Citation
Genetic and biochemical characterization of the β-D-xylosidase gene from Butyrivibrio fibrisolvens

Material Information

Title:
Genetic and biochemical characterization of the β-D-xylosidase gene from Butyrivibrio fibrisolvens
Creator:
Utt, Eric Andrew, 1961-
Publication Date:
Language:
English
Physical Description:
x, 151 leaves : ill., col. photos ; 29 cm.

Subjects

Subjects / Keywords:
Amino acids ( jstor )
Bacillus ( jstor )
DNA ( jstor )
Enzyme substrates ( jstor )
Enzymes ( jstor )
Gels ( jstor )
Mutant proteins ( jstor )
Open reading frames ( jstor )
Plasmids ( jstor )
Xylans ( jstor )
Dissertations, Academic -- Microbiology and Cell Science -- UF
Microbiology and Cell Science thesis Ph. D
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1991.
Bibliography:
Includes bibliographical references (leaves 142-150).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Eric Andrew Utt.

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:
027763195 ( ALEPH )
26483359 ( OCLC )

Downloads

This item has the following downloads:


Full Text












GENETIC AND BIOCHEMICAL CHARACTERIZATION OF
THE B-D-XYLOSIDASE GENE FROM BUTYRIVIBRIO FIBRISOLVENS














By

ERIC ANDREW UTT




















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 1991















ACKNOWLEDGEMENTS

I owe my development as a scientist as well as the successful completion of this dissertation to my major professor, Dr. Neal Ingram. I will be forever indebted to him for his uninhibited sharing of knowledge and expertise. I wish also to express my gratitude to the members of my graduate committee, Dr. Allen, Dr. Aldrich, Dr. Shanmugam, and Dr. Gander. Their contributions to my research and in the preparation and review of this manuscript are greatly appreciated. I must also thank my friends and comrades Jeff Mejia and David Beall for the friendship and helpful suggestions during the course of my doctoral work. To all the former postdocs, Dr. Christina Eddy, Dr. Terryl Conway, and Dr. Guy Sewell, from whom I learned much, I wish to express my thanks. Thanks are due to my parents for their love and support during my graduate education. Similarly, I wish to thank my wife's parents for their love and support. And lastly, I would like to thank my loving wife, Lisa, and my beautiful girls, Tina and Hannah, for making my life special. It is to them that I dedicate this dissertation.







ii
















TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS ...... ........................... ii

LIST OF TABLES....................................*** v

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

ABSTRACT ....................... ... .............. ix

CHAPTERS

I GENERAL INTRODUCTION ........................ 1

II REVIEW OF THE LITERATURE..................... 5

III CLONING, SEQUENCING, AND SEQUENCE ANALYSIS OF THE XYLOSIDASE GENE FROM
BUTYRIVIBRIO FIBROSOLVENS ................... 16

Introduction............................... 16
Materials and Methods...................... 17
Results and Discussion..................... 20
Conclusions................................ 50

IV MUTATIONAL ANALYSIS OF THE xvlB GENE......... 54

Introduction............................... 54
Materials and Methods ....................... 56
Results and Discussion ..................... 63
Conclusions................................ 84

V PARTIAL PURIFICATION AND CHARACTERIZATION
OF THE WILD TYPE GENE PRODUCT................. 86

Introduction............................... 86
Materials and Methods........................88
Results and Discussion..................... 91
Conclusions................................ 115






iii










CHAPTERS Page

VI PARTIAL PURIFICATION AND CHARACTERIZATION OF
THE L178F MUTATION.......................... 120

Introduction............................... 120
Materials and Methods ...................... 121
Results and Discussion .................... 121
Conclusions................................ 128

VII SUMMARY AND GENERAL CONCLUSIONS ............. 135


LITERATURE CITED .................................. 142

BIOGRAPHICAL SKETCH ............................... 151









































iv















LIST OF TABLES

Table Pace

1. Comparison of codon usage frequency for the three B. fibrisolvens ORF's .................... 36

2. Expression of enzyme activities in recombinant E. coli harboring xl.................. 40

3. Hydrolysis of different nitrophenylsubstituted glycosides by the xvlB gene
product ........................................ 42

4. Comparison of the translated amino acid
sequences of the three B. fibrisolvens ORF's
in pLOIl001 with those of selected
proteins ........................................... 46

5. Amino acid sequence alignment of conserved regions ............................................ 48

6. Localization of point mutations by
restriction fragment replacement analysis.......... 67

7. Enzymatic activities of recombinants
harboring point mutations in xylB
relative to the wild type protein.................. 82

8. Xylosidase activity of in vitro mutations with
varying substrate concentrations ................... 83

9. Purification scheme of the xvlB-encoded protein
from E. coli DH5a.................................. 97













v
















LIST OF FIGURES

Figure Page

1. Restriction maps of pUC18 derived plasmids that express B-D-xylosidase activity in
E. coli DH5a ................................... 22

2. Subclone analysis of pLOI1005 to localize the xvlB coding region......................... 24

3. Southern hybridizations of chromosomal DNA from B. fibrisolvens and E. coli ........... 27

4. Outline of sequencing strategy of pLOI1001 and subclone analysis of xvlB................... 29

5. The complete nucleotide sequence and translated amino acid sequence of the
4.2 kb insert from pLOI1001..................... 31

6. SDS-PAGE analysis of cytoplasmic extracts from recombinant E. coli DH5a harboring
selected constructs............................ 45

7. Assignment of domains to the xvlB gene.......... 59

8. Subcloning strategy used to localize in vitro mutations to one of five domains in the
xvlB gene ...................................... 61

9. Deletion analysis of the xvlB gene.............. 65

10. Localization and identification of point
mutations in xvlB by DNA sequencing ............ 69

11. SDS-PAGE analysis of wild type and mutant
proteins ............ ........................... 73

12. Native-PAGE comparison of W158UGA and L178F
mutations with the wild type stained with
Coomassie blue.................................. 75

13. Western hybridization of native-PAGE of wild
type and mutant proteins ....................... 77


vi









Figure Page

14. Substrate binding native gel western
hybridization assays of wild type and mutant
proteins ......................................... 80

15. Elution profile of the xvlB gene product
during preparative electrophoresis on the
BioRad-Prep Cell system ........................ 93

16. Elution profile of the xvlB gene product
during hydrophobic interaction chromatography.. 95

17. SDS-PAGE analysis of pooled xylosidasecontaining fractions from preparative
electrophoresis on 8% native-PAGE............... 99

18 Native-PAGE analysis of the B-D-xylosidase..... 101

19. Thermal inactivation profile of xylosidase and
arabinofuranosidase activities................. 103

20. Temperature optimum profile for xylosidase and
arabinofuranosidase activities.................. 105

21. pH activity profiles for xylosidase and
arabinofuranosidase activities................. 107

22. Double recipricol plots of xylosidase and
arabinofuranosidase activities of the native
protein ....................................... 110

23. Competitive inhibition of xylosidase activity
by 4-MU-a-L-arabinofuranoside................. 112

24. Competitive inhibition of arabinofuranosidase
activity by B-methyl-D-xyloside............... 114

25. Elution profile of the xvlB-encoded protein
harboring the L178F mutation during
preparative electrophoresis................... 123

26. SDS-PAGE analysis of partially purified
L178F mutant protein ......................... 125








vii









Figure Page

27. Thermal inactivation profiles of xylosidase
and arabinofuranosidase activities of the
L178F mutant protein.......................... 127

28. pH activity profiles for xylosidase and
arabinofuranosidase activities of the
L178F mutant protein ......................... 130

29. Double recipricol plots of xylosidase and
arabinofuranosidase activities for the
L178F mutant protein ......................... 132





































viii















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

GENETIC AND BIOCHEMICAL CHARACTERIZATION OF B-D-XYLOSIDASE
FROM BUTYRIVIBRIO FIBRISOLVENS By

Eric Andrew Utt

December 1991

Chairman: Lonnie O. Ingram
Major Department: Microbiology and Cell Science


The gene for B-D-xylosidase from the rumen bacterium Butyrivibrio fibrisolvens encodes a protein that exhibits hydrolytic activity against 8-D-xylopyranosides and a-Larabinofuranosides. This gene, xvlB, was cloned into E. coli as a 4.2 kilobase pairs (kbp) insert in pUC18 and sequenced in both directions. The xvlB gene is present as a single copy on the B. fibrisolvens chromosome and consists of a 1,551 base pair (bp) open reading frame (ORF) which encodes a protein of 517 amino acids. Insertion of a 10 bp linker into the coding region resulted in a frameshift that abolished both activities. Deletions from the 3' end and the 5' end of xvlB also resulted in inactive proteins. SDSPAGE analysis of cytoplasmic extracts from recombinant E. coli clones harboring xvlB confirmed the presence of a new protein with an apparent molecular weight of 60,000.

ix









Although the xylB gene did not exhibit a high degree of amino acid identity with other xylan-degrading enzymes or glycohydrolases, a conserved sequence was identified with significant identity to the active site region of hen egg white lysozyme and Aspergillus niger glucoamylase. No predictable stem loop structures or sequences resembling terminators were found on the xvlB gene fragment and this gene appears to be part of an operon. In vitro analysis of xylB mutants demonstrated structural and functional relationships between the two enzyme activities. All point mutations investigated in xvlB resulted in the reduction or loss of both enzymatic activities. Most of these mutations were clustered in a region near the proposed active site. The point mutations decreased the apparent affinity of the enzyme for xylan. The partially purified xvlB-encoded protein exhibited thermal inactivation kinetics and temperature optima that were essentially the same for both enzymatic activities. The pH optimum for both activities was 6.0. However, the arabinofuranosidase activity exhibited a broader pH range, retaining 90% of maximal activity up to pH 9.0. The apparent Km for p-nitrophenyl-8D-xylopyranoside and p-nitrophenyl-a-L-arabinofuranoside were 3.7 mM and 1.8 mM respectively. Substrate competition experiments corroborated the genetic evidence and demonstrated that the same active center was responsible for both enzymatic activities of the xvlB-encoded protein.


x















CHAPTER I

GENERAL INTRODUCTION


Plant cell walls represent the largest reserve of fixed carbon on earth. Plant cell walls are composed primarily of cellulose, hemicellulose, and lignin (Weinstein and Albersheim, 1979). Cellulose is the most abundant carbohydrate found in plant biomass (Coughlan, 1985) while hemicellulose is a major plant structural polymer that ranks second only to cellulose in natural abundance (Dekker and Richards, 1976). The amount of hemicellulose in dry wood is between 20% and 30% (Eriksson et al. 1990). The composition of hemicellulose varies between softwoods and hardwoods. The major hemicellulose in softwoods is galactoglucomannan (Eriksson et al. 1990). This polymer has a backbone composed of a linear chain of 1,4-linked B-D-glucopyranose and B-D-mannopyranose units. The mannose and glucose moieties of the backbone may be substituted with acetyl groups at the C-i and C-2 positions.

Glucouronoxylan (O-acetyl-4-O-methyl-glucurono-B-Dxylan) is the major hemicellulosic component of hardwoods and agricultural residues (Eriksson et al. 1990). The major structural feature of xylan is a linear chain consisting of



1









2

chain is often substituted with acetyl, arabinofuranosyl, ferulyl glucopyranosyl, and mannopyranosyl side chains to form a complex heterogeneous structure.

Several genera of bacteria and fungi are able to partially or completely depolymerize xylan in various habitats (Biely, 1985). Xylan depolymerization by microorganisms is a multistep process which involves the concerted activities of several different enzymes. Xylanases (1,4-B-D-xylan xylanohydrolase; EC 3.2.1.8) are extracellular enzymes which hydrolyze the internal B-1,4xylosidic linkages on the main chain. The resulting smaller oligosaccharides are transported into the microbial cells where xylosidases (1,4-B-D-xylan xylohydrolase; EC 3.2.1.37) continue the hydrolysis and release monosaccharides for glycolysis (Dekker and Richards, 1976). The hydrolysis and removal of side chain substituents requires additional enzyme activities including arabinofuranosidase, which removes substituted arabinofuranosyl residues from the xylan backbone. (Biely, 1985). This may be particularly important since arabinose substituents on the xylan chain have been shown to limit the complete enzymatic breakdown of xylan (Chesson et al. 1983). One of the organisms which is particularly adept at xylan depolymerization in Butyrivibrio fibrisolvens. B. fibrisolvens is a Gram variable,








3

obligately anaerobic bacillus that is frequently found in the rumen and anaerobic digesters (Dehority, 1966). This organism produces a cadre enzymes which enable it to degrade plant biomass, including cellulose and hemicellulose (Hespell, 1987). The genus Butyrivibrio contains only a single species but consists of many strains that vary in DNA relatedness between 20% to 100% (Mannarelli, 1988). This organism is also characterized as having a low (38% to 42%) guanine plus cytosine (mole percent) content. Butyrivibrio produces an extracellular polysaccharide (EPS) that contains an unusual 4-O-(l-carboxymethyl)-rhamnose sugar (Mannarelli et al. 1990). These investigators suggested that the unusual sugars found in the EPS of B. fibrisolvens serve to protect the organism from glycosidases and other enzymes found in the digestive tract of the host animal.

B. fibrisolvens GS113, an anaerobic digester isolate used in these studies, was shown to produce high levels of both xylanase and xylosidase (Sewell et al. 1988). These two enzymes were shown to be repressed by glucose and induced by xylan and xylose.

In a previous study, the xvlB gene encoding the B-Dxylosidase from B. fibrisolvens GS113 was isolated from a plasmid pUC18 genomic library (Sewell et al. 1989).

During the course of the current investigations, an aL-arabinofuranosidase activity was detected in all clones harboring the xvlB gene. This characteristic became the









4

principle focus of this research. The principle question that was addressed relates to the dual activity exhibited by this enzyme against B-D-xylopyranosides and a-Larabinofuranosides: do these two activities reside in the same active center of the enzyme or are they on separate domains?

The following research examined the structure and function of the xvylB gene and gene product and includes:

A) The complete nucleotide sequence of xvlB and

sequence comparisons with related enzymes from other

organisms.

B) Genetic evidence that the two enzymatic activities

are encoded by a single open reading frame.

C) Mutational analyses to investigate the genetic

interdependence of the two enzymatic activities of

xvlB.

D) Purification and characterization of the xvlBencoded protein.

E) Biochemical and kinetic experiments to investigate

the functional relationship between the two enzymatic

activities.














CHAPTER II

REVIEW OF THE LITERATURE

Many different bacteria have been characterized which are able to depolymerize xylan including: Bacteroides succinogenes (Forsberg et al. 1981), Clostridium acetobutylicum (Lee et al. 1987), Bacillus pumilus (Panbangred et al. 1983), Bacillus subtilis (Paice et al. 1986), Butyrivibrio fibrisolvens (Hespell et al. 1987), Caldocellum saccharolyticum (Luthi et al. 1990), and Clostridium thermocellum (Garcia-Martinez et al. 1980). These microorganisms which degrade xylan have been isolated from numerous environments. Varel (1987) demonstrated that pig large intestine contained xylanolytic Bacteroides succinogenes and Ruminococcus flavefaciens. Both of these organisms are also present in large numbers in the bovine rumen. Other studies demonstrated that up to 30% of the metabolic energy requirements of the pig can be met via the utilization of volatile fatty acids, products of microbial cellulose and hemicellulose digestion (Rerat et al. 1987). In addition, Salyers et al. (1981) isolated two species of human colonic Bacteroides that were able to utilize xylan as a carbon and energy source, thus producing volatile fatty acid products. However, only 5 to 10 % of the maintenance

5









6

energy requirements are derived from volatile fatty acids produced in the human colon (MacNeil, 1984). Xylanolytic enzymes have also been isolated and characterized from several fungi including Trichoderma reesei (Poutanen and Puls, 1988), Aspergillus niger (Fukumoto et al. 1970), and Fusarium roseum (Gascoigne and Gascoigne, 1980).

The rumen is the primary organ in which cattle, sheep, and other ruminants derive their energy and nutrition through breakdown of complex carbohydrates. Starch, cellulose and hemicellulose are degraded by enzymes that are secreted by resident microorganisms and metabolized to volatile fatty acids as the end products of fermentation (Hobson and Wallace, 1982). In general, cellulose and hemicellulose depolymerization by rumen microbial flora releases free monosaccharides and short chain oligosaccharides. The predominant metabolic waste products, volatile fatty acids, are released and either absorbed and utilized by the animal or used by other microorganisms in the rumen and other digestive organs (omasum, abomasum and the small intestine).

The rumen microbial community represents a diverse group of organisms, many of which have the ability to degrade hemicellulose (Dehority, 1966). The rumen ecosystem differs from other microbial ecosystems in substrate availability and product accumulation. The rumen is close to an industrial fermentation in that substrate availability









7

is very high and constant while product accumulation is low. It has been established that microbial cells in the rumen are present in high numbers and contains: 1011 bacteria ml"1, 106 ciliate protozoa ml"', and 104 fungi ml' (Patterson, 1989).

Butyrivibrio fibrisolvens is a Gram variable,

obligately anaerobic, motile bacillus (Hespell and Bryant, 1981). B. fibrisolvens is particularly abundant in the rumen and anaerobic digesters in which plant material serves as the primary substrate (Hespell et al. 1987). B. fibrisolvens converts hemicellulose to mono and oligosaccharides. These are transported and metabolized to yield butyric acid. Mannarelli et al. (1990) cloned and sequenced the gene encoding 8-D-xylanase from B. fibrisolvens strain 49. Sewell et al. (1988) isolated several strains of B. fibrisolvens that produced both xylanase and xylosidase. In this study, the synthesis of both enzymes were concurrently repressed by glucose and induced by xylan and xylose. This was surprising since earlier work on rumen isolates of B. fibrisolvens had reported that these enzymes were expressed constitutively (Hespell et al. 1987). Similarly, it was reported that xylose served as an inducer of the xylanase and 8-Dxylosidase in Pullularia pullulans ( Pou-Llinas and Driguez, 1987). In addition, B-D-xylosidase of Bacillus pumilus was found to be induced by xylose (Kersters-Hilderson et al.









8

1969). Xylobiose was found to induce synthesis of Bxylanase in Cryptococcus albidus (Biely et al. 1980) and in Streptomyces sp. (MacKenzie et al. 1987). Biely and Petrakova (1984), studying the xylan-degrading system in C. albidus, found that certain positional isomers of xylose and xylobiose, notably 1,4-8-xylobiose, could serve as inducers of B-xylanase and 8-D-xylosidase.

Some organisms produce multiple xylanase enzymes.

Esteban et al. (1982) reported that Bacillus circulans WL-12 secretes two endo-B-xylanases and one B-D-xylosidase when grown on xylan as a sole carbon source. Three distinct xylanase genes have been identified and cloned from Clostridium thermocellum (MacKenzie et al. 1989). A multiplicity of xylanases has also been reported in fungi including Aspergillus niger (Frederick et al. 1985), and Trichoderma harzianum ( Wong et al. 1986). It has been suggested that the multiplicity serves to enhance the ability of microbes to depolymerize a wide range of substituted xylans under different environmental conditions.

In some organisms, the xylanolytic and cellulolytic systems are combined. Recently Morag et al. (1990) demonstrated that, in addition to free xylanases, Clostridium thermocellum possessed a cellulosome-associated xylanase which exhibits endo-glucanase activity. However this organism was unable to utilize or grow on xylan. These investigators postulated that cellulosome-associated








9

xylanolytic enzymes act to increase the availability of cellulose to cellulases of the cellulosome through removal of associated xylan chains. In the rumen, cooperativity between xylanase and cellulase degrading enzymes is also apparent. In Bacteroides succinoQenes isolated from rumen fluid it was demonstrated that carboxymethylcellulase (B1,4-endo-glucanase), B-xylanase, and B-D-xylosidase were expressed by the organism when grown on media containing cellulose as a sole source of carbohydrate (Forsberg et al. 1981). These investigators postulated that cooperativity between the cellulose and hemicellulose degrading enzymes helps to enhance polymer breakdown and increase substrate availability for rumen microorganisms which lack these enzymes. This cooperativity among different organisms may serve to maintain a stable microbial population in the rumen.

Enzymatic cooperativity and synergism is also present within the hemicellulose-degrading systems. A synergistic action of B-xylanase and 8-D-xylosidase has been demonstrated in cultures of Neurospora crassa when grown on xylan (Deshpande et al. 1986). In this study the degree of hydrolysis of D-xylan by xylanase was increased 30% by the addition of B-D-xylosidase to a cell-free system.

Another example of enzymatic synergism involves the enzyme acetyl esterase. Acetyl esterase (EC 3.1.1.6) is active against esters of acetic acid and are widely








10

distributed in nature (Poutanen et al. 1991). The acetyl residues on the xylan backbone are removed by acetyl esterase (Biely et al. 1985). These enzymes have been found to act cooperatively with xylanases. The acetyl esterase serves to increase the rate of glycosidic bond cleavage by B-xylanase from Trichoderma reesei (Biely et al. 1986). These enzymes were also found to act synergistically to liberate acetyl residues. More recently it was demonstrated that the rate of liberation of acetic acid from acetyl-xylan by acetyl esterase of T. reesei was increased by the addition of endo-xylanase and B-D-xylosidase (Poutanen and Sundberg, 1988).

Also involved in enzyme synergism is the enzyme a-Larabinofuranosidase (Greve et al. 1984). This enzyme was purified from Ruminococcus albus 8 and had a pH optimum of

6.9 and a Km of 1.3 mM, both for p-nitrophenyl-a-Larabinofuranoside as a substrate. They showed that this enzyme enhanced the rate of hydrolysis of alfalfa cell wall hemicellulose when combined with other xylanolytic or pectinolytic enzymes. It was hypothesized that this enzyme functioned to provide rumen microbes with suitable substrates for xylanase.

The mechanism of xylan hydrolysis by microbial

xylanases has been studied extensively. Xylanases are usually small proteins having molecular weights ranging between 20,000 to 50,000 (Bastawde et al. 1991). Most









11

xylanases are in fact endo-xylanases by virtue of the fact that they attack the interior 8-(1,4)-D-xylosidic linkages of the xylan polymer rather than the exterior linkages (Ward and Moo-Young, 1989). The B-(1,4)-D-endo-xylanases have a pH optimum in the range of 3.5 to 6.5 while the temperature optima and thermal stabilities vary depending upon the source (Ward and Moo-Young, 1989). Xylanase from Bacillus pumilus IPO has a molecular weight of 22,000 and is a 8-D1,4-endo-xylanase (Panbangred et al. 1983). The pH and temperature optimum of this enzyme are 6.5 and 400C, respectively. Quantification of the hydrolysis end products from larchwood xylan indicated that the B. pumilus enzyme had the greatest affinity for the second and sixth Bxylosidic linkages of the polymer.

The xvnZ gene product from Clostridium thermocellum is also an endo-xylanase with pH and temperature optima of 6.0 and 650C, respectively (Grepinet et al. 1988). Lee et al. (1987) purified and characterized two different endoxylanases, xylanase A and xylanase B, from Clostridium acetobutylicum. Xylanase A has a molecular weight of 65,000, a pH optimum of 5.0, an optimum temperature of 500C, and is stable for up to 30 min at 400C. Xylanase B is a smaller protein having a molecular weight of 29,000. It had a pH optimum of 5.0 to 6.0, showed a temperature optimum of 600C, and is stable for 30 min at 500C. Both enzymes hydrolyze larchwood xylan randomly, however xylanase B









12

produced only xylotriose and xylobiose as products whereas xylanase A also yields xylohexose, -pentose, and -treaose as end products. Xylanase A was also active against carboxymethylcellulose, acid-swollen cellulose and lichenin. The two enzymes were antigenically different as judged by "Ouchterlony"-immunodiffusion assays. The two enzymes were therefore presumed to be encoded by separate genes.

Some xylanases, such as those that are produced by fungi, notably Aspergillus niger, produce endo-1,4-B-Dxylanases that can hydrolyze the 1,3-a-L-arabinofuranosyl side chains from arabinoxylans (Dekker, 1985). These enzymes have been termed "debranching" xylanases. Recently a unique "appendage-dependent" xylanase was isolated and purified from Bacillus subtilis (Nishitani and Nevins, 1991). This enzyme is classified as a B-(1,4)-xylan xylanohydrolase and has a prerequisite for glucuronosyl substituted side chains in order to initiate hydrolysis of the xylan backbone structure. Three novel xylanases were purified from B. subtilis which exhibited activity against ferulylolated arabinoxylans (Nishitani and Nevins, 1988). These enzymes acted on ferulic acid-substituted arabinoxylan and liberated the terminal arabinofuranosyl, terminal gluconopyranosyl, and ferulic acid moieties from the polymer. While much recent work has concentrated on the extracellular microbial xylanases, less is known concerning the molecular biology and properties of the intracellular B-









13

D-xylosidase component of the microbial xylanolytic system. Early mechanistic studies of the B. pumilus xylosidase indicated the enzyme contained several thiol groups and at least one of which is involved in the catalysis (Saman et al. 1975). Panbangred et al. (1983) first cloned the genes for B-xylanase and B-D-xylosidase from Bacillus pumilus IPO. Both cloned proteins were expressed in Escherichia coli from a hybrid plasmid and were immunologically and chemically identical to those of B. pumilus. The cloned genes from B. pumilus IPO were later sequenced by Moriyama et al. (1987). The gene for B-D-xylosidase was localized to a 1617 base pairs open reading frame encoding a deduced 62,607d protein. The N-terminal amino acid sequence agreed with that predicted from the DNA sequence and that obtained from the purified enzyme. It is interesting to note that the 8xylanase gene from the same organism was located 4,600 base pairs downstream from the 3'-end of the B-D-xylosidase. The B. pumilis enzyme was not reported to exhibit any additional enzymatic activities.

More recently, two xylosidase genes were cloned and sequenced from the obligately anaerobic, thermophilic organism Caldocellum saccharolyticum (Luthi et al. 1990). The protein encoded by one of these xylosidase genes was found to possess xylanase activity in addition to the expected xylosidase activity. These genes were also found to reside in close proximity to each other and to a gene









14

encoding a protein having acetyl esterase activity. Additional B-D-xylosidase enzymes with multiple activities have been reported earlier. Kinetic methods were used to investigate the active site of a A-D-xylosidase from Chaetomium trilaterale (Uziie et al. 1985). In this study, which employed substrate analogues as inhibitors, a single active center was postulated to function for both the B-Dxylosidase and B-glucosidase activities exhibited by this enzyme. A B-D-xylosidase purified from Trichoderma reesei was also found to exhibit an a-L-arabinofuranosidase activity (Poutanen and Puls, 1988). Additionally, a cloned gene cluster from Bacteroides ovatus was also found to exhibit B-D-xylosidase and a-L-arabinofuranosidase activities (Whitehead and Hespell, 1990). These dual activities co-purified and were encoded by a single open reading frame present in the cloned gene fragment. While the dual activities of B-D-xylanases and 8-D-xylosidases have been documented, little is known about the genetic and biochemical basis of this property. Multiple substrate activities can be attributed to the presence of more than one catalytic region on the enzyme. Another possibility is the presence of a single catalytic region with wide substrate specificity. It has been proposed that the more evolved an enzyme or protein is, the more narrow its specificity becomes (Knowles, 1988). Accordingly the more primitive proteins tend to have multiple functions. It has









15

also been proposed that the environment in which an enzyme evolves also contributes to the enzyme specificity (Robson and Gardier, 1988). The rumen, an environment with specialized substrate-hydrolyzing requirements, may exert selective pressures resulting in the evolution of organisms and enzyme systems that reflect the heterogeneous nature of available substrates.















CHAPTER III

CLONING, SEQUENCING, AND SEQUENCE ANALYSIS OF THE
XYLOSIDASE GENE FROM BUTYRIVIBRIO FIBRISOLVENS


Introduction

The synthesis of enzymes needed for xylan

depolymerization has been found to be constitutive in many ruminal isolates of B. fibrisolvens (Hespell et al. 1987). Recently, anaerobic digester isolates of B. fibrisolvens have been described in which the synthesis of xylanase and xylosidase were coordinately repressed by glucose and induced by xylans and xylose (Sewell et al. 1988). The gene for B-D-xylosidase from B. fibrisolvens GS113, in which this enzyme is inducible, has been cloned on a multicopy plasmid pUC18 in Escherichia coli (Sewell et al. 1989). Subcloning analysis localized the coding region to a 5.8 kilobase pairs (kbp) segment of cloned B. fibrisolvens DNA. The enzyme was found to be predominantly intracellular in B. fibrisolvens with 25% of the activity associated with the cell membrane fraction. The cloned xylosidase is primarily cytoplasmic with less than 2% of the active protein being membrane associated in E. coli.

This investigation has been extended by restriction

endonuclease mapping the B. fibrisolvens DNA insert in pUC18


16









17

and to further define the coding region of the B-Dxylosidase gene in the insert. The number of chromosomal copies of this gene was determined by Southern hybridization. Additional subclones and primers were generated to allow complete DNA sequencing of both strands. Finally, the DNA sequence was compared with other, related gene sequences.

Materials and Methods

Medium and growth conditions. Escherichia coli DH5a was propagated at 370C in Luria broth or on Luria agar supplemented with 50 mg of ampicillin per liter (Maniatis et al., 1982).

Genetic methods. Plasmid pUC18 was used as a cloning vector in all cloning and sequencing experiments unless otherwise noted. The plasmids pLOI1001 and pLOI1005 harbor the xylosidase coding region (Sewell et al., 1989). Analysis of restriction sites, plasmid purification, subcloning, DNA ligation, Southern hybridization and other DNA manipulations were performed using standard methods (Maniatis et al., 1982). Restriction enzymes (Bethesda Research Laboratories, Gathersburg, MD) were used according to the manufacturer's instructions. Transformed colonies were screened for xylosidase and arabinofuranosidase activity on Luria agar plates containing 20 gg/ml of the flurorogenic substrates 4-methylumbelliferyl-B-Dxylopyranoside or 4-methylumbelliferyl-a-L-arabinofuranoside









18

(Sigma Chemical Co., St. Louis, MO.). The internal Sau3A and large internal PstI fragment from pLOI1005 were utilized as probes in the

Southern hybridization analysis of digested chromosomal B. fibrisolvens and E. coli DNA.

DNA sequencing. Double-stranded DNA was sequenced in both directions using the dideoxy-chain termination method (Sanger, 1982) and Sequenase (United States Biochemical Corp.) according to the manufacturer's instructions. Additional sequencing primers were synthesized by the University of Florida Interdisiplinary Center for Biotechnology Research and the Department of Microbiology and Cell Science Nucleotide facility. The DNA sequences were assembled using the "GENEPRO" software package (Hoefer Scientific Instruments, San Francisco, Calif.) and the University of Wisconsin Genetics Computer Group GCG package, version 6.1 (Devereux et al. 1984) Primary sequence comparisons were made with GenBank and EMBL sequence libraries.

Preparation of cell extracts. E. coli cells harboring the recombinant plasmids were harvested while in midexponential phase of growth by centrifugation (10,000 g, 10 min, 40C) and washed twice with 5 mM phosphate buffer (pH

6.8). Cell pellets were stored at -700C, until needed. Cell pellets were thawed on ice and resuspended in an equal volume of 5 mM phosphate buffer (pH 6.8) containing 10 mM 8-









19
mercaptoethanol and were lysed by two passes through a French pressure cell at 20,000 lb in-2. Cell membranes and other debris were removed by centrifugation (100,000 g, 1 h, 40C). Supernatants containing the total cytoplasmic proteins were stored at -700C.

Enzyme assays. B-D-xylosidase and a-Larabinofuranosidase activities were determined by measuring the rate of hydrolysis of p-nitrophenyl-B-D-xylopyranoside and p-nitrophenyl-a-L-arabinofuranoside (1 mM final concentration), respectively, in 50mM phosphate buffer (pH

6.8) at 370C. The nitrophenyl derivatives of other monoand disaccharides were examined as possible substrates under the same conditions. All assays were conducted in a volume of 1 ml catalysis was terminated by the addition of 2 ml of 500 mM sodium carbonate. The hydrolysis of one nmole of substrate resulted in an increase of absorbance of 0.007 at 405nm. Specific activities are expressed as nmoles pnitrophenol released per minute per milligram of total protein. Carbohydrate derivatives were purchased from Sigma Chemical Co. Protein concentration was estimated by the method of Bradford (Bradford, 1976).

Sodium dodecyl sulfate-polyacrylamide gel

electrophoresis (SDS-PAGE). Cell proteins were separated in denaturing gels by the method of Laemmli (Laemmli, 1970). Protein bands were visualized by staining with Coomassie blue.









20

Nucleotide sequence accession number. The nucleotide sequence reported here has been assigned GenBank accession number M55537.

Results and discussion

Identification of the xylosidase coding region. Many restriction endonuclease sites on the original xylosidasepositive subclone, pLOI1005 (3.2 kb B. fibrisolvens insert), were mapped to facilitate the generation of futher subclones for analysis (Fig. 1). Plasmid DNA was purified using cesium chloride and digested with a battery of restriction enzymes. Restriction endonuclease sites were used to generate subclones in both directions with respect to the lac promoter in pUC18. Each subclone was examined for enzyme activity on 4-methylumbelliferyl-B-D-xylopyranoside (MUX) indicator plates. Based upon the results of these experiments (Fig. 2), the region encompassing the xylosidase gene was localized to a 2.1 kb DNA segment that spans an internal EcoRI site and the internal PstI site. The predicted gene size was in excess of the 1.4 kb DraI fragment. In addition to the indicated xylosidase a ctivity, arabinofuranosidase activity was also associated with all xylosidase-positive subclones. It seems possible that the gene, classified as xylB, encodes an enzyme that has activity against both substrates.

The number of chromosomal copies of xylB was examined using Southern hybridization. The large internal PstI and




























Figure 1. Restriction endonuclease digestion maps of plasmids pLOI1001 and pLOI1005 that express B-D-xylosidase activity in E. coli DH5a.








Pstl
Hndl
Xbal 22

O Pati



It


CD EcoRI " pLOI 1001 o a z
L 6.95 Kbp Hindll

Xbal

Indit
Kpnl
EcoR Xbal
EcoRI
PstI
Kpnl bal
bal
EcoRI Hindll

Sau3A KL ,Psti Sau3A
Sau3A 0







__ 7/ ba
ral Dral
Sau3A Hnl c
b EcoRI

HindIII


pLOI 1005 ? Xbal Dral EcoRV m 5.8 Kbp z Sau3A~ s\~~-~c

Kpni
ON 4 Sau3A Xbal
Pati
Dral


EcnRI !. .. Accl Xbal





























Figure 2. Subclone analysis of pLOI1005 to localize the 2xg coding region: xyl; xylosidase activity, araf; arabinofuranosidase activity. A "+" or "-" denotes the presence or absence of enzyme activity.











kilobase pairs 24 .0 1.0 .0 3.0 p s E H P H Xvyl Araf



I I I I I + +

H D H S D E


P P





E E H H





H H





H H





S S





D D






Coding Region









25

Sau3A fragments were used as probes against PstI and Sau3A digests of E. coli DH5a, B. fibrsolvens GS113 chromosomal DNA and plasmid pLOI1005 (Fig. 3). These probes did not bind to the DH5a chromosomal DNA but did bind to a single band in each B. fibrisolvens chromosomal digest. A single copy of the xvlB gene appears to be present in B. fibrisolvens GS113.

Multiple copies of xylosidase genes have been reported for Bacillus pumilus (Panbangred et al., 1984) and for Caldocellum saccharolyticum (Luthi et al., 1990). If there are additional xylosidase genes present in GS113, then they must share limited homology with xvlB.

DNA sequence of the xylosidase gene (xylB). Plasmid pLOI1001, the original GS113 library clone (Sewell et al., 1989), contained a 4.2 kb insert of B. fibrisolvens DNA. Both strands of this fragment were sequenced utilizing the strategy summarized in Fig. 4A. The complete nucleotide sequence of this fragment is outlined in Fig. 5. Analysis of the sequence revealed the presence of three open reading frames (ORF's) in this DNA segment. The first ORF, ORF1, was incomplete and is 1,340 bp in length. It lacks a ShineDalgarno sequence and an ATG initiation codon. ORF2, was 1,551 bp in length and was found 15 bp downstream from ORF1. ORF2 has a Shine-Dalgarno sequence located 6 bp upstream from the initiation codon and defines a complete gene. This gene spans the predicted xylosidase coding region and






























Figure 3. Southern hybridizations of chromosomal DNA from B. fibrisolvens GS113 (8 hr exposure) and E. coli DH5a (8 hr exposure) and plasmid DNA from pLOI1005 (2 hr exposure). Lanes 1, 2, and 3 contain PstI digests from E. coli, B. fibrisolvens, and pLOI1005, respectively, probed with the internal PstI fragment from pLOI1005. Lanes 4, 5, and 6 are Sau3A digests of E. coli DH5a, B. fibrisolvens GS113 and pLOI1005 which were probed with the internal Sau3A fragment from pLOI1005. The additional bands in lane 6 represent larger, incomplete restriction endonuclease digestion products.















N CNJ

































Figure 4. Outline of sequencing strategy of pLOI1001 and subclone analysis of xvlB. (A) Sequencing strategy of the complete 4.2 kb B. fibrisolvens insert in pLOI1001. Arrows indicate the direction of sequencing. Subclones were sequenced using universal pUC18 primers (vertical bars in front of arrows) and additional oligonucleotide primers (vertical bars absent) were both used. (B) Outline of the three ORF's sequenced and selected subclones and insertional inactivation used to identify xvlB. Enzyme activity was evaluated using MUG indicator plates. Double vertical bar in pLOI1040 indicates the site of insertion of a 10 bp NotI linker. Abbreviations: E; EcoRI, H; HindIII, A; AccI, X; XbaI, D; DraI, S; SsI, P; PstI, ERV; EcoRV, xyl; xylosidase activity, araf; arabinofuranosidase activity. A "+" or "-" denotes the presence or absence of enzyme activity, respectively.











A pLOI 1001 29 E E H AX D S P H ERV H E D S P P











---) I-B BASE PAIRS

0 1000 2000 3000 4000




ORF1

ORF2 ORF3



y araf
E p
pLOI 1005
+ +

S S
pLOI1043
+ +

E pLOI1040 P




NotI linker insertion


































Figure 5. The complete nucleotide sequence and translated amino acid sequence of the 4.2 kb insert from pLOI1001. Putative Shine-Dalgarno (S.D) sequences and initiation codons are underlined. Translational termination is indicated by an asterisk (*).











31
1 AATTGTGGATGCACATATGAAAAGCTGATTTATGCTTATAAGGCAGGTCTTGTCAAGGAA 60
N C G C T Y E K L I Y A Y K A G L V K E

61 GAGACCATCGATGAGGCTGTTACTCGACTTATGGAAATCAGACTTCGTCTAGGTACTATT 120
E T I D E A V T R L M E I R L R L G T I

121 CCAGAGAGAAAGAGTAAGTATGATGATATCCCATATGAAGTGGTCGAATGCAAAGAGCAT 180
P E R K S K Y D D I P Y E V V E C K E H

181 ATCAAACTTGCTCTTGACGCTGCAAAGGATAGCTTTGTCCTTTTGAAGAATGATGGTTTA 240 I K L A L D A' A K D S F V L L K N D G L

241 CTTCCACTGAATAAAAAGGATTATAAATCTATTGCTGTTATTG CTG ATTCA 300 L P L N K K D Y K S I A V I G P N S D S

301 AGAAGAGCTTTAATTGGAAATTATGAGGGCCTTTCTTCAGAGTATATTACAGTTTTAGAG 360
R R A L I G N Y E G L S S E Y I T V L E

361 GGGATTCGTCAGGTTGTCGGTGATGATATTAGATTATTCCACGCTGAGGGCACTCATCTT 420
G I R Q V V G D D I R L F H A E G T H L

421 TGGAAGGATAGAATTCACGTAATCAGTGAGCCAAAA GATTTGCCGAGGCTAAAATC 480
W K D R I H V I S E P K D G F A E A K I

481 GTGGCAGAGCATTCAGATTTAGTTGTGATGTGTCTTGGACTTGACGCATCAATCGAAGGA 540
V A E H S D L V V M C L G L D A S I E G

541 GAAGAAGGAGACGAGGGTAATCAGTTCGGTAGCGGAGACAAGCCTGGATTAAAGCTTACA 600
E E G D E G N Q F G S G D K P G L K L T

601 GGTTGTCAGCAAGAGCTACTTGAGGAAATTGCCAAAATCGGCAAGCCTGTTGTACTTCTT 660 G C Q Q E L L E E I A K I G K P V V L L

661 GTGCTTTCAGGTTCTGCTCTTGATTTATCATGGGCGCAGGAATCTAATAACGTAAATGCG 720
V L S G S A L D L S W A Q E S N N V N A

721 ATAATGCAGTGCTGGTATCCAGGCGCAAGAGGTGGACGTGCTATTGCAGAGGTTTTATTT 780 I M Q C W Y P G A R G G R A I A E V L F

781 GGCAAGGCCAGTCCAGGCGGTAAAATGCCTCTTACATTTTATGCCTCAGATGATGACCTT 840
G K A S P G G K M P L T F Y A S D D D L

841 CCTGATTTTTCTGATTATTCAATGGAAAATAGGACATACAGATATTTCAAGGGCACACCA 900
P D F S D Y S M E N R T Y R Y F K G T P

901 CTTTATCCATTTGGTTATGGACTAGGTTATTCTAAAATTGATTATCTATTTGCTTCTATT 960
L Y P F G Y G L G Y S K I D Y L F A S I

961 GATAAAGATAAGGGAGCAATTGGTGATACATTCAAGCTAAAGGTAGATGTTAAAAATACC 1020 D K D K G A I G D T F K L K V D V K N T

1021 GGTAAGTATACACAGCATGAGGCTGTTCAAGTATATGTAACGGACCTTGAGGCAACGACA 1080
G K Y T Q H E A V Q V Y V T D L E A T T

1081 AGAGTGCCTATTAGAAGCCTTAGAAAGGTTAAATGTCTAGAGCTTGAGCCTGGTGAAACA 1140
R V P I R S L R K V K C L E L E P G E T

1141 AAAGAGGTTGAATTTACCCTTTTTGCAAGAGATTTTGCCATTATTGATGAAAGGGGAAAA 1200
K E V E F T L F A R D F A I I D E R G K

1201 TGTATCATAGAGCCAGGCAAGTTTAAGATTTCTATTGGGGGACAACAGCCAGACGATAGA 1260
C I I E P G K F K I S I G G Q Q P D D R











32
1261 AGTAAAGAACTTATGGGCAGAGAGTGTGATATTTTTGAAATTGAATTAACAGGCTCTGTT 1320
S K E L M G R E C D I F E I E L T G S V

1321 ACAGAAGTTGAATATTAATTGAGAGGTGCATCATGGTTATAGCTAACAATCCAATTTTAA 1380
T E V E Y * M V I A N N P I L K

1381 AAGGTTTTTATCCAGACCCTTCTATCTGCAGAAAAGGGGATGATTTTTATCTAGTTTGTT 1440
G F Y P D P S I C R K G D D F Y L V C S

1441 CAAGTTTTGTGTATGCTCCGGGAGTACCGATTTTTCACACTAAGGATTTGGCACATTTTG 1500
S F V Y A P G V P I F H T K D L A H F E

1501 AGCAAATTGGAAATATATTAGACAGAGAAAGTCAACTTCCATTGTCGGGAGATATATCTA 1560
Q I G N I L D R E S Q L P L S G D I S R

1561 GAGGCATATTTGCCCCAACAATAAGAGAGCATAATGGAATCTTTTACATGATAACAACTA 1620
G I F A P T I R E H N G I F Y M I T T N

1621 ATGTAAGCTCTGGCGGCAACTTTATTGTTACTGCAAAAGATCCAGCTGGTCCTTGGTCAG 1680
V S S G G N F I V T A K D P A G P W S E

1681 AGCCATATTATTTAGGTGAAGATGAGGCGCCAGGTATTGATCCATCTCTGTTTTTTGATG 1740
P Y Y L G E D E A P G I D P S L F F D D

1741 ACGATGGCAAATGTTATTACGTTGGTACCAGACCAAATCCTGATGGAGTTCGTTACAACG 1800
D G K C Y Y V G T R P N P D G V R Y N G

1801 GTGATTGGGAGATATGGGTTCAAGAGCTGGATTTAGAGCAAATGAAACTTGTAGGTCCTT 1860
D W E I W V Q E L D L E Q M K L V G P S

1861 CGATGGCAATTTGGAAGGGCGCTCTTAAGGATGTTATTTGGCCAGAAGGACCACACCTTT 1920
M A I W K G A L K D V I W P E G P H L Y

1921 ATAAGAAAGATGGATATTATTATCTTTTACATGCAGAAGCTGGCACAAGCTTTGAACATG 1980
K K D G Y Y Y L L H A E A G T S F E H A

1981 CTATTTCTGTAGCTCGCTCAAAGGAGCTATTCAAATGGTTTGAGGGATGTCCTAGAAATC 2040
I S V A R S K E L F K W F E G C P R N P

2041 CTATATTTACCCATAGAAATTTAGGCAAGGATTATCCAGTATGCAATGTTGGACATGCTG 2100
I F T H R N L G K D Y P V C N V G H A D

2101 ATTTAGTTGATGATATCAATGGCAACTGGTATATGGTGATGCTGGCATCTAGACCATGCA 2160
L V D D I N G N W Y M V M L A S R P C K

2161 AGGGAAAGTGCAGCTTGGGACGAGAGACATTCCTTGCAAAAGTAATTTGGGAAGACGGAT 2220
G K C S L G R E T F L A K V I W E D G W

2221 GGCCAGTGGTTAATCCGGGAGTTGGTCGTTTGACTGATGAGGTGGAGATGGACCTTCCTG 2280
P V V N P G V G R L T D E V E M D L P E

2281 AATATAGATTCTCAAAAGAGATTACTACAAAGGATAAAATGACCTTTGAAGAGACAGTCC 2340
Y R F S K E I T T K D K M T F E E T V L

2341 TAGATGATAGATTTGTTGGAATTGAAAGAAGAAGTGAGGACTTTTATTCCCTTACTGACA 2400
D D R F V G I E R R S E D F Y S L T D N

2401 ATCCTGGATTCTTAAGATTAAAGCTTCGTCCTGAGGCCATAGAAAATACTGGCAATCCAT 2460
P G F L R L K L R P E A I E N T G N P S

2461 CTTACTTAGGAATTCGTCAAAAGACTCATTCGTTTAGAGCAAGCTGTGGCCTTAAGTTTA 2520
Y L G I R Q K T H S F R A S C G L K F T









33
2521 CACCAGCAAAAGATAATGAATGTGCAGGAATGGTGTTATTCCAGAATAATGAAAATCACT 2580
P A K D N E C A G M V L F Q N N E N H L

2581 TGGAGCTTTTAGTTGTAAAGAAGAAAGATAAGCTACAGTTTAAAGTAGGACCAGTTATTA 2640
E L L V V K K K D K L Q F K V G P V I K

2641 AAGGAACCAAAATCAGACTTGCTACTTTTGATATTTCATCAGGTGATTTAGAAATTATTC 2700
G T K I R L A T F D I S S G D L E I I L

2701 TTGAGGCAGCAAATCAGCTGGCTAATATCTATATTAAAAAGAATAATGAAAAGATTCTTG 2760
E A A N Q L A N I Y I K K N N E K I L V

2761 TGGCAGAATGTATTGATTTGAGCCCATACACTACCGAAGAATCAGGCGGATTCGTAGGAT 2820
A E C I D L S P Y T T E E S G G F V G C

2821 GTACCATTGGACTATATGCTTCTTCAAATGGAAAGACCAGTGATAACTATTGCGATTATT 2880
T I G L Y A S S N G K T S D N Y C D Y S

2881 CCTACTTTACAGTAGAAGAAGTATAGCATTTTCAATGAGCGAATTTGCAAGTTTTATATA 2940
Y F T V E E V *

2941 CGGGATTAATTGTACGTAAAAACCATACAGGTGTAAAATAGTTTCCAGAGAAAGTTTTTT 3000

3001 CTCTGGAATTTTTTATTATgg GGATTATGCTTCAGGAAAGTATTAAGAAGTTGGTAC 3060 M L Q E S I K K L V Q

3061 AGTACGGTATTGATATGGGGCTTACACCAGAATGTGAGAGAATATATACTACAAATCTTT 3120
Y G I D M G L T P E C E R I Y T T N L L

3121 TGCTTGAATGTATGAAAGAAGATGAGTACATAGATCCAGACTGTGATTTAAGCAATATTA 3180
L E C M K E D E Y I D P D C D L S N I I

3181 TACTTGAAGATGTATTAAAGGAACTTTTAGATGAGGCAGTTAATAGAGGTATCATAGAGG 3240
L E D V L K E L L D E A V N R G I I E D

3241 ATTCAGTTACACATAGGGATTTGTTTGATACAAAGCTAATGAATCAGCTATGCCCACGTC 3300
S V T H R D L F D T K L M N Q L C P R P

3301 CTAAACAGGTTATAGATGATTTTAACCGTATACGATAACCATGGTCCAATAGCTGCAA 3360
K Q V I D D F N R I Y D N H G P I A A T

3361 CAGATTATTTTTACAAGTTAAGCAAAGCCTCTGACTATATCCGTACTTACAGGGTAAAAA 3420
D Y F Y K L S K A S D Y I R T Y R V K K

3421 AGGACCTAAAATGGACATGCGATACAGAGTATGGCACTCTTGACATAACAATTAATCTCT 3480
D L K W T C D T E Y G T L D I T I N L S

3481 CTAAGCCAGAAAAAGACCCAAAGGCAATTGCTGCAGCTAAGAATGCAAAACAATCCACAT 3540
K P E K D P K A I A A A K N A K Q S T Y

3541 ATCCGAAGTGCCAATTATGTATGGAAAATGAAGGCTATGCTGGTCGCATTAATCATCCTG 3600
P K C Q L C M E N E G Y A G R I N H P A

3601 CTAGAGAGAATCATCGCATAATTCCTATAACTATAAATAACAGCAACTGGGGATTTCAAT 3660
R E N H R I I P I T I N N S N W G F Q Y

3661 ATAGCCCATACGTTTATTACAATGAGCATTGCATAGTCTTTAACGGAGAGCATACTCCTA 3720
S P Y V Y Y N E H C I V F N G E H T P M

3721 TGAAAATAGAGCGAGCTACTTTTGTTAAGCTATTTGATTTTATCAAACTATTTCCACACT 3780
K I E R A T F V K L F D F I K L F P H Y

3781 ATTTTTTAGGAAGCAATGCTGATTTACCAATTGTTGGAGGATCTATTTTAAGCCATGACC 3840
F L G S N A D L P I V G G S I L S H D H











34
3841 ATTTCCAAGGCGGCCATTACACATTTGCCATGGAAAAAGCCATGGAAAAAGAATATTCAGGAATTTA 3900
F Q G G H Y T F A M E K A P I I Q E F T

3901 CTGTAAAAGGATATGAGGATGTTAAGGCTGGTATAGTTAAATGGCCACTTTCAGTAATTA 3960
V K G Y E D V K A G I V K W P L S V I R

3961 GACTTCAGTGCAAGGATGAGACTAGACTTATTGATTTAGCGACTAATATATTAGACAAAT 4020
L Q C K D E T R L I D L A T N I L D K W

4021 GGAGAAATTACACCGATGAAGAGGCATATATTTTGCTGAAACAGATGGTGAGCCTCACA 4080
R N Y T D E E A Y I F A E T D G E P H N

4081 ATACGATTACACCTATTGCTAGAAAAAGAGGGGATTACTTTGAACTAGATCCTCTAGAGT 4140
T I T P I A R K R G D Y F E L D P L E S

4141 CGACCTGCAGGCATGCAAGCTTGGCACTGGCCGTCGTTTTACAACGTCGTGACTGGG* 4198
T C R H A S L A L A V V L Q R R D W









35

encodes a 517 amino acid protein having a calculated molecular weight of 58,421. ORF3, is located 123 bp downstream from ORF2 and includes a Shine-Dalgarno sequence with an initiation codon 5 bp downstream. This ORF continues for 1,173 bp until the end of the clone and is also incomplete. No predicted "stemloop" structures or sequences that resemble rho-independent transcriptional terminators were identified by computer analysis in the region between ORF2 and ORF3. It is therefore unlikely that this DNA functions as a transcriptional terminator in E. coli. It is also unlikely that any transcriptional terminators are present between ORF1 and ORF2 since these two ORF's are only separated by 15 bp and ORF2 is expressed in large amounts in E. coli in constructs that also contain ORF1. This evidence suggests that these three ORF's may constitute part of a xylan-degrading operon in B. fibrisolvens.

Codon usage. The codon usage of the three B. fibrisolvens ORF's is summarized in Table 1. For comparative purposes, codon usage for B. fibrisolvens 49 xvnA (Mannarelli et al. 1990) and the average codon usage for E. coli (Allf-Steinberger, 1984) are included. The three B. fibrisolvens ORF's have similar patterns of codon usage with each other and with strain 49 xvnA. The low guanine plus cytosine content of B. fibrisolvens is reflected in the three ORF's in the usage of an A or a T in












36




TABLE 1. Comparison of codon usage frequency for the three B. fibrisolvens ORF's.



Frequency (mol %) Codon Usaae Amino Acid Codon B. fibrisolvens E. coli B. fibrisolvens ORF 1 ORF 2 ORF 1 xvlA



Phe TTT 2.7 4.1 3.6 1.3 2.9

TTC 0.9 1.2 0.3 2.2 1.5 Leu TTA 2.0 2.5 2.8 0.7 0.7

TTG 0.2 1.2 1.0 0.9 0.5 CTT 5.6 2.7 2.5 0.8 4.4 CTC -0- -0- 0.3 0.8 -0CTA 1.4 1.0 1.8 0.2 0.2 CTG 0.5 0.8 0.3 6.8 1.0 Ile ATT 4.7 3.9 4.3 2.2 2.7

ATC 2.0 1.0 0.8 3.7 1.7 ATA 0.5 1.7 3.8 0.2 0.5

Met ATG 1.4 1.7 2.0 2.8 2.7 Val GTT 2.9 2.7 2.3 2.9 2.0

GTC 0.9 0.2 0.5 1.2 0.5 GTA 1.4 2.1 1.3 1.8 4.2

GTG 1.1 1.2 -0- 2.2 1.0 Ser TCT 2.3 1.5 0.8 1.3 1.2

TCC -0- 0.4 0.3 1.5 0.5 TCA 1.8 1.5 0.5 0.4 2.9 TCG -0- 0.6 0.3 0.6 0.2

AGT 0.9 0.8 0.3 0.3 -0AGC 0.7 1.0 1.8 1.4 1.2











37





TABLE 1. (continued)



Frequency (mol %) Codon Usage Amino Acid Codon B. fibrisolvens E. coli B. fibrisolvens ORF 1 ORF 2 ORF 1 xvlA



Pro CCT 1.4 1.7 1.8 0.5 1.2

CCC 0.2 -0- -0- 0.3 -0CCA 2.3 3.5 2.8 0.7 2.4 CCG -0- 0.6 0.3 2.5 -0Thr ACT 0.7 1.9 2.3 1.1 0.7

ACC 0.7 1.4 0.5 2.4 1.0 ACA 2.7 1.7 3.1 0.3 5.1 ACG 0.5 -0- 0.3 0.8 0.5 Ala GCT 2.7 2.1 2.8 2.6 1.5

GCC 1.1 0.4 0.8 2.2 1.2 GCA 2.0 2.3 2.0 2.3 4.1 GCG 0.5 0.2 0.2 3.2 -0Tyr TAT 4.3 3.5 2.8 1.0 3.9

TAC 0.2 1.2 2.5 1.5 3.4 His CAT 0.9 1.4 2.5 0.7 1.5

CAC 0.5 0.6 0.5 1.2 0.2

Gin CAA 0.7 1.0 1.3 1.0 0.5

CAG 1.6 0.6 1.5 3.2 2.2

Asn AAT 2.0 4.2 3.8 1.0 5.4

AAC 0.5 1.1 1.3 2.8 2.7











38





TABLE 1. (continued



Freauency (mol %) Codon Usage Amino Acid Codon B. fibrisolvens E. coli B. fibrisolvens ORF 1 ORF 2 ORF 1 xvlA



Lys AAA 3.4 3.1 3.6 4.1 3.4

AAG 4.7 3.5 3.3 1.3 3.9 Asp GAT 5.8 5.6 5.9 2.5 3.9

GAC 1.6 1.4 2.0 3.0 0.2 Glu GAA 4.3 3.7 3.7 4.9 1.7

GAG 4.9 3.5 3.8 1.8 2.7 Cys TGT 1.4 1.4 1.0 0.4 0.7

TGC 0.7 1.0 1.5 0.5 0.5 Trp TGG 1.3 1.7 1.3 0.7 2.0 Arg CGT 0.7 0.8 1.3 3.1 1.0

CGC -0- 0.2 0.5 2.0 0.2 CGA 0.2 0.2 0.2 0.2 -0CGG -0- -0- -0- 0.2 -0AGA 3.1 2.9 2.0 0.1 2.0 AGG 0.5 -0- 0.8 0.1 0.2 Gly GGT 3.4 1.7 1.5 3.8 2.4

GGC 2.3 2.1 1.0 3.1 1.5 GGA 3.1 4.4 1.5 0.4 2.7 GGG 0.5 0.2 0.8 0.6 0.5









39

the wobble position with the exception of CAG for Gln, AAG for Lys, and GAG for Glu.

Insertional inactivation and subclone analysis of

ORF2. Subclones were generated to investigate the

relationship between the xylosidase and arabinofuranosidase activities encoded by the xvlB gene (Fig. 4b). Retention or loss of enzymatic activity was initially asssayed on 4methylumbelliferyl-B-D-xylopyranoside (MUX) and 4methylumbelliferyl-a-L-arabinofuranoside (MUA) indicator plates. The SsAI fragment (1,843 bp) from pLOI1005, which contains 17 bp upstream and 274 bp downstream in addition to xvlB, was subcloned in both directions in pUC18. Both activities were concurrently expressed only when xvlB was cloned in the direction of transcription of the lac promoter (pLOI1043), indicating a dependence on this promoter in E. coli. The insertion of a 10 bp NotI linker into the unique EcoRV site of xvlB (pLOI1040) resulted in a frameshift mutation that abolished both enzymatic activities (Fig. 4b). The results of the indicator plate assays were confirmed by comparing the specific activities for xylosidase and arabinofuranosidase in cell free extracts (Table 2.). Using p-nitrophenol derivatives, arabinofuranosidase activity was approximately 1.7-times higher than the xylosidase activity. The ratio of these activities was the same for the three active subclones. The original subclone, pLOI1005, exhibited the highest specific activity. The smaller active











40





TABLE 2. Expression of enzyme activities in recombinant E. coli harboring xvlB.


Sp act'


plasmid xylosidase arabinofuranosidase Ratio ara/xylb pLOI1005 9.0 16.0 1.8 pLOI1040 0.2 0.2 pLOI1043 2.0 3.0 1.5 pLOI1050 6.8 10.1 1.5 pUCi8 0.2 0.2 ' Nanomoles per minute per milligram of cell protein. b Ratio calculated after subtraction of background values from the pUC18 control.









41

subclone, pLOI1043, exhibited a four-fold decrease in both activities but retained a similar ratio of arabinofuranosidase to xylosidase activities. An additional subclone, pLOI1050, contained two SsPI fragments each harboring xvlB oriented with the direction of transcription from the lac promoter. This subclone exhibited a three-fold increase in both enzymatic activities with respect to the single insertion (pLOI1043) but less than the original clone. Again the ratio of the two activities remained essentially the same as the wild type (pLOI1005). The results of these experiments demonstrate the dependence of both enzymatic activities on ORF2.

Presence of other glycosidic activities. The presence of additional hydrolytic activities was examined in the xvlB-encoded protein using various ortho- and paranitrophenyl glycosidic substrates (Table 3). No additional activity above the background levels was detected with 12 other pentose and hexose derivatives. Low levels of activity was detected against the o-NP-B-D-fucopyranoside. This may not be significant since the activity represents less than 5% of the activity against the B-D-xylopyranoside. A 19-fold higher activity was detected against onitrophenyl-B-D-xylopyranoside relative to the parasubstituted derivative. This phenomenon is analogous to B-galactosidase from E. coli. The para- and orthosubstituted substrates are known to have different rates for











42












TABLE 3. Hydrolysis of different nitrophenyl-substituted glycosides by the xvlB gene product.

Specific activity
substrate pLOI1005 pUCl8



p-NP-B-D-xylopyranoside 8.9 0.2 p-NP-a-L-arabinofuranoside 15.5 0.5 p-NP-a-L-arabinopyranoside 0.1 0.2 p-NP-a-D-galactopyranoside 0.2 0.2 p-NP-a-D-glucopyranoside 3.5 3.3 p-NP-a-L-fucopyranoside 0.2 0.2 p-NP-B-D-fucopyranoside 0.2 0.2 p-NP-B-L-fucopyranoside 0.4 0.4 p-NP-a-L-rhamnopyranoside 0.3 0.2 o-NP-B-D-fucopyranoside 1.4 1.0 o-NP-a-D-galactopyranoside 0.8 1.0 o-NP-8-D-galactopyranoside 1.0 1.0



' Nanomoles per minute per milligram of cell protein.









43

the glycosidic bond-breaking step (Martinez-Bilbao et al., 1991). Under the conditions of these assays it appears that the xvlB-encoded protein is limited to hydrolytic activity against B-D-xylopyranosides and a-L-arabinofuranosides only.

Electrophoretic analysis of cloned proteins. Using

SDS-PAGE analysis of cell-free cytoplasmic extracts, a new protein band with an apparent molecular weight of 60,000 was observed in cells harboring plasmid pLOI1005 (Fig. 6). This band was absent in extracts from cells containing the vector plasmid pUC18 alone. Extracts from which the gene was inactivated by a frameshift mutation (pLOI1040) also lacked this protein band. The observed levels of this protein band in the single (pLOI1043) and double (pLOI1050) SsgI subclones was consistent with the presence of the enzyme.

Primary sequence homology comparisons. Homologies of the B. fibrisolvens ORF's to other glycohydrolases were compared to determine evolutionary relatedness. The translated amino acid sequences of the three B. fibrisolvens ORF's exhibited 42 to 45% similarity (a conservative match) and 14 to 19% identity (an exact match) with each other (Table 4.). The xvlB was found to be most similar (44% similarity, 20% identity) to the 8-glucosidase from Kluvveromyces fragilis. Additional comparisons with other glycohydrolase sequences revealed no significant amino acid identities. Although the N-terminal sequence of the


































Figure 6. SDS-PAGE analysis of cytoplasmic extracts from recombinant E. coli DH5a harboring selected plasmids. Approximately 20 Ag of protein was loaded in each lane. Lanes 1 and 7; molecular weight markers, lane 2; pLOI1005, lane 3; pUC18, lane 4; pLOI1040, lane 5; pLOI1043, lane 6; pLOI1050. The band cooresponding to the xylosidase-arabinofuranosidase enzyme is indicated by an arrow. The numbers in the right represent the apparent molecular weight of the standards
(X 103).









45


















S94









12 345 87

12 3 4 5 67











46





TABLE 4. Comparison of the translated amino acid sequences of the three B. fibrisolvens ORF's in pLOI 1001 with those of selected proteins.


% Similarity (% Identity)
ORF 1 ORF 2 ORF 3 Reference Organism (gene) xvlB)


B. fibrisolvens (xvnA) 40 (16) 41 (16) 41 (13) Mannarelli et al. 1990 B. fibrisolvens (endl) 16 (8) 14 (6) 20 (11) Berger et al. 1989 B. pumilus (xynA) 46 (21) 44 (16) 12 (8) Fukusaki et al. 1984 B. pumilus (xvnB) 39 (16) 28 (21) 38 (17) Moriyama et al. 1987 D. subtilis (xvnA) 48 (21) 44 (16) 12 (6) Paice et al. 1986 Caldocellum 20 (14) 25 (12) 21 (12) Luthi et al. 1990 saccharolvticus (xvnB)

C. saccharolyticus 16 (8) 20 (9) 19 (10) Luthi et al. 1990
(xvnA / xvnB)

C. saccharolyticus 16 (8) 22 (11) 31 (3) Luthi et al. 1990
(xvnC)

Clostridium
thermocellum 18 (7) 40 (16) 15 (8) Grepinet et al. 1988

(xvnZ)

C. thermocellum 45 (22) 41 (18) 39 (15) Benguin et al. 1985
(celA)

C. thermocellum 44 (20) 44 (18) 43 (19) Grepinet et al. 1986
(celB)

C. thermocellum 45 (16) 39 (15) 43 (18) Joliff et al. 1986
(celD)

Asperaillus
nicer
(a-amylase) 40 (16) 41 (18) 43 (19) Boel et al. 1984 Kluyveromyces
fragilis 52 (31) 44 (20) 46 (22) Raynal et al. 1987 (8-glucosidase)


ORF1 100 44 (19) 45 (19) This study ORF2 44 (19) 100 42 (14) This study ORF3 45 (19) 42 (14) 100 This study









47

Bacillus pumilus B-xylosidase did exhibit strong amino acid identity in selected regions, the overall identity was only 21%. Thus the B. fibrisolvens xvlB gene is evolutionarily divergent from other glycohydrolases. The translated, primary sequences for ORF1 and ORF3 also exhibited similarity (52% and 46%, respectively) and identity (31% and 22%, respectively) to the K. fragilis B-glucosidase. This is consistant with these two ORF's also being involved in carbohydrate degradation.

It has been postulated that the hydrolytic mechanism of lysozyme (Teeri, et al. 1987) and cellulases (Knowles et al., 1987) can serve as a model for other carbohydratehydrolyzing enzymes. Studies of hen egg-white lysozyme (HEWL) indicate a general acid-base catalytic mechanism involving Glu-35 and Asp-52 as the catalytic residues (Quiocho, F. A., 1986). Subsequent studies have demonstrated that this catalytic region is conserved in some cellulases (Knowles et al. 1987). An analysis of the translated primary sequence from xylB reveal a region homologous to the active site region from HEWL, and glucoamylase from Aspergillus niger (Table 5). The conserved region from additional carbohydrate hydrolases are included for comparison. The xvlB region was most similar to the glucoamylase, with 38% identity between the amino acids in the catalytic region. The catalytically important










48





TABLE 5. Amino acid sequence alignment of conserved regions.

Protein' Concensus sequence Reference


35 44 52
HEWL F E S N F N T Q A T . N R N T D G S . . . T 2 Y Yaguchi et al. 1983

331 338 349
A.n. (a-aml) P . 2 T Y . Y N G .. N P W F L C T L A A A & Q Boel et al. 1984

342 350 362
B.f. xvlB, S D F Y S L T D . N P G F L R L K L R P A This study

144 149 160
B.f. xv l P 2 G V R Y ..... N G A W E I W V Q E L P L This study

320 328 337
B.f. endl G T S A T N R N .. N T A E R V K W A .. . Y Berger et al. 1989

355 362 369
B.f. xvnA N E K P L I W S . .. N I G V A K P A Y .. . E Mannarelli et al. 1990

769 785
B.f. bQll S D W W G F G E H Y K . .... E V L A G N D I Barnett et al. 1991

325 335 343
B.p. xvnB I E C T R L A Q L N W N T C S M Q F V . . . E Moriyama et al. 1987

SAbbreviations: Hen egg-white lysozyme (HEWL), Aspergillus niger a-amylase (aaml), B. fibrisolvens 8-D-xylosidase (xvlB), B. fibrisolvens endoglucanase 1 (endi), B. fibrisolvens 8-D-xylanase (xvlA), B. fibrosolvens 8-glucosidase (bQlI), Bacillus Dumillus B-D-xylosidase (yvlB), Clostridium thermocellum B-D-xylanase (xvnZ), C. thermocellum cellobiohydrolase B (celB), C. thermocellum cellobiohydrolase D (celD), C. cellulolvticum endoglucanase A (EGCCA), Caldocellum saccharolyticum B-Dxylosidase (xvnB), Cellulomonas fimi endoglucanase A (cen ), C. fimi exoglucanase A (cex), Trichoderma reesei cellobiohydrolase II (CBH II), Kluvveromyces fraailis cellobiohydrolase I (CBH I), Schizophvllum commune endoglucanase (EG I).











49





TABLE 5. (continued)


Protein' Concensus sequence Reference


458 466 475
C.t. xvnZ G I A L L R A D V .. H R S G K V D S . T P Y Grepinet et al. 1988

418 427 436
C.t. celB T 3 G G H P L L D L . H L K Y L R C M R . F Grepinet and Benguin, 1986
376 384 395
C.t. celD D E E Y L R D FE .. H R A A Q F S K K E A Q F Joliff et al. 1986

408 418 427
C.s. xvnB R I V F V E R I D E Y N A N P K R V W L . . E M Luthi et al. 1990

244 254 263
T.r. CBH II L 5 C I N Y A V T Q L N L P N V A M Y L . . Q A Rouvinen et al. 1990

586 596 605
K.f. CBH I G E W E T E G Y D R E N M D L P K R T N . . E L Raynal et al. 1987

33 42 50
S.c. EG I N E S C A E F G N Q . I P G V K N . . . T 2 Y Yaguchi et al. 1983


'Abbreviations: Hen egg-white lysozyme (HEWL), Aspergillus niger a-amylase (aaml), B. fibrisolvens B-D-xylosidase (xvlBI, B. fibrisolvens endoglucanase 1 (endl), B. fibrisolvens B-D-xylanase (xvlA), B. fibrosolvens B-glucosidase (bolI), Bacillus numillus B-D-xylosidase (xvlB), Clostridium thermocellum B-D-xylanase (xvnZ), C. thermocellum cellobiohydrolase B (celB), C. thermocellum cellobiohydrolase D (celD), C. cellulolyticum endoglucanase A (EGCCA), Caldocellum saccharolyticum 8-Dxylosidase (xvnB), Cellulomonas fimi endoglucanase A (cenA), C. fimi exoglucanase A (cex), Trichoderma reesei cellobiohydrolase II (CBH II), Kluyveromyces fragilis cellobiohydrolase I (CBH I), SchizoDhvllum commune endoglucanase (EG I).









50

glutamic and aspartic acid residues and the approximate spacing were found to be conserved conserved.

Conclusions

The xvlB gene, encoding B-D-xylosidase and a-Larabinofuranosidase activities is the first of its kind to be sequenced. The xvlB gene is 1,551 bp in length and encodes a 517 amino acids protein having a predicted molecular weight of 58,000. The absence of any significant stem-loop structures or terminators in the regions between ORF1, ORF2, and ORF3 as well as the strong expression of ORF2 in E. coli even when preceded by ORF1 suggests that these three genes may constitute a xylan-degrading operon. The subcloning analysis and insertional inactivation studies demonstrate the dependence of both activities on the intact x14 gene.

The codon usage of the three ORF's is consistent with the low guanine plus cytosine content of this organism in general (Mannarelli et al. 1990).

This enzyme exhibited B-D-xylopyranosidase and a-Larabinofuranosidase activities. No additional glycosidic bond cleavage activities were detected in the xvlB gene product.

The xvlB gene displayed limited homology to other reported xylosidase sequences and must therefore be considered to be evolutionarily divergent from genes encoding similar functions from other organisms. It did









51
however exhibit partial identity with the 8-glucosidase from Kluyveromyces fraQilis which is consistent with the similarity of the substrates which these two enzymes attack.

A single gene encoding xylanase/xylosidase activities has been cloned and sequenced from Caldocellum saccharolyticum (Luthi et al. 1990). This protein, however, lacked arabinofuranosidase activity. Recently, a xylosidase/arabinofuranosidase gene was proposed to reside in a gene cluster isolated from Bacteroides ovatus (Whitehead and Hespell, 1990). All clones exhibited both activities concurrently and both activities co-purified. Additionally, an enzyme having xylosidase and arabinofuranosidase activities has been purified from Trichoderma reesei (Poutanen and Puls, 1988). No sequence, however, for the encoding gene has been reported.

Substrate ambiguity between carboxymethylcellulase and xylanase enzymes is relatively common (Flint et al. 1989). The substrate ambiguity for other xylosidase enzymes has also been reported. By employing kinetic methods on the purified, bifunctional B-xylosidase/8-glucosidase from Chaetomium trilaterale, Uziie et al. (1985) demonstrated that this enzyme possessed a single active site with dual substrate-binding regions. More recently, a neopullulanase from Bacillus stearothermophilus was cloned, sequenced, and characterized (Kuriki and Imanaka, 1989). This enzyme possessed activity against a-(1,6)-glycosidic linkages in









52

addition to the usual hydrolysis of a-(1,4)-glycosidic linkages. Mutational analysis demonstrated that a single active center was involved in the catalysis of both these linkages (Kuriki et al. 1991).

It seems reasonable to speculate that bifunctionality and substrate ambiguities among the microbial carbohydrate hydrolases is common. The celB gene encoding a "true" bifunctional cellulase has been cloned from Caldocellum saccharolyticum and sequenced (Saul et al. 1990). This enzyme exhibited both endo-glucanase and exo-glucanase activities. The endo-glucanase activity was localized to the carboxy terminal domain and the exo-glucanase activity was localized to the amino terminal domain. This protein also exhibited homology with both endo- and exo-glucanase enzymes from other organisms. The organization of separate functions to separate domains has also been demonstrated with the endo-glucanase 2 from Bacteroides succinogenes (McGavin and Forsberg, 1989). These investigators used protease treatments to demonstrate that this enzyme possessed separate substrate binding and catalytic domains.

The structural similarities between the various B- and a-linked glycosyl residues may be responsible for the apparent evolution of enzymes with broad substrate specificity.










53

Since many enzymatic activities are required to completely depolymerize xylans and cellulose, the evolution of such enzymes could represent a selective advantage in the rumen and other environments.
















Chapter IV

Mutational analysis of the xvlB gene Introduction


Carbohydrate-degrading enzymes have been studied

extensively in the microbial world and form the basis of much of what we know about the cycling of carbon in the environment (Weinstein and Albersheim, .1979). Lysozyme, an enzyme which hydrolyzes bacterial cell wall carbohydrates, was one of the first such erzymes to be uitdied extensively and as a result much is known about this enzyme's catalytic mechanism and structure (Quiocho, 1986). The cellulases have also been extensively studied as they are responsible for the cycling of the most abundant natural polymer cellulose (Knowles et al., 1987). Studies of the mechanistic properties of these enzymes have been facilitated by the use of molecular genetic techniques. Gene cloning, sequencing, and oligonucleotide-directed mutagenesis have allowed mutations to be made in a site specific manner. A kinetic study of the mutant proteins can then be done and predictions about catalytic mechanisms tested.




54









55

The active center for a neopullulanase from Bacillus stearothermoDhilus has been recently examined using mutagenesis (Kuriki et al., 1991). This enzyme exhibits dual activities against a-(1,4) and a-(1,6) glycosidic linkages. The catalytically important amino acids were tentatively identified using sequence alignment and homology searches. The putative catalytic amino acids were changed using site-directed mutagenesis and activities were examined in the resulting mutants. This approach identified that one active center containing Glu-357 and Asp-424 was responsible for both catalytic activities.

Enzymes which exhibit substrate ambiguity are interesting both from mechanistic and evolutionary perspectives. The obvious question that arises with respect to bifunctionality is does this enzyme have two separate, specialized, catalytic sites? This situation would imply that the protein has evolved from a gene fusion to perform two separate functions. Another possibility involves the presence of a single active site in which two structurally similar substrates are bound and hydrolyzed. This situation would be analogous to a case of mistaken substrate identity which proves advantagous to the organism, and has been evolutionarily conserved. Examples of multifunctional xylosidases have been reported in the literature and are discussed in the previous chapters. The apparent bifunctionality of the xvlB gene was demonstrated in chapter









56

III. A series of genetic experiments were designed to investigate the presence or absence of two catalytic or functionally separate domains on the xvlB gene that are responsible for the dual activities exhibited by this enzyme.

Materials and Methods

Medium and growth conditions, genetic methods and DNA sequencing were done as described in chapter III.

In vitro nitrous acid mutaQenesis of xvlB. A total of 80 gg of pLOI1005 which contains the xvlB gene was resuspended in 50 Al Tris-EDTA (TE) buffer (pH 8.0). Mutagenesis was initiated by the addition of 10 Al of 2.5 M sodium acetate (pH 4.3) and 50 Ml 2.0 M sodium nitrite. Exposure times were zero, thirty seconds, one, two five and ten minutes. Mutagenesis reactions were terminated by the addition of 200 Al 100% ethanol. The precipitation step was repeated twice to ensure the complete removal of the mutagenic agent. The mutagen-treated plasmids were resuspended in 80 Al TE buffer (pH 8.0). A total of 5 Al of the plasmid was transformed into competent E. coli DH5a. Serial ten-fold dilutions of the transformed cells were plated in triplicate onto Luria agar supplemented with the fluorogenic substrates. A 99% reduction in transformation by the mutagenized plasmid was observed after ten minutes of mutagenesis.









57

Localization of point mutations. The entire xylB coding region was divided into five domains based upon restriction sites (Fig. 7). Three restriction fragments, the PstI, EcoRI, and HindIII fragments were isolated from each mutant plasmid. These fragments were used to replace the corresponding fragment in the wild type gene which had been modified to construct receiving vectors for each respective restriction fragment (pLOI1051, pLOI1052, and pLOI1053) to test the functionality of individual fragments. This strategy, outlined in Fig. 8, allowed the localization of point mutations to one or more of the five domains.

Exonuclease III deletion of xylB. Plasmid pLOI1043 containing the xvlB coding region was deleted from the 3' terminal region into the coding region by exonuclease III using the "Erase-a-Base" deletion kit (Promega Corporation, Madison, WI.) according to the manufacturer's instructions. The deleted plasmids were subsequently transformed into competent E. coli DH5a and screened in the same fashion as the in vitro-generated mutant plasmids.

5' deletion analysis and lacZ' fusions. The internal PstI fragment from pLOI1005 was subcloned in the original orientation into plasmid pUC18. This in-frame fusion with lacZ results in the subsequent deletion of the 5' terminal 54 base pairs (18 amino acids).

Preparation of cell extracts. Extracts were prepared as described in chapter III.






























Figure 7. The assignment of domains to the y gene (solid bar). Restriction endonuclease fragments used to localize mutations are shown below. Abbreviations: E; EcoRI, S; sgI, P; PstI, H; HindIII.











59











II III IV V


pUC18 LacZ--



E SP H H E S P EcoRI PetI


HindIII





























Figure 8. Subcloning strategy used to localize in vitro mutations to one of five domains on the xylB gene. Abbreviations: E; EcoRI, S; SsDI, P; PstI, H; HindIII.











kilobase pairs 61
1.0 2.0 3.0




pLOI1005
E SP H H E S P pUC18

LacZ====



PstI




EcoRI




HindIII






E E P E S P pUC18 pUC18 LacZ== I I Lac Z= =
pLOI1051 pLOI1052




SP H E S P pUC18


LacZ===p
pLOI1053









62

Enzyme assays. Enzyme assays were done as described in chapter III.

Sodium dodecyvl sulfate polvacrylamide gel

electrophoresis (SDS-PAGE). SDS-PAGE gels were done as described in chapter III.

Native polvacrylamide gel electrophoresis (native

PAGE). Cell proteins were separated in non-denaturing gels by the method of Ornstein and Davis (1964). Following electrophoresis, gels were equilibrated in 50 mM sodium phosphate buffer (pH 6.8). The equilibrated gels were then overlaid with Whatman #1 filter paper soaked with a solution of 20 mg per ml of the fluorogenic substrates, 4methylumbelliferyl-B-D-xylopyranoside or 4methylumbelliferyl-a-L-arabinofuranoside in 70% ethanol. Overlays were incubated at 370C for 15 min or until activity bands were visible under long-wave UV light. Native-PAGE gels were also stained for protein as outlined before.

Western hybridization analysis of wild type and mutant proteins. Native and SDS-PAGE protein gels were electroblotted using the Trans-Blot apparatus (BioRad Laboratories, Richmond, CA.) according to the manufacturer's instructions. "Western" hybridizations were done using polyclonal antisera raised to E. coli DH5a cell extracts harboring pLOI1005 in rabbits. Protein bands were visualized using alkaline phosphatase conjugated goat anti-









63

rabbit antisera. All procedures and conditions used have been described elsewhere (Harlow and Lane, 1988).

Native agarose-xvlan Qel electrophoresis. Separating gels consisted of 0.75% agarose with or without the inclusion of 0.75% birchwood xylan. Both the gels and the running buffer were standard Tris-borate-EDTA buffer (TBE-pH

8.0). Approximately 40 gg cell protein was added per well. Proteins were electrophoresed at 75 V and 21 mA in a horizontal electrophoresis unit until the dye front reached a point 1 cm from the end of the gel. Agarose gels were either directly stained for activity (zymograms) or electroblotted and analyzed immunologically as outlined above.

Results and discussion

Deletion analysis. Exonuclease III deletions from the 3' end of xvlB resulted in the concurrent loss of both enzymatic activities in all deletions (Fig. 9a). DNA sequencing of the deletion end points identified a minimal deletion of the terminal 27 base pairs (9 amino acids) resulted in the loss of both enzymatic activities. A "TAA" termination codon in the pUC18 polylinker immediately downstream from this deletion served to define the new 3' end of this mutant gene.

A LacZ' fusion of the internal PstI fragment of xvlB resulted in the deletion of the 5' terminal 54 base pairs (18 amino acids) including the Shine-Dalgarno sequence and




























Figure 9. Deletion analysis of xyA gene. (A) Exonuclease III deletion series from the 3' end of xylB. Shaded arrow denotes direction of transcription from the pUC18 lac promoter. Solid arrow denotes direction of deletion. Underlined "taa" indicates relative location of stop codon from the 3' end of deletion. Retention or loss of respective enzyme activites is indicated by a "+" or "-". (B) LacZ' fusion of 5' end of the large internal PstI fragment of xvlB resulting in a deletion of 56 base pairs.









65




kilobase pairs
0.5 1.0 1.5



pLOI1043
SalI
SP H H E S PstI
SphI
pUC18

LacZ== deletion I taa


xylfaraf

+/+ (undeleted)


-/- (27 bp./9 aa)

-I


-I/


-I


B

SP H H E S

pUC18

ILacZ==
P H H E S




LacZ'=== (56 bp/18 aa)








66

the "ATG" initiation codon with the loss of both enzymatic activities (Fig. 9b).

In vitro mutagenesis. Twelve mutants were isolated

using selective media on the basis of reduced or abolished enzyme activities. Two mutants could not be classified due to multiple mutations and were not analyzed further. A total of ten mutants were classified as negative for both enzymatic activities against the fluorogenic substrates based upon agar plate assays. Two mutants, number six and number ten, displayed reduced but significant fluorescence on fluorogenic indicator plates (Table 6).

Localization of point mutations. Most of the point mutations were determined to reside in domains II and III (Table 6). None of the mutations were localized in the lacZ' promoter of pUC18. The generation of point mutations is consistent with the deamination activity of nitrous acid. The point mutations were clustered in domains II and III near the region proposed to be the active site of this enzyme.

DNA sequencing of in vitro mutants. The amino acid substitutions resulting from in vitro mutagenesis were deduced from the base changes as determined by DNA sequencing (Fig. 10). The point mutations were all AT to GC transitions which is consistant with the mode of action for nitrous acid mutagenesis. Two frameshift mutations were identified, in domain I and domain V. These were not










67









Table 6. Localization of point mutations by restriction fragment replacement analysis.



MUTANT PHENOTYPE PstI EcoRI HindIII LOCUS


1 -1/-2 +/+ II fs3

2 - -- -- +/+ II 3 -/- -/- -/- +/+ II

4 -I- -I- -I- -I- III

5 -/- +/+ -/- +/+ I

6 w/w4 w/w +/+ +/+ II 7 w/w w/w +/+ +/+ II

8 -/- -- +/+ V fs

9 -/- -/- -/- +/+ II 10 -/- -/- -- -/- III

11- -/- -/- III 12 III


1 denotes presence or absence of xylosidase activity.
2 denotes presence or absence of arabinofuranosidase
activity.

3 denotes a frameshift mutation.

4 denotes "weak" activity.




























Figure 10. Location and identification of point mutations in W by DNA sequencing. Numbers above sequence indicate the position of the amino acids. Amino acids in parentheses below the sequence indicates the new amino acid inserted by mutation. A "-" denotes loss of enzymatic activity. A "w" denotes weak enzymatic activity.










69


Phenotype
Region of mutation xyl/ara Mutant

180 186 192
Asp.Val.Ile.Trp.Pro.Glu.Gly.Pro.His.Leu.Tyr.Lys.Lys
(Arg) -/- G186R

197 203 209 Tyr.Leu.Leu.His.Ala.Glu.Ala.Gly.Thr.Ser.Phe.Glu.His
(Thr) -/- A203T

204 210 216 Gly.Thr.Ser.Phe.Glu.His.Ala.Ile.Ser.Val.Ala.Arg.Ser
(Val) -/- A210V

1 7 13 Met.Val.Ile.Ala.Asn.Asn.Pro.Ile.Leu.Lys.Gly.Phe.Tyr
(Leu) -/- P7L

172 178 184 Ala.Ile.Trp.Lys.Gly.Ala.Leu.Lys.Asp.Val.Ile.Trp.Pro
(Phe) w/w L178F

146 152 158 Val.Arg.Tyr.Asn.Gly.Asp.Trp.Glu.Ile.Trp.Val.Gln.Glu
(UGA) w/w W158UGA

197 203 209 Thr.Leu.Leu.His.Ala.Glu.Ala.Gly.Thr.Ser.Phe.Glu.His
(Val) -/- A203V

232 238 244 Phe.Thr.His.Arg.Asn.Leu.Gly.Lys.Asp.Tyr.Pro.Val.Cys
(Asp) -/- G238D

204 210 216 Gly.Thr.Ser.Phe.Glu.His.Ala.Ile.Ser.Val.Ala.Arg.Ser
(Thr) A210T

197 203 209 Thr.Leu.Leu.His.Ala.Glu.Ala.Gly.Th.Ser.Phe.Glu.His
(Thr) A203T

23
Ser.ILe.Cys.Arg.Lys.Gly...
-/- fs









70

studied further. Ten mutations are clustered in an area of 60 amino acids. Six of these mutations; glycine 185 to arginine 185, alanine 202 to threonine 202, alanine 202 to valine 202, alanine 209 to valine 209, glycine 237 to aspartate 237, and alanine 209 to threonine 209 all resulted in an inactive protein. Each of these mutations represent nonconservative amino acid changes that would be expected to effect the function and/or conformation of the protein. Another mutation in this cluster, leucine 178 to phenylalanine 178, resulted in a mutant having one-tenth the enzymatic activities of the wild type. This is a conservative change in that both leucine and phenylalanine are hydrophobic amino acids and have similar structural properties. It is possible that the larger aromatic group on phenylalanine is affecting substrate binding and/or catalysis. In addition, the substitution of a "UGA" stop codon for tryptophan at position 152 also resulted in a protein with one-tenth the enzymatic activities of the wild type. The most probable explaination is that the "UGA" codon is functioning as tryptophan in this E. coli strain and the efficiency of read through is very low which yields a reduced expression of the xvlB protein. The existence of such suppressor mutations has been previously demonstrated (Hirsh, 1971). One mutation, proline (7) to leucine (7) occurred in the amino terminus of the protein and resulted in a negative phenotype. The amino terminus of the xvlB









71

protein therefore plays an important in the structural and/or catalytic role.

SDS-PAGE analysis of mutant proteins. The presence of the 60,000 molecular weight monomeric subunit encoded by xlB was confirmed for the various mutants (Fig. 11). In two cases, cell extracts from recombinants harboring the proline (7) to leucine (7) and glycine (185) to arginine (185), did not contain the xvlB-encoded protein band. The insertion of the plasmid containing the proline (7) to leucine (7) mutation into a lon-negative strain of E. coli, which is deficient in serine proteases, resulted in the restoration of the xvlB-encoded protein band on SDS-PAGE gels. It seems likely that proteolysis of an improperly folded protein is responsible for the absence of this protein in cell extracts of these two mutants.

Native polyacrylamide gel comparisons of mutant and

wild tvDe proteins. Zymograms of the wild type and mutants L178F and W158UGA proteins indicated that the mutant proteins are approximately the same size as the wild type protein (Fig. 12). Western hybridizations of blotted protein bands from native-PAGE indicated that all of the expressed mutations result in proteins that have unaltered electrophoretic mobilities and subunit assemblies relative to the wild type protein (Fig. 13). This evidence suggests that the point mutations that result in expressed protein do not induce any destabilizing secondary structural



























Figure 11. SDS-PAGE analysis of wild type and mutant proteins. Lane assignments: A and J; wild type (pLOI1005), B; pUCl8, C; G186R, D; A203T, E; A210V, F; P7L, G; L178F, H; W158UGA, I; molecular weight markers, K; A203V, L; G238D, M; G238D expressed in a lon" strain of E. coli N; A210T, 0; A203T. Molecular weight marker sizes (X 103): 1; 94, 2; 67, 3; 43, 4; 30, 5; 20.








73






















IA K L M






























Figure 12. Native-PAGE comparison of W158UGA and L178F mutants proteins with the wild type (pLOI1005) stained with Coomassie blue. Lane assignments: A; mutant L178F, B; pUC18 control, C; W158UGA, D; pLOI1005.
































75




































;~





























































































I








"'


raiiiiri
_iir3 . 21' i:i
si~i i"i; x:
lr
IC
































Figure 13. Western hybridization of native-PAGE of wild type and mutant proteins. Lane assignments: A; A203T, B; pUC18, C:; P7L, D; A210V, E; L178F, F, W158UGA, G; pLOI1005.










77









78

perturbations that would result in proteolysis by the cell or a major change in tertiary structure of the proteins.

The reduced levels of the xylosidase protein in the W158UGA mutant correlates to the reduced enzymatic activities for recombinants harboring this mutant. The "UGA" termination codon can be decoded as a tryptophan at low efficiency in E. coli (Hirsh, 1971). It is likely that this is the case also for the W158UGA mutant in xylB.

Substrate-binding comparisons of mutant proteins.

Electrophoresis of the wild type and mutant proteins on native gels that contained agarose alone and agarose plus birchwood xylan indicated a differential mobility between the wild type and mutant proteins (Fig. 14). Using agarose alone no differences between electrophoretic mobilities of the respective mutant proteins and the wild type were detected using Western hybridization analysis. The inclusion of birchwood xylan (0.75 %) resulted in a change in the mobility of the proteins. Without exception, all the mutant proteins exhibited faster electrophoretic mobilities relative to the wild type protein. It is possible that the xylan is functioning as a psuedo-substrate and the point mutations have affected the relative affinities of these proteins for the substrate.
































Figure 14. Substrate binding native gel Western hybridization assays of wild type and mutant proteins. Lane assignments: A F, and G; native (pLOI1005), B; G186R, C; A203T, D; W158UGA, E; L178F, G; A203V, H; G238D, I; A210T, J; A203T. Arrow indicates the direction of protein migration. The "+" and "-" indicated the relative location of the anode and cathode, respectively.







80



















AGH I JK









81

Analysis of enzymatic activities of expressed mutant proteins. In all cases, the point mutations affected both enzymatic activities concurrently (Table 7). All mutations resulted in an expressed phenotype in which enzymatic activities were reduced or abolished. The clustering of these mutations in the 60 amino acid region (12% of the coding region) which contains the catalytic consensus sequence is evidence that the two enzymatic activities expressed by this protein are not functionally confined to separated domains. There is a dependance of function relating both enzymatic activities to this region of the protein.

The effects of substrate concentration on reaction rate of the mutant enzymes was investigated using the crude extracts as a source of protein (Table 8). Mutants A203T, A210V, L178F, G238D, and A210T exhibited an increase in apparent increase in reaction rate relative to increasing substrate concentration. Increasing substrate concentration had no effect on reaction velocity for mutants G186R and A203V, however the activities for these two mutant proteins were above that for the pUC18 background. It is possible that the lowest concentration of substrate used, 3 mM, is at saturation with respect to these two mutant proteins.










82











Table 7. Enzymatic activities of recombinants harboring point mutations on xvlB relative to the wild type protein. Clone Xyl Sp.Ac.a Ara Sp.Ac. ara/xylo Rfc pLOI1005 10.5 17.4 1.7 .46 G186R 0.1 0.1 1.0 .51 A203T 0.1 0.3 3.0 .49 A210V 0.2 0.2 1.0 .49 P7L 0.2 0.2 1.0 N/D L178P 1.2 1.8 1.5 .53 W158AUG 2.1 2.9 1.4 .46 A203V 0.2 0.2 1.0 .49 G238D 0.2 0.2 1.0 N/D A210T 0.2 0.1 0.5 .50 pUC18 0.0 0.0 a Specific activity in nmoles p-nitrophenol released per min per mg protein.
b Ratio computed after subtraction of pUC18 background values. c Determined by comparing relative migration distances of each protein verses that for the dye front on agarose/xylan native gel electrophoresis using "Western" hybridization to visualize protein bands.










83

TABLE 8. Effects of substrate concentration on 8-Dxylosidase activity for xyLB in vitro mutantsb Clone 3 mM 6 mM 9 mM KmaPc

p-nitrophenyl-B-D-xylopyranoside


wild type 18.6 24.0 25.6 4 mM G186R 0.4 0.4 0.5 2 mM A203T 0.4 0.7 0.9 11 mM A210V 0.8 0.9 1.2 2 mM L178F 1.4 1.8 3.0 18 mM A203V 0.6 0.8 0.8 3 mM G238D 0.6 1.1 1.3 7 mM A210T 0.4 0.7 0.8 12 mM


a Specific activity expressed as nmoles p-nitrophenol
released per min per mg protein.
b Does not include the frameshift mutations or those mutant
proteins that are not expressed.
c Apparent Km values determined using the direct linear
method of Cornish and Bowden.









84

Conclusions

The mutation and deletion data are consistant with the proposal that both enzymatic activities exhibited by the xvlB-encoded protein are not functionally separate but depend upon the same region of the protein for complete activity. The clustering of mutations about the consensus sequence and the 60 amino acids region is strong evidence in favor of this region being important for substrate-binding and/or catalytic activity of the protein. This region of the protein is rich in aspartic acid, glutamic acid and histidine residues which have been previously implicated in the catalytic function of related proteins such as lysozyme (Quiocho, 1986), taka-amylase from Aspergillus niQer (Matsuura et al., 1984), cellobiohydrolase II from Trichoderma reesei (Rouvinen et al., 1990), and a neopullulanase from Bacillus stearothermophilus (Kuriki et al., 1991).

The point mutations that were expressed as complete proteins did not affect either subunit assembly or the apparent size of the native protein relative to the wild type. These mutations did, however, change the protein's mobility during electrophoresis on agarose-xylan native gels. It is possible that this change in mobility is due to a reduced affinity of the various mutant proteins for xylan, which is functioning as a surrogate substrate for the









85

enzyme. There appears to be a dependence of velocity on substrate concentration for at least some of the mutants.

The mutation evidence supports the hypothesis that one active center or domain is responsible for both enzymatic activities. It does not, however, totally rule out the possibility that the protein may contain separate catalytic sites or subsites which are spacially close together. Future kinetic experiments, including the investigation of substrate competition between arabinofuranosides and xylopyranosides, will allow further definition of the catalytic regions responsible for both activities.















Chapter V

Purification and characterization of the xvlB-encoded protein

Introduction

The existance of polysaccharide-hydrolyzing enzymes having broad substrate specificities is well documented (Ward and Moo-Young, 1989). One such example includes the exoglucanase, EXG, produced by Cellulomonas fimi (Beguin, 1991). This enzyme also exhibits B-D-xylanase activity. The xylanase from Clostridium thermocellum (XYNZ) also exhibits endo-glucanase activity towards carboxymethylcellulose (Grepinet et al. 1988). In particular, several xylanase and xylosidase enzymes have been characterized which exhibit substrate ambiguity (Flint et al. 1989). A recent example is a B-D-xylosidase that was cloned from Caldocellum saccharolyticum that also exhibits endoxylanase activity (Luthi et al. 1990). Enzymes which exhibit both endoxylanase and 8-glucosidase activities have been shown to be fairly common among micoorganisms (Gilkes et al. 1991).

Substrate ambiguity among the glycohydrolases has been attributed to the similarities between the various substrates involved. Upon closer examination, this phenomenon is not totally unexpected. The B-(1,4)-xylosidic bonds of xylan and the 8-(1,4)-glycosidic bonds of cellulose 86









87

are structurally related and have quite similar molecular configurations about the 8-(1,4) bonds with respect to the hydroxyl group on the a-carbon being in the axial or equatorial positions.

Some true bifunctional cellulases have been

demonstrated which contain separate active sites for each enzymatic activity. Saul et al. (1990) isolated a cellulase from Caldocellum saccharolyticum which exhibited both endoglucanase and exo-glucanase activities. These authors used DNA sequence homology comparisons and deletion analysis to demonstrate that the endoglucanase activity was located in the carboxy terminal domain and the exoglucanase activity was located at the amino terminal domain.

In an earlier study purified a B-xylosidase from

Chaetomium trilaterale that also exhibited 8-glucosidase activity (Uziie et al. 1985). These investigators used kinetic analysis employing substrate competition and inhibitors to demonstrate that a single active site was responsible for both enzymatic activities. It was also suggested that two kinetically separate substrate binding sites may reside in the active center of this enzyme.

In the previous chapter mutational analysis

demonstrated that the two enzymatic activities encoded by xvlB were not functionally separate but both appeared to be catalytically dependent upon the same region of the protein. The proposal that a single active center is responsible for









88

both enzymatic activities in this enzyme was tested using analogous kinetic experiments including substrate inhibition and competition of the enzyme with respect to both the xylopyranosyl and arabinofuranosyl substrates.

Materials and Methods

Medium and growth conditions. Medium and growth conditions were described in chapter III.

Preparation of cell extracts. Cell extracts were prepared as described in chapter III.

Partial purification of B-xvlosidase by Dreparative

electrophoresis. Extracts containing the total cytoplasmic proteins from pLOI1005 or L178F recombinant clones were used as a source of xylosidase. Proteins were fractionated in an 8% native polyacrylamide gel in a BioRad-Prep Cell preparative electrophoresis system. Gel and buffer formulations were: separating gel buffer; 240 mM Tris (pH

8.48), stacking gel buffer; 40 mM Tris (pH 6.9), lower tank buffer; 63 mM Tris/50 mM HCL (pH 7.5), upper tank buffer; 38 mM Tris/ 40 mM glycine (pH 8.9). A total of 50 mg/ml protein was loaded onto the gel. Electrophoresis was done at constant power of 31 W. Starting conditions were 250 V and 40 mA. Protein elution was monitored at 280 nm.

Fractions were collected and assayed for xylosidase

activity as described below. The most active fractions were pooled and concentrated using an Amicon Centriprep concentrator (Amicon Division, Danvers, MA). Xylosidase in










89

the concentrate was then precipitated by the addition of solid ammonium sulfate to 70% saturation. The enzyme was stored as an ammonium sulfate pellet at 50C until needed. No loss of enzymatic activity was detected after storage for one week under these conditions. Pellets were resuspended in 5 mM phosphate (pH 7.0) containing 10 mM 8mercaptoethanol immediately prior to use.

Hydrophobic interaction chromatography. The xylosidase-containing pellet from preparative electrophoresis was resuspended in 1 ml of 1.7 M (NH4)2SO4 in

5 mM phosphate buffer (pH 6.8) and loaded onto a 2.0 x 3.0 cm Pharmacia XK 16/20 chromatography column (Pharmacia LKB, Uppsala, Sweden) packed with Toyopearl "TSK-Gel" hydrophobic gel (Supelco, Inc., Bellefonte, PA.). The column was equilibrated with 1.7 M (NH4)2S04 in 5 mM phosphate buffer (pH 6.8) prior to the addition of sample. Xylosidase was eluted using a linear negative salt gradient starting with

1.7 M (NH4)2SO4 down to zero in a total volume of 200 ml. Fractions were collected in 3 ml volumes and analyzed as described below.

Enzyme assays. Xylosidase activity in each fraction was assayed using p-nitrophenyl-B-D-xylopyranoside (p-NP-X) at a final concentration of 2.5 mM unless otherwise noted and in 50 mM phosphate buffer (pH 6.8) at 370C. Assays were done in a total volume of 1 ml and allowed to continue until the yellow color indicating enzyme activity was detected.









90

Assays were then terminated by the addition of 2 ml 0.5 M carbonate. The p-nitrophenol released by hydrolysis was measured spectrophotometrically at 405 nm. The liberation of 1 nmole of p-nitrophenol results in an increase in absorbance of 0.0184 at 405 nm. Samples were assayed for aL-arabinofuranosidase activity under the same conditions using p-nitrophenyl-a-L-arabinofuranoside (p-NP-A) as a substrate.

Electrophoretic analysis of proteins. SDS-PAGE,

native-PAGE, activity stains, and "Western" hybridizations were done as described in chapter III.

Optimum activity pH. The optimum pH for both

activities was determined in duplicate using citrate (40 mM)-sodium phosphate (80 mM) buffer in the pH range of 3.0 to 7.4. Tricine (50 mM) was used for the 7.5 to 8.5 pH range and Bicine (50 mM) for the 8.6 to 9.0 pH range. All activities were determined using a 6 mM final concentration of p-NP-X or p-NP-A at 370C.

Thermal optimum and thermal inactivation

determinations. Thermal stability was determined by incubating the enzyme for 30 min at 10, 25, 35, 45, 55, and 650C. After 30 min the enzyme was placed in ice for 10 min prior to assaying both enzymatic activities. The optimum temperature for activity was determined by assaying the enzyme for both activities at 10, 25, 35, 45, 55, and 600C for 10 min. Assays were done in duplicate at 370C using 6




Full Text
Figure 3. Southern hybridizations of chromosomal DNA from B.
fibrisolvens GS113 (8 hr exposure) and E. coli DH5a (8 hr
exposure) and plasmid DNA from pLOI1005 (2 hr exposure).
Lanes 1, 2, and 3 contain PstI digests from E. coli. B.
fibrisolvens. and pLOI1005, respectively, probed with the
internal PstI fragment from pLOHOOS. Lanes 4, 5, and 6 are
Sau3A digests of E. coli DH5a, B. fibrisolvens GS113 and
pLOHOOS which were probed with the internal Sau3A fragment
from pLOI1005. The additional bands in lane 6 represent
larger, incomplete restriction endonuclease digestion
products.


Although the xvlB gene did not exhibit a high degree of
amino acid identity with other xylan-degrading enzymes or
glycohydrolases, a conserved sequence was identified with
significant identity to the active site region of hen egg
white lysozyme and Aspergillus niaer glucoamylase. No
predictable stem loop structures or sequences resembling
terminators were found on the xvlB gene fragment and this
gene appears to be part of an operon. In vitro analysis of
xvlB mutants demonstrated structural and functional
relationships between the two enzyme activities. All point
mutations investigated in xylB resulted in the reduction or
loss of both enzymatic activities. Most of these mutations
were clustered in a region near the proposed active site.
The point mutations decreased the apparent affinity of the
enzyme for xylan. The partially purified xylB-encoded
protein exhibited thermal inactivation kinetics and
temperature optima that were essentially the same for both
enzymatic activities. The pH optimum for both activities
was 6.0. However, the arabinofuranosidase activity
exhibited a broader pH range, retaining 90% of maximal
activity up to pH 9.0. The apparent Km for p-nitrophenyl-6-
D-xylopyranoside and p-nitrophenyl-a-L-arabinofuranoside
were 3.7 mM and 1.8 mM respectively. Substrate competition
experiments corroborated the genetic evidence and
demonstrated that the same active center was responsible for
both enzymatic activities of the xvlB-encoded protein.
x


13
D-xylosidase component of the microbial xylanolytic system.
Early mechanistic studies of the B. pumilus xylosidase
indicated the enzyme contained several thiol groups and at
least one of which is involved in the catalysis (Saman et
al. 1975). Panbangred et al. (1983) first cloned the genes
for B-xylanase and B-D-xylosidase from Bacillus pumilus IPO.
Both cloned proteins were expressed in Escherichia coli from
a hybrid plasmid and were immunologically and chemically
identical to those of B. pumilus. The cloned genes from B.
pumilus IPO were later sequenced by Moriyama et al. (1987).
The gene for B-D-xylosidase was localized to a 1617 base
pairs open reading frame encoding a deduced 62,607d protein.
The N-terminal amino acid sequence agreed with that
predicted from the DNA sequence and that obtained from the
purified enzyme. It is interesting to note that the B-
xylanase gene from the same organism was located 4,600 base
pairs downstream from the 3'-end of the 6-D-xylosidase. The
B. pumilis enzyme was not reported to exhibit any additional
enzymatic activities.
More recently, two xylosidase genes were cloned and
sequenced from the obligately anaerobic, thermophilic
organism Caldocellum saccharolvticum (Luthi et al. 1990).
The protein encoded by one of these xylosidase genes was
found to possess xylanase activity in addition to the
expected xylosidase activity. These genes were also found
to reside in close proximity to each other and to a gene


Figure 29.
Double recipricol plots of xylosidase and
arabinofuranosidase activities for the L178F
mutant protein: open squares;
arabinofuranosidase activity, closed squares
xylosidase activity.


90
Assays were then terminated by the addition of 2 ml 0.5 M
carbonate. The p-nitrophenol released by hydrolysis was
measured spectrophotometrically at 405 nm. The liberation
of 1 nmole of p-nitrophenol results in an increase in
absorbance of 0.0184 at 405 nm. Samples were assayed for a-
L-arabinofuranosidase activity under the same conditions
using p-nitrophenyl-a-L-arabinofuranoside (p-NP-A) as a
substrate.
Electrophoretic analysis of proteins. SDS-PAGE,
native-PAGE, activity stains, and "Western" hybridizations
were done as described in chapter III.
Optimum activity pH. The optimum pH for both
activities was determined in duplicate using citrate (40
mM)-sodium phosphate (80 mM) buffer in the pH range of 3.0
to 7.4. Tricine (50 mM) was used for the 7.5 to 8.5 pH
range and Bicine (50 mM) for the 8.6 to 9.0 pH range. All
activities were determined using a 6 mM final concentration
of p-NP-X or p-NP-A at 37C.
Thermal optimum and thermal inactivation
determinations. Thermal stability was determined by
incubating the enzyme for 30 min at 10, 25, 35, 45, 55, and
65C. After 30 rain the enzyme was placed in ice for 10 min
prior to assaying both enzymatic activities. The optimum
temperature for activity was determined by assaying the
enzyme for both activities at 10, 25, 35, 45, 55, and 60C
for 10 min. Assays were done in duplicate at 37C using 6


147
MacKenzie, C. R., R. C. A. Yang, G. B. Patal, D. Bilous, and
S. A. Nurang. 1989. Identification of three distinct C.
thermocellum xylanase genes by molecular cloning. Arch.
Microbiol. 152:377-381.
MacNeil, N. I. 1984. The contribution of the large intestine
to energy supplies in man. Amer. J. Clin. Nutr. 39:338-346.
Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982.
Molecular cloning: a laboratory manual. Cold Spring Harbor
Laboratory, Cold Spring Harbor, N. Y.
Mannarelli, B. M., R. J. Stack, D. Lee, and L. Ericsson.
1990. Taxonomic relatedness of Butvrivibrio. Lachnospira.
Roseburia and Eubacterium species as determined by DNA
hybridization and extracellular-polysaccharide analysis.
Int. J. Syst. Bacteriol. 40:535-544.
Mannarelli, B. M., S. Evans, and D. Lee. 1990. Cloning,
sequencing and expression of a xylanase gene from the
anaerobic rumen bacterium Butvrivibrio fibrisolvens. J.
Bacteriol. 172:4247-4254.
Mannerelli, B. M. 1988. Deoxyribonucleic acid relatedness
among strains of the species Butyrivibrio fibrisolvens. Int.
J. Syst. Bacteriol. 23:308-315.
McGavin, M., and C. W. Forsberg. 1989. Catalytic and
substrate-binding domains of endoglucanase 2 from
Bacteroides succinoqenes. J. Bacteriol. 171:3310-3315.
Morag. E., E. A. Bayer, and R. Lamed. 1990. Relationship of
cellulosomal and noncellulosomal xylanases of Clostridium
thermocellum to cellulose-degrading enzymes. J. Bacteriol.
172: 6098-6105.
Moriyama, H., E. Fufusaki, J. Cabrera Crespo, A. Shinmyo.,
and H. Okada. 1987. Structure and expression of genes coding
for xylan-degrading enzymes of Bacillus pumilus. Eur. J.
Biochem. 166:539-545.
Morosoli, R., C. Roy, and M. Yaguchi. 1986. Isolation and
partial primary sequence of a xylanase from the yeast
Cryptococcus albidus. Biochim. Biophys. Acta 870:473-478.
Nakanishi, K., T. Yasui, and T. Kobayashi. 1971. Inducers
for the xylanase production by Streptomvces sp. J. Ferment.
Technol. 54:801-807.
Nishitani, K., and D. J. Nevins. 1991. Glucuronoxylan
xylanohydrolase: A unique xylanase with the requirement for
appendant glucuronosyl units. J. Biol. Chem. 266:6539-6543.


46
TABLE 4. Comparison of the translated amino acid sequences of the three B.
fibrisolvens ORFs in pLOI 1001 with those of selected proteins.
% Similarity (% Identity)
Organism (gene)
ORF 1
ORF 2
xylB)
ORF 3
Reference
B. fibrisolvens x^
cnh)
40 (16)
41 (16)
41 (13)
Mannarelli
et al. 1990
B. fibrisolvens (endl)
16 (8)
14 (6)
20 (11)
Berger
et al. 1989
B. Dumilus (xvnA)
46 (21)
44 (16)
12 (8)
Fukusaki
et al. 1984
B. Dumilus (xvnB)
39 (16)
28 (21)
38 (17)
Moriyama
et al. 1987
B. subtilis (xvnA)
48 (21)
44 (16)
12 (6)
Paice
et al. 1986
Caldocellum
saccharolvticus (x\
rnB)
20 (14)
25 (12)
21 (12)
Luthi
et al. 1990
C. saccharolvticus
(xvnA / xvnB)
16 (8)
20 (9)
19 (10)
Luthi
et al. 1990
C. saccharolvticus
(xynC)
16 (8)
22 (11)
31 (3)
Luthi
et al. 1990
Clostridium
thermocellum
18 (7)
40 (16)
15 (8)
Grepinet
et al. 1988
(xynZ)
C. thermocellum
icelA)
45 (22)
41 (18)
39 (15)
Benguin
et al. 1985
C. thermocellum
(celBi
44 (20)
44 (18)
43 (19)
Grepinet
et al. 1986
C. thermocellum
(celD1
45 (16)
39 (15)
43 (18)
Joliff
et al. 1986
AsDeraillus
nicer
(a-amylase)
40 (16)
41 (18)
43 (19)
Boel
et al. 1984
Kluvveromvces
fraailis
(B-glucosidase)
52 (31)
44 (20)
46 (22)
Raynal et al. 1987
ORF1
100
44 (19)
45 (19)
This study
ORF2
44 (19)
100
42 (14)
This study
ORF3
45 (19)
42 (14)
100
This study


Figure 1. Restriction endonuclease digestion maps of
plasmids pLOHOOl and pLOI1005 that express 6-D-xylosidase
activity in E. coli DH5a.


Figure 21. pH profiles for xylosidase (closed circles) and
arabinofuranosidase (open circles) activities.


136
the shorter hydrolytic products of the endoxylanase
activity. No data is yet available concerning active site
studies of this enzyme as to whether or not one or two
catalytic centers are responsible for this phenomenon.
The present study is unique in that it examines the
substrate ambiguity of the B. fibrisolvens enzyme using
both genetic and biochemical techniques. This work
represents a complete study of the enzyme and has elucidated
a number of important aspects of the B-D-xylosidase from B.
fibrisolvens. (A) The gene that encodes this enzyme in B.
fibrisolvens. xvlB is present as a single copy in the
chromosome (Fig. 3). Any additional xylosidase genes
present in this organism must share limited homology with
xvlB. (B) The DNA sequence is 1,551 bp in length, encodes
517 amino acids and is located between two additional large
open reading frames (ORF's) each in excess of 1,000 bp (Fig.
5). The size of this enzyme is essentially the same as the
B. pumilus xylosidase. (C) No stem loop or rho-independant
teminators were identified between any of the three ORF's
and the expression data indicates that this region does not
function as a terminator in E. coli (Table 2). This result
suggests that these three ORF's represent part of a xylan-
degrading operon in B. fibrisolvens. The subcloning and
expression data also demonstrate that the single xvlB gene
is responsible for both the xylosidase and
arabinofuranosidase activities (Fig. 4b). (D) The codon


63
rabbit antisera. All procedures and conditions used have
been described elsewhere (Harlow and Lane, 1988) .
Native aoarose-xvlan gel electrophoresis. Separating
gels consisted of 0.75% agarose with or without the
inclusion of 0.75% birchwood xylan. Both the gels and the
running buffer were standard Txis-borate-EDTA buffer (TBE-pH
8.0). Approximately 40 nq cell protein was added per well.
Proteins were electrophoresed at 75 V and 21 mA in a
horizontal electrophoresis unit until the dye front reached
a point 1 cm from the end of the gel. Agarose gels were
either directly stained for activity (zymograms) or
electroblotted and analyzed immunologically as outlined
above.
Results and discussion
Deletion analysis. Exonuclease III deletions from the
3' end of xvlB resulted in the concurrent loss of both
enzymatic activities in all deletions (Fig. 9a). DNA
sequencing of the deletion end points identified a minimal
deletion of the terminal 27 base pairs (9 amino acids)
resulted in the loss of both enzymatic activities. A "TAA"
termination codon in the pUC18 polylinker immediately
downstream from this deletion served to define the new 3'
end of this mutant gene.
A LacZ1 fusion of the internal PstI fragment of xvlB
resulted in the deletion of the 5' terminal 54 base pairs
(18 amino acids) including the Shine-Dalgarno sequence and


37
TABLE 1. (continued)
Frequency (mol %> Codon Usage
Amino Acid
Pro
Thr
Ala
Tyr
His
Gin
Codon
B.
ORF 1
fibrisolvens E.
ORF 2 ORF 1
coli
B. fibrisolvens
xvlA
CCT
1.4
1.7
1.8
0.5
1.2
CCC
0.2
-0-
-0-
0.3
-0-
CCA
2.3
3.5
2.8
0.7
2.4
CCG
-0-
0.6
0.3
2.5
-0-
ACT
0.7
1.9
2.3
1.1
0.7
ACC
0.7
1.4
0.5
2.4
1.0
ACA
2.7
1.7
3.1
0.3
5.1
ACG
0.5
-0-
0.3
0.8
0.5
GCT
2.7
2.1
2.8
2.6
1.5
GCC
1.1
0.4
0.8
2.2
1.2
GCA
2.0
2.3
2.0
2.3
4.1
GCG
0.5
0.2
0.2
3.2
-0-
TAT
4.3
3.5
2.8
1.0
3.9
TAC
0.2
1.2
2.5
1.5
3.4
CAT
0.9
1.4
2.5
0.7
1.5
CAC
0.5
0.6
in
O
1.2
0.2
CAA
0.7
1.0
1.3
1.0
0.5
CAG
1.6
0.6
1.5
3.2
2.2
AAT
2.0
4.2
3.8
1.0
5.4
AAC
0.5
1.1
1.3
2.8
2.7
Asn


-Nitrophenol Released (nmoles)
93
250
200
0 50 100 150 200
Elution Volume (ml)
250


TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS
LIST OF TABLES V
LIST OF FIGURES vi
ABSTRACT ix
CHAPTERS
IGENERAL INTRODUCTION 1
IIREVIEW OF THE LITERATURE 5
IIICLONING, SEQUENCING, AND SEQUENCE
ANALYSIS OF THE XYLOSIDASE GENE FROM
BUTYRIVIBRIO FIBROSOLVENS 16
Introduction 16
Materials and Methods 17
Results and Discussion 20
Conclusions 50
IVMUTATIONAL ANALYSIS OF THE XVIB GENE 54
Introduction 54
Materials and Methods 56
Results and Discussion 63
Conclusions 84
VPARTIAL PURIFICATION AND CHARACTERIZATION
OF THE WILD TYPE GENE PRODUCT 86
Introduction 86
Materials and Methods 88
Results and Discussion 91
Conclusions 115
iii


133
"induced-fit" model of enzyme catalysis proposes that the
three-dimensional conformation of an enzyme will change as
it binds to the substrate to allow closer contacts to be
made between the substrate and the catalytically important
functional groups (Koshland, 1966). Increasing the
hydrophobic interactions in the core or active center of the
protein might also decrease the ability of the protein to
change conformation upon binding substrate. This would lead
to a loss of catalytic efficiency and possibly affect
substrate binding.
A more general explanation of the effects of the L178F
mutation would be that the protein is simply folded
incorrectly and this results in a less accessible active
center which is reflected in the increase in Km for p-NPX
and p-NPA in this mutant protein.
The addition of a large aromatic R-group by the
substitution of phenylalanine for leucine could also
introduce a steric hindrance factor which prevents optimal
enzyme-substrate interactions and therefore decreases the
affinity and/or catalytic efficiency of the enzyme. This
assumes, however, that the amino acid substitution has
occurred in an area within or near the active site.


45
1 2 3 4 5 6 7


kilobase pairs
24
0.0
1.0 2.0 3.0
P S
H
Xvl
Araf
H
H
H
H
S
S
Coding Region


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
GENETIC AND BIOCHEMICAL CHARACTERIZATION OF B-D-XYLOSIDASE
FROM BUTYRIVIBRIO FIBRISOLVENS
By
Eric Andrew Utt
December 1991
Chairman: Lonnie 0. Ingram
Major Department: Microbiology and Cell Science
The gene for B-D-xylosidase from the rumen bacterium
Butyrivibrio fibrisolvens encodes a protein that exhibits
hydrolytic activity against B-D-xylopyranosides and a-L-
arabinofuranosides. This gene, xvlB. was cloned into E.
coli as a 4.2 kilobase pairs (kbp) insert in pUC18 and
sequenced in both directions. The xvlB gene is present as a
single copy on the B. fibrisolvens chromosome and consists
of a 1,551 base pair (bp) open reading frame (ORF) which
encodes a protein of 517 amino acids. Insertion of a 10 bp
linker into the coding region resulted in a frameshift that
abolished both activities. Deletions from the 3' end and
the 5' end of xvlB also resulted in inactive proteins. SDS-
PAGE analysis of cytoplasmic extracts from recombinant E.
coli clones harboring xvlB confirmed the presence of a new
protein with an apparent molecular weight of 60,000.

IX


Figure 24. Competitive inhibition of arabinofuranosidase
activity by fi-D-methylxylopyranoside: closed circles; no
inhibitor, open circles; 1 mM inhibitor, closed squares; 2 mM
inhibitor.


Figure 18. Native-PAGE analysis of the -D-xylosidase. A;
Coomassie blue stained protein bands, B; Western hybridization
of electroblotted native-PAGE gel, C; activity stain
(xylosidase activity/arabinofuranosidase activity). Arrows
indicate approximate region where the apparent dimeric and
monomeric forms of the enzyme would be located.


95
Elution Volume


31
1 AATTGTGGATGCACATATGAAAAGCTGATTTATGCTTATAAGGCAGGTCTTGTCAAGGAA 60
NCGCTYEKLIYAYKAGLVKE
61 GAGACCATCGATGAGGCTGTTACTCGACTTATGGAAATCAGACTTCGTCTAGGTACTATT 120
ETIDEAVTRLMEIRLRLGTI
121 CCAGAGAGAAAGAGTAAGTATGATGATATCCCATATGAAGTGGTCGAATGCAAAGAGCAT 180
PERKSKYDDIPYEVVECKEH
181 ATCAAACTTGCTCTTGACGCTGCAAAGGATAGCTTTGTCCTTTTGAAGAATGATGGTTTA 240
I KLALDA AKD S FVLLKND GL
241 CTTCCACTGAATAAAAAGGATTATAAATCTATTGCTGTTATTGGACCCAACTCTGATTCA 300
LPLNKKDYKSIAVIGPNSDS
301 AGAAGAGCTTTAATTGGAAATTATGAGGGCCTTTCTTCAGAGTATATTACAGTTTTAGAG 360
RRALIGNYEGLSSEYITVLE
361 GGGATTCGTCAGGTTGTCGGTGATGATATTAGATTATTCCACGCTGAGGGCACTCATCTT 420
GIRQVVGDDIRLFHAEGTHL
421 TGGAAGGATAGAATTCACGTAATCAGTGAGCCAAAAGATGGATTTGCCGAGGCTAAAATC 480
WKDRIHVISEPKDGFAEAKI
481 GTGGCAGAGCATTCAGATTTAGTTGTGATGTGTCTTGGACTTGACGCATCAATCGAAGGA 540
VAEHSDLVVMCLGLDAS IEG
541 GAAGAAGGAGACGAGGGTAATCAGTTCGGTAGCGGAGACAAGCCTGGATTAAAGCTTACA 600
EEGDEGNQFGSGDKPGLKLT
601 GGTTGTCAGCAAGAGCTACTTGAGGAAATTGCCAAAATCGGCAAGCCTGTTGTACTTCTT 660
GCQQELLEE IAKIGKPVVLL
661 GTGCTTTCAGGTTCTGCTCTTGATTTATCATGGGCGCAGGAATCTAATAACGTAAATGCG 720
VLSGSALDLSWAQESNNVNA
721 ATAATGCAGTGCTGGTATCCAGGCGCAAGAGGTGGACGTGCTATTGCAGAGGTTTTATTT 780
IMQCWYPGARGGRAIAEVLF
781 GGCAAGGCCAGTCCAGGCGGTAAAATGCCTCTTACATTTTATGCCTCAGATGATGACCTT 840
GKASPGGKMPLTFYASDDDL
841 CCTGATTTTTCTGATTATTCAATGGAAAATAGGACATACAGATATTTCAAGGGCACACCA 900
PDFSDYSMENRTYRYFKGTP
901 CTTTATCCATTTGGTTATGGACTAGGTTATTCTAAAATTGATTATCTATTTGCTTCTATT 960
LYPFGYGLGYSKIDYLFAS I
961 GATAAAGATAAGGGAGCAATTGGTGATACATTCAAGCTAAAGGTAGATGTTAAAAATACC 1020
DKDKGAIGDTFKLKVDVKNT
1021 GGTAAGTATACACAGCATGAGGCTGTTCAAGTATATGTAACGGACCTTGAGGCAACGACA 1080
GKYTQHEAVQVYVTDLEATT
1081 AGAGTGCCTATTAGAAGCCTTAGAAAGGTTAAATGTCTAGAGCTTGAGCCTGGTGAAACA 1140
RVPIRSLRKVKCLELEPGET
1141 AAAGAGGTTGAATTTACCCTTTTTGCAAGAGATTTTGCCATTATTGATGAAAGGGGAAAA 1200
KEVEFTLFARDFAI IDERGK
1201 TGTATCATAGAGCCAGGCAAGTTTAAGATTTCTATTGGGGGACAACAGCCAGACGATAGA 1260
Cl IEPGKFKISIGGQQPDDR


Figure 19. Thermal inactivation profile of xylosidase (closed
circle) and arabinofuranosidase (open circle) activities.


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.
CLjl to. Ojia.
Charles M. Allen
Professor of Biochemistry and
Molecular Biology
This dissertation was submitted to the Graduate Faculty
of the College of Agriculture and to the Graduate School and
was accepted as partial fulfillment of the requirements for
the degree of Doctor of Philosophy.
December 1991
of.
gfjicu]
Dean,
lege of Agriculture
Dean, Graduate School


The substrate competition experiments, using model
substrates, unequivicaly demonstrated that xylopyranoside
and arabinofuranoside are competitive inhibitors of each
other.


41
subclone, pLOI1043, exhibited a four-fold decrease in both
activities but retained a similar ratio of
arabinofuranosidase to xylosidase activities. An additional
subclone, pLOI1050, contained two SspI fragments each
harboring xvlB oriented with the direction of transcription
from the lac promoter. This subclone exhibited a three-fold
increase in both enzymatic activities with respect to the
single insertion (pLOI1043) but less than the original
clone. Again the ratio of the two activities remained
essentially the same as the wild type (pLOHOOS) The
results of these experiments demonstrate the dependence of
both enzymatic activities on 0RF2.
Presence of other alvcosidic activities. The presence
of additional hydrolytic activities was examined in the
xvlB-encoded protein using various ortho- and para-
nitrophenyl glycosidic substrates (Table 3). No additional
activity above the background levels was detected with 12
other pentose and hexose derivatives. Low levels of
activity was detected against the o-NP-B-D-fucopyranoside.
This may not be significant since the activity represents
less than 5% of the activity against the B-D-xylopyranoside.
A 19-fold higher activity was detected against o-
nitrophenyl-6-D-xylopyranoside relative to the para-
substituted derivative. This phenomenon is analogous to
B-galactosidase from E. coli. The para- and ortho-
substituted substrates are known to have different rates for


146
Kersters-Hilderson, H., F. G. Loontiens, M. Claeyssens, and
C. K. De Bruyne. 1969. Partial purification and properties
of an induced B-D-xylosidase of Bacillus pumilus 12. Eur. J.
Biochem. 7:434-441.
Knowles, J. K. C., P. Lehtovaara, M. Murray, and M. L.
Sinnot. 1988. Stereochemical course of action of the
cellobioside hydrolases I and II of Trichoderma reesei. J.
Chem. Soc. Chem. Commun. 1988:1401-1402.
Knowles, J., P. Lehtovaara, and T. Teeri. 1987. Cellulase
families and their genes. Trends Biotechnol. 5:255-261.
Koshland, D. E., Jr., G. Nemethy, and D. Filmer. 1966.
Comparison of experimental binding data and theoretical
models in proteins containing subunits. Biochem. 5:365-376.
Kuriki, T., H. Takata, S. Okada, and T. Imanaka. 1991.
Analysis of the active center of Bacillus staerothermophilus
neopullulanase. J. Bacteriol. 173:6147-6152.
Laemmli, U. K. 1970. Cleavage of structural proteins during
the assembly of the head of bacteriophage T4. Nature
227:680-685.
Lee, S. F., and C. W. Forsberg. 1987. Purification and
characterization of an a-L-arabinofuranosidase from
Clostridium acetobutylicum ATCC 824. Can. J. Microbiol.
33:1011-1016.
Lee, S. F., C. W. Forsberg, and J. B. Rattray. 1987.
Purification and characterization of two endoxylaases from
Clostridium acetobutylicum ATCC 824. Appl. Environ
Micriobiol. 53:644-650.
Luthi, E., D. R. Love, J. McAnulty, C. Wallace, P. A.
Caughey., D. Saul, and P. Bergquist. 1990. Cloning, sequence
analysis, and expression of genes encoding xylan-degrading
enzymes from the thermophile Caldocellum saccharolvticum.
Appl. Environ. Microbiol. 56:1017-1024.
Lynch, J. M. 1987. Utilization of lignocellulosic wastes. In
Journal of Applied Bacteriology Symposium Supplement, 71S-
83S.
Mackenzie, C. R., D. Bilous, H. Schneider, and K. G.
Johnson. 1987. Induction of cellulolytic and xylanlolytic
enzyme systems in Streptomvces spp. Appl. Environ.
Microbiol. 53:2835-2839.


140
less of proton donating potential from involved hisidinyl
residues. This explaination may also be invoked to support
the possbility that separate amino acids and/or mechanisms
are involved in the two separate activities.
Also relating to substrate effects is the concept of
substrate-assisted catalysis (Carter and Wells, 1990).
Substrate-assisted catalysis was first demonstrated with
proteolytic enzymes. In the enzyme subtilisin, a serine
endopeptidase, it was demonstrated that a histidine in the
protein substrate could replace a catalytic histidine in the
active site of subtilisin that had been mutated to an
alanine by site-directed mutagenesis. It is possible that a
glycosyl carboxylate could serve as an alternate proton
acceptor under certain conditions.
The kinetic constants for the two activities were
comparable to other xylosidase and arabinofuranosidase
enzmes that have been reported. The catalytic efficiency or
Kcat for this enzyme is not as high as that reported for
other xylosidase or arabinofuranosidase enzymes. Since both
arabinofuranosidase and xylopyranosidase are substrates that
are indigenous to hemicellulose, this protein appears to
have saccrificed catalytic efficiency for a broader
substrate specificity. This may represent an evolutionary
adaptation to the highly competitive rumen environment.


67
Table 6. Localization of point mutations by restriction
fragment replacement analysis.
MUTANT
PHENOTYPE
PstI
EcoRI
Hindlll
LOCUS
1
-V-2
-/-
-/-
+/ +
II fs3
2
-/-
-/-
-/-
+/ +
II
3
-/-
-/-
-/-
+/ +
II
4
-/-
-/-
-/-
-/-
III
5
-/-
+/+
-/-
+/ +
I
6
w/w4
w/w
+/ +
+/ +
II
7
w/w
w/w
+/ +
+/ +
II
8
-/-
-/-
-/-
+/ +
V fs
9
-/-
-/-
-/-
+/ +
II
10
-/-
-/-
-/-
-/-
III
11
-/-
-/-
-/-
-/-
III
12
-/-
-/-
-/-
-/-
III
1 denotes presence or absence of xylosidase activity.
2 denotes presence or absence of arabinofuranosidase
activity.
3 denotes a frameshift mutation.
4 denotes "weak" activity.


CHAPTERS
Page
VI PARTIAL PURIFICATION AND CHARACTERIZATION OF
THE L178F MUTATION 120
Introduction 120
Materials and Methods 121
Results and Discussion 121
Conclusions 128
VII SUMMARY AND GENERAL CONCLUSIONS 13 5
LITERATURE CITED 142
BIOGRAPHICAL SKETCH 151
iv


Figure 15. Elution profile of the xvlB gene product during
preparative electophoresis on BioRad Prep Cell system.


LIST OF TABLES
Table Page
1.Comparison of codon usage frequency for
the three B. fibrisolvens ORF's 36
2.Expression of enzyme activities in
recombinant E. coli harboring xvlB 40
3. Hydrolysis of different nitrophenyl-
substituted glycosides by the xylB gene
product 42
4. Comparison of the translated amino acid
sequences of the three B. fibrisolvens ORF's
in pLOHOOl with those of selected
proteins 46
5. Amino acid sequence alignment of conserved
regions 48
6. Localization of point mutations by
restriction fragment replacement analysis 67
7. Enzymatic activities of recombinants
harboring point mutations in xvlB
relative to the wild type protein 82
8. Xylosidase activity of in vitro mutations with
varying substrate concentrations 83
9. Purification scheme of the xvlB-encoded protein
from E. coli DH5a 97
v


Figure 8. Subcloning strategy used to localize in vitro
mutations to one of five domains on the xvlB gene.
Abbreviations: E; EcoRI. S; SspI. P; PstI. H; Hindlll.


52
addition to the usual hydrolysis of a-(1,4)-glycosidic
linkages. Mutational analysis demonstrated that a single
active center was involved in the catalysis of both these
linkages (Kuriki et al. 1991).
It seems reasonable to speculate that bifunctionality
and substrate ambiguities among the microbial carbohydrate
hydrolases is common. The celB gene encoding a "true"
bifunctional cellulase has been cloned from Caldocellum
saccharolvticum and sequenced (Saul et al. 1990). This
enzyme exhibited both endo-glucanase and exo-glucanase
activities. The endo-glucanase activity was localized to
the carboxy terminal domain and the exo-glucanase activity
was localized to the amino terminal domain. This protein
also exhibited homology with both endo- and exo-glucanase
enzymes from other organisms. The organization of separate
functions to separate domains has also been demonstrated
with the endo-glucanase 2 from Bacteroides succinoaenes
(McGavin and Forsberg, 1989). These investigators used
protease treatments to demonstrate that this enzyme
possessed separate substrate binding and catalytic domains.
The structural similarities between the various 6- and
a-linked glycosyl residues may be responsible for the
apparent evolution of enzymes with broad substrate
specificity.


139
proceeds by a mechanism different from that of xylosidase
activity. This phenomenon, involving separate hydrolytic
mechanisms, was shown to exist with the Bacillus pumilus
xylosidase (Kasumi et al. 1987) These investigators
demonstrated kinetically that the enzyme was able to
catalyze the hydrolysis of a- and B-D-xylosylfluoride by a
mechanism entirely different from that for the hydrolysis of
p-nitrophenyl-B-D-xylopyranoside. It was postulated that
separate catalytic groups were responsible for the two
mechanisms. In a more recent study Kuriki et al. (1991)
demonstrated that a neopullulanase from Bacillus
stearothermophilus. exhibited separate a-1,4 and a-1,6 bond
cleavage activities. These investigators used site-directed
mutagenesis to demonstrate that the two activities were
catalyzed by separate amino acids in the same active center.
It is possible that an analogous process occurs in the
xvlB-encoded protein. A tyrosinyl residue (pKa 10.0) could
serve as a general acid and donate a proton at the pH range
where the arabinofuranosidase activity is present. There
are five tyrosine residues in the proposed active center
which may be involved.
Another explanation involves the relative stabilities
of the two substrates at elevated pH. If the
arabinofuranoside is less stable at higher pH values
relative to the xylopyranoside, it follows hydrolysis will
proceed more easily for the former and therefore requires


81
Analysis of enzymatic activities of expressed mutant
proteins. In all cases, the point mutations affected both
enzymatic activities concurrently (Table 7). All mutations
resulted in an expressed phenotype in which enzymatic
activities were reduced or abolished. The clustering of
these mutations in the 60 amino acid region (12% of the
coding region) which contains the catalytic consensus
sequence is evidence that the two enzymatic activities
expressed by this protein are not functionally confined to
separated domains. There is a dependance of function
relating both enzymatic activities to this region of the
protein.
The effects of substrate concentration on reaction rate
of the mutant enzymes was investigated using the crude
extracts as a source of protein (Table 8). Mutants A203T,
A210V, L178F, G238D, and A210T exhibited an increase in
apparent increase in reaction rate relative to increasing
substrate concentration. Increasing substrate concentration
had no effect on reaction velocity for mutants G186R and
A203V, however the activities for these two mutant proteins
were above that for the pUC18 background. It is possible
that the lowest concentration of substrate used, 3 mM, is at
saturation with respect to these two mutant proteins.


Figure 12. Native-PAGE comparison of W158UGA and L178F
mutants proteins with the wild type (pLOI1005) stained with
Coomassie blue. Lane assignments: A; mutant L178F, B; pUC18
control, C; W158UGA, D; pLOI1005.


91
mM final concentration of p-NP-X or p-NP-A in citrate (40
mM)-sodium phosphate (80 mM) buffer at pH 6.0.
Determination of Km and Vmax for both enzymatic
activites. Both p-NP substrates were tested at various
concentrations to determine the Km and Vmax values with
respect to both substrates. All assays were done in
duplicate at 37C and at pH 6.0. Kinetic parameters were
graphically determined using the Lineweaver-Burk and direct
linear methods.
Substrate competition experiments. The analogue B-
methyl-D-xylopyranoside was used as an inhibitor for cr-L-
arabinofuranosidase activity. The fluorogenic substrate 4-
methylumbelliferyl-a-L-arabinofuranoside was used as an
inhibitor for B-D-xylosidase activity.
Results and discussion
Purification of xvlosidase. The xylosidase-containing
fractions were separated as one large peak during
preparative electrophoresis (Fig. 15). This peak consisted
of 15, 3 ml fractions which contained both xylosidase and
arabinofuranosidase activities. The fraction which
corresponded to the middle eight fractions exhibited the
highest activities and were pooled, concentrated, and
precipitated with ammonium sulfate. The xylosidase was
further purified by hydrophobic interaction chromatography
and was separated as a single peak consisting of six 3 ml
fractions (Fig. 16). The four most active fractions were


CHAPTER II
REVIEW OF THE LITERATURE
Many different bacteria have been characterized which
are able to depolymerize xylan including: Bacteroides
succinoaenes (Forsberg et al. 1981), Clostridium
acetobutvlicum (Lee et al. 1987), Bacillus pumilus
(Panbangred et al. 1983), Bacillus subtilis (Paice et al.
1986), Butvrivibrio fibrisolvens (Hespell et al. 1987) ,
Caldocellum saccharolvticum (Luthi et al. 1990), and
Clostridium thermocellum (Garcia-Martinez et al. 1980) .
These microorganisms which degrade xylan have been isolated
from numerous environments. Varel (1987) demonstrated that
pig large intestine contained xylanolytic Bacteroides
succinoaenes and Ruminococcus flavefaciens. Both of these
organisms are also present in large numbers in the bovine
rumen. Other studies demonstrated that up to 30% of the
metabolic energy requirements of the pig can be met via the
utilization of volatile fatty acids, products of microbial
cellulose and hemicellulose digestion (Rerat et al. 1987).
In addition, Salyers et al. (1981) isolated two species of
human colonic Bacteroides that were able to utilize xylan as
a carbon and energy source, thus producing volatile fatty
acid products. However, only 5 to 10 % of the maintenance
5


36
TABLE 1. Comparison of codon usage frequency for the three B.
fibrisolvens ORF's.
Frequency (mol %) Codon Usage
Amino Acid
Codon
B.
ORF 1
fibrisolvens E.
ORF 2 ORF 1
coli
B. fibrisolvens
xv 1A
Phe
TTT
2.7
4.1
3.6
1.3
2.9
TTC
0.9
1.2
0.3
2.2
1.5
Leu
TTA
2.0
2.5
2.8
0.7
0.7
TTG
0.2
1.2
1.0
0.9
0.5
CTT
5.6
2.7
2.5
0.8
4.4
CTC
-0-
-0-
0.3
0.8
-0-
CTA
1.4
1.0
1.8
0.2
0.2
CTG
0.5
0.8
0.3
6.8
1.0
He
ATT
4.7
3.9
4.3
2.2
2.7
ATC
2.0
1.0
0.8
3.7
1.7
ATA
0.5
1.7
3.8
0.2
0.5
Met
ATG
1.4
1.7
2.0
2.8
2.7
Val
GTT
2.9
2.7
2.3
2.9
2.0
GTC
0.9
0.2
0.5
1.2
0.5
GTA
1.4
2.1
1.3
1.8
4.2
GTG
1.1
1.2
-0-
2.2
1.0
Ser
TCT
2.3
1.5
0.8
1.3
1.2
TCC
-0-
0.4
0.3
1.5
0.5
TCA
1.8
1.5
0.5
0.4
2.9
TCG
-0-
0.6
0.3
0.6
0.2
4
AGT
0.9
0.8
0.3
0.3
-0-
AGC
0.7
1.0
1.8
1.4
1.2


2
chain is often substituted with acetyl, arabinofuranosyl,
ferulyl glucopyranosyl, and mannopyranosyl side chains to
form a complex heterogeneous structure.
Several genera of bacteria and fungi are able to
partially or completely depolymerize xylan in various
habitats (Biely, 1985) Xylan depolymerization by
microorganisms is a multistep process which involves the
concerted activities of several different enzymes.
Xylaases (1,4-B-D-xylan xylanohydrolase; EC 3.2.1.8) are
extracellular enzymes which hydrolyze the internal B-1,4-
xylosidic linkages on the main chain. The resulting smaller
oligosaccharides are transported into the microbial cells
where xylosidases (1,4-B-D-xylan xylohydrolase; EC 3.2.1.37)
continue the hydrolysis and release monosaccharides for
glycolysis (Dekker and Richards, 1976). The hydrolysis and
removal of side chain substituents requires additional
enzyme activities including arabinofuranosidase, which
removes substituted arabinofuranosyl residues from the xylan
backbone. (Biely, 1985). This may be particularly important
since arabinose substituents on the xylan chain have been
shown to limit the complete enzymatic breakdown of xylan
(Chesson et al. 1983). One of the organisms which is
particularly adept at xylan depolymerization in Butvrivibrio
fibrisolvens. B. fibrisolvens is a Gram variable,


88
both enzymatic activities in this enzyme was tested using
analogous kinetic experiments including substrate inhibition
and competition of the enzyme with respect to both the
xylopyranosyl and arabinofuranosyl substrates.
Materials and Methods
Medium and growth conditions. Medium and growth
conditions were described in chapter III.
Preparation of cell extracts. Cell extracts were
prepared as described in chapter III.
Partial purification of fl-xvlosidase by preparative
electrophoresis. Extracts containing the total cytoplasmic
proteins from pLOHOOS or L178F recombinant clones were used
as a source of xylosidase. Proteins were fractionated in an
8% native polyacrylamide gel in a BioRad-Prep Cell
preparative electrophoresis system. Gel and buffer
formulations were: separating gel buffer; 240 mM Tris (pH
8.48), stacking gel buffer; 40 mM Tris (pH 6.9), lower tank
buffer; 63 mM Tris/50 mM HCL (pH 7.5), upper tank buffer; 38
mM Tris/ 40 mM glycine (pH 8.9). A total of 50 mg/ml
protein was loaded onto the gel. Electrophoresis was done
at constant power of 31 W. Starting conditions were 250 V
and 40 mA. Protein elution was monitored at 280 nm.
Fractions were collected and assayed for xylosidase
activity as described below. The most active fractions were
pooled and concentrated using an Amicon Centriprep
concentrator (Amicon Division, Danvers, MA). Xylosidase in


l/*(iu)
no
1/[S] (mU)


J>
§
¡I r
) i i I


148
Nishitani, k., and D. J. Nevins. 1988. Enzymatic analysis of
feruloylated arabinoxylans (Feraxan) derived from Zea mays
cell walls I. Plant Physiol. 87:883-890.
Ornstein, L., and B. J. Davis. 1964. Disc electrophoresis:
background and theory. Ann. N. Y. Acad. Sci. 121:321-349.
Paice, M. G., R. Bourbonnais, M. Desrochers, L. Jurasek, M.
Yaguchi. 1986. A xylanase gene from Bacillus subtilis:
nucleotide sequence and comparison with B. pumilus gene.
Arch. Microbiol. 144:201-206.
Panbangred, W., E. Fukusaki, E. C. Epifanio, A. Shinmyo, and
H. Okada. 1985. Expression of a xylanase gene of Bacillus
pumilus in Escherichia coli and Bacillus subtilis.
Appl. Microbiol. Biotechnol. 22:259-264.
Panbangred, W., A. Shinmyo, S. Kinoshita, and H. Okada.
1983a. Purification and properties of endoxylanase produced
by Bacillus pumilus. Agrie. Biol. Chem. 47:957-963.
Panbangred, W., T. Rondo, S. Negoro, A. Shinmyo, and H.
Okada. 1983b. Molecular cloning of the genes for xylan
degradation of Bacillus pumilus and their expression in
Escherichia coli. Mol. Gen. Genet. 192:335-341.
Patterson, J. A. 1989. Prospects for establishment of
genetically engineered microorganisms in the rumen. Enzyme
Microb. Technol. 11:187-189.
Puo-Llinas, J., and H. Driguez. 1987. D-Xylose as inducer of
the xylan-degrading system in the yeast Pullularia
pullulans. Appl. Microbiol. Biotechnol. 27:134-138.
Poutanen, K., M. Tenkanen, H. Korte, and J. Puls. 1991.
Accessory enzymes involved in the hydrolysis of xylans. In
Enzymes in biomass conversion, G. F. Leatham and M. E.
Himmel, ed. American Chemical Society Symposium series 460.
American Chemical Society, Washington, D. C.
Poutanen, K., and J. Puls. 1988. Characteristics of
Trichoderma reesei B-xylosidase and its use in the
hydrolysis of solubilized xylans. Appl. Microbiol.
Biotechnol. 28:425-432.
Poutanen, K., and M. Sundberg. 1988. An acetyl esterase
of Trichoderma reesei and it'srole in the hydrolysis of
acetyl xylans. Appl Microbiol. Biotechnol. 28:419-424.
Quiocho, F. A. 1986. Carbohydrate-binding proteins:
tertiary structures and protein-sugar interactions. Ann.
Rev. Biochem. 55:287-315.


121
Materials and methods
All methods used to purify and characterize the L178F
mutant enzyme were essentially the same as those described
for the native enzyme in Chapter V.
Results and Discussion
Partial purification of L178F protein. The elution
profile for the mutant protein was essentially the same as
that for the wild type protein, have a single broad peak
consisting of 15, 3 ml fractions (Fig. 25). The fractions
which corresponded to the middle six fractions exhibited the
highest relative activity and were pooled, concentrated, and
precipitated at 70 % saturation with ammonium sulfate. SDS-
PAGE analysis of the partially purified protein indicated it
was approximately 80 % to 90 % pure with low levels of
contaminating proteins (Fig. 26).
Enzyme optima. The mutant protein exhibited a higher
thermal stability relative to the wild type (Fig. 27). This
protein was stable up to 55C for 30 min and still retained
100 % relative activity with respect to both substrates.
Since the replacement of a phenylalanine for leucine
introduces a more hydrophobic residue at this position, it
is possible that this strengthened the hydrophobic
interactions at the core of the protein and thereby
increased the thermal stability. The optimal temperature
was not investigated.


83
TABLE 8. Effects of substrate concentration on B-D-
xylosidase activity for xvlB in vitro mutants6.
Clone
3 mM
6 mM
9 mM
Kin c
app
p-nitrophenyl
-B-D-
xylopyranoside
wild type
18.6
24.0
25.6
4 mM
G186R
0.4
0.4
0.5
2 mM
A203T
0.4
0.7
0.9
11 mM
A210V
0.8
0.9
1.2
2 mM
L178F
1.4
1.8
3.0
18 mM
A203V
0.6
0.8
0.8
3 mM
G238D
0.6
1.1
1.3
7 mM
A210T
0.4
0.7
0.8
12 mM
a Specific activity expressed as nmoles p-nitrophenol
released per min per mg protein.
b Does not include the frameshift mutations or those mutant
proteins that are not expressed.
c Apparent Km values determined using the direct linear
method of Cornish and Bowden.


132
1/[S] ( mM)


Chapter V
Purification and characterization of the
xvlB-encoded protein
Introduction
The existance of polysaccharide-hydrolyzing enzymes
having broad substrate specificities is well documented
(Ward and Moo-Young, 1989). One such example includes the
exoglucanase, EXG, produced by Cellulomonas fimi (Beguin,
1991). This enzyme also exhibits B-D-xylanase activity. The
xylanase from Clostridium thermocellum (XYNZ) also exhibits
endo-glucanase activity towards carboxymethylcellulose
(Grepinet et al. 1988). In particular, several xylanase and
xylosidase enzymes have been characterized which exhibit
substrate ambiguity (Flint et al. 1989) A recent example
is a fi-D-xylosidase that was cloned from Caldocellum
saccharolvticum that also exhibits endoxylanase activity
(Luthi et al. 1990). Enzymes which exhibit both
endoxylanase and B-glucosidase activities have been shown to
be fairly common among micoorganisms (Gilkes et al. 1991).
Substrate ambiguity among the glycohydrolases has been
attributed to the similarities between the various
substrates involved. Upon closer examination, this
phenomenon is not totally unexpected. The B-(1,4)-xylosidic
bonds of xylan and the B-(1,4)-glycosidic bonds of cellulose
86


71
protein therefore plays an important in the structural
and/or catalytic role.
SDS-PAGE analysis of mutant proteins. The presence of
the 60,000 molecular weight monomeric subunit encoded by
xvlB was confirmed for the various mutants (Fig. 11). In
two cases, cell extracts from recombinants harboring the
proline (7) to leucine (7) and glycine (185) to arginine
(185), did not contain the xvlB-encoded protein band. The
insertion of the plasmid containing the proline (7) to
leucine (7) mutation into a lon-negative strain of E. coli.
which is deficient in serine proteases, resulted in the
restoration of the xvlB-encoded protein band on SDS-PAGE
gels. It seems likely that proteolysis of an improperly
folded protein is responsible for the absence of this
protein in cell extracts of these two mutants.
Native polyacrylamide gel comparisons of mutant and
wild type proteins. Zymograms of the wild type and mutants
L178F and W158UGA proteins indicated that the mutant
proteins are approximately the same size as the wild type
protein (Fig. 12). Western hybridizations of blotted
protein bands from native-PAGE indicated that all of the
expressed mutations result in proteins that have unaltered
electrophoretic mobilities and subunit assemblies relative
to the wild type protein (Fig. 13). This evidence suggests
that the point mutations that result in expressed protein do
not induce any destabilizing secondary structural


145
Grepinet, 0., and P. Benguin. 1986. Sequence of the
cellulase gene of Clostridium thermocellum coding for
endoglucanase B. Nucleic Acids. Res. 14:1791-1799.
Greve, C. L., J. M. Labavitch, and R. E. Hungate. 1984. a-L-
arabinofuranosidase from Ruminococcus albus 8: Purification
and possible role in hydrolysis of alfalfa cell wall. Appl.
Environ. Microbiol. 47:1135-1140.
Hespell, R. B., and P. J. 0'Bryan-Shah. 1988. Esterase
activities in Butvrivibrio fibrisolvens strains. Appl.
Environ. Microbiol. 54:1917-1922.
Hespell, R. B., R. Wolf, and R. J. Bothast. 1987.
Fermentation of xylans by Butvrivibrio fibrisolvens and
other ruminal bacterial species. Appl. Environ. Microbiol.
53:000-000.
Hespell, R. B., and M. P. Bryant. 1981. The genera
Butvrivibrio. Succinivibrio. Succinomomas. Lachnospira. and
Selenomonas. p. 1479-1494. In M. P. Starr, H. Stolp, H. G.
Truper, A. Balows, and H. G. Schlegel (ed), The prokaryotes,
a handbook on habitates, isolation, and identification of
bacteria. Springer-Verlag, New York.
Hobson, P. N., and R. J. Wallace. 1982. Microbial ecology
and activities in the rumen. II. Crit. Rev. Microbiol.
9:253-320.
Hobson, P. N., and M. R. Purdom. 1961. Two types of xylan
fermenting bacteria from the sheep rumen. J. Appl.
Bacteriol. 24: 188- 193.
Joliff, G., P. Benguin, and J. P. Aubert. 1986. Nucleotide
sequence of the cellulase gene celD encoding endoglucanase D
of Clostridium thermocellum. Nucleic Acids Res. 14:8605-
8613.
Kasumi, T., Y. Tsumuraya, C. F. Brewer, H. Kersters-
Hilderson, M. Claeyssens, and E. J. Hehre. 1987. Catalytic
versatility of Bacillus pumilus B-xylosidase: Glucosyl
transfer and hydrolysis promoted with a- and B-D-xylosyl
fluoride. Biochem. 26:3010-3016.
Kelly, M. A., M. L. Sinnott, and M. Herrchen. 1987.
Purification and mechanistic properties of an extracellular
a-L-arabinofuranosidase from Monilinia fructigena. Biochem
J. 245:843-849.


Figure 13. Western hybridization of native-
and mutant proteins. Lane assignments: A;
C:; P7L, D; A210V, E; L178F, F, W158UGA, G;
PAGE of wild type
A203T, B; pUC18,
PLOI1005.


51
however exhibit partial identity with the B-glucosidase from
Kluyveromvces fraailis which is consistent with the
similarity of the substrates which these two enzymes attack.
A single gene encoding xylanase/xylosidase activities
has been cloned and seguenced from Caldocellum
saccharolvticum (Luthi et al. 1990). This protein, however,
lacked arabinofuranosidase activity. Recently, a
xylosidase/arabinofuranosidase gene was proposed to reside
in a gene cluster isolated from Bacteroides ovatus
(Whitehead and Hespell, 1990). All clones exhibited both
activities concurrently and both activities co-purified.
Additionally, an enzyme having xylosidase and
arabinofuranosidase activities has been purified from
Trichoderma reesei (Poutanen and Puls, 1988). No seguence,
however, for the encoding gene has been reported.
Substrate ambiguity between carboxymethylcellulase and
xylanase enzymes is relatively common (Flint et al. 1989).
The substrate ambiguity for other xylosidase enzymes has
also been reported. By employing kinetic methods on the
purified, bifunctional B-xylosidase/B-glucosidase from
Chaetomium trilaterale. Uziie et al. (1985) demonstrated
that this enzyme possessed a single active site with dual
substrate-binding regions. More recently, a neopullulanase
from Bacillus stearothermophilus was cloned, sequenced, and
characterized (Kuriki and Imanaka, 1989). This enzyme
possessed activity against a-(1,6)-glycosidic linkages in


82
Table 7. Enzymatic activities of recombinants harboring
point mutations on xvlB relative to the wild type protein.
Clone
Xyl Sp.Ac.a
Ara Sp.Ac.
ara/xylb
pLOI1005
10.5
17.4
1.7
.46
G186R
0.1
0.1
1.0
.51
A203T
0.1
0.3
3.0
.49
A210V
0.2
0.2
1.0
.49
P7L
0.2
0.2
1.0
N/D
L178P
1.2
1.8
1.5
.53
W158AUG
2.1
2.9
1.4
.46
A203V
0.2
0.2
1.0
.49
G238D
0.2
0.2
1.0
N/D
A210T
0.2
0.1
0.5
.50
pUC18
0.0
0.0
-
a Specific activity in nmoles p-nitrophenol released per min
per mg protein.
b Ratio computed after subtraction of pUC18 background values.
c Determined by comparing relative migration distances of each
protein verses that for the dye front on agarose/xylan native
gel electrophoresis using "Western" hybridization to visualize
protein bands.


Figure 6. SDS-PAGE analysis of cytoplasmic extracts from
recombinant E. coli DH5a harboring selected plasmids.
Approximately 20 /Ltg of protein was loaded in each lane. Lanes
1 and 7; molecular weight markers, lane 2; pLOI1005, lane 3;
pUC18, lane 4; pLOI1040, lane 5; pLOI1043, lane 6; pLOI1050.
The band cooresponding to the xylosidase-arabinofuranosidase
enzyme is indicated by an arrow. The numbers in the right
represent the apparent molecular weight of the standards
(X 103) .


59


Figure 26.
SDS-PAGE analysis of partially purified L178F
mutant protein: A; Prep-Cell purified
preparation, B; crude extract, C; molecular
weight markers.


Figure 25. Elution profile of the xylB-encoded protein
harboring the L178F mutation during preparative
electrophoresis:



PAGE 1

GENETIC AND BIOCHEMICAL CHARACTERIZATION OF THE 6-D-XYLOSIDASE GENE FROM BUTYRIVIBRIO FIBRISOLVENS By ERIC ANDREW UTT 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 1991

PAGE 2

ACKNOWLEDGEMENTS I owe my development as a scientist as well as the successful completion of this dissertation to my major professor. Dr. Neal Ingram. I will be forever indebted to him for his uninhibited sharing of knowledge and expertise. I wish also to express my gratitude to the members of my graduate committee, Dr. Allen, Dr. Aldrich, Dr. Shanmugam, and Dr. Gander. Their contributions to my research and in the preparation and review of this manuscript are greatly appreciated. I must also thank my friends and comrades Jeff Mejia and David Beall for the friendship and helpful suggestions during the course of my doctoral work. To all the former postdocs. Dr. Christina Eddy, Dr. Terry 1 Conway, and Dr. Guy Sewell, from whom I learned much, I wish to express my thanks. Thanks are due to my parents for their love and support during my graduate education. Similarly, I wish to thank my wife's parents for their love and support. And lastly, I would like to thank my loving wife, Lisa, and my beautiful girls, Tina and Hannah, for making my life special. It is to them that I dedicate this dissertation. ii

PAGE 3

TABLE OF CONTENTS ACKNOWLEDGEMENTS i i LIST OF TABLES V LIST OF FIGURES vi ABSTRACT ix CHAPTERS I GENERAL INTRODUCTION 1 II REVIEW OF THE LITERATURE 5 III CLONING, SEQUENCING, AND SEQUENCE ANALYSIS OF THE XYLOSIDASE GENE FROM BUTYRIVIBRIO FIBROSOLVENS 16 Introduction 16 Materials and Methods 17 Results and Discussion 20 Conclusions 50 IV MUTATIONAL ANALYSIS OF THE XV IB GENE 54 Introduction 54 Materials and Methods 56 Results and Discussion 63 Conclusions 84 V PARTIAL PURIFICATION AND CHARACTERIZATION OF THE WILD TYPE GENE PRODUCT 86 Introduction 86 Materials and Methods 88 Results and Discussion 91 Conclusions 115 iii

PAGE 4

CHAPTERS Page VI PARTIAL PURIFICATION AND CHARACTERIZATION OF THE L178F MUTATION 12 0 Introduction 120 Materials and Methods 121 Results and Discussion 121 Conclusions 128 VII SUMMARY AND GENERAL CONCLUSIONS 135 LITERATURE CITED 142 BIOGRAPHICAL SKETCH 151 iv

PAGE 5

LIST OF TABLES Table Page 1. Comparison of codon usage frequency for the three B. f ibrisolvens ORF's 3 6 2. Expression of enzyme activities in recombinant E. coli harboring xylB 40 3. Hydrolysis of different nitrophenylsubstituted glycosides by the xylB gene product 42 4. Comparison of the translated amino acid sequences of the three B. f ibrisolvens ORF's in pLOIlOOl with those of selected proteins 46 5. Amino acid sequence alignment of conserved regions 48 6. Localization of point mutations by restriction fragment replacement analysis 67 7. Enzymatic activities of recombinants harboring point mutations in xylB relative to the wild type protein 82 8. Xylosidase activity of in vitro mutations with varying substrate concentrations 83 9. Purification scheme of the xy IB -encoded protein from E. coli DH5a 97 V

PAGE 6

LIST OF FIGURES Figure Page 1. Restriction maps of pUC18 derived plasmids that express 6-D-xylosidase activity in E. coli DH5a 22 2. Subclone analysis of pLOIlOOS to localize the xylB coding region 24 3. Southern hybridizations of chromosomal DNA from B. f ibrisolvens and E. coli 27 4. Outline of sequencing strategy of pLOIlOOl and subclone analysis of xylB 29 5. The complete nucleotide sequence and translated amino acid sequence of the 4.2 kb insert from pLOIlOOl 31 6. SDS-PAGE analysis of cytoplasmic extracts from recombinant E. coli DH5a harboring selected constructs 45 7. Assignment of domains to the xylB gene 59 8. Subcloning strategy used to localize in vitro mutations to one of five domains in the xylB gene 61 9. Deletion analysis of the xylB gene 65 10. Localization and identification of point mutations in xylB by DNA sequencing 69 11. SDS-PAGE analysis of wild type and mutant proteins 73 12. Native-PAGE comparison of W158UGA and L178F mutations with the wild type stained with Coomassie blue 75 13. Western hybridization of native-PAGE of wild type and mutant proteins 77 vi

PAGE 7

Figure Page 14. Substrate binding native gel western hybridization assays of wild type and mutant proteins 80 15. Elution profile of the xylB gene product during preparative electrophoresis on the BioRad-Prep Cell system 93 16. Elution profile of the xylB gene product during hydrophobic interaction chromatography. . 95 17. SDS-PAGE analysis of pooled xylosidasecontaining fractions from preparative electrophoresis on 8% native-PAGE 99 18 Native-PAGE analysis of the B-D-xylosidase 101 19. Thermal inactivation profile of xylosidase and arabinof uranosidase activities 103 20. Temperature optimum profile for xylosidase and arabinof uranosidase activities 105 21. pH activity profiles for xylosidase and arabinof uranosidase activities 107 22. Double recipricol plots of xylosidase and arabinofuranosidase activities of the native protein 110 23. Competitive inhibition of xylosidase activity by 4-MU-a-L-arabinofuranoside 112 24. Competitive inhibition of arabinofuranosidase activity by 6-methyl-D-xyloside 114 25. Elution profile of the xylB -encoded protein harboring the L178F mutation during preparative electrophoresis 123 26. SDS-PAGE analysis of partially purified L178F mutant protein 125 vii

PAGE 8

Figure ^^^e 27. Thermal inactivation profiles of xylosidase and arabinofuranosidase activities of the L178F mutant protein 127 28. pH activity profiles for xylosidase and arabinofuranosidase activities of the L178F mutant protein 130 29. Double recipricol plots of xylosidase and arabinofuranosidase activities for the L178F mutant protein 132 viii

PAGE 9

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 GENETIC AND BIOCHEMICAL CHARACTERIZATION OF 6-D-XYLOSIDASE FROM BUTYRIVIBRIO FIBRISOLVENS By Eric Andrew Utt December 1991 Chairman: Lonnie O. Ingram Major Department: Microbiology and Cell Science The gene for 6-D-xylosidase from the rumen bacterium Butyrivibrio f ibrisolvens encodes a protein that exhibits hydrolytic activity against 6-D-xylopyranosides and a-Larabinofuranosides. This gene, xylB . was cloned into E. coli as a 4.2 kilobase pairs (kbp) insert in pUC18 and sequenced in both directions. The xylB gene is present as a single copy on the B. f ibrisolvens chromosome and consists of a 1,551 base pair (bp) open reading frame (ORF) which encodes a protein of 517 amino acids. Insertion of a 10 bp linker into the coding region resulted in a frameshift that abolished both activities. Deletions from the 3' end and the 5' end of xylB also resulted in inactive proteins. SDSPAGE analysis of cytoplasmic extracts from recombinant E. coli clones harboring xylB confirmed the presence of a new protein with an apparent molecular weight of 60,000. ix

PAGE 10

Although the xylB gene did not exhibit a high degree of amino acid identity with other xylan-degrading enzymes or glycohydrolases, a conserved sequence was identified with significant identity to the active site region of hen egg white lysozyme and Aspergillus niger glucoamylase. No predictable stem loop structures or sequences resembling terminators were found on the xylB gene fragment and this gene appears to be part of an operon. In vitro analysis of xylB mutants demonstrated structural and functional relationships between the two enzyme activities. All point mutations investigated in xylB resulted in the reduction or loss of both enzymatic activities. Most of these mutations were clustered in a region near the proposed active site. The point mutations decreased the apparent affinity of the enzyme for xylan. The partially purified xvlB -encoded protein exhibited thermal inactivation kinetics and temperature optima that were essentially the same for both enzymatic activities. The pH optimum for both activities was 6.0. However, the arabinofuranosidase activity exhibited a broader pH range, retaining 90% of maximal activity up to pH 9.0. The apparent Km for p-nitrophenyl-BD-xylopyranoside and p-nitrophenyl-a-L-arabinofuranoside were 3.7 mM and 1.8 mM respectively. Substrate competition experiments corroborated the genetic evidence and demonstrated that the same active center was responsible for both enzymatic activities of the xy IB -encoded protein. X

PAGE 11

CHAPTER I GENERAL INTRODUCTION Plant cell walls represent the largest reserve of fixed carbon on earth. Plant cell walls are composed primarily of cellulose, hemicellulose, and lignin (Weinstein and Albersheim, 1979) . Cellulose is the most abundant carbohydrate found in plant biomass (Coughlan, 1985) while hemicellulose is a major plant structural polymer that ranks second only to cellulose in natural abundance (Dekker and Richards, 1976) . The amount of hemicellulose in dry wood is between 20% and 30% (Eriksson et_al. 1990) . The composition of hemicellulose varies between softwoods and hardwoods. The major hemicellulose in softwoods is galactoglucomannan (Eriksson et al . 1990) . This polymer has a backbone composed of a linear chain of 1,4-linked 6-D-glucopyranose and 6-D-mannopyranose units. The mannose and glucose moieties of the backbone may be substituted with acetyl groups at the C-1 and C-2 positions. Glucouronoxylan (0-acetyl-4-0-methyl-glucurono-6-Dxylan) is the major hemicellulosic component of hardwoods and agricultural residues (Eriksson et al . 1990) . The major structural feature of xylan is a linear chain consisting of 1

PAGE 12

2 chain is often substituted with acetyl, arabinofuranosyl, ferulyl glucopyranosyl , and mannopyranosyl side chains to form a complex heterogeneous structure. Several genera of bacteria and fungi are able to partially or completely depolymerize xylan in various habitats (Biely, 1985) . Xylan depolymerization by microorganisms is a multistep process which involves the concerted activities of several different enzymes. Xylanases (1, 4-6-D-xylan xylanohydrolase; EC 3.2.1.8) are extracellular enzymes which hydrolyze the internal 6-1,4xylosidic linkages on the main chain. The resulting smaller oligosaccharides are transported into the microbial cells where xylosidases (1, 4-6-D-xylan xylohydrolase; EC 3.2.1.37) continue the hydrolysis and release monosaccharides for glycolysis (Dekker and Richards, 1976) . The hydrolysis and removal of side chain substituents requires additional enzyme activities including arabinofuranosidase, which removes substituted arabinofuranosyl residues from the xylan backbone. (Biely, 1985) . This may be particularly important since arabinose substituents on the xylan chain have been shown to limit the complete enzymatic breakdown of xylan (Chesson et al . 1983). One of the organisms which is particularly adept at xylan depolymerization in Butvrivibrio f ibrisolvens . B. f ibrisolvens is a Gram variable.

PAGE 13

obligately anaerobic bacillus that is frequently found in the rumen and anaerobic digesters (Dehority, 1966) . This organism produces a cadre enzymes which enable it to degrade plant biomass, including cellulose and hemicellulose (Hespell, 1987). The genus Butyrivibrio contains only a single species but consists of many strains that vary in DNA relatedness between 20% to 100% (Mannarelli, 1988). This organism is also characterized as having a low (38% to 42%) guanine plus cytosine (mole percent) content. Butyrivibrio produces an extracellular polysaccharide (EPS) that contains an unusual 4-0(1-carboxymethyl) -rhamnose sugar (Mannarelli et_al. 1990) . These investigators suggested that the . unusual sugars found in the EPS of B. f ibrisolvens serve to protect the organism from glycosidases and other enzymes found in the digestive tract of the host animal. B. f ibrisolvens GS113, an anaerobic digester isolate used in these studies, was shown to produce high levels of both xylanase and xylosidase (Sewell et al . 1988) . These two enzymes were shown to be repressed by glucose and induced by xylan and xylose. In a previous study, the xylB gene encoding the 6-Dxylosidase from B. f ibrisolvens GS113 was isolated from a plasmid pUC18 genomic library (Sewell et al . 1989) . During the course of the current investigations, an aL-arabinofuranosidase activity was detected in all clones harboring the xylB gene. This characteristic became the

PAGE 14

principle focus of this research. The principle question that was addressed relates to the dual activity exhibited by this enzyme against 6-D-xylopyranosides and a-Larabinofuranosides: do these two activities reside in the same active center of the enzyme or are they on separate domains? The following research examined the structure and function of the xylB gene and gene product and includes: A) The complete nucleotide sequence of xylB and sequence comparisons with related enzymes from other organisms . B) Genetic evidence that the two enzymatic activities are encoded by a single open reading frame. C) Mutational analyses to investigate the genetic interdependence of the two enzymatic activities of xylB . D) Purification and characterization of the xylB encoded protein. E) Biochemical and kinetic experiments to investigate the functional relationship between the two enzymatic activities.

PAGE 15

CHAPTER II REVIEW OF THE LITERATURE Many different bacteria have been characterized which are able to depolymerize xylan including: Bacteroides succinoaenes (Forsberg et__al. 1981) , Clostridium acetobutvlicum (Lee et al . 1987) , Bacillus pumilus (Panbangred et al . 1983), Bacillus subtilis (Paice et al . 1986) , Butvrivibrio f ibrisolvens (Hespell et al . 1987) , Caldocellum saccharolyticum (Luthi et al . 1990) , and Clostridium thermocellum (Garcia-Martinez et al . 1980) . These microorganisms which degrade xylan have been isolated from numerous environments. Varel (1987) demonstrated that pig large intestine contained xylanolytic Bacteroides succinoaenes and Ruminococcus f lavef aciens . Both of these organisms are also present in large numbers in the bovine rumen. Other studies demonstrated that up to 30% of the metabolic energy requirements of the pig can be met via the utilization of volatile fatty acids, products of microbial cellulose and hemicellulose digestion (Rerat et al . 1987) . In addition, Salyers et al . (1981) isolated two species of human colonic Bacteroides that were able to utilize xylan as a carbon and energy source, thus producing volatile fatty acid products. However, only 5 to 10 % of the maintenance 5

PAGE 16

6 energy requirements are derived from volatile fatty acids produced in the human colon (MacNeil, 1984). Xylanolytic enzymes have also been isolated and characterized from several fungi including Trichoderma reesei (Poutanen and Puis, 1988) , Aspergillus niger (Fukumoto et al . 1970) , and Fusarium roseum (Gascoigne and Gascoigne, 1980) . The rumen is the primary organ in which cattle, sheep, and other ruminants derive their energy and nutrition through breakdown of complex carbohydrates. Starch, cellulose and hemicellulose are degraded by enzymes that are secreted by resident microorganisms and metabolized to volatile fatty acids as the end products of fermentation (Hobson and Wallace, 1982). In general, cellulose and hemicellulose depolymerization by rumen microbial flora releases free monosaccharides and short chain oligosaccharides. The predominant metabolic waste products, volatile fatty acids, are released and either absorbed and utilized by the animal or used by other microorganisms in the rumen and other digestive organs (omasum, abomasum and the small intestine) . The rumen microbial community represents a diverse group of organisms, many of which have the ability to degrade hemicellulose (Dehority, 1966) . The rumen ecosystem differs from other microbial ecosystems in substrate availability and product accumulation. The rumen is close to an industrial fermentation in that substrate availability

PAGE 17

7 is very high and constant while product accumulation is low. It has been established that microbial cells in the rumen are present in high numbers and contains: lo" bacteria ml"\ 10* ciliate protozoa ml''', and 10^ fungi ml"'' (Patterson, 1989) . Butyrivibrio f ibrisolvens is a Gram variable, obligately anaerobic, motile bacillus (Hespell and Bryant, 1981) . Bj. f ibrisolvens is particularly abundant in the rumen and anaerobic digesters in which plant material serves as the primary substrate (Hespell et al . 1987) . B. f ibrisolvens converts hemicellulose to mono and oligosaccharides. These are transported and metabolized to yield butyric acid. Mannarelli et al . (1990) cloned and sequenced the gene encoding 6-D-xylanase from B. f ibrisolvens strain 49. Sewell et al . (1988) isolated several strains of B. f ibrisolvens that produced both xylanase and xylosidase. In this study, the synthesis of both enzymes were concurrently repressed by glucose and induced by xylan and xylose. This was surprising since earlier work on rumen isolates of B. f ibrisolvens had reported that these enzymes were expressed constitutively (Hespell et al . 1987). Similarly, it was reported that xylose served as an inducer of the xylanase and 6-Dxylosidase in Pullularia pullulans ( Pou-Llinas and Driguez, 1987) . In addition, 6-D-xylosidase of Bacillus pumilus was found to be induced by xylose (Kersters-Hilderson et al .

PAGE 18

1969) . Xylobiose was found to induce synthesis of 6xylanase in Cr yptococcus albidus (Biely et al . 1980) and in Streptomyces sp. (MacKenzie et al . 1987). Biely and Petrakova (1984), studying the xylan-degrading system in C. albidus . found that certain positional isomers of xylose and xylobiose, notably 1 , 4-6-xylobiose, could serve as inducers of 6-xylanase and 6-D-xylosidase. Some organisms produce multiple xylanase enzymes. Esteban et al . (1982) reported that Bacillus circulans WL-12 secretes two endo-6-xylanases and one 6-D-xylosidase when grown on xylan as a sole carbon source. Three distinct xylanase genes have been identified and cloned from Clostridium thermocellum (MacKenzie et al . 1989) . A multiplicity of xylanases has also been reported in fungi including Aspergillus niqer (Frederick et al . 1985) , and Trichoderma harzianum ( Wong et al . 1986) . It has been suggested that the multiplicity serves to enhance the ability of microbes to depolymerize a wide range of substituted xylans under different environmental conditions. In some organisms, the xylanolytic and cellulolytic systems are combined. Recently Morag et al . (1990) demonstrated that, in addition to free xylanases, Clostridium thermocellum possessed a cellulosome-associated xylanase which exhibits endo-glucanase activity. However this organism was unable to utilize or grow on xylan. These investigators postulated that cellulosome-associated

PAGE 19

xylanolytic enzymes act to increase the availability of cellulose to cellulases of the cellulosome through removal of associated xylan chains. In the rumen, cooperativity between xylanase and cellulase degrading enzymes is also apparent. In Bacteroides succinoqenes isolated from rumen fluid it was demonstrated that carboxymethylcellulase (61, 4-endo-glucanase) , 6-xylanase, and 6-D-xylosidase were expressed by the organism when grown on media containing cellulose as a sole source of carbohydrate (Forsberg et al . 1981) . These investigators postulated that cooperativity between the cellulose and hemicellulose degrading enzymes helps to enhance polymer breakdown and increase substrate availability for rumen microorganisms which lack these enzymes. This cooperativity among different organisms may serve to maintain a stable microbial population in the rumen . Enzymatic cooperativity and synergism is also present within the hemicellulose-degrading systems. A synergistic action of 6-xylanase and 6-D-xylosidase has been demonstrated in cultures of Neurospora crassa when grown on xylan (Deshpande et al . 1986) . In this study the degree of hydrolysis of D-xylan by xylanase was increased 30% by the addition of 6-D-xylosidase to a cell-free system. Another example of enzymatic synergism involves the enzyme acetyl esterase. Acetyl esterase (EC 3.1.1.6) is active against esters of acetic acid and are widely

PAGE 20

10 distributed in nature (Poutanen et_al. 1991) . The acetyl residues on the xylan backbone are removed by acetyl esterase (Biely et al . 1985) . These enzymes have been found to act cooperatively with xylanases. The acetyl esterase serves to increase the rate of glycosidic bond cleavage by 6-xylanase from Trichoderma reesei (Biely et al . 1986) . These enzymes were also found to act synergistically to liberate acetyl residues. More recently it was demonstrated that the rate of liberation of acetic acid from acetyl-xylan by acetyl esterase of T. reesei was increased by the addition of endo-xylanase and 6-D-xylosidase (Poutanen and Sundberg, 1988) . Also involved in enzyme synergism is the enzyme a-Larabinofuranosidase (Greve et al. 1984). This enzyme was purified from Ruminococcus albus 8 and had a pH optimum of 6.9 and a Km of 1.3 mM, both for p-nitrophenyl-a-Larabinofuranoside as a substrate. They showed that this enzyme enhanced the rate of hydrolysis of alfalfa cell wall hemicellulose when combined with other xylanolytic or pectinolytic enzymes. It was hypothesized that this enzyme functioned to provide rumen microbes with suitable substrates for xylanase. The mechanism of xylan hydrolysis by microbial xylanases has been studied extensively. Xylanases are usually small proteins having molecular weights ranging between 20,000 to 50,000 (Bastawde et al . 1991). Most

PAGE 21

11 xylanases are in fact endo-xylanases by virtue of the fact that they attack the interior 6( 1 , 4) -D-xylosidic linkages of the xylan polymer rather than the exterior linkages (Ward and Moo-Young, 1989). The 6( 1 , 4 ) -D-endo-xylanases have a pH optimum in the range of 3.5 to 6.5 while the temperature optima and thermal stabilities vary depending upon the source (Ward and Moo-Young, 1989) . Xylanase from Bacillus pumilus IPO has a molecular weight of 22,000 and is a 6-D1, 4-endo-xylanase (Panbangred et al . 1983). The pH and temperature optimum of this enzyme are .6.5 and 40''C, respectively. Quantification of the hydrolysis end products from larchwood xylan indicated that the B. pumilus enzyme had the greatest affinity for the second and sixth 6xylosidic linkages of the polymer. The xynZ gene product from Clostridium thermocellum is also an endo-xylanase with pH and temperature optima of 6.0 and 65°C, respectively (Grepinet et al . 1988) . Lee et al . (1987) purified and characterized two different endoxylanases, xylanase A and xylanase B, from Clostridium acetobutylicum . Xylanase A has a molecular weight of 65,000, a pH optimum of 5.0, an optimum temperature of 50°C, and is stable for up to 30 min at 40''C. Xylanase B is a smaller protein having a molecular weight of 29,000. It had a pH optimum of 5.0 to 6.0, showed a temperature optimum of 60°C, and is stable for 30 min at 50°C. Both enzymes hydrolyze larchwood xylan randomly, however xylanase B

PAGE 22

12 produced only xylotriose and xylobiose as products whereas xylanase A also yields xylohexose, -pentose, and -treaose as end products. Xylanase A was also active against carboxymethylcellulose, acid-swollen cellulose and lichenin. The two enzymes were antigenically different as judged by "Ouchterlony"-immunodif fusion assays. The two enzymes were therefore presumed to be encoded by separate genes. Some xylanases, such as those that are produced by fungi, notably Aspergillus niger, produce endo-1 , 4-6-Dxylanases that can hydrolyze the 1, 3-a-L-arabinofuranosyl side chains from arabinoxylans (Dekker, 1985) . These enzymes have been termed "debranching" xylanases. Recently a unique "appendage-dependent" xylanase was isolated and purified from Bacillus subtilis (Nishitani and Nevins, 1991). This enzyme is classified as a 6-(l,4)-xylan xylanohydrolase and has a prerequisite for glucuronosyl substituted side chains in order to initiate hydrolysis of the xylan backbone structure. Three novel xylanases were purified from B. subtilis which exhibited activity against f erulylolated arabinoxylans (Nishitani and Nevins, 1988) . These enzymes acted on ferulic acid-substituted arabinoxylan and liberated the terminal arabinofuranosyl , terminal gluconopyranosyl, and ferulic acid moieties from the polymer. While much recent work has concentrated on the extracellular microbial xylanases, less is known concerning the molecular biology and properties of the intracellular 6-

PAGE 23

D-xylosidase component of the microbial xylanolytic system. Early mechanistic studies of the B. pumilus xylosidase indicated the enzyme contained several thiol groups and at least one of which is involved in the catalysis (Saman et al. 1975) . Panbangred et al . (1983) first cloned the genes for 6-xylanase and 6-D-xylosidase from Bacillus pumilus IPO. Both cloned proteins were expressed in Escherichia coli from a hybrid plasmid and were immunologically and chemically identical to those of B. pumilus . The cloned genes from B. pumilus IPO were later sequenced by Moriyama et al . (1987) . The gene for 6-D-xylosidase was localized to a 1617 base pairs open reading frame encoding a deduced 62,607d protein. The N-terminal amino acid sequence agreed with that predicted from the DNA sequence and that obtained from the purified enzyme. It is interesting to note that the 6xylanase gene from the same organism was located 4,600 base pairs downstream from the 3 ' -end of the 6-D-xylosidase. The B. pumilis enzyme was not reported to exhibit any additional enzymatic activities. More recently, two xylosidase genes were cloned and sequenced from the obligately anaerobic, thermophilic organism Caldocellum saccharolvticum (Luthi et al . 1990) . The protein encoded by one of these xylosidase genes was found to possess xylanase activity in addition to the expected xylosidase activity. These genes were also found to reside in close proximity to each other and to a gene

PAGE 24

encoding a protein having acetyl esterase activity. Additional fi-D-xylosidase enzymes with multiple activities have been reported earlier. Kinetic methods were used to investigate the active site of a 6-D-xylosidase from Chaetomium trilaterale (Uziie et al . 1985) . In this study, which employed substrate analogues as inhibitors, a single active center was postulated to function for both the 6-Dxylosidase and 6-glucosidase activities exhibited by this enzyme. A B-D-xylosidase purified from Trichoderma reesei was also found to exhibit an a-L-arabinofuranosidase activity (Poutanen and Puis, 1988) . Additionally, a cloned gene cluster from Bacteroides ovatus was also found to exhibit 6-D-xylosidase and a-L-arabinofuranosidase activities (Whitehead and Hespell, 1990) . These dual activities co-purified and were encoded by a single open reading frame present in the cloned gene fragment. While the dual activities of 6-D-xylanases and 6-D-xylosidases have been documented, little is known about the genetic and biochemical basis of this property. Multiple substrate activities can be attributed to the presence of more than one catalytic region on the enzyme. Another possibility is the presence of a single catalytic region with wide substrate specificity. It has been proposed that the more evolved an enzyme or protein is, the more narrow its specificity becomes (Knowles, 1988) . Accordingly the more primitive proteins tend to have multiple functions. It has

PAGE 25

15 also been proposed that the environment in which an enzyme evolves also contributes to the enzyme specificity (Robson and Gardier, 1988) . The rumen, an environment with specialized substrate-hydrolyzing requirements, may exert selective pressures resulting in the evolution of organisms and enzyme systems that reflect the heterogeneous nature of available substrates.

PAGE 26

CHAPTER III CLONING, SEQUENCING, AND SEQUENCE ANALYSIS OF THE XYLOSIDASE GENE FROM BUTYRIVIBRIO FIERI SOLVENS Introduction The synthesis of enzymes needed for xylan depolymerization has been found to be constitutive in many ruminal isolates of B. f ibrisolvens (Hespell et al . 1987) . Recently, anaerobic digester isolates of B. f ibrisolvens have been described in which the synthesis of xylanase and xylosidase were coordinately repressed by glucose and induced by xylans and xylose (Sewell et al . 1988) . The gene for 6-D-xylosidase from B. f ibrisolvens GS113, in which this enzyme is inducible, has been cloned on a multicopy plasmid pUC18 in Escherichia coli (Sewell et al . 1989) . Subcloning analysis localized the coding region to a 5.8 kilobase pairs (kbp) segment of cloned B. f ibrisolvens DNA. The enzyme was found to be predominantly intracellular in B. f ibrisolvens with 25% of the activity associated with the cell membrane fraction. The cloned xylosidase is primarily cytoplasmic with less than 2% of the active protein being membrane associated in E. coli. This investigation has been extended by restriction endonuclease mapping the B. f ibrisolvens DNA insert in pUC18 16

PAGE 27

17 and to further define the coding region of the 6-Dxylosidase gene in the insert. The number of chromosomal copies of this gene was determined by Southern hybridization. Additional subclones and primers were generated to allow complete DNA sequencing of both strands. Finally, the DNA sequence was compared with other, related gene sequences. Materials and Methods Medium and growth conditions . Escherichia coli DH5a was propagated at 37°C in Luria broth or on Luria agar supplemented with 50 mg of ampicillin per liter (Maniatis et al. , 1982). Genetic methods . Plasmid pUC18 was used as a cloning vector in all cloning and sequencing experiments unless otherwise noted. The plasmids pLOIlOOl and pLOI1005 harbor the xylosidase coding region (Sewell et al . . 1989). Analysis of restriction sites, plasmid purification, subcloning, DNA ligation. Southern hybridization and other DNA manipulations were performed using standard methods (Maniatis et al . . 1982). Restriction enzymes (Bethesda Research Laboratories, Gathersburg, MD) were used according to the manufacturer's instructions. Transformed colonies were screened for xylosidase and arabinofuranosidase activity on Luria agar plates containing 20 Mg/ml of the flurorogenic substrates 4-methylumbellif eryl-6-Dxylopyranoside or 4-methylumbellif eryl-a-L-arabinofuranoside

PAGE 28

(Sigma Chemical Co., St. Louis, MO.). The internal Sau3A and large internal PstI fragment from pLOI1005 were utilized as probes in the Southern hybridization analysis of digested chromosomal B. 3 f ibrisolvens and E. coli DNA. DNA sequencing . Double-stranded DNA was sequenced in both directions using the dideoxy-chain termination method (Sanger, 1982) and Sequenase (United States Biochemical Corp.) according to the manufacturer's instructions. Additional sequencing primers were synthesized by the University of Florida Interdisiplinary Center for Biotechnology Research and the Department of Microbiology and Cell Science Nucleotide facility. The DNA sequences were assembled using the "GENEPRO" software package (Hoefer Scientific Instruments, San Francisco, Calif.) and the University of Wisconsin Genetics Computer Group GCG package, version 6.1 (Devereux et al . 1984) Primary sequence comparisons were made with GenBank and EMBL sequence libraries . Preparation of cell extracts . E. coli cells harboring the recombinant plasmids were harvested while in midexponential phase of growth by centrifugation (10,000 g, 10 min, 4°C) and washed twice with 5 mM phosphate buffer (pH 6.8). Cell pellets were stored at -70°C, until needed. Cell pellets were thawed on ice and resuspended in an equal volume of 5 mM phosphate buffer (pH 6.8) containing 10 mM 6-

PAGE 29

mercaptoethanol and were lysed by two passes through a French pressure cell at 20,000 lb in'^. Cell membranes and other debris were removed by centrif ugation (100,000 g, l h, i°C) . Supernatants containing the total cytoplasmic proteins were stored at -70°C. Enzyme assays . 6-D-xylosidase and a-Larabinofuranosidase activities were determined by measuring the rate of hydrolysis of p-nitrophenyl-6-D-xylopyranoside and p-nitrophenyl-a-L-arabinofuranoside (1 mM final concentration) , respectively, in 50mM phosphate buffer (pH 6.8) at 37°C. The nitrophenyl derivatives of other monoand disaccharides were examined as possible substrates under the same conditions. All assays were conducted in a volume of 1 ml catalysis was terminated by the addition of 2 ml of 500 mM sodium carbonate. The hydrolysis of one nmole of substrate resulted in an increase of absorbance of 0.007 at 405nm. Specific activities are expressed as nmoles pnitrophenol released per minute per milligram of total protein. Carbohydrate derivatives were purchased from Sigma Chemical Co. Protein concentration was estimated by the method of Bradford (Bradford, 1976) . Sodium do decvl sulf ate-polvacrvlamide ael electrophoresis r sDS-PAGEK Cell proteins were separated in denaturing gels by the method of Laemmli (Laemmli, 1970). Protein bands were visualized by staining with Coomassie blue.

PAGE 30

20 Nucleotide sequence accession number . The nucleotide sequence reported here has been assigned GenBank accession number M55537. Results and discussion i Identification of the xylosidase coding region . Many restriction endonuclease sites on the original xylosidasepositive subclone, pLOIlOOS (3.2 kb B. f ibrisolvens insert), were mapped to facilitate the generation of futher subclones for analysis (Fig. 1) . Plasmid DNA was purified using cesium chloride and digested with a battery of restriction enzymes. Restriction endonuclease sites were used to generate subclones in both directions with respect to the lac promoter in pUC18. Each subclone was examined for enzyme activity on 4-methylumbellif eryl-6-D-xylopyranoside (MUX) indicator plates. Based upon the results of these experiments (Fig. 2) , the region encompassing the xylosidase gene was localized to a 2.1 kb DNA segment that spans an internal EcoRI site and the internal PstI site. The predicted gene size was in excess of the 1.4 kb Dral fragment. In addition to the indicated xylosidase a ctivity, arabinofuranosidase activity was also associated with all xylosidase-positive subclones. It seems possible that the gene, classified as xylB, encodes an enzyme that has activity against both substrates. The number of chromosomal copies of xylB was examined using Southern hybridization. The large internal PstI and

PAGE 31

Figure 1. Restriction endonuclease digestion maps of plasmids pLOIlOOl and pLOIlOOS that express 6-D-xylosidase activity in E. coli DH5a.

PAGE 32

PstI

PAGE 33

Figure 2. Subclone analysis of pLOIlOOS to localize the xylB coding region: xyl; xylosidase activity, araf; arabinofuranosidase activity. A "+" or "-" denotes the presence or absence of enzyme activity.

PAGE 34

0.0 1.0 PS E H D H kilobase pairs 24 2.0 3.0 H P H S D B Xvl Araf + + H H H H H H Coding Region

PAGE 35

Sau3A fragments were used as probes against PstI and Sau3A digests of E. coli DH5a, B. fibrsolvens GS113 chromosomal DNA and plasmid pLOIlOOS (Fig. 3) . These probes did not bind to the DH5a chromosomal DNA but did bind to a single band in each B. f ibrisolvens chromosomal digest. A single copy of the xylB gene appears to be present in B. f ibrisolvens GS113. Multiple copies of xylosidase genes have been reported for Bacillus pumilus (Panbangred et al . . 1984) and for Caldocellum saccharolyticum (Luthi et al . . 1990). If there are additional xylosidase genes present in GS113, then they must share limited homology with xylB . DNA sequence of the xylosidase gene fxylB) . Plasmid pLOIlOOl, the original GS113 library clone (Sewell et al., 1989), contained a 4.2 kb insert of B. f ibrisolvens DNA. Both strands of this fragment were sequenced utilizing the strategy summarized in Fig. 4A. The complete nucleotide sequence of this fragment is outlined in Fig. 5. Analysis of the sequence revealed the presence of three open reading frames (ORF's) in this DNA segment. The first ORF, ORFl, was incomplete and is 1,340 bp in length. It lacks a ShineDalgarno sequence and an ATG initiation codon. 0RF2 , was 1,551 bp in length and was found 15 bp downstream from ORFl. 0RF2 has a Shine-Dalgarno sequence located 6 bp upstream from the initiation codon and defines a complete gene. This gfene spans the predicted xylosidase coding region and

PAGE 36

Figure 3. Southern hybridizations of chromosomal DNA from B. f ibrisolvens GS113 (8 hr exposure) and E. coli DH5a (8 hr exposure) and plasmid DNA from pLOIlOOS (2 hr exposure) . Lanes l, 2, and 3 contain PstI digests from E. coli . B. f ibrisolvens . and pLOIlOOS, respectively, probed with the internal PstI fragment from pLOIlOOS. Lanes 4, 5, and 6 are Sau3A digests of E. coli DH5a, B. f ibrisolvens GS113 and pLOIlOOS which were probed with the internal Sau3A fragment from pLOI1005. The additional bands in lane 6 represent larger, incomplete restriction endonuclease digestion products .

PAGE 37

27 3 4 5 6

PAGE 38

Figure 4. Outline of sequencing strategy of pLOIlOOl and subclone analysis of xylB . (A) Sequencing strategy of the complete 4.2 kb B. f ibrisolvens insert in pLOIlOOl. Arrows indicate the direction of sequencing. Subclones were sequenced using universal pUC18 primers (vertical bars in front of arrows) and additional oligonucleotide primers (vertical bars absent) were both used. (B) Outline of the three ORF's sequenced and selected subclones and insertional inactivation used to identify xylB . Enzyme activity was evaluated using MUG indicator plates. Double vertical bar in pLOI1040 indicates the site of insertion of a 10 bp NotI linker. Abbreviations: E; EcoRI . H; Hindlll . A; AccI, X; Xbal . D; Dral, S; Sspl . P; PstI . ERV; EcoRV, xyl; xylosidase activity, araf; arabinofuranosidase activity. A "+" or "-" denotes the presence or absence of enzyme activity, respectively.

PAGE 39

pLOI 1001 EH AX DSP H ERV H E D S P 29 P -> — *B 1000 BASE PAIRS 2000 3000 4000 ORFl 0RF2 0RF3 E pLOI 1005 xyl araf + + PLOI104 3 + + ! pLO] :i040 I NotI linker insertion

PAGE 40

Figure 5. The complete nucleotide sequence and translated amino acid sequence of the 4.2 kb insert from pLOIlOOl. Putative Shine-Dalgarno (S.D) sequences and initiation codons are underlined. Translational termination is indicated by an asterisk (*) .

PAGE 41

31 1 AATTGTGGATGCACATATGAAAAGCTGATTTATGCTTATAAGGCAGGTCTTGTCAAGGAA 60 NCGCTYEKLIYAYKAGLVKE 61 GAGACCATCGATGAGGCTGTTACTCGACTTATGGAAATCAGACTTCGTCTAGGTACTATT 120 ETIDEAVTRLMEIRLRLGTI 121 CCAGAGAGAAAGAGTAAGTATGATGATATCCCATATGAAGTGGTCGAATGCAAAGAGCAT 180 PERKSKYDDIPYEVVECKEH 181 ATCAAACTTGCTCTTGACGCTGCAAAGGATAGCTTTGTCCTTTTGAAGAATG ATGGTTTA 240 IKLALDA AKDSFVLLKNDGL 241 CTTCCACTGAATAAAAAGGATTATAAATCTATTGCTGTTATTGGACCCAACTCTGATTCA 300 LPLNKKDYKSIAVIGPNSDS 301 AGAAGAGCTTTAATTGGAAATTATGAGGGCCTTTCTTCAGAGTATATTACAGTTTTAGAG 360 RRALIGNYEGLSSEYITVLE 361 GGGATTCGTCAGGTTGTCGGTGATGATATTAGATTATTCCACGCTGAGGGCACTCATCTT 420 GIRQVVGDDIRLFHAEGTHL 421 TGGAAGGATAGAATTCACGTAATCAGTGAGCCAAAAGATGGATTTGCCGAGGCTAAAATC 480 WKDRIHVISEPKDGFAEAKI 481 GTGGCAGAGC ATTCAGATTTAGTTGTGATGTGTCTTGGACTTGACGCATCAATCGAAGGA 540 VAEHSDLVVMCLGLDASIEG 541 GAAGAAGGAGACGAGGGTAATCAGTTCGGTAGCGGAGACAAGCCTGGATTAAAGCTTACA 600 EEGDEGNQFGSGDKPGLKLT 601 GGTTGTCAGCAAGAGCTACTTGAGGAAATTGCCAAAATCGGCAAGCCTGTTGTACTTCTT 660 GCQQELLEEIAKIGKPVVLL 661 GTGCTTTCAGGTTCTGCTCTTGATTTATCATGGGCGCAGGAATCTAATAACGTAAATGCG 720 VLSGSALDLSWAQESNNVNA 721 ATAATGCAGTGCTGGTATCCAGGCGCAAGAGGTGGACGTGCTATTGCAGAGGTTTTATTT 780 IMQCWYPGARGGRAIAEVLF 781 GGCAAGGCCAGTCCAGGCGGTAAAATGCCTCTTACATTTTATGCCTCAGATGATGACCTT 840 GKASPGGKMPLTFYASDDDL 841 CCTGATTTTTCTGATTATTCAATGGAAAATAGGACATACAGATATTTCAAGGGCACACCA 900 PDFSDYSMENRTYRYFKGTP 901 CTTTATCCATTTGGTTATGGACTAGGTTATTCTAAAATTGATTATCTATTTGCTTCTATT 960 LYPFGYGLGYSKIDYLFASI 961 GATAAAGATAAGGGAGCAATTGGTGATACATTCAAGCTAAAGGTAGATGTTAAAAATACC 1020 DKDKGAIGDTFKLKVDVKNT 1021 GGTAAGTATACACAGCATGAGGCTGTTCAAGTATATGTAACGGACCTTGAGGCAACGACA 1080 GKYTQHEAVQVYVTDLEATT 1081 AGAGTGCCTATTAGAAGCCTTAGAAAGGTTAAATGTCTAGAGCTTGAGCCTGGTGAAACA 1140 RVPIRSLRKVKCLELEPGET 1141 AAAGAGGTTGAATTTACCCTTTTTGCAAGAGATTTTGCCATTATTGATGAAAGGGGAAAA 1200 KEVEFTLFARDFAIIDERGK 1201 TGTATCATAGAGCCAGGCAAGTTTAAGATTTCTATTGGGGGACAACAGCCAGACGATAGA 1260 CIIEPGKFKISIGGQQPDDR

PAGE 42

1261 AGTAAAGAACTTATGGGCAGAGAGTGTGATATTTTTGAAATTGAATTAACAGGCTCTGTT 1320 SKELMGRECDIFEIELTGSV 1321 ACAGAAGTTGAATATTAATTGAOAgOTGCATCATGGTTATAGCTAACAATCCAATTTTAA 1380 TEVEY* MVIANNPILK 1381 AAGGTTTTTATCCAGACCCTTCTATCTGCAGAAAAGGGGATGATTTTTATCTAGTTTGTT 1440 GFYPDPSICRKGDDFYLVCS 1441 CAAGTTTTGTGTATGCTCCGGGAGTACCGATTTTTCACACTAAGGATTTGGCACATTTTG 1500 SFVYAPGVPIFHTKDLAHFE 1501 AGCAAATTGGAAATATATTAGACAGAGAAAGTCAACTTCCATTGTCGGGAGATATATCTA 1560 QIGNILDRESQLPLSGDISR 1561 GAGGCATATTTGCCCCAACAATAAGAGAGCATAATGGAATCTTTTACATGATAACAACTA 1620 GIFAPTIREHNGIFYMITTN 1621 ATGTAAGCTCTGGCGGCAACTTTATTGTTACTGCAAAAGATCCAGCTGGTCCTTGGTCAG 1680 VSSGGNFIVTAKDPAGPWSE 1681 AGCCATATTATTTAGGTGAAGATGAGGCGCCAGGTATTGATCCATCTCTGTTTTTTGATG 1740 PYYLGEDEAPGIDPSLFFDD 1741 ACGATGGCAAATGTTATTACGTTGGTACCAGACCAAATCCTGATGGAGTTCGTTACAACG 1800 DGKCYYVGTRPNPDGVRYNG 1801 GTGATTGGGAGATATGGGTTCAAGAGCTGGATTTAGAGCAAATGAAACTTGTAGGTCCTT 1860 DWEIWVQELDLEQMKLVGPS 1861 CGATGGCAATTTGGAAGGGCGCTCTTAAGGATGTTATTTGGCCAGAAGGACCACACCTTT 1920 MAIWKGALKDVIWPEGPHLY 1921 ATAAGAAAGATGGATATTATTATCTTTTACATGCAGAAGCTGGCACAAGCTTTGAACATG 1980 KKDGYYYLLHAEAGTSFEHA 1981 CTATTTCTGTAGCTCGCTCAAAGGAGCTATTCAAATGGTTTGAGGGATGTCCTAGAAATC 2040 ISVARSKELFKWFEGCPRNP 2041 CTATATTTACCCATAGAAATTTAGGCAAGGATTATCCAGTATGCAATGTTGGACATGCTG 2100 IFTHRNLGKDYPVCNVGHAD 2101 ATTTAGTTGATGATATCAATGGCAACTGGTATATGGTGATGCTGGCATCTAGACCATGCA 2160 LVDDINGNWYMVMLASRPCK 2161 AGGGAAAGTGCAGCTTGGGACGAGAGACATTCCTTGCAAAAGTAATTTGGGAAGACGGAT 2220 GKCSLGRETFLAKVIWEDGW 2221 GGCCAGTGGTTAATCCGGGAGTTGGTCGTTTGACTGATGAGGTGGAGATGGACCTTCCTG 2280 PVVNPGVGRLTDEVEMDLPE 2281 AATATAGATTCTCAAAAGAGATTACTACAAAGGATAAAATGACCTTTGAAGAGACAGTCC 2340 YRFSKEITTKDKMTFEETVL 2341 TAGATGATAGATTTGTTGGAATTGAAAGAAGAAGTGAGGACTTTTATTCCCTTACTGACA 2400 DDRFVGIERRSEDFYSLTDN 2401 ATCCTGGATTCTTAAGATTAAAGCTTCGTCCTGAGGCCATAGAAAATACTGGCAATCCAT 2460 PGFLRLKLRPEAIENTGNPS 2461 CTTACTTAGGAATTCGTCAAAAGACTCATTCGTTTAGAGCAAGCTGTGGCCTTAAGTTTA 2520 YLGIRQKTHSFRASCGLKFT

PAGE 43

33 2521 CACCAGCAAAAGATAATGAATGTGCAGGAATGGTGTTATTCCAGAATAATGAAAATCACT 2580 PAKDNECAGMVLFQNNENHL 2581 TGGAGCTTTTAGTTGTAAAGAAGAAAGATAAGCTACAGTTTAAAGTAGGACCAGTTATTA 2640 ELLVVKKKDKLQFKVGPVIK 2641 AAGGAACCAAAATCAGACTTGCTACTTTTGATATTTCATCAGGTGATTTAGAAATTATTC 2700 GTKIRLATFDISSGDLEIIL 2701 TTGAGGCAGCAAATCAGCTGGCTAATATCTATATTAAAAAGAATAATGAAAAGATTCTTG 2760 EAANQLANIYIKKNNEKILV 2761 TGGCAGAATGTATTGATTTGAGCCCATACACTACCGAAGAATCAGGCGGATTCGTAGGAT 2820 AECIDLSPYTTEESGGFVGC 2821 GTACCATTGGACTATATGCTTCTTCAAATGGAAAGACCAGTGATAACTATTGCGATTATT 2880 TIGLYASSNGKTSDNYCDYS 2881 CCTACTTTACAGTAGAAGAAGTATAGCATTTTCAATGAGCGAATTTGCAAGTTTT ATATA 2940 YFTVEEV* 2941 CGGGATTAATTGTACGTAAAAACCATACAGGTGTAAAATAGTTTCCAGAGAAAGTTTTTT 3000 3001 CTCTGGAATTTTTTATTAT OOAGGG GATTATGCTTCAGGAAAGTATTAAGAAGTTGGTAC 3060 MLQES .IKKLVQ 3061 AGTACGGTATTGATATGGGGCTTACACCAGAATGTGAGAGAATATATACTACAAATCTTT 3120 YGIDMGLTPECERIYTTNLL 3121 TGCTTGAATGTATGAAAGAAGATGAGTACATAGATCCAGACTGTGATTTAAGCAATATTA 3180 LECMKEDEYIDPDCDLSNII 3181 TACTTGAAGATGTATTAAAGGAACTTTTAGATGAGGCAGTTAATAGAGGTATCATAGAGG 3240 LEDVLKELLDEAVNRGI lED 3241 ATTCAGTTACACATAGGGATTTGTTTGATACAAAGCTAATGAATCAGCTATGCCCACGTC 3300 SVTHRDLFDTKLMNQLCPRP 3301 CTAAACAGGTTATAGATGATTTTAACCGTATATACGATAACCATGGTCCAATAGCTGCAA 3360 KQVIDDFNRIYDNHGPIAAT 3361 CAGATTATTTTTACAAGTTAAGCAAAGCCTCTGACTATATCCGTACTTACAGGGTAAAAA 3420 DYFYKLSKASDYIRTYRVKK 3421 AGGACCT AAAATGGACATGCGATACAGAGTATGGCACTCTTGACATAACAATTAATCTCT 3480 DLKWTCDTEYGTLDITINLS 3481 CTAAGCCAGAAAAAGACCCAAAGGCAATTGCTGCAGCTAAGAATGCAAAACAATCCACAT 3540 KPEKDPKAIAAAKNAKQSTY 3541 ATCCGAAGTGCCAATTATGTATGGAAAATGAAGGCTATGCTGGTCGCATTAATCATCCTG 3600 PKCQLCMENEGYAGRINHPA 3601 CTAGAGAGAATCATCGCATAATTCCTATAACTATAAATAACAGCAACTGGGGATTTCAAT 3660 RENHRIIPITINNSNWGFQY 3661 ATAGCCCATACGTTTATTACAATGAGCATTGCATAGTCTTTAACGGAGAGCAT ACTCCT A 3720 SPYVYYNEHCIVFNGEHTPM 3721 TGAAAATAGAGCGAGCTACTTTTGTTAAGCTATTTGATTTTATCAAACTATTTCCACACT 3780 KIERATFVKLFDFIKLFPHY 3781 ATTTTTTAGGAAGCAATGCTGATTTACCAATTGTTGGAGGATCTATTTTAAGCCATGACC 3840 FLGSNADLPIVGGSILSHDH

PAGE 44

34 3841 ATTTCCAAGGCGGCCATTACACATTTGCCATGGAAAAAGCTCCAATTATTCAGGAATTTA 3900 FQGGHYTFAMEKAPI IQEFT 3901 CTGTAAAAGGATATGAGGATGTTAAGGCTGGTATAGTTAAATGGCCACTTTCAGTAATTA 3960 VKGYEDVKAGIVKWPLSVIR 3961 GACTTCAGTGCAAGGATGAGACTAGACTTATTGATTTAGCGACTAATAT ATTAGACAAAT 4020 LQCKDETRLIDLATNILDKW 4021 GGAGAAATTACACCGATGAAGAGGCATATATTTTTGCTGAAACAGATGGTGAGCCTCACA 4080 RNYTDEEAYIFAETDGEPHN 4081 ATACGATTACACCTATTGCTAGAAAAAGAGGGGATTACTTTGAACTAGATCCTCTAGAGT 4140 TITPIARKRGDYFELDPLES 4141 CGACCTGCAGGCATGCAAGCTTGGCACTGGCCGTCGTTTTACAACGTCGTGACTGGG* 4198 TCRHASLALAVVLQRRDW

PAGE 45

encodes a 517 amino acid protein having a calculated molecular weight of 58,421. 0RF3 , is located 12 3 bp downstream from 0RF2 and includes a Shine-Dalgarno sequence with an initiation codon 5 bp downstream. This ORF continues for 1,173 bp until the end of the clone and is also incomplete. No predicted "stemloop" structures or sequences that resemble rho independent transcriptional terminators were identified by computer analysis in the region between 0RF2 and 0RF3 . It is therefore unlikely that this DNA functions as a transcriptional terminator in E. coli. It is also unlikely that any transcriptional terminators are present between ORFl and 0RF2 since these two ORF's are only separated by 15 bp and 0RF2 is expressed in large amounts in E. coli in constructs that also contain ORFl. This evidence suggests that these three ORF's may constitute part of a xylan-degrading operon in B. f ibrisolvens . Codon usage . The codon usage of the three B. f ibrisolvens ORF's is summarized in Table 1. For comparative purposes, codon usage for B. f ibrisolvens 49 xynA (Mannarelli et al . 1990) and the average codon usage for E. coli (Allf-Steinberger, 1984) are included. The three B. f ibrisolvens ORF's have similar patterns of codon usage with each other and with strain 49 xynA . The low guanine plus cytosine content of B. f ibrisolvens is reflected in the three ORF's in the usage of an A or a T in

PAGE 46

36 TABLE 1. Comparison of codon usage frequency for the three B. f ibrisolvens ORF ' s . Frequency (mol Codon Usage Codon B. fibrisolvens E. coli B. f ibrisolvens ORF 1 ORF 2 ORF 1 xvlA Amino Acid Phe TTT 2.7 4.1 3.6 1.3 2.9 TTC 0.9 1.2 0.3 2.2 1.5 Leu TTA 2.0 2.5 2.8 0.7 0.7 TTG 0.2 1.2 1.0 0.9 0.5 CTT 5.6 2.7 2.5 0.8 4.4 CTC -0-00.3 0.8 -0CTA 1.4 1.0 1.8 0.2 0.2 CTG 0.5 0.8 0.3 6.8 1.0 He ATT 4.7 3.9 4.3 2.2 2.7 ATC 2.0 1.0 0.8 3.7 1.7 ATA 0.5 1.7 3.8 0.2 0.5 Met ATG 1.4 1.7 2.0 2.8 2.7 Val GTT 2.9 2.7 2.3 2.9 2.0 GTC 0.9 0.2 0.5 1.2 0.5 GTA 1.4 2.1 1.3 1.8 4.2 GTG 1.1 1.2 -02.2 1.0 Ser TCT 2.3 1.5 0.8 1.3 1.2 TCC -00.4 0.3 1.5 0.5 TCA 1.8 1.5 0.5 0.4 2.9 TCG -00.6 0.3 0.6 0.2 * AGT 0.9 0.8 0.3 0.3 -0AGC 0.7 1.0 1.8 1.4 1.2

PAGE 47

37 TABLE 1. (continued) Frecmency (mol %) Codon Usage Codon B. fibrisolvens E. coli B. fibrisolvens ORE 1 ORE 2 ORE 1 Amino Acid xvlA Pro CCT 1.4 1.7 1.8 0.5 1.2 CCC 0.2 -0-00.3 -0CCA 2.3 3.5 2.8 0.7 2.4 CCG -00.6 0.3 2.5 -0Thr ACT 0.7 1.9 2.3 1.1 0.7 ACC 0.7 1.4 0.5 2.4 1.0 ACA 2.7 1.7 3.1 0.3 5.1 ACG 0.5 -00.3 0.8 0.5 Ala GCT 2.7 2.1 2.8 2.6 1.5 GCC 1.1 0.4 0.8 2.2 1.2 GCA 2.0 2.3 2.0 2.3 4.1 GCG 0.5 0.2 0.2 3.2 -0Tyr TAT 4.3 3.5 2.8 1.0 3.9 TAC 0.2 1.2 2.5 1.5 3.4 His CAT 0.9 1.4 2.5 0.7 1.5 CAC 0.5 0.6 0.5 1.2 0.2 Gin CAA 0.7 1.0 1.3 1.0 0.5 CAG 1.6 0.6 1.5 3.2 2.2 Asn AAT 2.0 4.2 3.8 1.0 5.4 AAC 0.5 1.1 1.3 2.8 2.7

PAGE 48

TABLE 1. (continued Freauencv (mol % ) Codon Usaae Amino Acid Codon B. fibrisolvens E. coli B. fibrisolvens ORF 1 ORF 2 ORF 1 xvlA Lys AAA 3 . 4 3.1 3.6 4.1 3.4 AAG 4 . 7 3.5 3.3 1 . 3 3 . 9 Asp GAT 5 . 8 5.6 5.9 2 . 5 3 . 9 GAC 1 . 6 1.4 2.0 3 . 0 0 . 2 61u GAA 4 . 3 3.7 3.7 4 . 9 1 . 7 GAG 4 . 9 3.5 3.8 1 . 8 2 . 7 Cys TGT 1 . 4 1.4 1.0 0.4 0.7 TGC 0 . 7 1.0 1.5 0 . 5 0. 5 Trp TGG 1 . 3 1.7 1.3 0.7 2 . 0 Arg COT 0 . 7 0.8 1.3 3.1 1.0 CGC -00.2 0.5 2.0 0.2 CGA 0 . 2 0.2 0.2 0.2 -0CGG -0-0-00.2 -0AGA 3.1 2.9 2.0 0.1 2.0 AGG 0.5 -00.8 0.1 0.2 Gly GGT 3.4 1.7 1.5 3.8 2.4 GGC 2.3 2.1 1.0 3.1 1.5 GGA 3.1 4.4 1.5 0.4 2.7 GGG 0.5 0.2 0.8 0.6 0.5

PAGE 49

39 the wobble position with the exception of CAG for Gin, AAG for Lys, and GAG for Glu. Insertional inactivation and subclone analysis of 0RF2 . Subclones were generated to investigate the relationship between the xylosidase and arabinof uranosidase activities encoded by the xylB gene (Fig. 4b) . Retention or loss of enzymatic activity was initially asssayed on 4methylumbellif eryl-6-D-xylopyranoside (MUX) and 4methylumbellif eryl-a-L-arabinof uranoside (MUA) indicator plates. The Sspl fragment (1,843 bp) from pLOIlOOS, which contains 17 bp upstream and 274 bp downstream in addition to xylB . was subcloned in both directions in pUC18. Both activities were concurrently expressed only when xylB was cloned in the direction of transcription of the lac promoter (pLOI1043) , indicating a dependence on this promoter in E. coli . The insertion of a 10 bp NotI linker into the unique EcoRV site of xylB (pLOI1040) resulted in a frameshift mutation that abolished both enzymatic activities (Fig. 4b) . The results of the indicator plate assays were confirmed by comparing the specific activities for xylosidase and arabinofuranosidase in cell free extracts (Table 2.). Using p-nitrophenol derivatives, arabinofuranosidase activity was approximately 1.7-times higher than the xylosidase activity. The ratio of these activities was the same for the three active subclones. The original subclone, pLOIlOOS, exhibited the highest specific activity. The smaller active

PAGE 50

TABLE 2. Expression of enzyme activities in recombinant E. coli harboring xvlB . Sp act' plasmid xylosidase arabinofuranosidase Ratio ara/xyl*" pLOIlOOS 9.0 16.0 1.8 PLOI1040 0.2 0.2 PLOI1043 2.0 3.0 1.5 pLOIlOSO 6.8 10.1 1.5 pUClB 0.2 0.2 ' Nanomoles per minute per milligram of cell protein. ^ Ratio calculated after subtraction of background values from the pUClS control.

PAGE 51

41 subclone, pLOI1043, exhibited a four-fold decrease in both activities but retained a similar ratio of arabinofuranosidase to xylosidase activities. An additional subclone, pLOI1050, contained two Sspl fragments each harboring xylB oriented with the direction of transcription from the lac promoter. This subclone exhibited a three-fold increase in both enzymatic activities with respect to the single insertion (pLOI1043) but less than the original clone. Again the ratio of the two activities remained essentially the same as the wild type (pLOIlOOS) . The results of these experiments demonstrate the dependence of both enzymatic activities on 0RF2 . Presence of other qlycosidic activities . The presence of additional hydrolytic activities was examined in the xylB -encoded protein using various orthoand paranitrophenyl glycosidic substrates (Table 3). No additional activity above the background levels was detected with 12 other pentose and hexose derivatives. Low levels of activity was detected against the o-NP-B-D-fucopyranoside. This may not be significant since the activity represents less than 5% of the activity against the B-D-xylopyranoside. A 19-fold higher activity was detected against onitrophenyl-6-D-xylopyranoside relative to the parasubstituted derivative. This phenomenon is analogous to B-galactosidase from E. coli. The paraand orthosubstituted substrates are known to have different rates for

PAGE 52

42 TABLE 3. Hydrolysis of different nitrophenyl-substituted glycosides by the xvlB gene product. substrate Specific activity' pLOIlOOS pUClS p-NP-B-D-xylopyranoside 8.9 0.2 p-NP-a-L-arabinofuranoside 15.5 0.5 p-NP-a-L-arabinopyranoside 0.1 0.2 p-NP-a-D-galactopyranoside 0.2 0.2 p-NP-a-D-glucopyranos ide 3.5 3.3 p-NP-a-L-fucopyranoside 0.2 0.2 p-NP-fl-D-fucopyranoside 0.2 0.2 p-NP-fl-L-fucopyranoside 0.4 0.4 p-NP-a-L-rheunnopyranoside 0.3 0.2 o-NP-13-D-fucopyranoside 1.4 1.0 o-NP-a-D-galactopyranoside 0.8 1.0 o-NP-B-D-galactopyranoside 1,0 1.0 ' Nanomoles per minute per milligram of cell protein.

PAGE 53

the glycosidic bond-breaking step (Martinez-Bilbao et al . , 1991) . Under the conditions of these assays it appears that the xylB -encoded protein is limited to hydrolytic activity against 6-D-xylopyranosides and ot-L-arabinofuranosides only. Electrophoretic analysis of cloned proteins . Using SDS-PAGE analysis of cell-free cytoplasmic extracts, a new protein band with an apparent molecular weight of 60,000 was observed in cells harboring plasmid pLOIlOOS (Fig. 6) . This band was absent in extracts from cells containing the vector plasmid pUC18 alone. Extracts from which the gene was inactivated by a frameshift mutation (pLOI1040) also lacked this protein band. The observed levels of this protein band in the single (pLOI1043) and double (pLOIlOSO) Sspl subclones was consistent with the presence of the enzyme. Primary sequence homology comparisons . Homologies of the B. f ibrisolvens ORF's to other glycohydrolases were compared to determine evolutionary relatedness. The translated amino acid sequences of the three B. f ibrisolvens ORF's exhibited 42 to 45% similarity (a conservative match) and 14 to 19% identity (an exact match) with each other (Table 4.). The xvlB was found to be most similar (44% similarity, 20% identity) to the 6-glucosidase from Kluvveromvces fraqilis . Additional comparisons with other glycohydrolase sequences revealed no significant amino acid identities. Although the N-terminal sequence of the

PAGE 54

Figure 6. SDS-PAGE analysis of cytoplasmic extracts from recombinant E. coli DH5a harboring selected plasmids. Approximately 20 )ug of protein was loaded in each lane. Lanes 1 and 7; molecular weight markers, lane 2; pLOIlOOS, lane 3; pUC18, lane 4; pLOI1040, lane 5; pLOI1043, lane 6; pLOIlOSO. The band cooresponding to the xylosidase-arabinof uranosidase enzyme is indicated by an arrow. The numbers in the right represent the apparent molecular weight of the standards (X 10^) .

PAGE 55

45 1 2 3 4 5 6 7

PAGE 56

46 TABLE 4. Comparison of the translated amino acid sequences of the three B. fibrisolvens ORF's in pLOI 1001 with those of selected proteins. Organism (gene) wKr ± ORF 2 ORF 3 xylB) Reference B. fibrisolvens (xvnA) 40 (16) 41 (16) 41 (13) Mannarelli et al. 1990 B. fibrisolvens (endl> 16 (8) 14 (6) 20 (11) oercjer et ai • lyoy B. tiumiluH / wnA \ AC. / 5 1 \ 44 (16) 12 (8) Fukusaki et al. 1984 B. pumilus (xvnB) 39 (16) 28 (21) 38 (17) Morivama et al. 1987 B. subtilis (xynA) 48 (21) 44 (16) 12 (6) Paice et al. 1986 Caldocellum saccharolvticus (xvnBl 20 (14) 25 (12) 21 (12) Luthi et al. 1990 (xvnA / xvnB) 16 (8) 20 (9) 19 (10) Luthi et al. 1990 C. saccharolvticus ( xvnC ) 16 (8) 22 (11) 31 (3) Luthi et al. 1990 Clostridium thermocellum 18 (7) 40 (16) 15 (8) Grepinet et al. 1988 (xynZ) C. thermocellum ( celA) 45 (22) 41 ( lo ) 39 (15) Benquin et al. 1985 C. thermocellum (celB) 44 (20) 44 (18) 43 (19) orepinet et al. 1986 C. thermocellum (celD) 45 (16) 39 (15) 43 (18) Joliff et al. 1986 Asoeraillus niaer (a-eunylase) 40 (16) 41 (18) 43 (19) Boel et al. 1984 Kluvveromvces f raailis (i3-glucosidase) 52 (31) 44 (20) 46 (22) Raynal et al. 1987 ORFl 100 44 (19) 45 (19) This study 0RF2 44 (19) 100 42 (14) This study ORF3 45 (19) 42 (14) 100 This study

PAGE 57

47 Bacillus pumilus S-xylosidase did exhibit strong amino acid identity in selected regions, the overall identity was only 21%. Thus the B. f ibrisolvens xylB gene is evolutionarily divergent from other glycohydrolases. The translated, primary sequences for ORFl and 0RF3 also exhibited similarity (52% and 46%, respectively) and identity (31% and 22%, respectively) to the K. fraqilis 6-glucosidase. This is consistant with these two ORF's also beinq involved in carbohydrate deqradation. It has been postulated that the hydrolytic mechanism of lysozyme (Teeri, et_al. 1987) and cellulases (Knowles et al., 1987) can serve as a model for other carbohydratehydrolyzing enzymes. Studies of hen egg-white lysozyme (HEWL) indicate a general acid-base catalytic mechanism involving Glu-35 and Asp-52 as the catalytic residues (Quiocho, F. A., 1986). Subsequent studies have demonstrated that this catalytic region is conserved in some cellulases (Knowles et al . 1987) . An analysis of the translated primary sequence from xvlB reveal a reqion homoloqous to the active site reqion from HEWL, and qlucoamylase from Aspergillus niqer (Table 5) . The conserved reqion from additional carbohydrate hydrolases are included for comparison. The xylB reqion was most similar to the qlucoamylase, with 38% identity between the amino acids in the catalytic reqion. The catalytically important

PAGE 58

48 TABLE 5. Amino acid sequence alignment of conserved regions. Protein' Concensus sequence Reference 35 44 52 HEWL FESNFNTQAT.NRNTDGS. . .TDY Yaguchi et_al. 1983 331 338 349 A. n. (a-aml) PEDTY.YNG. .NPWFLCTLAAAEQ Boel et al . 1984 342 350 362 B. f. xvlB | SEDFYSLTD. .NPGFLRLKLRPEA This study 144 149 160 B.f. xvlB-, PDGVRY NGAWEIWVQELDL This study 320 328 337 B.f. endl GETSATNRN. .NTAERVKWA. .DY Berger et al . 1989 355 362 369 B.f. xvnA NEKPLIWS. . .NIGVAKPAY. .DE Mannarelli et al . 1990 769 785 B.f. bqll SDWWGFGEHYK BVLAGNDI Barnett et al . 1991 325 335 343 B.p. xvnB lECTRLAQLNWNTCSMQFV. . .EE Moriyama et al . 1987 ' Abbreviations: Hen egg-white lysozyme (HEWL), Aspergillus niger a-eunylase (aami)/ B. fibr isolvens B-D-xylosidase ( xvlB> . B. fibrisolvens endoglucanase 1 (endl), B. fibrisolvens B-D-xylanase ( xvlA ) . B. fibrosolvens fl-glucosidase ( bqll ) , Bacillus pumillus B-D-xylosidase ( xvlB > . Clostridium thermocellum B-D-xylanase ( xvnZ ) . C. thermocellum cellobiohydrolase B (celB) , C. thermocellum cellobiohydrolase D (celD) , C. cellu lolvticum endoglucanase A (EGCCA), Caldocellum saccharolvticum B-Dxylosidase ( xvnB ) . Cellulomonas fimi endoglucanase A (cenA) , C. fimi exoglucanase A (cex), Trichoderma reesei cellobiohydrolase II (CBH II), Kluweromvces fraailis cellobiohydrolase I (CBH I), Schizophvllum commune endoglucanase (EG I).

PAGE 59

V 49 TABLE 5. (continued) Protein* Concensus sequence Reference 458 466 475 C.t. xvnZ GEALLRADV. .NRSGKVDS. .TDY Grepinet et al . 1988 418 427 436 C.t. celB TEGGHPLLDL.NLKYLRCMR. .DF Grepinet and Benguin, 1986 376 384 395 C.t. celD DEEYLRDFE. .NRAAQFSKKEADF Joliff et al . 1986 408 418 427 C.8. xvnB REVFVERIDEYNANPKRVWL. .EM Luthi et al . 1990 244 254 263 T.r. CBH II LECINYAVTQLNLPNVAMYL. .DA Rouvinen et al . 1990 586 596 605 K.f. CBH I GEWETEGYDRENMDLPKRTN. .EL Raynal et al . 1987 33 42 50 S.c. EG I NESCAEFGNQ.NIPGVKN...TaY Yaguchi et al . 1983 * Abbreviations; Hen egg-white lysozyme (HEWL), Aspergillus niger a-amylase (aami), B. fibrisolvens fl-D-xylosidase ( xvlB) . B. fibrisolvens endoglucanase 1 (endl), B. fibrisolvens B-D-xylanase ( xvlA ) . B. fibrosolvens fl-glucosidase (ball), Bacillus pumillua fl-D-xylosidase ( xvlB i , Clostridium thermocellum B-D-xylanase ( xvnZ ) , C. thermocellum cellobiohydrolase B (celB) , C. thermocellum cellobiohydrolase D (celD) , C. cellulolvticum endoglucanase A (EGCCA), Caldocellum saccharolvticum fl-Dxylosidase ( xvnB ) , Cellulomonaa fimi endoglucanase A ( cenA ) , C. fimi exoglucanase A (cex), Trichoderma reesei cellobiohydrolase II (CBH II), Kluvveromvces fraoilis cellobiohydrolase I (CBH I), Schizophvllum commune endoglucanase (EG I).

PAGE 60

50 glutamic and aspartic acid residues and the approximate spacing were found to be conserved conserved. Conclusions The xvlB gene, encoding 6-D-xylosidase and a-Larabinofuranosidase activities is the first of its kind to be sequenced. The xylB gene is 1,551 bp in length and encodes a 517 amino acids protein having a predicted molecular weight of 58,000. The absence of any significant stem-loop structures or terminators in the regions between ORFl, 0RF2, and 0RF3 as well as the strong expression of 0RF2 in E. coli even when preceded by ORFl suggests that these three genes may constitute a xylan-degrading operon. The subcloning analysis and insertional inactivation studies demonstrate the dependence of both activities on the intact xylB gene. The codon usage of the three ORF's is consistent with the low guanine plus cytosine content of this organism in general (Mannarelli et al . 1990) . This enzyme exhibited 6-D-xylopyranosidase and a-Larabinofuranosidase activities. No additional glycosidic bond cleavage activities were detected in the xylB gene product . The xylB gene displayed limited homology to other reported xylosidase sequences and must therefore be considered to be evolutionarily divergent from genes encoding similar functions from other organisms. It did

PAGE 61

however exhibit partial identity with the 6-glucosidase from Kluyveromyces fragilis which is consistent with the similarity of the substrates which these two enzymes attack. A single gene encoding xylanase/xylosidase activities has been cloned and sequenced from Caldocellum saccharolyticum (Luthi et al . 1990) . This protein, however, lacked arabinofuranosidase activity. Recently, a xylosidase/arabinofuranosidase gene was proposed to reside in a gene cluster isolated from Bacteroides ovatus (Whitehead and Hespell, 1990). All clones exhibited both activities concurrently and both activities co-purified. Additionally, an enzyme having xylosidase and arabinofuranosidase activities has been purified from Trichoderma reesei (Poutanen and Puis, 1988) . No sequence, however, for the encoding gene has been reported. Substrate ambiguity between carboxymethylcellulase and xylanase enzymes is relatively common (Flint et al . 1989) . The substrate ambiguity for other xylosidase enzymes has also been reported. By employing kinetic methods on the purified, bifunctional 6-xylosidase/6-glucosidase from Chaetomium trilaterale . Uziie et al . (1985) demonstrated that this enzyme possessed a single active site with dual substrate-binding regions. More recently, a neopullulanase from Bacillus stearothermophilus was cloned, sequenced, and characterized (Kuriki and Imanaka, 1989). This enzyme possessed activity against a(1, 6) -glycosidic linkages in

PAGE 62

52 addition to the usual hydrolysis of a(1 , 4) -glycosidic linkages. Mutational analysis demonstrated that a single active center was involved in the catalysis of both these linkages (Kuriki et al . 1991) . It seems reasonable to speculate that bif unctionality and substrate ambiguities among the microbial carbohydrate hydrolases is common. The celB gene encoding a "true" bifunctional cellulase has been cloned from Caldocellum saccharolyticum and sequenced (Saul et al. 1990) . This enzyme exhibited both endo-glucanase and exo-glucanase activities. The endo-glucanase activity was localized to the carboxy terminal domain and the exo-glucanase activity was localized to the amino terminal domain. This protein also exhibited homology with both endoand exo-glucanase enzymes from other organisms. The organization of separate functions to separate domains has also been demonstrated with the endo-glucanase 2 from Bacteroides succinoqenes (McGavin and Forsberg, 1989) . These investigators used protease treatments to demonstrate that this enzyme possessed separate substrate binding and catalytic domains. The structural similarities between the various 6and a-linked glycosyl residues may be responsible for the apparent evolution of enzymes with broad substrate specificity.

PAGE 63

53 Since many enzymatic activities are required to completely depolymerize xylans and cellulose, the evolution of such enzymes could represent a selective advantage in the rumen and other environments.

PAGE 64

Chapter IV Mutational analysis of the xylB gene Introduction Carbohydrate-degrading enzymes have been studied extensively in the microbial world and form the basis of much of what we know about the cycling of carbon in the environment (Weinstein and Albersheim, .1979) . Lysozyme, an enzyme which hydrolyzes bacterial cell wall carbohydrates, was one of the first such enzymes to be s'Cudied extensively and as a result much is known about this enzyme's catalytic mechanism and structure (Quiocho, 1986) . The cellulases have also been extensively studied as they are responsible for the cycling of the most abundant natural polymer cellulose (Knowles et al . . 1987). Studies of the mechanistic properties of these enzymes have been facilitated by the use of molecular genetic techniques. Gene cloning, sequencing, and oligonucleotide-directed mutagenesis have allowed mutations to be made in a site specific manner. A kinetic study of the mutant proteins can then be done and predictions about catalytic mechanisms tested. 54

PAGE 65

55 The active center for a neopullulanase from Bacillus stearothermophilus has been recently examined using mutagenesis (Kuriki et__al. , 1991) . This enzyme exhibits dual activities against a(1,4) and a(1,6) glycosidic linkages. The catalytically important amino acids were tentatively identified using sequence alignment and homology searches. The putative catalytic amino acids were changed using site-directed mutagenesis and activities were examined in the resulting mutants. This approach identified that one active center containing Glu-357 and Asp-424 was responsible for both catalytic activities. Enzymes which exhibit substrate ambiguity are interesting both from mechanistic and evolutionary perspectives. The obvious question that arises with respect to bifunctionality is does this enzyme have two separate, specialized, catalytic sites? This situation would imply that the protein has evolved from a gene fusion to perform two separate functions. Another possibility involves the presence of a single active site in which two structurally similar substrates are bound and hydrolyzed. This situation would be analogous to a case of mistaken substrate identity which proves advantagous to the organism, and has been evolutionarily conserved. Examples of multifunctional xylosidases have been reported in the literature and are discussed in the previous chapters. The apparent bifunctionality of the xvlB gene was demonstrated in chapter

PAGE 66

56 III. A series of genetic experiments were designed to investigate the presence or absence of two catalytic or functionally separate domains on the xylB gene that are responsible for the dual activities exhibited by this enzyme. Materials and Methods Medium and growth conditions, genetic methods and DNA sequencing were done as described in chapter III. In vitro nitrous acid mutagenesis of xylB . A total of 80 of pLOIlOOS which contains the xylB gene was resuspended in 50 /xl Tris-EDTA (TE) buffer (pH 8.0). Mutagenesis was initiated by the addition of 10 nl of 2.5 M sodium acetate (pH 4.3) and 50 /xl 2.0 M sodium nitrite. Exposure times were zero, thirty seconds, one, two five and ten minutes. Mutagenesis reactions were terminated by the addition of 200 nl 100% ethanol. The precipitation step was repeated twice to ensure the complete removal of the mutagenic agent. The mutagen-treated plasmids were resuspended in 80 /il TE buffer (pH 8.0). A total of 5 /xl of the plasmid was transformed into competent E. coli DH5a. Serial ten-fold dilutions of the transformed cells were plated in triplicate onto Luria agar supplemented with the fluorogenic substrates. A 99% reduction in transformation by the mutagenized plasmid was observed after ten minutes of mutagenesis.

PAGE 67

57 Localization of point mutations . The entire xylB coding region was divided into five domains based upon restriction sites (Fig. 7) . Three restriction fragments, the PstI . EcoRI . and Hindlll fragments were isolated from each mutant plasmid. These fragments were used to replace the corresponding fragment in the wild type gene which had been modified to construct receiving vectors for each respective restriction fragment (pLOIlOSl, pLOI1052, and pLOI1053) to test the functionality of individual fragments. This strategy, outlined in Fig. 8, allowed the localization of point mutations to one or more of the five domains. Exonuclease III deletion of xylB . Plasmid pLOI1043 containing the xylB coding region was deleted from the 3 ' terminal region into the coding region by exonuclease III using the "Erase-a-Base" deletion kit (Promega Corporation, Madison, WI.) according to the manufacturer's instructions. The deleted plasmids were subsequently transformed into competent E. coli DH5a and screened in the same fashion as the in vitro -aenerated mutant plasmids. 5' deletion analvsis and lacZ' fusions . The internal PstI fragment from pLOIlOOS was subcloned in the original orientation into plasmid pUC18. This in-frame fusion with lacZ results in the subsequent deletion of the 5' terminal 54 base pairs (18 amino acids). Prepara tion of cell extracts . Extracts were prepared as described in chapter III.

PAGE 68

Figure 7. The assignment of domains to the xylB gene (solid bar) . Restriction endonuclease fragments used to localize mutations are shown below. Abbreviations: E; EcoRI . S; Sspl . P; Pstl . H; Hindlll .

PAGE 70

Figure 8. Subcloning strategy used to localize in vitro mutations to one of five domains on the xylB gene. Abbreviations: E; EcoRI . S; Sspl . P; PstI . H; Hindlll .

PAGE 71

PstI PLOI1053

PAGE 72

62 Kriz yme assays . Enzyme assays were done as described in chapter III. Sodium dodecvl sulfate poly anryl amide gel electrophoresis fSDS-PAGE) . SDS-PAGE gels were done as described in chapter III. Native polvacrvlamide ael ele ctrophoresis (native PAGE) . Cell proteins were separated in non-denaturing gels by the method of Ornstein and Davis (1964). Following electrophoresis, gels were equilibrated in 50 mM sodium phosphate buffer (pH 6.8). The equilibrated gels were then overlaid with Whatman #1 filter paper soaked with a solution of 20 mg per ml of the fluorogenic substrates, 4methylumbelliferyl-B-D-xylopyranoside or 4methylumbelliferyl-a-L-arabinofuranoside in 70% ethanol. Overlays were incubated at 37°C for 15 min or until activity bands were visible under long-wave UV light. Native-PAGE gels were also stained for protein as outlined before. Western hybridization analvsis of wild type and mutant proteins . Native and SDS-PAGE protein gels were electroblotted using the Trans-Blot apparatus (BioRad Laboratories, Richmond, CA.) according to the manufacturer's instructions. "Western" hybridizations were done using polyclonal antisera raised to E. coli DH5a cell extracts harboring pLOI1005 in rabbits. Protein bands were visualized using alkaline phosphatase conjugated goat anti-

PAGE 73

63 rabbit antisera. All procedures and conditions used have been described elsewhere (Harlow and Lane, 1988) . Native agarose-xylan gel electrophoresis . Separating gels consisted of 0.75% agarose with or without the inclusion of 0.75% birchwood xylan. Both the gels and the running buffer were standard Tris-borate-EDTA buffer (TBE-pH 8.0). Approximately 40 fig cell protein was added per well. Proteins were electrophoresed at 75 V and 21 mA in a horizontal electrophoresis unit until the dye front reached a point 1 cm from the end of the gel. Agarose gels were either directly stained for activity (zymograms) or electroblotted and analyzed immunologically as outlined above . Results and discussion Deletion analysis . Exonuclease III deletions from the 3 ' end of xylB resulted in the concurrent loss of both enzymatic activities in all deletions (Fig. 9a) . DNA sequencing of the deletion end points identified a minimal deletion of the terminal 27 base pairs (9 amino acids) resulted in the loss of both enzymatic activities. A "TAA" termination codon in the pUC18 polylinker immediately downstream from this deletion served to define the new 3 • end of this mutant gene. A LacZ ' fusion of the internal PstI fragment of xvlB resulted in the deletion of the 5' terminal 54 base pairs (18 amino acids) including the Shine-Dalgarno sequence and

PAGE 74

Figure 9. Deletion analysis of xylB gene. (A) Exonuclease III deletion series from the 3' end of xylB . Shaded arrow denotes direction of transcription from the pUC18 lac promoter. Solid arrow denotes direction of deletion. Underlined "taa" indicates relative location of stop codon from the 3 ' end of deletion. Retention or loss of respective enzyme activites is indicated by a "+'• or (B) LacZ' fusion of 5' end of the large internal PstI fragment of xylB resulting in a deletion of 56 base pairs.

PAGE 75

65 B kilobase pairs 0.5 1.0 1.5 Sail PstI SphI LacZ' = deletion taa xvl/araf +/+ (undeleted) -/(27 bp./9 aa) -/-/-/S ] ^ 1 i ] E i 3 (56 bp/18 aa) +/+ -/-

PAGE 76

66 the "ATG" initiation codon with the loss of both enzymatic activities (Fig. 9b) . In vitro mutagenesis . Twelve mutants were isolated using selective media on the basis of reduced or abolished enzyme activities. Two mutants could not be classified due to multiple mutations and were not analyzed further. A total of ten mutants were classified as negative for both enzymatic activities against the fluorogenic substrates based upon agar plate assays. Two mutants, number six and number ten, displayed reduced but significant fluorescence on fluorogenic indicator plates (Table 6) . Localization of point mutations . Most of the point mutations were determined to reside in domains II and III (Table 6) . None of the mutations were localized in the lacZ ' promoter of pUC18. The generation of point mutations is consistent with the deamination activity of nitrous acid. The point mutations were clustered in domains II and III near the region proposed to be the active site of this enzyme. DNA sequencing of in vitro mutants . The amino acid substitutions resulting from in vitro mutagenesis were deduced from the base changes as determined by DNA sequencing (Fig. 10) . The point mutations were all AT to GC transitions which is consistant with the mode of action for nitrous acid mutagenesis. Two frameshift mutations were identified, in domain I and domain V. These were not

PAGE 77

67 Table 6. Localization of point mutations by restriction fragment replacement analysis. MUTANT PHENOTYPE PstI EcoRI Hi ndTTT 1 -/-/+/ + II fs^ 2 -/-/-/+/ + II 3 -/-/-/+/ + II 4 -/-/-/III 5 -/+/+ -/+/ + I 6 w/v/* w/w +/+ +/ + II 7 w/w w/w +/+ +/ + II 8 -/-/-/+ / + V fs 9 -/+ / + II 10 -/-/-/-/III 11 -/' -/-/-/III 12 -/-/-/-/III denotes presence or absence of xylosidase activity. ^ denotes presence or absence of arabinof uranosidase activity. ^ denotes a frameshift mutation. ^ denotes "weak" activity.

PAGE 78

Figure 10. Location and identification of point mutations in xylB by DNA sequencing. Numbers above sequence indicate the position of the amino acids. Amino acids in parentheses below the sequence indicates the new amino acid inserted by mutation. A "-" denotes loss of enzymatic activity. A "w" denotes weak enzymatic activity.

PAGE 79

69 Region of autation 180 186 192 Asp . Val . I le . Trp . Pro. Glu . Gly . Pro. His . Leu . Tyr . Lys . Lys (Arg) 197 203 209 Tyr.Leu.Leu.His.Ala.Glu.Ala.Gly.Thr.Ser.Phe.Glu.His (Thr) 204 210 216 Gly.Thr.Ser.Phe.Glu.HiB.Ala.Ile.Ser.Val.Ala.Arg.Ser (Val) 1 7 13 Met . Val . I le . Ala . Asn . Asn . Pro . lie . Leu . Lys . Gly . Phe . Tyr (Leu) 172 178 184 Ala . He . Trp . Lys . Gly . Ala . Leu . Ly s . Asp . Val . I le . Trp . Pro (Phe) 1'*^ 152 158 Val . Arg . Tyr . Asn . Gly . Asp . Trp . Glu . He . Trp . Val . Gin . Glu (UGA) 197 203 209 Thr . Leu . Leu . His . Ala . Glu . Ala . Gly . Thr . Ser . Phe . Glu . His (Val) 232 238 244 Phe . Thr . His . Arg . Asn . Leu . Gly . Ly s . Asp . Tyr . Pro . Val . Cys (Asp) 204 210 216 Gly . Thr . Ser . Phe . Glu . His . Ala . He . Ser . Val . Ala . Arg . Ser (Thr) ^^^ 203 209 Thr . Leu . Leu . His . Ala . Glu . Ala . Gly . Thr . Ser . Phe . Glu . His (Thr) 23 Ser. ILe. Cys. Arg. Lys. Gly. . . Phenotype xyl/ara -/Mutant G186R -/-/-/w/w ' t . w/w -/-/-/-/-/A203T A210V P7L L178F W158UGA A203V G2380 A210T A203T fs

PAGE 80

70 Studied further. Ten mutations are clustered in an area of 60 amino acids. Six of these mutations; glycine 185 to arginine 185, alanine 202 to threonine 202, alanine 202 to valine 202, alanine 209 to valine 209, glycine 237 to aspartate 237, and alanine 209 to threonine 209 all resulted in an inactive protein. Each of these mutations represent nonconservative amino acid changes that would be expected to effect the function and/or conformation of the protein. Another mutation in this cluster, leucine 178 to phenylalanine 178, resulted in a mutant having one-tenth the enzymatic activities of the wild type. This is a conservative change in that both leucine and phenylalanine are hydrophobic amino acids and have similar structural properties. It is possible that the larger aromatic group on phenylalanine is affecting substrate binding and/or catalysis. In addition, the substitution of a "UGA" stop codon for tryptophan at position 152 also resulted in a protein with one-tenth the enzymatic activities of the wild type. The most probable explaination is that the "UGA" codon is functioning as tryptophan in this E. coli strain and the efficiency of read through is very low which yields a reduced expression of the xylB protein. The existence of such suppressor mutations has been previously demonstrated (Hirsh, 1971) . One mutation, proline (7) to leucine (7) occurred in the amino terminus of the protein and resulted in a negative phenotype. The amino terminus of the xylB

PAGE 81

71 protein therefore plays an important in the structural and/or catalytic role. SDS-PAGE analysis of mutant proteins . The presence of the 60,000 molecular weight monomer ic subunit encoded by xylB was confirmed for the various mutants (Fig. 11) . In two cases, cell extracts from recombinants harboring the proline (7) to leucine (7) and glycine (185) to arginine (185) , did not contain the xy IB -encoded protein band. The insertion of the plasmid containing the proline (7) to leucine (7) mutation into a Ion -negative strain of E. coli, which is deficient in serine proteases, resulted in the restoration of the xy IB -encoded protein band on SDS-PAGE gels. It seems likely that proteolysis of an improperly folded protein is responsible for the absence of this protein in cell extracts of these two mutants. Native polyacrylamide gel comparisons of mutant and wild type proteins . Zymograms of the wild type and mutants L178F and W158UGA proteins indicated that the mutant proteins are approximately the same size as the wild type protein (Fig. 12). Western hybridizations of blotted protein bands from native-PAGE indicated that all of the expressed mutations result in proteins that have unaltered electrophoretic mobilities and subunit assemblies relative to the wild type protein (Fig. 13). This evidence suggests that the point mutations that result in expressed protein do not induce any destabilizing secondary structural

PAGE 82

Figure 11. SDS-PAGE analysis of wild type and mutant proteins. Lane assignments: A and J; wild type (pLOIlOOS) , B; pUC18, C; G186R, D; A203T, E; A210V, F; P7L, G; L178F, H; W158UGA, I; molecular weight markers, K; A203V, L; G238D, M; G238D expressed in a Ion" strain of E. coli N; A210T, 0; A203T. Molecular weight marker sizes (X 10^) : 1; 94, 2; 67, 3; 43, 4; 30, 5; 20.

PAGE 83

73

PAGE 84

Figure 12. Native-PAGE comparison of W158UGA and L178F mutants proteins with the wild type (pLOIlOOS) stained with Coomassie blue. Lane assignments: A; mutant L178F, B; pUClB control, C; W158UGA, D; pLOIlOOS.

PAGE 85

BCD

PAGE 86

Figure 13. Western hybridization of native-PAGE of wild type and mutant proteins. Lane assignments: A; A203T, B; pUC18, C:; P7L, D; A210V, E; L178F, F, W158UGA, G; pLOIlOOS.

PAGE 88

78 perturbations that would result in proteolysis by the cell or a major change in tertiary structure of the proteins. The reduced levels of the xylosidase protein in the W158UGA mutant correlates to the reduced enzymatic activities for recombinants harboring this mutant. The "UGA" termination codon can be decoded as a tryptophan at low efficiency in E. coli (Hirsh, 1971) . It is likely that this is the case also for the W158UGA mutant in xylB . Substrate-binding comparisons of mutant proteins . Electrophoresis of the wild type and mutant proteins on native gels that contained agarose alone and agarose plus birchwood xylan indicated a differential mobility between the wild type and mutant proteins (Fig. 14) . Using agarose alone no differences between electrophoretic mobilities of the respective mutant proteins and the wild type were detected using Western hybridization analysis. The inclusion of birchwood xylan (0.75 %) resulted in a change in the mobility of the proteins. Without exception, all the mutant proteins exhibited faster electrophoretic mobilities relative to the wild type protein. It is possible that the xylan is functioning as a psuedo-substrate and the point mutations have affected the relative affinities of these proteins for the substrate.

PAGE 89

Figure 14. Substrate binding native gel Western hybridization assays of wild type and mutant proteins. Lane assignments: A , F, and G; native (pLOIlOOS) , B; G186R, C; A203T, D; W158UGA, E; L178F, G; A203V, H; G238D, I; A210T, J; A203T. Arrow indicates the direction of protein migration. The "+" and "-" indicated the relative location of the anode and cathode, respectively.

PAGE 90

80

PAGE 91

81 Analysis of enzvmatic activities of expressed mutant proteins . In all cases, the point mutations affected both enzymatic activities concurrently (Table 7). All mutations resulted in an expressed phenotype in which enzymatic activities were reduced or abolished. The clustering of these mutations in the 60 amino acid region (12% of the coding region) which contains the catalytic consensus sequence is evidence that the two enzymatic activities expressed by this protein are not functionally confined to separated domains. There is a dependance of function relating both enzymatic activities to this region of the protein. The effects of substrate concentration on reaction rate of the mutant enzymes was investigated using the crude extracts as a source of protein (Table 8) . Mutants A203T, A210V, L178F, G238D, and A210T exhibited an increase in apparent increase in reaction rate relative to increasing substrate concentration. Increasing substrate concentration had no effect on reaction velocity for mutants G186R and A203V, however the activities for these two mutant proteins were above that for the pUC18 background. It is possible that the lowest concentration of substrate used, 3 mM, is at saturation with respect to these two mutant proteins. 4

PAGE 92

82 Table 7. Enzymatic activities of recombinants harboring point mutations on xvlB relative to the wild type protein. Clone Xyl Sp.Ac. AJLCL op • AC* • R = pLOIlOOS 10.5 17.4 1.7 . 46 G186R 0.1 0.1 1.0 .51 A203T 0.1 0.3 3.0 .49 A210V 0.2 0.2 1.0 .49 P7L 0.2 0.2 1.0 N/D L178P 1.2 1.8 1.5 .53 W158AUG 2.1 2.9 1.4 .46 A203V 0.2 0.2 1.0 .49 G238D 0.2 0.2 1.0 N/D A210T 0.2 0.1 0.5 .50 pUC18 0.0 0.0 ^ Specific activity in nmoles p-nitrophenol released per min per mg protein. Ratio computed after subtraction of pUC18 background values. Determined by comparing relative migration distances of each protein verses that for the dye front on agarose/xylan native gel electrophoresis using "Western" hybridization to visualize protein bands.

PAGE 93

83 TABLE 8. Effects of substrate concentration on 6-Dxylosidase activity* for xylB in vitro mutants''. Clone 3 mM 6 mM 9 mM Km,_^ app p-nitrophenyl -6-Dxylopyranoside wild type 18.6 24.0 25.6 4 mM G186R 0.4 0.4 0.5 2 mM A203T 0.4 0.7 0.9 11 mM A210V 0.8 0.9 1.2 2 mM L178F 1.4 1.8 3.0 18 mM A203V 0.6 0.8 0.8 3 mM G238D 0.6 1.1 1.3 7 mM A210T 0.4 0.7 0.8 12 mM * Specific activity expressed as nmoles p-nitrophenol released per min per mg protein. Does not include the frameshift mutations or those mutant proteins that are not expressed. Apparent Km values determined using the direct linear method of Cornish and Bowden.

PAGE 94

84 Conclusions The mutation and deletion data are consistant with the proposal that both enzymatic activities exhibited by the xylB -encoded protein are not functionally separate but depend upon the same region of the protein for complete activity. The clustering of mutations about the consensus sequence and the 60 amino acids region is strong evidence in favor of this region being important for substrate-binding and/or catalytic activity of the protein. This region of the protein is rich in aspartic acid, glutamic acid and histidine residues which have been previously implicated in the catalytic function of related proteins such as lysozyme (Quiocho, 1986) , taka-amylase from Aspergillus niger (Matsuura et al., 1984), cellobiohydrolase II from Trichoderma reesei (Rouvinen et al., 1990), and a neopullulanase from Bacillus stearothermophilus (Kuriki et al., 1991). The point mutations that were expressed as complete proteins did not affect either subunit assembly or the apparent size of the native protein relative to the wild type. These mutations did, however, change the protein's mobility during electrophoresis on agarose-xylan native gels. It is possible that this change in mobility is due to a reduced affinity of the various mutant proteins for xylan, which is functioning as a surrogate substrate for the

PAGE 95

85 enzyme. There appears to be a dependence of velocity on substrate concentration for at least some of the mutants. The mutation evidence supports the hypothesis that one active center or domain is responsible for both enzymatic activities. It does not, however, totally rule out the possibility that the protein may contain separate catalytic sites or subsites which are spacially close together. Future kinetic experiments, including the investigation of substrate competition between arabinof uranosides and xylopyranosides, will allow further definition of the catalytic regions responsible for both activities.

PAGE 96

Chapter V Purification and characterization of the xylB -encoded protein Introduction The existance of polysaccharide-hydrolyzing enzymes having broad substrate specificities is well documented (Ward and Moo-Young, 1989) . One such example includes the exoglucanase, EXG, produced by Cellulomonas f imi (Beguin, 1991) . This enzyme also exhibits 6-D-xylanase activity. The xylanase from Clostridium thermocellum (XYNZ) also exhibits endo-glucanase activity towards carboxymethylcellulose (Grepinet et al . 1988) . In particular, several xylanase and xylosidase enzymes have been characterized which exhibit substrate ambiguity (Flint et al . 1989) . A recent example is a B-D-xylosidase that was cloned from Caldocellum saccharolyticum that also exhibits endoxylanase activity (Luthi et al . 1990) . Enzymes which exhibit both endoxylanase and 6-glucosidase activities have been shown to be fairly common among micoorganisms (Gilkes et al . 1991) . Substrate ambiguity among the glycohydrolases has been attributed to the similarities between the various substrates involved. Upon closer examination, this phenomenon is not totally unexpected. The 6(1, 4) -xylosidic bonds of xylan and the 6-(l,4)-glycosidic bonds of cellulose 86

PAGE 97

87 are structurally related and have quite similar molecular configurations about the 6(1,4) bonds with respect to the hydroxyl group on the a-carbon being in the axial or equatorial positions. Some true bifunctional cellulases have been demonstrated which contain separate active sites for each enzymatic activity. Saul et al . (1990) isolated a cellulase from Caldocellum saccharolyticum which exhibited both endoglucanase and exo-glucanase activities. These authors used DNA sequence homology comparisons and deletion analysis to demonstrate that the endoglucanase activity was located in the carboxy terminal domain and the exoglucanase activity was located at the amino terminal domain. In an earlier study purified a 6-xylosidase from Chaetomium trilaterale that also exhibited B-glucosidase activity (Uziie et al . 1985) . These investigators used kinetic analysis employing substrate competition and inhibitors to demonstrate that a single active site was responsible for both enzymatic activities. It was also suggested that two kinetically separate substrate binding sites may reside in the active center of this enzyme. In the previous chapter mutational analysis demonstrated that the two enzymatic activities encoded by xylB were not functionally separate but both appeared to be catalytically dependent upon the same region of the protein. The proposal that a single active center is responsible for

PAGE 98

88 both enzymatic activities in this enzyme was tested using analogous kinetic experiments including substrate inhibition and competition of the enzyme with respect to both the xylopyranosyl and arabinofuranosyl substrates. Materials and Methods Medium and growth conditions . Medium and growth conditions were described in chapter III. Preparation of cell extracts . Cell extracts were prepared as described in chapter III. Partial purification of fi-xvlosidase by preparative electrophoresis . Extracts containing the total cytoplasmic proteins from pLOIlOOS or L178F recombinant clones were used as a source of xylosidase. Proteins were fractionated in an 8% native polyacrylamide gel in a BioRad-Prep Cell preparative electrophoresis system. Gel and buffer formulations were: separating gel buffer; 240 mM Tris (pH 8.48), stacking gel buffer; 40 mM Tris (pH 6.9), lower tank buffer; 63 mM Tris/50 mM HCL (pH 7.5), upper tank buffer; 38 mM Tris/ 40 mM glycine (pH 8.9). A total of 50 mg/ml protein was loaded onto the gel. Electrophoresis was done at constant power of 31 W. Starting conditions were 250 V and 40 mA. Protein elution was monitored at 280 nm. Fractions were collected and assayed for xylosidase activity as described below. The most active fractions were pooled and concentrated using an Amicon Centriprep concentrator (Amicon Division, Danvers, MA) . Xylosidase in

PAGE 99

89 the concentrate was then precipitated by the addition of solid arnxnoniu. sulfate to 70% saturation. The enzyme was stored as an axtunonium sulfate pellet at S^C until needed. NO loss of enzyxnatic activity was detected after storage for one weeK under these conditions. Pellets were resuspended in 5 mM phosphate (pH 7.0) containing 10 mM Bmercaptoethanol immediately prior to use. uy^r-n phnbic -ir.^-or-;,nt-,ion rhroma toqr aphy . The xylosidase-containing pellet from preparative electrophoresis was resuspended inl ml of 1.7 M (NH,),SO, in 5 mM phosphate buffer (pH 6.8) and loaded onto a 2.0 x 3.0 cm Pharmacia XK 16/20 chromatography column (Pharmacia LKB, Uppsala, Sweden) packed with Toyopearl "TSK-Gel" hydrophobic gel (Supelco, Inc., Bellefonte, PA.). The column was equilibrated with 1.7 M (NH,),SO, in 5 mM phosphate buffer (pH 6.8) prior to the addition of sample. Xylosidase was eluted using a linear negative salt gradient starting with 1.7 M (NH,)2S0, down to zero in a total volume of 200 ml. Fractions were collected in 3 ml volumes and analyzed as described below. Pn.v^e assays , xylosidase activity in each fraction was assayed using p-nitrophenyl-B-D-xylopyranoside (p-NP-X) at a final concentration of 2.5 mM unless otherwise noted and in 50 mM phosphate buffer (pH 6.8) at 37^0. Assays were done in a total volume of 1 ml and allowed to continue until the yellow color indicating enzyme activity was detected.

PAGE 100

90 Assays were then terminated by the addition of 2 ml 0.5 M carbonate. The p-nitrophenol released by hydrolysis was measured spectrophotometrically at 405 nm. The liberation of 1 nmole of p-nitrophenol results in an increase in absorbance of 0.0184 at 405 nm. Samples were assayed for aL-arabinofuranosidase activity under the same conditions using p-nitrophenyl-a-L-arabinof uranoside (p-NP-A) as a substrate. Electrophoretic analysis of proteins . SDS-PAGE, native-PAGE, activity stains, and "Western" hybridizations were done as described in chapter III. Optimum activity pH . The optimum pH for both activities was determined in duplicate using citrate (40 mM) -sodium phosphate (80 mM) buffer in the pH range of 3.0 to 7.4. Tricine (50 mM) was used for the 7.5 to 8.5 pH range and Bicine (50 mM) for the 8.6 to 9.0 pH range. All activities were determined using a 6 mM final concentration of p-NP-X or p-NP-A at 37"'C. Thermal optimum and thermal inactivation determinations . Thermal stability was determined by incubating the enzyme for 30 min at 10, 25, 35, 45, 55, and 65°C. After 30 min the enzyme was placed in ice for 10 min prior to assaying both enzymatic activities. The optimum temperature for activity was determined by assaying the enzyme for both activities at 10, 25, 35, 45, 55, and 60°C for 10 min. Assays were done in duplicate at 37"'C using 6

PAGE 101

mM final concentration of p-NP-X or p-NP-A in citrate (40 mM) -sodium phosphate (80 mM) buffer at pH 6.0. Determination of Km and Vmax for both enzymatic activites . Both p-NP substrates were tested at various concentrations to determine the Km and Vmax values with respect to both substrates. All assays were done in duplicate at 37°C and at pH 6.0. Kinetic parameters were graphically determined using the Lineweaver-Burk and direct linear methods. Substrate competition experiments . The analogue 6methyl-D-xylopyranoside was used as an inhibitor for a-Larabinofuranosidase activity. The fluorogenic substrate 4methylumbellif eryl-a-L-arabinofuranoside was used as an inhibitor for 6-D-xylosidase activity. Results and discussion Purification of xylosidase . The xylosidase-containing fractions were separated as one large peak during preparative electrophoresis (Fig. 15) . This peak consisted of 15, 3 ml fractions which contained both xylosidase and arabinofuranosidase activities. The fraction which corresponded to the middle eight fractions exhibited the highest activities and were pooled, concentrated, and precipitated with ammonium sulfate. The xylosidase was further purified by hydrophobic interaction chromatography and was separated as a single peak consisting of six 3 ml fractions (Fig. 16) . The four most active fractions were

PAGE 102

Figure 15. Elution profile of the xylB gene product during preparative electophoresis on BioRad Prep Cell system.

PAGE 104

Figure 16. Elution profile of the xylB gene product during hydrophobic interaction chromatography.

PAGE 105

0 20 40 60 80 100 120 140 160 180 200 Elution Volume

PAGE 106

96 pooled and concentrated by ultrafiltration. A summary of this purification scheme is shown in Table 9. SDS-PAGE analysis of the pooled fractions from all the purification steps showed the xylosidase to be approximately 90% pure with low levels of contaminating proteins (Fig. 17) . Active enzyme conformation . Native-PAGE analysis of the enzyme using Coomassie blue, activity stains, and Western hybridization to visualize the protein indicated only one predominant form of the xylosidase was active (Fig. 18) . This active band had an apparent molecular weight of 120,000 which corresponds to the dimeric form of the enzyme. Enzyme optima . The temperature activity profiles for both activities were also essentially the same with 45°C being the optimum temperature for both the xylosidase and arabinofuranosidase activities (Fig. 19) . The thermal inactivation profiles for both activities in the wild type protein were essentially the same with activity diminishing rapidly after incubation for 30 min at 35°C (Fig. 20) . The pH profiles for both activities in the wild type protein showed marked differences (Fig. 21) . The xylosidase activity had a sharp peak at pH 6.0 which was followed by a rapid decline in activity with less than 50% relative activity remaining at pH 6.8. Arabinofuranosidase activity peaked at pH 6.0 but remained at 90% relative activity up to pH 9.0.

PAGE 107

97 TABLE 9. Purification of the B. f ibrisolvens xylB encoded xylosidase from E. coli DH5a (pLOI1005) . Fraction Vol (ml) Protein (mg) Activity (mU)* Sp Act (mU/mg) Yield (%) Crude extract 2.0 100 4600 46 100 Prep Cell 7.0 3.4 1068 314 23 Hydrophobic column 3.0 0.9 587 652 13 * Specific activity expressed as nmoles p-nitrophenol released per min per mg protein for B-D-xylosidase activity only.

PAGE 108

Figure 17. SDS-PAGE analysis of pooled xylosidase-containing fractions from preparative electrophoresis on 8% native PAGE: A and E; molecular weight markers, B; hydrophobic column purified preparation, C; Prep-Cell purified preparation, D; crude extract.

PAGE 109

r

PAGE 110

Figure 18. Native-PAGE analysis of the 6-D-xylosidase. A; Coomassie blue stained protein bands, B; Western hybridization of electroblotted native-PAGE gel, . C; activity stain (xylosidase activity/arabinofuranosidase activity) . Arrows indicate approximate region where the apparent dimeric and monomer ic forms of the enzyme would be located.

PAGE 111

dimer monomer

PAGE 112

Figure 19. Thermal inactivation profile of xylosidase (closed circle) and arabinofuranosidase (open circle) activities.

PAGE 113

103

PAGE 114

Figure 20. Temperature optimum profile for xylosidase (closed circle) and arabinofuranosidase (open circle) .

PAGE 115

105 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 Temperature ( C)

PAGE 116

Figure 21. pH profiles for xylosidase (closed circles) and arabinofuranosidase (open circles) activities.

PAGE 117

107

PAGE 118

108 Kinetic properties . A characteristic Michaelis-Menten relationship was observed for the native protein using p-NPX and p-NP-A as substrates. The Km and Vmax values for xylosidase and arabinofuranosidase activities were 3.8 itiM , 1.7 mM and 1,111 nmoles min"^ mg'^ protein, 833 nmoles min''' mg"'" protein, respectively (Fig. 22) . These values give a Kcat or turnover number per subunit of 1111 sec"^ for xylosidase activity and 833 sec"^ for arabinofuranosidase activity. Substrate competition experiments . The fluorogenic substrate 4-methylumbellif eryl-B-D-arabinofuranoside proved to be a potent competitive inhibitor of 6-D-xylosidase activity when p-NP-X was the substrate (Fig. 23). Alternatively, 6-methyl-D-xylopyranoside, a xylobioside analogue, was a competitive inhibitor of a-Larabinofuranosidase activity when p-NP-A was the substrate (Fig. 24). In separate experiments, the enzymatic activities were measured with the two substrates, p-NP-X and p-NP-A individually and together at 10 mM (saturation) concentration at 37°C for 10 min. Xylosidase activity alone was 1160 mU, and arabinofuranosidase activity alone was 809 mU. When both substrates were combined in the same assay the resultant activity was 948 mU which is essentially the average of the two activities separately. These individual activities are not additive when both substrates are present

PAGE 119

Figure 22. Double recipricol plots of xylosidase and arabinofuranosidase activities. Xylosidase; closed circles, arabinofuranosidase; open circles.

PAGE 120

2.0 / / 1.8 ; 1.6 ; 1.4 1.2 / / 5/4 /0.2 1 — I — 1 — ' — I'M ' — ^— 1 1 ' 1 ' 1 ' — h-«— 1 1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1/[S] (mM)

PAGE 121

Figure 23. Competitive inhibition of xylosidase activity by 4-inethylunibelliferyl-a-L-arabinofuranoside: closed circles; no inhibitor, open circles; 100 /xM inhibitor, open squares; 250 juM inhibitor.

PAGE 122

112 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1/[S1 (mM)

PAGE 123

Figure 24. Competitive inhibition of arabinofuranosidase activity by B-D-methylxylopyranoside: closed circles; no inhibitor, open circles; 1 mM inhibitor, closed squares; 2 mM inhibitor.

PAGE 125

115 together in the assay system. An additive result would have indicated that separate active sites might be present for the two enzymatic activities. This data is consistent with the results of the competitive inhibition experiments and demonstrate that both substrates are competing with each other for the same catalytic center on the enzyme. Conclusions The xylB gene encodes an enzyme that exhibits substrate ambiguity with respect to the two p-nitrophenol-derived xylopyranosides and arabonofuranosides. The activities against these two substrates co-purified during preparative electrophoresis and were both stable as ammonium sulfate pellets for up to one week at 5°C. Only a single active form of this enzyme is present on native-PAGE gels and appears to have a molecular weight of 120,000. The thermal stability and the optimal temperature were essentially the same for both activities with activities being stable up to 42°C after 30 min and the optimal temperature was 45°C. The thermal stability of this enzyme is relatively low when compared to xylosidases purified from other microorganisms. The temperature optimum for this enzyme is similar to xylosidases produced by Clostridium acetobu tvlicum (Lee and Forsberg, 1987) and Bacillus pumilus (Panbangred et al . 1984) . Investigations of optimal pH revealed a major difference with respect to both activities. Both activities

PAGE 126

116 exhibited an optimum pH of 6.0 which compares with that of the £. acetobutvlicum xylosidase. There are four histidine residues in the proposed active center for this enzyme. Since the pKa of histidine is 6.5, it is likely that one or more of these residues are involved in both catalytic activities. The proposed mechanism for glycohydrolase, and the related glycosyltransf erase catalysis, involves contributions from a hydrogen donor and a hydrogen acceptor moiety (Mooser et_al. 1991) . In the usual scheme, an aspartic or glutamic acid in deprotonated form serves as a hydrogen acceptor and stabilizes the hydrolytic intermediate. A general acid donates a proton to the glycosidic oxygen to facilitate the cleavage of the monomer ic glycoside. The pH dependence of the xy IB -encoded protein with respect to the B-D-xylosidase activity is probably due to one or more histidine residues serving as the general acid. The broad pH activity profile for the arabinofuranosidase activity is uncommon. At the higher pH ranges, histidine is completely deprotonated and could no longer serve as a general acid and donate a proton. It is possible that the arabinofuranoside substrate, once bound in the active pocket, protects the histidine from further deprotonation at high pH. This charge transfer phenomenon has been demonstrated to occur in other enzymatic systems

PAGE 127

117 (Robson and Garnier 1988) . The binding of the arabinofuranoside may also induce a conformational change in the enzyme which somehow leads to the protection of the histidyl moiety from deprotonation at high pH ranges. It should be noted that the effective pKa of these amino acids can vary markedly from the intrinsic pKa due to the electonic environment in the active center of the protein. The possibility that a different amino acid is contributing a proton at the higher pH ranges cannot be ruled out. Tyrosyl residues (pKa 10) in this region of the protein could possibly substitute for histidine as a general acid at the higher pH ranges. This phenomenon might be explained if the two activities were catalyzed by different mechanisms that in turn involve different amino acids present in the same active center. This was shown to be the case with the neopullulanase from Bacillus stearothermophilus in which two separate enzymatic activities were found to reside in the same active center but were catalyzed by different amino acids and thus different mechanisms (Kuriki et al . 1991) . In an earlier work, The 6-D-xylosidase from Bacillus pumilus was found to hydrolyze a-D-xylosylf luoride by a mechanism that was separate from that for the hydrolysis of 6-D-xylosylf luoride (Kasumi et al . 1987) . The number and nature of the active centers is not known for this protein.

PAGE 128

118 The kinetic properties of the xylB -encoded enzyme with respect to the xylosidase activity is similar to the xylosidase enzymes from C. acetobutylicum and B. pumilus which have Km's for p-NP-X of 3.8 mM, 3.7 mM, and 2.4 mM, respectively. The Km for the arabinofuranosidase activity against p-NP-A was 1.7 mM which compares favorably with 1.3 mM for the a-L-arabinof uranosidase isolated from another rumen bacteria, Ruminococcus albus (Greve et al . 1984) . The maximum velocity values for the xylosidase activity of the xy IB -encoded protein is approximately tenfold lower than that for the C. acetobuty 1 icum enzyme and approximately onehalf that for the B. pumilus enzyme. Lee and Forsberg (1987) purified and characterized an a-Larabinof uranosidase from C. acetobutylicum ATCC 824. This enzyme exhibited Km and Vm values of 4 mM and 36.4 /nmole min"^ mg'^ protein respectively against p-nitrophenyl-a-Larabinofuranoside. It had a pH optimum of 5.0 5.5 and exhibited no activity against other p-nitrophenylglycosides. High catalytic efficiency is usually associated with highly evolved enzymes having a narrow substrate specificity (Robson and Garnier, 1988) . It is possible that the xvlB encoded enzyme has sacrificed catalytic efficiency for a broader substrate range since neither the C. acetobutylicum or the B. pumilus xylosidase enzymes have additional activities.

PAGE 129

119 The substrate competition experiments, using model substrates, unequivicaly demonstrated that xylopyranoside and arabinof uranoside are competitive inhibitors of each other .

PAGE 130

CHAPTER VI PARTIAL PURIFICATION AND CHARACTERIZATION OF THE L178F MUTANT PROTEIN Introduction As outlined in chapter IV, point mutations were introduced into the xylB gene using in vitro mutagenesis. All the mutations resulted in proteins having a reduction or loss of both enzymatic activities concurrently. Additionally, 10 of these point mutations were clustered in a 60 amino acid region of the protein. Native agarose-xylan gel electrophoresis of the expressed wild type and mutant proteins revealed that the mutant proteins bind to the surrogate substrate xylan with less affinity relative to the native protein indicating an apparent increase in the Km values for these mutant enzymes. The L178F mutant appeared to have the least affinity (highest apparent Km) for the surrogate substrate. To further investigate the possibility of altered Km due to the L178F mutantion, the mutant protein was partially purified and characterized. 120

PAGE 131

121 Materials and methods All methods used to purify and characterize the L178F mutant enzyme were essentially the same as those described for the native enzyme in Chapter V. Results and Discussion Partial purification of L178F protein . The elution profile for the mutant protein was essentially the same as that for the wild type protein, have a single broad peak consisting of 15, 3 ml fractions (Fig. 25) . The fractions which corresponded to the middle six fractions exhibited the highest relative activity and were pooled, concentrated, and precipitated at 70 % saturation with ammonium sulfate. SDSPAGE analysis of the partially purified protein indicated it was approximately 80 % to 90 % pure with low levels of contaminating proteins (Fig. 26) . Enzyme optima . The mutant protein exhibited a higher thermal stability relative to the wild type (Fig. 27) . This protein was stable up to 55°C for 3 0 min and still retained 100 % relative activity with respect to both substrates. Since the replacement of a phenylalanine for leucine introduces a more hydrophobic residue at this position, it is possible that this strengthened the hydrophobic interactions at the core of the protein and thereby increased the thermal stability. The optimal temperature was not investigated.

PAGE 132

Figure 25. Elution profile of the xylB -encoded protein harboring the L178F mutation during preparative electrophoresis :

PAGE 133

123 250 50 100 150 200 Elulion Volume (ml) 250

PAGE 134

Figure 26. SDS-PAGE analysis of partially purified L178F mutant protein: A; Prep-Cell purified preparation, B; crude extract, C; molecular weight markers.

PAGE 135

125

PAGE 136

Figure 27. Thermal inactivation profiles of xylosidase and arabinofuranosidase activities of the L178 mutant protein: open squares; xylosidase activity, closed squares; arabinofurnanosidase activity.

PAGE 137

127

PAGE 138

128 The pH profile for both enzymatic activities for the mutant protein was essentially the same as that for the wild type protein (Fig. 28) . Xylosidase activity had a sharp activity peak at pH 6.0 which was followed by a rapid decline. Arabinof uranosidase activity also peaked at pH 6.0 but remained at 90 % relative activity up to pH 9.0. Kinetic properties . A characteristic Michaelis-Menten linear relationship was observed for the mutant protein with respect to both substrates (Fig. 29) . The Km values for the xylosidase and arabinof uranosidase activities were 16 mM and 3 3 mM respectively. This represents a decrease in the affinity of the enzyme for both substrates. The increase in Km agrees with the data in chapter IV using the agarosexylan gel binding assays. Conclusions It is apparent that the replacement of leucine with phenylalanine at position 178 decreases both the affinity of the enzyme for both substrates and the catalytic efficiency of the enzyme. This amino acid substitution also increases the thermal stability of the xylB protein. It is possible that these three properties are related. The increase in thermal stability could represent a general increase in the stability of the enzyme by strengthening hydrophobic interactions at the core of the protein. An increase in the overall stability of the protein could influence the plasticity of the active center. The

PAGE 139

Figure 28. pH activity profiles of xylosidase and arabinofuranosidase activies of the L178F mutant protein: open squares; arabinofuranosidase activity, closed squares; xylosidase activity.

PAGE 140

4> 100 75 50 25 0

PAGE 141

Figure 29. Double recipricol plots of xylosidase and arabinofuranosidase activities for the L178F mutant protein: open squares; arabinofuranosidase activity, closed squares xylosidase activity.

PAGE 142

• <* • ' 132 150125100: 75: 50: 25: f' ' 1 ' 1 1 1 ' ' 1 1 1 1 1 1 1 i/i (/ ' 1 ' 1 1 1 1 1 1 1 1 1 r 1 1 1 1 1 1 1 ' ' I ' " ' M ' ' ' I ' ' ' ' I ' ' ' ' I I I I I I -1.00 -0.75 -0.50 -0.25 0.00 0.25 0.50 0.75 1.00 1/[S] (mM)

PAGE 143

133 "induced-fit" model of enzyme catalysis proposes that the three-dimensional conformation of an enzyme will change as it binds to the substrate to allow closer contacts to be made between the substrate and the catalytically important functional groups (Koshland, 1966) . Increasing the hydrophobic interactions in the core or active center of the protein might also decrease the ability of the protein to change conformation upon binding substrate. This would lead to a loss of catalytic efficiency and possibly affect substrate binding. A more general explanation of the effects of the L178F mutation would be that the protein is simply folded incorrectly and this results in a less accessible active center which is reflected in the increase in Km for p-NPX and p-NPA in this mutant protein. The addition of a large aromatic R-group by the substitution of phenylalanine for leucine could also introduce a steric hindrance factor which prevents optimal enzyme-substrate interactions and therefore decreases the affinity and/or catalytic efficiency of the enzyme. This assumes, however, that the amino acid substitution has occurred in an area within or near the active site.

PAGE 144

Based upon the data in this chapter and in chapter IV it is reasonable to conclude that the L178F mutation is in an area of the xylB protein that is important in substrate binding and possibly catalysis with respect to both enzymatic activities.

PAGE 145

CHAPTER VII SUMMARY AND GENERAL CONCLUSIONS The studies presented here have characterized the 6-Dxylosidase from the rumen bacteria Butyrivibrio f ibrisolvens using genetic and biochemical techniques. This enzyme is important in the final steps of depolymerization of hemicellulose in that the products of hydrolysis, usually monomers, are used directly in the metabolism of the bacterium. This enzyme is particularly interesting in that it exhibits substrate ambiguity. Both 6-D-xylopyranosides and a-L-arabinofuranosides are hydrolyzed by this enzyme. Previous studies involving 6-D-xylosidase enzymes from other organisms have shown that substrate ambiguity among the xylosidase and related enzymes is not universal. The xylosidase from Bacillus pumilus only exhibits activity against aryl-6-D-xylopyranosides (Panbangred et al . 1983) . The same is true for the 6-D-xylosidase from B. subtilis (Paice et al., 1986). Notably a 6-D-xylosidase from Caldocellum saccharolyticum was cloned and sequenced that also exhibited endoxylanase activity (Luthi et al . 1990) . In this case the two substrates are structurally identical but vary in size with the endoxylanase acting on the longer chain length polymers while the xylosidase is specific for 135

PAGE 146

the shorter hydrolytic products of the endoxylanase activity. No data is yet available concerning active site studies of this enzyme as to whether or not one or two catalytic centers are responsible for this phenomenon. The present study is unique in that it examines the substrate ambiguity of the B. f ibrisolvens enzyme using both genetic and biochemical techniques. This work represents a complete study of the enzyme and has elucidated a number of important aspects of the 6-D-xylosidase from B. f ibrisolvens . (A) The gene that encodes this enzyme in B. f ibrisolvens . xylB is present as a single copy in the chromosome (Fig. 3) . Any additional xylosidase genes present in this organism must share limited homology with xylB . (B) The DNA sequence is 1,551 bp in length, encodes 517 amino acids and is located between two additional large open reading frames (ORF's) each in excess of 1,000 bp (Fig. 5) . The size of this enzyme is essentially the same as the B. pumilus xylosidase. (C) No stem loop or rho-independant teminators were identified between any of the three ORF's and the expression data indicates that this region does not function as a terminator in E. coli (Table 2) . This result suggests that these three ORF's represent part of a xylandegrading operon in B. f ibrisolvens . The subcloning and expression data also demonstrate that the single xylB gene is responsible for both the xylosidase and arabinofuranosidase activities (Fig. 4b) . (D) The codon

PAGE 147

137 bias of these three ORF's is consistent with the low guanine plus cytosine ratio of the DNA from this organism (Table l) . (E) The xylB -encoded protein shares limited amino acid identity with other published amino acid sequences of 6-Dxylosidase enzymes and related proteins (Table 4) . This is an indication that the xylB gene from B. f ibrisolvens is evolutionarily divergent from these other genes. (F) A consensus sequence has been identified in xylB that has significant identity to an amino acid sequence that has been previously implicated in catalytic function of other glycohydrolases (Table 5) . Mutational analysis of the xylB -encoded protein has revealed several important structural and functional relationships. (A) All point mutations within the xylB gene resulted in a protein in which both enzymatic activities were reduced or abolished (Table 6) . (B) All point mutations that were isolated are localized or clustered in single region of the protein that is near the proposed catalytic center (Fig. 10) . (C) All but two of the point mutations result in a stably expressed protein (Fig. 11) . (D) All of the point mutations that result in expressed protein exhibit a decrease of affinity in native gel assays (Fig. 14) and an increase in apparent Km. These results demonstrate that the catalytic center for these two enzymatic activities is not functionally independent but is localized on the same region of the xylB -encoded protein.

PAGE 148

138 Finally, a biochemical characterization of the purified enzyme was initiated to examine whether or not a single active center was responsible for both enzymatic activities. The partially purified enzyme exhibited thermal inactivation and temperature optimum profiles that were essentially the same for both activities (Figs. 19 and 20). The pH optima for both activities were essentially the same however the activity range with respect to pH differed for the two activities (Fig. 21) . Several hypotheses may be invoked to explain this difference in pH. Hydrolytic mechanisms usually involve the contributions of a proton donor and a proton acceptor to the catalysis (Knowles et_al. 1988) . The proton donor acts as a general acid in donating a proton to the glycosidic bond oxygen generating an oxycarbonium ion intermediate. The proton acceptor serves to stabilize the hydrolytic intermediate and acts as a general base. At pH 6.0, the pH optimum for this enzyme, carboxyl groups on aspartic and glutamic acid residues are fully charged. At this pH a histidinyl residue (pKa 6.5) may serve as a proton donor in that it is still protonated. At pH values above 6.0 the xylosidase activity diminishes reflecting the deprotonation of histidinyl residues in the active center demonstrating the dependence of xylosidase activity on this amino acid. The relatively high arabinof uranosidase activity at elevated pH may indicate that this catalytic activity

PAGE 149

139 proceeds by a mechanism different from that of xylosidase activity. This phenomenon, involving separate hydrolytic mechanisms, was shown to exist with the Bacillus pumilus xylosidase (Kasumi et al . 1987) . These investigators demonstrated kinetically that the enzyme was able to catalyze the hydrolysis of aand 6-D-xylosylf luoride by a mechanism entirely different from that for the hydrolysis of p-nitrophenyl-6-D-xylopyranoside. It was postulated that separate catalytic groups were responsible for the two mechanisms. In a more recent study Kuriki et al. (1991) demonstrated that a neopullulanase from Bacillus stearothermophilus . exhibited separate a-1,4 and a-1,6 bond cleavage activities. These investigators used site-directed mutagenesis to demonstrate that the two activities were catalyzed by separate amino acids in the same active center. It is possible that an analogous process occurs in the xy IB -encoded protein. A tyrosinyl residue (pKa 10.0) could serve as a general acid and donate a proton at the pH range where the arabinofuranosidase activity is present. There are five tyrosine residues in the proposed active center which may be involved. Another explanation involves the relative stabilities of the two substrates at elevated pH. If the arabinofuranoside is less stable at higher pH values relative to the xylopyranoside, it follows hydrolysis will proceed more easily for the former and therefore requires

PAGE 150

140 less of proton donating potential from involved hisidinyl residues. This explaination may also be invoked to support the possbility that separate amino acids and/or mechanisms are involved in the two separate activities. Also relating to substrate effects is the concept of substrate-assisted catalysis (Carter and Wells, 1990) . Substrate-assisted catalysis was first demonstrated with proteolytic enzymes. In the enzyme subtilisin, a serine endopeptidase, it was demonstrated that a histidine in the protein substrate could replace a catalytic histidine in the active site of subtilisin that had been mutated to an alanine by site-directed mutagenesis. It is possible that a glycosyl carboxylate could serve as an alternate proton acceptor under certain conditions. The kinetic constants for the two activities were comparable to other xylosidase and arabinofuranosidase enzmes that have been reported. The catalytic efficiency or Kcat for this enzyme is not as high as that reported for other xylosidase or arabinofuranosidase enzymes. Since both arabinofuranosidase and xylopyranosidase are substrates that are indigenous to hemicellulose, this protein appears to have saccrificed catalytic efficiency for a broader substrate specificity. This may represent an evolutionary adaptation to the highly competitive rumen environment.

PAGE 151

141 The substrate competition experiments corroborate the genetic evidence and demonstrated that the two activities reside in the same active center of the xylB -encoded enzyme. This study demonstrates that the xylB gene from Butyrivibrio f ibrisolvens encodes a single protein having 6D-xylosidase and a-L-arabinofuranosidase activities. These activities are localized in the same active site of the protein.

PAGE 152

LITERATURE CITED Allf-Steinberger , C. 1984. Evidence for coding pattern on the non-coding strand of the Escherichia coli genome. Nucleic Acids Res. 12:2235-2241. Armstrong, D.G., and H.J. Gilbert. 1985. Biotechnology and the rumen: A mini-review. J. Sci. Food Agric. 36:1039-1046. Barnett, C. C, R. M. Berka, and T. Fowler. 1991. Cloning and amplification of the gene encoding an extracellular 6glucosidase from Trichoderma reesei : evidence for improved rates of saccharif ication of cellulosic substrates. Biotechnol. 9:562-567. Bastawde, K. B. , L. B. Tabatabai, M. M. Meagher, M. C. Srinivasan, H. G. Vartak, M. V. Rele, and P. J. Reilly. 1991. Catalytic properties and partial amino acid sequence of an actinomycete endo(1-4) -6-D-xylanase from Chainia species. p417-425. In G. F. Leatham and M. E. Himmel (ed) , Enzymes in biomass conversion. Amer. Chem. Soc. Washington, D. C. . , , Biely, P. 1985. Microbial xylanolytic systems. Trends in Biotechnol. 3:286-290. Biely, P., C. R. MacKenzie, J. Puis, And H. Schneuder. 1986. Cooperativity of esterases and xylanases in the enzymatic degradation of acetyl xylan. Biotechnol. 4:731-733. Biely, P., and E. Petrakova. 1984. Novel inducers of the xylan degrading system of Crvptococcus albidus . J. Bacterid. 160:408-412. Biely, P., J. Puis, and Schneider. 1985. Acetyl xylan esterases in fungalcellulolytic systems. FEBS Lett. 186:8084. Biely, P., Z. Kratsky, M. Vranska, and D. Urmanicova. 1980. Induction and inducers of endo-1, 4-6-xylanase in the yeast Crvptococcus albidus . Eur. J. Biochem. 108:323-329. Benguin, P., P. Cornet, and J. P. Aubert. 1985. Sequence of a cellulase gene of the thermophilic bacterium Clostridium thermocellum . J. Bacterid. 162:102-105. 142

PAGE 153

143 Berger, E. , W. A. Jones, D. T. Jones, and D. R. Woods. 1989. Cloning and sequencing of an endoglucanase (endl) gene from Butyrivibrio fibrisolvens H17c. Mol. Gen. Genet. 219:193198. Boel, E., M.T. Hansen, I. Hjort, and N.P. Fiil. 1984. Two different types of intervening sequences in the glucoamylase gene from Aspergillus niaer . EMBO J. 3:1581-1585. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of proteins utilizing the principle of protein-dye binding. Anal. Biochem. 72:248254. Chesson, A., A. H. Gordon, and J. A. Lomax. 1983. Substituent groups linked by alkali labile bondsto arabinose and xylose residues of legume, grass, and cereal straw walls and their fate during digestion by rumen microorganisms. J. Sci. Food. Agri. 34:1330-1340. Cotta, M. A., and R. B. Hespell. 1986. Proteolytic activity of the ruminal bacteria Butyrivibrio fibrisolvens . Appl. Environ. Microbiol. 52:51-58. Coughlan, M. P. 1985. Properties of fungal and bacteria cellulases with comment on thier production and application. Biotechnol. Genet. Eng. Rev. 3:39-109. Dehority, B. A. 1968. Mechanism of isolated hemicellulose and xylan degradation by cellulolytic rumen bacteria. Appl. Microbiol. 16:781-786. Dehority, B. A. 1966. Characterization of several bovine rumen bacteria isolated with a xylan medium. J. Bacterid. 91: 1724-1729. Dekker, R. F. H. 1985. Hemicellulose degradation, p 505. In Higuchi, T. (ed) , Biosynthesis and biodegradation of wood components. Academic Press, Orlando. Dekker, R. F. H. , and G. N. Richards. 1976. Hemicellulases: Their occurrence, purification, properties, and mode of action. Adv. Carbohydr. Chem. Biochem. 32:277-352. Deshpande, V., A. Lachke, C. Mishra, S. Keskar, and M. Rao. 1986. Mode of action and properties of xylanase and 6xylosidase from Neurospora crassa. Biotechnol. Bioengin. 28:1832-1837. Devereux, J., P. Haeberli, and 0. Smithies. 1984. A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12:387-395.

PAGE 154

144 Eriksson, K. E. L. , R. A. Blanchette, and P. Ander. 1990. In Microbial and enzymatic degradation of wood and wood components. Springer series in wood science. Springer-Verlag Berlin, Heideilberg. pp 407. Esteban, R. , J. R. Villanueva, and T. G. Villa. 1982. 6-Dxylanases of Bacillus circulans WL-12. Can J. Microbiol. 28:733-739. Frederick, M. M. , C. Kiang, J. R. Frederick, and P. J. Reilly. 1985. Purification and characterization of endoxylanases from Aspergillus niger . I. Two isozymes active on xylan backbones near branch points. Biotechnol. Bioengin. 27:525-532. Flint, H. J., C. A. McPherson, and J. Bisset. 1989. Molecular cloning of genes from Ruminococcus flavefaciens encoding xylanase and 6(1-3, 1-4) glucanase activities. Appl. Environ. Microbiol. 55:1230-1233. Forsberg, C. W. , B. Crosby, and D. Y. Thomas. 1986. Potential for manipulation of the rumen fermentation through the use of recombinant DNA techniques. J. Anim. Sci. 63:310325. Forsberg, C. W. , T. J. Beveridge, and A. Hellstrom. 1981. Cellulase and xylanase release from Bacteroides succinogenes and its importance in the rumen environment. Appl. Environ. Microbiol. 42:886-896. Fukumoto, J., Y. Tsujisaka, and S. Takenishi. 1970. Studies on hemicellulases . I. Purification and properties of hemicellulases from Aspergillus niger var. Tieghem sp. Nippon Nogei Kogaku Kaishi. 44:447-475. Fukusaki, E. , W. Panbangred, A. Shinmyo, and H. Okada. 1984. The complete nucleotide sequence of the xylanase gene (xylA) of Bacillus pumilus . FEBS Lett. 171:197-201. Garcia-Martinez, D. V., A. Shinmyo, A Madia, and A. L. Demain. 1980. Studies on cellulase production by Clostridium thermocellum . Appl. Microbiol. Biotechnol. 9:189-197. Gascoigne, J. A., and M. M. Gascoigne. 1980. The xylanases of Fusarium roseum . J. Gen. Microbiol. 22:242-248. Grepinet, O. , Marie-Christine Chebrou, and P. Beguin. 1988. Nucleotide sequence and deletion analysis of the xylanase gene ( xylZ ) of Clostridium thermocellum . J. Bacterid. 170:4582-4588.

PAGE 155

145 Grepinet, O., and P. Benguin. 1986. Sequence of the cellulase gene of Clostridium thermocellum coding for endoglucanase B. Nucleic Acids. Res. 14:1791-1799. Greve, C. L. , J. M. Labavitch, and R. E. Hungate. 1984. ot-Larabinofuranosidase from Ruminococcus albus 8: Purification and possible role in hydrolysis of alfalfa cell wall. Appl. Environ. Microbiol. 47:1135-1140. Hespell, R. B., and P. J. 0 ' Bryan-Shah. 1988. Esterase activities in Butyrivibrio fibrisolvens strains. Appl. Environ. Microbiol. 54:1917-1922. Hespell, R. B., R. Wolf, and R. J. Bothast. 1987. Fermentation of xylans by Butyrivibrio fibrisolvens and other ruminal bacterial species. Appl. Environ. Microbiol. 53:000-000. Hespell, R. B., and M. P. Bryant. 1981. The genera Butyrivibrio . Succinivibrio . Succinomomas . Lachnospira . and Selenomonas . p. 1479-1494. In M. P. Starr, H. Stolp, H. G. Truper, A. Balows, and H. G. Schlegel (ed) , The prokaryotes, a handbook on habitates, isolation, and identification of bacteria. Springer-Verlag, New York. Hobson, P. N., and R. J. Wallace. 1982. Microbial ecology and activities in the rumen. II. Crit. Rev. Microbiol. 9:253-320. Hobson, P. N., and M. R. Purdom. 1961. Two types of xylan fermenting bacteria from the sheep rumen. J. Appl. Bacteriol. 24: 188193. Joliff, G. , P. Benguin, and J. P. Aubert. 1986. Nucleotide sequence of the cellulase gene celD encoding endoglucanase D of Clostridium thermocellum . Nucleic Acids Res. 14:86058613. Kasumi, T. , Y. Tsumuraya, C. F. Brewer, H. KerstersHilderson, M. Claeyssens, and E. J. Hehre. 1987. Catalytic versatility of Bacillus pumilus 6-xylosidase: Glucosyl transfer and hydrolysis promoted with aand 6-D-xylosyl fluoride. Biochem. 26:3010-3016. Kelly, M. A., M. L. Sinnott, and M. Herrchen. 1987. Purification and mechanistic properties of an extracellular a-L-arabinofuranosidase from Monilinia fructiaena . Biochem J. 245:843-849.

PAGE 156

146 Kersters-Hilderson, H., F. G. Loontiens, M. Claeyssens, and C. K. De Bruyne. 1969. Partial purification and properties of an induced 6-D-xylosidase of Bacillus pumilus 12. Eur. J. Biochem. 7:434-441. Knowles, J. K. C. , P. Lehtovaara, M. Murray, and M. L. Sinnot. 1988. Stereochemical course of action of the cellobioside hydrolases I and II of Trichoderma reesei . J. Chem. Soc. Chem. Conunun. 1988:1401-1402. Knowles, J., P. Lehtovaara, and T. Teeri. 1987. Cellulase families and their genes. Trends Biotechnol. 5:255-261. Koshland, D. E., Jr., G. Nemethy, and D. Filmer. 1966. Comparison of experimental binding data and theoretical models in proteins containing subunits. Biochem. 5:365-376. Kuriki, T. , H. Takata, S. Okada, and T. Imanaka. 1991. Analysis of the active center of Bacillus staerothermophilus neopullulanase. J. Bacterid. 173:6147-6152. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4 . Nature 227:680-685. Lee, S. F., and C. W. Forsberg. 1987. Purification and characterization of an o-L-arabinof uranosidase from Clostridium acetobutylicum ATCC 824. Can. J. Microbiol. 33: 1011-1016. Lee, S. F. , C. W. Forsberg, and J. B. Rattray. 1987. Purification and characterization of two endoxylanases from Clostridium acetobutylicum ATCC 824. Appl. Environ Micriobiol. 53:644-650. Luthi, E. , D. R. Love, J. McAnulty, C. Wallace, P. A. Caughey. , D. Saul, and P. Bergquist. 1990. Cloning, sequence analysis, and expression of genes encoding xylan-degrading enzymes from the thermophile Caldocellum saccharolyticum . Appl. Environ. Microbiol. 56:1017-1024. Lynch, J. M. 1987. Utilization of lignocellulosic wastes. In Journal of Applied Bacteriology Symposium Supplement, 71S83S. MacKenzie, C. R. , D. Bilous, H. Schneider, and K. G. Johnson. 1987. Induction of cellulolytic and xylanlolytic enzyme systems in Streptomyces spp. Appl. Environ. Microbiol. 53:2835-2839.

PAGE 157

147 MacKenzie, C. R. , R. C. A. Yang, G. B. Fatal, D. Bilous, and S. A. Nurang. 1989. Identification of three distinct C. thermocellum xylanase genes by molecular cloning. Arch. Microbiol. 152:377-381. MacNeil, N. I. 1984. The contribution of the large intestine to energy supplies in man. Amer. J. Clin. Nutr. 39:338-346. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y. Mannarelli, B. M. , R. J. Stack, D. Lee, and L. Ericsson. 1990. Taxonomic relatedness of Butvrivibrio . Lachnospira ^ Roseburia and Eubacterium species as determined by DNA hybridization and extracellular-polysaccharide analysis. Int. J. Syst. Bacterid. 40:535-544. Mannarelli, B. M. , S. Evans, and D. Lee. 1990. Cloning, sequencing and expression of a xylanase gene from the anaerobic rumen bacterium Butyrivibrio f ibrisolvens . J. Bacterid. 172:4247-4254. Mannerelli, B. M. 1988. Deoxyribonucleic acid relatedness among strains of the species Butyrivibrio f ibrisolvens. Int. J. Syst. Bacterid. 23:308-315. McGavin, M. , and C. W. Forsberg. 1989. Catalytic and substrate-binding domains of endoglucanase 2 from Bacteroides succinoaenes . J. Bacterid. 171:3310-3315. Morag. E. , E. A. Bayer, and R. Lamed. 1990. Relationship of cellulosomal and noncellulosomal xylanases of Clostridium thermocellum to cellulose-degrading enzymes. J. Bacterid. 172: 6098-6105. Moriyama, H. , E. Fufusaki, J. Cabrera Crespo, A. Shinmyo. , and H. Okada. 1987. Structure and expression of genes coding for xylan-degrading enzymes of Bacillus pumilus . Eur. J. Biochem. 166:539-545. Morosoli, R., c. Roy, and M. Yaguchi. 1986. Isolation and partial primary sequence of a xylanase from the yeast Crvptoc occus albidus . Biochim. Biophys. Acta 870:473-478. Nakanishi, K. , T. Yasui, and T. Kobayashi. 1971. Inducers for the xylanase production by Streptomyces sp. J. Ferment. Technol. 54:801-807. Nishitani, K. , and D. J. Nevins. 1991. Glucuronoxylan xylanohydrolase: A unique xylanase with the requirement for appendant glucuronosyl units. J. Biol. Chem. 266:6539-6543.

PAGE 158

148 Nishitani, k. , and D. J. Nevins. 1988. Enzymatic analysis of feruloylated arabinoxylans (Feraxan) derived from Zea mays cell walls I. Plant Physiol. 87:883-890. Ornstein, L. , and B. J. Davis. 1964. Disc electrophoresis: background and theory. Ann. N. Y. Acad. Sci. 121:321-349. Paice, M. G. , R. Bourbonnais, M. Desrochers, L. Jurasek, M. Yaguchi. 1986. A xylanase gene from Bacillus subtilis ; nucleotide sequence and comparison with B. pumilus gene. Arch. Microbiol. 144:201-206. Panbangred, W. , E. Fukusaki, E. C. Epifanio, A. Shinmyo, and H. Okada. 1985. Expression of a xylanase gene of Bacillus pumilus in Escherichia coli and Bacillus subtilis . Appl. Microbiol. Biotechnol. 22:259-264. Panbangred, W. , A. Shinmyo, S. Kinoshita, and H. Okada. 1983a. Purification and properties of endoxylanase produced by Bacillus pumilus . Agric. Biol. Chem. 47:957-963. Panbangred, W. , T. Kondo, S. Negoro, A. Shinmyo, and H. Okada. 1983b. Molecular cloning of the genes for xylan degradation of Bacillus pumilus and their expression in Escherichia coli. Mol. Gen. Genet. 192:335-341. Patterson, J. A. 1989. Prospects for establishment of genetically engineered microorganisms in the rumen. Enzyme Microb. Technol. 11:187-189. Puo-Llinas, J., and H. Driguez. 1987. D-Xylose as inducer of the xylan-degrading system in the yeast Pullularia pullulans . Appl. Microbiol. Biotechnol. 27:134-138. Poutanen, K. , M. Tenkanen, H. Korte, and J. Puis. 1991. Accessory enzymes involved in the hydrolysis of xylans. In Enzymes in biomass conversion, G. F. Leatham and M. E. Himmel, ed. American Chemical Society Symposium series 460. American Chemical Society, Washington, D. C. Poutanen, K., and J. Puis. 1988. Characteristics of Trichoderma reesei 6-xylosidase and its use in the hydrolysis of solubilized xylans. Appl. Microbiol. Biotechnol. 28:425-432. Poutanen, K. , and M. Sundberg. 1988. An acetyl esterase of Trichoderma reesei and it'srole in the hydrolysis of acetyl xylans. Appl Microbiol. Biotechnol. 28:419-424. Quiocho, F. A. 1986. Carbohydrate-binding proteins: tertiary structures and protein-sugar interactions. Ann. Rev. Biochem. 55:287-315.

PAGE 159

149 Raynal,A., C. Gerbaud, M. C. Francingues, and M. Guerineau. 1987. Sequence and transcription of the 6-glucosidase gene of Kluyveromyces fraqilis cloned in Saccharomvces cerevisiae . Curr. Genet. 12:175-184. Rerat, A., M. Fiszlewicz, A.Giusi, and P. Vaugelade. 1987. Influence of meal frequency on postprandial variations in the production of volatile fatty acids in the digestive tract of conscious pigs. J. Anim. Sci. 64:448-453. Robson and Gardier. 1988. Proteins and protein engineering, p 699. Elsevier Science Publishers B. V. Rouvinen, J., T. Bergfors, T. Teeri, J. K. C. Knowles, and T. A. Jones. 1990. Three-dimensional structure of cellobiohydrolase II from Trichoderma reesei . Science 249:380-386. Salyers, A. A., J. R. Balascio, and J. K. Palmer. 1981. Breakdown of xylan by enzymes from Human colonic bacteria. J. Food Biochem. 6:39-55. Saman, E. , M. Claeyssens and C. K. Bruyne. 1975. Bacillus pumilus 6-D-xylosidase: study of thiol groups. In Biochemical Society Transactions, 558th Meeting, Edinburgh. 3:998-999. Saul, D. J., L. C. Williams, R. A. Grayling, L. W. Chamley, D. R. Love, and P. L. Bergquist. 1990. celB . A gene coding for a bifunctional cellulase from the extreme thermophile " Caldocellum saccharolvticum " . Appl. Environ. Microbiol. 56:3117-3124. Sewell, G. W., H. C. Aldrich, D. Williams, B. Mannarelli, A. Wilkie, R. B. Hespell, P. H. Smith, and L. O. Ingram. 1988. Isolation and characterization of xylan-degrading strains of Butyri vibrio fibrisolvens from a Napier grass-fed anaerobic digester. Appl. Environ. Microbiol. 54:1085-1090. Sewell, G. W., E. A. Utt, R. B. Hespell, K. F. MacKenzie, and L. 0. Ingram. 1989. Identification of the Butvrivibrio fibrisolvens xylosidase gene f xylB ) coding region. Appl. Environ. Microbiol. 55:306-311. Teeri, T. T. , P. Lehtovaara, S. Kauppinen, I. Salovouri, and J. Knowles. 1987. Homologous domains in Trichoderma reesei cellulolytic enzymes: gene sequence and expression of cellobiohydrolase. Gene. 51:43-52. Uziie, M. , M. Matsuo, and T. Yasui. 1985. Possible identity of B-xylosidase and 6-glucosidase of Chaetomium triaterale . Agric. Biol. Chem. 49:1167-1173.

PAGE 160

150 Varel, V. H. 1987. Activity of fiber-degrading microorganism in the pig large intestine. J. Anim. Sci. 65:488-496. Ward, 0. P., and M. Moo-Young. 1989. Enzymatic degradation of cell wall and related plant polysaccharides, p. 237-274. In CRC CRitical Reviews in Biotechnology vol 8 . Weinstein, L. , and P. Albersheim. 1979. Structure of plant cell walls. Plant. Physiol. 63:425-432. Whitehead, T. R. , and R. B. Hespell. 1990. The gene for xylan-degrading activities from Bacteroides ovatus are clustered in a 3.8 kilobase region. J. Bacterid. 172:24082412. Wong, K. Y., L. U. L. Tan, and J. N. Saddler. 1986. Purification of a third distinct xylanase from the xylanolytic system of Trichoderma harzianum. Can J. Microbiol. 32:570-576. Yaguchi, M. , C. Roy, C. F. Rollin, M. G. Paice, and L. Jurasek. 1983. A fungal cellulase shows sequence homology with the active site of hen egg-white lysozyme. Biochem. Biophy. Res. Comm. 116:408-411.

PAGE 161

BIOGRAPHICAL SKETCH Eric Andrew Utt was born at the dawn of the space age, on February 20,1961, in Fort Lauderdale, Florida, to Harold and Anita Utt. He grew up as the youngest of two sons in Cocoa Beach, Florida. He attended Cocoa Beach High School where he graduated in 1979. In the fall of that year, he attended Brevard Community College before transferring to the University of Central Florida. He was awarded the Bachelor of Science degree in microbiology in 1984 and went on to pursue and receive a Master of Science degree in microbiology with Dr. Rudy J. Wodzinski at the University of Central Florida. In 1987, he began doctoral studies in the laboratory of Dr. Neal Ingram in the Department of Microbiology and Cell Science at the University of Florida in Gainesville. During his tenure in Dr. Ingram's laboratory, he became interested in the structural and functional relationships of proteins and microbial evolution. This interest expanded during his doctoral research and is an area in which he would like to continue to work. Dr. Utt has been awarded a National Research Council Postdoctoral Fellowship and will pursue research in the genetic characterization of virulence factors in Listeria monocytogenes at the Centers for Disease Control, Atlanta, Georgia. '»' , ' 151

PAGE 162

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. /X ^T^yu^ L' L^i^T-'Lonnie 0. Ingram / Chairman, Professor of Microbiology and Cell Science 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. Cell Science 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. Keelnatham T. Shanmugfem Professor of Microbiology and Cell Science 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. Henry
PAGE 163

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. CRarles M. Allen Professor of Biochemistry and Molecular Biology This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December 1991 Dean, /Cpliege of Agriculture Dean, Graduate School


Figure 17. SDS-PAGE analysis of pooled xylosidase-containing
fractions from preparative electrophoresis on 8% native PAGE:
A and E; molecular weight markers, B; hydrophobic column
purified preparation, C; Prep-Cell purified preparation, D;
crude extract.


112
1 /[ S] (mM)


84
Conclusions
The mutation and deletion data are consistant with the
proposal that both enzymatic activities exhibited by the
xvlB-encoded protein are not functionally separate but
depend upon the same region of the protein for complete
activity. The clustering of mutations about the consensus
sequence and the 60 amino acids region is strong evidence in
favor of this region being important for substrate-binding
and/or catalytic activity of the protein. This region of
the protein is rich in aspartic acid, glutamic acid and
histidine residues which have been previously implicated in
the catalytic function of related proteins such as lysozyme
(Quiocho, 1986), taka-amylase from Aspergillus nicer
(Matsuura et al., 1984), cellobiohydrolase II from
Trichoderma reesei (Rouvinen et al., 1990), and a
neopullulanase from Bacillus stearothermophilus (Kuriki et
al., 1991).
The point mutations that were expressed as complete
proteins did not affect either subunit assembly or the
apparent size of the native protein relative to the wild
type. These mutations did, however, change the protein's
mobility during electrophoresis on agarose-xylan native
gels. It is possible that this change in mobility is due to
a reduced affinity of the various mutant proteins for xylan,
which is functioning as a surrogate substrate for the


130
pH


GENETIC AND BIOCHEMICAL CHARACTERIZATION OF
THE B-D-XYLOSIDASE GENE FROM BUTYRIVIBRIO FIBRISOLVENS
By
ERIC ANDREW UTT
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
1991


35
encodes a 517 amino acid protein having a calculated
molecular weight of 58,421. 0RF3, is located 123 bp
downstream from ORF2 and includes a Shine-Dalgarno seguence
with an initiation codon 5 bp downstream. This ORF
continues for 1,173 bp until the end of the clone and is
also incomplete. No predicted "stemloop" structures or
sequences that resemble rho-independent transcriptional
terminators were identified by computer analysis in the
region between 0RF2 and 0RF3. It is therefore unlikely that
this DNA functions as a transcriptional terminator in E.
coli. It is also unlikely that any transcriptional
terminators are present between 0RF1 and 0RF2 since these
two ORF's are only separated by 15 bp and 0RF2 is expressed
in large amounts in E. coli in constructs that also contain
0RF1. This evidence suggests that these three ORF's may
constitute part of a xylan-degrading operon in B.
fibrisolvens.
Codon usage. The codon usage of the three B.
fibrisolvens ORF's is summarized in Table 1. For
comparative purposes, codon usage for B. fibrisolvens 49
xynA (Mannarelli et_al. 1990) and the average codon usage
for E. coli (Allf-Steinberger, 1984) are included. The
three B. fibrisolvens ORF's have similar patterns of codon
usage with each other and with strain 49 xvnA. The low
guanine plus cytosine content of B. fibrisolvens is
reflected in the three ORF's in the usage of an A or a T in


80


8
1969). Xylobiose was found to induce synthesis of 6-
xylanase in Crvptococcus albidus (Biely et al. 1980) and in
Streptomvces sp. (MacKenzie et al. 1987). Biely and
Petrakova (1984), studying the xylan-degrading system in C.
albidus. found that certain positional isomers of xylose and
xylobiose, notably 1,4-B-xylobiose, could serve as inducers
of B-xylanase and B-D-xylosidase.
Some organisms produce multiple xylanase enzymes.
Esteban et al. (1982) reported that Bacillus circulans WL-12
secretes two endo-6-xylanases and one B-D-xylosidase when
grown on xylan as a sole carbon source. Three distinct
xylanase genes have been identified and cloned from
Clostridium thermocellum (MacKenzie et al. 1989) A
multiplicity of xylanases has also been reported in fungi
including Aspergillus niaer (Frederick et al. 1985), and
Trichoderma harzianum ( Wong et al. 1986). It has been
suggested that the multiplicity serves to enhance the
ability of microbes to depolymerize a wide range of
substituted xylans under different environmental conditions.
In some organisms, the xylanolytic and cellulolytic
systems are combined. Recently Morag et al. (1990)
demonstrated that, in addition to free xylanases,
Clostridium thermocellum possessed a cellulosome-associated
xylanase which exhibits endo-glucanase activity. However
this organism was unable to utilize or grow on xylan. These
investigators postulated that cellulosome-associated


123


115
together in the assay system. An additive result would have
indicated that separate active sites might be present for
the two enzymatic activities. This data is consistent with
the results of the competitive inhibition experiments and
demonstrate that both substrates are competing with each
other for the same catalytic center on the enzyme.
Conclusions
The xvlB gene encodes an enzyme that exhibits substrate
ambiguity with respect to the two p-nitrophenol-derived
xylopyranosides and arabonofuranosides. The activities
against these two substrates co-purified during preparative
electrophoresis and were both stable as ammonium sulfate
pellets for up to one week at 5C. Only a single active
form of this enzyme is present on native-PAGE gels and
appears to have a molecular weight of 120,000.
The thermal stability and the optimal temperature were
essentially the same for both activities with activities
being stable up to 42C after 30 min and the optimal
temperature was 45C. The thermal stability of this enzyme
is relatively low when compared to xylosidases purified from
other microorganisms. The temperature optimum for this
enzyme is similar to xylosidases produced by Clostridium
acetobutylicum (Lee and Forsberg, 1987) and Bacillus pumilus
(Panbangred et al. 1984).
Investigations of optimal pH revealed a major
difference with respect to both activities. Both activities


4
principle focus of this research. The principle question
that was addressed relates to the dual activity exhibited by
this enzyme against B-D-xylopyranosides and a-L-
arabinofuranosides: do these two activities reside in the
same active center of the enzyme or are they on separate
domains?
The following research examined the structure and
function of the xvlB gene and gene product and includes:
A) The complete nucleotide sequence of xvlB and
sequence comparisons with related enzymes from other
organisms.
B) Genetic evidence that the two enzymatic activities
are encoded by a single open reading frame.
C) Mutational analyses to investigate the genetic
interdependence of the two enzymatic activities of
xvlB.
D) Purification and characterization of the xvlB-
encoded protein.
E) Biochemical and kinetic experiments to investigate
the functional relationship between the two enzymatic
activities.


Figure 10. Location and identification of point mutations in
xylB by DNA sequencing. Numbers above sequence indicate the
position of the amino acids. Amino acids in parentheses below
the sequence indicates the new amino acid inserted by
mutation. A denotes loss of enzymatic activity. A "w"
denotes weak enzymatic activity.


LITERATURE CITED
Allf-Steinberger, C. 1984. Evidence for coding pattern on
the non-coding strand of the Escherichia coli genome.
Nucleic Acids Res. 12:2235-2241.
Armstrong, D.G., and H.J. Gilbert. 1985. Biotechnology and
the rumen: A mini-review. J. Sci. Food Agrie.
36:1039-1046.
Barnett, C. C., R. M. Berka, and T. Fowler. 1991. Cloning
and amplification of the gene encoding an extracellular B-
glucosidase from Trichoderma reesei : evidence for improved
rates of saccharification of cellulosic substrates.
Biotechnol. 9:562-567.
Bastawde, K. B., L. B. Tabatabai, M. M. Meagher, M. C.
Srinivasan, H. G. Vartak, M. V. Rele, and P. J. Reilly.
1991. Catalytic properties and partial amino acid sequence
of an actinomycete endo-(l-4)-B-D-xylanase from Chainia
species. p417-425. In G. F. Leatham and M. E. Himmel (ed),
Enzymes in biomass conversion. Amer. Chem. Soc. Washington,
D. C.
Biely, P. 1985. Microbial xylanolytic systems. Trends in
Biotechnol. 3:286-290.
Biely, P., C. R. MacKenzie, J. Puls, And H. Schneuder. 1986.
Cooperativity of esterases and xylanases in the enzymatic
degradation of acetyl xylan. Biotechnol. 4:731-733.
Biely, P., and E. Petrakova. 1984. Novel inducers of the
xylan degrading system of Crvotococcus albidus. J.
Bacteriol. 160:408-412.
Biely, P., J. Puls, and Schneider. 1985. Acetyl xylan
esterases in fungalcellulolytic systems. FEBS Lett. 186:80-
84.
Biely, P., Z. Kratsky, M. Vranska, and D. Urmanicova. 1980.
Induction and inducers of endo-1,4-6-xylanase in the yeast
Crvotococcus albidus. Eur. J. Biochem. 108:323-329.
Benguin, P., P. Cornet, and J. P. Aubert. 1985. Sequence of
a cellulase gene of the thermophilic bacterium Clostridium
thermocellum. J. Bacteriol. 162:102-105.
142


14
encoding a protein having acetyl esterase activity.
Additional 6-D-xylosidase enzymes with multiple activities
have been reported earlier. Kinetic methods were used to
investigate the active site of a B-D-xylosidase from
Chaetomium trilaterale (Uziie et al. 1985). In this study,
which employed substrate analogues as inhibitors, a single
active center was postulated to function for both the B-D-
xylosidase and B-glucosidase activities exhibited by this
enzyme. A 6-D-xylosidase purified from Trichoderma reesei
was also found to exhibit an a-L-arabinofuranosidase
activity (Poutanen and Puls, 1988). Additionally, a cloned
gene cluster from Bacteroides ovatus was also found to
exhibit B-D-xylosidase and a-L-arabinofuranosidase
activities (Whitehead and Hespell, 1990) These dual
activities co-purified and were encoded by a single open
reading frame present in the cloned gene fragment. While
the dual activities of B-D-xylanases and B-D-xylosidases
have been documented, little is known about the genetic and
biochemical basis of this property. Multiple substrate
activities can be attributed to the presence of more than
one catalytic region on the enzyme. Another possibility is
the presence of a single catalytic region with wide
substrate specificity. It has been proposed that the more
evolved an enzyme or protein is, the more narrow its
specificity becomes (Knowles, 1988). Accordingly the more
primitive proteins tend to have multiple functions. It has


56
III. A series of genetic experiments were designed to
investigate the presence or absence of two catalytic or
functionally separate domains on the xvlB gene that are
responsible for the dual activities exhibited by this
enzyme.
Materials and Methods
Medium and growth conditions, genetic methods and DNA
sequencing were done as described in chapter III.
In vitro nitrous acid mutagenesis of xvlB. A total of
80 Mg of pLOI1005 which contains the xvlB gene was
resuspended in 50 Ml Tris-EDTA (TE) buffer (pH 8.0).
Mutagenesis was initiated by the addition of 10 Ml of 2.5 M
sodium acetate (pH 4.3) and 50 Ml 2.0 M sodium nitrite.
Exposure times were zero, thirty seconds, one, two five and
ten minutes. Mutagenesis reactions were terminated by the
addition of 200 m! 100% ethanol. The precipitation step was
repeated twice to ensure the complete removal of the
mutagenic agent. The mutagen-treated plasmids were
resuspended in 80 Ml TE buffer (pH 8.0). A total of 5 Ml of
the plasmid was transformed into competent E. coli DH5a.
Serial ten-fold dilutions of the transformed cells were
plated in triplicate onto Luria agar supplemented with the
fluorogenic substrates. A 99% reduction in transformation
by the mutagenized plasmid was observed after ten minutes of
mutagenesis.


149
Raynal,A., C. Gerbaud, M. C. Francingues, and M. Guerineau.
1987. Sequence and transcription of the B-glucosidase gene
of Kluvveromvces fraailis cloned in Saccharomvces
cerevisiae. Curr. Genet. 12:175-184.
Rerat, A., M. Fiszlewicz, A.Giusi, and P. Vaugelade. 1987.
Influence of meal frequency on postprandial variations in
the production of volatile fatty acids in the digestive
tract of conscious pigs. J. Anim. Sci. 64:448-453.
Robson and Gardier. 1988. Proteins and protein engineering,
p 699. Elsevier Science Publishers B. V.
Rouvinen, J., T. Bergfors, T. Teeri, J. K. C. Knowles, and
T. A. Jones. 1990. Three-dimensional structure of cellobio-
hydrolase II from Trichoderma reesei. Science 249:380-386.
Salyers, A. A., J. R. Balascio, and J. K. Palmer. 1981.
Breakdown of xylan by enzymes from Human colonic bacteria.
J. Food Biochem. 6:39-55.
Saman, E., M. Claeyssens and C. K. Bruyne. 1975. Bacillus
pumilus B-D-xylosidase: study of thiol groups. In
Biochemical Society Transactions, 558th Meeting, Edinburgh.
3:998-999.
Saul, D. J., L. C. Williams, R. A. Grayling, L. W. Chamley,
D. R. Love, and P. L. Bergquist. 1990. celB, A gene coding
for a bifunctional cellulase from the extreme thermophile
MCaldocellum saccharolvticum". Appl. Environ. Microbiol.
56:3117-3124.
Sewell, G. W., H. C. Aldrich, D. Williams, B. Mannarelli, A.
Wilkie, R. B. Hespell, P. H. Smith, and L. 0. Ingram. 1988.
Isolation and characterization of xylan-degrading strains of
Butvrivibrio fibrisolvens from a Napier grass-fed anaerobic
digester. Appl. Environ. Microbiol. 54:1085-1090.
Sewell, G. W., E. A. Utt, R. B. Hespell, K. F. MacKenzie,
and L. 0. Ingram. 1989. Identification of the Butvrivibrio
fibrisolvens xylosidase gene (xylB) coding region. Appl.
Environ. Microbiol. 55:306-311.
Teeri, T. T., P. Lehtovaara, S. Kauppinen, I. Salovouri, and
J. Knowles. 1987. Homologous domains in Trichoderma reesei
cellulolytic enzymes: gene sequence and expression of
cellobiohydrolase. Gene. 51:43-52.
Uziie, M., M. Matsuo, and T. Yasui. 1985. Possible identity
of 6-xylosidase and B-glucosidase of Chaetomium triaterale.
Agrie. Biol. Chem. 49:1167-1173.


7
is very high and constant while product accumulation is low.
It has been established that microbial cells in the rumen
are present in high numbers and contains: 1011 bacteria ml'1,
106 ciliate protozoa ml'1, and 104 fungi ml'1 (Patterson,
1989) .
Butvrivibrio fibrisolvens is a Gram variable,
obligately anaerobic, motile bacillus (Hespell and Bryant,
1981) B_;_ fibrisolvens is particularly abundant in the
rumen and anaerobic digesters in which plant material serves
as the primary substrate (Hespell et al. 1987). B.
fibrisolvens converts hemicellulose to mono and
oligosaccharides. These are transported and metabolized to
yield butyric acid. Mannarelli et al. (1990) cloned and
sequenced the gene encoding B-D-xylaase from B.
fibrisolvens strain 49. Sewell et al. (1988) isolated
several strains of B. fibrisolvens that produced both
xylanase and xylosidase. In this study, the synthesis of
both enzymes were concurrently repressed by glucose and
induced by xylan and xylose. This was surprising since
earlier work on rumen isolates of B. fibrisolvens had
reported that these enzymes were expressed constitutively
(Hespell et al. 1987). Similarly, it was reported that
xylose served as an inducer of the xylanase and fl-D-
xylosidase in Pullularia pullulans ( Pou-Llinas and Driguez,
1987). In addition, B-D-xylosidase of Bacillus pumi1 ns was
found to be induced by xylose (Kersters-Hilderson et al.


48
TABLE 5. Amino acid sequence alignment of conserved regions.
Protein* Concensus sequence Reference
HEWL
A. n.
B.f.
B.f.
B.f.
B.f.
B.f.
B.p.
35 44
FES
N
F N
T
Q
A
T
. N R
N
T D
G
S


331
338
(a-aml)
P E D
T
y .
y
N
G

. N P
W
F L
C
T
L A
A
342
350
xyl.Bi
SED
F
y s
L
T
D

. N P
G
F L
R
L
K L
R
144
149
xylJ,
P D G
V
r y
A
W E
I
W
V Q
E
320
328
endl
GET
S
A T
N
R
N

. N T
A
E R
V
K
W A

355
362
xvnA
NEK
P
L I
W
S


. N I
G
V A
K
P
a y

769
ball
S D W
W
G F
G
E
H
y
K .
E
V
L A
G
325
335
xvnB
I E C
T
R L
A
Q
L
N
W N T
C
S M
Q
F
V .

52
t d y
Yaauchi et al. 1983
349
A E Q
Boel et al. 1984
362
PEA
This study
160
L D L
This study
337
. D y
Beraer et al. 1989
369
. D E
Mannarelli et al. 1990
785
N D I
Barnett et al. 1991
343
. E E
Morivama et al. 1987
* Abbreviations: Hen egg-white lysozyme (HEWL), Aspergillus niger a-amylase (a-
aml)/ B. fibrisolvens fl-D-xylosidase (xvlBl. B. fibrisolvens endoglucanase 1 (endl),
B. fibrisolvens B-D-xylanase (xvlA). B. fibrosolvens B-glucosidase (ball). Bacillus
pumillus B-D-xylosidase (xylB), Clostridium thermocellum fl-D-xylanase (xvnZ). C.
thermocellum cellobiohydrolase B (celB), C. thermocellum cellobiohydrolase D (celD),
C. cellulolyticum endoglucanase A (EGCCA), Caldocellum saccharolvticum B-D-
xylosidase (xynB), Cellulomonas fimi endoglucanase A (cenA), C. fimi exoglucanase A
(cex), Trichoderma reesei cellobiohydrolase II (CBH II), Kluvveromvces fraailis
cellobiohydrolase I (CBH I), Schizophvllum commune endoglucanase (EG I).


11
xylanases are in fact endo-xylanases by virtue of the fact
that they attack the interior B-(1,4)-D-xylosidic linkages
of the xylan polymer rather than the exterior linkages (Ward
and Moo-Young, 1989). The fl-(1,4)-D-endo-xylanases have a
pH optimum in the range of 3.5 to 6.5 while the temperature
optima and thermal stabilities vary depending upon the
source (Ward and Moo-Young, 1989). Xylaase from Bacillus
pumilus IPO has a molecular weight of 22,000 and is a B-D-
1,4-endo-xylanase (Panbangred et al. 1983). The pH and
temperature optimum of this enzyme are 6.5 and 40C,
respectively. Quantification of the hydrolysis end products
from larchwood xylan indicated that the B. pumilus enzyme
had the greatest affinity for the second and sixth B-
xylosidic linkages of the polymer.
The xvnZ gene product from Clostridium thermocellum is
also an endo-xylanase with pH and temperature optima of 6.0
and 65C, respectively (Grepinet et al. 1988). Lee et al.
(1987) purified and characterized two different endo-
xylanases, xylanase A and xylanase B, from Clostridium
acetobutvlicum. Xylanase A has a molecular weight of
65,000, a pH optimum of 5.0, an optimum temperature of 50C,
and is stable for up to 30 min at 40C. Xylanase B is a
smaller protein having a molecular weight of 29,000. It had
a pH optimum of 5.0 to 6.0, showed a temperature optimum of
60C, and is stable for 30 min at 50C. Both enzymes
hydrolyze larchwood xylan randomly, however xylanase B


89
the concentrate was then precipitated by the addition of
solid ammonium sulfate to 70% saturation. The enzyme was
stored as an ammonium sulfate pellet at 5C until needed.
No loss of enzymatic activity was detected after storage for
one week under these conditions. Pellets were resuspended
in 5 mM phosphate (pH 7.0) containing 10 mM R-
mercaptoethanol immediately prior to use.
Hydrophobic interaction chromatography. The
xylosidase-containing pellet from preparative
electrophoresis was resuspended in 1 ml of 1.7 M (NH4)2S04 in
5 mM phosphate buffer (pH 6.8) and loaded onto a 2.0 x 3.0
cm Pharmacia XK 16/20 chromatography column (Pharmacia LKB,
Uppsala, Sweden) packed with Toyopearl "TSK-Gel" hydrophobic
gel (Supelco, Inc., Bellefonte, PA.). The column was
equilibrated with 1.7 M (NH4)2S04 in 5 mM phosphate buffer
(pH 6.8) prior to the addition of sample. Xylosidase was
eluted using a linear negative salt gradient starting with
1.7 M (NH4)2S04 down to zero in a total volume of 200 ml.
Fractions were collected in 3 ml volumes and analyzed as
described below.
Enzyme assays. Xylosidase activity in each fraction
was assayed using p-nitrophenyl-B-D-xylopyranoside (p-NP-X)
at a final concentration of 2.5 mM unless otherwise noted
and in 50 mM phosphate buffer (pH 6.8) at 37C. Assays were
done in a total volume of 1 ml and allowed to continue until
the yellow color indicating enzyme activity was detected.


62
Enzvme assays. Enzyme assays were done as described in
chapter III.
Sodium dodecvl sulfate polyacrylamide gel
electrophoresis (SDS-PAGE). SDS-PAGE gels were done as
described in chapter III.
Native polyacrylamide gel electrophoresis (native
PAGE). Cell proteins were separated in non-denaturing gels
by the method of Ornstein and Davis (1964). Following
electrophoresis, gels were equilibrated in 50 mM sodium
phosphate buffer (pH 6.8). The equilibrated gels were then
overlaid with Whatman #1 filter paper soaked with a solution
of 20 mg per ml of the fluorogenic substrates, 4-
methylumbelliferyl-B-D-xylopyranoside or 4-
methylumbelliferyl-a-L-arabinofuranoside in 70% ethanol.
Overlays were incubated at 37C for 15 min or until activity
bands were visible under long-wave UV light. Native-PAGE
gels were also stained for protein as outlined before.
Western hybridization analysis of wild type and mutant
proteins. Native and SDS-PAGE protein gels were
electroblotted using the Trans-Blot apparatus (BioRad
Laboratories, Richmond, CA.) according to the manufacturer's
instructions. "Western" hybridizations were done using
polyclonal antisera raised to E. coli DH5a cell extracts
harboring pLOI1005 in rabbits. Protein bands were
visualized using alkaline phosphatase conjugated goat anti-


55
The active center for a neopullulanase from Bacillus
stearothermophilus has been recently examined using
mutagenesis (Kuriki et al.. 1991). This enzyme exhibits
dual activities against a-(1,4) and a-(1,6) glycosidic
linkages. The catalytically important amino acids were
tentatively identified using seguence alignment and homology
searches. The putative catalytic amino acids were changed
using site-directed mutagenesis and activities were examined
in the resulting mutants. This approach identified that one
active center containing Glu-357 and Asp-424 was responsible
for both catalytic activities.
Enzymes which exhibit substrate ambiguity are
interesting both from mechanistic and evolutionary
perspectives. The obvious question that arises with respect
to bifunctionality is does this enzyme have two separate,
specialized, catalytic sites? This situation would imply
that the protein has evolved from a gene fusion to perform
two separate functions. Another possibility involves the
presence of a single active site in which two structurally
similar substrates are bound and hydrolyzed. This situation
would be analogous to a case of mistaken substrate identity
which proves advantagous to the organism, and has been
evolutionarily conserved. Examples of multifunctional
xylosidases have been reported in the literature and are
discussed in the previous chapters. The apparent
bifunctionality of the xvlB gene was demonstrated in chapter


53
Since many enzymatic activities are required to completely
depolymerize xylans and cellulose, the evolution of such
enzymes could represent a selective advantage in the rumen
and other environments.


19
mercaptoethanol and were lysed by two passes through a
French pressure cell at 20,000 lb in'2. Cell membranes and
other debris were removed by centrifugation (100,000 g, 1 h,
4C). Supernatants containing the total cytoplasmic
proteins were stored at -70C.
Enzyme assays. B-D-xylosidase and a-L-
arabinofuranosidase activities were determined by measuring
the rate of hydrolysis of p-nitrophenyl-B-D-xylopyranoside
and p-nitrophenyl-a-L-arabinofuranoside (1 mM final
concentration), respectively, in 50mM phosphate buffer (pH
6.8) at 37C. The nitrophenyl derivatives of other mono-
and disaccharides were examined as possible substrates under
the same conditions. All assays were conducted in a volume
of 1 ml catalysis was terminated by the addition of 2 ml of
500 mM sodium carbonate. The hydrolysis of one nmole of
substrate resulted in an increase of absorbance of 0.007 at
405nm. Specific activities are expressed as nmoles p-
nitrophenol released per minute per milligram of total
protein. Carbohydrate derivatives were purchased from Sigma
Chemical Co. Protein concentration was estimated by the
method of Bradford (Bradford, 1976).
Sodium dodecvl sulfate-polvacrvlamide gel
electrophoresis (SDS-PAGE). Cell proteins were separated in
denaturing gels by the method of Laemmli (Laemmli, 1970).
Protein bands were visualized by staining with Coomassie
blue.


61
kilobase pairs
1.0 2.0 3.0
pUC18
LacZ=-
E
SP
pLOI1005
H
S P
PstI
EcoRI
Hindlll
pUC18
7
!
LacZ=^
PLOI1051
pUC18
S P
LacZ=
PLOI1052
PLOI1053


43
the glycosidic bond-breaking step (Martinez-Bilbao et al.,
1991). Under the conditions of these assays it appears that
the xvlB-encoded protein is limited to hydrolytic activity
against 6-D-xylopyranosides and a-L-arabinofuranosides only.
Electrophoretic analysis of cloned proteins. Using
SDS-PAGE analysis of cell-free cytoplasmic extracts, a new
protein band with an apparent molecular weight of 60,000 was
observed in cells harboring plasmid pLOI1005 (Fig. 6).
This band was absent in extracts from cells containing the
vector plasmid pUC18 alone. Extracts from which the gene
was inactivated by a frameshift mutation (pLOI1040) also
lacked this protein band. The observed levels of this
protein band in the single (pLOI1043) and double (pLOI1050)
SspI subclones was consistent with the presence of the
enzyme.
Primary sequence homology comparisons. Homologies of
the B. fibrisolvens ORF's to other glycohydrolases were
compared to determine evolutionary relatedness. The
translated amino acid sequences of the three B. fibrisolvens
ORF's exhibited 42 to 45% similarity (a conservative match)
and 14 to 19% identity (an exact match) with each other
(Table 4.). The xvlB was found to be most similar (44%
similarity, 20% identity) to the B-glucosidase from
Kluyveromyces fragilis. Additional comparisons with other
glycohydrolase sequences revealed no significant amino acid
identities. Although the N-terminal sequence of the


141
The substrate competition experiments corroborate the
genetic evidence and demonstrated that the two activities
reside in the same active center of the xvlB-encoded enzyme.
This study demonstrates that the xvlB gene from
Butvrivibrio fibrisolvens encodes a single protein having B-
D-xylosidase and a-L-arabinofuranosidase activities. These
activities are localized in the same active site of the
protein.


47
Bacillus pumilus B-xylosidase did exhibit strong amino acid
identity in selected regions, the overall identity was only
21%. Thus the B. fibrisolvens xvlB gene is evolutionarily
divergent from other glycohydrolases. The translated,
primary sequences for 0RF1 and 0RF3 also exhibited
similarity (52% and 46%, respectively) and identity (31% and
22%, respectively) to the K. fraqilis S-glucosidase. This
is consistant with these two ORF's also being involved in
carbohydrate degradation.
It has been postulated that the hydrolytic mechanism of
lysozyme (Teeri, et al. 1987) and cellulases (Knowles et
al., 1987) can serve as a model for other carbohydrate
hydrolyzing enzymes. Studies of hen egg-white lysozyme
(HEWL) indicate a general acid-base catalytic mechanism
involving Glu-35 and Asp-52 as the catalytic residues
(Quiocho, F. A., 1986). Subsequent studies have
demonstrated that this catalytic region is conserved in some
cellulases (Knowles et al. 1987). An analysis of the
translated primary sequence from xvlB reveal a region
homologous to the active site region from HEWL, and
glucoamylase from Aspergillus niqer (Table 5). The
conserved region from additional carbohydrate hydrolases are
included for comparison. The xvlB region was most similar
to the glucoamylase, with 38% identity between the amino
acids in the catalytic region. The catalytically important


CHAPTER I
GENERAL INTRODUCTION
Plant cell walls represent the largest reserve of fixed
carbon on earth. Plant cell walls are composed primarily of
cellulose, hemicellulose, and lignin (Weinstein and
Albersheim, 1979). Cellulose is the most abundant
carbohydrate found in plant biomass (Coughlan, 1985) while
hemicellulose is a major plant structural polymer that ranks
second only to cellulose in natural abundance (Dekker and
Richards, 1976). The amount of hemicellulose in dry wood is
between 20% and 30% (Eriksson et al. 1990). The composition
of hemicellulose varies between softwoods and hardwoods.
The major hemicellulose in softwoods is galactoglucomannan
(Eriksson et al. 1990). This polymer has a backbone
composed of a linear chain of 1,4-linked fi-D-glucopyranose
and B-D-mannopyranose units. The mannose and glucose
moieties of the backbone may be substituted with acetyl
groups at the C-l and C-2 positions.
Glucouronoxylan (O-acetyl-4-O-methyl-glucurono--D-
xylan) is the major hemicellulosic component of hardwoods
and agricultural residues (Eriksson et al. 1990). The major
structural feature of xylan is a linear chain consisting of
1


LIST OF FIGURES
Figure Page
1.Restriction maps of pUC18 derived plasmids
that express fi-D-xylosidase activity in
E. coli DH5a 22
2.Subclone analysis of pLOI1005 to localize
the xylB coding region 24
3.Southern hybridizations of chromosomal
DNA from B. fibrisolvens and E. coli 27
4.Outline of sequencing strategy of pLOHOOl
and subclone analysis of xvlB 29
5. The complete nucleotide sequence and
translated amino acid sequence of the
4.2 kb insert from pLOHOOl 31
6. SDS-PAGE analysis of cytoplasmic extracts
from recombinant E. coli DH5a harboring
selected constructs 4 5
7. Assignment of domains to the xvlB gene 59
8.Subcloning strategy used to localize in vitro
mutations to one of five domains in the
xvlB gene 61
9.Deletion analysis of the xvlB gene 65
10.Localization and identification of point
mutations in xvlB by DNA sequencing 69
11. SDS-PAGE analysis of wild type and mutant
proteins 73
12. Native-PAGE comparison of W158UGA and L178F
mutations with the wild type stained with
Coomassie blue 75
13. Western hybridization of native-PAGE of wild
type and mutant proteins 77
vi


114
1/[S] ( mM)


137
bias of these three ORF's is consistent with the low guanine
plus cytosine ratio of the DNA from this organism (Table 1).
(E) The xvlB-encoded protein shares limited amino acid
identity with other published amino acid sequences of B-D-
xylosidase enzymes and related proteins (Table 4). This is
an indication that the xvlB gene from B. fibrisolvens is
evolutionarily divergent from these other genes. (F) A
consensus sequence has been identified in xvlB that has
significant identity to an amino acid sequence that has been
previously implicated in catalytic function of other
glycohydrolases (Table 5).
Mutational analysis of the xvlB-encoded protein has
revealed several important structural and functional
relationships. (A) All point mutations within the xvlB gene
resulted in a protein in which both enzymatic activities
were reduced or abolished (Table 6). (B) All point
mutations that were isolated are localized or clustered in
single region of the protein that is near the proposed
catalytic center (Fig. 10). (C) All but two of the point
mutations result in a stably expressed protein (Fig. 11).
(D) All of the point mutations that result in expressed
protein exhibit a decrease of affinity in native gel assays
(Fig. 14) and an increase in apparent Km. These results
demonstrate that the catalytic center for these two
enzymatic activities is not functionally independent but is
Idealized on the same region of the xvlB-encoded protein.


Figure 23. Competitive inhibition of xylosidase activity by
4-methylumbelliferyl-Q!-L-arabinofuranoside: closed circles; no
inhibitor, open circles; 100 ;M inhibitor, open squares; 250
/LtM inhibitor.


Figure 5. The complete nucleotide sequence and translated
amino acid sequence of the 4.2 kb insert from pLOHOOl.
Putative Shine-Dalgarno (S.D) sequences and initiation codons
are underlined. Translational termination is indicated by an
asterisk (*) .


49
TABLE 5. (continued)
Protein*
Concensus sequence
Reference
458
466
475
C.t. xvnZ
G E A
L
L
R
A
D
V

. N R
S
G
K
V
D
S


T D Y
418
427
436
C.t. celB
TEG
G
H
P
L
L
D
L
. N L
K
Y
L
R
C
M
R

. D F
376
384
395
C.t. celD
DEE
Y
L
R
D
F
E

. N R
A
A
Q
F
S
K
K
E
A D F
408
418
427
C.s. xvnB
REV
F
V
E
R
I
D
E
Y N A
N
P
K
R
V
W
L

. E M
244
254
263
T.r. CBH II
L E C
I
N
Y
A
V
T
Q
L N L
P
N
V
A
M
Y
L

. D A
586
596
605
K.f. CBH I
G E W
E
T
E
G
Y
D
R
E N M
D
L
P
K
R
T
N

. E L
33
42
50
SC EG X
N E S
C
A
E
F
G
N
Q
. N I
P
G
V
K
N



T D Y
Grepinet et al. 1988
Grepinet and Benguin,
1986
Joliff et al. 1986
Luthi et al. 1990
Rouvinen et al. 1990
Raynal et al. 1987
Yaguchi et al. 1983
* Abbreviations: Hen egg-white lysozyme (HEWL), Aspergillus niger a-amylase (a-
aml), B. fibrisolvens fi-D-xylosidase (xvlB). B. fibrisolvens endoglucanase 1 (endl),
B. fibrisolvens fi-D-xylanase (xylA), B. fibrosolvens B-glucosidase (ball 1. Bacillus
pumillus fi-D-xylosidase (xvlBl. Clostridium thermocellum B-D-xylanase (xvnZ). C.
thermocellum cellobiohydrolase B (celB), C. thermocellum cellobiohydrolase D (celD),
C. cellulolvticum endoglucanase A (EGCCA), Caldocellum saccharolvticum B-D-
xylosidase (xvnB), Cellulomonas fimi endoglucanase A (cenA), C. fimi exoglucanase A
(cex), Trichoderma reesei cellobiohydrolase II (CBH II), Kluvveromvces frgilis
cellobiohydrolase I (CBH I), Schizophvllum commune endoglucanase (EG I).


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EM74NPG46_RQBIIS INGEST_TIME 2015-02-19T21:07:19Z PACKAGE AA00028764_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


18
(Sigma Chemical Co., St. Louis, MO.). The internal Sau3A
and large internal PstI fragment from pLOI1005 were utilized
as probes in the
Southern hybridization analysis of digested chromosomal B.
fibrisolvens and E. coli DNA.
DNA sequencing. Double-stranded DNA was sequenced in
both directions using the dideoxy-chain termination method
(Sanger, 1982) and Sequenase (United States Biochemical
Corp.) according to the manufacturer's instructions.
Additional sequencing primers were synthesized by the
University of Florida Interdisiplinary Center for
Biotechnology Research and the Department of Microbiology
and Cell Science Nucleotide facility. The DNA sequences
were assembled using the "GENEPRO" software package (Hoefer
Scientific Instruments, San Francisco, Calif.) and the
University of Wisconsin Genetics Computer Group GCG package,
version 6.1 (Devereux et al. 1984) Primary sequence
comparisons were made with GenBank and EMBL sequence
libraries.
Preparation of cell extracts. E. coli cells harboring
the recombinant plasmids were harvested while in mid
exponential phase of growth by centrifugation (10,000 g, 10
min, 4C) and washed twice with 5 mM phosphate buffer (pH
6.8). Cell pellets were stored at -70C, until needed.
Cell pellets were thawed on ice and resuspended in an equal
volume of 5 mM phosphate buffer (pH 6.8) containing 10 mM B-


Figure Page
14. Substrate binding native gel western
hybridization assays of wild type and mutant
proteins 80
15. Elution profile of the xvlB gene product
during preparative electrophoresis on the
BioRad-Prep Cell system 93
16. Elution profile of the xvlB gene product
during hydrophobic interaction chromatography.. 95
17. SDS-PAGE analysis of pooled xylosidase-
containing fractions from preparative
electrophoresis on 8% native-PAGE 99
18 Native-PAGE analysis of the B-D-xylosidase 101
19. Thermal inactivation profile of xylosidase and
arabinofuranosidase activities 103
20. Temperature optimum profile for xylosidase and
arabinofuranosidase activities 105
21. pH activity profiles for xylosidase and
arabinofuranosidase activities 107
22. Double recipricol plots of xylosidase and
arabinofuranosidase activities of the native
protein 110
23. Competitive inhibition of xylosidase activity
by 4-MU-a-L-arabinofuranoside 112
24. Competitive inhibition of arabinofuranosidase
activity by B-methyl-D-xyloside 114
25. Elution profile of the xvlB-encoded protein
harboring the L178F mutation during
preparative electrophoresis 12 3
26. SDS-PAGE analysis of partially purified
L178F mutant protein 125
vii


1261
1321
1381
1441
1501
1561
1621
1681
1741
1801
1861
1921
1981
2041
2101
2161
2221
2281
2341
2401
2461
32
1320
1380
1440
1500
1560
1620
1680
1740
1800
1860
1920
1980
2040
2100
2160
2220
2280
2340
2400
2460
2520
AGTAAAGAACTTATGGGCAGAGAGTGTGATATTTTTGAAATTGAATTAACAGGCTCTGTT
SKELMGRECDIFEIELTGSV
ACAGAAGTTGAATATTAATTGAQAOOTGCATCATGGTTATAGCTAACAATCCAATTTTAA
T E V E Y MVIANNPILK
AAGGTTTTTATCCAGACCCTTCTATCTGCAGAAAAGGGGATGATTTTTATCTAGTTTGTT
GFYPDPSICRKGDDFYLVCS
CAAGTTTTGTGTATGCTCCGGGAGTACCGATTTTTCACACTAAGGATTTGGCACATTTTG
SFVYAPGVPIFHTKDLAHFE
AGCAAATTGGAAATATATTAGACAGAGAAAGTCAACTTCCATTGTCGGGAGATATATCTA
QIGNILDRESQLPLSGD I SR
GAGGCATATTTGCCCCAACAATAAGAGAGCATAATGGAATCTTTTACATGATAACAACTA
GIFAPTIREHNGIFYMITTN
ATGTAAGCTCTGGCGGCAACTTTATTGTTACTGCAAAAGATCCAGCTGGTCCTTGGTCAG
VSSGGNFIVTAKDPAGPWSE
AGCCATATTATTTAGGTGAAGATGAGGCGCCAGGTATTGATCCATCTCTGTTTTTTGATG
PYYLGEDEAPGIDPSLFFDD
ACGATGGCAAATGTTATTACGTTGGTACCAGACCAAATCCTGATGGAGTTCGTTACAACG
DGKCYYVGTRPNPDGVRYNG
GTGATTGGGAGATATGGGTTCAAGAGCTGGATTTAGAGCAAATGAAACTTGTAGGTCCTT
DWE IWVQELDLEQMKLVGPS
CGATGGCAATTTGGAAGGGCGCTCTTAAGGATGTTATTTGGCCAGAAGGACCACACCTTT
MAIWKGALKDVIWPEGPHLY
ATAAGAAAGATGGATATTATTATCTTTTACATGCAGAAGCTGGCACAAGCTTTGAACATG
KKDGYYYLLHAEAGTSFEHA
CTATTTCTGTAGCTCGCTCAAAGGAGCTATTCAAATGGTTTGAGGGATGTCCTAGAAATC
I SVARSKELFKWFEGCPRNP
CTATATTTACCCATAGAAATTTAGGCAAGGATTATCCAGTATGCAATGTTGGACATGCTG
I FTHRNLGKDYPVCNVGHAD
ATTTAGTTGATGATATCAATGGCAACTGGTATATGGTGATGCTGGCATCTAGACCATGCA
LVDDINGNWYMVMLASRPCK
AGGGAAAGTGCAGCTTGGGACGAGAGACATTCCTTGCAAAAGTAATTTGGGAAGACGGAT
GKCSLGRETFLAKVIWEDGW
GGCCAGTGGTTAATCCGGGAGTTGGTCGTTTGACTGATGAGGTGGAGATGGACCTTCCTG
PVVNPGVGRLTDEVEMDLPE
AATATAGATTCTCAAAAGAGATTACTACAAAGGATAAAATGACCTTTGAAGAGACAGTCC
YRFSKEITTKDKMTFEETVL
TAGATGATAGATTTGTTGGAATTGAAAGAAGAAGTGAGGACTTTTATTCCCTTACTGACA
DDRFVGIERRSEDFYSLTDN
ATCCTGGATTCTTAAGATTAAAGCTTCGTCCTGAGGCCATAGAAAATACTGGCAATCCAT
PGFLRLKLRPEAIENTGNPS
CTTACTTAGGAATTCGTCAAAAGACTCATTCGTTTAGAGCAAGCTGTGGCCTTAAGTTTA
YLGIRQKTHSFRASCGLKFT


27
6


Figure 27.
Thermal inactivation profiles of xylosidase and
arabinofuranosidase activities of the L178
mutant protein: open squares; xylosidase
activity, closed squares; arabinofurnanosidase
activity.


101


Figure 9. Deletion analysis of xvlB gene. (A) Exonuclease III
deletion series from the 3' end of xvlB. Shaded arrow denotes
direction of transcription from the pUC18 lac promoter. Solid
arrow denotes direction of deletion. Underlined "taa"
indicates relative location of stop codon from the 31 end of
deletion. Retention or loss of respective enzyme activites is
indicated by a "+" or (B) LacZ' fusion of 5' end of the
large internal PstI fragment of xvlB resulting in a deletion
of 56 base pairs.


Figure 14. Substrate binding native gel Western hybridization
assays of wild type and mutant proteins. Lane assignments: A
, F, and G; native (pLOI1005), B; G186R, C; A203T, D; W158UGA,
E; L178F, G; A203V, H; G238D, I; A210T, J; A203T. Arrow
indicates the direction of protein migration. The "+" and
indicated the relative location of the anode and cathode,
respectively.


CHAPTER VII
SUMMARY AND GENERAL CONCLUSIONS
The studies presented here have characterized the B-D-
xylosidase from the rumen bacteria Butvrivibrio fibrisolvens
using genetic and biochemical techniques. This enzyme is
important in the final steps of depolymerization of
hemicellulose in that the products of hydrolysis, usually
monomers, are used directly in the metabolism of the
bacterium. This enzyme is particularly interesting in that
it exhibits substrate ambiguity. Both B-D-xylopyranosides
and a-L-arabinofuranosides are hydrolyzed by this enzyme.
Previous studies involving B-D-xylosidase enzymes from
other organisms have shown that substrate ambiguity among
the xylosidase and related enzymes is not universal. The
xylosidase from Bacillus pumilus only exhibits activity
against aryl-B-D-xylopyranosides (Panbangred et al. 1983) .
The same is true for the B-D-xylosidase from B. subtilis
(Paice et al., 1986). Notably a B-D-xylosidase from
Caldocellum saccharolvticum was cloned and sequenced that
also exhibited endoxylanase activity (Luthi et al. 1990).
In this case the two substrates are structurally identical
but vary in size with the endoxylanase acting on the longer
chain length polymers while the xylosidase is specific for
135


12
produced only xylotriose and xylobiose as products whereas
xylanase A also yields xylohexose, -pentose, and -treaose as
end products. Xylanase A was also active against
carboxymethylcellulose, acid-swollen cellulose and lichenin.
The two enzymes were antigenically different as judged by
"Ouchterlony"-immunodiffusion assays. The two enzymes were
therefore presumed to be encoded by separate genes.
Some xylaases, such as those that are produced by
fungi, notably Aspergillus niger, produce endo-1,4-6-D-
xylanases that can hydrolyze the 1,3-a-L-arabinofuranosyl
side chains from arabinoxylans (Dekker, 1985). These
enzymes have been termed "debranching" xylanases. Recently
a unique "appendage-dependent" xylanase was isolated and
purified from Bacillus subtilis (Nishitani and Nevins,
1991). This enzyme is classified as a B-(1,4)-xylan
xylanohydrolase and has a prerequisite for glucuronosyl
substituted side chains in order to initiate hydrolysis of
the xylan backbone structure. Three novel xylanases were
purified from B. subtilis which exhibited activity against
ferulylolated arabinoxylans (Nishitani and Nevins, 1988).
These enzymes acted on ferulic acid-substituted arabinoxylan
and liberated the terminal arabinofuranosyl, terminal
gluconopyranosyl, and ferulic acid moieties from the
polymer. While much recent work has concentrated on the
extracellular microbial xylanases, less is known concerning
the molecular biology and properties of the intracellular 6-


Pst!


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.
' * l 0
/a
Lonnie 0. Ingram /
Chairman, Professor of
Microbiology and Cell Science
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.
Cell Science
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.
- \ c ,
Keelnatham T. Shanmugam
Professor of Microbiology and
Cell Science
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.
Henry ti. Aldrich
Professor of Microbiology and
Cell Science


128
The pH profile for both enzymatic activities for the
mutant protein was essentially the same as that for the wild
type protein (Fig. 28). Xylosidase activity had a sharp
activity peak at pH 6.0 which was followed by a rapid
decline. Arabinofuranosidase activity also peaked at pH 6.0
but remained at 90 % relative activity up to pH 9.0.
Kinetic properties. A characteristic Michaelis-Menten
linear relationship was observed for the mutant protein with
respect to both substrates (Fig. 29). The Km values for the
xylosidase and arabinofuranosidase activities were 16 mM and
33 mM respectively. This represents a decrease in the
affinity of the enzyme for both substrates. The increase in
Km agrees with the data in chapter IV using the agarose-
xylan gel binding assays.
Conclusions
It is apparent that the replacement of leucine with
phenylalanine at position 178 decreases both the affinity of
the enzyme for both substrates and the catalytic efficiency
of the enzyme. This amino acid substitution also increases
the thermal stability of the xvlB protein. It is possible
that these three properties are related. The increase in
thermal stability could represent a general increase in the
stability of the enzyme by strengthening hydrophobic
interactions at the core of the protein.
An increase in the overall stability of the protein
could influence the plasticity of the active center. The


69
Phenotype
Region of nutation xyl/ara
180 186 192
Asp.Val.Ile.Trp.Pro.Glu.Gly.Pro.His.Leu.Tyr.Lys.Lys
(Arg) -/-
197 203 209
Tyr.Leu.Leu.His.Ala.Glu.Ala.Gly.Thr.Ser.Phe.Glu.His
(Thr)
204 210 216
Gly.Thr.Ser.Phe.Glu.His.Ala.Ile.Ser.Val.Ala.Arg.Ser
(Val)
1 7 13
Met.Val.Ile.Ala.Asn.Asn.Pro.lie.Leu.Lys.Gly.Phe.Tyr
(Leu)
172 178 184
Ala.lie.Trp.Lys.Gly.Ala.Leu.Lys.Asp.Val.lie.Trp.Pro
(Phe)
146 152 158
Val.Arg.Tyr.Asn.Gly.Asp.Trp.Glu.lie.Trp.Val.Gin.Glu
(UGA)
197 203 209
Thr.Leu.Leu.His.Ala.Glu.Ala.Gly.Thr.Ser.Phe.Glu.His
(Val)
232 238 244
Phe.Thr.His.Arg.Asn.Leu.Gly.Lys.Asp.Tyr.Pro.Val.Cys
(Asp)
204 210 216
Gly.Thr.Ser.Phe.Glu.His.Ala.He.Ser.Val.Ala.Arg.Ser
(Thr)
197 203 209
Thr.Leu.Leu.His.Ala.Glu.Ala.Gly.Thr.Ser.Phe.Glu.His
(Thr)
23
Ser.ILe.Cys.Arg.Lys.Gly...
-/-
-/-
w/w
w/w
-/-
-/-
-/-
-/-
-/-
Mutant
G186R
A203T
A210V
P7L
L178F
W158UGA
A203V
G238D
A210T
A203T
f s


125


143
Berger, E., W. A. Jones, D. T. Jones, and D. R. Woods. 1989.
Cloning and sequencing of an endoglucanase (endl) gene from
Butvrivibrio fibrisolvens H17c. Mol. Gen. Genet. 219:193-
198.
Boel, E., M.T. Hansen, I. Hjort, and N.P. Fiil. 1984. Two
different types of intervening sequences in the glucoamylase
gene from Aspergillus niaer. EMBO J. 3:1581-1585.
Bradford,M. M. 1976. A rapid and sensitive method for the
quantitation of microgram quantities of proteins utilizing
the principle of protein-dye binding. Anal. Biochem. 72:248-
254.
Chesson, A., A. H. Gordon, and J. A. Lomax. 1983.
Substituent groups linked by alkali labile bondsto arabinose
and xylose residues of legume, grass, and cereal straw walls
and their fate during digestion by rumen microorganisms. J.
Sci. Food. Agri. 34:1330-1340.
Cotta, M. A., and R. B. Hespell. 1986. Proteolytic activity
of the ruminal bacteria Butvrivibrio fibrisolvens. Appl.
Environ. Microbiol. 52:51-58.
Coughlan, M. P. 1985. Properties of fungal and bacteria
cellulases with comment on thier production and application.
Biotechnol. Genet. Eng. Rev. 3:39-109.
Dehority, B. A. 1968. Mechanism of isolated hemicellulose
and xylan degradation by cellulolytic rumen bacteria. Appl.
Microbiol. 16:781-786.
Dehority, B. A. 1966. Characterization of several bovine
rumen bacteria isolated with a xylan medium. J. Bacteriol.
91:1724-1729.
Dekker, R. F. H. 1985. Hemicellulose degradation, p 505. In
Higuchi, T. (ed), Biosynthesis and biodegradation of wood
components. Academic Press, Orlando.
Dekker, R. F. H., and G. N. Richards. 1976. Hemicellulases:
Their occurrence, purification, properties, and mode of
action. Adv. Carbohydr. Chem. Biochem. 32:277-352.
Deshpande, V., A. Lachke, C. Mishra, S. Keskar, and M. Rao.
1986. Mode of action and properties of xylanase and B-
xylosidase from Neurospora crassa. Biotechnol. Bioengin.
28:1832-1837.
Devereux, J., P. Haeberli, and 0. Smithies. 1984. A
comprehensive set of sequence analysis programs for the VAX.
Nucleic Acids Res. 12:387-395.


138
Finally, a biochemical characterization of the purified
enzyme was initiated to examine whether or not a single
active center was responsible for both enzymatic activities.
The partially purified enzyme exhibited thermal inactivation
and temperature optimum profiles that were essentially the
same for both activities (Figs. 19 and 20).
The pH optima for both activities were essentially
the same however the activity range with respect to pH
differed for the two activities (Fig. 21). Several
hypotheses may be invoked to explain this difference in pH.
Hydrolytic mechanisms usually involve the contributions of a
proton donor and a proton acceptor to the catalysis (Knowles
et al. 1988). The proton donor acts as a general acid in
donating a proton to the glycosidic bond oxygen generating
an oxycarbonium ion intermediate. The proton acceptor
serves to stabilize the hydrolytic intermediate and acts as
a general base. At pH 6.0, the pH optimum for this enzyme,
carboxyl groups on aspartic and glutamic acid residues are
fully charged. At this pH a histidinyl residue (pKa 6.5)
may serve as a proton donor in that it is still protonated.
At pH values above 6.0 the xylosidase activity diminishes
reflecting the deprotonation of histidinyl residues in the
active center demonstrating the dependence of xylosidase
activity on this amino acid.
The relatively high arabinofuranosidase activity at
elevated pH may indicate that this catalytic activity


2521
2581
2641
2701
2761
2821
2881
2941
3001
3061
3121
3181
3241
3301
3361
3421
3481
3541
3601
3661
3721
3781
3;
2580
2640
2700
2760
2820
2880
2940
3000
3060
3120
3180
3240
3300
3360
3420
3480
3540
3600
3660
3720
3780
3840
CACCAGCAAAAGATAATGAATGTGCAGGAATGGTGTTATTCCAGAATAATGAAAATCACT
PAKDNECAGMVLFQNNENHL
TGGAGCTTTTAGTTGTAAAGAAGAAAGATAAGCTACAGTTTAAAGTAGGACCAGTTATTA
ELLVVKKKDKLQFKVGPVIK
AAGGAACCAAAATCAGACTTGCTACTTTTGATATTTCATCAGGTGATTTAGAAATTATTC
GTKIRLATFDI SSGDLE I IL
TTGAGGCAGCAAATCAGCTGGCTAATATCTATATTAAAAAGAATAATGAAAAGATTCTTG
EAANQLANIYIKKNNEKILV
TGGCAGAATGTATTGATTTGAGCCCATACACTACCGAAGAATCAGGCGGATTCGTAGGAT
AECIDLSPYTTEESGGFVGC
GTACCATTGGACTATATGCTTCTTCAAATGGAAAGACCAGTGATAACTATTGCGATTATT
TIGLYASSNGKTSDNYCDYS
CCTACTTTACAGTAGAAGAAGTATAGCATTTTCAATGAGCGAATTTGCAAGTTTTATATA
YFTVEEV*
CGGGATTAATTGTACGTAAAAACCATACAGGTGTAAAATAGTTTCCAGAGAAAGTTTTTT
CTCTGGAATTTTTTATTATQOAQOGGATTATGCTTCAGGAAAGTATTAAGAAGTTGGTAC
MLQESJKKLVQ
AGTACGGTATTGATATGGGGCTTACACCAGAATGTGAGAGAATATATACTACAAATCTTT
YGIDMGLTPECERIYTTNLL
TGCTTGAATGTATGAAAGAAGATGAGTACATAGATCCAGACTGTGATTTAAGCAATATTA
LECMKEDEYIDPDCDLSNI I
TACTTGAAGATGTATTAAAGGAACTTTTAGATGAGGCAGTTAATAGAGGTATCATAGAGG
LEDVLKELLDEAVNRGI IED
ATTCAGTTACACATAGGGATTTGTTTGATACAAAGCTAATGAATCAGCTATGCCCACGTC
SVTHRDLFDTKLMNQLCPRP
CTAAACAGGTTATAGATGATTTTAACCGTATATACGATAACCATGGTCCAATAGCTGCAA
KQVIDDFNRIYDNHGPIAAT
CAGATTATTTTTACAAGTTAAGCAAAGCCTCTGACTATATCCGTACTTACAGGGTAAAAA
DYFYKLSKASDYIRTYRVKK
AGGACCTAAAATGGACATGCGATACAGAGTATGGCACTCTTGACATAACAATTAATCTCT
DLKWTCDTEYGTLDITINLS
CTAAGCCAGAAAAAGACCCAAAGGCAATTGCTGCAGCTAAGAATGCAAAACAATCCACAT
KPEKDPKAIAAAKNAKQSTY
ATCCGAAGTGCCAATTATGTATGGAAAATGAAGGCTATGCTGGTCGCATTAATCATCCTG
PKCQLCMENEGYAGRINHPA
CTAGAGAGAATCATCGCATAATTCCTATAACTATAAATAACAGCAACTGGGGATTTCAAT
RENHRI IPITINNSNWGFQY
ATAGCCCATACGTTTATTACAATGAGCATTGCATAGTCTTTAACGGAGAGCATACTCCTA
SPYVYYNEHCIVFNGEHTPM
TGAAAAXAGAGCGAGCTACTTTTGTTAAGCTATTTGATTTTATCAAACTATTTCCACACT
KIERATFVKLFDFIKLFPHY
ATTTTTTAGGAAGCAATGCTGATTTACCAATTGTTGGAGGATCTATTTTAAGCCATGACC
FLGSNADLPIVGGSILSHDH


PLOT 1001
EH AX DSP H ERV H E D S P
29
P
-><*-
> <
-> *-
B
1000
BASE PAIRS
2000 3000
4000
ORF1
ORF2
ORF3
pLOI 1005
xv1 araf
+ +
PLOI1043
pLOI1040
+ +
Notl linker insertion


103
o
<
Q>
O
o>
o:
o>
o
o>
Q-
0
30 40
Temperature ( C)
10
20
50
60
70


20
Nucleotide sequence accession number. The nucleotide
sequence reported here has been assigned GenBank accession
number M55537.
Results and discussion
Identification of the xvlosidase coding region. Many
restriction endonuclease sites on the original xylosidase-
positive subclone, pLOI1005 (3.2 kb B. fibrisolvens insert),
were mapped to facilitate the generation of futher subclones
for analysis (Fig. 1). Plasmid DNA was purified using
cesium chloride and digested with a battery of restriction
enzymes. Restriction endonuclease sites were used to
generate subclones in both directions with respect to the
lac promoter in pUC18. Each subclone was examined for
enzyme activity on 4-methylumbelliferyl-B-D-xylopyranoside
(MUX) indicator plates. Based upon the results of these
experiments (Fig. 2), the region encompassing the xylosidase
gene was localized to a 2.1 kb DNA segment that spans an
internal EcoRI site and the internal PstI site. The
predicted gene size was in excess of the 1.4 kb Dral
fragment. In addition to the indicated xylosidase a
ctivity, arabinofuranosidase activity was also associated
with all xylosidase-positive subclones. It seems possible
that the gene, classified as xvlB. encodes an enzyme that
has activity against both substrates.
The number of chromosomal copies of xvlB was examined
using Southern hybridization. The large internal PstI and


o n mi h m y as
ZL


Figure 7. The assignment of domains to the xylB gene (solid
bar) Restriction endonuclease fragments used to localize
mutations are shown below. Abbreviations: E; EcoRI. S; SspI.
P; Pstl. H; Hindlll.


Figure 28. pH activity profiles of xylosidase and
arabinofuranosidase activies of the L178F mutant
protein: open squares; arabinofuranosidase
activity, closed squares; xylosidase activity.


CHAPTER III
CLONING, SEQUENCING, AND SEQUENCE ANALYSIS OF THE
XYLOSIDASE GENE FROM BUTYRIVIBRIO FIBRISOLVENS
Introduction
The synthesis of enzymes needed for xylan
depolymerization has been found to be constitutive in many
ruminal isolates of B. fibrisolvens (Hespell et al. 1987).
Recently, anaerobic digester isolates of B. fibrisolvens
have been described in which the synthesis of xylanase and
xylosidase were coordinately repressed by glucose and
induced by xylans and xylose (Sewell et al. 1988). The gene
for B-D-xylosidase from B. fibrisolvens GS113, in which this
enzyme is inducible, has been cloned on a multicopy plasmid
pUC18 in Escherichia coli (Sewell et al. 1989). Subcloning
analysis localized the coding region to a 5.8 kilobase pairs
(kbp) segment of cloned B. fibrisolvens DNA. The enzyme was
found to be predominantly intracellular in B. fibrisolvens
with 25% of the activity associated with the cell membrane
fraction. The cloned xylosidase is primarily cytoplasmic
with less than 2% of the active protein being membrane
associated in E. coli.
This investigation has been extended by restriction
endonuclease mapping the B. fibrisolvens DNA insert in pUC18
16


17
and to further define the coding region of the 6-D-
xylosidase gene in the insert. The number of chromosomal
copies of this gene was determined by Southern
hybridization. Additional subclones and primers were
generated to allow complete DNA sequencing of both strands.
Finally, the DNA sequence was compared with other, related
gene sequences.
Materials and Methods
Medium and growth conditions. Escherichia coli DH5a
was propagated at 37C in Luria broth or on Luria agar
supplemented with 50 mg of ampicillin per liter (Maniatis et
al., 1982).
Genetic methods. Plasmid pUC18 was used as a cloning
vector in all cloning and sequencing experiments unless
otherwise noted. The plasmids pLOHOOl and pLOHOOS harbor
the xylosidase coding region (Sewell et al.. 1989).
Analysis of restriction sites, plasmid purification,
subcloning, DNA ligation, Southern hybridization and other
DNA manipulations were performed using standard methods
(Maniatis et al.. 1982). Restriction enzymes (Bethesda
Research Laboratories, Gathersburg, MD) were used according
to the manufacturer's instructions. Transformed colonies
were screened for xylosidase and arabinofuranosidase
activity on Luria agar plates containing 20 /g/ml of the
flurorogenic substrates 4-methylumbelliferyl-B-D-
xylopyranoside or 4-methylumbelliferyl-a-L-arabinofuranoside


65
SP
pUC18
LacZ^=-
kilobase pairs
0.5 1.0 1.5
pLOI1043
H H
Sail
PstI
SphI
deletion
taa
xvl/araf
+/+ (undeleted)
-/- (27 bp./9 aa)
(56 bp/18 aa)


96
pooled and concentrated by ultrafiltration. A summary of
this purification scheme is shown in Table 9.
SDS-PAGE analysis of the pooled fractions from all the
purification steps showed the xylosidase to be approximately
90% pure with low levels of contaminating proteins (Fig.
17) .
Active enzyme conformation. Native-PAGE analysis of
the enzyme using Coomassie blue, activity stains, and
Western hybridization to visualize the protein indicated
only one predominant form of the xylosidase was active (Fig.
18). This active band had an apparent molecular weight of
120,000 which corresponds to the dimeric form of the enzyme.
Enzyme optima. The temperature activity profiles for
both activities were also essentially the same with 45C
being the optimum temperature for both the xylosidase and
arabinofuranosidase activities (Fig. 19). The thermal
inactivation profiles for both activities in the wild type
protein were essentially the same with activity diminishing
rapidly after incubation for 30 min at 35C (Fig. 20) .
The pH profiles for both activities in the wild type
protein showed marked differences (Fig. 21). The xylosidase
activity had a sharp peak at pH 6.0 which was followed by a
rapid decline in activity with less than 50% relative
activity remaining at pH 6.8. Arabinofuranosidase activity
peaked at pH 6.0 but remained at 90% relative activity up to
pH 9.0.


Figure 4. Outline of sequencing strategy of pLOHOOl and
subclone analysis of xvlB. (A) Sequencing strategy of the
complete 4.2 kb B. fibrisolvens insert in pLOHOOl. Arrows
indicate the direction of sequencing. Subclones were
sequenced using universal pUC18 primers (vertical bars in
front of arrows) and additional oligonucleotide primers
(vertical bars absent) were both used. (B) Outline of the
three ORF's sequenced and selected subclones and insertional
inactivation used to identify xvlB. Enzyme activity was
evaluated using MUG indicator plates. Double vertical bar in
pLOI1040 indicates the site of insertion of a 10 bp Notl
linker. Abbreviations: E; EcoRI. H; Hindlll. A; AccI. X;
Xbal. D; Dral, S; SspI. P; PstI. ERV; EcoRV. xyl; xylosidase
activity, araf; arabinofuranosidase activity. A "+" or
denotes the presence or absence of enzyme activity,
respectively.


15
also been proposed that the environment in which an enzyme
evolves also contributes to the enzyme specificity (Robson
and Gardier, 1988). The rumen, an environment with
specialized substrate-hydrolyzing requirements, may exert
selective pressures resulting in the evolution of organisms
and enzyme systems that reflect the heterogeneous nature of
available substrates.


CHAPTER VI
PARTIAL PURIFICATION AND CHARACTERIZATION OF
THE L178F MUTANT PROTEIN
Introduction
As outlined in chapter IV, point mutations were
introduced into the xvlB gene using in vitro mutagenesis.
All the mutations resulted in proteins having a reduction or
loss of both enzymatic activities concurrently.
Additionally, 10 of these point mutations were clustered in
a 60 amino acid region of the protein. Native agarose-xylan
gel electrophoresis of the expressed wild type and mutant
proteins revealed that the mutant proteins bind to the
surrogate substrate xylan with less affinity relative to the
native protein indicating an apparent increase in the Km
values for these mutant enzymes. The L178F mutant appeared
to have the least affinity (highest apparent Km) for the
surrogate substrate.
To further investigate the possibility of altered Km
due to the L178F mutantion, the mutant protein was partially
purified and characterized.
120


87
are structurally related and have quite similar molecular
configurations about the B-(l,4) bonds with respect to the
hydroxyl group on the a-carbon being in the axial or
equatorial positions.
Some true bifunctional cellulases have been
demonstrated which contain separate active sites for each
enzymatic activity. Saul et al. (1990) isolated a cellulase
from Caldocellum saccharolvticum which exhibited both endo-
glucanase and exo-glucanase activities. These authors used
DNA sequence homology comparisons and deletion analysis to
demonstrate that the endoglucanase activity was located in
the carboxy terminal domain and the exoglucanase activity
was located at the amino terminal domain.
In an earlier study purified a B-xylosidase from
Chaetomium trilaterale that also exhibited S-glucosidase
activity (Uziie et al. 1985). These investigators used
kinetic analysis employing substrate competition and
inhibitors to demonstrate that a single active site was
responsible for both enzymatic activities. It was also
suggested that two kinetically separate substrate binding
sites may reside in the active center of this enzyme.
In the previous chapter mutational analysis
demonstrated that the two enzymatic activities encoded by
xylB were not functionally separate but both appeared to be
catalytically dependent upon the same region of the protein.
The proposal that a single active center is responsible for


38
TABLE 1. (continued
Frequency (mol %> Codon Usage
Amino Acid
Codon
B.
ORF 1
fibrisolvens E.
ORF 2 ORF 1
coli
B. fibrisolvens
xvlA
Lys
AAA
3.4
3.1
3.6
4.1
3.4
AAG
4.7
3.5
3.3
1.3
3.9
Asp
GAT
5.8
5.6
5.9
2.5
3.9
GAC
1.6
1.4
2.0
3.0
0.2
Glu
GAA
4.3
3.7
3.7
4.9
1.7
GAG
4.9
3.5
3.8
1.8
2.7
Cys
TGT
1.4
1.4
1.0
0.4
0.7
TGC
0.7
1.0
1.5
0.5
0.5
Trp
TGG
1.3
1.7
1.3
0.7
2.0
Arg
CGT
0.7
0.8
1.3
3.1
1.0
CGC
-0-
0.2
0.5
2.0
0.2
CGA
0.2
0.2
0.2
0.2
-0-
CGG
-0-
-0-
-0-
0.2
-0-
AGA
3.1
2.9
2.0
0.1
2.0
AGG
0.5
-0-
0.8
0.1
0.2
Gly
GGT
3.4
1.7
1.5
3.8
2.4
GGC
2.3
2.1
1.0
3.1
1.5
GGA
3.1
4.4
1.5
0.4
2.7
GGG
0.5
0.2
0.8
0.6
0.5


34
3900
3841
3901
3961
4021
4081
4141
ATTTCCAAGGCGGCCATTACACATTTGCCATGGAAAAAGCTCCAATTATTCAGGAATTTA
FQGGHYTFAMEKAP I IQEFT
CTGTAAAAGGATATGAGGATGTTAAGGCTGGTATAGTTAAATGGCCACTTTCAGTAATTA
VKGYEDVKAGIVKWPLSVIR
3960
GACTTCAGTGCAAGGATGAGACTAGACTTATTGATTTAGCGACTAATATATTAGACAAAT
LQCKDETRLIDLATNILDKW
4020
GGAGAAATTACACCGATGAAGAGGCATATATTTTTGCTGAAACAGATGGTGAGCCTCACA
RNYTDEEAYIFAETDGEPHN
4080
ATACGATTACACCTATTGCTAGAAAAAGAGGGGATTACTTTGAACTAGATCCTCTAGAGT
TITPIARKRGDYFELDPLES
4140
CGACCTGCAGGCATGCAAGCTTGGCACTGGCCGTCGTTTTACAACGTCGTGACTGGG* 4198
TCRHASLALAVVLQRRDW


97
TABLE 9. Purification of the B. fibrisolvens xvlB encoded
xylosidase from E. coli DH5a (pLOI1005).
Fraction
Vol
(ml)
Protein
(mg)
Activity
(mU)a
Sp Act
(mU/mg)
Yield
(%)
Crude
extract
2.0
100
4600
46
100
Prep
Cell
7.0
3.4
1068
314
23
Hydrophobic
column
3.0
0.9
587
652
13
3 Specific activity expressed as nmoles p-nitrophenol
released per min per mg protein for B-D-xylosidase
activity only.


105
0 51015 20 25 30 35 40 45 50 55 60 65 70
Temperature ( C)


xylanolytic enzymes act to increase the availability of
cellulose to cellulases of the cellulosome through removal
of associated xylan chains. In the rumen, cooperativity
between xylanase and cellulase degrading enzymes is also
apparent. In Bacteroides succinoaenes isolated from rumen
fluid it was demonstrated that carboxymethylcellulase (B-
1,4-endo-glucanase), B-xylanase, and 6-D-xylosidase were
expressed by the organism when grown on media containing
cellulose as a sole source of carbohydrate (Forsberg et al.
1981). These investigators postulated that cooperativity
between the cellulose and hemicellulose degrading enzymes
helps to enhance polymer breakdown and increase substrate
availability for rumen microorganisms which lack these
enzymes. This cooperativity among different organisms may
serve to maintain a stable microbial population in the
rumen.
Enzymatic cooperativity and synergism is also present
within the hemicellulose-degrading systems. A synergistic
action of 6-xylanase and fl-D-xylosidase has been
demonstrated in cultures of Neurospora crassa when grown on
xylan (Deshpande et al. 1986). In this study the degree of
hydrolysis of D-xylan by xylanase was increased 30% by the
addition of B-D-xylosidase to a cell-free system.
Another example of enzymatic synergism involves the
enzyme acetyl esterase. Acetyl esterase (EC 3.1.1.6) is
active against esters of acetic acid and are widely


77


134
Based upon the data in this chapter and in chapter IV
it is reasonable to conclude that the L178F mutation is in
an area of the xvlB protein that is important in substrate
binding and possibly catalysis with respect to both
enzymatic activities.


6
energy requirements are derived from volatile fatty acids
produced in the human colon (MacNeil, 1984). Xylanolytic
enzymes have also been isolated and characterized from
several fungi including Trichoderma reesei (Poutanen and
Puls, 1988), Aspergillus niaer (Fukumoto et al. 1970), and
Fusarium roseum (Gascoigne and Gascoigne, 1980).
The rumen is the primary organ in which cattle, sheep,
and other ruminants derive their energy and nutrition
through breakdown of complex carbohydrates. Starch,
cellulose and hemicellulose are degraded by enzymes that are
secreted by resident microorganisms and metabolized to
volatile fatty acids as the end products of fermentation
(Hobson and Wallace, 1982). In general, cellulose and
hemicellulose depolymerization by rumen microbial flora
releases free monosaccharides and short chain
oligosaccharides. The predominant metabolic waste products,
volatile fatty acids, are released and either absorbed and
utilized by the animal or used by other microorganisms in
the rumen and other digestive organs (omasum, abomasum and
the small intestine).
The rumen microbial community represents a diverse
group of organisms, many of which have the ability to
degrade hemicellulose (Dehority, 1966). The rumen ecosystem
differs from other microbial ecosystems in substrate
availability and product accumulation. The rumen is close
to an industrial fermentation in that substrate availability


GENETIC AND BIOCHEMICAL CHARACTERIZATION OF
THE B-D-XYLOSIDASE GENE FROM BUTYRIVIBRIO FIBRISOLVENS
By
ERIC ANDREW UTT
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
1991

ACKNOWLEDGEMENTS
I owe my development as a scientist as well as the
successful completion of this dissertation to my major
professor, Dr. Neal Ingram. I will be forever indebted to
him for his uninhibited sharing of knowledge and expertise.
I wish also to express my gratitude to the members of my
graduate committee, Dr. Allen, Dr. Aldrich, Dr. Shanmugam,
and Dr. Gander. Their contributions to my research and in
the preparation and review of this manuscript are greatly
appreciated. I must also thank my friends and comrades Jeff
Mejia and David Beall for the friendship and helpful
suggestions during the course of my doctoral work. To all
the former postdocs, Dr. Christina Eddy, Dr. Terryl Conway,
and Dr. Guy Sewell, from whom I learned much, I wish to
express my thanks. Thanks are due to my parents for their
love and support during my graduate education. Similarly, I
wish to thank my wife's parents for their love and support.
And lastly, I would like to thank my loving wife, Lisa, and
my beautiful girls, Tina and Hannah, for making my life
special. It is to them that I dedicate this dissertation.
ii

TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS
LIST OF TABLES V
LIST OF FIGURES vi
ABSTRACT ix
CHAPTERS
IGENERAL INTRODUCTION 1
IIREVIEW OF THE LITERATURE 5
IIICLONING, SEQUENCING, AND SEQUENCE
ANALYSIS OF THE XYLOSIDASE GENE FROM
BUTYRIVIBRIO FIBROSOLVENS 16
Introduction 16
Materials and Methods 17
Results and Discussion 20
Conclusions 50
IVMUTATIONAL ANALYSIS OF THE XVIB GENE 54
Introduction 54
Materials and Methods 56
Results and Discussion 63
Conclusions 84
VPARTIAL PURIFICATION AND CHARACTERIZATION
OF THE WILD TYPE GENE PRODUCT 86
Introduction 86
Materials and Methods 88
Results and Discussion 91
Conclusions 115
iii

CHAPTERS
Page
VI PARTIAL PURIFICATION AND CHARACTERIZATION OF
THE L178F MUTATION 120
Introduction 120
Materials and Methods 121
Results and Discussion 121
Conclusions 128
VII SUMMARY AND GENERAL CONCLUSIONS 13 5
LITERATURE CITED 142
BIOGRAPHICAL SKETCH 151
iv

LIST OF TABLES
Table Page
1.Comparison of codon usage frequency for
the three B. fibrisolvens ORF's 36
2.Expression of enzyme activities in
recombinant E. coli harboring xvlB 40
3. Hydrolysis of different nitrophenyl-
substituted glycosides by the xylB gene
product 42
4. Comparison of the translated amino acid
sequences of the three B. fibrisolvens ORF's
in pLOHOOl with those of selected
proteins 46
5. Amino acid sequence alignment of conserved
regions 48
6. Localization of point mutations by
restriction fragment replacement analysis 67
7. Enzymatic activities of recombinants
harboring point mutations in xvlB
relative to the wild type protein 82
8. Xylosidase activity of in vitro mutations with
varying substrate concentrations 83
9. Purification scheme of the xvlB-encoded protein
from E. coli DH5a 97
v

LIST OF FIGURES
Figure Page
1.Restriction maps of pUC18 derived plasmids
that express fi-D-xylosidase activity in
E. coli DH5a 22
2.Subclone analysis of pLOI1005 to localize
the xylB coding region 24
3.Southern hybridizations of chromosomal
DNA from B. fibrisolvens and E. coli 27
4.Outline of sequencing strategy of pLOHOOl
and subclone analysis of xvlB 29
5. The complete nucleotide sequence and
translated amino acid sequence of the
4.2 kb insert from pLOHOOl 31
6. SDS-PAGE analysis of cytoplasmic extracts
from recombinant E. coli DH5a harboring
selected constructs 4 5
7. Assignment of domains to the xvlB gene 59
8.Subcloning strategy used to localize in vitro
mutations to one of five domains in the
xvlB gene 61
9.Deletion analysis of the xvlB gene 65
10.Localization and identification of point
mutations in xvlB by DNA sequencing 69
11. SDS-PAGE analysis of wild type and mutant
proteins 73
12. Native-PAGE comparison of W158UGA and L178F
mutations with the wild type stained with
Coomassie blue 75
13. Western hybridization of native-PAGE of wild
type and mutant proteins 77
vi

Figure Page
14. Substrate binding native gel western
hybridization assays of wild type and mutant
proteins 80
15. Elution profile of the xvlB gene product
during preparative electrophoresis on the
BioRad-Prep Cell system 93
16. Elution profile of the xvlB gene product
during hydrophobic interaction chromatography.. 95
17. SDS-PAGE analysis of pooled xylosidase-
containing fractions from preparative
electrophoresis on 8% native-PAGE 99
18 Native-PAGE analysis of the B-D-xylosidase 101
19. Thermal inactivation profile of xylosidase and
arabinofuranosidase activities 103
20. Temperature optimum profile for xylosidase and
arabinofuranosidase activities 105
21. pH activity profiles for xylosidase and
arabinofuranosidase activities 107
22. Double recipricol plots of xylosidase and
arabinofuranosidase activities of the native
protein 110
23. Competitive inhibition of xylosidase activity
by 4-MU-a-L-arabinofuranoside 112
24. Competitive inhibition of arabinofuranosidase
activity by B-methyl-D-xyloside 114
25. Elution profile of the xvlB-encoded protein
harboring the L178F mutation during
preparative electrophoresis 12 3
26. SDS-PAGE analysis of partially purified
L178F mutant protein 125
vii

Figure
Page
27. Thermal inactivation profiles of xylosidase
and arabinofuranosidase activities of the
L178F mutant protein 127
28. pH activity profiles for xylosidase and
arabinofuranosidase activities of the
L178F mutant protein 130
29. Double recipricol plots of xylosidase and
arabinofuranosidase activities for the
L178F mutant protein 132
viii

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
GENETIC AND BIOCHEMICAL CHARACTERIZATION OF B-D-XYLOSIDASE
FROM BUTYRIVIBRIO FIBRISOLVENS
By
Eric Andrew Utt
December 1991
Chairman: Lonnie 0. Ingram
Major Department: Microbiology and Cell Science
The gene for B-D-xylosidase from the rumen bacterium
Butyrivibrio fibrisolvens encodes a protein that exhibits
hydrolytic activity against B-D-xylopyranosides and a-L-
arabinofuranosides. This gene, xvlB. was cloned into E.
coli as a 4.2 kilobase pairs (kbp) insert in pUC18 and
sequenced in both directions. The xvlB gene is present as a
single copy on the B. fibrisolvens chromosome and consists
of a 1,551 base pair (bp) open reading frame (ORF) which
encodes a protein of 517 amino acids. Insertion of a 10 bp
linker into the coding region resulted in a frameshift that
abolished both activities. Deletions from the 3' end and
the 5' end of xvlB also resulted in inactive proteins. SDS-
PAGE analysis of cytoplasmic extracts from recombinant E.
coli clones harboring xvlB confirmed the presence of a new
protein with an apparent molecular weight of 60,000.

IX

Although the xvlB gene did not exhibit a high degree of
amino acid identity with other xylan-degrading enzymes or
glycohydrolases, a conserved sequence was identified with
significant identity to the active site region of hen egg
white lysozyme and Aspergillus niaer glucoamylase. No
predictable stem loop structures or sequences resembling
terminators were found on the xvlB gene fragment and this
gene appears to be part of an operon. In vitro analysis of
xvlB mutants demonstrated structural and functional
relationships between the two enzyme activities. All point
mutations investigated in xylB resulted in the reduction or
loss of both enzymatic activities. Most of these mutations
were clustered in a region near the proposed active site.
The point mutations decreased the apparent affinity of the
enzyme for xylan. The partially purified xylB-encoded
protein exhibited thermal inactivation kinetics and
temperature optima that were essentially the same for both
enzymatic activities. The pH optimum for both activities
was 6.0. However, the arabinofuranosidase activity
exhibited a broader pH range, retaining 90% of maximal
activity up to pH 9.0. The apparent Km for p-nitrophenyl-6-
D-xylopyranoside and p-nitrophenyl-a-L-arabinofuranoside
were 3.7 mM and 1.8 mM respectively. Substrate competition
experiments corroborated the genetic evidence and
demonstrated that the same active center was responsible for
both enzymatic activities of the xvlB-encoded protein.
x

CHAPTER I
GENERAL INTRODUCTION
Plant cell walls represent the largest reserve of fixed
carbon on earth. Plant cell walls are composed primarily of
cellulose, hemicellulose, and lignin (Weinstein and
Albersheim, 1979). Cellulose is the most abundant
carbohydrate found in plant biomass (Coughlan, 1985) while
hemicellulose is a major plant structural polymer that ranks
second only to cellulose in natural abundance (Dekker and
Richards, 1976). The amount of hemicellulose in dry wood is
between 20% and 30% (Eriksson et al. 1990). The composition
of hemicellulose varies between softwoods and hardwoods.
The major hemicellulose in softwoods is galactoglucomannan
(Eriksson et al. 1990). This polymer has a backbone
composed of a linear chain of 1,4-linked fi-D-glucopyranose
and B-D-mannopyranose units. The mannose and glucose
moieties of the backbone may be substituted with acetyl
groups at the C-l and C-2 positions.
Glucouronoxylan (O-acetyl-4-O-methyl-glucurono--D-
xylan) is the major hemicellulosic component of hardwoods
and agricultural residues (Eriksson et al. 1990). The major
structural feature of xylan is a linear chain consisting of
1

2
chain is often substituted with acetyl, arabinofuranosyl,
ferulyl glucopyranosyl, and mannopyranosyl side chains to
form a complex heterogeneous structure.
Several genera of bacteria and fungi are able to
partially or completely depolymerize xylan in various
habitats (Biely, 1985) Xylan depolymerization by
microorganisms is a multistep process which involves the
concerted activities of several different enzymes.
Xylaases (1,4-B-D-xylan xylanohydrolase; EC 3.2.1.8) are
extracellular enzymes which hydrolyze the internal B-1,4-
xylosidic linkages on the main chain. The resulting smaller
oligosaccharides are transported into the microbial cells
where xylosidases (1,4-B-D-xylan xylohydrolase; EC 3.2.1.37)
continue the hydrolysis and release monosaccharides for
glycolysis (Dekker and Richards, 1976). The hydrolysis and
removal of side chain substituents requires additional
enzyme activities including arabinofuranosidase, which
removes substituted arabinofuranosyl residues from the xylan
backbone. (Biely, 1985). This may be particularly important
since arabinose substituents on the xylan chain have been
shown to limit the complete enzymatic breakdown of xylan
(Chesson et al. 1983). One of the organisms which is
particularly adept at xylan depolymerization in Butvrivibrio
fibrisolvens. B. fibrisolvens is a Gram variable,

3
obligately anaerobic bacillus that is frequently found in
the rumen and anaerobic digesters (Dehority, 1966). This
organism produces a cadre enzymes which enable it to degrade
plant biomass, including cellulose and hemicellulose
(Hespell, 1987). The genus Butyrivibrio contains only a
single species but consists of many strains that vary in DNA
relatedness between 20% to 100% (Mannarelli, 1988). This
organism is also characterized as having a low (38% to 42%)
guanine plus cytosine (mole percent) content. Butyrivibrio
produces an extracellular polysaccharide (EPS) that contains
an unusual 4-0-(1-carboxymethyl)-rhamnose sugar (Mannarelli
et al. 1990). These investigators suggested that the
unusual sugars found in the EPS of B. fibrisolvens serve to
protect the organism from glycosidases and other enzymes
found in the digestive tract of the host animal.
B. fibrisolvens GS113, an anaerobic digester isolate
used in these studies, was shown to produce high levels of
both xylanase and xylosidase (Sewell et al. 1988) These
two enzymes were shown to be repressed by glucose and
induced by xylan and xylose.
In a previous study, the xvlB gene encoding the B-D-
xylosidase from B. fibrisolvens GS113 was isolated from a
plasmid pUC18 genomic library (Sewell et al. 1989).
During the course of the current investigations, an a-
L-arabinofuranosidase activity was detected in all clones
harboring the xvlB gene. This characteristic became the

4
principle focus of this research. The principle question
that was addressed relates to the dual activity exhibited by
this enzyme against B-D-xylopyranosides and a-L-
arabinofuranosides: do these two activities reside in the
same active center of the enzyme or are they on separate
domains?
The following research examined the structure and
function of the xvlB gene and gene product and includes:
A) The complete nucleotide sequence of xvlB and
sequence comparisons with related enzymes from other
organisms.
B) Genetic evidence that the two enzymatic activities
are encoded by a single open reading frame.
C) Mutational analyses to investigate the genetic
interdependence of the two enzymatic activities of
xvlB.
D) Purification and characterization of the xvlB-
encoded protein.
E) Biochemical and kinetic experiments to investigate
the functional relationship between the two enzymatic
activities.

CHAPTER II
REVIEW OF THE LITERATURE
Many different bacteria have been characterized which
are able to depolymerize xylan including: Bacteroides
succinoaenes (Forsberg et al. 1981), Clostridium
acetobutvlicum (Lee et al. 1987), Bacillus pumilus
(Panbangred et al. 1983), Bacillus subtilis (Paice et al.
1986), Butvrivibrio fibrisolvens (Hespell et al. 1987) ,
Caldocellum saccharolvticum (Luthi et al. 1990), and
Clostridium thermocellum (Garcia-Martinez et al. 1980) .
These microorganisms which degrade xylan have been isolated
from numerous environments. Varel (1987) demonstrated that
pig large intestine contained xylanolytic Bacteroides
succinoaenes and Ruminococcus flavefaciens. Both of these
organisms are also present in large numbers in the bovine
rumen. Other studies demonstrated that up to 30% of the
metabolic energy requirements of the pig can be met via the
utilization of volatile fatty acids, products of microbial
cellulose and hemicellulose digestion (Rerat et al. 1987).
In addition, Salyers et al. (1981) isolated two species of
human colonic Bacteroides that were able to utilize xylan as
a carbon and energy source, thus producing volatile fatty
acid products. However, only 5 to 10 % of the maintenance
5

6
energy requirements are derived from volatile fatty acids
produced in the human colon (MacNeil, 1984). Xylanolytic
enzymes have also been isolated and characterized from
several fungi including Trichoderma reesei (Poutanen and
Puls, 1988), Aspergillus niaer (Fukumoto et al. 1970), and
Fusarium roseum (Gascoigne and Gascoigne, 1980).
The rumen is the primary organ in which cattle, sheep,
and other ruminants derive their energy and nutrition
through breakdown of complex carbohydrates. Starch,
cellulose and hemicellulose are degraded by enzymes that are
secreted by resident microorganisms and metabolized to
volatile fatty acids as the end products of fermentation
(Hobson and Wallace, 1982). In general, cellulose and
hemicellulose depolymerization by rumen microbial flora
releases free monosaccharides and short chain
oligosaccharides. The predominant metabolic waste products,
volatile fatty acids, are released and either absorbed and
utilized by the animal or used by other microorganisms in
the rumen and other digestive organs (omasum, abomasum and
the small intestine).
The rumen microbial community represents a diverse
group of organisms, many of which have the ability to
degrade hemicellulose (Dehority, 1966). The rumen ecosystem
differs from other microbial ecosystems in substrate
availability and product accumulation. The rumen is close
to an industrial fermentation in that substrate availability

7
is very high and constant while product accumulation is low.
It has been established that microbial cells in the rumen
are present in high numbers and contains: 1011 bacteria ml'1,
106 ciliate protozoa ml'1, and 104 fungi ml'1 (Patterson,
1989) .
Butvrivibrio fibrisolvens is a Gram variable,
obligately anaerobic, motile bacillus (Hespell and Bryant,
1981) B_;_ fibrisolvens is particularly abundant in the
rumen and anaerobic digesters in which plant material serves
as the primary substrate (Hespell et al. 1987). B.
fibrisolvens converts hemicellulose to mono and
oligosaccharides. These are transported and metabolized to
yield butyric acid. Mannarelli et al. (1990) cloned and
sequenced the gene encoding B-D-xylaase from B.
fibrisolvens strain 49. Sewell et al. (1988) isolated
several strains of B. fibrisolvens that produced both
xylanase and xylosidase. In this study, the synthesis of
both enzymes were concurrently repressed by glucose and
induced by xylan and xylose. This was surprising since
earlier work on rumen isolates of B. fibrisolvens had
reported that these enzymes were expressed constitutively
(Hespell et al. 1987). Similarly, it was reported that
xylose served as an inducer of the xylanase and fl-D-
xylosidase in Pullularia pullulans ( Pou-Llinas and Driguez,
1987). In addition, B-D-xylosidase of Bacillus pumi1 ns was
found to be induced by xylose (Kersters-Hilderson et al.

8
1969). Xylobiose was found to induce synthesis of 6-
xylanase in Crvptococcus albidus (Biely et al. 1980) and in
Streptomvces sp. (MacKenzie et al. 1987). Biely and
Petrakova (1984), studying the xylan-degrading system in C.
albidus. found that certain positional isomers of xylose and
xylobiose, notably 1,4-B-xylobiose, could serve as inducers
of B-xylanase and B-D-xylosidase.
Some organisms produce multiple xylanase enzymes.
Esteban et al. (1982) reported that Bacillus circulans WL-12
secretes two endo-6-xylanases and one B-D-xylosidase when
grown on xylan as a sole carbon source. Three distinct
xylanase genes have been identified and cloned from
Clostridium thermocellum (MacKenzie et al. 1989) A
multiplicity of xylanases has also been reported in fungi
including Aspergillus niaer (Frederick et al. 1985), and
Trichoderma harzianum ( Wong et al. 1986). It has been
suggested that the multiplicity serves to enhance the
ability of microbes to depolymerize a wide range of
substituted xylans under different environmental conditions.
In some organisms, the xylanolytic and cellulolytic
systems are combined. Recently Morag et al. (1990)
demonstrated that, in addition to free xylanases,
Clostridium thermocellum possessed a cellulosome-associated
xylanase which exhibits endo-glucanase activity. However
this organism was unable to utilize or grow on xylan. These
investigators postulated that cellulosome-associated

xylanolytic enzymes act to increase the availability of
cellulose to cellulases of the cellulosome through removal
of associated xylan chains. In the rumen, cooperativity
between xylanase and cellulase degrading enzymes is also
apparent. In Bacteroides succinoaenes isolated from rumen
fluid it was demonstrated that carboxymethylcellulase (B-
1,4-endo-glucanase), B-xylanase, and 6-D-xylosidase were
expressed by the organism when grown on media containing
cellulose as a sole source of carbohydrate (Forsberg et al.
1981). These investigators postulated that cooperativity
between the cellulose and hemicellulose degrading enzymes
helps to enhance polymer breakdown and increase substrate
availability for rumen microorganisms which lack these
enzymes. This cooperativity among different organisms may
serve to maintain a stable microbial population in the
rumen.
Enzymatic cooperativity and synergism is also present
within the hemicellulose-degrading systems. A synergistic
action of 6-xylanase and fl-D-xylosidase has been
demonstrated in cultures of Neurospora crassa when grown on
xylan (Deshpande et al. 1986). In this study the degree of
hydrolysis of D-xylan by xylanase was increased 30% by the
addition of B-D-xylosidase to a cell-free system.
Another example of enzymatic synergism involves the
enzyme acetyl esterase. Acetyl esterase (EC 3.1.1.6) is
active against esters of acetic acid and are widely

10
distributed in nature (Poutanen et al. 1991). The acetyl
residues on the xylan backbone are removed by acetyl
esterase (Biely et al. 1985). These enzymes have been found
to act cooperatively with xylanases. The acetyl esterase
serves to increase the rate of glycosidic bond cleavage by
B-xylanase from Trichoderma reesei (Biely et al. 1986).
These enzymes were also found to act synergistically to
liberate acetyl residues. More recently it was demonstrated
that the rate of liberation of acetic acid from acetyl-xylan
by acetyl esterase of T. reesei was increased by the
addition of endo-xylanase and B-D-xylosidase (Poutanen and
Sundberg, 1988).
Also involved in enzyme synergism is the enzyme a-L-
arabinofuranosidase (Greve et al. 1984). This enzyme was
purified from Ruminococcus albus 8 and had a pH optimum of
6.9 and a Km of 1.3 mM, both for p-nitrophenyl-a-L-
arabinofuranoside as a substrate. They showed that this
enzyme enhanced the rate of hydrolysis of alfalfa cell wall
hemicellulose when combined with other xylanolytic or
pectinolytic enzymes. It was hypothesized that this enzyme
functioned to provide rumen microbes with suitable
substrates for xylanase.
The mechanism of xylan hydrolysis by microbial
xylanases has been studied extensively. Xylanases are
usually small proteins having molecular weights ranging
between 20,000 to 50,000 (Bastawde et al. 1991). Most

11
xylanases are in fact endo-xylanases by virtue of the fact
that they attack the interior B-(1,4)-D-xylosidic linkages
of the xylan polymer rather than the exterior linkages (Ward
and Moo-Young, 1989). The fl-(1,4)-D-endo-xylanases have a
pH optimum in the range of 3.5 to 6.5 while the temperature
optima and thermal stabilities vary depending upon the
source (Ward and Moo-Young, 1989). Xylaase from Bacillus
pumilus IPO has a molecular weight of 22,000 and is a B-D-
1,4-endo-xylanase (Panbangred et al. 1983). The pH and
temperature optimum of this enzyme are 6.5 and 40C,
respectively. Quantification of the hydrolysis end products
from larchwood xylan indicated that the B. pumilus enzyme
had the greatest affinity for the second and sixth B-
xylosidic linkages of the polymer.
The xvnZ gene product from Clostridium thermocellum is
also an endo-xylanase with pH and temperature optima of 6.0
and 65C, respectively (Grepinet et al. 1988). Lee et al.
(1987) purified and characterized two different endo-
xylanases, xylanase A and xylanase B, from Clostridium
acetobutvlicum. Xylanase A has a molecular weight of
65,000, a pH optimum of 5.0, an optimum temperature of 50C,
and is stable for up to 30 min at 40C. Xylanase B is a
smaller protein having a molecular weight of 29,000. It had
a pH optimum of 5.0 to 6.0, showed a temperature optimum of
60C, and is stable for 30 min at 50C. Both enzymes
hydrolyze larchwood xylan randomly, however xylanase B

12
produced only xylotriose and xylobiose as products whereas
xylanase A also yields xylohexose, -pentose, and -treaose as
end products. Xylanase A was also active against
carboxymethylcellulose, acid-swollen cellulose and lichenin.
The two enzymes were antigenically different as judged by
"Ouchterlony"-immunodiffusion assays. The two enzymes were
therefore presumed to be encoded by separate genes.
Some xylaases, such as those that are produced by
fungi, notably Aspergillus niger, produce endo-1,4-6-D-
xylanases that can hydrolyze the 1,3-a-L-arabinofuranosyl
side chains from arabinoxylans (Dekker, 1985). These
enzymes have been termed "debranching" xylanases. Recently
a unique "appendage-dependent" xylanase was isolated and
purified from Bacillus subtilis (Nishitani and Nevins,
1991). This enzyme is classified as a B-(1,4)-xylan
xylanohydrolase and has a prerequisite for glucuronosyl
substituted side chains in order to initiate hydrolysis of
the xylan backbone structure. Three novel xylanases were
purified from B. subtilis which exhibited activity against
ferulylolated arabinoxylans (Nishitani and Nevins, 1988).
These enzymes acted on ferulic acid-substituted arabinoxylan
and liberated the terminal arabinofuranosyl, terminal
gluconopyranosyl, and ferulic acid moieties from the
polymer. While much recent work has concentrated on the
extracellular microbial xylanases, less is known concerning
the molecular biology and properties of the intracellular 6-

13
D-xylosidase component of the microbial xylanolytic system.
Early mechanistic studies of the B. pumilus xylosidase
indicated the enzyme contained several thiol groups and at
least one of which is involved in the catalysis (Saman et
al. 1975). Panbangred et al. (1983) first cloned the genes
for B-xylanase and B-D-xylosidase from Bacillus pumilus IPO.
Both cloned proteins were expressed in Escherichia coli from
a hybrid plasmid and were immunologically and chemically
identical to those of B. pumilus. The cloned genes from B.
pumilus IPO were later sequenced by Moriyama et al. (1987).
The gene for B-D-xylosidase was localized to a 1617 base
pairs open reading frame encoding a deduced 62,607d protein.
The N-terminal amino acid sequence agreed with that
predicted from the DNA sequence and that obtained from the
purified enzyme. It is interesting to note that the B-
xylanase gene from the same organism was located 4,600 base
pairs downstream from the 3'-end of the 6-D-xylosidase. The
B. pumilis enzyme was not reported to exhibit any additional
enzymatic activities.
More recently, two xylosidase genes were cloned and
sequenced from the obligately anaerobic, thermophilic
organism Caldocellum saccharolvticum (Luthi et al. 1990).
The protein encoded by one of these xylosidase genes was
found to possess xylanase activity in addition to the
expected xylosidase activity. These genes were also found
to reside in close proximity to each other and to a gene

14
encoding a protein having acetyl esterase activity.
Additional 6-D-xylosidase enzymes with multiple activities
have been reported earlier. Kinetic methods were used to
investigate the active site of a B-D-xylosidase from
Chaetomium trilaterale (Uziie et al. 1985). In this study,
which employed substrate analogues as inhibitors, a single
active center was postulated to function for both the B-D-
xylosidase and B-glucosidase activities exhibited by this
enzyme. A 6-D-xylosidase purified from Trichoderma reesei
was also found to exhibit an a-L-arabinofuranosidase
activity (Poutanen and Puls, 1988). Additionally, a cloned
gene cluster from Bacteroides ovatus was also found to
exhibit B-D-xylosidase and a-L-arabinofuranosidase
activities (Whitehead and Hespell, 1990) These dual
activities co-purified and were encoded by a single open
reading frame present in the cloned gene fragment. While
the dual activities of B-D-xylanases and B-D-xylosidases
have been documented, little is known about the genetic and
biochemical basis of this property. Multiple substrate
activities can be attributed to the presence of more than
one catalytic region on the enzyme. Another possibility is
the presence of a single catalytic region with wide
substrate specificity. It has been proposed that the more
evolved an enzyme or protein is, the more narrow its
specificity becomes (Knowles, 1988). Accordingly the more
primitive proteins tend to have multiple functions. It has

15
also been proposed that the environment in which an enzyme
evolves also contributes to the enzyme specificity (Robson
and Gardier, 1988). The rumen, an environment with
specialized substrate-hydrolyzing requirements, may exert
selective pressures resulting in the evolution of organisms
and enzyme systems that reflect the heterogeneous nature of
available substrates.

CHAPTER III
CLONING, SEQUENCING, AND SEQUENCE ANALYSIS OF THE
XYLOSIDASE GENE FROM BUTYRIVIBRIO FIBRISOLVENS
Introduction
The synthesis of enzymes needed for xylan
depolymerization has been found to be constitutive in many
ruminal isolates of B. fibrisolvens (Hespell et al. 1987).
Recently, anaerobic digester isolates of B. fibrisolvens
have been described in which the synthesis of xylanase and
xylosidase were coordinately repressed by glucose and
induced by xylans and xylose (Sewell et al. 1988). The gene
for B-D-xylosidase from B. fibrisolvens GS113, in which this
enzyme is inducible, has been cloned on a multicopy plasmid
pUC18 in Escherichia coli (Sewell et al. 1989). Subcloning
analysis localized the coding region to a 5.8 kilobase pairs
(kbp) segment of cloned B. fibrisolvens DNA. The enzyme was
found to be predominantly intracellular in B. fibrisolvens
with 25% of the activity associated with the cell membrane
fraction. The cloned xylosidase is primarily cytoplasmic
with less than 2% of the active protein being membrane
associated in E. coli.
This investigation has been extended by restriction
endonuclease mapping the B. fibrisolvens DNA insert in pUC18
16

17
and to further define the coding region of the 6-D-
xylosidase gene in the insert. The number of chromosomal
copies of this gene was determined by Southern
hybridization. Additional subclones and primers were
generated to allow complete DNA sequencing of both strands.
Finally, the DNA sequence was compared with other, related
gene sequences.
Materials and Methods
Medium and growth conditions. Escherichia coli DH5a
was propagated at 37C in Luria broth or on Luria agar
supplemented with 50 mg of ampicillin per liter (Maniatis et
al., 1982).
Genetic methods. Plasmid pUC18 was used as a cloning
vector in all cloning and sequencing experiments unless
otherwise noted. The plasmids pLOHOOl and pLOHOOS harbor
the xylosidase coding region (Sewell et al.. 1989).
Analysis of restriction sites, plasmid purification,
subcloning, DNA ligation, Southern hybridization and other
DNA manipulations were performed using standard methods
(Maniatis et al.. 1982). Restriction enzymes (Bethesda
Research Laboratories, Gathersburg, MD) were used according
to the manufacturer's instructions. Transformed colonies
were screened for xylosidase and arabinofuranosidase
activity on Luria agar plates containing 20 /g/ml of the
flurorogenic substrates 4-methylumbelliferyl-B-D-
xylopyranoside or 4-methylumbelliferyl-a-L-arabinofuranoside

18
(Sigma Chemical Co., St. Louis, MO.). The internal Sau3A
and large internal PstI fragment from pLOI1005 were utilized
as probes in the
Southern hybridization analysis of digested chromosomal B.
fibrisolvens and E. coli DNA.
DNA sequencing. Double-stranded DNA was sequenced in
both directions using the dideoxy-chain termination method
(Sanger, 1982) and Sequenase (United States Biochemical
Corp.) according to the manufacturer's instructions.
Additional sequencing primers were synthesized by the
University of Florida Interdisiplinary Center for
Biotechnology Research and the Department of Microbiology
and Cell Science Nucleotide facility. The DNA sequences
were assembled using the "GENEPRO" software package (Hoefer
Scientific Instruments, San Francisco, Calif.) and the
University of Wisconsin Genetics Computer Group GCG package,
version 6.1 (Devereux et al. 1984) Primary sequence
comparisons were made with GenBank and EMBL sequence
libraries.
Preparation of cell extracts. E. coli cells harboring
the recombinant plasmids were harvested while in mid
exponential phase of growth by centrifugation (10,000 g, 10
min, 4C) and washed twice with 5 mM phosphate buffer (pH
6.8). Cell pellets were stored at -70C, until needed.
Cell pellets were thawed on ice and resuspended in an equal
volume of 5 mM phosphate buffer (pH 6.8) containing 10 mM B-

19
mercaptoethanol and were lysed by two passes through a
French pressure cell at 20,000 lb in'2. Cell membranes and
other debris were removed by centrifugation (100,000 g, 1 h,
4C). Supernatants containing the total cytoplasmic
proteins were stored at -70C.
Enzyme assays. B-D-xylosidase and a-L-
arabinofuranosidase activities were determined by measuring
the rate of hydrolysis of p-nitrophenyl-B-D-xylopyranoside
and p-nitrophenyl-a-L-arabinofuranoside (1 mM final
concentration), respectively, in 50mM phosphate buffer (pH
6.8) at 37C. The nitrophenyl derivatives of other mono-
and disaccharides were examined as possible substrates under
the same conditions. All assays were conducted in a volume
of 1 ml catalysis was terminated by the addition of 2 ml of
500 mM sodium carbonate. The hydrolysis of one nmole of
substrate resulted in an increase of absorbance of 0.007 at
405nm. Specific activities are expressed as nmoles p-
nitrophenol released per minute per milligram of total
protein. Carbohydrate derivatives were purchased from Sigma
Chemical Co. Protein concentration was estimated by the
method of Bradford (Bradford, 1976).
Sodium dodecvl sulfate-polvacrvlamide gel
electrophoresis (SDS-PAGE). Cell proteins were separated in
denaturing gels by the method of Laemmli (Laemmli, 1970).
Protein bands were visualized by staining with Coomassie
blue.

20
Nucleotide sequence accession number. The nucleotide
sequence reported here has been assigned GenBank accession
number M55537.
Results and discussion
Identification of the xvlosidase coding region. Many
restriction endonuclease sites on the original xylosidase-
positive subclone, pLOI1005 (3.2 kb B. fibrisolvens insert),
were mapped to facilitate the generation of futher subclones
for analysis (Fig. 1). Plasmid DNA was purified using
cesium chloride and digested with a battery of restriction
enzymes. Restriction endonuclease sites were used to
generate subclones in both directions with respect to the
lac promoter in pUC18. Each subclone was examined for
enzyme activity on 4-methylumbelliferyl-B-D-xylopyranoside
(MUX) indicator plates. Based upon the results of these
experiments (Fig. 2), the region encompassing the xylosidase
gene was localized to a 2.1 kb DNA segment that spans an
internal EcoRI site and the internal PstI site. The
predicted gene size was in excess of the 1.4 kb Dral
fragment. In addition to the indicated xylosidase a
ctivity, arabinofuranosidase activity was also associated
with all xylosidase-positive subclones. It seems possible
that the gene, classified as xvlB. encodes an enzyme that
has activity against both substrates.
The number of chromosomal copies of xvlB was examined
using Southern hybridization. The large internal PstI and

Figure 1. Restriction endonuclease digestion maps of
plasmids pLOHOOl and pLOI1005 that express 6-D-xylosidase
activity in E. coli DH5a.

Pst!

Figure 2. Subclone analysis of pLOI1005 to localize the xylB
coding region: xyl; xylosidase activity, araf;
arabinofuranosidase activity. A "+" or denotes the
presence or absence of enzyme activity.

kilobase pairs
24
0.0
1.0 2.0 3.0
P S
H
Xvl
Araf
H
H
H
H
S
S
Coding Region

25
Sau3A fragments were used as probes against PstI and Sau3A
digests of E. coli DH5a, B. fibrsolvens GS113 chromosomal
DNA and plasmid pLOI1005 (Fig. 3). These probes did not
bind to the DH5a chromosomal DNA but did bind to a single
band in each B. fibrisolvens chromosomal digest. A single
copy of the xvlB gene appears to be present in B.
fibrisolvens GS113.
Multiple copies of xylosidase genes have been reported
for Bacillus pumilus (Panbangred et al.. 1984) and for
Caldocellum saccharolvticum (Luthi et al.. 1990). If there
are additional xylosidase genes present in GS113, then they
must share limited homology with xvlB.
DNA sequence of the xylosidase gene (xvlB). Plasmid
pLOHOOl, the original GS113 library clone (Sewell et al.,
1989), contained a 4.2 kb insert of B. fibrisolvens DNA.
Both strands of this fragment were sequenced utilizing the
strategy summarized in Fig. 4A. The complete nucleotide
sequence of this fragment is outlined in Fig. 5. Analysis
of the sequence revealed the presence of three open reading
frames (ORF's) in this DNA segment. The first ORF, ORF1,
was incomplete and is 1,340 bp in length. It lacks a Shine-
Dalgarno sequence and an ATG initiation codon. 0RF2, was
1,551 bp in length and was found 15 bp downstream from ORF1.
ORF2 has a Shine-Dalgarno sequence located 6 bp upstream
from the initiation codon and defines a complete gene. This
gne spans the predicted xylosidase coding region and

Figure 3. Southern hybridizations of chromosomal DNA from B.
fibrisolvens GS113 (8 hr exposure) and E. coli DH5a (8 hr
exposure) and plasmid DNA from pLOI1005 (2 hr exposure).
Lanes 1, 2, and 3 contain PstI digests from E. coli. B.
fibrisolvens. and pLOI1005, respectively, probed with the
internal PstI fragment from pLOHOOS. Lanes 4, 5, and 6 are
Sau3A digests of E. coli DH5a, B. fibrisolvens GS113 and
pLOHOOS which were probed with the internal Sau3A fragment
from pLOI1005. The additional bands in lane 6 represent
larger, incomplete restriction endonuclease digestion
products.

27
6

Figure 4. Outline of sequencing strategy of pLOHOOl and
subclone analysis of xvlB. (A) Sequencing strategy of the
complete 4.2 kb B. fibrisolvens insert in pLOHOOl. Arrows
indicate the direction of sequencing. Subclones were
sequenced using universal pUC18 primers (vertical bars in
front of arrows) and additional oligonucleotide primers
(vertical bars absent) were both used. (B) Outline of the
three ORF's sequenced and selected subclones and insertional
inactivation used to identify xvlB. Enzyme activity was
evaluated using MUG indicator plates. Double vertical bar in
pLOI1040 indicates the site of insertion of a 10 bp Notl
linker. Abbreviations: E; EcoRI. H; Hindlll. A; AccI. X;
Xbal. D; Dral, S; SspI. P; PstI. ERV; EcoRV. xyl; xylosidase
activity, araf; arabinofuranosidase activity. A "+" or
denotes the presence or absence of enzyme activity,
respectively.

PLOT 1001
EH AX DSP H ERV H E D S P
29
P
-><*-
> <
-> *-
B
1000
BASE PAIRS
2000 3000
4000
ORF1
ORF2
ORF3
pLOI 1005
xv1 araf
+ +
PLOI1043
pLOI1040
+ +
Notl linker insertion

Figure 5. The complete nucleotide sequence and translated
amino acid sequence of the 4.2 kb insert from pLOHOOl.
Putative Shine-Dalgarno (S.D) sequences and initiation codons
are underlined. Translational termination is indicated by an
asterisk (*) .

31
1 AATTGTGGATGCACATATGAAAAGCTGATTTATGCTTATAAGGCAGGTCTTGTCAAGGAA 60
NCGCTYEKLIYAYKAGLVKE
61 GAGACCATCGATGAGGCTGTTACTCGACTTATGGAAATCAGACTTCGTCTAGGTACTATT 120
ETIDEAVTRLMEIRLRLGTI
121 CCAGAGAGAAAGAGTAAGTATGATGATATCCCATATGAAGTGGTCGAATGCAAAGAGCAT 180
PERKSKYDDIPYEVVECKEH
181 ATCAAACTTGCTCTTGACGCTGCAAAGGATAGCTTTGTCCTTTTGAAGAATGATGGTTTA 240
I KLALDA AKD S FVLLKND GL
241 CTTCCACTGAATAAAAAGGATTATAAATCTATTGCTGTTATTGGACCCAACTCTGATTCA 300
LPLNKKDYKSIAVIGPNSDS
301 AGAAGAGCTTTAATTGGAAATTATGAGGGCCTTTCTTCAGAGTATATTACAGTTTTAGAG 360
RRALIGNYEGLSSEYITVLE
361 GGGATTCGTCAGGTTGTCGGTGATGATATTAGATTATTCCACGCTGAGGGCACTCATCTT 420
GIRQVVGDDIRLFHAEGTHL
421 TGGAAGGATAGAATTCACGTAATCAGTGAGCCAAAAGATGGATTTGCCGAGGCTAAAATC 480
WKDRIHVISEPKDGFAEAKI
481 GTGGCAGAGCATTCAGATTTAGTTGTGATGTGTCTTGGACTTGACGCATCAATCGAAGGA 540
VAEHSDLVVMCLGLDAS IEG
541 GAAGAAGGAGACGAGGGTAATCAGTTCGGTAGCGGAGACAAGCCTGGATTAAAGCTTACA 600
EEGDEGNQFGSGDKPGLKLT
601 GGTTGTCAGCAAGAGCTACTTGAGGAAATTGCCAAAATCGGCAAGCCTGTTGTACTTCTT 660
GCQQELLEE IAKIGKPVVLL
661 GTGCTTTCAGGTTCTGCTCTTGATTTATCATGGGCGCAGGAATCTAATAACGTAAATGCG 720
VLSGSALDLSWAQESNNVNA
721 ATAATGCAGTGCTGGTATCCAGGCGCAAGAGGTGGACGTGCTATTGCAGAGGTTTTATTT 780
IMQCWYPGARGGRAIAEVLF
781 GGCAAGGCCAGTCCAGGCGGTAAAATGCCTCTTACATTTTATGCCTCAGATGATGACCTT 840
GKASPGGKMPLTFYASDDDL
841 CCTGATTTTTCTGATTATTCAATGGAAAATAGGACATACAGATATTTCAAGGGCACACCA 900
PDFSDYSMENRTYRYFKGTP
901 CTTTATCCATTTGGTTATGGACTAGGTTATTCTAAAATTGATTATCTATTTGCTTCTATT 960
LYPFGYGLGYSKIDYLFAS I
961 GATAAAGATAAGGGAGCAATTGGTGATACATTCAAGCTAAAGGTAGATGTTAAAAATACC 1020
DKDKGAIGDTFKLKVDVKNT
1021 GGTAAGTATACACAGCATGAGGCTGTTCAAGTATATGTAACGGACCTTGAGGCAACGACA 1080
GKYTQHEAVQVYVTDLEATT
1081 AGAGTGCCTATTAGAAGCCTTAGAAAGGTTAAATGTCTAGAGCTTGAGCCTGGTGAAACA 1140
RVPIRSLRKVKCLELEPGET
1141 AAAGAGGTTGAATTTACCCTTTTTGCAAGAGATTTTGCCATTATTGATGAAAGGGGAAAA 1200
KEVEFTLFARDFAI IDERGK
1201 TGTATCATAGAGCCAGGCAAGTTTAAGATTTCTATTGGGGGACAACAGCCAGACGATAGA 1260
Cl IEPGKFKISIGGQQPDDR

1261
1321
1381
1441
1501
1561
1621
1681
1741
1801
1861
1921
1981
2041
2101
2161
2221
2281
2341
2401
2461
32
1320
1380
1440
1500
1560
1620
1680
1740
1800
1860
1920
1980
2040
2100
2160
2220
2280
2340
2400
2460
2520
AGTAAAGAACTTATGGGCAGAGAGTGTGATATTTTTGAAATTGAATTAACAGGCTCTGTT
SKELMGRECDIFEIELTGSV
ACAGAAGTTGAATATTAATTGAQAOOTGCATCATGGTTATAGCTAACAATCCAATTTTAA
T E V E Y MVIANNPILK
AAGGTTTTTATCCAGACCCTTCTATCTGCAGAAAAGGGGATGATTTTTATCTAGTTTGTT
GFYPDPSICRKGDDFYLVCS
CAAGTTTTGTGTATGCTCCGGGAGTACCGATTTTTCACACTAAGGATTTGGCACATTTTG
SFVYAPGVPIFHTKDLAHFE
AGCAAATTGGAAATATATTAGACAGAGAAAGTCAACTTCCATTGTCGGGAGATATATCTA
QIGNILDRESQLPLSGD I SR
GAGGCATATTTGCCCCAACAATAAGAGAGCATAATGGAATCTTTTACATGATAACAACTA
GIFAPTIREHNGIFYMITTN
ATGTAAGCTCTGGCGGCAACTTTATTGTTACTGCAAAAGATCCAGCTGGTCCTTGGTCAG
VSSGGNFIVTAKDPAGPWSE
AGCCATATTATTTAGGTGAAGATGAGGCGCCAGGTATTGATCCATCTCTGTTTTTTGATG
PYYLGEDEAPGIDPSLFFDD
ACGATGGCAAATGTTATTACGTTGGTACCAGACCAAATCCTGATGGAGTTCGTTACAACG
DGKCYYVGTRPNPDGVRYNG
GTGATTGGGAGATATGGGTTCAAGAGCTGGATTTAGAGCAAATGAAACTTGTAGGTCCTT
DWE IWVQELDLEQMKLVGPS
CGATGGCAATTTGGAAGGGCGCTCTTAAGGATGTTATTTGGCCAGAAGGACCACACCTTT
MAIWKGALKDVIWPEGPHLY
ATAAGAAAGATGGATATTATTATCTTTTACATGCAGAAGCTGGCACAAGCTTTGAACATG
KKDGYYYLLHAEAGTSFEHA
CTATTTCTGTAGCTCGCTCAAAGGAGCTATTCAAATGGTTTGAGGGATGTCCTAGAAATC
I SVARSKELFKWFEGCPRNP
CTATATTTACCCATAGAAATTTAGGCAAGGATTATCCAGTATGCAATGTTGGACATGCTG
I FTHRNLGKDYPVCNVGHAD
ATTTAGTTGATGATATCAATGGCAACTGGTATATGGTGATGCTGGCATCTAGACCATGCA
LVDDINGNWYMVMLASRPCK
AGGGAAAGTGCAGCTTGGGACGAGAGACATTCCTTGCAAAAGTAATTTGGGAAGACGGAT
GKCSLGRETFLAKVIWEDGW
GGCCAGTGGTTAATCCGGGAGTTGGTCGTTTGACTGATGAGGTGGAGATGGACCTTCCTG
PVVNPGVGRLTDEVEMDLPE
AATATAGATTCTCAAAAGAGATTACTACAAAGGATAAAATGACCTTTGAAGAGACAGTCC
YRFSKEITTKDKMTFEETVL
TAGATGATAGATTTGTTGGAATTGAAAGAAGAAGTGAGGACTTTTATTCCCTTACTGACA
DDRFVGIERRSEDFYSLTDN
ATCCTGGATTCTTAAGATTAAAGCTTCGTCCTGAGGCCATAGAAAATACTGGCAATCCAT
PGFLRLKLRPEAIENTGNPS
CTTACTTAGGAATTCGTCAAAAGACTCATTCGTTTAGAGCAAGCTGTGGCCTTAAGTTTA
YLGIRQKTHSFRASCGLKFT

2521
2581
2641
2701
2761
2821
2881
2941
3001
3061
3121
3181
3241
3301
3361
3421
3481
3541
3601
3661
3721
3781
3;
2580
2640
2700
2760
2820
2880
2940
3000
3060
3120
3180
3240
3300
3360
3420
3480
3540
3600
3660
3720
3780
3840
CACCAGCAAAAGATAATGAATGTGCAGGAATGGTGTTATTCCAGAATAATGAAAATCACT
PAKDNECAGMVLFQNNENHL
TGGAGCTTTTAGTTGTAAAGAAGAAAGATAAGCTACAGTTTAAAGTAGGACCAGTTATTA
ELLVVKKKDKLQFKVGPVIK
AAGGAACCAAAATCAGACTTGCTACTTTTGATATTTCATCAGGTGATTTAGAAATTATTC
GTKIRLATFDI SSGDLE I IL
TTGAGGCAGCAAATCAGCTGGCTAATATCTATATTAAAAAGAATAATGAAAAGATTCTTG
EAANQLANIYIKKNNEKILV
TGGCAGAATGTATTGATTTGAGCCCATACACTACCGAAGAATCAGGCGGATTCGTAGGAT
AECIDLSPYTTEESGGFVGC
GTACCATTGGACTATATGCTTCTTCAAATGGAAAGACCAGTGATAACTATTGCGATTATT
TIGLYASSNGKTSDNYCDYS
CCTACTTTACAGTAGAAGAAGTATAGCATTTTCAATGAGCGAATTTGCAAGTTTTATATA
YFTVEEV*
CGGGATTAATTGTACGTAAAAACCATACAGGTGTAAAATAGTTTCCAGAGAAAGTTTTTT
CTCTGGAATTTTTTATTATQOAQOGGATTATGCTTCAGGAAAGTATTAAGAAGTTGGTAC
MLQESJKKLVQ
AGTACGGTATTGATATGGGGCTTACACCAGAATGTGAGAGAATATATACTACAAATCTTT
YGIDMGLTPECERIYTTNLL
TGCTTGAATGTATGAAAGAAGATGAGTACATAGATCCAGACTGTGATTTAAGCAATATTA
LECMKEDEYIDPDCDLSNI I
TACTTGAAGATGTATTAAAGGAACTTTTAGATGAGGCAGTTAATAGAGGTATCATAGAGG
LEDVLKELLDEAVNRGI IED
ATTCAGTTACACATAGGGATTTGTTTGATACAAAGCTAATGAATCAGCTATGCCCACGTC
SVTHRDLFDTKLMNQLCPRP
CTAAACAGGTTATAGATGATTTTAACCGTATATACGATAACCATGGTCCAATAGCTGCAA
KQVIDDFNRIYDNHGPIAAT
CAGATTATTTTTACAAGTTAAGCAAAGCCTCTGACTATATCCGTACTTACAGGGTAAAAA
DYFYKLSKASDYIRTYRVKK
AGGACCTAAAATGGACATGCGATACAGAGTATGGCACTCTTGACATAACAATTAATCTCT
DLKWTCDTEYGTLDITINLS
CTAAGCCAGAAAAAGACCCAAAGGCAATTGCTGCAGCTAAGAATGCAAAACAATCCACAT
KPEKDPKAIAAAKNAKQSTY
ATCCGAAGTGCCAATTATGTATGGAAAATGAAGGCTATGCTGGTCGCATTAATCATCCTG
PKCQLCMENEGYAGRINHPA
CTAGAGAGAATCATCGCATAATTCCTATAACTATAAATAACAGCAACTGGGGATTTCAAT
RENHRI IPITINNSNWGFQY
ATAGCCCATACGTTTATTACAATGAGCATTGCATAGTCTTTAACGGAGAGCATACTCCTA
SPYVYYNEHCIVFNGEHTPM
TGAAAAXAGAGCGAGCTACTTTTGTTAAGCTATTTGATTTTATCAAACTATTTCCACACT
KIERATFVKLFDFIKLFPHY
ATTTTTTAGGAAGCAATGCTGATTTACCAATTGTTGGAGGATCTATTTTAAGCCATGACC
FLGSNADLPIVGGSILSHDH

34
3900
3841
3901
3961
4021
4081
4141
ATTTCCAAGGCGGCCATTACACATTTGCCATGGAAAAAGCTCCAATTATTCAGGAATTTA
FQGGHYTFAMEKAP I IQEFT
CTGTAAAAGGATATGAGGATGTTAAGGCTGGTATAGTTAAATGGCCACTTTCAGTAATTA
VKGYEDVKAGIVKWPLSVIR
3960
GACTTCAGTGCAAGGATGAGACTAGACTTATTGATTTAGCGACTAATATATTAGACAAAT
LQCKDETRLIDLATNILDKW
4020
GGAGAAATTACACCGATGAAGAGGCATATATTTTTGCTGAAACAGATGGTGAGCCTCACA
RNYTDEEAYIFAETDGEPHN
4080
ATACGATTACACCTATTGCTAGAAAAAGAGGGGATTACTTTGAACTAGATCCTCTAGAGT
TITPIARKRGDYFELDPLES
4140
CGACCTGCAGGCATGCAAGCTTGGCACTGGCCGTCGTTTTACAACGTCGTGACTGGG* 4198
TCRHASLALAVVLQRRDW

35
encodes a 517 amino acid protein having a calculated
molecular weight of 58,421. 0RF3, is located 123 bp
downstream from ORF2 and includes a Shine-Dalgarno seguence
with an initiation codon 5 bp downstream. This ORF
continues for 1,173 bp until the end of the clone and is
also incomplete. No predicted "stemloop" structures or
sequences that resemble rho-independent transcriptional
terminators were identified by computer analysis in the
region between 0RF2 and 0RF3. It is therefore unlikely that
this DNA functions as a transcriptional terminator in E.
coli. It is also unlikely that any transcriptional
terminators are present between 0RF1 and 0RF2 since these
two ORF's are only separated by 15 bp and 0RF2 is expressed
in large amounts in E. coli in constructs that also contain
0RF1. This evidence suggests that these three ORF's may
constitute part of a xylan-degrading operon in B.
fibrisolvens.
Codon usage. The codon usage of the three B.
fibrisolvens ORF's is summarized in Table 1. For
comparative purposes, codon usage for B. fibrisolvens 49
xynA (Mannarelli et_al. 1990) and the average codon usage
for E. coli (Allf-Steinberger, 1984) are included. The
three B. fibrisolvens ORF's have similar patterns of codon
usage with each other and with strain 49 xvnA. The low
guanine plus cytosine content of B. fibrisolvens is
reflected in the three ORF's in the usage of an A or a T in

36
TABLE 1. Comparison of codon usage frequency for the three B.
fibrisolvens ORF's.
Frequency (mol %) Codon Usage
Amino Acid
Codon
B.
ORF 1
fibrisolvens E.
ORF 2 ORF 1
coli
B. fibrisolvens
xv 1A
Phe
TTT
2.7
4.1
3.6
1.3
2.9
TTC
0.9
1.2
0.3
2.2
1.5
Leu
TTA
2.0
2.5
2.8
0.7
0.7
TTG
0.2
1.2
1.0
0.9
0.5
CTT
5.6
2.7
2.5
0.8
4.4
CTC
-0-
-0-
0.3
0.8
-0-
CTA
1.4
1.0
1.8
0.2
0.2
CTG
0.5
0.8
0.3
6.8
1.0
He
ATT
4.7
3.9
4.3
2.2
2.7
ATC
2.0
1.0
0.8
3.7
1.7
ATA
0.5
1.7
3.8
0.2
0.5
Met
ATG
1.4
1.7
2.0
2.8
2.7
Val
GTT
2.9
2.7
2.3
2.9
2.0
GTC
0.9
0.2
0.5
1.2
0.5
GTA
1.4
2.1
1.3
1.8
4.2
GTG
1.1
1.2
-0-
2.2
1.0
Ser
TCT
2.3
1.5
0.8
1.3
1.2
TCC
-0-
0.4
0.3
1.5
0.5
TCA
1.8
1.5
0.5
0.4
2.9
TCG
-0-
0.6
0.3
0.6
0.2
4
AGT
0.9
0.8
0.3
0.3
-0-
AGC
0.7
1.0
1.8
1.4
1.2

37
TABLE 1. (continued)
Frequency (mol %> Codon Usage
Amino Acid
Pro
Thr
Ala
Tyr
His
Gin
Codon
B.
ORF 1
fibrisolvens E.
ORF 2 ORF 1
coli
B. fibrisolvens
xvlA
CCT
1.4
1.7
1.8
0.5
1.2
CCC
0.2
-0-
-0-
0.3
-0-
CCA
2.3
3.5
2.8
0.7
2.4
CCG
-0-
0.6
0.3
2.5
-0-
ACT
0.7
1.9
2.3
1.1
0.7
ACC
0.7
1.4
0.5
2.4
1.0
ACA
2.7
1.7
3.1
0.3
5.1
ACG
0.5
-0-
0.3
0.8
0.5
GCT
2.7
2.1
2.8
2.6
1.5
GCC
1.1
0.4
0.8
2.2
1.2
GCA
2.0
2.3
2.0
2.3
4.1
GCG
0.5
0.2
0.2
3.2
-0-
TAT
4.3
3.5
2.8
1.0
3.9
TAC
0.2
1.2
2.5
1.5
3.4
CAT
0.9
1.4
2.5
0.7
1.5
CAC
0.5
0.6
in
O
1.2
0.2
CAA
0.7
1.0
1.3
1.0
0.5
CAG
1.6
0.6
1.5
3.2
2.2
AAT
2.0
4.2
3.8
1.0
5.4
AAC
0.5
1.1
1.3
2.8
2.7
Asn

38
TABLE 1. (continued
Frequency (mol %> Codon Usage
Amino Acid
Codon
B.
ORF 1
fibrisolvens E.
ORF 2 ORF 1
coli
B. fibrisolvens
xvlA
Lys
AAA
3.4
3.1
3.6
4.1
3.4
AAG
4.7
3.5
3.3
1.3
3.9
Asp
GAT
5.8
5.6
5.9
2.5
3.9
GAC
1.6
1.4
2.0
3.0
0.2
Glu
GAA
4.3
3.7
3.7
4.9
1.7
GAG
4.9
3.5
3.8
1.8
2.7
Cys
TGT
1.4
1.4
1.0
0.4
0.7
TGC
0.7
1.0
1.5
0.5
0.5
Trp
TGG
1.3
1.7
1.3
0.7
2.0
Arg
CGT
0.7
0.8
1.3
3.1
1.0
CGC
-0-
0.2
0.5
2.0
0.2
CGA
0.2
0.2
0.2
0.2
-0-
CGG
-0-
-0-
-0-
0.2
-0-
AGA
3.1
2.9
2.0
0.1
2.0
AGG
0.5
-0-
0.8
0.1
0.2
Gly
GGT
3.4
1.7
1.5
3.8
2.4
GGC
2.3
2.1
1.0
3.1
1.5
GGA
3.1
4.4
1.5
0.4
2.7
GGG
0.5
0.2
0.8
0.6
0.5

39
the wobble position with the exception of CAG for Gin, AAG
for Lys, and GAG for Glu.
Insertional inactivation and subclone analysis of
0RF2. Subclones were generated to investigate the
relationship between the xylosidase and arabinofuranosidase
activities encoded by the xylB gene (Fig. 4b). Retention or
loss of enzymatic activity was initially asssayed on 4-
methylumbelliferyl-B-D-xylopyranoside (MUX) and 4-
methylumbelliferyl-a-L-arabinofuranoside (MUA) indicator
plates. The SspI fragment (1,843 bp) from pLOI1005, which
contains 17 bp upstream and 274 bp downstream in addition to
xvIB. was subcloned in both directions in pUC18. Both
activities were concurrently expressed only when xvlB was
cloned in the direction of transcription of the lac promoter
(pLOI1043), indicating a dependence on this promoter in E.
coli. The insertion of a 10 bp Notl linker into the unique
EcoRV site of xvlB (pLOI1040) resulted in a frameshift
mutation that abolished both enzymatic activities (Fig. 4b).
The results of the indicator plate assays were confirmed by
comparing the specific activities for xylosidase and
arabinofuranosidase in cell free extracts (Table 2.). Using
p-nitrophenol derivatives, arabinofuranosidase activity was
approximately 1.7-times higher than the xylosidase activity.
The ratio of these activities was the same for the three
active subclones. The original subclone, pLOI1005,
exhibited the highest specific activity. The smaller active

40
TABLE 2. Expression of enzyme activities in recombinant E. coli
harboring xvlB.
plasmid
Sp act*
xylosidase
arabinofuranosidase
Ratio ara/xylb
pLOI1005
9.0
16.0
1.8
pLOI1040
0.2
0.2
-
pLOI1043
2.0
3.0
1.5
PLOI1050
6.8
10.1
1.5
pUC18
0.2
0.2
-
1 Nanomoles per minute per milligram of cell protein.
b Ratio calculated after subtraction of background values from
the pUC18 control.

41
subclone, pLOI1043, exhibited a four-fold decrease in both
activities but retained a similar ratio of
arabinofuranosidase to xylosidase activities. An additional
subclone, pLOI1050, contained two SspI fragments each
harboring xvlB oriented with the direction of transcription
from the lac promoter. This subclone exhibited a three-fold
increase in both enzymatic activities with respect to the
single insertion (pLOI1043) but less than the original
clone. Again the ratio of the two activities remained
essentially the same as the wild type (pLOHOOS) The
results of these experiments demonstrate the dependence of
both enzymatic activities on 0RF2.
Presence of other alvcosidic activities. The presence
of additional hydrolytic activities was examined in the
xvlB-encoded protein using various ortho- and para-
nitrophenyl glycosidic substrates (Table 3). No additional
activity above the background levels was detected with 12
other pentose and hexose derivatives. Low levels of
activity was detected against the o-NP-B-D-fucopyranoside.
This may not be significant since the activity represents
less than 5% of the activity against the B-D-xylopyranoside.
A 19-fold higher activity was detected against o-
nitrophenyl-6-D-xylopyranoside relative to the para-
substituted derivative. This phenomenon is analogous to
B-galactosidase from E. coli. The para- and ortho-
substituted substrates are known to have different rates for

42
TABLE 3. Hydrolysis of different nitrophenyl-substituted glycosides by the xvlB
gene product.
substrate
SDecific activity*
pLOI1005
pUC18
p-NP-B-D-xylopyranoside
8.9
0.2
p-NP-a-L-arabinofuranoside
15.5
0.5
p-NP-a-L-arabinopyranoside
0.1
0.2
p-NP-a-D-galactopyranoside
0.2
CM
O
p-NP-a-D-glucopyranoside
3.5
3.3
p-NP-a-L-fucopyranoside
0.2
0.2
p-NP-fi-D-fucopyranoside
0.2
0.2
p-NP-B-L-fucopyranoside
0.4
0.4
p-NP-a-L-rhamnopyranoside
0.3
0.2
o-NP-fi-D-fucopyranoside
1.4
1.0
o-NP--D-galactopyranoside
0.8
1.0
o-NP-B-D-galactopyranoside
1.0
1.0
* Nanomoles per minute per milligram of cell protein.

43
the glycosidic bond-breaking step (Martinez-Bilbao et al.,
1991). Under the conditions of these assays it appears that
the xvlB-encoded protein is limited to hydrolytic activity
against 6-D-xylopyranosides and a-L-arabinofuranosides only.
Electrophoretic analysis of cloned proteins. Using
SDS-PAGE analysis of cell-free cytoplasmic extracts, a new
protein band with an apparent molecular weight of 60,000 was
observed in cells harboring plasmid pLOI1005 (Fig. 6).
This band was absent in extracts from cells containing the
vector plasmid pUC18 alone. Extracts from which the gene
was inactivated by a frameshift mutation (pLOI1040) also
lacked this protein band. The observed levels of this
protein band in the single (pLOI1043) and double (pLOI1050)
SspI subclones was consistent with the presence of the
enzyme.
Primary sequence homology comparisons. Homologies of
the B. fibrisolvens ORF's to other glycohydrolases were
compared to determine evolutionary relatedness. The
translated amino acid sequences of the three B. fibrisolvens
ORF's exhibited 42 to 45% similarity (a conservative match)
and 14 to 19% identity (an exact match) with each other
(Table 4.). The xvlB was found to be most similar (44%
similarity, 20% identity) to the B-glucosidase from
Kluyveromyces fragilis. Additional comparisons with other
glycohydrolase sequences revealed no significant amino acid
identities. Although the N-terminal sequence of the

Figure 6. SDS-PAGE analysis of cytoplasmic extracts from
recombinant E. coli DH5a harboring selected plasmids.
Approximately 20 /Ltg of protein was loaded in each lane. Lanes
1 and 7; molecular weight markers, lane 2; pLOI1005, lane 3;
pUC18, lane 4; pLOI1040, lane 5; pLOI1043, lane 6; pLOI1050.
The band cooresponding to the xylosidase-arabinofuranosidase
enzyme is indicated by an arrow. The numbers in the right
represent the apparent molecular weight of the standards
(X 103) .

45
1 2 3 4 5 6 7

46
TABLE 4. Comparison of the translated amino acid sequences of the three B.
fibrisolvens ORFs in pLOI 1001 with those of selected proteins.
% Similarity (% Identity)
Organism (gene)
ORF 1
ORF 2
xylB)
ORF 3
Reference
B. fibrisolvens x^
cnh)
40 (16)
41 (16)
41 (13)
Mannarelli
et al. 1990
B. fibrisolvens (endl)
16 (8)
14 (6)
20 (11)
Berger
et al. 1989
B. Dumilus (xvnA)
46 (21)
44 (16)
12 (8)
Fukusaki
et al. 1984
B. Dumilus (xvnB)
39 (16)
28 (21)
38 (17)
Moriyama
et al. 1987
B. subtilis (xvnA)
48 (21)
44 (16)
12 (6)
Paice
et al. 1986
Caldocellum
saccharolvticus (x\
rnB)
20 (14)
25 (12)
21 (12)
Luthi
et al. 1990
C. saccharolvticus
(xvnA / xvnB)
16 (8)
20 (9)
19 (10)
Luthi
et al. 1990
C. saccharolvticus
(xynC)
16 (8)
22 (11)
31 (3)
Luthi
et al. 1990
Clostridium
thermocellum
18 (7)
40 (16)
15 (8)
Grepinet
et al. 1988
(xynZ)
C. thermocellum
icelA)
45 (22)
41 (18)
39 (15)
Benguin
et al. 1985
C. thermocellum
(celBi
44 (20)
44 (18)
43 (19)
Grepinet
et al. 1986
C. thermocellum
(celD1
45 (16)
39 (15)
43 (18)
Joliff
et al. 1986
AsDeraillus
nicer
(a-amylase)
40 (16)
41 (18)
43 (19)
Boel
et al. 1984
Kluvveromvces
fraailis
(B-glucosidase)
52 (31)
44 (20)
46 (22)
Raynal et al. 1987
ORF1
100
44 (19)
45 (19)
This study
ORF2
44 (19)
100
42 (14)
This study
ORF3
45 (19)
42 (14)
100
This study

47
Bacillus pumilus B-xylosidase did exhibit strong amino acid
identity in selected regions, the overall identity was only
21%. Thus the B. fibrisolvens xvlB gene is evolutionarily
divergent from other glycohydrolases. The translated,
primary sequences for 0RF1 and 0RF3 also exhibited
similarity (52% and 46%, respectively) and identity (31% and
22%, respectively) to the K. fraqilis S-glucosidase. This
is consistant with these two ORF's also being involved in
carbohydrate degradation.
It has been postulated that the hydrolytic mechanism of
lysozyme (Teeri, et al. 1987) and cellulases (Knowles et
al., 1987) can serve as a model for other carbohydrate
hydrolyzing enzymes. Studies of hen egg-white lysozyme
(HEWL) indicate a general acid-base catalytic mechanism
involving Glu-35 and Asp-52 as the catalytic residues
(Quiocho, F. A., 1986). Subsequent studies have
demonstrated that this catalytic region is conserved in some
cellulases (Knowles et al. 1987). An analysis of the
translated primary sequence from xvlB reveal a region
homologous to the active site region from HEWL, and
glucoamylase from Aspergillus niqer (Table 5). The
conserved region from additional carbohydrate hydrolases are
included for comparison. The xvlB region was most similar
to the glucoamylase, with 38% identity between the amino
acids in the catalytic region. The catalytically important

48
TABLE 5. Amino acid sequence alignment of conserved regions.
Protein* Concensus sequence Reference
HEWL
A. n.
B.f.
B.f.
B.f.
B.f.
B.f.
B.p.
35 44
FES
N
F N
T
Q
A
T
. N R
N
T D
G
S


331
338
(a-aml)
P E D
T
y .
y
N
G

. N P
W
F L
C
T
L A
A
342
350
xyl.Bi
SED
F
y s
L
T
D

. N P
G
F L
R
L
K L
R
144
149
xylJ,
P D G
V
r y
A
W E
I
W
V Q
E
320
328
endl
GET
S
A T
N
R
N

. N T
A
E R
V
K
W A

355
362
xvnA
NEK
P
L I
W
S


. N I
G
V A
K
P
a y

769
ball
S D W
W
G F
G
E
H
y
K .
E
V
L A
G
325
335
xvnB
I E C
T
R L
A
Q
L
N
W N T
C
S M
Q
F
V .

52
t d y
Yaauchi et al. 1983
349
A E Q
Boel et al. 1984
362
PEA
This study
160
L D L
This study
337
. D y
Beraer et al. 1989
369
. D E
Mannarelli et al. 1990
785
N D I
Barnett et al. 1991
343
. E E
Morivama et al. 1987
* Abbreviations: Hen egg-white lysozyme (HEWL), Aspergillus niger a-amylase (a-
aml)/ B. fibrisolvens fl-D-xylosidase (xvlBl. B. fibrisolvens endoglucanase 1 (endl),
B. fibrisolvens B-D-xylanase (xvlA). B. fibrosolvens B-glucosidase (ball). Bacillus
pumillus B-D-xylosidase (xylB), Clostridium thermocellum fl-D-xylanase (xvnZ). C.
thermocellum cellobiohydrolase B (celB), C. thermocellum cellobiohydrolase D (celD),
C. cellulolyticum endoglucanase A (EGCCA), Caldocellum saccharolvticum B-D-
xylosidase (xynB), Cellulomonas fimi endoglucanase A (cenA), C. fimi exoglucanase A
(cex), Trichoderma reesei cellobiohydrolase II (CBH II), Kluvveromvces fraailis
cellobiohydrolase I (CBH I), Schizophvllum commune endoglucanase (EG I).

49
TABLE 5. (continued)
Protein*
Concensus sequence
Reference
458
466
475
C.t. xvnZ
G E A
L
L
R
A
D
V

. N R
S
G
K
V
D
S


T D Y
418
427
436
C.t. celB
TEG
G
H
P
L
L
D
L
. N L
K
Y
L
R
C
M
R

. D F
376
384
395
C.t. celD
DEE
Y
L
R
D
F
E

. N R
A
A
Q
F
S
K
K
E
A D F
408
418
427
C.s. xvnB
REV
F
V
E
R
I
D
E
Y N A
N
P
K
R
V
W
L

. E M
244
254
263
T.r. CBH II
L E C
I
N
Y
A
V
T
Q
L N L
P
N
V
A
M
Y
L

. D A
586
596
605
K.f. CBH I
G E W
E
T
E
G
Y
D
R
E N M
D
L
P
K
R
T
N

. E L
33
42
50
SC EG X
N E S
C
A
E
F
G
N
Q
. N I
P
G
V
K
N



T D Y
Grepinet et al. 1988
Grepinet and Benguin,
1986
Joliff et al. 1986
Luthi et al. 1990
Rouvinen et al. 1990
Raynal et al. 1987
Yaguchi et al. 1983
* Abbreviations: Hen egg-white lysozyme (HEWL), Aspergillus niger a-amylase (a-
aml), B. fibrisolvens fi-D-xylosidase (xvlB). B. fibrisolvens endoglucanase 1 (endl),
B. fibrisolvens fi-D-xylanase (xylA), B. fibrosolvens B-glucosidase (ball 1. Bacillus
pumillus fi-D-xylosidase (xvlBl. Clostridium thermocellum B-D-xylanase (xvnZ). C.
thermocellum cellobiohydrolase B (celB), C. thermocellum cellobiohydrolase D (celD),
C. cellulolvticum endoglucanase A (EGCCA), Caldocellum saccharolvticum B-D-
xylosidase (xvnB), Cellulomonas fimi endoglucanase A (cenA), C. fimi exoglucanase A
(cex), Trichoderma reesei cellobiohydrolase II (CBH II), Kluvveromvces frgilis
cellobiohydrolase I (CBH I), Schizophvllum commune endoglucanase (EG I).

50
glutamic and aspartic acid residues and the approximate
spacing were found to be conserved conserved.
Conclusions
The xvlB gene, encoding 6-D-xylosidase and a-L-
arabinofuranosidase activities is the first of its kind to
be sequenced. The xvlB gene is 1,551 bp in length and
encodes a 517 amino acids protein having a predicted
molecular weight of 58,000. The absence of any significant
stem-loop structures or terminators in the regions between
ORF1, ORF2, and ORF3 as well as the strong expression of
ORF2 in E. coli even when preceded by ORF1 suggests that
these three genes may constitute a xylan-degrading operon.
The subcloning analysis and insertional inactivation studies
demonstrate the dependence of both activities on the intact
xvlB gene.
The codon usage of the three ORF's is consistent with
the low guanine plus cytosine content of this organism in
general (Mannarelli et al. 1990).
This enzyme exhibited B-D-xylopyranosidase and a-L-
arabinofuranosidase activities. No additional glycosidic
bond cleavage activities were detected in the xvlB gene
product.
The xvlB gene displayed limited homology to other
reported xylosidase sequences and must therefore be
considered to be evolutionarily divergent from genes
encoding similar functions from other organisms. It did

51
however exhibit partial identity with the B-glucosidase from
Kluyveromvces fraailis which is consistent with the
similarity of the substrates which these two enzymes attack.
A single gene encoding xylanase/xylosidase activities
has been cloned and seguenced from Caldocellum
saccharolvticum (Luthi et al. 1990). This protein, however,
lacked arabinofuranosidase activity. Recently, a
xylosidase/arabinofuranosidase gene was proposed to reside
in a gene cluster isolated from Bacteroides ovatus
(Whitehead and Hespell, 1990). All clones exhibited both
activities concurrently and both activities co-purified.
Additionally, an enzyme having xylosidase and
arabinofuranosidase activities has been purified from
Trichoderma reesei (Poutanen and Puls, 1988). No seguence,
however, for the encoding gene has been reported.
Substrate ambiguity between carboxymethylcellulase and
xylanase enzymes is relatively common (Flint et al. 1989).
The substrate ambiguity for other xylosidase enzymes has
also been reported. By employing kinetic methods on the
purified, bifunctional B-xylosidase/B-glucosidase from
Chaetomium trilaterale. Uziie et al. (1985) demonstrated
that this enzyme possessed a single active site with dual
substrate-binding regions. More recently, a neopullulanase
from Bacillus stearothermophilus was cloned, sequenced, and
characterized (Kuriki and Imanaka, 1989). This enzyme
possessed activity against a-(1,6)-glycosidic linkages in

52
addition to the usual hydrolysis of a-(1,4)-glycosidic
linkages. Mutational analysis demonstrated that a single
active center was involved in the catalysis of both these
linkages (Kuriki et al. 1991).
It seems reasonable to speculate that bifunctionality
and substrate ambiguities among the microbial carbohydrate
hydrolases is common. The celB gene encoding a "true"
bifunctional cellulase has been cloned from Caldocellum
saccharolvticum and sequenced (Saul et al. 1990). This
enzyme exhibited both endo-glucanase and exo-glucanase
activities. The endo-glucanase activity was localized to
the carboxy terminal domain and the exo-glucanase activity
was localized to the amino terminal domain. This protein
also exhibited homology with both endo- and exo-glucanase
enzymes from other organisms. The organization of separate
functions to separate domains has also been demonstrated
with the endo-glucanase 2 from Bacteroides succinoaenes
(McGavin and Forsberg, 1989). These investigators used
protease treatments to demonstrate that this enzyme
possessed separate substrate binding and catalytic domains.
The structural similarities between the various 6- and
a-linked glycosyl residues may be responsible for the
apparent evolution of enzymes with broad substrate
specificity.

53
Since many enzymatic activities are required to completely
depolymerize xylans and cellulose, the evolution of such
enzymes could represent a selective advantage in the rumen
and other environments.

Chapter IV
Mutational analysis of the xvlB gene
Introduction
Carbohydrate-degrading enzymes have been studied
extensively in the microbial world and form the basis of
much of what we know about the cycling of carbon in the
environment (Weinstein and Albersheim, .1979). Lysozyme, an
enzyme which hydrolyzes bacterial cell wall carbohydrates,
was one of the first such enzymes to be studied extensively
and as a result much is known about this enzyme's catalytic
mechanism and structure (Quiocho, 1986). The cellulases
have also been extensively studied as they are responsible
for the cycling of the most abundant natural polymer
cellulose (Knowles et al.. 1987). Studies of the
mechanistic properties of these enzymes have been
facilitated by the use of molecular genetic techniques.
Gene cloning, sequencing, and oligonucleotide-directed
mutagenesis have allowed mutations to be made in a site
specific manner. A kinetic study of the mutant proteins can
then be done and predictions about catalytic mechanisms
tested.
54

55
The active center for a neopullulanase from Bacillus
stearothermophilus has been recently examined using
mutagenesis (Kuriki et al.. 1991). This enzyme exhibits
dual activities against a-(1,4) and a-(1,6) glycosidic
linkages. The catalytically important amino acids were
tentatively identified using seguence alignment and homology
searches. The putative catalytic amino acids were changed
using site-directed mutagenesis and activities were examined
in the resulting mutants. This approach identified that one
active center containing Glu-357 and Asp-424 was responsible
for both catalytic activities.
Enzymes which exhibit substrate ambiguity are
interesting both from mechanistic and evolutionary
perspectives. The obvious question that arises with respect
to bifunctionality is does this enzyme have two separate,
specialized, catalytic sites? This situation would imply
that the protein has evolved from a gene fusion to perform
two separate functions. Another possibility involves the
presence of a single active site in which two structurally
similar substrates are bound and hydrolyzed. This situation
would be analogous to a case of mistaken substrate identity
which proves advantagous to the organism, and has been
evolutionarily conserved. Examples of multifunctional
xylosidases have been reported in the literature and are
discussed in the previous chapters. The apparent
bifunctionality of the xvlB gene was demonstrated in chapter

56
III. A series of genetic experiments were designed to
investigate the presence or absence of two catalytic or
functionally separate domains on the xvlB gene that are
responsible for the dual activities exhibited by this
enzyme.
Materials and Methods
Medium and growth conditions, genetic methods and DNA
sequencing were done as described in chapter III.
In vitro nitrous acid mutagenesis of xvlB. A total of
80 Mg of pLOI1005 which contains the xvlB gene was
resuspended in 50 Ml Tris-EDTA (TE) buffer (pH 8.0).
Mutagenesis was initiated by the addition of 10 Ml of 2.5 M
sodium acetate (pH 4.3) and 50 Ml 2.0 M sodium nitrite.
Exposure times were zero, thirty seconds, one, two five and
ten minutes. Mutagenesis reactions were terminated by the
addition of 200 m! 100% ethanol. The precipitation step was
repeated twice to ensure the complete removal of the
mutagenic agent. The mutagen-treated plasmids were
resuspended in 80 Ml TE buffer (pH 8.0). A total of 5 Ml of
the plasmid was transformed into competent E. coli DH5a.
Serial ten-fold dilutions of the transformed cells were
plated in triplicate onto Luria agar supplemented with the
fluorogenic substrates. A 99% reduction in transformation
by the mutagenized plasmid was observed after ten minutes of
mutagenesis.

57
Localization of point mutations. The entire xvlB
coding region was divided into five domains based upon
restriction sites (Fig. 7). Three restriction fragments,
the PstI. EcoRI. and Hindlll fragments were isolated from
each mutant plasmid. These fragments were used to replace
the corresponding fragment in the wild type gene which had
been modified to construct receiving vectors for each
respective restriction fragment (pLOI1051, pLOI1052, and
pLOI1053) to test the functionality of individual fragments.
This strategy, outlined in Fig. 8, allowed the localization
of point mutations to one or more of the five domains.
Exonuclease III deletion of xvlB. Plasmid pLOI1043
containing the xvlB coding region was deleted from the 3'
terminal region into the coding region by exonuclease III
using the "Erase-a-Base" deletion kit (Promega Corporation,
Madison, WI.) according to the manufacturer's instructions.
The deleted plasmids were subsequently transformed into
competent E. coli DH5a and screened in the same fashion as
the in vitro-qenerated mutant plasmids.
5' deletion analysis and lacZ1 fusions. The internal
PstI fragment from pLOI1005 was subcloned in the original
orientation into plasmid pUC18. This in-frame fusion with
lacZ results in the subsequent deletion of the 5' terminal
54 base pairs (18 amino acids).
Preparation of cell extracts. Extracts were prepared
as described in chapter III.

Figure 7. The assignment of domains to the xylB gene (solid
bar) Restriction endonuclease fragments used to localize
mutations are shown below. Abbreviations: E; EcoRI. S; SspI.
P; Pstl. H; Hindlll.

59

Figure 8. Subcloning strategy used to localize in vitro
mutations to one of five domains on the xvlB gene.
Abbreviations: E; EcoRI. S; SspI. P; PstI. H; Hindlll.

61
kilobase pairs
1.0 2.0 3.0
pUC18
LacZ=-
E
SP
pLOI1005
H
S P
PstI
EcoRI
Hindlll
pUC18
7
!
LacZ=^
PLOI1051
pUC18
S P
LacZ=
PLOI1052
PLOI1053

62
Enzvme assays. Enzyme assays were done as described in
chapter III.
Sodium dodecvl sulfate polyacrylamide gel
electrophoresis (SDS-PAGE). SDS-PAGE gels were done as
described in chapter III.
Native polyacrylamide gel electrophoresis (native
PAGE). Cell proteins were separated in non-denaturing gels
by the method of Ornstein and Davis (1964). Following
electrophoresis, gels were equilibrated in 50 mM sodium
phosphate buffer (pH 6.8). The equilibrated gels were then
overlaid with Whatman #1 filter paper soaked with a solution
of 20 mg per ml of the fluorogenic substrates, 4-
methylumbelliferyl-B-D-xylopyranoside or 4-
methylumbelliferyl-a-L-arabinofuranoside in 70% ethanol.
Overlays were incubated at 37C for 15 min or until activity
bands were visible under long-wave UV light. Native-PAGE
gels were also stained for protein as outlined before.
Western hybridization analysis of wild type and mutant
proteins. Native and SDS-PAGE protein gels were
electroblotted using the Trans-Blot apparatus (BioRad
Laboratories, Richmond, CA.) according to the manufacturer's
instructions. "Western" hybridizations were done using
polyclonal antisera raised to E. coli DH5a cell extracts
harboring pLOI1005 in rabbits. Protein bands were
visualized using alkaline phosphatase conjugated goat anti-

63
rabbit antisera. All procedures and conditions used have
been described elsewhere (Harlow and Lane, 1988) .
Native aoarose-xvlan gel electrophoresis. Separating
gels consisted of 0.75% agarose with or without the
inclusion of 0.75% birchwood xylan. Both the gels and the
running buffer were standard Txis-borate-EDTA buffer (TBE-pH
8.0). Approximately 40 nq cell protein was added per well.
Proteins were electrophoresed at 75 V and 21 mA in a
horizontal electrophoresis unit until the dye front reached
a point 1 cm from the end of the gel. Agarose gels were
either directly stained for activity (zymograms) or
electroblotted and analyzed immunologically as outlined
above.
Results and discussion
Deletion analysis. Exonuclease III deletions from the
3' end of xvlB resulted in the concurrent loss of both
enzymatic activities in all deletions (Fig. 9a). DNA
sequencing of the deletion end points identified a minimal
deletion of the terminal 27 base pairs (9 amino acids)
resulted in the loss of both enzymatic activities. A "TAA"
termination codon in the pUC18 polylinker immediately
downstream from this deletion served to define the new 3'
end of this mutant gene.
A LacZ1 fusion of the internal PstI fragment of xvlB
resulted in the deletion of the 5' terminal 54 base pairs
(18 amino acids) including the Shine-Dalgarno sequence and

Figure 9. Deletion analysis of xvlB gene. (A) Exonuclease III
deletion series from the 3' end of xvlB. Shaded arrow denotes
direction of transcription from the pUC18 lac promoter. Solid
arrow denotes direction of deletion. Underlined "taa"
indicates relative location of stop codon from the 31 end of
deletion. Retention or loss of respective enzyme activites is
indicated by a "+" or (B) LacZ' fusion of 5' end of the
large internal PstI fragment of xvlB resulting in a deletion
of 56 base pairs.

65
SP
pUC18
LacZ^=-
kilobase pairs
0.5 1.0 1.5
pLOI1043
H H
Sail
PstI
SphI
deletion
taa
xvl/araf
+/+ (undeleted)
-/- (27 bp./9 aa)
(56 bp/18 aa)

66
the "ATG" initiation codon with the loss of both enzymatic
activities (Fig. 9b).
In vitro mutagenesis. Twelve mutants were isolated
using selective media on the basis of reduced or abolished
enzyme activities. Two mutants could not be classified due
to multiple mutations and were not analyzed further. A
total of ten mutants were classified as negative for both
enzymatic activities against the fluorogenic substrates
based upon agar plate assays. Two mutants, number six and
number ten, displayed reduced but significant fluorescence
on fluorogenic indicator plates (Table 6).
Localization of point mutations. Most of the point
mutations were determined to reside in domains II and III
(Table 6). None of the mutations were localized in the
lacZ1 promoter of pUC18. The generation of point mutations
is consistent with the deamination activity of nitrous acid.
The point mutations were clustered in domains II and III
near the region proposed to be the active site of this
enzyme.
DNA sequencing of in vitro mutants. The amino acid
substitutions resulting from in vitro mutagenesis were
deduced from the base changes as determined by DNA
sequencing (Fig. 10). The point mutations were all AT to GC
transitions which is consistant with the mode of action for
nitrous acid mutagenesis. Two frameshift mutations were
identified, in domain I and domain V. These were not

67
Table 6. Localization of point mutations by restriction
fragment replacement analysis.
MUTANT
PHENOTYPE
PstI
EcoRI
Hindlll
LOCUS
1
-V-2
-/-
-/-
+/ +
II fs3
2
-/-
-/-
-/-
+/ +
II
3
-/-
-/-
-/-
+/ +
II
4
-/-
-/-
-/-
-/-
III
5
-/-
+/+
-/-
+/ +
I
6
w/w4
w/w
+/ +
+/ +
II
7
w/w
w/w
+/ +
+/ +
II
8
-/-
-/-
-/-
+/ +
V fs
9
-/-
-/-
-/-
+/ +
II
10
-/-
-/-
-/-
-/-
III
11
-/-
-/-
-/-
-/-
III
12
-/-
-/-
-/-
-/-
III
1 denotes presence or absence of xylosidase activity.
2 denotes presence or absence of arabinofuranosidase
activity.
3 denotes a frameshift mutation.
4 denotes "weak" activity.

Figure 10. Location and identification of point mutations in
xylB by DNA sequencing. Numbers above sequence indicate the
position of the amino acids. Amino acids in parentheses below
the sequence indicates the new amino acid inserted by
mutation. A denotes loss of enzymatic activity. A "w"
denotes weak enzymatic activity.

69
Phenotype
Region of nutation xyl/ara
180 186 192
Asp.Val.Ile.Trp.Pro.Glu.Gly.Pro.His.Leu.Tyr.Lys.Lys
(Arg) -/-
197 203 209
Tyr.Leu.Leu.His.Ala.Glu.Ala.Gly.Thr.Ser.Phe.Glu.His
(Thr)
204 210 216
Gly.Thr.Ser.Phe.Glu.His.Ala.Ile.Ser.Val.Ala.Arg.Ser
(Val)
1 7 13
Met.Val.Ile.Ala.Asn.Asn.Pro.lie.Leu.Lys.Gly.Phe.Tyr
(Leu)
172 178 184
Ala.lie.Trp.Lys.Gly.Ala.Leu.Lys.Asp.Val.lie.Trp.Pro
(Phe)
146 152 158
Val.Arg.Tyr.Asn.Gly.Asp.Trp.Glu.lie.Trp.Val.Gin.Glu
(UGA)
197 203 209
Thr.Leu.Leu.His.Ala.Glu.Ala.Gly.Thr.Ser.Phe.Glu.His
(Val)
232 238 244
Phe.Thr.His.Arg.Asn.Leu.Gly.Lys.Asp.Tyr.Pro.Val.Cys
(Asp)
204 210 216
Gly.Thr.Ser.Phe.Glu.His.Ala.He.Ser.Val.Ala.Arg.Ser
(Thr)
197 203 209
Thr.Leu.Leu.His.Ala.Glu.Ala.Gly.Thr.Ser.Phe.Glu.His
(Thr)
23
Ser.ILe.Cys.Arg.Lys.Gly...
-/-
-/-
w/w
w/w
-/-
-/-
-/-
-/-
-/-
Mutant
G186R
A203T
A210V
P7L
L178F
W158UGA
A203V
G238D
A210T
A203T
f s

70
studied further. Ten mutations are clustered in an area of
60 amino acids. Six of these mutations; glycine 185 to
arginine 185, alanine 202 to threonine 202, alanine 202 to
valine 202, alanine 209 to valine 209, glycine 237 to
aspartate 237, and alanine 209 to threonine 209 all resulted
in an inactive protein. Each of these mutations represent
nonconservative amino acid changes that would be expected to
effect the function and/or conformation of the protein.
Another mutation in this cluster, leucine 178 to
phenylalanine 178, resulted in a mutant having one-tenth the
enzymatic activities of the wild type. This is a
conservative change in that both leucine and phenylalanine
are hydrophobic amino acids and have similar structural
properties. It is possible that the larger aromatic group
on phenylalanine is affecting substrate binding and/or
catalysis. In addition, the substitution of a "UGA" stop
codon for tryptophan at position 152 also resulted in a
protein with one-tenth the enzymatic activities of the wild
type. The most probable explaination is that the "UGA"
codon is functioning as tryptophan in this E. coli strain
and the efficiency of read through is very low which yields
a reduced expression of the xvlB protein. The existence of
such suppressor mutations has been previously demonstrated
(Hirsh, 1971). One mutation, proline (7) to leucine (7)
occurred in the amino terminus of the protein and resulted
in a negative phenotype. The amino terminus of the xvlB

71
protein therefore plays an important in the structural
and/or catalytic role.
SDS-PAGE analysis of mutant proteins. The presence of
the 60,000 molecular weight monomeric subunit encoded by
xvlB was confirmed for the various mutants (Fig. 11). In
two cases, cell extracts from recombinants harboring the
proline (7) to leucine (7) and glycine (185) to arginine
(185), did not contain the xvlB-encoded protein band. The
insertion of the plasmid containing the proline (7) to
leucine (7) mutation into a lon-negative strain of E. coli.
which is deficient in serine proteases, resulted in the
restoration of the xvlB-encoded protein band on SDS-PAGE
gels. It seems likely that proteolysis of an improperly
folded protein is responsible for the absence of this
protein in cell extracts of these two mutants.
Native polyacrylamide gel comparisons of mutant and
wild type proteins. Zymograms of the wild type and mutants
L178F and W158UGA proteins indicated that the mutant
proteins are approximately the same size as the wild type
protein (Fig. 12). Western hybridizations of blotted
protein bands from native-PAGE indicated that all of the
expressed mutations result in proteins that have unaltered
electrophoretic mobilities and subunit assemblies relative
to the wild type protein (Fig. 13). This evidence suggests
that the point mutations that result in expressed protein do
not induce any destabilizing secondary structural

Figure 11. SDS-PAGE analysis of wild type and mutant
proteins. Lane assignments: A and J; wild type (pLOI1005) B;
pUC18, C; G186R, D; A203T, E; A210V, F; P7L, G; L178F, H;
W158UGA, I; molecular weight markers, K; A203V, L; G238D, M;
G238D expressed in a Ion' strain of E. coli N; A210T, 0;
A203T. Molecular weight marker sizes (X 103) : 1; 94, 2; 67,
3; 43, 4; 30, 5; 20.

o n mi h m y as
ZL

Figure 12. Native-PAGE comparison of W158UGA and L178F
mutants proteins with the wild type (pLOI1005) stained with
Coomassie blue. Lane assignments: A; mutant L178F, B; pUC18
control, C; W158UGA, D; pLOI1005.

LO

Figure 13. Western hybridization of native-
and mutant proteins. Lane assignments: A;
C:; P7L, D; A210V, E; L178F, F, W158UGA, G;
PAGE of wild type
A203T, B; pUC18,
PLOI1005.

77

78
perturbations that would result in proteolysis by the cell
or a major change in tertiary structure of the proteins.
The reduced levels of the xylosidase protein in the
W158UGA mutant correlates to the reduced enzymatic
activities for recombinants harboring this mutant. The
"UGA" termination codon can be decoded as a tryptophan at
low efficiency in E. coli (Hirsh, 1971). It is likely that
this is the case also for the W158UGA mutant in xvlB.
Substrate-binding comparisons of mutant proteins.
Electrophoresis of the wild type and mutant proteins on
native gels that contained agarose alone and agarose plus
birchwood xylan indicated a differential mobility between
the wild type and mutant proteins (Fig. 14). Using agarose
alone no differences between electrophoretic mobilities of
the respective mutant proteins and the wild type were
detected using Western hybridization analysis. The
inclusion of birchwood xylan (0.75 %) resulted in a change
in the mobility of the proteins. Without exception, all the
mutant proteins exhibited faster electrophoretic mobilities
relative to the wild type protein. It is possible that the
xylan is functioning as a psuedo-substrate and the point
mutations have affected the relative affinities of these
proteins for the substrate.

Figure 14. Substrate binding native gel Western hybridization
assays of wild type and mutant proteins. Lane assignments: A
, F, and G; native (pLOI1005), B; G186R, C; A203T, D; W158UGA,
E; L178F, G; A203V, H; G238D, I; A210T, J; A203T. Arrow
indicates the direction of protein migration. The "+" and
indicated the relative location of the anode and cathode,
respectively.

80

81
Analysis of enzymatic activities of expressed mutant
proteins. In all cases, the point mutations affected both
enzymatic activities concurrently (Table 7). All mutations
resulted in an expressed phenotype in which enzymatic
activities were reduced or abolished. The clustering of
these mutations in the 60 amino acid region (12% of the
coding region) which contains the catalytic consensus
sequence is evidence that the two enzymatic activities
expressed by this protein are not functionally confined to
separated domains. There is a dependance of function
relating both enzymatic activities to this region of the
protein.
The effects of substrate concentration on reaction rate
of the mutant enzymes was investigated using the crude
extracts as a source of protein (Table 8). Mutants A203T,
A210V, L178F, G238D, and A210T exhibited an increase in
apparent increase in reaction rate relative to increasing
substrate concentration. Increasing substrate concentration
had no effect on reaction velocity for mutants G186R and
A203V, however the activities for these two mutant proteins
were above that for the pUC18 background. It is possible
that the lowest concentration of substrate used, 3 mM, is at
saturation with respect to these two mutant proteins.

82
Table 7. Enzymatic activities of recombinants harboring
point mutations on xvlB relative to the wild type protein.
Clone
Xyl Sp.Ac.a
Ara Sp.Ac.
ara/xylb
pLOI1005
10.5
17.4
1.7
.46
G186R
0.1
0.1
1.0
.51
A203T
0.1
0.3
3.0
.49
A210V
0.2
0.2
1.0
.49
P7L
0.2
0.2
1.0
N/D
L178P
1.2
1.8
1.5
.53
W158AUG
2.1
2.9
1.4
.46
A203V
0.2
0.2
1.0
.49
G238D
0.2
0.2
1.0
N/D
A210T
0.2
0.1
0.5
.50
pUC18
0.0
0.0
-
a Specific activity in nmoles p-nitrophenol released per min
per mg protein.
b Ratio computed after subtraction of pUC18 background values.
c Determined by comparing relative migration distances of each
protein verses that for the dye front on agarose/xylan native
gel electrophoresis using "Western" hybridization to visualize
protein bands.

83
TABLE 8. Effects of substrate concentration on B-D-
xylosidase activity for xvlB in vitro mutants6.
Clone
3 mM
6 mM
9 mM
Kin c
app
p-nitrophenyl
-B-D-
xylopyranoside
wild type
18.6
24.0
25.6
4 mM
G186R
0.4
0.4
0.5
2 mM
A203T
0.4
0.7
0.9
11 mM
A210V
0.8
0.9
1.2
2 mM
L178F
1.4
1.8
3.0
18 mM
A203V
0.6
0.8
0.8
3 mM
G238D
0.6
1.1
1.3
7 mM
A210T
0.4
0.7
0.8
12 mM
a Specific activity expressed as nmoles p-nitrophenol
released per min per mg protein.
b Does not include the frameshift mutations or those mutant
proteins that are not expressed.
c Apparent Km values determined using the direct linear
method of Cornish and Bowden.

84
Conclusions
The mutation and deletion data are consistant with the
proposal that both enzymatic activities exhibited by the
xvlB-encoded protein are not functionally separate but
depend upon the same region of the protein for complete
activity. The clustering of mutations about the consensus
sequence and the 60 amino acids region is strong evidence in
favor of this region being important for substrate-binding
and/or catalytic activity of the protein. This region of
the protein is rich in aspartic acid, glutamic acid and
histidine residues which have been previously implicated in
the catalytic function of related proteins such as lysozyme
(Quiocho, 1986), taka-amylase from Aspergillus nicer
(Matsuura et al., 1984), cellobiohydrolase II from
Trichoderma reesei (Rouvinen et al., 1990), and a
neopullulanase from Bacillus stearothermophilus (Kuriki et
al., 1991).
The point mutations that were expressed as complete
proteins did not affect either subunit assembly or the
apparent size of the native protein relative to the wild
type. These mutations did, however, change the protein's
mobility during electrophoresis on agarose-xylan native
gels. It is possible that this change in mobility is due to
a reduced affinity of the various mutant proteins for xylan,
which is functioning as a surrogate substrate for the

85
enzyme. There appears to be a dependence of velocity on
substrate concentration for at least some of the mutants.
The mutation evidence supports the hypothesis that one
active center or domain is responsible for both enzymatic
activities. It does not, however, totally rule out the
possibility that the protein may contain separate catalytic
sites or subsites which are spacially close together.
Future kinetic experiments, including the investigation of
substrate competition between arabinofuranosides and
xylopyranosides, will allow further definition of the
catalytic regions responsible for both activities.

Chapter V
Purification and characterization of the
xvlB-encoded protein
Introduction
The existance of polysaccharide-hydrolyzing enzymes
having broad substrate specificities is well documented
(Ward and Moo-Young, 1989). One such example includes the
exoglucanase, EXG, produced by Cellulomonas fimi (Beguin,
1991). This enzyme also exhibits B-D-xylanase activity. The
xylanase from Clostridium thermocellum (XYNZ) also exhibits
endo-glucanase activity towards carboxymethylcellulose
(Grepinet et al. 1988). In particular, several xylanase and
xylosidase enzymes have been characterized which exhibit
substrate ambiguity (Flint et al. 1989) A recent example
is a fi-D-xylosidase that was cloned from Caldocellum
saccharolvticum that also exhibits endoxylanase activity
(Luthi et al. 1990). Enzymes which exhibit both
endoxylanase and B-glucosidase activities have been shown to
be fairly common among micoorganisms (Gilkes et al. 1991).
Substrate ambiguity among the glycohydrolases has been
attributed to the similarities between the various
substrates involved. Upon closer examination, this
phenomenon is not totally unexpected. The B-(1,4)-xylosidic
bonds of xylan and the B-(1,4)-glycosidic bonds of cellulose
86

87
are structurally related and have quite similar molecular
configurations about the B-(l,4) bonds with respect to the
hydroxyl group on the a-carbon being in the axial or
equatorial positions.
Some true bifunctional cellulases have been
demonstrated which contain separate active sites for each
enzymatic activity. Saul et al. (1990) isolated a cellulase
from Caldocellum saccharolvticum which exhibited both endo-
glucanase and exo-glucanase activities. These authors used
DNA sequence homology comparisons and deletion analysis to
demonstrate that the endoglucanase activity was located in
the carboxy terminal domain and the exoglucanase activity
was located at the amino terminal domain.
In an earlier study purified a B-xylosidase from
Chaetomium trilaterale that also exhibited S-glucosidase
activity (Uziie et al. 1985). These investigators used
kinetic analysis employing substrate competition and
inhibitors to demonstrate that a single active site was
responsible for both enzymatic activities. It was also
suggested that two kinetically separate substrate binding
sites may reside in the active center of this enzyme.
In the previous chapter mutational analysis
demonstrated that the two enzymatic activities encoded by
xylB were not functionally separate but both appeared to be
catalytically dependent upon the same region of the protein.
The proposal that a single active center is responsible for

88
both enzymatic activities in this enzyme was tested using
analogous kinetic experiments including substrate inhibition
and competition of the enzyme with respect to both the
xylopyranosyl and arabinofuranosyl substrates.
Materials and Methods
Medium and growth conditions. Medium and growth
conditions were described in chapter III.
Preparation of cell extracts. Cell extracts were
prepared as described in chapter III.
Partial purification of fl-xvlosidase by preparative
electrophoresis. Extracts containing the total cytoplasmic
proteins from pLOHOOS or L178F recombinant clones were used
as a source of xylosidase. Proteins were fractionated in an
8% native polyacrylamide gel in a BioRad-Prep Cell
preparative electrophoresis system. Gel and buffer
formulations were: separating gel buffer; 240 mM Tris (pH
8.48), stacking gel buffer; 40 mM Tris (pH 6.9), lower tank
buffer; 63 mM Tris/50 mM HCL (pH 7.5), upper tank buffer; 38
mM Tris/ 40 mM glycine (pH 8.9). A total of 50 mg/ml
protein was loaded onto the gel. Electrophoresis was done
at constant power of 31 W. Starting conditions were 250 V
and 40 mA. Protein elution was monitored at 280 nm.
Fractions were collected and assayed for xylosidase
activity as described below. The most active fractions were
pooled and concentrated using an Amicon Centriprep
concentrator (Amicon Division, Danvers, MA). Xylosidase in

89
the concentrate was then precipitated by the addition of
solid ammonium sulfate to 70% saturation. The enzyme was
stored as an ammonium sulfate pellet at 5C until needed.
No loss of enzymatic activity was detected after storage for
one week under these conditions. Pellets were resuspended
in 5 mM phosphate (pH 7.0) containing 10 mM R-
mercaptoethanol immediately prior to use.
Hydrophobic interaction chromatography. The
xylosidase-containing pellet from preparative
electrophoresis was resuspended in 1 ml of 1.7 M (NH4)2S04 in
5 mM phosphate buffer (pH 6.8) and loaded onto a 2.0 x 3.0
cm Pharmacia XK 16/20 chromatography column (Pharmacia LKB,
Uppsala, Sweden) packed with Toyopearl "TSK-Gel" hydrophobic
gel (Supelco, Inc., Bellefonte, PA.). The column was
equilibrated with 1.7 M (NH4)2S04 in 5 mM phosphate buffer
(pH 6.8) prior to the addition of sample. Xylosidase was
eluted using a linear negative salt gradient starting with
1.7 M (NH4)2S04 down to zero in a total volume of 200 ml.
Fractions were collected in 3 ml volumes and analyzed as
described below.
Enzyme assays. Xylosidase activity in each fraction
was assayed using p-nitrophenyl-B-D-xylopyranoside (p-NP-X)
at a final concentration of 2.5 mM unless otherwise noted
and in 50 mM phosphate buffer (pH 6.8) at 37C. Assays were
done in a total volume of 1 ml and allowed to continue until
the yellow color indicating enzyme activity was detected.

90
Assays were then terminated by the addition of 2 ml 0.5 M
carbonate. The p-nitrophenol released by hydrolysis was
measured spectrophotometrically at 405 nm. The liberation
of 1 nmole of p-nitrophenol results in an increase in
absorbance of 0.0184 at 405 nm. Samples were assayed for a-
L-arabinofuranosidase activity under the same conditions
using p-nitrophenyl-a-L-arabinofuranoside (p-NP-A) as a
substrate.
Electrophoretic analysis of proteins. SDS-PAGE,
native-PAGE, activity stains, and "Western" hybridizations
were done as described in chapter III.
Optimum activity pH. The optimum pH for both
activities was determined in duplicate using citrate (40
mM)-sodium phosphate (80 mM) buffer in the pH range of 3.0
to 7.4. Tricine (50 mM) was used for the 7.5 to 8.5 pH
range and Bicine (50 mM) for the 8.6 to 9.0 pH range. All
activities were determined using a 6 mM final concentration
of p-NP-X or p-NP-A at 37C.
Thermal optimum and thermal inactivation
determinations. Thermal stability was determined by
incubating the enzyme for 30 min at 10, 25, 35, 45, 55, and
65C. After 30 rain the enzyme was placed in ice for 10 min
prior to assaying both enzymatic activities. The optimum
temperature for activity was determined by assaying the
enzyme for both activities at 10, 25, 35, 45, 55, and 60C
for 10 min. Assays were done in duplicate at 37C using 6

91
mM final concentration of p-NP-X or p-NP-A in citrate (40
mM)-sodium phosphate (80 mM) buffer at pH 6.0.
Determination of Km and Vmax for both enzymatic
activites. Both p-NP substrates were tested at various
concentrations to determine the Km and Vmax values with
respect to both substrates. All assays were done in
duplicate at 37C and at pH 6.0. Kinetic parameters were
graphically determined using the Lineweaver-Burk and direct
linear methods.
Substrate competition experiments. The analogue B-
methyl-D-xylopyranoside was used as an inhibitor for cr-L-
arabinofuranosidase activity. The fluorogenic substrate 4-
methylumbelliferyl-a-L-arabinofuranoside was used as an
inhibitor for B-D-xylosidase activity.
Results and discussion
Purification of xvlosidase. The xylosidase-containing
fractions were separated as one large peak during
preparative electrophoresis (Fig. 15). This peak consisted
of 15, 3 ml fractions which contained both xylosidase and
arabinofuranosidase activities. The fraction which
corresponded to the middle eight fractions exhibited the
highest activities and were pooled, concentrated, and
precipitated with ammonium sulfate. The xylosidase was
further purified by hydrophobic interaction chromatography
and was separated as a single peak consisting of six 3 ml
fractions (Fig. 16). The four most active fractions were

Figure 15. Elution profile of the xvlB gene product during
preparative electophoresis on BioRad Prep Cell system.

-Nitrophenol Released (nmoles)
93
250
200
0 50 100 150 200
Elution Volume (ml)
250

Figure 16. Elution profile of the xvlB gene product during
hydrophobic interaction chromatography.

95
Elution Volume

96
pooled and concentrated by ultrafiltration. A summary of
this purification scheme is shown in Table 9.
SDS-PAGE analysis of the pooled fractions from all the
purification steps showed the xylosidase to be approximately
90% pure with low levels of contaminating proteins (Fig.
17) .
Active enzyme conformation. Native-PAGE analysis of
the enzyme using Coomassie blue, activity stains, and
Western hybridization to visualize the protein indicated
only one predominant form of the xylosidase was active (Fig.
18). This active band had an apparent molecular weight of
120,000 which corresponds to the dimeric form of the enzyme.
Enzyme optima. The temperature activity profiles for
both activities were also essentially the same with 45C
being the optimum temperature for both the xylosidase and
arabinofuranosidase activities (Fig. 19). The thermal
inactivation profiles for both activities in the wild type
protein were essentially the same with activity diminishing
rapidly after incubation for 30 min at 35C (Fig. 20) .
The pH profiles for both activities in the wild type
protein showed marked differences (Fig. 21). The xylosidase
activity had a sharp peak at pH 6.0 which was followed by a
rapid decline in activity with less than 50% relative
activity remaining at pH 6.8. Arabinofuranosidase activity
peaked at pH 6.0 but remained at 90% relative activity up to
pH 9.0.

97
TABLE 9. Purification of the B. fibrisolvens xvlB encoded
xylosidase from E. coli DH5a (pLOI1005).
Fraction
Vol
(ml)
Protein
(mg)
Activity
(mU)a
Sp Act
(mU/mg)
Yield
(%)
Crude
extract
2.0
100
4600
46
100
Prep
Cell
7.0
3.4
1068
314
23
Hydrophobic
column
3.0
0.9
587
652
13
3 Specific activity expressed as nmoles p-nitrophenol
released per min per mg protein for B-D-xylosidase
activity only.

Figure 17. SDS-PAGE analysis of pooled xylosidase-containing
fractions from preparative electrophoresis on 8% native PAGE:
A and E; molecular weight markers, B; hydrophobic column
purified preparation, C; Prep-Cell purified preparation, D;
crude extract.

J>
§
¡I r
) i i I

Figure 18. Native-PAGE analysis of the -D-xylosidase. A;
Coomassie blue stained protein bands, B; Western hybridization
of electroblotted native-PAGE gel, C; activity stain
(xylosidase activity/arabinofuranosidase activity). Arrows
indicate approximate region where the apparent dimeric and
monomeric forms of the enzyme would be located.

101

Figure 19. Thermal inactivation profile of xylosidase (closed
circle) and arabinofuranosidase (open circle) activities.

103
o
<
Q>
O
o>
o:
o>
o
o>
Q-
0
30 40
Temperature ( C)
10
20
50
60
70

Figure 20. Temperature optimum profile for xylosidase (closed
circle) and arabinofuranosidase (open circle).

105
0 51015 20 25 30 35 40 45 50 55 60 65 70
Temperature ( C)

Figure 21. pH profiles for xylosidase (closed circles) and
arabinofuranosidase (open circles) activities.

Percent Relative Activity
f'O 4*. CT> CO O
O O O O CD O
107

108
Kinetic properties. A characteristic Michaelis-Menten
relationship was observed for the native protein using p-NP-
X and p-NP-A as substrates. The Km and Vmax values for
xylosidase and arabinofuranosidase activities were 3.8 mM ,
1.7 mM and 1,111 nmoles min'1 mg'1 protein, 833 nmoles min'1
mg'1 protein, respectively (Fig. 22). These values give a
Kcat or turnover number per subunit of 1111 sec'1 for
xylosidase activity and 833 sec'1 for arabinofuranosidase
activity.
Substrate competition experiments. The florogenic
substrate 4-methylumbelliferyl-B-D-arabinofuranoside proved
to be a potent competitive inhibitor of S-D-xylosidase
activity when p-NP-X was the substrate (Fig. 23).
Alternatively, fi-methyl-D-xylopyranoside, a xylobioside
analogue, was a competitive inhibitor of a-L-
arabinofuranosidase activity when p-NP-A was the substrate
(Fig. 24).
In separate experiments, the enzymatic activities were
measured with the two substrates, p-NP-X and p-NP-A
individually and together at 10 mM (saturation)
concentration at 37C for 10 min. Xylosidase activity alone
was 1160 mU, and arabinofuranosidase activity alone was 809
mU. When both substrates were combined in the same assay
the resultant activity was 948 mU which is essentially the
average of the two activities separately. These individual
activities are not additive when both substrates are present

Figure 22. Double recipricol plots of xylosidase and
arabinofuranosidase activities. Xylosidase; closed circles,
arabinofuranosidase; open circles.

l/*(iu)
no
1/[S] (mU)

Figure 23. Competitive inhibition of xylosidase activity by
4-methylumbelliferyl-Q!-L-arabinofuranoside: closed circles; no
inhibitor, open circles; 100 ;M inhibitor, open squares; 250
/LtM inhibitor.

112
1 /[ S] (mM)

Figure 24. Competitive inhibition of arabinofuranosidase
activity by fi-D-methylxylopyranoside: closed circles; no
inhibitor, open circles; 1 mM inhibitor, closed squares; 2 mM
inhibitor.

114
1/[S] ( mM)

115
together in the assay system. An additive result would have
indicated that separate active sites might be present for
the two enzymatic activities. This data is consistent with
the results of the competitive inhibition experiments and
demonstrate that both substrates are competing with each
other for the same catalytic center on the enzyme.
Conclusions
The xvlB gene encodes an enzyme that exhibits substrate
ambiguity with respect to the two p-nitrophenol-derived
xylopyranosides and arabonofuranosides. The activities
against these two substrates co-purified during preparative
electrophoresis and were both stable as ammonium sulfate
pellets for up to one week at 5C. Only a single active
form of this enzyme is present on native-PAGE gels and
appears to have a molecular weight of 120,000.
The thermal stability and the optimal temperature were
essentially the same for both activities with activities
being stable up to 42C after 30 min and the optimal
temperature was 45C. The thermal stability of this enzyme
is relatively low when compared to xylosidases purified from
other microorganisms. The temperature optimum for this
enzyme is similar to xylosidases produced by Clostridium
acetobutylicum (Lee and Forsberg, 1987) and Bacillus pumilus
(Panbangred et al. 1984).
Investigations of optimal pH revealed a major
difference with respect to both activities. Both activities

116
exhibited an optimum pH of 6.0 which compares with that of
the C. acetobutvlicum xylosidase. There are four histidine
residues in the proposed active center for this enzyme.
Since the pKa of histidine is 6.5, it is likely that one or
more of these residues are involved in both catalytic
activities.
The proposed mechanism for glycohydrolase, and the
related glycosyltransferase catalysis, involves
contributions from a hydrogen donor and a hydrogen acceptor
moiety (Mooser et al. 1991). In the usual scheme, an
aspartic or glutamic acid in deprotonated form serves as a
hydrogen acceptor and stabilizes the hydrolytic
intermediate. A general acid donates a proton to the
glycosidic oxygen to facilitate the cleavage of the
monomeric glycoside. The pH dependence of the xvlB-encoded
protein with respect to the 6-D-xylosidase activity is
probably due to one or more histidine residues serving as
the general acid.
The broad pH activity profile for the
arabinofuranosidase activity is uncommon. At the higher pH
ranges, histidine is completely deprotonated and could no
longer serve as a general acid and donate a proton. It is
possible that the arabinofuranoside substrate, once bound in
the active pocket, protects the histidine from further
deprotonation at high pH. This charge transfer phenomenon
has been demonstrated to occur in other enzymatic systems

117
(Robson and Gamier 1988). The binding of the
arabinofuranoside may also induce a conformational change in
the enzyme which somehow leads to the protection of the
histidyl moiety from deprotonation at high pH ranges.
It should be noted that the effective pKa of these amino
acids can vary markedly from the intrinsic pKa due to the
electonic environment in the active center of the protein.
The possibility that a different amino acid is
contributing a proton at the higher pH ranges cannot be
ruled out. Tyrosyl residues (pKa 10) in this region of the
protein could possibly substitute for histidine as a general
acid at the higher pH ranges. This phenomenon might be
explained if the two activities were catalyzed by different
mechanisms that in turn involve different amino acids
present in the same active center. This was shown to be the
case with the neopullulanase from Bacillus
stearothermophilus in which two separate enzymatic
activities were found to reside in the same active center
but were catalyzed by different amino acids and thus
different mechanisms (Kuriki et al. 1991). In an earlier
work, The B-D-xylosidase from Bacillus pumilus was found to
hydrolyze a-D-xylosylfluoride by a mechanism that was
separate from that for the hydrolysis of B-D-xylosylfluoride
(Kasumi et al. 1987). The number and nature of the active
centers is not known for this protein.

118
The kinetic properties of the xvlB-encoded enzyme with
respect to the xylosidase activity is similar to the
xylosidase enzymes from C. acetobutvlicum and B. pumilus
which have Km's for p-NP-X of 3.8 mM, 3.7 mM, and 2.4 mM,
respectively. The Km for the arabinofuranosidase activity
against p-NP-A was 1.7 mM which compares favorably with 1.3
mM for the a-L-arabinofuranosidase isolated from another
rumen bacteria, Ruminococcus albus (Greve et al. 1984) The
maximum velocity values for the xylosidase activity of the
xvlB-encoded protein is approximately tenfold lower than
that for the C. acetobutvlicum enzyme and approximately one-
half that for the B. pumilus enzyme. Lee and Forsberg
(1987) purified and characterized an a-L-
arabinofuranosidase from C. acetobutvlicum ATCC 824. This
enzyme exhibited Km and Vm values of 4 mM and 3 6.4 mole
min'1 mg'1 protein respectively against p-nitrophenyl-a-L-
arabinofuranoside. It had a pH optimum of 5.0 5.5 and
exhibited no activity against other p-nitrophenylglycosides.
High catalytic efficiency is usually associated with highly
evolved enzymes having a narrow substrate specificity
(Robson and Gamier, 1988) It is possible that the xvlB-
encoded enzyme has sacrificed catalytic efficiency for a
broader substrate range since neither the C. acetobutvlicum
or the B. pumilus xylosidase enzymes have additional
activities.

The substrate competition experiments, using model
substrates, unequivicaly demonstrated that xylopyranoside
and arabinofuranoside are competitive inhibitors of each
other.

CHAPTER VI
PARTIAL PURIFICATION AND CHARACTERIZATION OF
THE L178F MUTANT PROTEIN
Introduction
As outlined in chapter IV, point mutations were
introduced into the xvlB gene using in vitro mutagenesis.
All the mutations resulted in proteins having a reduction or
loss of both enzymatic activities concurrently.
Additionally, 10 of these point mutations were clustered in
a 60 amino acid region of the protein. Native agarose-xylan
gel electrophoresis of the expressed wild type and mutant
proteins revealed that the mutant proteins bind to the
surrogate substrate xylan with less affinity relative to the
native protein indicating an apparent increase in the Km
values for these mutant enzymes. The L178F mutant appeared
to have the least affinity (highest apparent Km) for the
surrogate substrate.
To further investigate the possibility of altered Km
due to the L178F mutantion, the mutant protein was partially
purified and characterized.
120

121
Materials and methods
All methods used to purify and characterize the L178F
mutant enzyme were essentially the same as those described
for the native enzyme in Chapter V.
Results and Discussion
Partial purification of L178F protein. The elution
profile for the mutant protein was essentially the same as
that for the wild type protein, have a single broad peak
consisting of 15, 3 ml fractions (Fig. 25). The fractions
which corresponded to the middle six fractions exhibited the
highest relative activity and were pooled, concentrated, and
precipitated at 70 % saturation with ammonium sulfate. SDS-
PAGE analysis of the partially purified protein indicated it
was approximately 80 % to 90 % pure with low levels of
contaminating proteins (Fig. 26).
Enzyme optima. The mutant protein exhibited a higher
thermal stability relative to the wild type (Fig. 27). This
protein was stable up to 55C for 30 min and still retained
100 % relative activity with respect to both substrates.
Since the replacement of a phenylalanine for leucine
introduces a more hydrophobic residue at this position, it
is possible that this strengthened the hydrophobic
interactions at the core of the protein and thereby
increased the thermal stability. The optimal temperature
was not investigated.

Figure 25. Elution profile of the xylB-encoded protein
harboring the L178F mutation during preparative
electrophoresis:

123

Figure 26.
SDS-PAGE analysis of partially purified L178F
mutant protein: A; Prep-Cell purified
preparation, B; crude extract, C; molecular
weight markers.

125

Figure 27.
Thermal inactivation profiles of xylosidase and
arabinofuranosidase activities of the L178
mutant protein: open squares; xylosidase
activity, closed squares; arabinofurnanosidase
activity.

127

128
The pH profile for both enzymatic activities for the
mutant protein was essentially the same as that for the wild
type protein (Fig. 28). Xylosidase activity had a sharp
activity peak at pH 6.0 which was followed by a rapid
decline. Arabinofuranosidase activity also peaked at pH 6.0
but remained at 90 % relative activity up to pH 9.0.
Kinetic properties. A characteristic Michaelis-Menten
linear relationship was observed for the mutant protein with
respect to both substrates (Fig. 29). The Km values for the
xylosidase and arabinofuranosidase activities were 16 mM and
33 mM respectively. This represents a decrease in the
affinity of the enzyme for both substrates. The increase in
Km agrees with the data in chapter IV using the agarose-
xylan gel binding assays.
Conclusions
It is apparent that the replacement of leucine with
phenylalanine at position 178 decreases both the affinity of
the enzyme for both substrates and the catalytic efficiency
of the enzyme. This amino acid substitution also increases
the thermal stability of the xvlB protein. It is possible
that these three properties are related. The increase in
thermal stability could represent a general increase in the
stability of the enzyme by strengthening hydrophobic
interactions at the core of the protein.
An increase in the overall stability of the protein
could influence the plasticity of the active center. The

Figure 28. pH activity profiles of xylosidase and
arabinofuranosidase activies of the L178F mutant
protein: open squares; arabinofuranosidase
activity, closed squares; xylosidase activity.

130
pH

Figure 29.
Double recipricol plots of xylosidase and
arabinofuranosidase activities for the L178F
mutant protein: open squares;
arabinofuranosidase activity, closed squares
xylosidase activity.

132
1/[S] ( mM)

133
"induced-fit" model of enzyme catalysis proposes that the
three-dimensional conformation of an enzyme will change as
it binds to the substrate to allow closer contacts to be
made between the substrate and the catalytically important
functional groups (Koshland, 1966). Increasing the
hydrophobic interactions in the core or active center of the
protein might also decrease the ability of the protein to
change conformation upon binding substrate. This would lead
to a loss of catalytic efficiency and possibly affect
substrate binding.
A more general explanation of the effects of the L178F
mutation would be that the protein is simply folded
incorrectly and this results in a less accessible active
center which is reflected in the increase in Km for p-NPX
and p-NPA in this mutant protein.
The addition of a large aromatic R-group by the
substitution of phenylalanine for leucine could also
introduce a steric hindrance factor which prevents optimal
enzyme-substrate interactions and therefore decreases the
affinity and/or catalytic efficiency of the enzyme. This
assumes, however, that the amino acid substitution has
occurred in an area within or near the active site.

134
Based upon the data in this chapter and in chapter IV
it is reasonable to conclude that the L178F mutation is in
an area of the xvlB protein that is important in substrate
binding and possibly catalysis with respect to both
enzymatic activities.

CHAPTER VII
SUMMARY AND GENERAL CONCLUSIONS
The studies presented here have characterized the B-D-
xylosidase from the rumen bacteria Butvrivibrio fibrisolvens
using genetic and biochemical techniques. This enzyme is
important in the final steps of depolymerization of
hemicellulose in that the products of hydrolysis, usually
monomers, are used directly in the metabolism of the
bacterium. This enzyme is particularly interesting in that
it exhibits substrate ambiguity. Both B-D-xylopyranosides
and a-L-arabinofuranosides are hydrolyzed by this enzyme.
Previous studies involving B-D-xylosidase enzymes from
other organisms have shown that substrate ambiguity among
the xylosidase and related enzymes is not universal. The
xylosidase from Bacillus pumilus only exhibits activity
against aryl-B-D-xylopyranosides (Panbangred et al. 1983) .
The same is true for the B-D-xylosidase from B. subtilis
(Paice et al., 1986). Notably a B-D-xylosidase from
Caldocellum saccharolvticum was cloned and sequenced that
also exhibited endoxylanase activity (Luthi et al. 1990).
In this case the two substrates are structurally identical
but vary in size with the endoxylanase acting on the longer
chain length polymers while the xylosidase is specific for
135

136
the shorter hydrolytic products of the endoxylanase
activity. No data is yet available concerning active site
studies of this enzyme as to whether or not one or two
catalytic centers are responsible for this phenomenon.
The present study is unique in that it examines the
substrate ambiguity of the B. fibrisolvens enzyme using
both genetic and biochemical techniques. This work
represents a complete study of the enzyme and has elucidated
a number of important aspects of the B-D-xylosidase from B.
fibrisolvens. (A) The gene that encodes this enzyme in B.
fibrisolvens. xvlB is present as a single copy in the
chromosome (Fig. 3). Any additional xylosidase genes
present in this organism must share limited homology with
xvlB. (B) The DNA sequence is 1,551 bp in length, encodes
517 amino acids and is located between two additional large
open reading frames (ORF's) each in excess of 1,000 bp (Fig.
5). The size of this enzyme is essentially the same as the
B. pumilus xylosidase. (C) No stem loop or rho-independant
teminators were identified between any of the three ORF's
and the expression data indicates that this region does not
function as a terminator in E. coli (Table 2). This result
suggests that these three ORF's represent part of a xylan-
degrading operon in B. fibrisolvens. The subcloning and
expression data also demonstrate that the single xvlB gene
is responsible for both the xylosidase and
arabinofuranosidase activities (Fig. 4b). (D) The codon

137
bias of these three ORF's is consistent with the low guanine
plus cytosine ratio of the DNA from this organism (Table 1).
(E) The xvlB-encoded protein shares limited amino acid
identity with other published amino acid sequences of B-D-
xylosidase enzymes and related proteins (Table 4). This is
an indication that the xvlB gene from B. fibrisolvens is
evolutionarily divergent from these other genes. (F) A
consensus sequence has been identified in xvlB that has
significant identity to an amino acid sequence that has been
previously implicated in catalytic function of other
glycohydrolases (Table 5).
Mutational analysis of the xvlB-encoded protein has
revealed several important structural and functional
relationships. (A) All point mutations within the xvlB gene
resulted in a protein in which both enzymatic activities
were reduced or abolished (Table 6). (B) All point
mutations that were isolated are localized or clustered in
single region of the protein that is near the proposed
catalytic center (Fig. 10). (C) All but two of the point
mutations result in a stably expressed protein (Fig. 11).
(D) All of the point mutations that result in expressed
protein exhibit a decrease of affinity in native gel assays
(Fig. 14) and an increase in apparent Km. These results
demonstrate that the catalytic center for these two
enzymatic activities is not functionally independent but is
Idealized on the same region of the xvlB-encoded protein.

138
Finally, a biochemical characterization of the purified
enzyme was initiated to examine whether or not a single
active center was responsible for both enzymatic activities.
The partially purified enzyme exhibited thermal inactivation
and temperature optimum profiles that were essentially the
same for both activities (Figs. 19 and 20).
The pH optima for both activities were essentially
the same however the activity range with respect to pH
differed for the two activities (Fig. 21). Several
hypotheses may be invoked to explain this difference in pH.
Hydrolytic mechanisms usually involve the contributions of a
proton donor and a proton acceptor to the catalysis (Knowles
et al. 1988). The proton donor acts as a general acid in
donating a proton to the glycosidic bond oxygen generating
an oxycarbonium ion intermediate. The proton acceptor
serves to stabilize the hydrolytic intermediate and acts as
a general base. At pH 6.0, the pH optimum for this enzyme,
carboxyl groups on aspartic and glutamic acid residues are
fully charged. At this pH a histidinyl residue (pKa 6.5)
may serve as a proton donor in that it is still protonated.
At pH values above 6.0 the xylosidase activity diminishes
reflecting the deprotonation of histidinyl residues in the
active center demonstrating the dependence of xylosidase
activity on this amino acid.
The relatively high arabinofuranosidase activity at
elevated pH may indicate that this catalytic activity

139
proceeds by a mechanism different from that of xylosidase
activity. This phenomenon, involving separate hydrolytic
mechanisms, was shown to exist with the Bacillus pumilus
xylosidase (Kasumi et al. 1987) These investigators
demonstrated kinetically that the enzyme was able to
catalyze the hydrolysis of a- and B-D-xylosylfluoride by a
mechanism entirely different from that for the hydrolysis of
p-nitrophenyl-B-D-xylopyranoside. It was postulated that
separate catalytic groups were responsible for the two
mechanisms. In a more recent study Kuriki et al. (1991)
demonstrated that a neopullulanase from Bacillus
stearothermophilus. exhibited separate a-1,4 and a-1,6 bond
cleavage activities. These investigators used site-directed
mutagenesis to demonstrate that the two activities were
catalyzed by separate amino acids in the same active center.
It is possible that an analogous process occurs in the
xvlB-encoded protein. A tyrosinyl residue (pKa 10.0) could
serve as a general acid and donate a proton at the pH range
where the arabinofuranosidase activity is present. There
are five tyrosine residues in the proposed active center
which may be involved.
Another explanation involves the relative stabilities
of the two substrates at elevated pH. If the
arabinofuranoside is less stable at higher pH values
relative to the xylopyranoside, it follows hydrolysis will
proceed more easily for the former and therefore requires

140
less of proton donating potential from involved hisidinyl
residues. This explaination may also be invoked to support
the possbility that separate amino acids and/or mechanisms
are involved in the two separate activities.
Also relating to substrate effects is the concept of
substrate-assisted catalysis (Carter and Wells, 1990).
Substrate-assisted catalysis was first demonstrated with
proteolytic enzymes. In the enzyme subtilisin, a serine
endopeptidase, it was demonstrated that a histidine in the
protein substrate could replace a catalytic histidine in the
active site of subtilisin that had been mutated to an
alanine by site-directed mutagenesis. It is possible that a
glycosyl carboxylate could serve as an alternate proton
acceptor under certain conditions.
The kinetic constants for the two activities were
comparable to other xylosidase and arabinofuranosidase
enzmes that have been reported. The catalytic efficiency or
Kcat for this enzyme is not as high as that reported for
other xylosidase or arabinofuranosidase enzymes. Since both
arabinofuranosidase and xylopyranosidase are substrates that
are indigenous to hemicellulose, this protein appears to
have saccrificed catalytic efficiency for a broader
substrate specificity. This may represent an evolutionary
adaptation to the highly competitive rumen environment.

141
The substrate competition experiments corroborate the
genetic evidence and demonstrated that the two activities
reside in the same active center of the xvlB-encoded enzyme.
This study demonstrates that the xvlB gene from
Butvrivibrio fibrisolvens encodes a single protein having B-
D-xylosidase and a-L-arabinofuranosidase activities. These
activities are localized in the same active site of the
protein.

LITERATURE CITED
Allf-Steinberger, C. 1984. Evidence for coding pattern on
the non-coding strand of the Escherichia coli genome.
Nucleic Acids Res. 12:2235-2241.
Armstrong, D.G., and H.J. Gilbert. 1985. Biotechnology and
the rumen: A mini-review. J. Sci. Food Agrie.
36:1039-1046.
Barnett, C. C., R. M. Berka, and T. Fowler. 1991. Cloning
and amplification of the gene encoding an extracellular B-
glucosidase from Trichoderma reesei : evidence for improved
rates of saccharification of cellulosic substrates.
Biotechnol. 9:562-567.
Bastawde, K. B., L. B. Tabatabai, M. M. Meagher, M. C.
Srinivasan, H. G. Vartak, M. V. Rele, and P. J. Reilly.
1991. Catalytic properties and partial amino acid sequence
of an actinomycete endo-(l-4)-B-D-xylanase from Chainia
species. p417-425. In G. F. Leatham and M. E. Himmel (ed),
Enzymes in biomass conversion. Amer. Chem. Soc. Washington,
D. C.
Biely, P. 1985. Microbial xylanolytic systems. Trends in
Biotechnol. 3:286-290.
Biely, P., C. R. MacKenzie, J. Puls, And H. Schneuder. 1986.
Cooperativity of esterases and xylanases in the enzymatic
degradation of acetyl xylan. Biotechnol. 4:731-733.
Biely, P., and E. Petrakova. 1984. Novel inducers of the
xylan degrading system of Crvotococcus albidus. J.
Bacteriol. 160:408-412.
Biely, P., J. Puls, and Schneider. 1985. Acetyl xylan
esterases in fungalcellulolytic systems. FEBS Lett. 186:80-
84.
Biely, P., Z. Kratsky, M. Vranska, and D. Urmanicova. 1980.
Induction and inducers of endo-1,4-6-xylanase in the yeast
Crvotococcus albidus. Eur. J. Biochem. 108:323-329.
Benguin, P., P. Cornet, and J. P. Aubert. 1985. Sequence of
a cellulase gene of the thermophilic bacterium Clostridium
thermocellum. J. Bacteriol. 162:102-105.
142

143
Berger, E., W. A. Jones, D. T. Jones, and D. R. Woods. 1989.
Cloning and sequencing of an endoglucanase (endl) gene from
Butvrivibrio fibrisolvens H17c. Mol. Gen. Genet. 219:193-
198.
Boel, E., M.T. Hansen, I. Hjort, and N.P. Fiil. 1984. Two
different types of intervening sequences in the glucoamylase
gene from Aspergillus niaer. EMBO J. 3:1581-1585.
Bradford,M. M. 1976. A rapid and sensitive method for the
quantitation of microgram quantities of proteins utilizing
the principle of protein-dye binding. Anal. Biochem. 72:248-
254.
Chesson, A., A. H. Gordon, and J. A. Lomax. 1983.
Substituent groups linked by alkali labile bondsto arabinose
and xylose residues of legume, grass, and cereal straw walls
and their fate during digestion by rumen microorganisms. J.
Sci. Food. Agri. 34:1330-1340.
Cotta, M. A., and R. B. Hespell. 1986. Proteolytic activity
of the ruminal bacteria Butvrivibrio fibrisolvens. Appl.
Environ. Microbiol. 52:51-58.
Coughlan, M. P. 1985. Properties of fungal and bacteria
cellulases with comment on thier production and application.
Biotechnol. Genet. Eng. Rev. 3:39-109.
Dehority, B. A. 1968. Mechanism of isolated hemicellulose
and xylan degradation by cellulolytic rumen bacteria. Appl.
Microbiol. 16:781-786.
Dehority, B. A. 1966. Characterization of several bovine
rumen bacteria isolated with a xylan medium. J. Bacteriol.
91:1724-1729.
Dekker, R. F. H. 1985. Hemicellulose degradation, p 505. In
Higuchi, T. (ed), Biosynthesis and biodegradation of wood
components. Academic Press, Orlando.
Dekker, R. F. H., and G. N. Richards. 1976. Hemicellulases:
Their occurrence, purification, properties, and mode of
action. Adv. Carbohydr. Chem. Biochem. 32:277-352.
Deshpande, V., A. Lachke, C. Mishra, S. Keskar, and M. Rao.
1986. Mode of action and properties of xylanase and B-
xylosidase from Neurospora crassa. Biotechnol. Bioengin.
28:1832-1837.
Devereux, J., P. Haeberli, and 0. Smithies. 1984. A
comprehensive set of sequence analysis programs for the VAX.
Nucleic Acids Res. 12:387-395.

144
Eriksson, K. E. L., R. A. Blanchette, and P. Ander. 1990. In
Microbial and enzymatic degradation of wood and wood
components. Springer series in wood science. Springer-Verlag
Berlin, Heideilberg. pp 407.
Esteban, R., J. R. Villanueva, and T. G. Villa. 1982. 6-D-
xylanases of Bacillus circulans WL-12. Can J. Microbiol.
28:733-739.
Frederick, M. M., C. Kiang, J. R. Frederick, and P. J.
Reilly. 1985. Purification and characterization of endo-
xylanases from Aspergillus niaer. I. Two isozymes active on
xylan backbones near branch points. Biotechnol. Bioengin.
27:525-532.
Flint, H. J., C. A. McPherson, and J. Bisset. 1989.
Molecular cloning of genes from Ruminococcus flavefaciens
encoding xylanase and J3 (13,1-4) glucanase activities. Appl.
Environ. Microbiol. 55:1230-1233.
Forsberg, C. W., B. Crosby, and D. Y. Thomas. 1986.
Potential for manipulation of the rumen fermentation through
the use of recombinant DNA techniques. J. Anim. Sci. 63:310-
325.
Forsberg, C. W., T. J. Beveridge, and A. Hellstrom. 1981.
Cellulase and xylanase release from Bacteroides succinogenes
and its importance in the rumen environment. Appl. Environ.
Microbiol. 42:886-896.
Fukumoto, J., Y. Tsujisaka, and S. Takenishi. 1970. Studies
on hemicellulases. I. Purification and properties of
hemicellulases from Aspergillus niger var. Tieghem sp.
Nippon Nogei Kogaku Kaishi. 44:447-475.
Fukusaki, E., W. Panbangred, A. Shinmyo, and H. Okada. 1984.
The complete nucleotide sequence of the xylanase gene (xylA)
of Bacillus pumilus. FEBS Lett. 171:197-201.
Garcia-Martinez, D. V., A. Shinmyo, A Madia, and A. L.
Demain. 1980. Studies on cellulase production by Clostridium
thermocellum. Appl. Microbiol. Biotechnol. 9:189-197.
Gascoigne, J. A., and M. M. Gascoigne. 1980. The xylanases
of Fusarium roseum. J. Gen. Microbiol. 22:242-248.
Grepinet, 0., Marie-Christine Chebrou, and P. Beguin. 1988.
Nucleotide sequence and deletion analysis of the xylanase
gene (xylZ) of Clostridium thermocellum. J. Bacteriol.
170:4582-4588.

145
Grepinet, 0., and P. Benguin. 1986. Sequence of the
cellulase gene of Clostridium thermocellum coding for
endoglucanase B. Nucleic Acids. Res. 14:1791-1799.
Greve, C. L., J. M. Labavitch, and R. E. Hungate. 1984. a-L-
arabinofuranosidase from Ruminococcus albus 8: Purification
and possible role in hydrolysis of alfalfa cell wall. Appl.
Environ. Microbiol. 47:1135-1140.
Hespell, R. B., and P. J. 0'Bryan-Shah. 1988. Esterase
activities in Butvrivibrio fibrisolvens strains. Appl.
Environ. Microbiol. 54:1917-1922.
Hespell, R. B., R. Wolf, and R. J. Bothast. 1987.
Fermentation of xylans by Butvrivibrio fibrisolvens and
other ruminal bacterial species. Appl. Environ. Microbiol.
53:000-000.
Hespell, R. B., and M. P. Bryant. 1981. The genera
Butvrivibrio. Succinivibrio. Succinomomas. Lachnospira. and
Selenomonas. p. 1479-1494. In M. P. Starr, H. Stolp, H. G.
Truper, A. Balows, and H. G. Schlegel (ed), The prokaryotes,
a handbook on habitates, isolation, and identification of
bacteria. Springer-Verlag, New York.
Hobson, P. N., and R. J. Wallace. 1982. Microbial ecology
and activities in the rumen. II. Crit. Rev. Microbiol.
9:253-320.
Hobson, P. N., and M. R. Purdom. 1961. Two types of xylan
fermenting bacteria from the sheep rumen. J. Appl.
Bacteriol. 24: 188- 193.
Joliff, G., P. Benguin, and J. P. Aubert. 1986. Nucleotide
sequence of the cellulase gene celD encoding endoglucanase D
of Clostridium thermocellum. Nucleic Acids Res. 14:8605-
8613.
Kasumi, T., Y. Tsumuraya, C. F. Brewer, H. Kersters-
Hilderson, M. Claeyssens, and E. J. Hehre. 1987. Catalytic
versatility of Bacillus pumilus B-xylosidase: Glucosyl
transfer and hydrolysis promoted with a- and B-D-xylosyl
fluoride. Biochem. 26:3010-3016.
Kelly, M. A., M. L. Sinnott, and M. Herrchen. 1987.
Purification and mechanistic properties of an extracellular
a-L-arabinofuranosidase from Monilinia fructigena. Biochem
J. 245:843-849.

146
Kersters-Hilderson, H., F. G. Loontiens, M. Claeyssens, and
C. K. De Bruyne. 1969. Partial purification and properties
of an induced B-D-xylosidase of Bacillus pumilus 12. Eur. J.
Biochem. 7:434-441.
Knowles, J. K. C., P. Lehtovaara, M. Murray, and M. L.
Sinnot. 1988. Stereochemical course of action of the
cellobioside hydrolases I and II of Trichoderma reesei. J.
Chem. Soc. Chem. Commun. 1988:1401-1402.
Knowles, J., P. Lehtovaara, and T. Teeri. 1987. Cellulase
families and their genes. Trends Biotechnol. 5:255-261.
Koshland, D. E., Jr., G. Nemethy, and D. Filmer. 1966.
Comparison of experimental binding data and theoretical
models in proteins containing subunits. Biochem. 5:365-376.
Kuriki, T., H. Takata, S. Okada, and T. Imanaka. 1991.
Analysis of the active center of Bacillus staerothermophilus
neopullulanase. J. Bacteriol. 173:6147-6152.
Laemmli, U. K. 1970. Cleavage of structural proteins during
the assembly of the head of bacteriophage T4. Nature
227:680-685.
Lee, S. F., and C. W. Forsberg. 1987. Purification and
characterization of an a-L-arabinofuranosidase from
Clostridium acetobutylicum ATCC 824. Can. J. Microbiol.
33:1011-1016.
Lee, S. F., C. W. Forsberg, and J. B. Rattray. 1987.
Purification and characterization of two endoxylaases from
Clostridium acetobutylicum ATCC 824. Appl. Environ
Micriobiol. 53:644-650.
Luthi, E., D. R. Love, J. McAnulty, C. Wallace, P. A.
Caughey., D. Saul, and P. Bergquist. 1990. Cloning, sequence
analysis, and expression of genes encoding xylan-degrading
enzymes from the thermophile Caldocellum saccharolvticum.
Appl. Environ. Microbiol. 56:1017-1024.
Lynch, J. M. 1987. Utilization of lignocellulosic wastes. In
Journal of Applied Bacteriology Symposium Supplement, 71S-
83S.
Mackenzie, C. R., D. Bilous, H. Schneider, and K. G.
Johnson. 1987. Induction of cellulolytic and xylanlolytic
enzyme systems in Streptomvces spp. Appl. Environ.
Microbiol. 53:2835-2839.

147
MacKenzie, C. R., R. C. A. Yang, G. B. Patal, D. Bilous, and
S. A. Nurang. 1989. Identification of three distinct C.
thermocellum xylanase genes by molecular cloning. Arch.
Microbiol. 152:377-381.
MacNeil, N. I. 1984. The contribution of the large intestine
to energy supplies in man. Amer. J. Clin. Nutr. 39:338-346.
Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982.
Molecular cloning: a laboratory manual. Cold Spring Harbor
Laboratory, Cold Spring Harbor, N. Y.
Mannarelli, B. M., R. J. Stack, D. Lee, and L. Ericsson.
1990. Taxonomic relatedness of Butvrivibrio. Lachnospira.
Roseburia and Eubacterium species as determined by DNA
hybridization and extracellular-polysaccharide analysis.
Int. J. Syst. Bacteriol. 40:535-544.
Mannarelli, B. M., S. Evans, and D. Lee. 1990. Cloning,
sequencing and expression of a xylanase gene from the
anaerobic rumen bacterium Butvrivibrio fibrisolvens. J.
Bacteriol. 172:4247-4254.
Mannerelli, B. M. 1988. Deoxyribonucleic acid relatedness
among strains of the species Butyrivibrio fibrisolvens. Int.
J. Syst. Bacteriol. 23:308-315.
McGavin, M., and C. W. Forsberg. 1989. Catalytic and
substrate-binding domains of endoglucanase 2 from
Bacteroides succinoqenes. J. Bacteriol. 171:3310-3315.
Morag. E., E. A. Bayer, and R. Lamed. 1990. Relationship of
cellulosomal and noncellulosomal xylanases of Clostridium
thermocellum to cellulose-degrading enzymes. J. Bacteriol.
172: 6098-6105.
Moriyama, H., E. Fufusaki, J. Cabrera Crespo, A. Shinmyo.,
and H. Okada. 1987. Structure and expression of genes coding
for xylan-degrading enzymes of Bacillus pumilus. Eur. J.
Biochem. 166:539-545.
Morosoli, R., C. Roy, and M. Yaguchi. 1986. Isolation and
partial primary sequence of a xylanase from the yeast
Cryptococcus albidus. Biochim. Biophys. Acta 870:473-478.
Nakanishi, K., T. Yasui, and T. Kobayashi. 1971. Inducers
for the xylanase production by Streptomvces sp. J. Ferment.
Technol. 54:801-807.
Nishitani, K., and D. J. Nevins. 1991. Glucuronoxylan
xylanohydrolase: A unique xylanase with the requirement for
appendant glucuronosyl units. J. Biol. Chem. 266:6539-6543.

148
Nishitani, k., and D. J. Nevins. 1988. Enzymatic analysis of
feruloylated arabinoxylans (Feraxan) derived from Zea mays
cell walls I. Plant Physiol. 87:883-890.
Ornstein, L., and B. J. Davis. 1964. Disc electrophoresis:
background and theory. Ann. N. Y. Acad. Sci. 121:321-349.
Paice, M. G., R. Bourbonnais, M. Desrochers, L. Jurasek, M.
Yaguchi. 1986. A xylanase gene from Bacillus subtilis:
nucleotide sequence and comparison with B. pumilus gene.
Arch. Microbiol. 144:201-206.
Panbangred, W., E. Fukusaki, E. C. Epifanio, A. Shinmyo, and
H. Okada. 1985. Expression of a xylanase gene of Bacillus
pumilus in Escherichia coli and Bacillus subtilis.
Appl. Microbiol. Biotechnol. 22:259-264.
Panbangred, W., A. Shinmyo, S. Kinoshita, and H. Okada.
1983a. Purification and properties of endoxylanase produced
by Bacillus pumilus. Agrie. Biol. Chem. 47:957-963.
Panbangred, W., T. Rondo, S. Negoro, A. Shinmyo, and H.
Okada. 1983b. Molecular cloning of the genes for xylan
degradation of Bacillus pumilus and their expression in
Escherichia coli. Mol. Gen. Genet. 192:335-341.
Patterson, J. A. 1989. Prospects for establishment of
genetically engineered microorganisms in the rumen. Enzyme
Microb. Technol. 11:187-189.
Puo-Llinas, J., and H. Driguez. 1987. D-Xylose as inducer of
the xylan-degrading system in the yeast Pullularia
pullulans. Appl. Microbiol. Biotechnol. 27:134-138.
Poutanen, K., M. Tenkanen, H. Korte, and J. Puls. 1991.
Accessory enzymes involved in the hydrolysis of xylans. In
Enzymes in biomass conversion, G. F. Leatham and M. E.
Himmel, ed. American Chemical Society Symposium series 460.
American Chemical Society, Washington, D. C.
Poutanen, K., and J. Puls. 1988. Characteristics of
Trichoderma reesei B-xylosidase and its use in the
hydrolysis of solubilized xylans. Appl. Microbiol.
Biotechnol. 28:425-432.
Poutanen, K., and M. Sundberg. 1988. An acetyl esterase
of Trichoderma reesei and it'srole in the hydrolysis of
acetyl xylans. Appl Microbiol. Biotechnol. 28:419-424.
Quiocho, F. A. 1986. Carbohydrate-binding proteins:
tertiary structures and protein-sugar interactions. Ann.
Rev. Biochem. 55:287-315.

149
Raynal,A., C. Gerbaud, M. C. Francingues, and M. Guerineau.
1987. Sequence and transcription of the B-glucosidase gene
of Kluvveromvces fraailis cloned in Saccharomvces
cerevisiae. Curr. Genet. 12:175-184.
Rerat, A., M. Fiszlewicz, A.Giusi, and P. Vaugelade. 1987.
Influence of meal frequency on postprandial variations in
the production of volatile fatty acids in the digestive
tract of conscious pigs. J. Anim. Sci. 64:448-453.
Robson and Gardier. 1988. Proteins and protein engineering,
p 699. Elsevier Science Publishers B. V.
Rouvinen, J., T. Bergfors, T. Teeri, J. K. C. Knowles, and
T. A. Jones. 1990. Three-dimensional structure of cellobio-
hydrolase II from Trichoderma reesei. Science 249:380-386.
Salyers, A. A., J. R. Balascio, and J. K. Palmer. 1981.
Breakdown of xylan by enzymes from Human colonic bacteria.
J. Food Biochem. 6:39-55.
Saman, E., M. Claeyssens and C. K. Bruyne. 1975. Bacillus
pumilus B-D-xylosidase: study of thiol groups. In
Biochemical Society Transactions, 558th Meeting, Edinburgh.
3:998-999.
Saul, D. J., L. C. Williams, R. A. Grayling, L. W. Chamley,
D. R. Love, and P. L. Bergquist. 1990. celB, A gene coding
for a bifunctional cellulase from the extreme thermophile
MCaldocellum saccharolvticum". Appl. Environ. Microbiol.
56:3117-3124.
Sewell, G. W., H. C. Aldrich, D. Williams, B. Mannarelli, A.
Wilkie, R. B. Hespell, P. H. Smith, and L. 0. Ingram. 1988.
Isolation and characterization of xylan-degrading strains of
Butvrivibrio fibrisolvens from a Napier grass-fed anaerobic
digester. Appl. Environ. Microbiol. 54:1085-1090.
Sewell, G. W., E. A. Utt, R. B. Hespell, K. F. MacKenzie,
and L. 0. Ingram. 1989. Identification of the Butvrivibrio
fibrisolvens xylosidase gene (xylB) coding region. Appl.
Environ. Microbiol. 55:306-311.
Teeri, T. T., P. Lehtovaara, S. Kauppinen, I. Salovouri, and
J. Knowles. 1987. Homologous domains in Trichoderma reesei
cellulolytic enzymes: gene sequence and expression of
cellobiohydrolase. Gene. 51:43-52.
Uziie, M., M. Matsuo, and T. Yasui. 1985. Possible identity
of 6-xylosidase and B-glucosidase of Chaetomium triaterale.
Agrie. Biol. Chem. 49:1167-1173.

150
Varel, V. H. 1987. Activity of fiber-degrading microorganism
in the pig large intestine. J. Anim. Sci. 65:488-496.
Ward, 0. P., and M. Moo-Young. 1989. Enzymatic degradation
of cell wall and related plant polysaccharides, p. 237-274.
In CRC CRitical Reviews in Biotechnology vol 8.
Weinstein, L., and P. Albersheim. 1979. Structure of plant
cell walls. Plant. Physiol. 63:425-432.
Whitehead, T. R., and R. B. Hespell. 1990. The gene for
xylan-degrading activities from Bacteroides ovatus are
clustered in a 3.8 kilobase region. J. Bacteriol. 172:2408-
2412.
Wong, K. Y., L. U. L. Tan, and J. N. Saddler. 1986.
Purification of a third distinct xylanase from the
xylanolytic system of Trichoderma harzianum. Can J.
Microbiol. 32:570-576.
Yaguchi, M., C. Roy, C. F. Rollin, M. G. Paice, and L.
Jurasek. 1983. A fungal cellulase shows sequence homology
with the active site of hen egg-white lysozyme. Biochem.
Biophy. Res. Comm. 116:408-411.

BIOGRAPHICAL SKETCH
Eric Andrew Utt was born at the dawn of the space age, on
February 20,1961, in Fort Lauderdale, Florida, to Harold and
Anita Utt. He grew up as the youngest of two sons in Cocoa
Beach, Florida. He attended Cocoa Beach High School where he
graduated in 1979. In the fall of that year, he attended
Brevard Community College before transferring to the
University of Central Florida. He was awarded the Bachelor of
Science degree in microbiology in 1984 and went on to pursue
and receive a Master of Science degree in microbiology with
Dr. Rudy J. Wodzinski at the University of Central Florida.
In 1987, he began doctoral studies in the laboratory of Dr.
Neal Ingram in the Department of Microbiology and Cell Science
at the University of Florida in Gainesville. During his
tenure in Dr. Ingram's laboratory, he became interested in the
structural and functional relationships of proteins and
microbial evolution. This interest expanded during his
doctoral research and is an area in which he would like to
continue to work. Dr. Utt has been awarded a National
Research Council Postdoctoral Fellowship and will pursue
research in the genetic characterization of virulence factors
in Listeria monocytogenes at the Centers for Disease Control,
Atlanta, Georgia.
151

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.
' * l 0
/a
Lonnie 0. Ingram /
Chairman, Professor of
Microbiology and Cell Science
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.
Cell Science
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.
- \ c ,
Keelnatham T. Shanmugam
Professor of Microbiology and
Cell Science
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.
Henry ti. Aldrich
Professor of Microbiology and
Cell Science

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.
CLjl to. Ojia.
Charles M. Allen
Professor of Biochemistry and
Molecular Biology
This dissertation was submitted to the Graduate Faculty
of the College of Agriculture and to the Graduate School and
was accepted as partial fulfillment of the requirements for
the degree of Doctor of Philosophy.
December 1991
of.
gfjicu]
Dean,
lege of Agriculture
Dean, Graduate School



Percent Relative Activity
f'O 4*. CT> CO O
O O O O CD O
107


10
distributed in nature (Poutanen et al. 1991). The acetyl
residues on the xylan backbone are removed by acetyl
esterase (Biely et al. 1985). These enzymes have been found
to act cooperatively with xylanases. The acetyl esterase
serves to increase the rate of glycosidic bond cleavage by
B-xylanase from Trichoderma reesei (Biely et al. 1986).
These enzymes were also found to act synergistically to
liberate acetyl residues. More recently it was demonstrated
that the rate of liberation of acetic acid from acetyl-xylan
by acetyl esterase of T. reesei was increased by the
addition of endo-xylanase and B-D-xylosidase (Poutanen and
Sundberg, 1988).
Also involved in enzyme synergism is the enzyme a-L-
arabinofuranosidase (Greve et al. 1984). This enzyme was
purified from Ruminococcus albus 8 and had a pH optimum of
6.9 and a Km of 1.3 mM, both for p-nitrophenyl-a-L-
arabinofuranoside as a substrate. They showed that this
enzyme enhanced the rate of hydrolysis of alfalfa cell wall
hemicellulose when combined with other xylanolytic or
pectinolytic enzymes. It was hypothesized that this enzyme
functioned to provide rumen microbes with suitable
substrates for xylanase.
The mechanism of xylan hydrolysis by microbial
xylanases has been studied extensively. Xylanases are
usually small proteins having molecular weights ranging
between 20,000 to 50,000 (Bastawde et al. 1991). Most


40
TABLE 2. Expression of enzyme activities in recombinant E. coli
harboring xvlB.
plasmid
Sp act*
xylosidase
arabinofuranosidase
Ratio ara/xylb
pLOI1005
9.0
16.0
1.8
pLOI1040
0.2
0.2
-
pLOI1043
2.0
3.0
1.5
PLOI1050
6.8
10.1
1.5
pUC18
0.2
0.2
-
1 Nanomoles per minute per milligram of cell protein.
b Ratio calculated after subtraction of background values from
the pUC18 control.


78
perturbations that would result in proteolysis by the cell
or a major change in tertiary structure of the proteins.
The reduced levels of the xylosidase protein in the
W158UGA mutant correlates to the reduced enzymatic
activities for recombinants harboring this mutant. The
"UGA" termination codon can be decoded as a tryptophan at
low efficiency in E. coli (Hirsh, 1971). It is likely that
this is the case also for the W158UGA mutant in xvlB.
Substrate-binding comparisons of mutant proteins.
Electrophoresis of the wild type and mutant proteins on
native gels that contained agarose alone and agarose plus
birchwood xylan indicated a differential mobility between
the wild type and mutant proteins (Fig. 14). Using agarose
alone no differences between electrophoretic mobilities of
the respective mutant proteins and the wild type were
detected using Western hybridization analysis. The
inclusion of birchwood xylan (0.75 %) resulted in a change
in the mobility of the proteins. Without exception, all the
mutant proteins exhibited faster electrophoretic mobilities
relative to the wild type protein. It is possible that the
xylan is functioning as a psuedo-substrate and the point
mutations have affected the relative affinities of these
proteins for the substrate.


118
The kinetic properties of the xvlB-encoded enzyme with
respect to the xylosidase activity is similar to the
xylosidase enzymes from C. acetobutvlicum and B. pumilus
which have Km's for p-NP-X of 3.8 mM, 3.7 mM, and 2.4 mM,
respectively. The Km for the arabinofuranosidase activity
against p-NP-A was 1.7 mM which compares favorably with 1.3
mM for the a-L-arabinofuranosidase isolated from another
rumen bacteria, Ruminococcus albus (Greve et al. 1984) The
maximum velocity values for the xylosidase activity of the
xvlB-encoded protein is approximately tenfold lower than
that for the C. acetobutvlicum enzyme and approximately one-
half that for the B. pumilus enzyme. Lee and Forsberg
(1987) purified and characterized an a-L-
arabinofuranosidase from C. acetobutvlicum ATCC 824. This
enzyme exhibited Km and Vm values of 4 mM and 3 6.4 mole
min'1 mg'1 protein respectively against p-nitrophenyl-a-L-
arabinofuranoside. It had a pH optimum of 5.0 5.5 and
exhibited no activity against other p-nitrophenylglycosides.
High catalytic efficiency is usually associated with highly
evolved enzymes having a narrow substrate specificity
(Robson and Gamier, 1988) It is possible that the xvlB-
encoded enzyme has sacrificed catalytic efficiency for a
broader substrate range since neither the C. acetobutvlicum
or the B. pumilus xylosidase enzymes have additional
activities.


3
obligately anaerobic bacillus that is frequently found in
the rumen and anaerobic digesters (Dehority, 1966). This
organism produces a cadre enzymes which enable it to degrade
plant biomass, including cellulose and hemicellulose
(Hespell, 1987). The genus Butyrivibrio contains only a
single species but consists of many strains that vary in DNA
relatedness between 20% to 100% (Mannarelli, 1988). This
organism is also characterized as having a low (38% to 42%)
guanine plus cytosine (mole percent) content. Butyrivibrio
produces an extracellular polysaccharide (EPS) that contains
an unusual 4-0-(1-carboxymethyl)-rhamnose sugar (Mannarelli
et al. 1990). These investigators suggested that the
unusual sugars found in the EPS of B. fibrisolvens serve to
protect the organism from glycosidases and other enzymes
found in the digestive tract of the host animal.
B. fibrisolvens GS113, an anaerobic digester isolate
used in these studies, was shown to produce high levels of
both xylanase and xylosidase (Sewell et al. 1988) These
two enzymes were shown to be repressed by glucose and
induced by xylan and xylose.
In a previous study, the xvlB gene encoding the B-D-
xylosidase from B. fibrisolvens GS113 was isolated from a
plasmid pUC18 genomic library (Sewell et al. 1989).
During the course of the current investigations, an a-
L-arabinofuranosidase activity was detected in all clones
harboring the xvlB gene. This characteristic became the


42
TABLE 3. Hydrolysis of different nitrophenyl-substituted glycosides by the xvlB
gene product.
substrate
SDecific activity*
pLOI1005
pUC18
p-NP-B-D-xylopyranoside
8.9
0.2
p-NP-a-L-arabinofuranoside
15.5
0.5
p-NP-a-L-arabinopyranoside
0.1
0.2
p-NP-a-D-galactopyranoside
0.2
CM
O
p-NP-a-D-glucopyranoside
3.5
3.3
p-NP-a-L-fucopyranoside
0.2
0.2
p-NP-fi-D-fucopyranoside
0.2
0.2
p-NP-B-L-fucopyranoside
0.4
0.4
p-NP-a-L-rhamnopyranoside
0.3
0.2
o-NP-fi-D-fucopyranoside
1.4
1.0
o-NP--D-galactopyranoside
0.8
1.0
o-NP-B-D-galactopyranoside
1.0
1.0
* Nanomoles per minute per milligram of cell protein.


116
exhibited an optimum pH of 6.0 which compares with that of
the C. acetobutvlicum xylosidase. There are four histidine
residues in the proposed active center for this enzyme.
Since the pKa of histidine is 6.5, it is likely that one or
more of these residues are involved in both catalytic
activities.
The proposed mechanism for glycohydrolase, and the
related glycosyltransferase catalysis, involves
contributions from a hydrogen donor and a hydrogen acceptor
moiety (Mooser et al. 1991). In the usual scheme, an
aspartic or glutamic acid in deprotonated form serves as a
hydrogen acceptor and stabilizes the hydrolytic
intermediate. A general acid donates a proton to the
glycosidic oxygen to facilitate the cleavage of the
monomeric glycoside. The pH dependence of the xvlB-encoded
protein with respect to the 6-D-xylosidase activity is
probably due to one or more histidine residues serving as
the general acid.
The broad pH activity profile for the
arabinofuranosidase activity is uncommon. At the higher pH
ranges, histidine is completely deprotonated and could no
longer serve as a general acid and donate a proton. It is
possible that the arabinofuranoside substrate, once bound in
the active pocket, protects the histidine from further
deprotonation at high pH. This charge transfer phenomenon
has been demonstrated to occur in other enzymatic systems


Figure 22. Double recipricol plots of xylosidase and
arabinofuranosidase activities. Xylosidase; closed circles,
arabinofuranosidase; open circles.


127


Figure 2. Subclone analysis of pLOI1005 to localize the xylB
coding region: xyl; xylosidase activity, araf;
arabinofuranosidase activity. A "+" or denotes the
presence or absence of enzyme activity.


66
the "ATG" initiation codon with the loss of both enzymatic
activities (Fig. 9b).
In vitro mutagenesis. Twelve mutants were isolated
using selective media on the basis of reduced or abolished
enzyme activities. Two mutants could not be classified due
to multiple mutations and were not analyzed further. A
total of ten mutants were classified as negative for both
enzymatic activities against the fluorogenic substrates
based upon agar plate assays. Two mutants, number six and
number ten, displayed reduced but significant fluorescence
on fluorogenic indicator plates (Table 6).
Localization of point mutations. Most of the point
mutations were determined to reside in domains II and III
(Table 6). None of the mutations were localized in the
lacZ1 promoter of pUC18. The generation of point mutations
is consistent with the deamination activity of nitrous acid.
The point mutations were clustered in domains II and III
near the region proposed to be the active site of this
enzyme.
DNA sequencing of in vitro mutants. The amino acid
substitutions resulting from in vitro mutagenesis were
deduced from the base changes as determined by DNA
sequencing (Fig. 10). The point mutations were all AT to GC
transitions which is consistant with the mode of action for
nitrous acid mutagenesis. Two frameshift mutations were
identified, in domain I and domain V. These were not


ACKNOWLEDGEMENTS
I owe my development as a scientist as well as the
successful completion of this dissertation to my major
professor, Dr. Neal Ingram. I will be forever indebted to
him for his uninhibited sharing of knowledge and expertise.
I wish also to express my gratitude to the members of my
graduate committee, Dr. Allen, Dr. Aldrich, Dr. Shanmugam,
and Dr. Gander. Their contributions to my research and in
the preparation and review of this manuscript are greatly
appreciated. I must also thank my friends and comrades Jeff
Mejia and David Beall for the friendship and helpful
suggestions during the course of my doctoral work. To all
the former postdocs, Dr. Christina Eddy, Dr. Terryl Conway,
and Dr. Guy Sewell, from whom I learned much, I wish to
express my thanks. Thanks are due to my parents for their
love and support during my graduate education. Similarly, I
wish to thank my wife's parents for their love and support.
And lastly, I would like to thank my loving wife, Lisa, and
my beautiful girls, Tina and Hannah, for making my life
special. It is to them that I dedicate this dissertation.
ii


Figure 16. Elution profile of the xvlB gene product during
hydrophobic interaction chromatography.


144
Eriksson, K. E. L., R. A. Blanchette, and P. Ander. 1990. In
Microbial and enzymatic degradation of wood and wood
components. Springer series in wood science. Springer-Verlag
Berlin, Heideilberg. pp 407.
Esteban, R., J. R. Villanueva, and T. G. Villa. 1982. 6-D-
xylanases of Bacillus circulans WL-12. Can J. Microbiol.
28:733-739.
Frederick, M. M., C. Kiang, J. R. Frederick, and P. J.
Reilly. 1985. Purification and characterization of endo-
xylanases from Aspergillus niaer. I. Two isozymes active on
xylan backbones near branch points. Biotechnol. Bioengin.
27:525-532.
Flint, H. J., C. A. McPherson, and J. Bisset. 1989.
Molecular cloning of genes from Ruminococcus flavefaciens
encoding xylanase and J3 (13,1-4) glucanase activities. Appl.
Environ. Microbiol. 55:1230-1233.
Forsberg, C. W., B. Crosby, and D. Y. Thomas. 1986.
Potential for manipulation of the rumen fermentation through
the use of recombinant DNA techniques. J. Anim. Sci. 63:310-
325.
Forsberg, C. W., T. J. Beveridge, and A. Hellstrom. 1981.
Cellulase and xylanase release from Bacteroides succinogenes
and its importance in the rumen environment. Appl. Environ.
Microbiol. 42:886-896.
Fukumoto, J., Y. Tsujisaka, and S. Takenishi. 1970. Studies
on hemicellulases. I. Purification and properties of
hemicellulases from Aspergillus niger var. Tieghem sp.
Nippon Nogei Kogaku Kaishi. 44:447-475.
Fukusaki, E., W. Panbangred, A. Shinmyo, and H. Okada. 1984.
The complete nucleotide sequence of the xylanase gene (xylA)
of Bacillus pumilus. FEBS Lett. 171:197-201.
Garcia-Martinez, D. V., A. Shinmyo, A Madia, and A. L.
Demain. 1980. Studies on cellulase production by Clostridium
thermocellum. Appl. Microbiol. Biotechnol. 9:189-197.
Gascoigne, J. A., and M. M. Gascoigne. 1980. The xylanases
of Fusarium roseum. J. Gen. Microbiol. 22:242-248.
Grepinet, 0., Marie-Christine Chebrou, and P. Beguin. 1988.
Nucleotide sequence and deletion analysis of the xylanase
gene (xylZ) of Clostridium thermocellum. J. Bacteriol.
170:4582-4588.


Figure
Page
27. Thermal inactivation profiles of xylosidase
and arabinofuranosidase activities of the
L178F mutant protein 127
28. pH activity profiles for xylosidase and
arabinofuranosidase activities of the
L178F mutant protein 130
29. Double recipricol plots of xylosidase and
arabinofuranosidase activities for the
L178F mutant protein 132
viii


108
Kinetic properties. A characteristic Michaelis-Menten
relationship was observed for the native protein using p-NP-
X and p-NP-A as substrates. The Km and Vmax values for
xylosidase and arabinofuranosidase activities were 3.8 mM ,
1.7 mM and 1,111 nmoles min'1 mg'1 protein, 833 nmoles min'1
mg'1 protein, respectively (Fig. 22). These values give a
Kcat or turnover number per subunit of 1111 sec'1 for
xylosidase activity and 833 sec'1 for arabinofuranosidase
activity.
Substrate competition experiments. The florogenic
substrate 4-methylumbelliferyl-B-D-arabinofuranoside proved
to be a potent competitive inhibitor of S-D-xylosidase
activity when p-NP-X was the substrate (Fig. 23).
Alternatively, fi-methyl-D-xylopyranoside, a xylobioside
analogue, was a competitive inhibitor of a-L-
arabinofuranosidase activity when p-NP-A was the substrate
(Fig. 24).
In separate experiments, the enzymatic activities were
measured with the two substrates, p-NP-X and p-NP-A
individually and together at 10 mM (saturation)
concentration at 37C for 10 min. Xylosidase activity alone
was 1160 mU, and arabinofuranosidase activity alone was 809
mU. When both substrates were combined in the same assay
the resultant activity was 948 mU which is essentially the
average of the two activities separately. These individual
activities are not additive when both substrates are present


85
enzyme. There appears to be a dependence of velocity on
substrate concentration for at least some of the mutants.
The mutation evidence supports the hypothesis that one
active center or domain is responsible for both enzymatic
activities. It does not, however, totally rule out the
possibility that the protein may contain separate catalytic
sites or subsites which are spacially close together.
Future kinetic experiments, including the investigation of
substrate competition between arabinofuranosides and
xylopyranosides, will allow further definition of the
catalytic regions responsible for both activities.


57
Localization of point mutations. The entire xvlB
coding region was divided into five domains based upon
restriction sites (Fig. 7). Three restriction fragments,
the PstI. EcoRI. and Hindlll fragments were isolated from
each mutant plasmid. These fragments were used to replace
the corresponding fragment in the wild type gene which had
been modified to construct receiving vectors for each
respective restriction fragment (pLOI1051, pLOI1052, and
pLOI1053) to test the functionality of individual fragments.
This strategy, outlined in Fig. 8, allowed the localization
of point mutations to one or more of the five domains.
Exonuclease III deletion of xvlB. Plasmid pLOI1043
containing the xvlB coding region was deleted from the 3'
terminal region into the coding region by exonuclease III
using the "Erase-a-Base" deletion kit (Promega Corporation,
Madison, WI.) according to the manufacturer's instructions.
The deleted plasmids were subsequently transformed into
competent E. coli DH5a and screened in the same fashion as
the in vitro-qenerated mutant plasmids.
5' deletion analysis and lacZ1 fusions. The internal
PstI fragment from pLOI1005 was subcloned in the original
orientation into plasmid pUC18. This in-frame fusion with
lacZ results in the subsequent deletion of the 5' terminal
54 base pairs (18 amino acids).
Preparation of cell extracts. Extracts were prepared
as described in chapter III.


25
Sau3A fragments were used as probes against PstI and Sau3A
digests of E. coli DH5a, B. fibrsolvens GS113 chromosomal
DNA and plasmid pLOI1005 (Fig. 3). These probes did not
bind to the DH5a chromosomal DNA but did bind to a single
band in each B. fibrisolvens chromosomal digest. A single
copy of the xvlB gene appears to be present in B.
fibrisolvens GS113.
Multiple copies of xylosidase genes have been reported
for Bacillus pumilus (Panbangred et al.. 1984) and for
Caldocellum saccharolvticum (Luthi et al.. 1990). If there
are additional xylosidase genes present in GS113, then they
must share limited homology with xvlB.
DNA sequence of the xylosidase gene (xvlB). Plasmid
pLOHOOl, the original GS113 library clone (Sewell et al.,
1989), contained a 4.2 kb insert of B. fibrisolvens DNA.
Both strands of this fragment were sequenced utilizing the
strategy summarized in Fig. 4A. The complete nucleotide
sequence of this fragment is outlined in Fig. 5. Analysis
of the sequence revealed the presence of three open reading
frames (ORF's) in this DNA segment. The first ORF, ORF1,
was incomplete and is 1,340 bp in length. It lacks a Shine-
Dalgarno sequence and an ATG initiation codon. 0RF2, was
1,551 bp in length and was found 15 bp downstream from ORF1.
ORF2 has a Shine-Dalgarno sequence located 6 bp upstream
from the initiation codon and defines a complete gene. This
gne spans the predicted xylosidase coding region and


Figure 20. Temperature optimum profile for xylosidase (closed
circle) and arabinofuranosidase (open circle).


Figure 11. SDS-PAGE analysis of wild type and mutant
proteins. Lane assignments: A and J; wild type (pLOI1005) B;
pUC18, C; G186R, D; A203T, E; A210V, F; P7L, G; L178F, H;
W158UGA, I; molecular weight markers, K; A203V, L; G238D, M;
G238D expressed in a Ion' strain of E. coli N; A210T, 0;
A203T. Molecular weight marker sizes (X 103) : 1; 94, 2; 67,
3; 43, 4; 30, 5; 20.


70
studied further. Ten mutations are clustered in an area of
60 amino acids. Six of these mutations; glycine 185 to
arginine 185, alanine 202 to threonine 202, alanine 202 to
valine 202, alanine 209 to valine 209, glycine 237 to
aspartate 237, and alanine 209 to threonine 209 all resulted
in an inactive protein. Each of these mutations represent
nonconservative amino acid changes that would be expected to
effect the function and/or conformation of the protein.
Another mutation in this cluster, leucine 178 to
phenylalanine 178, resulted in a mutant having one-tenth the
enzymatic activities of the wild type. This is a
conservative change in that both leucine and phenylalanine
are hydrophobic amino acids and have similar structural
properties. It is possible that the larger aromatic group
on phenylalanine is affecting substrate binding and/or
catalysis. In addition, the substitution of a "UGA" stop
codon for tryptophan at position 152 also resulted in a
protein with one-tenth the enzymatic activities of the wild
type. The most probable explaination is that the "UGA"
codon is functioning as tryptophan in this E. coli strain
and the efficiency of read through is very low which yields
a reduced expression of the xvlB protein. The existence of
such suppressor mutations has been previously demonstrated
(Hirsh, 1971). One mutation, proline (7) to leucine (7)
occurred in the amino terminus of the protein and resulted
in a negative phenotype. The amino terminus of the xvlB


LO


BIOGRAPHICAL SKETCH
Eric Andrew Utt was born at the dawn of the space age, on
February 20,1961, in Fort Lauderdale, Florida, to Harold and
Anita Utt. He grew up as the youngest of two sons in Cocoa
Beach, Florida. He attended Cocoa Beach High School where he
graduated in 1979. In the fall of that year, he attended
Brevard Community College before transferring to the
University of Central Florida. He was awarded the Bachelor of
Science degree in microbiology in 1984 and went on to pursue
and receive a Master of Science degree in microbiology with
Dr. Rudy J. Wodzinski at the University of Central Florida.
In 1987, he began doctoral studies in the laboratory of Dr.
Neal Ingram in the Department of Microbiology and Cell Science
at the University of Florida in Gainesville. During his
tenure in Dr. Ingram's laboratory, he became interested in the
structural and functional relationships of proteins and
microbial evolution. This interest expanded during his
doctoral research and is an area in which he would like to
continue to work. Dr. Utt has been awarded a National
Research Council Postdoctoral Fellowship and will pursue
research in the genetic characterization of virulence factors
in Listeria monocytogenes at the Centers for Disease Control,
Atlanta, Georgia.
151


50
glutamic and aspartic acid residues and the approximate
spacing were found to be conserved conserved.
Conclusions
The xvlB gene, encoding 6-D-xylosidase and a-L-
arabinofuranosidase activities is the first of its kind to
be sequenced. The xvlB gene is 1,551 bp in length and
encodes a 517 amino acids protein having a predicted
molecular weight of 58,000. The absence of any significant
stem-loop structures or terminators in the regions between
ORF1, ORF2, and ORF3 as well as the strong expression of
ORF2 in E. coli even when preceded by ORF1 suggests that
these three genes may constitute a xylan-degrading operon.
The subcloning analysis and insertional inactivation studies
demonstrate the dependence of both activities on the intact
xvlB gene.
The codon usage of the three ORF's is consistent with
the low guanine plus cytosine content of this organism in
general (Mannarelli et al. 1990).
This enzyme exhibited B-D-xylopyranosidase and a-L-
arabinofuranosidase activities. No additional glycosidic
bond cleavage activities were detected in the xvlB gene
product.
The xvlB gene displayed limited homology to other
reported xylosidase sequences and must therefore be
considered to be evolutionarily divergent from genes
encoding similar functions from other organisms. It did


39
the wobble position with the exception of CAG for Gin, AAG
for Lys, and GAG for Glu.
Insertional inactivation and subclone analysis of
0RF2. Subclones were generated to investigate the
relationship between the xylosidase and arabinofuranosidase
activities encoded by the xylB gene (Fig. 4b). Retention or
loss of enzymatic activity was initially asssayed on 4-
methylumbelliferyl-B-D-xylopyranoside (MUX) and 4-
methylumbelliferyl-a-L-arabinofuranoside (MUA) indicator
plates. The SspI fragment (1,843 bp) from pLOI1005, which
contains 17 bp upstream and 274 bp downstream in addition to
xvIB. was subcloned in both directions in pUC18. Both
activities were concurrently expressed only when xvlB was
cloned in the direction of transcription of the lac promoter
(pLOI1043), indicating a dependence on this promoter in E.
coli. The insertion of a 10 bp Notl linker into the unique
EcoRV site of xvlB (pLOI1040) resulted in a frameshift
mutation that abolished both enzymatic activities (Fig. 4b).
The results of the indicator plate assays were confirmed by
comparing the specific activities for xylosidase and
arabinofuranosidase in cell free extracts (Table 2.). Using
p-nitrophenol derivatives, arabinofuranosidase activity was
approximately 1.7-times higher than the xylosidase activity.
The ratio of these activities was the same for the three
active subclones. The original subclone, pLOI1005,
exhibited the highest specific activity. The smaller active


Chapter IV
Mutational analysis of the xvlB gene
Introduction
Carbohydrate-degrading enzymes have been studied
extensively in the microbial world and form the basis of
much of what we know about the cycling of carbon in the
environment (Weinstein and Albersheim, .1979). Lysozyme, an
enzyme which hydrolyzes bacterial cell wall carbohydrates,
was one of the first such enzymes to be studied extensively
and as a result much is known about this enzyme's catalytic
mechanism and structure (Quiocho, 1986). The cellulases
have also been extensively studied as they are responsible
for the cycling of the most abundant natural polymer
cellulose (Knowles et al.. 1987). Studies of the
mechanistic properties of these enzymes have been
facilitated by the use of molecular genetic techniques.
Gene cloning, sequencing, and oligonucleotide-directed
mutagenesis have allowed mutations to be made in a site
specific manner. A kinetic study of the mutant proteins can
then be done and predictions about catalytic mechanisms
tested.
54


150
Varel, V. H. 1987. Activity of fiber-degrading microorganism
in the pig large intestine. J. Anim. Sci. 65:488-496.
Ward, 0. P., and M. Moo-Young. 1989. Enzymatic degradation
of cell wall and related plant polysaccharides, p. 237-274.
In CRC CRitical Reviews in Biotechnology vol 8.
Weinstein, L., and P. Albersheim. 1979. Structure of plant
cell walls. Plant. Physiol. 63:425-432.
Whitehead, T. R., and R. B. Hespell. 1990. The gene for
xylan-degrading activities from Bacteroides ovatus are
clustered in a 3.8 kilobase region. J. Bacteriol. 172:2408-
2412.
Wong, K. Y., L. U. L. Tan, and J. N. Saddler. 1986.
Purification of a third distinct xylanase from the
xylanolytic system of Trichoderma harzianum. Can J.
Microbiol. 32:570-576.
Yaguchi, M., C. Roy, C. F. Rollin, M. G. Paice, and L.
Jurasek. 1983. A fungal cellulase shows sequence homology
with the active site of hen egg-white lysozyme. Biochem.
Biophy. Res. Comm. 116:408-411.


117
(Robson and Gamier 1988). The binding of the
arabinofuranoside may also induce a conformational change in
the enzyme which somehow leads to the protection of the
histidyl moiety from deprotonation at high pH ranges.
It should be noted that the effective pKa of these amino
acids can vary markedly from the intrinsic pKa due to the
electonic environment in the active center of the protein.
The possibility that a different amino acid is
contributing a proton at the higher pH ranges cannot be
ruled out. Tyrosyl residues (pKa 10) in this region of the
protein could possibly substitute for histidine as a general
acid at the higher pH ranges. This phenomenon might be
explained if the two activities were catalyzed by different
mechanisms that in turn involve different amino acids
present in the same active center. This was shown to be the
case with the neopullulanase from Bacillus
stearothermophilus in which two separate enzymatic
activities were found to reside in the same active center
but were catalyzed by different amino acids and thus
different mechanisms (Kuriki et al. 1991). In an earlier
work, The B-D-xylosidase from Bacillus pumilus was found to
hydrolyze a-D-xylosylfluoride by a mechanism that was
separate from that for the hydrolysis of B-D-xylosylfluoride
(Kasumi et al. 1987). The number and nature of the active
centers is not known for this protein.