosaccharides was similar regardless of chain length. Concentrations of methanol higher than those in solvent G increased the spreading of bands. Lower concentrations of methanol increased the sharpness of bands, but decreased the difference between Rp values. Although Baker-flex IB2 silica gel plates were used to evaluate most solvents and impregnants, Whatman K5 and K6 series plates provided higher resolution of cellooligosaccharides. The LK5D series is especially suited for separation of [1--^H]cellooligosaccharides, because (1) plates contain a preadsorbent zone which increases resolution, (2) preadsorbent zone allows addition of up to 50 ul of sample, (3) prechanneled plates prevent contamination of samples and (4) 150 A pore size provides excellent separation of polar compounds. The best separation of cellooligosaccharides was obtained using Whatman LK5D silica gel plates with three ascents of solvent G and is shown in Fig. 4-1. The ability of preadsorbent zone to focus cellooligosaccharides is reflected in the sharpness of the bands. Similarly, [1--^H] glucose and through [1-^H]cellohexaose were separated with high resolution as shown by the distribution of radioactivity in Fig. 4-2. The Rp values are listed in Table 4-2. Preliminary experiments indicated that approximately 510 percent of [1--^H]cellooligosaccharides adsorbed

PAGE 82

Figure 4-1 Thin-layer chromatographic pattern of cellooligosaccha rides Cellooligosacchar ides were separated on Whatman LK5D TLC plates using 3 ascents of solvent G (ethyl acetate:water :methanol ; 40:15:20; v:v:v). The TLC plate was stained for carbohydrate using _£.-ai^isaldehyde reagent as described in Experimental Procedures. Lanes, from left to right, contain 1 ug of glucose, cellobiose, cellotriose, cellotetraose cellopentaose, cellohexaose and a mixture of glucose-cellohexaose (1 ug each).

PAGE 84

Figure 4-2 Thin-layer chromatographic pattern of [1-^H]cellooligo saccharides Reducing end-labeled cellooligosaccharides were separated on Whatman LK5D TLC plates using 3 ascents of solvent G (ethyl acetate : water : methanol; 40:15:20; v:v:v). The lane was divijded into 2 mm zones from which [1 — 'Hjcellooligosaccharides were eluted and quantitated by scintillation counting as described in Experimental Procedures. The mixture applied contained 6500 dpm of each [ 1 -^Hjcel looligosaccharide The peaks represent, from left to right. [1 — 'H j eel lohexaose [ cell open taose [ ] ce 1 lotetraose [ 1-^H]cellotriose [ 1 -^Hjcellobiose and [l-^Hjglucose. The arrow represents tne end of the preadsorbent zone.

PAGE 85

67 8 12 16 20 DISTANCE FROM BOTTOM (CM)

PAGE 86

68 Table 4-2 Values of the Migration Parameter, Rp Cellooligosaccharides were separated using Whatman LK5D TLC plates with 3 developments in solvent G (ethyl acetate .-water rmethanol ; 40:15:20; v:v:v). [1 -^HjCellooligosaccharide Glucose 0.59 Cellobiose 0.50 Cellotriose 0.39 Cellotetraose 0.30 Cellopentaose 0.20 Cellohexaose 0.13 Rp values were based on distance migrated by cellooligosaccharides and solvent from top of preadsorbent zone

PAGE 87

69 irreversibly, when plates were dried at 110C after application [1--^H]cellooligosaccharides. To avoid irreversible adsorption at the origin or at intermediate positions of migration, plates must be dried at room temperature instead of at 110C. Extraction of [ 1 -^HjCellool igosaccharides Four methods of extracting a mixture containing equal amounts of radioactivity of [ 1 --^H] glucose through cellohexaose from Whatman LK5 TLC plates were evaluated and the results are shown in Table 4-3. When plates were dried at room temperature before extraction, cellooligosaccharides were recovered 10-14 percent more efficiently than when plates were dried at 110C. Extracting cellooligosaccharides before adding Aquasol improved recoveries by 42-45 percent with plates dried at room temperature and 32-37 percent with plates dried at IIO^C. Percent recovery of cellooligosaccharides was approximately equal for each extraction procedure, when plates were dried at room temperature. However, extraction using dimethyl sulfoxide : water solvent yielded slightly lower recoveries than other extraction procedures when plates were dried at 110C. Low recoveries of radioactivity from silica gel suspended in Aquasol may result from physical quenching elicited by silica gel. Aqueous solvents may serve to desorb cellooligosaccharides and disperse silica gel, thereby promoting contact of [1-^H]cellooligosaccharides with -.1

PAGE 88

70 I 0) r-H ja CO EH 0) •a •H £4 (0 s: o O CO CO (U w -P O CO •H a, I— I o o 0 -H i-H -C 1—1 Qi QJ CO O t, —1 ?iO a:: o 1 CD ^ e ^ o o o o >, +J CO X r-H •H I s c •H CO x: o a o c so •H -P o •P X a tit: c CO e -p , -P o •H -P > •P 0 O -H CO '—I O Q. •H Q, X3 CO -P 0) CO x: -p o CO Ch CO CO Ch is o CO 03 -P c D O CO o o 0) -P CO Li (U a e 0) -p a o o Li -P CO CO -p CO c c •H O CO £ D -P (U C -P CD O CD to L. .-H 4-> CD X OJ > CO .-H T3 CD (U'~ •H TD rH > -a B CO c in Q< CO--' x: CO -P S r-H o rH T3 '-s to >, C rH CO > C CD a 3 P CT O OJ CO in >, > +J Li O O 0 I (50 O CD C rC— 15 •Hi — i-H a: c >o •^ •H x: L. I a; CD (aOTD T+J 3 u-j C O Li O t< 0) •> CO o x; -p c XJ -P H ^ o 0) < a x: Li 3 CO Lr\ (u 4-' o • Ta X o to •H 3 (U Li tiO a ^ -p o c bO CD ^ -H a: Q.O o O iH I O O -H J To >— iEH TCH Li >— I T(u a -P L, CO Lr\ o 5 • Ch O bfjw Li C cu •H '-^-P rH > tD •H •• 15 o > c c ^ -H U Li o — Ch C •h T3 a (u -p LO CD TO cu "O CD 0 L4 0 o CD o •H +3 0 O CD •• I— I > T3 w 0 O Li CO 0 •-I Q. CO 0 .. Li 0 0 T3 3 CO bO C •H Q, CD L. O CO bO CO 3-H T3 Ch c o CD CO Q, CD X O rH Ch 0 > CO i-H Li T3 H 3 0 O C > CO S +J CO CO -p c o o CD TD 0 C\J C\J CNJ L. -H H -H -H 0 CTi 'g> CD O00 0 • • 0 0 OS -P 0 CO rH r0 0 0 a 0 0 0 0 0 Li i-H in in CO 0 3 CO H -H -H -H -p -P -P S 0 0 0 0 CD 0 a, 0 0 0 0 TD L. Eh Q CO CD CO 0 0 0 T} Ql m m CM T3 B' CO 0 Eh CD TD >. -P bO 0 •H C 0 Li CM m T> •H Li 0 H H -H +1 •H >i 3 > ^^^ \o -P Li -P 0 cr in 0 Q CO L4 0 a E 0 % Rec radioa^ Eh 0 0 0 0 I-H 0 0 0 0 rH a CD CD CO 0 -P S -P 0 0 O, +1 H -H -H 0 cc Eh Q 0 0 0 0 -P ir\ OCM c r— 0 CM CM 0 CM sed 0 CO Li 3 TJ 0 0 Li 0 0 0 Li o, es c •H 0 Li •H 0 -P > 0 0 CO 0 ( Li 0 t 4-> CC X 0 Q CO

PAGE 89

71 scintillation cocktail. Scott (88) eluted glucose, mannose, galactose, arabinose, xylose and sucrose from silica gel with 95 percent recovery. To assure equal recoveries of [1--^H]cellooligosaccharides from silica gel, methods for extracting individual [1--^H]cellooligosaccharides from Whatman LK5D plates were evaluated. The results are shown in Table 4-4. Although recoveries of [1 -^Hjcellooligosaccharides using methods A, B and C were similar, the range of recoveries for method B was less than those of methods A and C. Thus, method B, by which [1-^H]cellooligosaccharides are counted with an efficiency of 883, was chosen as the extraction procedure. The efficiency of extraction varies slightly from plate to plate and with length of heating at 110C. To efficiently stain cellooligosaccharide standards, plates were heated at 110C after the final development. Conclusions A thin-layer chromatographic method for quantitating and separating [1 --^Hjcellooligosaccharides with high resolution on silica gel has been developed. Although TLC methods for separating cellooligosaccharides have been reported (19,63,86), resolution and quantitation of cellooligosaccharides comparable to that reported here have not been demonstrated. The separation of [1 -^H]cellooligosaccharides on Whatman LK5D silica gel plates with 3 ascents of ethyl acetate : water : methanol (40:15:20; v:v:v) will facilitate the analysis of products formed from the action of cellulolytic enzymes on radioactive cellooligosaccharides.

PAGE 90

72 I J2 03 Eh U Eh O 0) tH L. to CO CU CO O T3 i-H C -H Q -H CD (U C LPi -H O isj CO > -P CQ J x: O O T3 CQ C O -P C ^ W CO CO CO CO 0) e CQ T3 .H "O -P O dJ CQ > •H CO bOTJ "-H J2 -H -O CO C CD !S ^ CO -H 0) o > x: O CO O O CO -P 0 0) -P r-H CO O CO +J .-H S +J X3 CQ CO tD 0) C O r-H 0) O --^T3 CO HOCL X}' — ir-H dj •H TD I e X3 to rH O CDK^ rH o -H I in CO CO ox: (u T• r-l Q, U I 1 O CO > i-H CO CU — CO so O tiO C -P — I O OJ -H d)-^ X -p -o e +J r-i TD CO -H CO S QJ IS t4 o :* T3 o CO in T3 t-) -C • • CD x; o •< TrH O O O w to CO CO 'H CO D sir CO "(US CU O O CO 0 •H >, tJOO -O -P ^ > CD -H O O CO I— t •H i-H r-i ^ x: s e T3 I O r-P c c o (u ~ir\ M -H --H -P S CQ • X: r-H CO o 'H Eh oj 60 '-^ O O CO C CQ S — I 0) -H U CO a: -p s o 0) O Li CO O K> -P •H I 1-H rH CO -P TarH SL. s O 1 — 1 O O CO tiO^ ^ •• x: c o -P O -H 0) 73 X CO >,x; (u .~ W 0) -P X CO 73 (D H tiO-P O O U C U OJ -H O CO +J CQ > tn (U tH 3 O r-H < CO tH Qi X3 r-H E OJ CO TD CO • -P -H CI) • CO CQ O > !! C 0 CO 0) -H -^-> t-i c -p e CO -P CU L. --H x x; o Lo Q. 0) +J > T3 CU CU > o o CU Qi a. Q CU CU r-H Q a a < 0) TD I -H 0 lu >-l CO r-H x: CD o o o I — CO CO o 1 to o O o o O O o O o o o O I>~ (Tl H -H -H -H -H H O o o O O o O o o o O o tr\ "^ CNJ in r— O '"^ C\J rn eg T — o O O O O o o O O o o o CNJ K^ H -H +1 -H -H o O O o o o o O o o o o CD o rm r— o CNJ CO >X) m m CU CO o o f-H a 0 CO o •H XI o CU o CU 0 CO CO OJ CU o o CO CO CO CO o o u CO •H -p c X L. CU 0 0 -P -p a x: o o o o rH l—t rH I-H 0 0 0 0 o o o o T3 CNJ CNJ rn Cvj rn CVI 0 -H -H -H -H -H -H CTi m CM r— 0 > CO cn 00 cr\ cr\ CTi o o u 0 O o O o o O ex: O o O o o o v£) CNI CNJ in V-H H -H -H •H -H O O o o o o O o o o o o s \o C7^ CNI a. CNI O fn m Q m CNI VCNJ rn •a CNJ m CNJ CNJ CNJ fn o CU H -H -H H -H H -H x: Li m in CO CT\ CTs 00 -p 0 CO CO CO CO CO CO 00 0 > o o c CQ 0 o o o O o o o o o o O o o •H o m cn -p o -H -H H -H -H -H CO o o o o O o o o o o o o •p CU o r— m o o X Q CNI o in CNI m w rn CNJ CNI rn eg CNI m fn m -H -H -H H -H •H m CO o 00 CO cr\ CO CO 00 OS 00

PAGE 91

CHAPTER 5 MODE OF ACTION OF S-GLUCOSIDASE Introduction The cellulase system of T. reesei is comprised of endoglucanases, cellobiohydrolase I (D), celloblohydrolase II and 8-glucosidases. Cellulolytic enzymes act sequentially and cooperatively to degrade crystalline cellulose to glucose. The role of the 6-glucosidase in saccharif ication of cellulose is to degrade cellobiose, an inhibitor of the depolymerizing enzymes, and cellooligosaccharides to glucose (5,21,36). Although extracellular S-glucosidases have been purified and partially characterized from Trichoderma reesei (8), Trichoderma viride (16,34,35), Trichoderma koningii (19), Irpex lacteus (20) and Botryodiplodia theobromae (89), little information is available regarding the direction from which e-glucosidases remove glucosyl residues, mode of attack (repetitive or multiple attack), kinetic constants governing the binding of cellooligosaccharides and topography of active sites. A variety of aglycones linked through 6-glycosidic bonds are hydrolyzed by 6-glucosidases (32). The nature of the aglycone influences rates of hydrolysis. Cellulolytic 8-glucosidases are specific for the C-4 hydroxyl group and do not hydrolyze 8-galactosides (90). In general, as cellooligosaccharide chain length increases, decreases 73

PAGE 92

74 Indicating that cellooligosaccharides are good substrates for 8-glucosidases (90). Rates of hydrolysis of cellooligosaccharides correspondingly increase with increasing chain length. The dependence of rates of hydrolysis and K^j on chain length suggested to Shewale (90) that in saccharification of cellulose, B-glucosidases are engaged more in hydrolysis of cellooligosaccharides than in hydrolysis of cellobiose. Hydrolysis of reduced cellooligosaccharides by 8-glucosidases indicates that glycosyl residues are sequentially removed from the nonreducing end; however, the effects of modifying the reducing glucosyl residue on enzymic activity have not been investigated (91). Although the arrangement of subsites in the active site of 6-glucosidases is poorly understood, subsite affinities' have been evaluated for glucoamylase from Rhizopus delemar (44) and Aspergillus salto (47). The subsite maps, which were constructed using the method of Hiromi et al. (44), show that the active site of each enzyme consists of 7 subsites. Subsite II (numbered from the nonreducing end) contributes the greatest proportion of binding energy, whereas the affinity of subsite I is negligible. Subsite affinities decrease in order from subsites III through VII. Since only one productive complex is possible with glucoamylases, the catalytic groups are positioned between subsites I and II. To determine the mode of action, topography of the active site and to more clearly define the role of the

PAGE 93

75 6-glucosidase in the cellulase system of T. reesei the activity of the S-glucosidases on [1-^H]cellooligosaccharides is examined in this report. The use of [1-^H]cellooligosaccharides permits measurement of initial rates of hydrolysis at micromolar substrate concentrations. In addition, these asymmetrically labeled cellooligosaccharides permit identification of glycosidic bonds susceptible to enzymic hydrolysis. The 8-gluco'sidase was shown to remove glucosyl residues from the nonreducing end of cellooligosaccharides in a multi-chain mode of attack. Maximum velocities were determined to be independent of chain length, whereas K^j's decrease as chain length increases. Cellooligosaccharides were also used as inhibitors of Sglucosidase-catalyzed hydrolysis of methylumbelliferyl-8-_Dglucopyranoside (MUG). Inhibition constants of cellooligosaccharides were shown to be similar to the corresponding Michaelis constants. Two approaches were used to evaluate subsite affinities of the 8-glucosidase and revealed that the active site consists primarily of 3 subsites. The results are discussed in terms of the role of the 6-glucosidase in saccharif ication of cellulose. Results and Discussion Hydrolysis of [l-^HjCellooligosaccharides To determine initial rates and to identify products formed from 8-glucosidase-catalyzed hydrolysis of [1-^H]cellooligosaccharides, samples were removed from assay mixtures at various times and products separated using

PAGE 94

76 TLC. Michaelis parameters were obtained by measuring initial rates of hydrolysis at various concentrations of [ 1 --^Hjcellooligosaccharides. The activity of 3-glucosidase on [1 --^Hjcellobiose is shown in Fig. 5-1. Reducing end-labeled cellobiose was hydrolyzed with a corresponding increase of [1-^H]glucose. Since only reducing end-labeled products were monitored in the assay, glucose at the nonreducing end was not observed. Initial bond cleavage frequencies were determined from the slopes of product ratios versus extent of reaction plots. Deviation from linearity of the slope indicates a change in mode of action. Bond cleavage frequency analysis of 6-glucosidase catalyzed hydrolysis of [1-^H]cellobiose indicated that only one bond was cleaved since the slope of [1-^H]glucose product ratio versus extent of reaction curve is equal to 1.0 (Fig. 5-2). Recovery of [1-%]cellobiose and [1-^H]glucose from TLC plates was constant for each time point and at each concentration of substrate assayed indicating that transfer reactions are unlikely to occur from 0.25 to 10.0 mM [1-^H]cellobiose. Linear Lineweaver-Burk and Eadie-Hof stee plots were obtained using initial rates of hydrolysis from 0.25 to 10.0 mM [1-^H]cellobiose (Fig. 5-3). Values of and V^^^ were determined to be 88040 uM and 17.70.2 umol/min/mg, respectively (Table 5-1 ). Reducing end-labeled cellobiose and [ 1 -^H]glucose were formed during S-giucosidase hydrolysis of [1 -^Hjcellotriose (Fig. 5-4). The initial appearance of only [1-%]cellobiose

PAGE 95

Figure 5-1 Time course hydrolysis of [1 --^Hjcellobiose by 8-glucosidase The 6-glucosidase (1.05x10"'^ mg) was incubated in 100 ul of 5 mM sodium acetate buffer, pH 5.0, containing 250 [1 -^Hjcellobiose Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The early, linear region of [1 --^Hjcellobiose degradation curve yielded the initial velocity. The curves represent the distribution of [1-^H]glucose (o) and [1-^H]cellobiose (•).

PAGE 96

78 TIME ( MIN )

PAGE 97

Figure 5-2 Bond cleavage frequency plot for hydrolysis of [1--^H]cellobiose by e-glucosidase The e-glucosidase (1. 05x10"^ rag) was incubated in 100 yl of 5 mM sodium aqetate buffer, pH 5.0, containing 250 [1-^H]cellobiose. Samples were removed from the reaction mixture and analyzed using TLC as described in Experimental Procedures. The initial slope of each line is the bond cleavage frequency of the substrate bond yielding the product (Gi). The curves represent the product ratio of [1-^H]gl ucose (o) and substrate ratio of [l-^Hjcellobiose (•),

PAGE 98

80 0.1 0.2 0.3 0.4 0.5 Gl / (Gl +G2)

PAGE 99

Figure 5-3 Lineweaver-Burk and Eadie-Hol'stee plots for hydrolysis of [ 1 --'Hjcellobiose by s-glucosidase Initial rates for hydrolysis of [ 1--^H]cellobiose (0.25-10 mM) by 8-glucosidase were used to construct Lineweaver-Burk (A) and EadieHofstee (B) plots.

PAGE 100

82 12 3 4 UJ O ir Q. V[^H G2]-^ C^MOL • MIN-'MG PROTEIN"'mM"

PAGE 101

83 Table 5-1 Kinetic Constants for [1--^H]Cellooligosaccharide Hydrolysis of 6-Glucosidase of T. reesei Activity of S-glucosidase was assayed at various concentrations of [1-^H]cellooligosaccharides (0.1 > 0.5 uCi) in the presence of 5 mM sodium acetate buffer, pH 5.0, containing 0.3 mM sodium azide. Kinetic constants were determined from initial rates of hydrolysis using Celand's HYPER program (70). [1-^H]Cellooligosaccharide V /K max'^^m Kju (uM) "max (u mole/min mg of protein) Cellobiose Cellotriose Cellotetraose Cellopentaose Cellohexaose 0.0200 0.0001 880 40 17.7 0.2 0.22 0.02 74 11 16.2 0.9 0.60 0.06 35 5 21 1 0.52 0.04 35 4 18.2 0.8 0.50 0.09 37 10 18 2

PAGE 102

Figure 5-4 Time course hydrolysis of [ 1 -^H J ce llotr lose by 6-glucosidase The g-glucosidase (8. 6x10"^ mg) was incubated in 100 yl of 5 niM sodium acetate buffer, pH 5.0, containing 55 u^'^ [l-^H]cellotriose. Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The early, linear region of [ l-^'H jcellotriose degradation curve yielded the initial velocity. The curves represent the distribution of [1-|H]glucose (o), [1-^H]cellobiose (•) and [1-^Hjcellotriose (n )

PAGE 103

85

PAGE 104

86 indicated that g-glucosidase removes glucosyl residues from the nonreducing end. The lag in the appearance of [1-^H]glucose indicated that [1-^H]cellobiose must accumulate before it is hydrolyzed. Reducing end-labeled glucose appeared after approximately 80% hydrolysis of [1-^H]cellotriose, where [1 -^Hjcellobiose concentration has reached 200 yM. Bond cleavage frequency analysis showed that hydrolysis occurs exclusively at bond 1 (numbered from the nonreducing end) resulting in the initial formation of only [1-^H]cellobiose (Fig. 5-5). Deviation from linearity of the slopes of cellobiose and [ 1 -^Hjglucose after approximately 80 percent hydrolysis of [1-^H]cellotriose indicates that [1-^H]glucose is formed from hydrolysis of [1-^H]cellobiose, instead of cleavage at bond 2 of [1-^H]cellotriose. When initial rates of hydrolysis were determined from 10 to 300 yM [1-^H]cellotriose, linear Lineweaver-Burk and Eadie-Hofstee plots were obtained (Fig. 5-6). Values of Kjjj ^max hydrolysis of [1 -%]cellotriose were determined to be 7411 uM and 16.20.9 ymol/min/mg, respectively (Table 5-1). The time course of g-glucosidase hydrolysis of [1-^H]cellotetraose is shown in Fig. 5-7. Initially, [1-^H]cellotriose was the only product formed indicating that hydrolysis occurred at the nonreducing end. Reducing endlabeled cellobiose appeared after approximately 13 percent degradation of [1-^H]cellotetraose. Bond cleavage

PAGE 105

Figure 5-5 Bond cleavage frequency plot for hydrolysis of [ 1 --^H Jcellotriose by S-glucosidase The B-glucosidase (8. 6x10" mg) was incubated in 100 yl of 5 mM sodium acetate buffer, pH 5.0, containing 55 [ 1 --^Hjcellotriose Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The initial slope of each line is the bond cleavage frequency, of the substrate bond yielding the product (Gi). The curves represent the product ratio of [ 1 -^H Jglucose (o) and [l-^Hjcellobiose (•) and the substrate ratio of [1-^H]cellotriose (),

PAGE 106

88 0 0.2 0.4 0.6 0.8 (GI + G2) / (GI+G2+G3)

PAGE 107

Figure 5-6 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [ 1 -^Hjcellotriose by 0-glucosidase Initial rates for hydrolysis of [1-^K]cellotriose (10-300 uM) by 8-glucosidase were used to construct Lineweaver-Burk (A) and EadieHofstee (B) plots.

PAGE 108

9C T r V • [^H G3]-^ (^MOLMIN-^MG PROTEIN"^mlvT')

PAGE 109

Figure 5-7 Time course hydrolysis of [1-^H]cellotetraose by 8-glucosidase The S_giucosidase (7.08x10"^ mg) was incubated in 100 of 5 mM sodium acetate buffer, pH 5.0, containing 40 [1-^H]cellotetraose. Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The early, linear region of [1-^H]cellotetraose degradation curve yielded the initial velocity. The curves represent the distribution of [1-^H]glucose (o), [1-^H]cell9biose (•), [l-^H]cellotriose ( ) and [1--^H]cellotetraose ().

PAGE 110

92

PAGE 111

93 frequency analysis confirmed that bond 1 was exclusively hydrolyzed; however, deviation from linearity after approximately 13 percent degradation of [ 1 -^Hjcellotetraose indicated that [1-^H]cellotriose is hydrolyzed to [1-^H]cellobiose and glucose (Fig. 5-8). No [ 1 -^Hjglucose was produced by 60 percent hydrolysis of substrate. Initial velocities of [1-^H]cellotetraose hydrolysis obtained from the early, linear portion of time course plots were used to construct linear Lineweaver-Burk and Eadie-Hofstee plots (Fig. 5-9). Values of and V^^^ for hydrolysis of [1-^H]cellotetraose were determined to be 355 yM and 211 ymol/min/mg, respectively (Table 5-1). Products formed during hydrolysis of [1--^H]cellopentaose catalyzed by 6-glucosidase are shown in Fig. 5-10. Consistent with the pattern established by [1-^H]cellotriose and [1-^H]cellotetraose degradation, S-glucosidase initiated attack at the nonreducing end of [1--^H]cellopentaose yielding [1-^H]cellotetraose. After approximately 8 percent degradation of [1--^H]cellopentaose, [1-^H]cellotriose appeared. Reducing end-labeled cellobiose appeared after 25 percent degradation of [1-^H]cellopentaose, whereas no [1-^H]glucose is produced after 80 percent hydrolysis of [1--^H]cellopentaose. Bond cleavage frequency analysis showed a linear decrease of [1-^H]cellopentaose with a corresponding initial, linear increase of [l-^H]cellotetraose (Fig. 5-11). Deviation from linearity of [1-^Hjcellotetraose curve indicated that [1 -^Hjcellotetraose

PAGE 112

1 Figure i?-8 Bond cleavage frequency plot for hydrolysis of [1-^H]cellotetraose by s-glucosidase The 6-glucosidase (7.08x10"^ mg) was incubated in 100 yl of 5 mW sodium acetate buffer, pH 5.0, containing 40 yM [1-^HJcellotetraose. Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The initial slope of each line is the bond cleavage frequency of the substrate bond yielding the product (Gi). The curves represent the product ratios of [ 1-^H ] glucose (o), [ 1 -^H j cellobiose (•) and [1-^H]cellotriose (o) and the substrate ratio of [1--^H]cellotetraose ().

PAGE 113

S5 ( G I+G2+G3 ) / ( G I +G2+63+G4 )

PAGE 114

i Figure 5-9 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [ 1 -^Hjcellotetraose by g-glucosidase Initial rates for hydrolysis of [ 1 --^Hjcellotetraose (10-250 uM) were used to construct Lineweaver-Burk (A) and Eadie-Hof stee (B) plots.

PAGE 115

97 T 1 1 r 0 I 1 1 I 1 L_ 0 100 200 300 400 500 V • [^H G4]-' C^MOL • MIN-'MG PROTEIN mM"')

PAGE 116

i-igure 5-10 Time course hydrolysis ol [1-^H]cellopentaose by S_giucosidase The S-glucosidase (8.33x10" mg) was incubated in 100 ul of 5 niM sodium acetate buffer, pH 5.0, containing 30 yM [ 1 --^H] ce 1 lopentaose Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The early, linear region of the [1-^Hjcellopentacse degradation curve yielded the initial velocity^ The curves represent the distribution of [ 1-£H]glucose (o), [l--^H]c^llobiose (•), [1-^Hjcellotriose (o), [ 1 -^H jcellotetraose () and [1-^H jcellopentaose (^).

PAGE 117

T 1 1 \ ~r TIME ( MIN )

PAGE 118

Figure 5-11 Bond cleavage frequency plot for hydrolysis of [l-^H]cellopentaose by S.giucosidase The S-giucosidase (8.33x10" mg) was incubated in 100 ^1 of 5 mh sodium acetate buffer, pH 5.0, containing 30 UM [1-^H]cellopentaose Samples were removed from the reaction mixture and analyzed using TLC as described in Experimental Procedures. The initial slope of each line is the bond cleavage frequency of the substrate yielding the product (Gij. The curves represent the produt ratios of [ 1-^H]glucose (o), [1--5H]cellobiose (•) and [l-^H]cellotriose (o ) and [ 1-^Hjcellotetraose () and the substrate ratio of [1-^H]cellopentaose (^).

PAGE 119

101 0 0.2 0.4 0.6 0.8 (Gl + G2 + G3 + G4) / ( G I + G2+ G3+ G4+ G5 )

PAGE 120

102 serves as substrate for 0-glucosidase forming [1--^H]cellotriose. Similarly, [1--^Hjcellotriose served as a substrate forming [1-^H]cellobiose. No [1-^H]glucose was detected after 80 percent hydrolysis of [ 1 --^Hjcellopentaose The initial, linear region of the decrease in [1--^H]cellopentaose concentration was used to determine initial rates of hydrolysis. Lineweaver-Burk and Eadie-Hof stee plots constructed using initial rates are linear (Fig. 5-12). Values of and V^g^^ for hydrolysis of [1-^H] cellopentaose were determined to be 354 uM and 18.20.8 uraol/min/mg, respectively (Table 5-1). The 8-glucosidase also catalyzed the hydrolysis of [1--^H]cellohexaose producing [1-^H]cellopentaose as the first detectable product (Fig. 5-13). After approximately 6 percent hydrolysis of [1-^H]cellohexaose, [1--^H]cellotetraose appeared. Reducing end-labeled cellotriose and [1--^H]cellobiose appeared after 16 percent and 50 percent disappearance of cellohexaose respectively. No detectable [1-^H] glucose was detected after 50 percent hydrolysis of cellohexaose. Bond cleavage frequency analysis confirmed that [1--^H]cellopentaose was the first product formed from hydrolysis of [1--^H]cellohexaose (Fig. 5-14). Deviation from linearity of the bond cleavage frequency curve of [1 -^Hjcellopentaose after 6 percent hydrolysis of [1--^H]cellohexaose indicated that [1--^H]cellopentaose serves as a substrate for g-glucosidase Similarly,

PAGE 121

Figure 5-12 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [1 -^Hjcellopentaose by 8-glucosi dase Initial rates for hydrolysis of [ 1-^H]cellopen taose (7.5-150 uM) by 6-glucosidase were used to construct Lineweaver-Burk (A) and EadieHofstee (B) plots.

PAGE 122

104 T 1 1 1 r I 1 1 1 1 I I U 0 20 40 60 80 100 120 140 [^H-G5]"^ CmM)"'' 1 r 0 100 200 300 400 V • [^H G5]"^ C^MOL • MIN-^' MG PROTEIN'^* mM"^)

PAGE 123

Figure 5-13 Time course hydrolysis of [1 -^Hjcellohexaose by 8-glucosidase The S-glucosidase (3.17x10*5 mg) was incubated in 100 pi of 5 mM sodium acetate buffer, pH 5.0, containing 250 [1-^H]cellohexaose. Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The early, linear region of the [ 1 -^Hjcellohexaose degradation curve yielded the initial velocityThe curves represent the distribution of [1-^H]glucose (o), [1-^H]cellobiose (•), [1-:^H]cellotriose (), [1-^H]cellotetraose (), [1--^H]cellopentaose (A) and [1-^H]cellohexaose

PAGE 124

1C6 10 20 30 40 TIME ( MIN ) 50 60

PAGE 125

Figure 5-14 Bon(4 cleavage frequency plot for hydrolysis of [1---'Hjcellohexaose by g-glucosidase The g-glucosidase (3.17x10"5 rng) was incubated in 100 pi of 5 mM sodium acetate buffer, pH 5.0, containing 250 yM [ 1 -^H] cellohexaose Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The initial slope of each line is the bond cleavage frequency of the substrate yielding the product (Gi). The curves represent the product ratios of M -^Hjglucose (o), [ 1 -^Hjcellobiose (•) and L1-"Hjcellotriose (), [1-^H jcellotetraose () and [ 1 --^Hjcellopentaose (a) and the substrate ratio of [ 1 --^Hjcellohexaose (a).

PAGE 126

1C£

PAGE 127

109 [ 1 --^Hjcellotetraose and [ 1 --^Hjcellotriose also serve as substrates for B-glucosidase Linear Lineweaver-Burk and Eadie-Hof stee plots constructed using initial velocities derived from time course plots of [1-^H]cellohexaose hydrolysis are shown in Fig. 5-15. Values of and V^j^^^ were determined to be 5711 yM and 182 ymol/min/mg, respectively (Table 5-1). The action patterns of 8-glucosidase from _T^ reesei on [1-^H]cellooligosaccharides indicate that the g-glucosidase removes glucosyl residues from the nonreducing end. Furthermore, products are released from the enzyme after hydrolysis of substrate indicating multi-chain attack. The 8-glucosidase exhibited little tendency to repetitively attack [1--^H]cellooligosaccharides Hirayama et al. (92) and Cole and King (91) reported that 3-glucosidases from oryzae and Aj_ niger respectively, sequentially remove glucosyl residues from the nonreducing end of reduced cellooligosaccharides. Whether modifying the reducing end glucosyl residue of cellooligosaccharides affects the mode of action of g-glucosidases has not been demonstrated. Glucoamylase from R. delemar has also been shown to remove glucosyl residues from nonreducing ends of oligosaccharides (93); however, whether mode of action is repetitive or multi-chain was not determined. Kinetic constants determined from initial rates of hydrolysis of s-glucosidase-catalyzed hydrolysis of cellobiose through [1 -^Hjcellohexaose are summarized

PAGE 128

Figure 5-15 Lineweaver-Burk aQd Eadie-Hof stee plots for hydrolysis of [l-'^Hjcellohexaose by g-glucosidase Initial rates for hydrolysis of [1-^Hjcellohexaose (7.5-150 uM) by s-glucosidase were used to construct Lineweaver-Burk (A) and EadieHofstee (B) plots.

PAGE 129

111

PAGE 130

112 in Table 5-1. Maximal rates of [1--^H]cellooligosaccharide hydrolysis vary by less than 25 percent with chain length; however, K.^ decreases approximately 25-fold from [1-^H] cellobiose to [1 -^Hjcellohexaose. The decrease in K_ as chain length increases is reflected in V/K values, which are approximately 25to 30-fold greater for [ 1 --^Hjcellotriose through [1 --^Hjcellohexaose than for [1--^H]cellobiose. Michaelis constants for cellooligosaccharide hydrolysis by 8-glucosidases from A. niger (94) and S. rolfsii (95) decrease from 4to 20-fold as chain length increases. In contrast, Michaelis constants determined for two 6-glucosidases purified from T. koningii by Wood and McCrae (19) increase as chain length increases. The dependence of Michaelis constants on chain length of cellooligosaccharides suggested to Shewale (90) that the role of 6-glucosidases in saccharif ication of cellulose is to hydrolyze cellooligosaccharides instead of cellobiose. However, the role of the S-glucosidases in cellulase systems should be assessed in terms of the roles of other cellulolytic components-not exclusively by kinetic parameters of one component. Although the decrease in Michaelis constants with increasing chain length for [ 1 -^Hjcellooligosaccharides hydrolyzed by 6-giucosidase of T. reesei is similar to those reported for other fungal S-glucosidases, values of Michaelis constants for [1-^H]cellotriose through [1-^H]cellohexaose are approximately 3to 20-fold less. For example, the K^j's for hydrolysis of cellobiose

PAGE 131

113 through cellohexaose by 6-glucosidase BG-2 from S. rolfsii were determined to be 3.07, 1.23, 0.85, 0.40, and 0.37 mM, respectively (95). When hydrolysis of cellotriose was monitored using Statzyme glucose assay, was determined to be 750 yM (data not shown). However, initial rates of glucose production from cellotriose were not measured accurately, because the amount of glucose in assay mixtures before 10 percent hydrolysis of cellotriose was below the sensitivity of the Statzyme assay. Therefore, the 760 uM for cellotriose is a poor estimate. Since the extent of cellooligosaccharide hydrolysis by B-glucosidases from A. niger (91), S. rolfsii (95) and T. koningii (19) were not reported, the accuracy and comparability of those Michaelis constants can not be evaluated. Hydrolysis of [ ^ ^C-U]Cellooligosaccharides and L 1 --"H J Cell 00 ligo saccharides" Since experiments using reducing-end labeled cellooligosaccharides provide little information regarding the fate of unlabeled products, [''^c-Ujcellooligosaccharides, which were generated from ['' ^C-Ujcellulose, were used in conjunction with [1-^H]cellooligosaccharides. Products formed during the hydrolysis of [1-^H jcellopentaose and ["^^C-Ujcellopentaose are listed in Table 5-2. Although the amount of [1--^H]cellooligosaccharides is similar to that of the corresponding [^'^C-Ujcellooligosaccharides at each time interval, glucose is almost exclusively labeled with '''^C confirming that all glucose arises from hydrolysis at

PAGE 132

114 C\J I \r\ rH XI CO EH T3 (U X3 O 1 — I 0) ac CO ^f^ o 1 CO T-P >— c (U o o r-l CQ 1— I •H 0) 03 U >.'-n r-l 3 O I Io >ro o C -H ^ r-l O CO -p o 3. o o in LH CD • 01 o o CO 60 -P c c H OJ c a •H o CD r-l -P rH c I • to o o s CO 3. S LTn1 — itO, CX) s o p CD T3 0) -p CO X2 a o c in I o t-. -H CM cr\ in in TCO m o CM Tin in in • in o O CTi CO in O O o o o CM O H -H +1 -H -H -H -H HH +1 -H • Tm o o o o TO CM CO l>^ m vo O cn CO O 0[>-vO CM rn CD O • O CO o o o o 00 rnv£) -H -H -H -H -H -H H -H cr\ 4 O • O CO CM O o o o o (J> cnvo in in oo oo oo oo cr> o -H -H U3 O r\VO CO >-

PAGE 133

115 bond 1. The predominant reducing end-labeled product is cellotetraose Initial rates of formation of products and hydrolysis of cellopentaose for 6-glucosidase are listed in Table 5-3. The hydrolysis of [1-^H]cellopentaose and [''^C-U]cellopentaose are within experimental error indicating absence of a detectable kinetic isotope effect. The rate of formation of glucose is dominated by C'' ^C-U]glucose confirming that 8-glucosidase removes glucose from the nonreducing end of cellopentaose. Initial rates of formation of [1-^H] cellooligosaccharides are dominated by that of [1-^H] cellotetraose. Products formed during hydrolysis of [ 1 -^H Jcellohexaose and [^"^C-Ujcellohexaose are shown in Table 5-4. Consistent with the pattern established by hydrolysis of cellopentaose, glucose is almost exclusively labeled with ^^C indicating that the 6-glucosidase removes glucosyl residues from the nonreducing end of cellohexaose In contrast, the predominant reducing end-labeled product is [1-^H]cellopentaose. Initial rates of formation of products and hydrolysis of [1-^H]cellohexaose and ["^"^C-Ujcellohexaose are listed in Table 5-5. The initial rates of hydrolysis of [1-^H]cellohexaose and [''^C-Ujcellohexaose are within experimental error indicating the absence of a detectable kinetic isotope effect. The rate of formation of [^'^C-U]glucose is similar to the rates of hydrolysis of [ 1 -^Hjcellohexaose and

PAGE 134

116 Table 5-3 Initial Rates for 8-Glucosidase Activity on [1-^H]Cellopentaose and [l^c-UjCellopentaose The S-glucosidase (LIxlO^^ mg) was incubated at 40C in 100 ^1 of 5 mM sodium acetate buffer, pH 5.0, containing 0.55 uCi of [1-^H]cellopentaose and 0.22 yCi of [ ^C-U]cellopentaose. Total concentration of cellopentaose was 83 ^M, Initial Velocity (pmol/min) Ce 1 1 00 li go saccharide 3h 14c Glucose 1 .2 0.8 165 8 Cellobiose 0.3 0.2 2 2 Cellotriose 9 1 8 3 Cellotetraose 144 4 133 7 Cellopentaose -154 6 -146 10

PAGE 135

117 I in 0 i-H -Q CO Eh c to CO CO T3 0) -H CO CO o o CO o X P OJ I— I Si o 0 I r-4 oa I— I cu >> O J3 X 0) CO 1 O TCO — I X 0) o o rH CO I— t •H 0) CO O O I tu o O H rO r-H C -H CO •H U +J 3O O Eh o in o in o -'4J O CO (U -P TJ , — ^ — ^ o o o o C\l Cvi (J CJ> CM 44 44 1 1 ij "H "Tl •H "H •H -H m I ^ ^ J — s o o 1 C\i in o o ^ rn in (30 c\J O O d o m in o o r— X) in -H -H -H -H -H -H -H •H +( m • O o m • o rn o o in o o in m m o CT CM m tncM C\J in 'tjo o -H +1 o o o o o o o o o o o in V£) o in in 3 r-l 3 o X 1 X 1 nc 1 n: 1 X 1 X 1 I-H ^n o fn u m o fn o m o m o I-H 1 1
PAGE 136

118 Table 5-5 Initial Rates for g-Glucosidase Activity on [1-^H] Cellohexaose and [ '^C-UjCellohexaose The 8-glucosidase(8.7x10" mg) was incubated at 40C in 100 ul of 5 raM sodium acetate buffer, pH 5.0, containing 0.55 pCi of [1-^H]cellohexaose and 0.19 uCi of [ 4c-U]cellohexaose. Total concentration of cellohexaose was 59.5 uM. Initial Velocity (pmol/min) Ce 1 1 00 li go saccharide 3h 14c Glucose 1 .2 0.4 88 7 Cellobiose 0.07 0.69 0. 7 1 Cellotriose -0.2 0.4 3. 2 0.9 Cellotetraose 7.6 0.8 6 2 Cellopentaose 76 10 68 15 Cellohexaose -84 10 -77 13

PAGE 137

119 [ '^C-U]cellohexaose. The negligible rate formation of [1--^H]glucose supports that cleavage occurs from the nonreducing end of cellohexaose. Rates of reducing end-labeled product formation are dominated by that of [1--^H]cellopentaose. Inhibition of Methylumbelliferylglucoside (MUG) Hydrolysis by Cellooligosaccharides To determine the affinity of 3-glucosidase for each cellooligosaccharide, purified cellobiose through cellohexaose were used to inhibit B-glucosidase-catalyzed hydrolysis of MUG in separate experiments. Figures 5-16, 5-17, 5-18, 5-19 and 5-20 illustrate the inhibition patterns obtained for cellobiose, cellotriose, cellotetraose and cellopentaose and cellohexaose, respectively, during hydrolysis of MUG. Lineweaver-Burk and Dixon plots indicate that each cellooligosaccharide competitively inhibits the 0glucosidase. Straight lines whose intercepts approach the origin were obtained when Dixon replots were constructed for cellobiose, cellotriose and cellotetraose (Fig. 5-21) and for cellopentaose and cellohexaose (Fig. 5-22). A straight line through the origin of a Dixon replot indicates purely competitive inhibition. Competitive inhibition constants determined for each cellooligosaccharide are listed in Table 5-6. The affinity of S-glucosidase for cellooligosaccharides increases from cellobiose through cellohexaose. In contrast, the S-glucosidase has a higher affinity for the product glucose than for cellobiose as substrate. Since

PAGE 138

Figure 5-16 Inhibition of 8-glucosidase by cellobiose using 4-methylumbelliferyl-S-^-glucopyranoside as the substrate The reaction mixture contained 0.17 ug of Sglucosidase with 4-methylumbellif eryl-B-_Dglucopyranoside as the substrate in 50 mM sodium acetate buffer, pH 5.0, as described in Experimental Procedures. A. Lineweaver-Burk plot: the fixed levels of inhibitor were 0 (o), 0.64 (•), 13 (n) and 3.8 raM B. Dixon plot: the levels of substrate were 12 (a), 19 (a), 24 (), 48 (), 59 (•) and 95 uM (o).

PAGE 139

121 I I I r -20 0 20 40 60 80 100 [mug]''' ( mM ) [G2] ( mM )

PAGE 140

Figure 5-17 Inhibition of e-glucosidase by cellotriose using 4-methylumbelliferyl-s-^-glucopyranoside as the substrate The reaction mixture contained 0.17 ug of 8glucosidase with 4-methylumbelliferyl-e-_Dglucopyranoside as the substrate in 50 mM sodium acetate buffer, pH 5.0, as described in Experimental Procedures. A. Lineweaver-Burk plot: the fixed levels of inhibitor were 0 (o), 26 (•), 53 (a) and 130 yM (). B. Dixon plot: the levels of substrate were 12 (a), 19 (a), 24 (), 48 (o), 59 (•) and 95 pM (o).

PAGE 141

-20 0 20 40 60 80 [mug]"'' C mM -T 1 1 1 i \ r [G3] C^M)

PAGE 142

Figure 5-18 Inhibition of g-glucosidase by cellotetraose using 4-methylumbellif eryl-g-^-glucopyranoside as the substrate The reaction mixture contained 0.17 ug of bglucosidase with 4-methylumbellif eryl-g-Dglucopyranoside as the substrate in 50 m sodium acetate buffer, pH 5.0, as described in Experimental Procedures. A. Lineweaver-Burk plot: the fixed levels of inhibitor were 0 (o), 25 (•), 49 (n) and 120 yM (). B. Dixon plot: the levels of substrate were 12 (4), 19 (a), 24 (), 48 (), 59 (•) and 95 yM (o).

PAGE 143

123

PAGE 144

Figure 5-19 Inhibition of s-glucosidase by cellopentaose using 4-niethylumbelliferyl-8-^-glucopyranoside as the substrate The reaction mixture contained 0.17 yg of gglucosidase with 4-methylumbelliferyl-6-D_glucopyranoside as the substrate in 50 mM sodium acetate buffer, pH 5.0, as described in Experimental Procedures. A. Lineweaver-Burk plot: the fixed levels of inhibitor were 0 (o), 9.2 (•), 18 () and 56 yM (). B. Dixon plot: the levels of substrate were 12 (a), 19 (a), 24 (), 48 (o), 59 (•) and 95 yM (o).

PAGE 145

127 -40 -20 0 20 40 60 [G5] CmM)

PAGE 146

Figure 5-20 Inhibition of S -glucosidase by cellohexaose using 4-methylumbelliferyl-S -D;-glucopyranoside as the substrate The reaction mixture contained 0.17 ug of gglucosidase with 4-methylumbelllferyl-8 glucopyranoside as the substrate in 50 mM sodium acetate buffer, pH 5.0, as described in Experimental Procedures. A. Lineweaver-Burk plot: the fixed levels of inhibitor were 0 (o), 15 (•), 30 () and 64 uM (). B. Dixon plot: the levels of substrate were 12 (a), 18 (A), 23 (), 46 (n), 58 (•) and 93 uM (o).

PAGE 147

-20 0 20 40 60 80 [mug]"' ( mM

PAGE 148

Figure 5-21 Dixon replots of inhibition of g-glucosidase by cellobiose, cellotriose and cellotetraose Dixon replots were obtained using (A) cellobiose, (B) cellotriose and (C) cellotetraose as inhibitors with 4-inethylumbelliferyl-6-_D_glucopyranoside as the substrate.

PAGE 149

151

PAGE 150

Figure 5-22 Dixon replots of inhibition of g-glucosidase by cellopentaose and cellohexaose Dixon replots were obtained using (A) cellopentaose and (B) cellohexaose as inhibitors with 4-methylumbelliferyl-B-D-glucopyranoside as the substrate.

PAGE 151

155

PAGE 152

134 Table 5-6 Inhibition of 8-Glucosidase by Cellooligosaccharides Cellooligosaccharides were used to inhibit the qglucosidase hydrolysis of methylumbellifery Iglucoside (MUG). Competitive inhibition constants were determined using computer analysis as described in Experimental Procedures. Cellooligosaccharide Kj^ (yM) Glucose^ 700 50 Cellobiose 1040 10 Cellotriose 58 3 Cellotetraose 46 2 Cellopentaose 35 3 Cellohexaose 31 2 Reference (8).

PAGE 153

135 the major role of the 6-glucosidase in saccharif ication of cellulose is to convert cellobiose, an inhibitor of the depolymerizing enzymes, to glucose, the relative affinities of the 8-glucosidase for glucose and cellobiose serve as a sensitive point of control. The inhibition constants for cellobiose through cellohexaose are similar to Michaelis constants for [1 --^Hjcellobiose through [1--^H]cellohexaose, suggesting that Michaelis constants are good approximations of dissociation constants for cellooligosaccharides. Subsite Mapping of 8-Glucosidase The decrease in Kjjj and for cellooligosaccharides as well as the dependence of and Vj^^^/Kj^ for cellooligosaccharide hydrolysis by 3-giucosidase on chain length suggests that the active site of the 6-glucosidase is composed of subsites which specifically interact with glucosyl residues of cellooligosaccharides. Hiromi et al. (44) formulated a theory by which the binding affinities of subsites of glucoamylase from R. delemar were estimated from the dependence of V^g^/Kjj, on chain length. If hydrolytic rate coefficients for all bonds are independent of chain length, Vjjjgjj/K^j values are directly proportional to the association constant of the single productive complex. Similar to glucoamylase, the activity of the 8-giucosidase from T. reesei on [ 1 -^Hjcellooligosaccharides indicates that the S-giucosidase binds cellooligosaccharides in one productive complex and cleaves glucosyl residues from the

PAGE 154

136 nonreducing end. Therefore, the method of Hiromi et al. (44) was used to construct a subsite map from values of ^max/^m listed in Table 5-1 for the e-glucosidase Equations relevant to the subsite mapping theory of Hiromi et al. (44) are outlined in Appendix A. Affinities for subsites III-VII of glucamylase were evaluated by comparing the ^it^ax^^m values for (n+1)-mer and n-mer substrates. The affinity of subsite I and (hydrolytic rate coefficient) for glucoamylase were obtained from the vertical and horizontal intercepts, respectively, of a plot of exp(An^.>|/RT) versus i^/^^raax^n' linearity of their plot indicates that is independent of chain length (n). The affinity of subsite II for glucoamylases was then calculated from the known affinities of subsites I and III-VII. However, a plot of exp(A^^^/RT) versus ^''/^max^n using Y^^^ values for the 6-glucosidase (Table 5-1) was not linear. Thus, affinity of subsite 1 could not be evaluated. Since determination of the affinity of subsite 2 requires values for affinities of subsites 1 and 3-6, the affinity of subsite 2 could not be evaluated. Therefore, only the summation affinities of subsites 1 and 2 can be determined. The subsite map for the B-glucosidase indicates that the active site is composed primarily of 4 subsites (subsites 1-4)(Fig. 5-23). The affinities of subsites 5 and 6 are negligible. The g-glucosidase binds cellooligosaccharides in one productive complex and cleaves glucosyl

PAGE 155

Figure 5-23 Subsite map for 8-glucosidase constructed using values of ^^g^y^/^^ for cellooligosaccharides Subsite affinities were estimated using the method of Hiromi et al. (44) and values of ^max/^m [ 1 -^H] cellooligosaccharide hydrolysis by 6-glucosidase. The arrow shows the position of the catalytic groups. Since individual affinities of subsites 1 and 2 could not be evaluated using the method of Hiromi et al. (44), the sum of affinities of subsites 1 and 2 is shown.

PAGE 156

158 2 3 4 5 6 SUBSITE

PAGE 157

139 residues from the nonreducing end indicating that the catalytic groups are located between subsites 1 and 2. Although subsites 1 and 2 provide the greatest proportion of binding energy, the individual contributions of subsites 1 and 2 cannot be determined using the method of Hiromi et al. (44). Although Michaelis constants of [1-^H]cellooligosaccharide hydrolysis by the g-glucosidase decrease with increasing chain length, the subsite map indicates that the S-glucosidase preferentially binds glucose through cellotetraose. If -0.8 kcal/mol is taken as the average free energy decrease caused by formation of one hydrogen bond (96), then subsites 1 and 2 could involve the formation of approximately 9 hydrogen bonds (Appendix D). Subsites 3 and 4 could contribute 2 and 1 hydrogen bonds, respectively. The affinities of subsites 5 and 6 are negligible. An alternative method for evaluating subsite binding affinities of the g-glucosidase is to calculate the contribution of glucosyl residues to the free energy of binding from the ratio of inhibition constants for a series of cellooligosaccharides. Roeser and Legler (71) estimated the contribution of hydroxyl groups of g-glucopyranosides to hydrolysis by 0-glucosidase A3 of Aspergill us wentii by a similar comparison of inhibition constants. A subsite map constructed from the competitive inhibition constants of cellooligosaccharides (Table 5-2), using the method of Roeser and Legler (71), is shown in Fig. 5-24. The active

PAGE 158

Figure 5-24 Subsite map for S-glucosidase constructed using values of inhibition constants for ce 1 1 00 li go saccharides Subsite affinities were estimated using the method of Roeser and Legler (71) and values of inhibition constants for cellooligosaccharides. The arrow shows the position of the catalytic groups.

PAGE 159

141 SUBSITE

PAGE 160

142 site of 6-glucosidase comprises primarily 3 subsites, with subsite 1 contributing the greatest proportion of binding energy. Binding at subsite 2 is slightly unfavorable, whereas binding at subsites 4-6 is negligible. Subsite 3 contributes significant binding suggesting that the S-glucosidase scavenges cellotriose formed from the action of the depolymerizing enzymes on cellulose. Subsites 1 and 3 could provide approximately 9 and 2 hydrogen bonds, respectively, for binding of glucosyl residues of cellooligosaccharides (Appendix D). To construct the subsite map using the method of Roeser and Legler (71), the assumption that glucose binds predominantly at subsite 1 was made. Glucose is believed to bind predominantly at subsite 1, because (1) substrate specificity studies yielded Michaelis constants for g-linked glucosyl disaccharides (Appendix E) that are similar to the determined for glucose (8), whereas Michaelis constants for B-glucosides containing hydrophobic aglycones, such as ^-nitrophenyl-e-^-glucoside and 4-methylumbellif eryl-6-_Dglucoside (8), are 10-fold lower than for glucose, and (2) the 6-glucosidase catalyzes the formation of transfer products (8). The formation of transfer products indicates that binding of glucose is sufficiently strong at subsite 1 to maintain a glucosyl-enzyme intermediate. The relative affinities of glucose, 6-linked glucosyl disaccharides and g-glucosides containing hydrophobic aglycones suggests that

PAGE 161

143 subsite 1 is specific for glucose, whereas subsite 2 prefers hydrophobic compounds. The subsite map (Fig. 5-23) constructed using the method of Hirorai et al. (44) is incomplete, because the individual contributions of subsites 1 and 2 cannot be evaluated. However, the method of Roeser and Legler (71) provides a complete subsite map. Both subsite maps indicate that binding at subsites 5 and 6 is negligible and that binding at subsite 3 could contribute to the binding of cellooligosaccharides. Although significant affinity of subsite 4 for glucosyl residues was determined using the method of Hiromi et al. (44), binding at subsite 4 was determined to be negligible using the method of Roser and Legler (71). The V^g^^ of [1-^H]cellotetra ose is approximately 13 percent higher than that of [1-^H]cellopentaose whereas the corresponding values of K^, are indistinguishable. Since the method of Hiromi et al. (44) uses values of ^max/^m evaluating subsite affinities, the V„^^ will significantly alter apparent subsite affinities. Nevertheless, the contribution of subsite 4 to binding of cellooligosaccharides is minor. Subsites affinities have been evaluated for several glucoamylases (41). The subsite adjacent to the catalytic groups and toward the reducing end contributes the greatest proportion of binding energy. The subsite to the left of catalytic groups provides, in general, unfavorable binding energy. In contrast, the corresponding subsites in the

PAGE 162

144 active site of the e-glucosidase of T. reesei exhibit the opposite pattern: subsite to the right of catalytic groups provides unfavorable binding energy, whereas the subsite to the left contributes the greatest proportion of binding energy. The affinities of subsites adjacent to catalytic groups of glucoamylase were evaluated from the linear relationship of exp(A^^^/RT) versus (W^jax^n"we ver, examination of this plot reported by Hiromi et al. (44) for glucoamylase of R. delemar shows that linearity is primarily dependent on ^max maltose. A linear dependence between these parameters is not obtained, if the value of V^j^^ for maltose is omitted. Thus, determination of the binding affinity of the subsite to the left of the catalytic groups may be inaccurate using the method of Hiromi et al. (44). The subsite to which glucose primarily binds must be identified by other experimental approaches. Although fluorometric studies on the binding of gluconolactone and glucose to glucoamylase reportedly supports that gluconolactone and glucose bind to subsites 1 and 2, respectively, the displacement of gluconolactone from glucoamylase by glucose was partly competitive (97). This suggests that subsite 1 has significant affinity for a glucosyl residue. Conclusions. The extracellular s-glucosidase from T. reesai binds [1-^H]cellooligosaccharides in one productive mode and removes glucosyl residues from the nonreducing end. Products are released from the 3-glucosidase after

PAGE 163

145 hydrolysis indicating multi-chain mode of attack. No evidence for repetitive attack was obtained. Michaelis constants for cellooligosaccharide hydrolysis by the 8-glucosidase decrease with increasing chain length. In contrast, values of V^jg^ essentially independent of chain length. Inhibition constants for cellooligosaccharides are similar to Michaelis constants indicating that approximates Kg. A subsite map constructed using the method of Hiromi et al. (44) shows that the active site of the 8glucosidase is composed primarily of 3 subsites; however, a fourth subsite toward the reducing end (subsite 4) provides some affinity for glucosyl residues. Catalytic groups are located between subsites 1 and 2. The active site of the 6glucosidase was also shown to consist primarily of 3 subsites using the method of Roeser and Legler (71). The number of subsites and the relative binding affinities suggests that the role of this 8-glucosidase is to convert cellobiose, an inhibitor of depolymerizing enzymes, and cellotriose to glucose. Additional subsites may be required to promote efficient hydrolysis of longer cellooligosaccharides.

PAGE 164

CHAPTER 6 MODE OF ACTION OF CELLOBIOHYDROLASE 1(D) Introduction Cellobiohydrolases have been purified from the extracellular matrix of several cellulolytic fungi including T. feesei (5,7), T. viride (13), T. konlngii (18) and I. lacteus (20). The role of cellobiohydrolases in the saccharif ication of cellulose is thought to be the removal of cellobiosyl residues from nonreducing chain ends of cellulose (36). Chain ends are generated by endoglucanases which hydrolyze internal glycosidic bonds of cellulose. Cellobiose, an inhibitor of the depolymerizing enzymes, is converted to glucose by S-glucosidases. Component enzymes of cellulase systems act synergistically to convert native, crystalline cellulose to glucose (18). Although the role of cellobiohydrolases in saccharif ication of cellulose has been established, the mode of action, the topography of the active sites, the direction from which enzymes attack and the mechanism by which cellobiohydrolases synergistically degrade crystalline cellulose have not been described. Two cellobiohydrolases have been purified and partially characterized from T. reesei (6,7). Hsu et al. (29) estimated values of and Y^^^ for cellotriose and cellotetraose hydrolysis catalyzed by a cellobiohydrolase 146

PAGE 165

147 isolated from T. reesei Hydrolysis of cellotetraose yielded predominantly cellobiose, whereas glucose and cellobiose were produced from hydrolysis of cellotriose. Glucose and cellobiose inhibited the enzyme. Hsu et al. (29) estimated kinetic constants from an analysis of the entire time course of the reaction. Initial rate studies were precluded due to low sensitivity of HPLC analysis of cellooligosaccharides. The action patterns of cellooligosaccharide hydrolysis by cellobiohydrolase 1(D) [CBH 1(D)] were investigated by Gritzali (6) using HPLC. Cellobiohydrolase 1(D) produces cellobiose and increasing proportions of glucose and cellotriose as the chain length of the oligosaccharide substrate increases. Cellohexaose is hydrolyzed by CBH 1(D) in a repetitive attack mode. The rate of hydrolysis of cellohexaose is 2-fold higher than that of cellotetraose and cellopentaose. Since hydrolysis at different glycosidic bonds of cellooligosaccharides may yield identical action patterns, the precise bond hydrolyzed cannot be identified. To overcome this problem, reduced cellooligosaccharides have been used to investigate cellobiohydrolases (98). Cellobiohydrolase 1(D) of T. viride showed little tendency to hydrolyze bonds linking the sorbityl residue at the former reducing end of the substrate. An alternative approach to introducing asymmetry into cellooligosaccharides is to convert cellooligosaccharides to 4-methylumbellif eryl glycosides (30). High performance liquid chromatographic analysis of

PAGE 166

148 reaction products of series of 4-fflethylumbelliferyl glycosides from cellooligosaccharides indicated that CBH 1(D) lacks specificity for glycosidic bonds of the glycosides. Two cellobiohydrolases purified from T. reesei by Fagerstam and Pettersson (7) were shown to synergistically degrade crystalline cellulose. This type of synergism is referred to as exo-exo synergism. The mechanism by which CBH 1(D) and CBH II exert exo-exo synergism is unknown. In this report, [ 1 -^Hjcellooligosaccharides are used as substrates to identify glycosidic bonds cleaved and to determine initial rates of cellooligosaccharide hydrolysis by CBH 1(D). Michaelis constants were determined using the initial rates of [1-^H]cellooligosaccharide hydrolysis. Subsite binding affinities for CBH 1(D) were evaluated from bond cleavage frequencies, and Y^^^^ for [1-^H]cellooligosaccharides. Action patterns of CBH 1(D) on [1-^H]cellooligosaccharides and [''^C-Ujcellooligosaccharides indicate that CBH 1(D) repetitively attacks longer cellooligosaccharides and proceeds from the reducing terminus. The results are discussed in terms of the role of CBH 1(D) in the saccharification of cellulose. Results and Discussion Hydrolysis of [1-^HlCellooligosaccharides To determine initial rates of hydrolysis and to identify glycosidic bonds cleaved by CBH 1(D) [ 1 -^Hjcellooligosaccharides were used as substrates. Samples were removed from assay mixtures and products separated using TLC.

PAGE 167

149 Reducing end-labeled products formed during hydrolysis of [1-^H]cellotriose are shown in Fig. 6-1. Cellobiohydrolase 1(D) cleaves both bonds of [1-^H]cellotriose as shown by the formation of [1-^Hjglucose and [1-^H]cellobiose. Bond cleavage frequency analysis indicates that bond 1 and 2 are cleaved with a frequency of 0.30 and 0.70 at 1 mM [1-^H]cellotriose, respectively (Fig. 6-2). Furthermore, the ratio of [1-^H]glucose to [1 -^Hjcellobiose remains constant through 80 percent hydrolysis of [1-^H]cellotriose. This indicates that [1-^H]cellobiose is not further degraded by CBH 1(D). However, the ratio of [ 1 -^Hjglucose to [1 -^Hjcellobiose changes as a function of concentration of [1-^H]cellotriose. At 15 uM [1 -^Hjcellotriose, the bond cleavage frequencies yielding [ 1 -^Hjglucose and [1-^H]cellobiose are 0.94 and 0.06, respectively; whereas at 50 mM cellotriose, the corresponding frequencies are 0.67 and 0.33 (Fig. 6-3). Initial rates of hydrolysis determined from 15 yM to 50 mM [1-^H]cellotriose were used to construct the EadieHofstee plot shown in Fig. 6-4. Values of K^^ and V^^^ were determined to be 9.0 4.5 uM and 0.016 0.002 umol/min/mg, respectively, for the high affinity site. Values of and ^max affinity site are 3400 600 uM and 0.26 0.02 umol/min/fflg, respectively. The presence of high and low affinity sites may represent (1) the presence of two enzymes in the preparation, (2) two binding sites on CBH 1(D) or (3) the formation of

PAGE 168

Figure 6-1 Time course hydrolysis of [1 -•^Hjcellotriose by CBH 1(D) Cellobiohydrolase 1(D) (13 yg) was incubated in ICO yl of 5 miM sodium acetate buffer, pH 5.0, containing 1 mM [ 1 -^H jcellotriose Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Pcocedures. The early, linear region of [1-^H]cellotriose degradation curve yielded the initial velocity. The curves represent the distribution of J-^H]glucose (o), [ 1 -^H jcellobiose (•) and J-^Hjcellotriose ().

PAGE 170

Figure 6-2 Bond cleavage frequency plot for hydrolysis of [l-^Hjcellotriose by CBH 1(D) Cellobiohydrolase 1(D) (13 ug) was incubated in 1C0 ul of 5 mM sodium acetate buffer, pH 5-0, containing 1 mM [1---'Hjcellotriose Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The slope of each line is the bond cleavage frequency of the substrate bond yielding the product (Gi). The curves represent the product ratios of [ 1 -^Hjglucose (o) and [ 1 --^K ]cellobio§e (•) and the substrate ratio of jcellotriose (a).

PAGE 171

155

PAGE 172

Figure 6-3 Effect of [1-^H]cellotriose concentration on product ratios The initial product ratios from [1 --^Hjcellotriose degradation by C^H 1(D) are shown as a function of initial [1-^H]cellotriose concentration (0.C15-50 mM). The curves represent the initial product ratios of [1-^H]glucose (o) and [1-^H]cellobiose (•).

PAGE 173

1^5 0 ^ 2 3 4 5 LOG [G3] jM

PAGE 174

Figure 6-4 Eadie-Hof stee plot for hydrolysis of [1--5h jcellotriose hydrolysis by CBH 1(D) Initial rates for hydrolysis of [l-^H]cellotriose (0.015-50 mM) by CBH 1(D) were used to construct the Eadie-Hof stee plot.

PAGE 175

157 0 0.2 0.4 0.6 0.8 V • [^H G3]"^ (yJVlOL • MIN"^MG PROTEIN"^mM"^ )

PAGE 176

158 transfer or condensation products at high concentrations of substrate. Cellobiohydrolase 1(D) has been shown to be pure by polyacrylamide disc gel electrophoresis, sodium dodecyl sulfate polyacrylamide disc gel electrophoresis, isoelectric focusing and sedimentation equilibrium centrif ugation (6). Cellobiohydrolase 1(D) also migrates as a single, symmetric peak on a Sephadex G-100 gel filtration column (2.5x100 cm) (data not shown). Nevertheless, the presence of a contaminating enzyme is difficult to rule out completely. Since recovery of [1-^H]cellooligosaccharides was constant at each time interval, the formation of transfer or condensation products is unlikely. However, the TLC method may not resolve putative transfer or condensation products. Transfer and condensation products have also not been detected during HPLC analysis of cellotriose hydrolysis catalyzed by CBH 1(D). Since the active site of cellobiohydrolase 1(D) may consist of a series of subsites, it is possible that at high concentrations of [1-^H]cellotriose (>5 mM) another molecule of [ 1 -^H]cellotriose may bind nonproductively and influence both the rate of catalysis and bond cleavage frequencies. However, concentration of cellotriose during saccharif ication of cellulose has been estimated to be less than 1 mM (14). The binding of 2 molecules of trisaccharide has been proposed to explain similar dependence of bond cleavage frequencies on concentration of trisaccharide for A. oryzae ct-amylase (99) and an endo-1,4-3xylanase from C. albidus (82).

PAGE 177

159 Cellobiohydrolase 1(D) primarily cleaves [1-^H]cellotetraose at bonds 2 and 3 forming [1-^H]cellobiose and [1-^H]gl ucose (Fig. 6-5). Bond cleavage frequency analysis showing a linear relationship between product ratios and extent of reaction throughout the reaction demonstrates that the mode of action does not change with time (Fig. 6-6). Bond cleavage frequencies yielding [1-^H]glucose, [1-^H]cellobiose and [ 1 -^Hjcellotriose were determined to be 0.19, 0.80 and 0.01, respectively. In contrast to the dependence of bond cleavage frequencies on concentration of [1-^H]cellotriose, bond cleavage frequencies were independent of the concentration of [1-^H]cellotretraose. This suggests that the active site of CBH 1(D) comprises less than eight subsites, and may not accomodate 2 molecules of cellotetraose. Linear Lineweaver-Burk and Eadie-Hof stee plots were constructed using initial rates of hydrolysis of [1-^H]cellotetraose at concentrations from 0.5 to 15 (Fig. 6-7). Values of and Y^^^ for hydrolysis of [1-^H]cellotetraose are 3.1 i 0.2 uM and 1.51 0.05 umol/min/mg, respectively. In contrast to the high and low affinity sites evaluated during hydrolysis of [1-^H]cellotriose, the binding of [1-^H]cellotetraose is described by one set of Kj^ and Vj^^^ values. No evidence was obtained for a low affinity site at substrate concentration up to 1 mM. The time course of CBH 1(D) hydrolysis of [1-^H]cellopentaose is shown in Fig. 6-8. Cellobiohydrolase 1(D)

PAGE 178

Figure 6-5 Time course hydrolysis of [ 1 -^Hjcellotetraose by CBH 1(D) Cellobiohydrolase 1(D) (4.5 ng) was incubated in 100 ul of 5 mM sodium acetate buffer, pH 5.0, containing 1 uM [1-^H]cellotetraose. Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The early, linear region of [ 1 -^'Hjcellotetraose degradation curve yielded the initial velocity. The curves represent the distribution of [1-^H]glucose (o), [l-^HjcellQbiose (•), (l-^H]cellotriose (a) and [ 1 -^Hjcellotetraose ().

PAGE 179

161 0 10 20 30 40 50 60 TIME ( MIN )

PAGE 180

Figure 6-6 Bond cleavage frequency plot for hydrolysis of [1-^H]cellotetraose by CBH 1(D) Cellobiohydrolase 1(D) (4.5 ng) was incubated in 100 vl of 5 mM sodium^acetate buffer, pH 5.0, containing 1 uM [1-^H]cellotetraose. Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The slope of each line is the bond cleavage frequency of the substrate bond yielding the product (Gi). The curves represent the product (•) and [1-^H]cellotriose () and the substrate ratio of [1-^H]cellotetraose (). ratios of

PAGE 182

Figure 6-7 Linewea ver-Burk and Eadie-Hof stee plots for hydrolysis of [ 1 -^Hjcellotetraose by CBHI(D) Initial rates for hydrolysis of [1-^Hjcellotetraose (0.5-15 uM) by CBH 1(D) were used to construct Lineweaver-Burk (A) and Eadie-Hof stee (B) plots.

PAGE 184

Figure 5-8 Tiice course hydrolysis of [1--^H]cellopentaose by CBH 1(D) Cellobiohydrolase 1(D) (3.8 ng) was incubated in 100 \il of 5 mM sodium agetate buffer, pH 5.0, containing 2.5 y^i [ 1 --^Hjcellopentaose Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The early, linear region of [ 1 -^H]cellopentaose degradation curve yielded the initial velocity. The curves represent the distribution of [1-^H]glucose (o), [1-^H]cellobiose (•), [ 1-^H]cellotriose (a), [1-^H]cellotetraose () and [ 1 --^Hjcellopentaose (a).

PAGE 186

168 cleaves primarily bonds 3 and 4 forming [ 1 -^Hjglucose and cellobiose, respectively. Initial bond cleavage frequencies yielding [1-^H]glucose, [1-^H]cellobiose, [1-^H]cellotriose and [1 -^Hjcellotetraose were determined to be 0.480, 0.513, 0.02 and <0.005, respectively (Fig. 6-9). Bond cleavage frequencies were essentially constant during the time course of the reaction and did not change from 0.25 to 5.0 m '[l-^Hjcellopentaose. Initial velocities obtained from the early, linear region of the time course of hydrolysis of [1-^H]cellopentaose were used to construct linear Lineweaver-Burk and Eadie-Hofstee plots (Fig. 6-10). Values of and V^j^^^ were determined to be 0.52 0.04 uM and 1.26 0.03 umol/min/mg. Products formed during CBH I (D)-catalyzed hydrolysis of [1-^H]cellohexaose is shown in Fig. 6-11. This cellobiohydrolase preferentially cleaves bonds 4 and 5 producing [1-^H]cellobiose and [1-^H]glucose. Initial bond cleavage frequencies derived from the slopes of product ratios versus extent of reaction for [1-^H]glucose, [1-^H]cellobiose, [1-^H]cellotriose, [1-^H]cellotetraose and [1 -^Hjcellopentaose are 0.62, 0.35, 0.025, 0.005 and <0.005, respectively (Fig. 6-12). Bond cleavage frequencies did not appreciably change during the time course of hydrolysis. Initial bond cleavage frequencies were also constant over a range of substrate concentration from 0.25 WiM to 2.5 uM. The early, linear region of the time course of [1-^H]cellohexaose hydrolysis yielded initial rates used to

PAGE 187

Figure 6-S bon^ cleavage frequency plot for hydrolysis of [1-^H]cellopentaose by CEH 1(D) Cellobiohydrolase 1(D) (3-3 ng) was incubated in 100 ul of 5 iTiM sodium acetate buffer, pH 5.0, containing 2.5 yM [ 1 --"Hjcellopentaose Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The slope of each line is the bond cleavage frequency of the substrate bond yielding the product (Gi). The curves represent. the product ratios of [ 1 -^Hjglucose (o), [ 1 -^H jcellobiose (•), [1--^H]cellotriose () and [1-^Hjcellotetraose ( ) and the substrate ratio of [l--5H]cellopentaose (A).

PAGE 188

1 7C T 1 1 1 1 1 r 0 0.1 0.2 0.3 0.4 0.5 0.6 ( G I + G2 + G3 + G4 ) / ( G I + G2 + G3 + G4 + G5 )

PAGE 189

Figure 6-10 Lineweaver-Burk and Eadie-Hoistee plots for hydrolysis of [1-^H]cellopentaose by CBH 1(D) Initial rates for hydrolysis of [ 1-^H]cellopentaose (0.25-5 yM) were used to construct Lineweaver-Burk (A) and Eadie-Hof stee (B) plots.

PAGE 190

172 0 0.5 1.0 1.5 V-[^H G5]"^ 9zM0LMIN"''-MG PROTEIN"'' yM'b

PAGE 191

Figure 6-11 Time course hydrolysis of [ 1 -^Hjcellohexaose by CBH 1(D) Cellobiohydrolase 1(D) (3.2 ng) was incubated in 100 Ml of 5 mM sodium acetate buffer, pH 5.0, containing 2.5 [1--'H]cellohexaose. Samples were removed from the reaction mixture after various intervals and analyzed using TLC early, linear region of [1-^H]cellohexaose degradation curve yielded the initial velocity. The curves represent the distribution of [l-^H]glucose (o), [l-^Hjcellobiose (•), [1-|H]cellotriose ( ) [1-^Hjcellotetraose (), as described in Procedures. The (A).

PAGE 192

174 250 200 X CO 100 10 20 30 40 TIME ( MIN ) 50 60

PAGE 193

Figure 6-12 Bond cleavage frequency plot for hydrolysis of [1-^Hjcellohexaose by CBH 1(D) Cellobiohydrolase 1(D) (3.2 ng) was incubated in 100 ul of 5 mM sodium acetate buffer, pH 5.0, containing 2.5 [1-^H]cellohexaose. Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The slope of each line is the bond cleavage frequency of the substrate bond yielding the product (Gi). The curves represent the product ratios of [ 1 -^Hjglucose (o), [ 1 -^Hjcellobiose (•), [1-^H]cellotriose ( ) [1-^Hjceilotetraose () and [1-^H]celloDentaose (a) and the substrate ratio of [ 1 -^Hjcellohexaose (a).

PAGE 194

17c 0 0.2 0.4 0.6 ( G I + G2 + G3 + G4 + G5 ) / ( G I + G2 + G3 + G4 + G5 + G6 )

PAGE 195

177 construct linear Lineweaver-Burk and Eadie-Hof stee plots (Fig. 6-13). Values of K^^ and Y^^^ for hydrolysis of [1-^H]cellohexaose are 1.1 0.2 yM and 2.7 0.2 ymol/min/mg, respectively. Initial bond cleavage frequencies of CBH 1(D) catalyzed hydrolysis of [ 1 -^Hjcellotriose through [ 1 -^Hjcellohexaose are summarized in Fig. 6-14. Although the major point of cleavage changes with each [1-^Hjcellooligosaccharide, CBH 1(D) shows a strong preference for the reducing end. Cellobiohydrolases reportedly remove cellobiosyl residues from the nonreducing end (36); however, CBH 1(D) may attack cellulosic substrates from the opposite direction. Since nonlabeled products were not monitored in these experiments, the assignment of processivity to one and only one direction cannot be demonstrated conclusively. Double label experiments with combined [1-^H]cellooligosaccharides and [''^C-Ujcellooligosaccharides described below provide more information regarding processivity. The change seen in bond cleavage frequencies with concentration of [1-^H]cellotriose may result from the binding of a second molecule of [ 1 -^Hjcellotriose to subsites which bind glucosyl residues toward the nonreducing end of cellooligosaccharides. In order to favor cleavage at bond 1, [1-^H]cellotriose must shift one subsite over toward the reducing end. At high concentrations, [l-^Hjcellotriose may compete for the mutual subsite and influence the bond cleavage frequencies. If each glucosyl residue of the

PAGE 196

Figure 6-13 Lineweaver-Burk aQd Eadie-Hof stee plots for hydrolysis of [1-^H]cellohexaose by CBH 1(D) Initial rates for hydrolysis of [1--^H]cellohexaose (0.25-2.5 uM) by CBH 1(D) were used to construct Lineweaver-Burk (A) and Eadie-Hof stee (B) plots.

PAGE 197

0 1.0 2.0 3.0 4.0 [^H-GS]"' UMr 0 0.5 1.0 1.5 2.0 V [^H G6]-^ Cy.MOL • MiN-''MG PROTEIN "''vM'')

PAGE 198

Figure 6-14 Bond cleavage frequencies of [ 1 --^Hjcellooligosaccharide hydrolysis by CBH 1(D) Values shown are initial bond cleavage frequencies of [ 1-^H]cellotriose (A^and B), [1--'H]cellotetraose (C), [ 1 -^H Jcellopentaose (D) and [ 1 -^H ] ce 1 lohexaose (E) hydrolysis by CBH 1(D). Bond cleavage frequencies for [1-^H]cellotriose were obtained at 0.015 (A) and 50 (E). The reducing end glucosyl residue is represented by (0).

PAGE 199

1d1 0.01 0.06 ^ 0,94 O p A 0.33 0.67 / O p B 0.80 0.19 / O p C <0.005 0.02 0.513 0.480 / o O-^^ o p D <0.005 0.005 0.025 ^ 0.35 0.62 / O O — O O p E

PAGE 200

182 second molecule of [1-^H]cellotriose binds to a specific subsite and the cleavage is catalyzed at bond 1 of the molecule bound at the above site, then a minimum of 4 subsites are located on the nonreducing side of the catalytic groups. Two molecules of substrate of chain length greater than 3 may not physically fit into the active site; therefore, initial bond cleavage frequencies of [1-^H]cellotretraose through [ 1 -^Hjcellohexaose are independent of concentration of substrate. Kinetic constants determined from initial rates of hydrolysis of [1-^H]cellooligosaccharides are summarized in Table 6-1. Values of Kjj, decrease with increasing chain length; however, the K„j of [1-^H]cellohexaose is approximately 2-fold greater than that of [1-^H]cellopentaose. Values of V^j^^ increase with increasing chain length. A similar dependence is reflected in the "^^^y^/K^ values, which apparently plateau at [1-^H]cellopentaose. Reducing endlabeled cellotriose, whose value for y^j^six^K^ is 250-fold lower than that of [1-^H]cellotetraose, is a poor substrate for CBH 1(D). Thus, values of K^^, V^^^ and V^^^/K^ indicate that cellobiohydrolase 1(D) preferentially cleaves longer cellooligosaccharides. In contrast, the value of k^g^/K^, for [1-^H]cellotriose hydrolysis by g-glucosidase is 170fold higher than that for hydrolysis by CBH 1(D) (Appendix F). Values of k^^^/K^^^ for [1-^H]cellopentaose and [1-^H]cellohexaose hydrolysis by CBH 1(D) are approximately 3-fold higher than those for hydrolysis of S-glucosidase

PAGE 201

183 Table 6-1 Kinetic Constants for [ 1 -^HjCellooligosaccharide Hydrolysis by Cellobiohydrolase 1(D) Activity of cellobiohydrolase 1(D) was measured at various concentrations of [1 -^Hjcellooligosaccharides in the presence of 5 mM sodium acetate buffer, pH 5.0, containing 3 mM sodium azide. Kinetic constants were determined from initial rates of hydrolysis using Cleland's HYPER program (70). [1-^H]Cellooligo^max saccharide K^j (uM) (umole/min mg) max /Km Cellotriose 1. 9.0 4.5 2. 3400 600 Cellotetraose 3.1 0.2 0.0160.002 0.26 0.02 1.51 0.05 Cellopentaose Cellohexaose 0.52 0.04 1.26 0.03 1.1 0.2 2.7 0.2 (1.80.9)x10-^ (7.41 .4)x10~5 (4.80.2)x10~'' 2.40.2 2.40.3

PAGE 202

184 (Appendix F). Thus, comparing values of k^g^/Kjjj for [1-^H]cellooligosaccharide hydrolysis by CBH 1(D) and 8glucosidase suggests that during the saccharif ication of cellulose the role of CBH 1(D) is to hydrolyze longer cellooligosaccharides, whereas the role of the 6-glucosidase is to hydrolyze cellobiose and cellotriose. Hydrolysis of [1-^H]Cellooligosaccharides and ''^C-U ]Cel loo ligosacchar ides Since activity of CBH 1(D) on [1-^H]cellooligosaccharides does not provide information regarding the fate of unlabeled products, [^'^C-U]cellooligosaccharides were used in conjunction with [ 1 --^Hjcellooligosaccharides as substrates for CBH 1(D). Products formed during the time course of hydrolysis of [ 1 -^Hjcellopentaose and [^"^C-Ujcellopentaose are listed in Table 6-2. The ratio of [''^C-Ujgl ucose/[1-^H]glucose is approximately 1 indicating that glucose produced from cellopentaose by CBH 1(D) arises exclusively from cleavage at bond 4. However, the amount of [''^c-ujcellotetraose produced does not equal that of [1-^H]glucose or [''^C-Ujglucose suggesting that CBH 1(D) repetitively attacks cellopentaose. The efficiency of repetitive attack estimated from the concentrations of [''^C-Ujcellotetraose and [ ''^C-U]glucose at 4 min is 97 percent for successive cleavage of bonds 4 and 2 of cellopentaose. Since essentially all glucose arises from cleavage at bond 4 and the resulting cellotetraose is immediately

PAGE 203

1—1 X •H in Q o 1 o a. o • M •H o S c\J 4J -H Oi CD m • cd —i CO 1 i-i r \ <~J • o H T3 u O C •o CD >t 1— 1 jC 3(D • o to S o o ft O O a O CO o T-p m + -H iH C CO o o fH C 0) CD in , O -H XI O OJ 0) O CO O O (U 1 — 1 o m 03 •P X CO H -H o cs<^ -p in o o as 1 c in -p TJ T(U \o t>~ c O rH •H 1— t 3 CU rH E o o 1-1 O -H O o (D c o E -H -H o •H a ch m CO o o o C\J O to in M "ciin 1 CO in c EH o s • o O -H r— ^ -p i_j tiO tiO CD B C S-< o o TJ •H -P TCM C in E c •H -H CO 1 -H CO CO o CO 0) CO > o JZ o Si i rH O E3 GO 3 rH rH rH CO 1 •H -H O o CU o rH Or1 L. o T3 3, •H 1 — 1 1— ( m o X rH O tH CU to o O 185 O o • o • o r— o H -H CO +1 a z • CM 00 • o O O \o O Vo o o 00 o CM •5r\J CM T7— rn CM H -H -H +1 -H -H -H -H o o TO CM O o o o in CM TCO VD cM in CM o o O o o o 03 CM CM a\ in H +1 -H -H -H H -H o o CO O CM CM o o 00 'sr CTvvO V 00 CTi 00 T00 O CM CM O CM O O O O CM CM -H -H H -H H a\ o o o CM o o in CM >^ in in rH V£) CT\0 o O O in -H -H -H -H H H -H vo rn CM o o cntV ^ CM CM f-CVJ in To -H o o o o o o in o CM oo • (U CU o-— > Or— 1 r-lZD -—in r-— 1=1 II X 1 X 1 X 1 X 1 m o m o m o Q 1 n 1 CO

PAGE 204

186 cleaved at bond 2, the direction of attack must be from the reducing end (Fig. 6-15a). Repetitive attack from the nonreducing end beginning at bond 1 would have resulted in a higher ratio of [''^C-U]glucose/[1 -^Hjglucose (Fig. 6-15b). Repetitive attack from the nonreducing end beginning at bond 2 would furnish [^^C-U]/[^-^H] ratios identical to those for repetitive attack beginning at bond 4. However, repetitive attack beginning at bond 2 would produce [1-^H]cellotricse, a poor substrate for CBH 1(D). Cellobiose arises from initial cleavage at bonds 3 or 2 of cellopentaose and from cleavage of cellotetraose which was produced from initial cleavage of cellopentaose. Thus, the ratio of [ '''^C-U] cellobiose/[1-^H]cellobiose (2.4) results from the sum of the concentrations of [1-^H]cellobiose, [ 1 -^Hjcellotriose and twice that of [1-^H]glucose divided by the concentration of cellobiose. Cellotriose is produced from cleavages at either bonds 2 or 3. Initial rates of formation of [1-^H] products and [''^C-Ujproducts from CBH I (D)-catalyzed hydrolysis of [1-^H]cellopentaose and ['''^C-Ujcellopentaose are listed in Table 6-3. The rates of [ 1 -^Hjcellopentaose and ['''^C-Ujcellopentaose hydrolysis are within experimental error indicating the absence of a detectable kinetic isotope effect. The ratio of ^^C/^H rates of formation of products are similar to ['^''^C-U]/[1-^H] product distributions supporting the idea that these distributions of products result from the same bond cleavage frequencies.

PAGE 205

Figure 6-15 Possible types of repetitive attack of CBH 1(D) with [1--5H]cellopentaose and [ ^C-U J cellopentaose Cellobiohydrolase 1(D) may repetitively attack cellopentaose in several possible modes some of which are shown: (A) attack toward the nonreducing end initiated at bond 4 and followed by attack at bond 2, (E) attack toward the reducing end initiated at bond 1 and followed by attack at bond 3, (C) attack toward the reducing end initiated at bond 2 and followed by attack at bond 4. Bonds are numbered from the nonreducing end. The reducing end glucosyl residue is represented by (^).

PAGE 206

186 ^ ^ d B O — O — O — 9 0 Q Q ^

PAGE 207

189 Table 6-3 Initial Rates of Cellobiohydrolase 1(D) Activity on [1-^H]Cellopentaose and [ '+C-U]Cellopentaose Cellobiohydrolase 1(D) (8.62x10"^ mg) was incubated at 40C in 100 ul of 5 mM sodium acetate buffer, pH 5.0, containing 0.55 uCi of [1-^H]cellopentaose and 0.22 uCi of [ ^C-U]cellopentaose. Total concentration of cellopentaose was 83 i^M. Initial Velocity Ratio Cellooligosaccharide (pmol/min) O^C-U]/[^-^li] 1.12 0.09 1-^H]Glucose 58 3 ^4c-U]Glucose 65 4 1-^H]Cell obiose 74 +2 ^ ^ ^4c-U]Cellobiose 178 8 ^'^^ ^ ^'^^ 1-^H]Cellotriose 9.1 0.3 a n + n ^ ^4c-U]Cellotriose 73 2 ^ iT^HjCellotetraose -0.55 0.05 Mna ^^C-UjCellotetraose 2+3 1-^H]Cellopentaose -141 6 o qi + n 07 ^^C-U]Cellopentaose -129 8 *^^ ND = Not determined.

PAGE 208

190 The distribution of products formed from hydrolysis of [''4c-U]cellohexaose and [1-^H]cellohexaose is shown in Table 6-4. The ratio of ['^^C-U]glucose/[1--^H]glucose is approximately 1 indicating that glucose is produced exclusively from hydrolysis at bond 5 of cellohexaose Formation of [ 4c-U]cellopentaose was negligible and did not correspond with the formation of [1-^H]glucose indicating that CBH 1(D) repetitively attacks cellohexaose after initiating cleavage at bond 5. The distribution of [''^C-Ujcellotriose is similar to that of [1-^H]glucose and ['' ^C-U] glucose indicating that cellopentaose which results from initial cleavage of cellohexaose at bond 5, is cleaved to cellobiose and cellotriose (Fig. 6-l6a). Although cellobiose and cellotriose may be produced from cleavage at either bond 2 or 3 of the resulting cellopentaose, bond cleavage frequencies of cellopentaose (Fig. 6-14) suggest that bond 3 is more likely to be cleaved than bond 2. The accumulation of cellotriose supports the idea that it is a poor substrate even when formed in the active site of CBH 1(D). Cellobiose is formed from cellohexaose by three routes: (1) cleavage of cellopentaose, which was initially formed from cleavage at bond 5 of cellohexaose (Fig. 6-l6a), (2) cleavage at bond 4 of cellohexaose (Fig. 6-l6b) and (3) cleavage of cellotetraose, which was initially formed from cleavage at bond 4 of cellohexaose (Fig. 6-l6b). The ratio of 4.6 for ['''*C-U]cellobiose/[1-^H]cellobiose results from the sum of concentrations of [1-^H]glucose and 3 times that of

PAGE 209

191 0) 09 O OS X (U s: o 0) I 0) r-t CO E-i T3 C CO CD CQ O CO X CD s: o CQ CO I— ( o T3 >> x: o •H J3 O CU O I o XI o CQ •H m > 1-1 o o >l r— t o 3: o CM CM CM o •H 1 O o o o o o o o 3. •H -H -H -H +1 -H P o S CTi CO 00 in CO 1 fn o o o o O T5 rE — 1— 1 CD • 3 o o o o O O o in o o (U in m in c\j ^n O CQ o o O O -H -H H -H -H -H -H -H -H -H CM O TCD O o rT— X • • -H -H o o O O O O o o o o •H x: 3 CTi en m cr\ CM in CM o v£) CO o CM CO rn o -H a^ m CM CM CM o rH m O Q) O CQ 1 — 1 CD O ^ o o ^ o o o o o +-> x s T — ^ — V coo in -H -H -H -H -H -H +1 -H H -H 1 O O O o o CO o m O O O O T3 ^ CQ in in o in vCM 00 CO QJi — O co ^ — ^ — +-> CO m CO X C XI O QJ •tH D x: 1 — 1 (— ^ — CM O -H O /-\ • • C U rH s o o o O U3 TO o o •H 3 .— t LxJ CM CM CM OJ -H -H -H -H -H -H -H -H -H -H -H -H CQ in o M 00 o CD in Eh • • S • CH o o V£) O o cr\ CM O O o o CM m T<:1as vo c in VD ~ O wO o • o o '-sin.-H o -H -H Q CO o o o o o o O O o o O ^-'X -P "^ CTi M Q, O CO TEH in in QJ CQ CD 1—1 O T3 >. x: o •H XI o OJ o 0) CQ o CO X Q) s: o QJ O e 13 D I •H U OtCQ u_. OJ OJ -o QJ CQ QJ CQ OJ •H 0) CQ O CQ O 0) CQ (-, d} OJ CQ O CD O CD CQ O CO QJ CQ CQ O CO t-i CD -P O CO x: CQ O O -H {-, -P -P c CO X o Q) O -H •H SL, -P QJ C QJ X OJ o QJ CQ •H xa £-1 -P QJ -P OJ O. QJ x: CD CQ O o -P O -P O a o x: o CQ O O O -1 O rH O rH O r-4 O O O 3 tjO D rH i—t 0) 1— 1 QJ 1—1 QJ 1—1 QJ 1—1 0) •H rH O OJ o QJ O Q) O QJ O QJ O 1— 1 CJr-, O"— Or—. o-— Or— 1 O r— o r— ,3 —.3 1— iH) r— .n o X 1 X 1 X 1 X 1 n: 1 X 1 1—1 fn cj m o m o m u m o r— 1 IT 0) T— [— o 1 t 1 — II — 1

PAGE 210

Figure 6-16 Possible types of repetitive attack of CBH 1(D) with [1-^H]cellohexaose and [ ^C-U]cellohexaose Cellobiohydrolase 1(D) repetitively attacks cellohexaose in 2 possible modes: (A) attack toward the nonreducing end initiated at bond 5 and followed by attack at bond 3, (B) attack toward the nonreducing end initiated at bond 4 and followed by attack at bond 2. Bonds are numbered from the nonreducing end. The reducing end glucosyl residue is represented by

PAGE 211

195 -O — o — o — o — ^ B "O — o — o — o — ^

PAGE 212

194 cellobiose divided by the concentration of [1--^H]cellobiose. The efficiency of repetitive attack initiated at bond 5 of cellohexaose and estimated from the concentration of [ ^C-U]cellopentaose at 60 min is 97 percent. The efficiency of repetitive attack initiated at bond 4 of cellohexaose and estimated from the concentration of [''^c-Ujcellotetraose excluding [1-^H]cellotetraose is approximately 82 percent. "The [''^C-Ujcellotriose/ cellotriose ratio of 14.7 results from the sum of the concentrations of 2 times [ 1 -^Hjcellotriose and [1-^H]glucose divided by [ 1 -^Hjcellotriose. Cellobiohydrolase 1(D) exhibits little tendency to cleave bonds 1 and 2 of cellohexaose Initial rates of formation of products from CBH 1(D)catalyzed hydrolysis of [1-%]cellohexaose and [''^C-Ujcellohexaose are within experimental error indicating the absence of a detectable kinetic isotope effect (Table 6-5). The ratios of formation of [''4c-U]/[1-^H] products are similar to ratios of [''^C-U] /[1-^H] product distributions supporting that the distribution of products results from the same initial cleavage frequencies. The activity of CBH 1(D) on [''^c-u] and [1-^H]cellooligosaccharides indicates that CBH 1(D) repetitively attacks long cellooligosaccharides from the reducing end. Cellobiohydrolase 1(D) was shown by van Tilbeurgh et al. (30) to preferentially hydrolyze 4-methylumbelliferyl glycosides of cellooligosaccharides at the reducing end. Since i

PAGE 213

195 Table 6-5 Initial Rates of Hydrolysis of CellqMohydrolase 1(D) Activity on [ 1 -^HjCellohexaose and [ ''^C-UjCellohexaose Cellobiohydrolase 1(D) (8.62x10"^ mg) was incubated at 40C in 100 ul of 5 mM sodium acetate buffer, pH 5.0, containing 0.55 uCi of [1-^H]cellohexaose and 0.19 yCi of [ ^C-U]cellohexaose. Total concentration of cellohexaose was 59 uM. Initial Velocity Ratio Cellooligosaccharide (pmol/min) [''^C-U]/[1-^H] 1-^H]Glucose X v.u 1 nfl ^ n ^4c-UjGlucose 76 5 ^'^^ ^ *'^ 1-^H]Cellobiose ^4c-U]Cellobiose 1-^H]Cellotr iose ^^C-UjCellotriose 1-:^H]Cellotetraose '^C-U]Cellotetraose (pmol/min) 70.0 0.6 76 5 35 2 161 4 5.5 0.3 84.6 0.8 2.5 0.2 9.6 4 18 14 10 17 130 10 120 20 4.6 0.3 15.4 0.8 3.8 1.6 1-^H]Cellopentaose 18 14 r^r. ^4c-U]CellSpentaose -53 0.99 1-^H]Cellohexaose -ij^u lu n nA n ^^C-U]Cellohexaose -^-^^ • ^'^^

PAGE 214

196 no evidence for repetitive attack or direction of attack was reported, products resulting from 4-methylumbellif eryl glycoside hydrolysis may not represent initial products. Examination of the frequency at which CBH 1(D) of T. viride cleaves reduced cellooligosaccharides revealed little tendency to hydrolyze bonds linking the sorbityl residue at the former reducing end of the substrate (98). High performance liquid chromatographic analysis of cellooligosaccharide hydrolysis also showed that CBH 1(D) repetitively attacks longer cellooligosaccharides (6). Subsite Mapping of Cellobiohydrolase 1(D) Two methods are available in the literature to evaluate subsite affinities of polysaccharide depolymerases which bind substates in more than one productive mode. The method of Suganuma et al. (45) uses V^j^^/K^j values and bond cleavage frequencies for a series of cellooligosaccharides to estimate subsite affinities. The method of Allen and Thoraa (49) uses bond cleavage frequencies to estimate subsite affinities; however, they also require values for ^m ^max ^inf ^'^^^ dissociation constant for a substrate that completely spans the catalytic site is termed ^inf) Equations pertinent to the method of Suganuma et al. (45) and Allen and Thoma (49) are presented in Appendix A. Subsite affinities calculated by the method of Suganuma et al. (45) using the major bond cleavage frequencies and values of ^^^ax^^m ^ series of cellooligosaccharides are listed in Table 6-6. Only major bond cleavage frequencies

PAGE 215

197 Table 6-6 Subsite Affinities of CBH 1(D) Calculated Using Values of Vj^j /K^ and Bond Cleavage Frequencies From 11 -'^JCellooligosaccharides Numbers in brackets indicate mode of calculation using bond cleavage frequencies of cellooligosaccharides at glycosidic bonds denoted by superscripts in parentheses. Subsite Mode of Calculation Subsite Affinity (cal/mol ) 1 [Gl4)/Gp)] -160 80 2 [Gp)/G^4)j -1580 30 [Gp)/G^4)j 230 90 3 [G^2)/Gp)j -2500300 [Gi2)/Gp)] -730 50 4 [g^^Vg^^)] -5100400 7 [G^2)/g|2)j -3400300 [Gp)/Gp)] -1620 80 [G^4)/G^4)] 190 30

PAGE 216

198 were used because they are less susceptible to error. The active site of CBH 1(D) consists of 7 subsites with catalytic groups located between subsites 5 and 6 (Fig. 6-17). Affinities of subsites 2, 3 and 7 were calculated using more than one combination of ^^aax^^m ^^^^^s and bond cleavage frequencies for cellooligosaccharides differing in length by one glucosyl residue. The various combinations used to calculate subsite affinities are listed in Table 6-6 and subsite affinities are shown in Fig. 6-17. Since subsites 5 and 6 are common to all productive complexes, their subsite affinities cannot be evaluated by this method. The significant affinity of subsites 1 through 4 for glucosyl residues is consistent with a repetitive attack mode of CBH 1(D) toward the nonreducing end of cellooligosaccharides. For example, after hydrolysis of bond 5 of cellohexaose subsites 1-4 would strongly bind cellopentaose When product glucose dissociates from the active site, the bound cellopentaose would relocate into subsites 3-7 without leaving the active site. Cellopentaose would then be cleaved at bond 3 forming cellobiose and cellotriose. The method of Suganuma et al. (45) permits calculation of a particular subsite affinity from various combinations of bond cleavage frequencies and V^j^^j^/Kj^ values for n-mer and n+1-mer oligosaccharides. However, differences of up to 1800 cal/mole were obtained when affinities for subsites 2, 3 and 7 were calculated for CBH 1(D). Since the method of Suganuma et al. (45) provides no means to evaluate the

PAGE 217

Figure 6-17 Subsite map for CBH 1(D) constructed using values gf ^^^y/^ia ^'^^ bond cleavage frequencies for [1-^H]ceIlooligosaccharides Subsite affinities were estimated using the method of Suganuma et al. (45) from values of V-jg„/K^ and bond cleavage frequencies for [T-*H]cellooligosaccharide hydrolysis by CBH 1(D). The arrow shows the position of the catalytic groups. Since subsites adjacent to catalytic groups are common to all productive complexes, their affinities cannot be evaluated using this method. Shaded bars indicate that more than one combination of Vj_ /Kj^^ values and bond cleavage frequencies for LT-^Hjcellooligosaccharides differing in chain length by one glucosyl residue were used to evaluate affinities at a particular subsite (Table 6-6).

PAGE 218

2C0

PAGE 219

201 reliability of a particular affinity, the three estimates which range from -176 to 3431 cal/mol, for the affinity of subsite 7 suggest that the method of Suganuma et al. (45) is not adequate to evaluate subsite affinities of CBH 1(D). The subsite map constructed using the computer model for depolymerases developed by Allen and Thoma (49) was first optimized to account for bond cleavage frequencies and location of catalytic groups. Then, subsite affinities were optimized using experimental values of and Vjjj^j^ for [1--^H]cellotriose through [1 -^Hjcellohexaose Values of Kj^ ^max l^igh affinity site detected during [1 -^Hjcellotriose hydrolysis were used in the calculations because they are more relevant to concentrations of cellotriose usually encountered during depolymerization of cellulose. Previously determined inhibition constants for glucose (0.476 M) and cellobiose (90.4 uM) (31) and apparent ^raax hydrolysis of cellobiose (8.84x10~'^ umol/rain/mg, data not shown) were also used in the evaluation of subsite affinities. The value of K^^ observed for cellohexaose hydrolysis was used as an estimate of K^^^. The subsite map for CBH 1(D) that gave the lowest residual error for bond cleavage frequencies, K_,, V^^^ and K,.„^ is shown in Fig. 6-18. A minimum of 7 subsites with the catalytic groups located between subsites 5 and 6 were required to account for the bond cleavage frequencies. Affinities decrease in order from subsites 5-3 with significant affinity at subsites 1 and 2. The potential of

PAGE 220

Figure 6-18 Subsite map for CBH 1(D) constructed using values of Kjj^, Vjj,„„ and bond cleavage frequencies for t*Hjcellooligosaccharides The number of subsites and the position of the catalytic groups were determined using the method of Allen and Thoma (48) from values of K_. V and bond cleavage frequencies for [i--^Hjcellooligosaccharides. The arrow shows the position of the catalytic groups. The bars depict the estimated binding energies including the contribution of the acceleration factor (551 cal/mol).

PAGE 221

20;

PAGE 222

204 subsites 1-5 to bind glucosyl residues is consistent with the repetitive attack toward the nonreducing end catalyzed by CBH 1(D). Unfavorable binding at subsite 6 may promote catalysis by distorting the glucosyl residue adjacent to the catalytic groups. The method of Allen and Thoma (49) permits the hydrolytic rate coefficient ^^^V ^ith chain length in order to obtain good agreement between predicted and experimentally derived values of "^^g^^, Kjjj and bond cleavage frequencies. The approximate decrease in the activation energy caused by the binding of a glycosyl residue and enhancing the rate of bond cleavage is termed the acceleration factor. Although the acceleration factor for subsites was determined to be 551 cal/mol, it only slightly improved the residuals. The subsite maps constructed using the methods of Suganuma et al. (45) or Allen and Thoma (49) indicate that the active site of CBH 1(D) is composed of 7 subsites (Figs. 6-17 and 5-18). Although each map shows favorable binding at subsites 1-4, the map constructed using the method of Suganuma et al. (45) results in different and sometimes conflicting affinities for subsites 2, 3 and 7. Discrepancies in subsite affinities calculated using the method of Suganuma et al. (45) may indicate that the hydrolytic rate coefficient (k^2) cellooligosaccharide hydrolysis is dependent on chain length of cellooligosaccharides. Suganuma et al. (45) assume that hydrolytic rate coefficients are equal.

PAGE 223

205 Conclusions Bond cleavage frequencies determined for CBH 1(D)catalyzed hydrolysis of [1-^H]cellooligosaccharides indicate that CBH 1(D) binds cellooligosaccharides in multiple productive modes. Furthermore, CBH 1(D) preferentially hydrolyzes bonds at the reducing end of [1 -^Hjcellotriose through [1--^H]cellohexaose. Michaelis parameters for hydrolysis of cellooligosaccharides decrease with increasing chain length demonstrating preferential binding of longer cellooligosaccharides. Activity of CBH 1(D) on ['''^C-U J cellooligosaccharides in conjunction with [1 -^Hjcellooligosaccharides indicates that CBH 1(D) repetitively attacks longer cellooligosaccharides towards the nonreducing end. Subsite maps constructed by two methods using kinetic parameters and bond cleavage frequencies show that the active site of CBH 1(D) is composed of 7 subsites with catalytic groups located between subsites 5 and 6.

PAGE 224

CHAPTER 7 MODE OF ACTION OF CELLOBIOHYDROLASE II Introduction The cellulase system of T. reesel comprises endoglucanases, cellobiohydrolases and 3-glucosidases which act sequentially and cooperatively to convert crystalline cellulose to glucose. Endoglucanases preferentially hydrolyze internal glycosidic bonds of cellulose producing chain ends from which cellobiohydrolases remove cellobiosyl residues. The role of the S-glucosidase is to convert cellobiose, an inhibitor of depo^ymerizing enzymes, and oligosaccharides to glucose. Although the role of cellobiohydrolases in saccharif ication of cellulose has been demonstrated, the mode of action, the topography of the active site, the direction from which enzymes attack and the mechanism by which cellobiohydrolases synergistically degrade crystalline cellulose have not been described. Two cellobiohydrolases [CBH 1(D) and CBH II] have been purified and partially characterized from extracellular culture filtrates of T. reesei (6,7). The mode of action of CBH II has been investigated using HPLC analysis of products formed during enzymic hydrolysis of cellooligosaccharides (6). Cellobiohydrolase II exhibits a 2-fold increase in the rate of hydrolysis of cellohexaose compared to cellotetraose 206

PAGE 225

207 and cellopentaose. Cellobiohydrolase II yields exclusively cellobiose from cellotetraose or a mixture of cellobiose and cellotriose from cellopentaose or cellohexaose. Cellobiose is produced from hydrolysis of cellohexaose without concomitant formation of cellotetraose, indicating a sequential cleavage of cellobiosyl residues. Strong synergism has been shown to exist between CBH II and CBH 1(D) in depolymerizing crystalline cellulose; however, the mechanism by which synergism occurs has not been described (7). Fagerstam and Pettersson (7) have referred to this synergism as exo-exo synergism in order to distinguish it from the welldocumented endo-exo synergism. The identity of some glycosidic bonds cleaved by CBH II cannot be determined by examination of products formed during enzymic hydrolysis of cellooligosaccharides, because identical product ratios can be obtained from hydrolysis at different bonds. In this report, the action patterns resulting from CBH II hydrolysis of a series of [1-^H]cellooligosaccharides are described. In addition, the sensitivity afforded by the [1 -^Hjcellooligosaccharides permits measurement of initial rates of hydrolysis at micromolar substrate concentrations and evaluation of Michaelis parameters. The action patterns of CBH II on [ 1 -^Hjcellooligosaccharides indicate that CBH II preferentially hydrolyzes glycosidic bonds at the nonreducing end. Action patterns of CBH II on [1-^H]cellohexaose suggest that cellobiosyl residues are sequentially removed from the nonreducing end.

PAGE 226

208 Values of Kjjj, Vjjjg^ and bond cleavage frequencies are used to construct a subsite map of CBH II. Results and Discussion Hydrolysis of [1-^H]Cellooligosaccharides Initial bond cleavage frequencies and initial rates of CBH II catalyzed hydrolysis of [1 -^Hjcellooligosaccharides were determined by removing samples from enzymic digests and separating products using TLC followed by scintillation counting of zones removed from TLC plates. Cellobiohydrolase II catalyzes the hydrolysis of [1-^H] cellotriose at bond 2 (numbered from the nonreducing end) forming predominantly [1 -^Hjglucose (Fig. 7-1). A negligible amount of [1-^H]cellobiose is also produced. Bond cleavage frequency analysis indicates that bond 1 and 2 are cleaved with a frequency of 0.01 and 0.99, respectively (Fig. 7-2). Furthermore, the linear relationship between product ratios and extent of reaction indicates that the mode of action of CBH II is maintained during the time course of reaction. In contrast to the dependence of bond cleavage frequencies on initial concentration of [1-^H]cellotrlose observed during CBH I (D)-catalyzed hydrolysis, bond cleavage frequencies resulting from CBH II hydrolysis of [1-^H]cellotriose remain essentially constant for substrate concentrations ranging from 2.5 m to 1 mM. Lineweaver-Burk and Eadie-Hof stee plots constructed from initial rates of hydrolysis of [1-^H]cellotriose were linear from 2.5 to 50 m (Fig. 7-3). In contrast, CBH 1(D)

PAGE 227

Figure 7-1 Time course hydrolysis of [1-^H]cellotriose by CBH II Cellobiohydrolase II (1.2 ug) was Incubated in 100 ul of 5 mM sodium acetate buffer, pH 5.0, containing 25 uM [ 1-^H]cellotriose Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The early, linear region of [ 1-^H]cellotriose degradation curve yielded the initial velocity. The curves represent the distribution of [1-^H]glucose (•), [1-^H]cellobiose (o) and [1-^H]cellotriose (n )

PAGE 229

Figure 7-2 Bond cleavage frequency plot for hydrolysis of [1-^H]cellotriose by CBH II Cellobiohydrolase II (1.2 ug) was incubated in 100 yl of 5 mM sodium acetate buffer, pH 5.0, containing 25 uM [ 1 -^H] cellotriose Samples were removed from the reaction mixture and analyzed using TLC as described in Experimental Procedures. The initial slope of each line is the bond cleavage frequency of the substrate bond yielding the product (Gi). The cwves represent the product ratios of [l-^H]glucose (•), [1-^H]cellobiose (o) and [1-^H]cellotriose ().

PAGE 230

212 (GI+G2) / (GI+G2+G3)

PAGE 231

Figure 7-3 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [1-^H]cellotriose by CBH II Initial rates of hydrolysis of [ 1 --^Hjcellotriose (2.5-50 uM) by CBH II were used to construct Lineweaver-Burk (A) and Eadie-Hof stee (B) plots.

PAGE 232

214 0 0.5 1 .0 V-[^H G3]"^ C^MOLMIN'MG PROTEIN"^mM'^

PAGE 233

215 activity on [ 1 --^H] cellotriose indicated the presence of high and low affinity binding sites. Values of and ^^^^ were determined to be 18 3 yM and 0.026 0.002 ymol/min/mg, respectively. The time course of hydrolysis of [ 1 -^Hjcellotetraose by CBH II is shown in Fig. 7-4. Reducing end-labeled cellobiose formed from hydrolysis at bond 2 is the predominant product. A negligible amount of [ 1 --^Hjcellotriose accumulates by 50 min. Bond cleavage frequency analysis showed a linear relationship between product ratios and extent of reaction indicating no change in mode of action (Fig. 7-5). Bond cleavage frequencies estimated from the slopes for [1-^H]gl ucose, [1-^H]cellobiose and [1 -^Hjcellotriose are <0.005, 0.99 and 0.01, respectively. In contrast, CBH 1(D) produces [ 1 -^Hjglucose from hydrolysis of [1-^H]cellotetraose with a frequency of 0.19. Initial velocities obtained from the early, linear region of the time course of hydrolysis of [1--^H]cellotetraose were used to construct linear Lineweaver-Burk and Eadie-Hofstee plots (Fig. 7-6). Values of and V^jj^^^ were determined to be 2.6 + 0.3 yM and 7.7 + 0.3 umol/min/mg, respectively. Reducing end-labeled cellobiose and [1 -^Hjcellotriose were produced from CBH II hydrolysis of [1-^Hjcellopentaose indicating that bonds 2 and 3 were cleaved (Fig. 7-7). Initial bond cleavage frequencies derived from the slope of product ratio versus extent of reaction for [ 1 -^Hjglucose

PAGE 234

Figure 7-4 Time course hydrolysis of [1 -^Hjcellotetraose by CBH II Cellobiohydrolase II (2.0 ng) was incubated in 100 ul of 5 mM sodium acetate buffer, pH 5.0, containing 3.8 uM [1-^Hjcellotetraose. Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Pcocedures. The early, linear region of [ 1 -^Hjcellotetraose degradation curve yielded the initial velocity. The curves represent the distribution of [l-:^H]glucose (o), [1-^H]cellobiose (•), [ 1---'H]cellotriose ( ) and [1--^H]cellotetraose ().

PAGE 235

217 '0 20 30 40 50 60 TIME ( MIN )

PAGE 236

Figure 7-5 Bond cleavage frequency plot for hydrolysis of [1-^H]cellotetraose by CBH II Cellobiohydrolase II (2.0 ng) was incubated in 100 yl of 5 mM sodium acetate buffer, pH 5.0, containing 3.8 yM [1-^H]cellotetraose. Samples were removed from the reaction mixture and analyzed using TLC as described in Experimental Procedures. The initial slope of each line is the bond cleavage frequency of the substrate bond yielding the product (Gi). The curves represent the product ratios of [ 1 -^H Jglucose (o), [1-^H]cellobiose (•), [ 1 -^H Jcellotriose () and the substrate ratio of [1-^H]cellotetraose ().

PAGE 237

219 0 0.2 0.4 0.6 CGI + G2+G3) / (GI + G2+G3+G4)

PAGE 238

Figure 7-6 Lineweaver-Burk and Eadie-Hof stee plots for hydrolysis of [1 -^Hjcellotetraose by CBH II Initial rates of hydrolysis of [1-^H]cellotetraose (0.5-15 uM) by CBH II were used to construct Lineweaver-Burk (A) and Eadie-Hof stee (B) plots.

PAGE 239

221 0 I ^ \ \ \ L_ 0 0.5 1.0 1.5 2.0 2.5 V-[^H G4]"'' C^MOL • MIN"^MG PROTEIN-^yM"'')

PAGE 240

Figure 7-7 Time course hydrolysis of [ 1 --^H Jcellopentaose by CBH II Cellobiohydrolase II (0.32 ng) was incubated in 100 wl of 5 mM sodium acetate buffer, pH 5.0, containing 0.25 uM [1-^H]cellopentaose. Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The early, linear region of [1 -^Hjcellopentaose degradation curve yielded the initial velocity. The curves represent the distribution of [l-^H]glucose (o), [1-^H]cellobiose (•), [1-^Hjcellotriose ( ) [1-^H]cellotetraose ( ) and [ 1-^H]cellopentaose (A).

PAGE 241

22i

PAGE 242

224 [1-^H]cellobiose, [1-^H]cellotriose and [1-^H]cellotetraose are <0.005, 0.483, 0.524 and <0.005, respectively (Fig. 7-8). Bond cleavage frequencies were essentially constant during the time course of reaction and were also constant for substrate concentrations ranging from 0.25 to 2.5 uM. Cellobiohydrolase 1(D), however, predominantly forms [1-^H]glucose and [1--^H]cellobiose from [1 -^Hjcellopentaose. Linear Lineweaver-Burk and Eadie-Hof stee plots were obtained using initial velocities derived from the early, linear region of time course of hydrolysis of [1-^H]cellopentaose (Fig. 7-9). Values of K^, and Vj^^^^ for hydrolysis of [1-^H]cellopentaose are 0.98 0.10 and 2.20 0.09 ymol/min/mg, respectively. The time course of hydrolysis of [1-^H]cellohexaose by CBH II is shown in Fig. 7-10. Two predominant products, [1-%]cellobiose and [1-%]cellotriose, are formed by the action of CBH II. Bond cleavage frequency analysis indicates that the frequencies for production of [1-^H]glucose, [1-^H]cellobiose, [1-^H]cellotriose, [1-^H]cellotetraose and [1-^H]cellopentaose are <0.005, 0.43, 0.54, 0.035 and <0.005, respectively (Fig. 7-11). Bond cleavage frequencies were essentially constant during the time course of hydrolysis and showed little tendency to vary from 0.125 to 2.5 yM [1-^H]cellohexaose. In contrast, CBH 1(D) predominantly

PAGE 243

Figure 7-8 Bond cleavage frequency plot for hydrolysis of [1-^H]cellopentaose by CBH II Cellobiohydrolase II (0.32 ng) was incubated in 100 lil of 5 mM sodium acetate buffer, pH 5.0, containing 0.25 yM [1-^H]cellopentaose. Samples were removed from the reaction mixture and analyzed using TLC as described in Experimental Procedures. The initial slope of each line is the bond cleavage frequency of the substrate bond yielding the product (Gi). The curves represent the product ratios of [1-|H]glucose (o), [1-^H]cellobiose (•), [l-^H]cellotriose ( ) [1-^H]cellotetraose () and the substrate ratio of [ 1 -^H]cellopentaose (A).

PAGE 244

226

PAGE 245

Figure 7-9 Lineweaver-Burk aad Eadie-Hof stee plots for hydrolysis of [1--^H]cellopentaose by CBH II Initial rates for hydrolysis of [l-^H]cellopentaose (0.25-5 uM) by CBH II were used to construct Lineweaver-Burk (A) and Eadie-Hof stee (B) plots.

PAGE 246

226

PAGE 247

Figure 7-10 Time course hydrolysis of Ll--^H]cellohexaose by CBH II Cellobiohydrolase II (0.29 ng) was incubated in 100 ul of 5 mM sodium gcetate buffer, pH 5.0, containing 0.25 yM [1--^H]cellohexaose. Samples were removed from the reaction mixture after various intervals and analyzed using TLC as described in Experimental Procedures. The early, linear region of [1-^H]cellohexaose degradation curve yielded the initial velocity. The curves represent the distribution of [1-^H]glucose (o), [1-^H]c§llobiose (•), [1-^H]cellotriose ( ) [1-^H]cellotetraose (), (A).

PAGE 248

250

PAGE 249

Figure 7-11 Bonc^ cleavage frequency plot for hydrolysis of [1--'H]cellohexaose by CBH II Cellobiohydrolase II (0.29 ng) was incubated in 100 ul of 5 mM sodium acetate buffer, pH 5.0, containing 0.25 uM [1-^H]cellohexaose. Samples were removed from the reaction mixture and analyzed using TLC as described in Experimental Procedures. The initial slope of each line is the bond cleavage frequency of the substrate bond yielding the product (Gi). The curves represent the product ratios of [1-^H]glucose (o), [ 1-^H]cellobiose (•), [ 1 -^Hjcellotriose (), [1-^H]cellotetraose ( ) and [l-^H]cellopentaose (a) and the substrate ratio of [1-^H]cellohexaose (a).

PAGE 250

232 (GI + G2-^G3+G4+G5) / ( G I + G2+G3+G4+G5+G6 )

PAGE 251

233 hydrolyzes bonds 4 and 5 of [ 1-^H]cellohexaose producing [1--^H]cellobiose and [1--^H]glucose, respectively. The early, linear region of time course hydrolysis of [1-^H]cellohexaose yielded initial rates used to construct linear Lineweaver-Burk and Eadie-Hof stee plots (Fig. 7-12). Values of and V^^^^ for hydrolysis of [1-^H]cellohexaose were determined to be 0.57 0.06 uM and 4.4 0.2 ymol/min/mg, respectively. A summary of bond cleavage frequencies for hydrolysis of [1-^H]cellooligosaccharides by CBH II is presented in Fig. 7-13. Cellobiohydrolase II produces [ 1 -^Hjglucose with a frequency of <0.005 from [ 1 -^Hjcellotetraose through [1--^H]cellohexaose whereas the bond cleavage frequency of [1-^H]glucose generated by CBH 1(D) increases from 0.19 to 0.62 from [ 1 -^Hjcellotetraose through [ 1 -^Hjcellohexaose During CBH Il-catalyzed hydrolysis of [1-^H]cellotriose or [1-^Hjcellotetraose bond 2 is cleaved predominantly; however, the bond at the nonreducing end of [1-^H]cellotriose and [1--^H]cellotetraose was cleaved with a frequency of 0.01. Cellobiohydrolase II binds [1-^H]cellopentaose and [1-%]cellohexaose in predominantly two modes. Cleavage at bond 2 is slightly favored over cleavage at bond 3 of [1-^H]cellopentaose. However, cleavage at bond 2 of [1-^H]cellohexaose is 0.035 compared to bonds 3 and 4, which have frequencies of 0.54 and 0.43, respectively. Since CBH II repetitively hydrolyzes cellohexaose the direction of attack significantly influences bond cleavage frequencies.

PAGE 252

Figure 7-12 Linewea ver-Burk and Eadie-Hof stee plots for hydrolysis of [ 1 -•^Hjcellohexaose by CEH II Initial rates of hydrolysis of [1-^H]cellohexaose (0.125-2.5 uM) by CEH II were used to construct Lineweaver-Burk (A) and Eadie-Hof stee (B) plots.

PAGE 253

255

PAGE 254

Figure 7-13 Bond cleavage frequencies of [1-^H]cellooligosaccharide hydrolysis by CBH II Values shown are initial bond cleavage frequencies of [1-^H]cellotriose (A), [1-^HJcellotetraose (B), [ 1 -^Hjcellopentaose (C), [1-^H]cellohexaose (B) and [1 -^Hjcellohexaose (assuming repetitive attack toward the reducing end) (E). The reducing end glucosyl residue is represented by

PAGE 255

237 0.01 0.99 O p 0.01 0.99 <0.005 <0.005 0.524 0.483 <0.005 o — a<0.005 0.035 ^ 0.54 0.43 <0.005 • o<0.005 0.465 0.54

PAGE 256

238 If CBH II repetitively attacks [1-^H]cellohexaose from the reducing end, the bond cleavage frequency analysis would accurately reflect initial points of cleavage. However, if attack is from the nonreducing end, then products of initial cleavages would not be detected, because they would not be radioactive. Evidence provided below using ['''^C-Ujcellohexaose in conjunction with [ 1 --^H] cellohexaose suggests that CBH II repetitively attacks cellohexaose removing cellobiosyl residues from the nonreducing end. Therefore, the true initial bond cleavage frequencies for CBH II activity on [1-^H]cellohexaose are shown in Fig. 7-13e. Bond cleavage frequency analysis of CBH Il-catalyzed hydrolysis of [1--^H]cellooligosaccharides indicates a preference for the nonreducing end, whereas that for CBH 1(D) indicates a preference for the reducing end. Kinetic constants for hydrolysis of [1-^H]cellooligosaccharides by CBH II are summarized in Table 7-1. Values of K^j decrease approximately 26-fold from [ 1 -^H]cellotriose to [1-^H]cellohexaose indicating that CBH II preferentially binds longer cellooligosaccharides. However, a similar trend is not observed for values of V„^^. Since CBH II repetitively attacks [ 1 -^Hjcellohexaose the V„^^ for [1-^H]cellohexaose hydrolysis does not reflect the maximum number of glycosidic bonds cleaved per substrate encounter with enzyme. Approximately 3 moles of cellobiose are produced per mole of cellohexaose suggesting that the V for max total glycosidic bond cleavage is 8.8 umol/min/mg. The

PAGE 257

239 Table 7-1 Kinetic Constants for [1--^H]Cellooligosaccharide Hydrolysis by Ce llobiohydrolase II Activity of cellobiohydrolase II was measured at various concentrations of [ 1 -^Hjcellooligosaccharides in the presence of 5 mM sodium acetate buffer, pH 5.0, containing 0.3 mM sodium azide. Kinetic constants were determined from initial rates of hydrolysis using Cleland's HYPER program (70). [1-^H]Cellooligosaccharide K m max (uM) (umole/min mg) V /K max' m Cellotriose Cellotetraose Cellopentaose Cellohexaose 18 3 2.6 0.3 0.980.10 0.670.06 0.0260.002 7.7 0.3 2.20 0.09 4.4 0.2 0.00150.0001 2.9 0.2 2.3 0.2 6.6 0.4

PAGE 258

240 comparatively low ^^^^ for [1--^H]cellopentaose hydrolysis may result from cellotriose (value of is 18 pM) "sticking" to the active site after hydrolysis of [1--^H]cellopentaose. The general increase in values of Vjjjg^/Kjjj as chain length increases confirms the preferential cleavage of longer cellooligosaccharides by CBH II. Cellobiohydrolase II is more catalytically effective with longer cellooligosaccharides than is CBH 1(D) as indicated by their respective values for k^^^/Kjjj of the cellooligosaccharide series (Appendix F), However, repetitive attack by CBH 1(D) and CBH II makes comparison difficult. Values of k^^^/Kj^ for S-glucosidase action on [1 -•^H]cellooligosaccharides of chain length >3 are greater than 3-fold lower than those reported for CBH 1(D) and CBH II indicating that the role of the e-glucosidase in saccharif ication of cellulose is to convert cellobiose and cellooligosaccharides (especially cellotriose) to glucose; cellooligosaccharides of longer chain length are rapidly hydrolyzed by CBH 1(D) and CBH II (Appendix F). Hydrolysis of [1-^H]Cellooligosaccharides and [^^C-U] Cellooligosaccharides The activity of CBH II on [1 -^Hjcellooligosaccharides provides little information regarding the fate of unlabeled products. Therefore, [ ^'^C-U]cellooligosaccharides were used in conjunction with [1-3H]cellooligosaccharides as substrates for CBH II.

PAGE 259

241 The distribution of products formed from hydrolysis of [''^C-Ujcellopentaose and [1 -^Hjcellopentaose by CBH II is shown in Table 7-2. The formation of [ ^C-U]glucose and the corresponding absence of [ 1--^H]glucose suggests that CBH II hyrolyzes bond 1. The absence of a corresponding amount of [1-^H] cellotetraose suggests that CBH II repetitively attacks cellopentaose initiating attack at bond 1 followed by attack at bond 3. The efficiency of repetitive attack estimated from the concentrations of [''^C-Ujgl ucose and [1 --^Hjcellotetraose is 80 percent. Since cellobiose and cellotriose are produced from cleavages at bonds 2 and 3, the [''^C-U]cellobiose/['''^C-U]cellotriose ratio is 1. The [''^C-U] cellobiose/ [1-^H] cellobiose ratio of 2 corresponds with the sum of concentrations of [1--^H]cellobiose and [1--^H]cellotriose divided by [1--^H]cellobiose. Similarly, the [''^C-U]cellotriose/[1-^H]cellotriose ratio of 1.6 corresponds with the sum of the concentrations of [1-^H] cellobiose and [1-^H]cellotriose divided by [1-^H]cellotriose. Thus, the distribution of ["^"^C-U] products and [1-^H] products resulting from hydrolysis of bond 2 and 3 indicate that one bond is hydrolyzed per encounter with CBH II. Initial rates of formation of products from CBH II catalyzed hydrolysis of ['''^C-Ujcellopentaose and [1-%]cellopentaose are listed in Table 7-3. Initial rates of hydrolysis of [''^C-Ujcellopentaose and [ 1 -^Hjcellopentaose are within experimental error indicating no detectable kinetic isotope effect. Ratios of formation of ["^^C-U]/

PAGE 260

242 I 1 x: o •H o o I — I X I U o CO •H CO >, 1— t o Li T3 >. CO CD S 3 •H T3 O cn ch o S -H o LO =^ tH CM O CVJ rH O 3. T3 • O C S O CD =^ cu C CO CO •H O CO O -P o c o 0) O CO +J O CO r-l CO cu -P O C 0.—. Q) -p X a coro o JQ I t-i C— I QJ c o •H H O Ch CO O CO -H ^ o c =• o boin -p E in CO • u o +J c c o I o X CO c CVJ -H • CD KO -P ^ c o M O CD o 8H 0 o CQ • • CO in Q) rH CQ o a: o IQ, CD T) -P >) ~ C j:: u — > O -P O CO I -P O Oreo I — -P CO o s a. 1 — 1 X m 1 CM • 1 1 • • CO O o O Q -H •H -H -H o 1 • • o CVJ O ro, o o o o o o o VD O o o in CM >~ CM rn CM O -H -H +1 -H -H -H -H -H -H in o o o O O ^~ m o o rO CM O m V O O rrm in CM CM ^ CM O O o o o o CM VO CM CO rn ^ CM rCM CM C\j in •H -H H -H -H -H -H -H -H O VO O O o o rn o o T•^o o TV C\J 00 o \o CM (T\ rv uD in m m o o o o in o o o rin -H -H -H -H -H -H -H -H -H CO in rn CM O O O rO o o in CM 00 VD CM r— T— in O O (7^ in r— C\J m o o o o o ^ 00 in -H -H H -H -H -H -H H -H -51m VO O o o \o rn o o CO o CO 00 CM KO fn in o fn o -H -H o o o o o o o o TCO Tm CO • cu (U (U a) -a (U CO cu CO •H cu CO o CQ O H u cu cu CO O CD O CD B CO CU CO CO o CD t-> CO -P sz CO O O -H l4 -P -P c — 1 o O 3 -P to 3 rH 1— t CU rH (U >— 1 Q) f-H Q) o •H rH C^ o-— Or— r U'—> o f— ':3 r— in -— .=) r— .3 r— .3 II o X 1 X 1 X 1 X 1 X 1 m o m o Q 1—1 2: 0) o 1 R— J CO

PAGE 261

243 Table 7-3 InitiaJ Rates of Cellobiohydrolase II Activity on [1-^H]Cellopentaose and [''^C-UjCellopentaose Cellobiohydrolase II (6.28x10~-'mg) was incubated at 40C in 100 yl of 5 mM sodium acetate buffer, pH 5.0, containing 0.55 uCi of [1-^H]cellopentaose and 0.22 uCi of [ '^C-Ujcellopentaose Total concentration of cellopentaose was 83 yM. Cellooligosaccharide Initial Velocity (pmol/min) -^HjGlucose "^C-UjGlucose 0.8 0.4 5 4 6.255.9 -^H]Cellobiose ^C-UjCellobiose 76 6 149 20 2.0 0.3 -^HjCellotriose 4c-U]Cellotriose 94 10 144 17 1.5 0.2 -^HjCelltetraose ^C-U]Celltetraose 2 2 -4 2 -^H] Cellopentaose 4c-U]Cellopentaose -172 18 -144 17 0.840.13 ^ ND = Not determined

PAGE 262

244 [1-^H] products agree well with ratios of [^^C-U] product distributions supporting the contention that products result from hydrolysis at the same bonds. The distribution of products formed from hydrolysis of [^^C-(J]cellohexaose and [1--^H]cellohexaose are listed in Table 7-4. Although bond cleavage frequency at bond 1 resulting in the formation of [1--^H]cellopentaose (<0.005) could not be precisely determined because of high background, the formation of ['''^C-U] glucose may indicate that cleavage occurs at bond 1. However, [ '''^C-Ujglucose may arise from cleavage at other bonds. The formation of [1 -^Hjcellobiose indicates that CBH II catalyzes hydrolysis at bond 4; however, the ["^^C-Ujcellobiose/[1 -^Hjcellobiose concentration ratio of 2.9 indicates that cellobiose also arises from cleavage at bonds other than bond 4. The 2.9-fold greater concentration of [''^C-Ujcellobiose compared to cellohexaose during the reaction indicates that CBH II repetitively attacks cellohexaose. Cellobiose may arise from cleavage at bond 2 and then at bond 4 (Fig. 7-14a) or from cleavage at bond 4 and then at bond 2 (Fig. 7-14b). Since repetitive attack from either direction results in the same ratio of ['''^C-U] cellobiose to [1-^H]cellobiose, the direction of repetitive attack must be ascertained from the [^^C-U]/[1-^H] ratios of other products, especially cellotetraose. Initial cleavage at a bond resulting in the formation of cellobiose will also result In the formation of an equal amount of

PAGE 263

245 X3 C CD (1) 03 O CO X cu s: o 03 CO i-H o u xs >, jC o •H O I 03 rH CO EH 8^ to* o 03 •H 03 (U 03 O CO X (U x: o 0) o r— 0 o 0 s o • • • • 3 •H CM o O O 0 •H -P -H -H •H -H CD -H CO Q VO o cd Q CO 03 tH 1 cr\ CO o 00 o U # • Q CM S -H rLPl — (h (T* O T— o o o o o o 1— 1 o in 00 CM O T— r — 0 CO CM CM 1^ C\J -H 44 -H -H -H -H 4) -H -H -H ^ — o o O O O O O O 0 0 O CO 3. t>o Q> LPs CO V V 0 CM O CO CM CM 00 VD CM ^— ^™ C 03 lA *H O CO 03 T— ''J" o o O O rO CO 0 0 O X CO ir\ CD VO o ^— LA O (US — O £ -H -H +f 41 -H -H -H -H 44 -H -H ^ O d) in r— m a o C~) O lO O CO T— (—1 03 T— CM V CO -P r-H O CO O CT^ ^ ^ — CO (U CO CM O -P T) — C J— d) X •H -PK^ Q. C2 O CO CO 1 O o CM CM LTN CM 'sT ^ ^ rH I—* B -H -H ft -H 41 44 41 44 44 3 1 1 rH U-i O (U • C 'H O M in CM t<^ O V — f ^ / CD O '^-^ / •H O EH o CD C — CO IP\ o 03 tH O ^ — CO U o LO "iH CM V.D o o ^ \j ^ bOiA -P B • CO r*\ 1 -H -H -H -H -H -H 41 41 •H -H lA O t-i CO o 1 -P • • • O 60 C C — CO OTC C r-H O 44 44 "1 -I O CO O O O O O O O O O O M O -P 1— 1 o •> Eh 0) o 03 • CO • d) d) dJ rH d) TJ d) CO CU O X 03 •H 03 O CQ O £Q. O C-i d) 03 O CO O CO CQ 0 T3 CO CO dJ 03 03 O — ^ >> X s: (Ti Srf* o CD O *H *r-] lU ^ \U O (U -C o d) 03 •H X2 C-I -P m C iH Ch O CO 03 O n o \J Am A 0 3 rH rH 0 r-H dJ r-H OJ ^ (1) rH O •H rH a dj O dJ O d) o O-— o-— Or-i O -P o r— .=3 >— 1 1=) r— ,3 r— ,0 CO 1 o as 1 X 1 X 1 X 1 a: 1 DC 1 -P O rH o O roi 0 arrt f— I Ord) r-T— \ T eo-—' o 1

PAGE 264

Figure 7-14 Possible types of repetitive attack of CBH II with [1-^H]cellohexaose and [ '^C-U]cellohexaose Cellobiohydrolase II may repetitively attack cellohexaose in either of two possible directions: (A) attack toward the reducing end initiated at bond 2 and followed by attack at bond 4, (B) attack toward the nonreducing end initiated at bond 4 and followed by attack at bond 2. Bonds are numbered from the nonreducing end. The reducing end glucosyl residue is represented by (j^).

PAGE 265

247

PAGE 266

248 cellotetraose. If all resulting cellotetraose remains associated with CBH II and is immediately converted to cellobiose, then repetitive attack is 100 percent efficient. The efficiency of repetitive attack in either direction estimated from the concentration of [^"^C-Ujcellotetraose and [1 -^Hjcellobiose at 60 min is 83 percent. If repetitive attack occurs toward the nonreducing end and is initiated at bond 4, then no [ 1 -^Hjcellotetraose should be formed. However, if repetitive attack occurs toward the reducing end and is initiated at bond 2, then [1 -^H]cellotetraose should be formed and the [''^C-U]cellotetraose/[1-^H]cellotetraose ratio should be equal to 1. The ratio of [''4c-U]cellotetraose/[1-^H]cellotetraose concentration is found to be equal to 1 supporting the model (Fig. 7-14a) that the direction of repetitive attack is toward the reducing end of cellohexaose. Distribution of [''^C-U] products and [1--^H] products resulting from CBH Il-catalyzed hydrolysis of ['''^C-Ujcellopentaose and [1--^H]cellopentaose is also consistent with repetitive attack toward the reducing end. The preference of CBH II for the nonreducing end is supported by the accumulation of ['''^C-U] glucose arising from cleavage at bond 1. The negligible accumulation of [1-%]glucose indicates that CBH II has little tendency to cleave the terminal reducing end bond (bond 5). Initial rates of formation of products from CBH II catalyzed hydrolysis of [^^C-U]cellohexaose and

PAGE 267

249 [1-^H]cellohexaose are listed In Table 7-5. Initial rates of hydrolysis of [''^C-Ujcellohexaose and [1 -^Hjcellohexaose are within experimental error indicating no detectable kinetic isotope effect. Ratios of initial rates of formation of products agree well with ratios of [''4c-U]/[1-^H] product distribution supporting the contention that products result from hydrolysis at the same bonds. In contrast to CBH 1(D), evidence from hydrolysis of cellohexaose in conjunction with cellohexaose supports that CBH II repetitively attacks cellohexaose from the nonreducing end. The e-glucosidase also prefers the nonreducing end of cellooligosaccharides ; however, no evidence for repetitive attack by the s-glucosidase has been obtained. The difference between CBH 1(D) and CBH II with respect to their directions of repetitive attack suggests a synergistic mechanism by which CBH 1(D) and CBH II may degrade crystalline cellulose. Possible mechanisms to explain exo-exo synergism are presented in the following chapter. Subsite Mapping of Cellobiohydrolase II The methods of Suganuma et al. (45) and Allen and Thoma (49) were used to construct subsite maps of CBH II. The method of Suganuma et al. (45) requires bond cleavage frequencies and values of Viaax^^m ^ series of oligosaccharides to estimate subsite affinities. The method of Allen and Thoma (49) requires bond cleavage frequencies;

PAGE 268

250 Table 7-5 Initial Rates of Cellobiohydrolase II Activity on [1-^H]Cellohexaose and [ '^C-U]Cellohexaose Cellobiohydrolase II (4. 52x10~--'mg) was incubated at 40C in 100 Ml of 5 mM sodium acetate buffer, pH 5.0, containing 0.55 uCi of [1-^H]cellohexaose and 0.19 uCi of [^^C-Ujcellohexaose Total concentration of cellohexaose was 59 uM. Initial Velocity Ratio Cellooligosaccharide (pmol/min) [^'+C-U]/[1-^H] 1-r^HjGlucose ^4c-U]Glucose 1 15. 4 8 0.1 0.2 11 .40.8 1-^H]Cellobiose ^^C-UjCellobiose 85 240 2 10 2.9 0.1 Ir^HjCellotriose ^^C-UjCellotriose 107 196 2 6 1.830.06 l-r^HjCelltetraose ^^C-UjCelltetraose 13 17 1 3 1.3 0.2 1 -•^HjCellopentaose ^^C-U]Cellopentaose 18 16 4 5 0.9 0.3 1-^H] Cellohexaose ''4c-U]Cellohexaose -224 -207 8 13 0.920.07

PAGE 269

251 however, values for Kjjj, V^g^^ and K^^^ are also required. Equations essential to the formulation of both methods are presented in Appendix A. Subsite affinities calculated by the method of Suganuma et al. (45) using major bond cleavage frequencies and values of Vjjjg^/Kjj, for [1-^H]cellotriose through [ 1 -^Hjcellohexaose are listed in Table 7-6. The active site of CBH II consists of 7 subsites with catalytic groups located between subsites 3 and 4 (Fig. 7-15). Subsite 2 contributes the greatest proportion of binding energy, whereas subsites 1, 6 and 7 provide maximum binding energies of approximately -800 cal/mol/subsite Subsites 3 and 4 are common to all productive modes, therefore individual affinities cannot be evaluated. More than one combination of bond cleavage frequencies and Vjjgjj/Kjjj values are used to calculate affinities for subsites 1 and 6. However, affinities for subsite 6 were determined to be 560 cal/mol and -740 cal/mol. Unfavorable and favorable binding affinities for subsite 1 were also obtained: 600 and -680 cal/mol. Since the importance of subsites 1 and 6 cannot be ascertained using bond cleavage frequencies and values of V_^^/K„, the method of Suganuma et al. (45) is not adequate for mapping subsites in the active site of CBH II. The subsite map constructed using the computer model for depolymerases developed by Allen and Thoma (49) was first optimized for experimental bond cleavage frequencies and location of catalytic groups. Bond cleavage frequencies

PAGE 270

252 Table 7-6 Subsite Affinities of CBH II Calculated Using Values of V/K ^nd Bond Cleavage Frequencies From [1 -^HjCellooligosaccharides Numbers in brackets indicate mode of calculation using bond-cleavage frequencies of cellooligosaccharides at glycosidic bonds denoted by superscripts in parentheses. Subsite Mode of Calculation Subsite Affinity (cal/mol ) 1 600 70 -680 70 2 -76001200 5 [a^2)/Gi2)] -4720 70 6 [g|2)/G^2)] 560 70 [G^3'/Gp)] -740 70 7 [G^2)/0^2)] -590 70

PAGE 271

Figure 7-15 Subsite map for CBH II constructed using values of Ynjax/^m ^'^^ bond cleavage frequencies for [1 --^H jcellooligosaccharides Subsite affinities were estimated using the method of Suganuma et al. (45) from values of V_|3„/Kjjj and bond cleavage frrequencies for [T-^H]cellooligosaccharides. The arrow shows the position of catalytic groups. Since subsites adjacent to catalytic groups are common to all productive complexes, their affinities cannot be evaluated using this method. Shaded bars indicate that more than one combination of Vjjj /K values and bond cleavage frequencies for [1Hjcellooligosaccharides differing in chain length by one glucosyl residue were used to evaluate affinities at a particular subsite (Table 7-6).

PAGE 272

254 J 1 \ I 1 I I I 2 3 4 5 6 7 SUBSITE

PAGE 273

255 calculated from the optimized subsite map agreed within 5 percent of experimentally derived bond cleavage frequencies. The resulting subsite map was then optimized for values of and '^^^^ for [ 1 -^Hjcellotriose through [1-^H]cellohexoase. Values of and V^^^ calculated from the optimized subsite map for 7 of 11 constants agreed within 20 percent of experimentally derived values of Kj^ and Vjjjgjj. Inhibition constants for glucose (24 mM) and cellobiose (204 uM) and apparent ^j^g^^ for hydrolysis of cellobiose (1.13x10"-^ umol/min/mg, data not shown) were also used in the evaluation of subsite affinities. Inhibition constants for glucose and cellobiose were determined during inhibition of CBH II hydrolysis of [1 -^Hjcellotetraose. A minimum of 7 subsites with the catalytic groups located between subsites 3 and 4 was required to account for the bond cleavage frequencies. The value of ^^Yit estimated from the K^j^ of cellohexaose. The subsite map for CBH II yielding the lowest residual error for bond cleavage frequencies, K^j, Vjjjgjj and K^^^^ is shown in Fig. 7-16. Subsite 2 contributes the greatest proportion of binding energy, whereas the affinity at subsite 4 is unfavorable. Subsites 5-7 favorably bind glucosyl residues and may retain products of cellooligosaccharide hydrolysis in the active site for another cleavage. Unfavorable binding at subsite 4 may promote catalysis by straining the glycosidic bond positioned near catalytic groups. When the approximation was made that the binding of a glucosyl residue

PAGE 274

Figure 7-16 Subsite map for CBH II constructed using values of Kp, ^max bond cleavage frequencies for [ 1 -^H ] ce 1 loo ligosaccha rides The number of subsites and the position of the catalytic groups were determined using the method of Allen and Thoraa (48) from values of K_, V_g^ and bond cleavage frequencies for [ I --^H Jcel looligosaccharides. The arrow shows the position of the catalytic groups. The bars depict the estimated binding energies including the contribution of the acceleration factor (500 cal/mol).

PAGE 275

257 SUBSITE

PAGE 276

258 decreases the activation energy for hydrolysis by CBH II by a constant amount enhancing the rate of bond cleavage, agreement between predicted and experimentally derived values significantly improved. The approximation, which is termed the acceleration factor, was estimated to be 500 cal/mol. The subsite maps constructed for CBH II using the method of Suganuma et al. (45) and Allen and Thoma (49) indicate that the active site of CBH II consists of 7 subsites with subsite 2 contributing the greatest proportion of binding energy. Although subsites 5-7 on each map indicate favorable binding, subsite 6 on the map constructed using the method of Suganuma et al. (45) also shows unfavorable binding. Subsite maps reveal that the active sites of CBH 1(D) and CBH II consists of 7 subsites; however, the location of catalytic groups is different. The catalytic groups are located between subsites 5 and 6 of CBH 1(D), whereas they are located between subsites 3 and 4 of CBH II. Favorable binding affinities at subsites 1-5 may enable CBH 1(D) to repetitively attack cellohexaose toward the nonreducing end. Similarly, subsites 5-7 of CBH II may promote repetitive attack of cellohexaose toward the reducing end. Conclusions Action patterns of [1--^H]cellooligosaccharides indicate that CBH II preferentially hydrolyzes cellooligosaccharides in two productive modes. Michaelis parameters for

PAGE 277

259 [1--^H]cellooligosaccharide hydrolysis decrease with increasing chain length indicating that CBH II preferentially hydrolyzes longer cellooligosaccharides. Examination of action patterns of [1-^H]cellohexaose and ['''^C-Ujcellohexaose revealed that CBH II repetitively attacks cellohexaose from the nonreducing end. Subsite maps constructed by two methods using kinetic parameters of bond cleavage frequencies showed that the active site of CBH II consists of 7 subsites with the catalytic groups located between subsites 3 and 4.

PAGE 278

CHAPTER 8 GENERAL DISCUSSION Hydrolysis of [ 1 -^H] Cellooligosaccharides The synthesis of [ 1 -^Hjcellooligosaccharides and their subsequent use with g -glucosidase CBH 1(D) and CBH II permitted identification of precise glycosidic bonds cleaved by these enzymes and estimation of Kjjj and Vjjjg^ for substrate hydrolysis. Asymmetrically modified cellooligosaccharides, such as 4-niethylumbellif eryl glycosides of cellooligo. saccharides (30) and reduced cellooligosaccharides (98), have been used to investigate the action patterns of cellulolytic enzymes. However, the influence of the derivatization on the activity of cellulolytic enzymes has not been determined. van Tilbeurgh et al. (30) reported that the Kjj, for cellotriose hydrolysis by CBH 1(D) is 12 mM, which is approximately 1000-fold and 4-fold higher than the high and low affinity sites, respectively, reported here. The discrepancies may be due to the difference in the assay used in this report and that used by van Tilbeurgh et al. (30). The formation of glucose was measured by van Tilbeurgh et al. (30) using the glucose oxidase-peroxidase reagent, the sensitivity of which does not permit measurement of initial rates of glucose formation at micromolar substrate

PAGE 279

261 concentrations. Hydrolysis of the aglycone bond of MeUmb(Glc_2_)2 was shown to have a of 35 uM, approximately 4-fold higher than Kjjj for the high affinity site of cellotriose reported here. This may indicate that subsite 5 of CBH 1(D) has less affinity for hydrophobic aglycones than for glucosyl residues. A comparison of the action patterns resulting from CBH 1(D) hydrolysis of MeUmb(Glc_2_)2 through MeUmb(Glc_2_)g with those of [1-^H]cellotriose through [1--^H]cellohexaose indicates that the 4-methylumbellif eryl aglycone alters the action patterns of CBH 1(D). Hydrolysis of MeUmb(Glc_£_)4 produces bond cleavage frequencies of bonds 2, 3 and 4 of 0.30, 0.53 and 0.13 estimated by this author using peak heights reported in (30), respectively, whereas the corresponding bond cleavage frequencies for [1-^H]cellopentaose are 0.02, 0.513, 0.480, respectively. The lower bond cleavage frequency at bond 4 of MeUmb(Glc2.)4 is consistent with the idea that subsite 5 is more specific for glucosyl residues than for hydrophobic methylumbellif eryl groups. However, hydrolysis of MeUmb(Glc_£^)5 results in a bond cleavage frequency at bond 5 of approximately 0.75. Bond cleavage frequency at bond 5 of [ 1 -^Hjcellohexaose is 0.62 suggesting that subsite 1 is crucial to the positioning of cellooligosaccharides as well as MeUmb(Glc_£_) glycosides in the active site of CBH 1(D). The change in bond cleavage frequencies with derivatizatlon confirms the utility of using [1-^H]cellooligosaccharides to investigate the action patterns of cellulolytic enzymes.

PAGE 280

262 Gum (98) investigated the action patterns of CBH 1(D) of T. viride using reduced cellooligosaccharides. When reduced cellohexaose was used as substrate, CBH 1(D) exhibited little tendency to cleave bond 5, which links the sorbityl residue to previous reducing end of the cellooligosaccharide. The decrease in the formation of glucose may be explained in two ways: (1) if bond 5 is normally hydrolyzed by CBH 1(D), then the subsite, which is adjacent to the catalytic groups and may position the glucosyl residue properly, rejects the sorbityl residue or (2) if bond 1 is normally hydrolyzed, then the subsite, which is 5 subsites away from the catalytic groups and may be essential to positioning cellohexaose properly, rejects the sorbityl residue and bond 1 is no longer hydrolyzed. Cellobiohydrolase 1(D) produces [ 1 -^H]glucose by hydrolysis at bond 5, therefore explanation (1) above is correct. The loss in specificity of cleavage at bond 5 resulting from the reduction of cellohexaose emphasizes the importance of using [1--^H]cellooligosaccharides to investigate the action patterns of cellulolytic enzymes. Gritzali (6) investigated the action patterns of CBH 1(D) and CBH II on cellotriose through cellohexaose. Since the observed molar ratios of products could have arisen from more than one combination of cleavages, the precise identity of glycosidic bonds cleaved could not be determined.

PAGE 281

263 Synergism of Gellulolytic Enzymes Mixtures of purified endoglucanases, cellobiohydrolases and e-glucosidases of T. koningii F. solani and P. funiculosum synergistically degrade various types of cellulose (100). Cotton is degraded with a high degree of synergism, whereas phosphoric acid-swollen cellulose shows less synergism and CM-cellulose is not synergistically degraded. Mixtures of component cellulolytic enzymes of T. reesei have also been shown to synergistically degrade cellulose (6,7). Furthermore, Fagerstara and Pettersson (7) have shown that CBH 1(D) and CBH II synergistically degrade cellulose and have referred to this synergism as exo-exo synergism. Endo-exo synergism is believed to result from the concerted action of endoglucanases and cellobiohydrolases. Endoglucanases hydrolyze internal glycosidic bonds of cellulose producing chain ends from which cellobiohydrolases remove cellobiosyl residues. Wood and McCrae (18) have suggested that a "loose complex" between cellobiohydrolases and endoglucanases may form at the surface of cellulose. After endoglucanase cleaves a glycosidic bond, cellobiohydrolase immediately binds to a chain end and begins removing cellobiosyl residues before the glycosidic bond reforms. The difference in the specificity of endoglucanases and cellobiohydrolases provides a basis for synergism. The exo-exo synergism described by Fagerstam and Pettersson (7) for CBH 1(D) and CBH II is not readily

PAGE 282

264 explained by a difference in specificity, because each enzyme removes cellobiosyl residues from chain ends. However, evidence provided in this report suggests that CBH 1(D) removes cellobiosyl and glucosyl residues from the reducing end of [1 -^Hjcellooligosaccharides, whereas CBH II removes cellobiosyl residues from the nonreducing end. Since no evidence for the formation of a complex between CBH 1(D) and CBH II has been provided, alternative explanations must be offered for synergism. To explain how CBH 1(D) and CBH II by removing cellobiosyl residues from chain ends in opposite directions results in synergism, the structure of cellulose needs to be considered. Two major forms of cellulose, cellulose I and cellulose II, have been described (9). Cellulose I, native crystalline cellulose, is considered from X-ray diffraction studies to be arranged in parallel chains, whereas the arrangement of chains in cellulose II are antiparallel However, Atalla and VanderHart (11) has recently reported that native celluloses isolated from various sources contain different degrees of cellulose I and cellulose II. The amount of cellulose I decreases in order of Acetobacter xylinium Valonia ventricosa cotton, ramie and cellulose II (regenerated cellulose I). Avicel, which was the substrate used during the assay showing exoexo synergism, is isolated from a-cellulose and is presumably cellulose II (18), Microfibrils of cellulose I and cellulose II may be diagrammatically represented as in Fig. 8-1. Cellulose I

PAGE 283

Figure 8-1 Structure of cellulose I and cellulose II The structure of cellulose I (A) and cellulose II (B) have parallel and antiparallel chains, respectively. Reducing end glucosyl residues are represented by i).

PAGE 284

2c6

PAGE 285

267 may be considered as a block with nonreducing chain ends protruding into solution on one side and on the other side reducing chain ends protruding into solution. If CBH 1(D) and CBH II acted on cellulose I, the product would be a truncated version of that shown in Fig. 8-1 a; however, the activity of either enzyme would not produce appropriate chain ends for the other. Therefore, synergism would not be expected. If CBH 1(D) acted on cellulose II (Fig."8-1b), then CBH 1(D) would expose reducing chain ends from which cellobiosyl residues may be removed by CBH II. Cellobiohydrolase II, on the other hand, would expose reducing chain ends from which CBH 1(D) would remove cellobiosyl residues. The production of more accessible chain ends by either enzyme would result in synergism. Another important feature of cellulose is the cellobiose repeating unit. Wood (100) has noted that the glycosidic bonds of contiguous glucosyl residues, which are held in position by hydrogen bonds, are in different planes. An enzyme proceeding down a cellulose chain encounters an array of hydroxyl groups and hydrogen bonding which is different from those of the adjacent residue (Fig. 8-2). The ability of CBH 1(D) and CBH II to bind [ 1 -^Hjcellooligosaccharides in two productive modes allows them to hydrolyze the appropriate chain ends regardless of the orientation of the first glycosidic bond they encounter. The alternating position of the glycosidic bonds may also explain the action pattern of CBH 1(D) on [1-^Hjcellohexaose. The bond cleavage

PAGE 286

CM I 00 a> a 00 0) u CD m o 0) o 0) o 0} (U 3 TJ •H CO (1> u >> CQ O O D i—t 00 CO 3 O 3 60 O -H c o o o CO T3 c o CO o o > r-l >'H I— t -H

PAGE 287

269 o o J.

PAGE 288

270 frequencies for bonds 1, 2, 3, 4 and 5 of [1 --^Hjcellohexaose are <0.005, 0.005, 0.025, 0.35 and 0.62, respectively. The action patterns of CBH 1(D) on [1 -^Hjcellohexaose and ^C-U]cellohexaose indicate that attack at bond 5 is immediately followed by attack at bond 3. Since the bond cleavage frequency at bond 4 of [1-^H]cellopentaose yielding glucose is 0.48, bond 4 of the original cellohexaose might be more likely cleaved than bond 3 after hydrolysis at bond 5. However, cleavage of the resulting cellopentaose at bond 4 would produce a significantly higher ucose/ [1--^H]glucose ratio than observed. When the stereochemistry of the cellulosic substrate is considered, bonds 1, 3 and 5 have the same orientation, whereas bonds 2 and 4 have the opposite orientation. Therefore, repetitive attack initiated at bond 5 must immediately proceed to bond 3. Individual Roles of Cellulolytic Enzymes The cellulase system of T. reesei comprises endoglucanases, CBH 1(D), CBH II and 8-glucosidases which together act sequentially and cooperatively to degrade crystalline cellulose to glucose. Through the use of [1-^H]cellooligosaccharides and ['''^C-U]cellooligosaccharides the roles of CBH 1(D), CBH II and 6-glucosidase are more clearly understood. The increase in the affinity of e-glucosidases for cellooligosaccharides as chain length increases suggested to Shewale (90) that the primary role of the 8-glucosidase is to convert cellooligosaccharides to glucose during the saccharif ication of cellulose. Comparison of k„„^/K-, values

PAGE 289

271 for hydrolysis of [1-^H]cellotriose by 6-glucosidase CBH 1(D) and CBH II indicates that [ 1 -^H]cellotriose is hydrolyzed 170and 200-fold more efficiently by s-glucosidase than by CBH 1(D) and CBH II, respectively (Appendix F). Although the l^cat/^m value of [1-^H]cellotetraose hydrolysis by e-glucosidase is approximately 2-fold higher than that of CBH 1(D), the k^at/^m value for [1 -^Hjcellotetraose hydrolysis by CBH II is approximately 3.6-fold higher than that of 6-glucosidase. The ^Q^t^^m values for hydrolysis of [1-^H]cellopentaose and [1-^H]cellohexaose by CBH 1(D) and CBH II are greater than 3-fold higher than the corresponding values for the 8-glucosidase. Furthermore, CBH 1(D) and CBH II show little tendency to hydrolyze cellobiose. Thus, the role of the 6-glucosidase is primarily to hydrolyze cellobiose and cellotriose. The role of the 6-glucosidase is reflected in the topography of its active site, which consists essentially of three subsites. The first subsite may serve to anchor substrate and position the scissile bond across the catalytic groups. Binding affinity for glucosyl residues at subsite 2 is negligible; however, subsite 2 has strong affinity for hydrophobic aglycones of glucosides; such as 2_-nitrophenyle-D-glucoside and methylumbelliferyl-e-^-glucoside. Values ^ ^max/^m p-nitrophenyl-6-D-glucoside and raethylumbelliferyl-6-D-glucoside have been reported to be 30and 125-fold higher than that for cellobiose (8). Low affinity for glucosyl residues at subsite 2 may promote

PAGE 290

272 diffusion of product after hydrolysis. Significant affinity for glucosyl residues at subslte 3 permits the B-glucosldase to scavenge cellotriose, which is a poor substrate for CBH 1(D) and CBH II, produced during saccharif Ication of cellulose. Affinities at subsites 4, 5 and 6 may represent nonspecific interactions between S-glucosidase and longer cellooligosaccharides Cellooligosaccharides of chain length greater than three will be rapidly hydrolyzed by CBH 1(D) or CBH II. The S-glucosidase exhibits little tendency to repetitively attack cellooligosaccharides supporting the idea that its role in the cellulase system of T. reesei is primarily to convert cellobiose and cellotriose to glucose rather than to depolymerize cellooligosaccharides. The roles of CBH 1(D) and CBH II are to remove cellobiosyl residues from reducing and nonreducing chain ends of cellulose, respectively. Furthermore, CBH 1(D) and CBH II repetitively attack [1-^H]cellooligosaccharides. Repetitive attack of cellooligosaccharides by CBH 1(D) and CBH II has been shown previously by Gritzali (6). Berghem and Pettersson (101) have shown that a cellobiohydrolase from T. viride successively removes cellobiosyl residues from phosphoric acid-swollen cellulose. The lower activity of CBH 1(D) on [1-^H]cellooligosaccharides compared to CBH II may explain why T. reesei secretes more CBH 1(D) than CBH II; the amount of CBH 1(D) and CBH II in extracellular filtrates has been estimated at 60 and 25 percent (weight of protein), respectively (5).

PAGE 291

273 The subsite maps of CBH 1(D) constructed using the methods of Suganuma et al. (45) and Allen and Thoma (48) show significant binding affinity for glucosyl residues at subsites 1-4 and 7. The catalytic groups are located between subsites 5 and 6. Subsites distant from the catalytic groups and toward the nonreducing end may permit CBH 1(D) to hold a portion of the substrate in the active site after hydrolysis. The absence of significant binding beyond subsite 7 may indicate the end of the active site. A physical block at the nonreducing end of the active site of glucoamylase has been proposed to explain the formation of only one productive complex in glucoamylase-catalyzed hydrolysis of maltooligosaccharides (44). Unfavorable affinity at subsite 6 may promote catalysis by introducing strain into the scissile glycosidic bond or by promoting diffusion of product away from the active site after hydrolysis. Unfavorable binding at one of the binding sites adjacent to catalytic groups has been reported for glucoamylases (44,47), B. amyloliquefaciens a-amvlase (49) and lysozyme (40). Aspergillus oryzae a-amylase is a repetitive attack enzyme for which a subsite map has been constructed using the method of Allen and Thoma (49). The substrate binding region is composed of 7 subsites of which subsite 2 contributes the greatest proportion of binding energy. The catalytic groups are located between subsites 3 and 4. Subsite 7 has unfavorable binding and acts as a "barrier subsite" reducing hydrolysis of certain binding modes.

PAGE 292

274 The unusual dependence of bond cleavage frequencies on concentration of [1--^H]cellotriose may be explained by the topography of the active site of CBH 1(D). Since the major product of [1-^H]cellotriose hydrolysis at low concentrations is [1-^H]glucose, [1-^H]cellotriose must reside in subsites 4, 5 and 6 with the reducing end glucosyl residue bound in subsite 6. At high concentrations of [1-^H]cellotriose, a second molecule of [1-^H]cellotriose may bind to subsites 2, 3 and 4, which have significant affinity for glucosyl residues, and shift [ 1 --^Hjcellotriose in subsites 4, 5 and 6 to subsites 5, 6 and 7. Thus a larger proportion of [1-^H]cellobiose would be produced. Subsite maps, constructed using the methods of Suganuma et al. (48) and Allen and Thoma (45), for CBH II show that catalytic groups are located between subsites 3 and 4, Subsites 5, 6 and 7 at the reducing end of the active site have significant affinity for glucosyl residues. Significant binding affinity for glucosyl residues at subsites 5, 6 and 7 located toward the reducing end of substrates may serve to retain cellooligosaccharides in the active site for subsequent hydrolysis. Unfavorable binding at subsite 4 may promote catalysis by straining the scissile bond or facilitating diffusion of products from the active site. Methods of Subsite Mapping There are several methods available in the literature to evaluate subsite affinities of polysaccharide depolymerizing enzymes (45,45,48,71). The application of a method

PAGE 293

275 depends on the number of productive binding modes that an enzyme forms with oligosaccharide substrates. Since the method of Hiromi et al. (44) can be applied to enzymes that bind substrates in only one productive complex, this method was used to evaluate subsite affinities of the S-glucosidase. The method of Roeser and Legler (71) was originally used to assess the contribution of individual parts of substrate to binding in glycoside hydrolysis by 3-glucosidase from A. wentii However, the method of Roeser and Legler (71) may also be applied to evaluating the contributions of individual glucosyl residues of cellooligosaccharides When the formation of more than one productive complex is detected, the methods of Suganuma et al. (45) and Allen and Thoma (48) are more appropriate, because they partition affinities of different subsites according to bond cleavage frequencies. Since CBH 1(D) and CBH II bind cellooligosaccharides in more than one productive complex, the methods of Suganuma et al. (45) and Allen and Thoma (48) were used to evaluate their subsite affinities. The equations essential to the formulation of each subsite mapping procedure are presented in Appendix A. The method of Roeser and Legler (71) was used to evaluate the contribution of glucosyl residues of cellooligosaccharides to binding by measuring cellobiose through cellohexaose inhibition of methylumbelliferyl-8-D_-glucoside hydrolysis catalyzed by the 3-glucosidase. Subsite affinities will be accurate if inhibition constants

PAGE 294

276 represent true dissociation constants. Since cellooligosaccharides are substrates for the g-glucosidase their inhibition constants may not be true dissociation constants. However, Lineweaver-Burk and Dixon plots of inhibition by cellobiose through cellohexaose of MUG hydrolysis are linear indicating that the inhibitors are not significantly degraded during the time course of the assay. The stability of cellooligosaccharides in the assay mixture is expected, because the value of ^^^/K^j for MUG hydrolysis (8) by 8-glucosidase is approximately 4-fold higher than that for cellooligosaccharide hydrolysis. Each cellooligosaccharide competitively inhibits the 3-glucosidase suggesting that cellooligosaccharides bind to the same subsites. Thus, the ratio of inhibition constants for an n-mer and n+1-mer provides an estimate of the subsite affinity for the n+1-mer glucosyl residue. Since previously reported substrate specificity studies (3) suggest that glucose primarily binds toward the nonreducing side of the catalytic groups and that the region of the g-glucosidase which binds the aglycone is hydrophobic, the binding affinity of subsite 1 was evaluated from the for glucose. Affinity at subsite 2 was evaluated from the ratio of inhibition constants for glucose and cellobiose. Hiromi et al. (44) evaluated subsite affinities of glucoamylase from R. delemar using first order rate constants (Vnj5^x/Kni) ^or hydrolysis of a series of maltooligosaccharides. Three important assumptions made by the

PAGE 295

277 authors are (1) 1/Km for each oligosaccharide approximates its association constant, (2) the value of the hydrolytic rate coefficient (k+2) is independent of chain length and (3) subsite affinities are additive. The values of for cellooligosaccharide hydrolysis by the 6-glucosidase are similar to inhibition constants supporting the validity of assumption (1) for S-glucosidase The values of Vjjjg^ for cellobiose through cellohexaose are similar supporting the assumption (2) that k+2 is independent of chain length. Assumption (3) is made by all subsite mapping procedures and is necessary to calculate individuals subsite affinities. Values of ^^aax^^m [1-^H]cellobiose through [1-^H]cellohexaose provided estimates for affinities of subsites 3-6 using the method of Hiromi et al. (44). Affinity of subsite 1 and the value of were evaluated from the intercepts of a linear plot of exp( -AGj^+^ /RT) versus [Eo/^n-1* However, this plot was not linear when corresponding values for the B-glucosidase were used. Since the affinity at subsite 1 cannot be estimated for the 6-glucosidase by Hiromi 's method; the affinity at subsite 2 can not be evaluated. Examination of the plot reported for glucoamylase reveals that the linearity of the plot is completely dependent on the value of V^j^^^ for maltose (44). The value of V^j^^^ for maltose is 5-fold higher than any of the other maltooligosaccharide substrates. Thus, the method of Hiromi et al. (44) does not provide an estimate for ^G^ when Vj^g^^ is constant. Affinities of subsite 1 and 2 must be evaluated

PAGE 296

278 using other experimental approaches. The methods of Roeser and Legler (71) and Hiromi et al. (44) provide similar affinities for subsites 3-6, supporting the validity of using either ratios of Vj^^^/K^j or K^^ to determine these subsite affinities. The method of Suganuma et al. (45) was used to construct subsite maps for CBH 1(D) and CBH II using bond cleavage frequencies and Vj^g^/Kj^ values for [1-^H]cellotriose through [1 -^Hjcellohexaose. The three assumptions made by Hiromi et al. (44) are also made by Suganuma et al. (45). Values of ^^aax^^m* ^^ich provide a direct estimate of subsite affinities, are partitioned using bond cleavage frequencies. Comparisons of hydrolytic rates for pairs of substrates differing in chain length by one glucosyl residue provide estimates for affinities of all subsites except the two subsites adjacent to catalytic groups. Since V^jj^^/K^j values provide information about productive complexes only and subsites adjacent to catalytic groups are common to all productive complexes, affinities of subsites adjacent to catalytic groups can not be evaluated. The method of Suganuma et al. (45) does not require values of and V„^^, which provide information about nonproductive complexes and are essential to evaluating affinities of subsites adjacent to catalytic groups. The method of Suganuma et al. (45) provides estimates of affinities of a particular subsite from more than one combination of substrate and bond cleavage frequencies. The

PAGE 297

279 authors suggest that the most reliable value is that obtained for the major products. However, equally reliable favorable and unfavorable affinities for subsites 2, 3 and 7 of CBH 1(D) and 1 and 6 of CBH II were calculated using the method of Suganuma et al. (45). The discrepancies in subsite affinities may indicate that hydrolytic rate coefficients (k+2) for cellooligosaccharide hydrolysis by CBH 1(D) and CBH II are dependent on chain length. Another disadvantage of the method of Suganuma et al. (45) is that it does not provide estimates of affinities of subsites adjacent to catalytic groups. The subsite mapping method of Allen and Thoraa (48) requires quantitative determination of bond cleavage frequencies and Michaelis parameters (K_ and V_^^) for a series Ul lUci X of oligosaccharides to evaluate a complete subsite map. Substrates of chain length equal to or greater than the entire binding region are used to measure the sum of the subsite binding energies. Since the method of Allen and Thoma (48) does not assume that hydrolytic rate coefficients (1^+2) are independent of chain length, ratio of bond cleavage frequencies for a pair of oligosaccharides provide a measure of the difference between 2 subsite energies and the ratio of hydrolytic rate coefficients. The sum of binding energies for subsites adjacent to the catalytic t groups can be partitioned between the two subsites using ^int' (niicroscopic dissociation constant for a binding mode in which the entire binding region is occupied) and

PAGE 298

280 Michaelis parameters for all other sites. Allen and Thoma (48) also use an acceleration factor, which is the average contribution a glucosyl residue makes to the binding energy, in order to obtain good fits between experimental and predicted bond cleavage frequencies, K--„^., K_ and V„_^. Subsite maps for CBH 1(D) and CBH II were constructed using the method of Allen and Thoma (48) and experimental values of bond cleavage frequencies, and ^^^y^ for [1 --^Hjcellotriose through [ 1 -^Hjcellohexaose The minimum number of subsites required to account for the bond cleavage frequencies of each enzyme is 7. The value of Kjjj for [1-^H]cellohexaose was used as an approximation of K^^^^i. However, this is not an adequate approximation, because [1-^H]cellohexaose is not equal to or greater than the substrate binding region. Cellooligosaccharides of chain length greater than 6 show limited solubility in water. Therefore, the subsite maps for CBH 1(D) and CBH II must be considered approximations of true subsite maps. Subsite mapping is also limited, because CBH 1(D) and CBH II repetitively attack cellooligosaccharides. For example, the apparent V^j^^^^ for hydrolysis of [1--^H]cellohexaose underestimates the catalytic abilities of CBH 1(D) and CBH II, because more than one bond is hydrolyzed per enzymesubstrate encounter. Although values of Kjjj and V^j^^^ derived from initial rates of [1-^H]cellooligosaccharide hydrolysis by CBH 1(D) and CBH II and those predicted from subsite binding energies agree within 20 percent, several differ by

PAGE 299

281 5-fold. The lack of agreement probably reflects the need for values of bond cleavage frequencies, and Vj^^^ for cellooligosaccharides of chain length greater than the substrate binding region. Thus, refinement of subsite maps for CBH 1(D) and CBH II requires (1) experimental conditions where repetitive attack is minimized and (2) the use of [1 -^H]cellooligosaccharides of chain length greater than 6. The synthesis and use of longer [1-^H]cellooligosaccharides may require solvents other than water. Future Experiments The topography of the active sites of S-glucosidase, CBH 1(D) and CBH II may be further investigated using the sensitive assay afforded by [1--^H]cellooligosaccharides. The specificity of particular subsites may be examined using different glycosyl inhibitors. Lactones of cellooligosaccharide aldonic acids may also be used as inhibitors to provide information about subsites which bind nonreducing ends of substrates. Although evidence provided in this report suggests that CBH 1(D) and CBH II attack cellooligosaccharides from reducing and nonreducing ends, respectively, evidence confirming the direction of processivity may be obtained by measuring their ability to hydrolyze derivatized reducing and nonreducing ends of cellulose. The mechanism of exo-exo synergism may be further investigated by measuring the activities of CBH 1(D) and CBH II on celluloses of different forms and crystallinities

PAGE 300

282 Subslte maps of CBH 1(D) and CBH II may be refined by synthesizing [ 1 --^Hjcelloheptaose and [ 1 -^Hjcellooctaose and then measuring bond cleavage frequencies, K^, and ^^g^^ for their hydrolysis.

PAGE 301

CHAPTER 9 SUMMARY To understand the process by which component enzymes of the cellulase system of T. reesei convert native crystalline cellulose to glucose, cellooligosaccharides that are asymmetrically and radioactively labeled are required. In this report, a method was developed for synthesizing cellooligosaccharides labeled at carbon one of the reducing end glucosyl residue with tritium. Asymmetry in [1-^H]cellooligosaccharides permits identification of precise glycosidic bonds cleaved by cellulolytic enzymes. Sensitivity afforded by radioactively labeled cellooligosaccharides allows determination of initial velocities of cellooligosaccharide hydrolysis at microraolar substrate concentrations. A TLC method for separating [ 1 -^Hjglucose through cellohexaose with high resolution and extracting cellooligosaccharides quantitatively is described. This method is superior to previously reported TLC methods for separating cellooligosaccharides and will facilitate analysis of cellulolytic enzymes. Activity on [1-^H]cellooligosaccharides indicated that the S-glucosidase binds [1-^H]cellooligosaccharides in one productive mode and removes glucosyl residues from the nonreducing end. The 0-glucosidase exhibits little tendency to

PAGE 302

284 repetitively attack [1-^H]cellooligosaccharides. Values of V_„ for [1-^H] cellooligosaccharide hydrolysis remain essentially constant as chain length increases, whereas values of Kjjj decrease. Cellooligosaccharides were used to inhibit inethylumbelliferyl-6-D_-glucoside hydrolysis by 8-glucosidase. Inhibition constants for cellooligosaccharides are similar to corresponding values of Kj^ indicating that 1 /Kjj, approximates association constant. Subsite maps constructed using the methods of Hiromi et al. (44) and Roeser and Legler (71) show that the active site consists primarily of 3 subsites. Subsite 1 contributes the greatest proportion of binding energy and the catalytic groups are located between subsites 1 and 2. Thus, the role of the S-glucosidase during saccharif ication of cellulose is to convert primarily cellobiose and cellotriose to glucose. Cellobiohydrolase 1(D) was shown to bind [1-^H]cellooligosaccharides in more than one productive mode. Values of Vjjj33j/Kjjj increase with increasing chain length. Activity on [1-^H]cellooligosaccharides and [^^C-U]cellooligosaccharides indicate that CBH 1(D) repetitively attacks cellooligosaccharides from the reducing end. Subsite maps constructed for CBH 1(D) using the methods of Suganuma et al. (45) and Allen and Thoma (48) show that the active site consists of at least 7 subsites and the catalytic groups are located between subsites 5 and 6. Cellobiohydrolase II also binds [1-^H]cellooligosaccharides in more than one productive mode. Values of V /K ma Y m

PAGE 303

285 increase with increasing chain length. Action patterns resulting from cleavage of [1--^H]cellooligosaccharides and ^C-Ujcellooligosaccharides suggest that CBH II repetitively attacks cellooligosaccharides from the nonreducing end. Subsite maps constructed for CBH II using methods of Suganuma et al. (45) and Allen and Thoma (48) show that the active site consists of at least 7 subsites and that the catalytic groups are located between subsites 3 and 4. The difference in the direction from which cellobiosyl residues are repetitively removed from cellooligosaccharides by CBH 1(D) and CBH II suggests that exo-exo synergism may result from the ability of CBH 1(D) to expose nonreducing chain ends for CBH II by hydrolyzing surrounding reducing ends. Similarly, CBH II may expose reducing chain ends for CBH 1(D) by hydrolyzing surrounding nonreducing chain ends.

PAGE 304

APPENDIX A SUBSITE MAPPING OF ENZYMES A synopsis of equations used to evaluate subsite affinities of polysaccharide depolymerases is presented below. Detailed theoretical treatments of subsite mapping have been reported previously (37,41,43-46,48). Although the subsite mapping procedures of Hiromi et al. (44), Suganuma et al. (45) and Allen and Thoma (48) originate from a common theoretical base, some assumptions, which are described below, distinguish the methods. To facilitate comparisons among the pethods, the nomenclature of Allen (37) will be used below. A list of symbols and notations is presented at the end of this section. The object of subsite mapping is to partition enzymic processes into measurable macroscopic constants, which describe the net association of substrate with enzyme, and microscopic constants, which describe processes associated with one particular binding mode. For example, the positional binding modes of a tr isaccharide with a foursubsite enzyme are shown in Fig. A-1 Each binding mode (1,3-6,3) is characterized by a microscopic association constant and each productive E-S complex (3,3 and 4,3) has, in addition, a microscopic hydrolytic rate coefficient. A combination of microscopic association constants will define

PAGE 305

Figure A-1 Binding modes of cellotriose on a hypothetical four-subsite enzyme Cellotriose may bind to a four-subsite enzyme in several binding modes. The arrow shows the position of catalytic groups. Reducing end glucosyl residues are represented by The broken line shows the position of bond cleavage. The nomenclature of Allen (37) is used to describe the binding modes.

PAGE 306

288 BINDING MODE INDEX ENZYME SUBSTRATE POSITIONAL ISOMERS SUBSITE HYDROUO RATE OOEFFICIENTS PRODUCTS k 2 4.3 >• ^2 3.3

PAGE 307

289 each measurable macroscopic association constant, whereas a specific combination of microscopic rate coefficients and microscopic association constants may define a measurable macroscopic velocity. The Michaelis scheme for binding and hydrolysis of a polymeric substrate may be desribed as (37,41,43-46,48): k k +1,r,n +2,r,n ^ + =ZZZI E „ E + P + P (1) n r,n r,n-m r,ni -1 ,r,n where E is enzyme, is substrate of n monoraeric units in length, Pr,n-m product of chain length n-m arising from the nonreducing end and Pj,^^ is product of chain length m arising from the reducing end of substrate. The microscopic rate coefficients k^.>,^j.^^ and k_-,^j.^^ describe the binding and dissociation of substrate and enzyme in binding mode r. The term r specifies a binding mode in which the real or virtual subsite occupied by the reducing end glycosyl residue. The microscopic hydrolytic rate coefficient is ^+2,r,n* ^ feature which distinguishes the method of Hiromi et al. (44) from that of Allen and Thoma (48) is the assumption by Hirmoi et al. (44) that the hydrolytic rate coefficient is independent of chain length n. When rapid equilibrium between E and A^^ is assumed ^^-1 r,n^''^+2, r,n^ ^ '^^^^ equation of the same form as the Michaelis-Menton type is obtained where the inverse of the macroscopic Michaelis constant (K^^^) is the sum of the

PAGE 308

290 inverse of microscopic Michaelis constants. The inverse of the microscopic Michaelis constant (Kj^^j^) approaches an association constant (Kjlj^^): ^ l+n-1 l+n-1 ^ -ET-^ = ^ K' (2) K,_ ^ ^ h.n ^ ^m,n "h'" h=1 where h is the general binding-mode index, specifying the real or virtual subsite occupied by the reducing end glycosyl residue and 1 is the number of real subsites comprising the binding region of the enzyme. The macroscopic maximum velocity (V^/Eq) is expressed as: c+n-2 c+n-2 2 +2. h.n 2 k ^ K' V„ h=c k' h=c +2,h,n h,n n h,n l+n-1 l+n-1 o l+n-1 h=1 *^h,n h=1 where c is the index position of the catalytic site, specifiying the subsite to the right of the position of bond cleavage. The first order rate constant for enzymic hydrolysis at low concentration of substrate (A„<
PAGE 309

291 productive and nonproductive binding. The macroscopic constants Kjj,^^, and are measurable and are related to microscopic constants by equations (2)-(4). The free energy of binding of a glycosyl residue to a particular subsite is assumed to be constant and not affected by binding at another subsite. Therefore, the macroscopic association constant for each binding mode of an n-mer (K^^^) is related to the sum of the binding energies of occupied subsites by: r ^ AG + 2500 = -RT In K' (5) i=r-n+1 ^ where aq^ is the unitary free energy of binding for subsite i. The 2500 cal/mol arises from the entropy of mixing in water at 40C and is referred to as the cratic free energy contribution to binding (102,103). Application of Subsite Mapping by Hiromi et al. (44) Since a true exoenzyme binds substrate in only one productive mode, equation (4) may be expressed as: n K^^^ ^+2,h,n ^r,n insofar as the inverse of the Michaelis constant approximates the association constant. The ratio of first order rate constants for an n-mer and n+1-mer are expressed in terms of subsite binding energies by using equations (5) and (6):

PAGE 310

292 exp (AG /RT) (7) '^+2,r + 1 ,n+1 n+1 n+1 Ratios of first order rate constants provide an estimate of affinities of subsites c+1 and larger, if hydrolytic rate coefficients are independent of n. Affinities of subsites 3-7 of glucoamylase from R. delemar (44) and A. saito (47) have been evaluated using equation (7). Hiromi et al. (44) showed that the binding energy for maltose calculated using equations (2) and (5) is higher than could be accounted for by subsites 3-7. Therefore, Hiromi et al. (44) assumed that only subsites 1 and 2 contribute significantly to the binding of oligosaccharides and calculated the binding energy for subsite 1 by expressing equation (3) in terms of subsite binding affinities to yield: If only subsites 1 and 2 contribute significantly to binding of oligosaccharides and hydrolytic rate coefficients are independent of n, then a plot of exp ( aG^^., /RT ) versus [EqJ/V^ should be linear. Vertical and horizontal intercepts give AG^ and k^2,r,n' respectively. The affinity at subsite 2 can be evaluated using two methods. The Michaelis constant for each n-mer can be used exp (-AG^_^^/RT) = [ 1 ] exp (-AG^/RT) (8)

PAGE 311

293 to estimate the sum of affinities of occupied subsites. Since the binding energy for all subsites except subsite 2 have been evaluated, they can be used to evaluate the binding energy at subsite 2. Alternatively, first order rate constants can be used with equations (5) and (6) to calculate the affinity at subsite 2. Affinities for subsites 3-6 of the S-giucosidase were calculated using first order rate constants by the method of Hiromi et al. (44). For example, the free energy of binding to subsite 3 was calculated using the first order rate constants of [1-^H]cellobiose and [1 -^Hjcellotriose hydrolysis and equation (7): = exp (AG^ ./RT) n+1 ^n+1 ^" ^0.217 ^ = ""^n+l -1480 cal/mol = AG^^^ However, AG>, could not be evaluated using the method of Hiromi et al. (44), because a plot of exp( AG^^^^ /RT) versus [E^j/V^ was not linear [equation (8)]. Since ag^ and k^2 could not be estimated, AG2 could not be evaluated. Application of Subsite Mapping by Allen and Thoma (48) The method of Allen and Thoma (48) uses four approximately accessible paramters: K^, V, bond cleavage

PAGE 312

294 frequencies and K^^^^ to evaluate subsite affinities. Substrates that cover the entire binding region of the active site can be used to measure the sum of subsite binding affinities. All binding modes in which subsites are completely occupied are assumed to have the same dissociation constant, K^^^^^. Equation (2) can be written in terms of K^^^ by partitioning it into: 1 1-1 1 n+1-1 \n ^int ^="^^ The term /^yi^^ accounts for binding modes in which subsites on the right side of the binding region are vacant; the middle term accounts for binding in which all subsites are occupied and the last term accounts for binding modes when subsites on the left side of the binding region are vacant. Equation (9) is in a linear form and Vk^^^^ is obtained from a plot of (n-1+1) versus VK-, „ for n>l. Information about subsites can also be obtained from the relative rates of formation of each product, i.e., bond cleavage frequencies. Specific productive bonding modes of a subsite are characterized by formation of specific products. The relative rate of formation of a product resulting from binding in a particular mode is proportional to the population of that enzyme-substrate complex multiplied by the corresponding hydrolytic rate coefficient. The relative rate of formation of product from

PAGE 313

295 two adjacent binding modes is ^'^'"^ '-^--'"^'y (10) where [Pr,m^=^LPp^ni]/dt. Since the k^2 ^^"^^ goes to zero in the numerator of equation (1) for nonproductive complexes, bond cleavage frequencies provide information about onlyproductive complexes. Bond cleavage frequencies also provide information about subsite-binding affinities and hydrolytic rate coefficients. Equation (5) may be solved for ^r,n ^^'^ then substituted into equation (10): ""'"TT^^^ = ^^r.1 ^^r-n.l ^ ^^Ini^^f^) (11) ^^r+1,m+1 J +2,r+1,n Thus, a bond cleavage frequency provides information about the difference between two subsite affinities and the ratio of hydrolytic rate coefficients. Affinities of all subsites with the exception of the two adjacent to the catalytic groups can be evaluated using bond cleavage frequencies for a series of oligosaccharides. Since subsites adjacent to catalytic groups are common to all productive complexes, their affinities cannot be evaluated. If hydrolytic rate coefficients (k^2) assumed equal, then bond cleavage frequencies provide a direct estimate of subsite affinities [equation (11)]. The method

PAGE 314

296 of Suganuma et al. (45) assumes that hydrolytic rate coefficients are equal. However, Thoma et al. (43) found that binding energies estimated from equation (11) and used in equations (2), (3) and (5) yielded values of i^j^^^ and Vjj/Eq that did not agree well with values of Kjn^^ and Vj^/E^ calculated from experimental initial velocities. Agreement between predicted and experimentally derived values significantly improved when the approximation was made that the binding of a glycosyl residue decreases the activation energy by a constant amount enhancing the rate of bond cleavage Thus, the hydrolytic rate coefficient be expressed as a function of the number of occupied subsites: r-n + 1 ^2,r,n %xp ^ 'G^,i/f^T) (12) where '^Gg^j^ is the contribution of the i-th real subsite to acceleration of bond cleavage. Equation (11) can be partitioned as: [P ] RTln-, ^ = (AG AG ^)-(AG AG ) rp 1 r + 1 a,r + V ^ ""r-n+l a,r-n + V r+l.m+l-" = 'r.1 ^^r-n.l (13) revealing that apparent subsite affinities must be increased by an average acceleration factor to provide true binding

PAGE 315

297 affinities When an acceleration factor was used by Thoraa et al. (43) values of Kjjj^^ and V^/Eq predicted from subsite energies agreed well with corresponding values estimated using experimental initial velocities. Allen and Thoma (48) have incorporated the basic equations of subsite mapping into a depolymerase computer model which generates a subsite map that gives an optimum fit to experimentally derived paramters: ^m,n' ^n> t)ond cleavage frequencies and Kj^^^^. Bond cleavage frequencies are first used to estimate apparent subsite binding energies and the location of the catalytic groups [equations (5) and (10)]. The next step is to optimize binding energies for the two subsites adjacent to the catalytic groups while binding energy of the other subsites are constrained at values estimated from bond cleavage frequencies. All four parameters are then optimized by varying binding energies for all subsites, while the acceleration factor is not allowed to vary. Finally, the acceleration factor is varied and an optimum subsite map is obtained providing the minimum error between predicted values of K^^^, V^, bond cleavage frequencies and K^^^^ and those derived from experimental initial velocities. Optimized subsite maps for CBH 1(D) and CBH II were generated using the depolymerase computer model, of which a tape and a test data set were kindly provided by J. D. Allen, Biotrack, Inc., Sunnyvale, OA.

PAGE 316

298 Application of Subsite Mapping by Suganuma et al. (45) The subsite mapping procedure for exoenzyraes described by Hiromi et al. (44) was adapted for use with endoenzyraes by Suganuma et al. (45). Since endoenzyraes can bind substrate in more than one productive mode, the values of first order rate constants (kj!^) alone do not provide estimates of subsite affinities [equation (6)]. However, comparison of rate constants for the reactions yielding labeled m-mer and m+1-mer products from reducing end-labeled n-mer and n+1-mer substrates provides a basis for estimation of the affinity of the subsite to which the n+1 glycosyl residue binds. Since Suganuma et al. (45) assume that the hydrolytic rate coefficent is independent of n, equation (10) reduces to: r, m-* r,n K• (14) fPr.1,m.lJ %.1.n.1 Equation (14) used in conjunction with equation (5) yields: [P ] RT In (—. — ^ ) = AG (15) r + l.m+V where AG^^^ is the free energy of binding for subsite n+1. Comparison of first order rate constants ik^) and bond cleavage frequencies for pairs of substrates differing in

PAGE 317

299 length by one glycosyl residue yields affinities of each subsite except those adjacent to the catalytic groups. Subsite affinities of CBH 1(D) and CBH II were evaluated using the method of Suganuma et al. (45). For example, subsites 4 and 7 of CBH 1(D) were evaluated using first order rate constants and bond cleavage frequencies of [1 --^Hjcellotriose and [1 --^Hjcellotetraose in equation (16) derived from equations (4) and (5) (Figure A-2): k' [P ] + 1 = 1" ^ ) (16) ^n+1 "-^r + l.m+V AG^ = RT m ( 0.479x0.78 ^ = cal/mol = RT In (OiOmTT^O^) ^ _5^qq cal/mol Calculated bond cleavage frequencies of bonds 1 and 2 of [1-^H]cellotriose are 0.06 and 0.94, respectively (Fig. 5-14). Bond cleavage frequency of bond 2 of [1-^H]cellotetraose is found to be 0.79 (Fig. 6-14). Values of V„^^/K„ for [1--^H]cellotriose and [1--^H]cellotetraose hydrolysis of CBH 1(D) are 0.00177 and 0.479 umol/min. mg. uM, respectively (Table 6-1 ) Application of Subsite Mapping by Roser and Legler (71) Roeser and Legler (71) calculated the contribution of individual parts of substrates to the free energy of binding of sugar analogs by S-glucosidase from A. wentii using

PAGE 318

Figure A-2 Combination of V x/K values and bond cleavage frequencies for iT-^HJcellotetraose and [1--'H]cellotriose used to calculate subsite affinities Schematic model for the productive ES complexes used to evaluate the affinities of subsites 4 (A) and 7 (B) of CBH 1(D). The arrow shows the position of catalytic groups.

PAGE 319

301 3 4 SUBSITE B ^ O O — jG^ O O 0 12 3 4 5 SUBSITE 6 7

PAGE 320

302 equation (17): K. AAG = -RT In {-77 — ) (17) ^i(ref ) Equation (16) was applied to determine affinities of subsites in the binding region of the 3-glucosidase from T. reesei For example, subsite 3 was calculated as follows: AAG = -RT In (ir^: ^ ^i,3 = -RT In (^'^P-) = -1800 cal/mol DO Affinity at subsite 1 was calculated using the inhibition constant of glucose (700 yM) (8) and equation (5): AG = -RT In (1/K^) 2500 = -RT In (1/7x10"^M) 2500 = -7000 cal/mol Symbols and Notations Used in Subsite Mapping A Substrate c Index of the position of the catalytic site, specifying the subsite to the right of the position of bond cleavage E Enzyme AG Unitary free energy of binding AGg Acceleration factor h General binding-mode index, specifying the real or virtual subsite occupied by the reducing-end glycosyl residue i Subsite index

PAGE 321

303 j Maximum chain length substrate for which experimental data are available K Microscopic dissociation constant K' Microscopic association constant Inhibition constant Kjjj Mlchaells constant K^^^ Microscopic dissociation constant for a binding mode In which the entire binding region Is occupied k The first-order rate constant for enzymlc hydrolysis k^.| Microscopic association rate constant k_>| Microscopic dissociation rate constant k^2 Hydrolytic rate coefficient 1 Number of real subsites comprising the binding region of an enzyme m Chain-length index for product n Chain-length index for substrate P Product R Gas constant r Specific binding-mode index, specifying the real or virtual subsite occupied by the reducing-end glycosyl unit T Absolute temperature t Sugar chain length V Maximum velocity Vq Measured velocity [ ] Concentration Measured or apparent value Time derivative

PAGE 322

APPENDIX B SAMPLE CALCULATIONS FOR BOND CLEAVAGE FREQUENCIES AND INITIAL RATES OF HYDROLYSIS OF [ 1 -^H] CELLOOLIGOSACCHARIDES Bond cleavage frequencies and initial rates for hydrolysis of [1--^H]cellooligosaccharides by exoglucanases were determined as described below for hydrolysis of [1-^H]cellotetraose (15 yM) by CBH 1(D): 1. The radioactivity of products ( [1 -^Hjglucose [1-^H]cellobiose and [1-^H]cellotriose) and substrate ( [1-^H]cellotetraose) in TLC scrapings was measured, at various time intervals. The sum of the radioactivity of products and substrate at each time interval was then calculated (Table B-1). 2. Product ratio for each product was determined at each time interval by dividing the radioactivity of each product by the sum of radioactivity of all products and substrate. Extent of reaction was calculated by dividing the sum of the radioactivity of the products by the sum of the radioactivity of products and substrate (Table B-2). 3. The amount of each product and substrate at each time interval was determined by multiplying the product and substrate ratios by the initial amount

PAGE 323

305 Table B-1 Distribution of [1 -^HjCellooligosaccharides from Hydrolysis of [1 -^HjCellotetraose by CBH 1(D) Cellobiohydrolase 1(D) (2.32x10"^ mg) was incubated at 40C in 100 ui of 5 5M sodium acetate buffer, pH 5.0, containing 15 [1--^H]cellotetraose Time (min) DPM [1-^H]G1 [1-^H]G2 [1-^H]G3 [1-^H]G4 Sum 0 531 798 672 45505 47506 2 958 2055 658 45307 48978 4 1281 3375 896 44551 50103 6 1599 4565 770 42823 49757 8 1 904 5696 712 40556 48868 15 2727 9054 876 36902 49559 30 3810 14057 906 29202 47975 60 5825 21485 1103 21598 50011

PAGE 324

306 Table B-2 Ratios for [1-^H]Cello Hydrolysis by CBH 1(D) „. Product Ratio Time Extent of (mm) [i_3h]gi [1-^H]G2 [^-^H]G3 [1-^H]G4 Reaction 0 0.0112 0.0168 0.0141 0.958 0.0421 2 0.0196 0.0420 0.0134 0.925 0.0750 4 0.0256 0.0674 0.0179 0.889 0.111 6 0.0321 0.0917 0.0155 0.861 0.139 8 0.0390 0.117 0.0146 0.830 0.170 15 0.0550 0.183 0.0177 0.745 0.255 30 0.0794 0.293 0.0189 0.609 0.391 60 0.116 0.430 0.0220 0.432 0.568

PAGE 325

307 of substrate; in this example, 1500 pmol (Table B-3). Bond cleavage frequencies for each product were determined by plotting each product ratio versus extent of reaction (69). The slope of the line is the bond cleavage frequency. Linear regression analysis yielded bond cleavage frequencies for [1-^H]G1, [1-^H]G2, [1-^H]G3 and [1-^H]G4 of 0.1970.002, 0.7890.003, 0.01450.003 and -1.00.0, respectively. The corresponding Rsquared values were found to be 99.9, 100.0, 80.8 and 100.0 percent, respectively. Initial velocities for product formation and substrate degradation were determined by plotting the amount of product and substrate (pmol) versus time (min). Linear regression analysis of early, linear region yielded velocities determined from the slope for [1-^H]G1, [1-^H]G2, [1-^H]G3 and [1-^H]G4 of 5.400.51, 19.00.05, 1.401.12 and -25.80.6 pmol/min, respectively. The corresponding R-squared values were found to be 99.1, 100.0, 61.2 and 99.9 percent, respectively.

PAGE 326

308 Table B-3 Concentration of [1--^H]Cellooligosaccharides from Hydrolysis of [ 1 -%]Cellotetraose by CBH 1(D) Time (min) Pmol [1-^H]G1 [1-%]G2 [1-%]G3 [1-^H]G4 0 16.8 25.2 21 .2 1437 2 29.3 62.9 20.2 1387 4 38.4 101 26.8 1333 6 48.2 138 23.2 1291 8 58.4 175 21 .8 1245 15 82.5 274 26.5 1117 30 119 440 28.3 913 60 175 644 33.1 648

PAGE 327

APPENDIX C KINETIC EQUATIONS 1. Cleland's HYPER program (70) was used to fit initial rates of [1-^H]cellooligosaccharide hydrolysis to the hyperbola described by: V A max K + A m 2. Cleland's TWO/ONE program was used to fit initial velocities determined for CBH 1(D) hydrolysis of [1-^H]cellotriose to:

PAGE 328

APPENDIX D SUBSITE AFFINITIES FOR 8-GLUCOSIDASE Subslte Binding Energy (cal/mol) Determined from L 1 --^H J ce 11 00 ligo saccharides Determined from values of ce 1 1 00 I igo saccharides 1 -6880 30 -702050 2 25050 3 -1480 70 -180030 4 -630 90 -14040 5 90 70 -17060 6 20120 -7070

PAGE 329

APPENDIX E KINETIC CONSTANTS FOR HYDROLYSIS OF 8-LINKED DISACCHARIDES OF GLUCOSE BY 8-GLUCOSIDASE Activity of 8-glucosidase was assayed at various concentrations of disaccharide in 50 mM sodium acetate buffer, pH 5.0, containing 3 mM sodium azide as described in Experimental Procedures. Kinetic constants were determined from initial rates of hydrolysis using Cleland's HYPER program (70). Disaccharide Linkage Kjjj (yM) ( pmole/min mg) ^max^^ V, max Sophorose 8-1 ,2 107090 27 1 0.0250.001 Laminaribiose 6-1,3 46020 22.60.4 0.0480.002 Gentiobiose 8-1 ,6 96060 8.30.2 0.00870.0004

PAGE 330

APPENDIX F KINETIC CONSTANTS FOR HYDROLYSIS OF [1-^H]CELL00LIG0SACCHARIDES BY 8 -GLUCOSIDASE CBH 1(D), AND CBH II To facilitate comparison of kinetic constants for hydrolysis of [ 1 -^Hjcellooligosaccharides by the exoglucanases kinetic constants found in Tables 5-1, 6-1, and 7-1 are expressed below as per mole of enzyme instead of per mg of protein. The molecular weights of 8-glucosidase CBH 1(D), and CBH II determined using sedimentation equilibrium analysis are reported to be 74600 (8), 53200 (6), and 54700 (6), respectively. [1-3H]Cellooligo^ ^ ""^^^ ^ saccharide (M)x10 (s"'') (M"'' s''' )x10" 8-Glucosidase G2 G3 G4 G5 G6 CBH 1(D) G3 G4 G5 G6 CBH II S3 G4 G5 G6 880 40 22.0 0.2 0.02500.001 74 11 20.1 1.1 0.272 0.04 35 5 26.1 1 .2 0.746 0.11 35 4 22.6 1 .0 0.646 0.079 37 10 22.4 2.5 0.605 0.18 9 4.5 3400 600 3.1 0.2 0.52 0.04 1.1 0.2 18 3 2.6 0.3 0.98 0.10 0.67 0.06 0.0140.002 0.23 0.02 1.34 0.04 1.12 0.03 2.4 0.2 0.0240.002 7.0 0.3 2.01 0.08 4.0 0.2 (1.56 0.8)x10"| (6.76 1.3)x10-5 0.432 0.03 2.15 0.18 2.18 0.44 0.00130.0002 2.69 0.3 2.05 0.2 5.97 0.6

PAGE 331

REFERENCES 1. Ghose, T. K. (1977) Cellulase biosynthesis and hydrolysis of cellulosic substances. Adv. Biochem. Eng. 6, 39-76. 2. Humphrey, A. E. (1979) The hydrolysis of cellulosic materials to useful products. In: Hydrolysis of Cellulose; Mechanisms of Enzymatic and Acid Catalysis (Brown, R. D. Jr., and Jurasek, L. Eds.) pp. 25-53, American Chemical Society, Washington. 3. Ryu, D. D. Y., and Mandels, M. (1980) Cellulases: Biosynthesis and applications. Enzyme Microb. Technol. 2, 91-102. 4. Mandels, M. (1975) Microbial sources of cellulase. Biotechnol. Bioeng. Symp. 6, 81-105. 5. Gritzali, M., and Brown, R. D., Jr. (1979) The cellulase system of Trichoderma Relationships between purified extracellular enzymes from induced or cellulose-grown cells. In: Hydrolysis of Cellulose: Mechanisms of Enzymatic and Acid Catalysis (Brown, R. D., Jr., and Jurasek, L. Eds.) pp. 237-250, American Chemical Society, Washington. 6. Gritzali, M. (1979) Purification and characterization of an endo-1 ,4-S-_D-glucanase and two exo-1,4-8-^glucanases from the cellulase system of Trichoderma reesei QM 9414. Ph.D. Dissertation, VPI and SU, Blacksburg, Virginia. 7. Fagerstam, L. G. and Pettersson, L. G. (1980) The 1,4-S-glucan cellobiohydrolases of Trichoderma reesei QM 9414. FEBS Lett. 119, 97-100. 8. Chirico, W. J. (1980) Purification and characterization of an extracellular S-glucosidase from Trichoderma reesei QM 9414. Master's Thesis, VPI and SU, Blacksburg, Virginia. 9. Blackwell, J. (1982) The marcomolecular organization of cellulose and chitin. In: Cellulose and Other Natural Polymer Systems (Brown, R. M., Jr., Ed.) pp. 403-428, Plenum Press, New York.

PAGE 332

314 10. Ward, K. Jr., and Seib, P. A. (1970) Cellulose, lichenan, and chitin. In: The Carbohydrates (Pigman, W., and Horton, D., Eds.) Vol. IIA, pp. 413-298, Academic Press, New York. 11. Atalla, R. H., and VanderHart, D. L. (1984) Native cellulose: A composite of two distinct crystalline forms. Science 223, 283-285. 12. Gum, E. K., Jr., and Brown, R. D., Jr. (1976) Structural characterization of a glycoprotein cellulase, 1 4-3-_D-glucan cellobiohydrolase C from Trlchoderma viride Biochim. Biophys. Acta 446, 37138^'^ 13. Gum, E. K., Jr., and Brown, R. D., Jr. (1977) Comparison of four purified extracellular 1,4-6-^glucan cellobiohydrolase enzymes from Trichoderma viride Biochim. Biophys. Acta 492, 225-231. 14. Shoemaker, S. P., and Brown, R. D. Jr. (1978) Enzymic activities of endo-1 4-3-^-glucanases purified from Trichoderma viride Biochim. Biophys. Acta 523, 133146. 15. Shoemaker, S. P., and Brown, R. D. Jr. (1978) Characterization of endo-1 4-3 -D_-glucanases purified from Trichoderma viride Biochim. Biophys. Acta 523, 147-lFr: 16. Berghem, L. E. R. and Pettersson, L. G. (1974) The mechanism of enzymatic cellulose degradation. Isolation and some properties of a 8 -glucosidase from Trichoderma viride Eur. J. Biochem. 46, 295-305. 17. Hakansson, U., Fagerstam, L. E., Pettersson, L. G., and Anderson, L. (1979). A 1,4-6-glucan glucanohydrolase from the cellulolytic fungus Trichoderma viride QM 9414. Biochem. J. 179, 141-149. 18. Wood, T. M., and McCrae, S. I. (1979) Synergism between enzymes involved in the solubilization of native cellulose. In: Hydrolysis of Cellulose: Mechanisms of Enzymic and Acid Catalysis (Brown, R. D., Jr., and Jurasek, L., Eds.) pp. 181-209, American Chemical Society, Washington. 19. Wood, T. M., and McCrae, S. I. (1982) Purification and some properties of the extracellular 8 -D-glucosidase of the cellulolytic fungus Trichoderma konTngii J. Gen. Micro. 128, 2973-2982.

PAGE 333

315 20. Kanda, T., Nakakubo, S. Wakabayashi, K., and Nisizawa, K. (1979) The mode of enzymic degradation of cellulose based on properties of cellulase components. In: Hydrolysis of Cellulose: Mechanisms of Enzymic and Acid Catalysis (BrowrTj RT DT] Jr. and Jurasek, L., Eds. ) pp. 211-236, American Chemical Society, Washington. 21. Emert, G. H., Gum, E. K., Jr., Lang, J. A., Liu, T. H., and Brown, R. D., Jr. (1974) Cellulases. In: Food Related Enzymes (Whi taker, J. R., Ed.) pp. 79-100, American Chemical Society, Washington. 22. Reese, E. T., Maguire, A. H., and Parrish, F. W. (1968) Glucosidases and exo-glucanases. Can. J. Biochem. 46, 25-34. 23. Ladisch, M. R. Gong, C. S., and Tsao, G. T. (1980) Cellobiose hydrolysis of endoglucanase (glucan glucanohydrolase ) from Trichoderma reesei ; Kinetics and mechanism. Biotechnol. Bioeng. 22, 1107-1126. 24. duMee, C. (1984) Structural characterization of carbohydrate attached to the glycoprotein cellulase enzymes of Trichoderma reesei QM 9414. Ph.D. Dissertation, Univ. of Florida, Gainesville, Florida. 25. Bhikhabhai, R. and Pettersson, L. G. (1984) The cellulolytic enzymes of Trichoderma reesei as a system of homologous proteins. FEBS Lett. 167. 301-308. 26. Mandels, M. Parrish, F. W. and Reese, E. T. (1962) Sophorose as an inducer of cellulase in Trichoderma viride J. Bacterid. 83, 406-408. 27. Fagerstam, L. G., Pettersson, L. G. and Engstrr>m, J. A. (1984) The primary structure of a 1,4-6 -glucan cellobiohydrolase from the fungus Trichoderma reesei QM 9414. FEBS Lett. 167, 309-315. 28. Shoemaker, S. Schweickart, V., Ladner, M. Gelfand, D., Kwok, S., Myambo, K., and Innish, M. (1983) Molecular cloning of exo-cellobiohydrolase I derived from Trichoderma reesei strain L27. Bio/Technoloey 1, 691-69^: ^ 29. Hsu, T., Gong, C. and Tsao, G. T. (1980) Kinetic studies of cellodextrins hydrolyses by exocellulase from Trichoderma reesei Biotech nol Bioeng. 22, 23052320. ^

PAGE 334

316 30. van Tilbeurgh, H. Claeyssens, M. and de Bruyne, C.K. (1982) The use of 4-methylumbellif eryl and other chromophoric glycosides in the study of cellulolytic enzymes. FEBS Lett. 149, 152-156. 31. Brown, R. D. Jr., and Greenberg, B. (1980) Unpublished observations. 32. Nisizawa, K. and Hashimoto, Y. (1970) Glycoside hydrolases and glycosyl transferases. In: The Carbohydrates (Pigman, W. and Horton, D. Eds.) Vol. IIA, pp. 241-300, Academic Press, New York. 33. Sternberg, D. (1976) S -Glucosidase of Trichoderma : Its biosynthesis and role in saccharif ication of cellulose. Appl. Environ. Microbiol. 31, 648-654. 34. Gong, C. S., Ladisch, R., and Tsao, G. T. (1977) Cellobiase from Trichoderma viride ; Purification, properties, kinetics and mechanism. Biotechnol Bioeng. 19, 959-981. 35. Emert, G. H. (1974) Purification and characterization of cellobiase from Trichoderma viride Ph.D. Dissertation, VPI and SU, Blacksburg, Virginia. 36. Mandels, M. (1982) Cellulases. In: Annual Reports on Fermentation Processes (Tsao, G. T. Ed.) Vol. 5, pp. 35-78, Academic Press, New York. 37. Allen, J. D. (1980) Subsite mapping of enzymes: Application to polysaccharide depolymerases. In: Methods in Enzymology (Purich, D. L., Ed.) Vol. 64, pp. 248-277, Academic Press, New York. 38. Schechter, I., and Berger, A. (1967) On the size of the active site in proteases. I. Papain. Biochem. Biophys. Res. Commun. 27, 157-162. 39. Walz, F. G., Jr., and Terenna, B. (1976) Subsite interactions of Ribonuclease T-i: Binding studies of dimeric substrate analogues. Biochemistry 15, 28372842. 40. Chipman, D. M. Grisaro, V., and Sharon, N. (1967) The binding of oligosaccharides containing N-acetylglucosamine and N-acetylmuramic acid to lysozyme. J. Biol. Chem. 242, 4388-4394. 41. Hiromi, K., Ohnishi, M. and Tanaka, A. (1983) Subsite structure and ligand binding mechanism of glucoamylase. Molecular and Cellu lar Biochemistrv 51, 79-95. —

PAGE 335

317 42. Biely, P., Kratkj^, Z., and Vrsanska, M. (1981) Substrate-binding site of endo-1 4-6 -xylanase of the yeast Cryptococcus albidus Eur. J. Biochem. 119, 559564. 43. Thoma, J. A., Rao, G. V. K., Brothers, C. and Spradlin, J. (1971) Subsite mapping of enzymes. J. Biol. Chem. 246, 5621-5635. 44. Hiromi, K., Nitta, Y., Numata, C, and Ono, S. (1973) Subsite affinities of glucoamylase : Examination of the validity of the subsite theory. Biochem. Biophys. Acta 302, 362-375. 45. Suganuma, T. Matsuno, R. Ohnishi, M. and Hiromi, K. (1978) A study of the mechanism of action of Takaamylase A on linear oligosaccharides by product analysis and computer simulation. J. Biochem. 84, 293316. 46. Hiromi, K. (1970) Interpretation of dependency of rate parameters on the degree of polymerization of substrate in enzyme-catalyzed reactions. Evaluation of subsite affinities of exo-enzyme. Biochem. Biophys. Res. Commun. 40, 1-6. 47. Koyama, T., Inokuchi, N., Kikuchi, Y., Shimada, H., Iwama, M., Takahashi, T., and Irie, M. (1984) Subsite affinity of a glucoamylase from Aspergillus saitoi Chem. Pharm. Bull. 32, 757-761. 48. Allen, J. D., and Thoma, J. A. (1976) Subsite mapping of enzymes: Depolymerase computer modelling. Biochem. J. 159, 105-120. 49. Allen, J. D., and Thoma, J. A. (1976) Subsite mapping of enzymes. Application of the depolymerase computer model to two a -amy lases Biochem. J. 159. 121-132. 50. Dubois, M., Gilles, K. A., Hamilton, S. K., Rebers, P. A. and Smith F. (1956) Colorimetric method for determination of sugars and related substances. Anal Chem. 28, 350-356. 51. Nelson, N. J. (1944) A photometric adaptation of the Somogyi method for determination of glucose. J. Biol. Chem. 153, 375-380. 52. Somogyi, M. J. (1952) Notes on sugar determination. J. Biol. Chem. 195, 19-23.

PAGE 336

318 53. Lowry, 0. H. Rosebrough, N. J., Furr, A. L., and Randall, R. J. (1951) Protein measurement with the Folin-phenol reagent. J. Biol. Chem. 193, 265-273. 54. Bailey, J. L. (1967) Techniques in Protein Chemistry 2nd Ed., pp. 345-346, Elsevier, Amsterdam. 55. Gum, E. K., Jr., and Brown, R. D., Jr. (1977) Two alternative HPLC separation methods for reduced and normal cellooligosaccharides. Anal. Biochem. 82, 372375. 56. Miller, G. L., Dean, J., and Blum, R. (I960) A study of methods for preparing oligosaccharides from cellulose. Arch. Biochem. Biophys. 91, 21-26, 57. Diehl, H. W. Polorny, M., Zissis, E., Ness, R. K., and Fletcher, H. G., Jr. (1974) Crystalline cellobiono1,5-lactone: Its preparation and growth-inhibitory activity in the Avena coleoptile section test. Carbohyd. Res. 38, 364-368. 58. Hudson, C. S., and Isbell, H. S. (1929) Relations between rotary power and structure in the sugar group. XIX. Improvements in the preparation of aldonic acids. J. Amer. Chem. Soc. 51. 2225-2229. 59. Conchie, J., Hay, A. J., Strachan, I., and Levvy, G. A. (1967) Inhibition of glycosidases by aldonolactones of corresponding configuration. Biochem. J. 102. 929-941. 60. Isbell, H. S., and Frush, H. L. (1963) Lactonization of aldonic acids. In: Methods in Carbohydrate Chemistry (Whistler, R. L. and Wolfrom, M. L. Eds.) Vol. II, pp. 16-18, Academic Press, New York. 61. Frush, H. L., and Isbell, H. S. (1956) Sodium bicarbonate reduction of aldonic lactones to glycitols. J. Amer. Chem. Soc. 78, 2844-2846. 62. Wolfrom, M. L., and Wood, H. B. (1951) Sodium borohydride as a reducing agent for sugar lactones. J. Amer. Chem. Soc. 73, 2933-2934. — 63. Saif-ur-Rahman, S. Krishnamurti C. R. and Kitts, W. D. (1968) Separation of cello-oligosaccharides by thin-layer chromatography. J. Chromatog. 38, 400-402. 64. Krebs, K. G., Heusser, D., and Wimmer, H. (I969) Spray reagents. In: Thin-Layer Chromatography (Stahl, E. Ed.) pp. 854-909, Springer-Verlag, New York.

PAGE 337

319 65. Ghebregzabher M. Rufini, S., Monaldo, B. and Lato, M. (1976) Thin-layer chromatography of carbohydrates. J. Chromatog. 127, 133-162. 66. Jermyn, M, A. (1955) Fungal cellulases. Australian J. Biol. Sci. 8, 541-602. 67. STATZYME. Worthington Diagnostics (1976) Worthington Statzyme Glucose Freedhold, New Jersey. 68. Rosenthal, A. L., and Saifer, A. (1973) Continuous uv monitoring of fluorogenic substrates I. Kinetic analysis of N-acetyl-3 -^-hexosaminidases. Anal Biochem. 55, 85-92. 69. Allen, J. D. (1975) Bond-cleavage frequency analysis. Carbohyd. Res. 39, 303-315. 70. Cleland, W. (1979) Statistical analysis of enzyme kinetic data. In: Methods in Enzymology (Purich, D. L., Ed.) Vol. 63, pp. 103-138, Academic Press, New York. 71. Roeser, K-R., and Legler, G. (1981) Role of sugar hydroxyl groups in glycoside hydrolysis. Biochem. Biophys. Acta 657. 321-333. 72. Robyt, J. F., and French, D. (1970) The action pattern of porcine pancreatic "-amylase in relationship to the substrate binding site of the enzyme. J. Biol. Chem. 245, 3917-3927. 73. Whitaker, D. R. and Merler, E. (1956) Cleavage of cellotriose by Myrothecium cellulase. Can. J. Biochem. Physiol. 34, 83-89. 74. Vaheri, M. Leisola, M. and Kauppinen, V. (1979) Translycosylation products of cellulase system of Trichoderma reesei Biotechnol. Lett. 1. 41-46. 75. Evans, E. A., Sheppard, H. C, Turner, J. C, and Warrell, D. C. (1974) A new approach to specific labelling of organic compounds with tritium: Catalyzed exchange in solution with tritium gas. J. Labelled Compounds 10, 569-587. 76. Friedberg, F. and Kaplan, L. (1957) The absence of an Amadori rearrangement in glucosazone formation. J. Am. Chem. Soc. 79. 2600-2601. 77. Isbell, H. S., Frush, H. L., Holt, N. B. and Moyer, J. D. (1960) Tritium-labeled compounds III. Aldoses-1t^. J. Research NSB 64A, 177-180.

PAGE 338

320 78. Green, J. W. (1980) Reduction of carbohydrates. In: The Carbohydrates (Pigman, W. and Horton, D., Eds.) Vol. IB, pp. 989-1011, Academic Press, New York. 79. Takeo, K., Okushio, K., Fukuyama, K., and Kuge, T. (1983) Synthesis of cellobiose, cellotriose, cellotetraose and lactose. Carbohydr. Res. 121, 163173. 80. John, M., Trenel, G., and Dellweg, H. (1969) Quantitative chromatography of homologous glucose oligomers and other saccharides using polyacrylamide gel. J. Chromatog. 42, 476-484. 81. Tipson, R. S. and Parker, F. S. (1980) Infrared spectroscopy. In: The Carbohydrates (Pigman, W. and Horton, D., Eds.) Vol. IB, pp. 1394-1436, Academic Press, New York. 82. Biely, P., Vrsanska, M. and Kratky, Z. (1981) Mechanisms of substrate digestion by endo-1,4-Sxylanase of Cryptococcus albidus Eur. J. Biochem. 119, 565-571. 83. Ladisch, M. R. and Tsao, G. T. (1978) Theory and practice of rapid liquid chromatography at moderate pressures using water as eluent. J. Chromatogr. 166, 85-100. 84. Vratny, P., Coupek, J., and Vozka, S. (1983) Accelerated reversed-phase chromatography of carbohydrate oligomers. J. Chromatogr. 254, 143-155. 85. Chen, C, and McGinnis, G. (1983) High-performance liquid chromatography of sugar oximes. Carbohydr. Res. 122, 322-326. 86. Brown, W. and Andersson, 0. (1971) Preparation and xylodextrins and their separation by gel chromatography. J. Chromatogr. 57, 255-263. 87. Honda, S. (1984) High-performance liquid chromatography of monoand oligosaccharides. Anal Biochem. 140, 1-47. 88. Scott, R. W. (1970) Quantitative recovery of sugars from silica gel thin layers. J. Chromatogr. 49, 473481 89. Umezurike, G. M. (1972) The purification and properties of extracellular S -glucosidase from Botryodiplodia theobromae Pat. Biochem Biophys. Acta 227, 419-428.

PAGE 339

321 90. Shewale, J. G. (1982) 6 -Glucosidase : Its role in cellulase synthesis and hydrolysis of cellulose. Int. J. Biochem. 14, 435-443. 91. Cole, F. E., and King, K. W. (1964) Site of hydrolysis of cellulodextrins and reduced cellulodextrins by purified cellulase components. Biochim. Biophys. Acta 81, 122-129. 92. Hirayama, T., Horie, S., Nagayama, H. and Matsuda, K. (1978) Studies on cellulases of phytopathogenic fungus, Pyricularia oryzae Cavara. J. Biochem. 84, 2737. 93. Tsujisaka, Y., Fukumota, J., and Yamamoto, T. (1958) Specificity of crystalline saccharogenic amylase of moulds. Nature 181, 770-771. 94. King, K. W., and Smibert, R. M. (1963) Distinctive properties of 3 -glucosidases and related enzymes derived from a commercial Aspergillus niger cellulase. Appl. Microbiol. 11, 315-319. 95. Shewale, J. G., and Sadana, J. (1981) Purification, characterization and properties of 6 -glucosidase enzymes from Sclerotium rolfsii Arch. Biochem. Biophys. 207, 185-196. 96. Jencks, W. P. (1969) Catalysis in Chemistry and Enzymology McGraw-Hill, New York, pp. 323-350. 97. Hiromi, K., Tanaka, A., and Ohnishi, M. (1982) Fluorometric studies on the binding of gluconolactone, glucose, and glucosides to the subsites of glucoamylase. Biochemistry 21 102-107. 98. Gum, E. (1978) Unpublished observations. 99. Allen, J. D., and Thoma, J. A. (1978) Model for carbohydrase action. Aspergillus oryzae g -amylase degradation of maltotriose. Biochemist ry 17, 23452350. 100. Wood, T. M. (1981) Enzyme interactions involved in fungal degradation of cellulosic materials. Ekman-Days Symp. 3, 31-38. 101. Eerghem, L. E. R. Pettersson, L. G., and AxioFrederiksson, U. (1975) The mechanism of enzymatic cellulose degradation. Eur. J. Biochem. 53, 55-62.

PAGE 340

322 102. Gurney, R. W. (1953) Ionic Processes in Solution McGraw-Hill, New York, pp 80-92. 103. Kauzman, W. (1959) Some factors in the interpretation of protein denaturation. In; Advances in Protein Chemistry (Antinsen, C. B. Jr., Bailey, K., Anson, M. L. and Edsall, J. T., Eds.) Vol. XIV, pp. 1-64, Academic Press, New York.

PAGE 341

BIOGRAPHICAL SKETCH William J. Chirico was born on April 2, 1955, in Brooklyn, New York. After completing his education at Xaverian High School in Brooklyn, he attended Bucknell University in Lewisburg, PA, where he received a B.S. degree in biology. In September of 1978, he began his graduate studies at Virginia Tech, in the Department of Biochemistry and Nutrition, working with Dr. Ross D. Brown, Jr. He continued his graduate studies with Dr. Ross D. Brown, Jr., at the University of Florida, in the Department of Biochemistry and Molecular Biology. He has accepted a postdoctoral position in Dr. Giiinter Blobel's laboratory at The Rockefeller University.

PAGE 342

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. 5£^ Ross D. Brown, Jr Ch/irman Associate Professor oJ Biochemistry and Molecular Biology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Charles M. 'Allen, Jr.// Professor of BiochemiBtry and Molecular Biology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Kiraon J. Angel id^j Assistant Professor of Biochemistry and Molecular Biology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Assistant Professor of Biochemistry and Molecular Biology

PAGE 343

I certify that I have read this study and that in ray opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. R. Michael Roberts Professor of Biochemistry and Molecular Biology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. j^sfrr^^ Associate Professor of Food Science and Human Nutrition This dissertation was submitted to the Graduate Faculty of the College of Medicine and to the Graduate School, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December 1984 0^anVco liege of Med^icine' Dean for Graduate Studies and Research