n... . ..

        

 

    

'9«~~- "Hug-u".

    
 

...,..-.....

 

 

 

 

 

 

 

 

 

 

 

   

 

 

. > ‘ ’ r r > ‘ . . I A. ‘ . ; ‘
‘1
' -' ‘) .‘

 

 

 

. .r
l

 

«111111:WT?mm'I will ”/19 0‘! g7 9‘7

31293 00574 4515

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

LIBRARY
Michigan State
”Mum

 

This is to certify that the

dissertation entitled

BIOCHEMICAL AND MOLECULAR CHARACTERIZATION
OF GLUCOSE ISOMERASE FROM THERMOANAEROBES

presented by

Chanyong LEE

has been accepted towards fulfillment
of the requirements for

ph .D . degree in Biochemistry

 

 

 

Date October 31, 1989

 

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or boron one due.

   

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ll

MSU Is An Affirmative Action/Equal Opportunity Institution

 

 

 

BIOCHEMICAL AND MOLECULAR CHARACTERIZATION

OF GLUCOSE ISOMERASE FROM THERMOANAEROBES

By

Chan yong Lee

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Biochemistry

1989

l

4

~—

(00545

ABSTRACT

BIOCHEMICAL AND MOLECULAR CHARACTERIZATION

OF GLUCOSE ISOMERASE FROM THERMOANAEROBES

By

Chanyon g Lee

Biochemical and molecular structural properties of glucose isomerase were
investigated in Clostridium thermosulfurogenes and Thermanaerobacter strain 86A.
Both organisms produced intracellular thermostable glucose isomerase, and synthesis
of the enzyme was induced by xylose. In addition, Thermoanaerobacter produced
intracellular glucogenic amylase and B-galactosidase activities constitutively, which
were environmentally compatible with glucose isomerase activity (i.e. active at the
same pH and temperature). The xylose grown Thermoanaerobacter cells produced
these three enzymes simultaneously, and converted starch or lactose directly into a
fructose syrup mixture in a single step process. Glucose isomerases were purified to
homogeneity from C. thermosulfurogenes and Thermoanaerobacter, and both enzymes
displayed very similar physicochemical and enzymatic properties. Long and flat
crystals were formed from glucose isomerase purified from C. thermosulfurogenes.
The gene (xyIA) encoding therrnostable glucose isomerase of C. thermosulfurogenes
was cloned and expressed constitutively in both Escherichia coli and Bacillus subtilis.

The recombinant glucose isomerase produced in these mesophilic hosts retained the

To

My parents, wife, and sons

with love

iii

ACKNOWLEDGMENT

I have been extremely fortunate to meet the right people at the right place and
time that have made the completion of this dissertation possible. I would like to
thank all of those and only a few are named here due to the limited space.

I would like to express my sincere appreciation to Professor J. Gregory Zeikus,
my scientific advisor, philosophical mentor, and competitive racket ball partner, for
his guidance, encouragement, and patience through my graduate school years. I
would also like to thank all of the members of my guidance committee, Drs. John
Wilson, Arnold Revzin, Robert Hausinger, and Michael Bagdasarian, who taught me
molecular genetic techniques, for their advice and invaluable time.

I would like to acknowledge the support of my co-workers and scientists who
shared techniques, ideas, and friendship; Drs. R. Lamed, S. Lowe, B. Saha, V.
Morales, Mr. J. Kemner, S. Mathupala, Y. Lee, W. Wu, M. Meng, and former and
all present lab members from the time in Madison, Wisconsin till now.

I would like to thank Michigan Biotechnology Institute for providing me with
nice facilities, scientific equipment, and financial support.

Finally, I am most grateful to my parents, brother, and sister for their
dedication and encouragement, and to Eunha, Maximilian, and Edmond who shared
the happiness and frustration, and strengthened me with love and smile. I owe this

dissertation to all of them, "My Big Family".

iv

TABLE OF CONTENTS

page
LIST OF TABLES .................................... iix
LIST OF FIGURES .................................... x
ABBREVIATIONS ................................... xiii
INTRODUCI ION ..................................... 1
CHAPTER 1. LITERATURE REVIEW ...................... 5
Literature Review .............................. 6
References .................................. 27
CHAPTER II. REGULATION AND CHARACTERIZATION OF
Thermoanaerobacter GLUCOSE ISOMERASE IN RELATIONSHIP
TO THERMOSTABLE SACCHARIDASE SYNTHESIS AND
DEVELOPMENT OF A SINGLE STEP PROCESS FOR
SWEETENER PRODUCTION ........................ 33
Abstract ................................... 34
Introduction ................................. 35
Materials and Methods .......................... 37
Results .................................... 41
Discussion .................................. 59
Literature Cited ............................... 62

CHAPTER III. PURIFICATION AND CHARACTERIZATION OF
THERMOSTABLE GLUCOSE ISOMERASE FROM Clostridium

thermosulfurogenes and Thermaanaerabacter ................ 65
Abstract ................................... 66
Introduction ................................. 67
Materials and Methods .......................... 69
Results .................................... 74
Discussion .................................. 95
Literature Cited ............................... 98

CHAPTER IV. CLONING AND EXPRESSION OF THE Clostridium

thermosulfurogenes GLUCOSE ISOMERASE GENE IN Escherichia

cali and Bacillus subtilis ............................ 102
Abstract ................................... 103
Introduction ................................. 104
Materials and Methods .......................... 106
Results .................................... 114
Discussion .................................. 129
Literature Cited ............................... 132

CHAPTER V. MOLECULAR CHARACTERIZATION OF THE
XY LOSE (GLUCOSE) ISOMERASE GENE FROM Clostridium

thermosulfurogenes. ROLE OF HIS101 IN ENZYMATIC

vi

CATALYSIS ................................... 135

Abstract ................................... 136
Introduction ................................. 138
Material and Methods ........................... 140
Results and Discussion .......................... 151
References .................................. 177
CHAPTER VI. CONCLUSIONS AND PERSPECTIVES .......... 180

vii

LIST OF TABLES

page
Chapter I
1. Some properties of glucose isomerase ................ 8
Chapter II
1. Cellular location of saccharidase activities in
Thermoanaerobacter strain B6A .................... 42
2. Effect of growth substrate on saccharidase synthesis
in Thermoanaerobacter strain B6A ................... 43
3. Effect of reaction conditions on the final monosaccharide
product ratio during a starch or lactose conversion
process with a Thermanaerobacter enzyme preparation ..... 57
Chapter [H
1. Biochemical properties of glucose isomerase acrivity in
crude cell extracts of therrnoanaerobes ................. 75
2. Summary of glucose isomerase purification steps ......... 76
3. Amino acid composition of thermostable glucose isomerase
from C. thernwsulfurogenes and Thermoanaerobacter ....... 80
4. Comparison of thermostable glucose isomerase kinetic
properties of therrnoanaerobes ..................... 90
5. Effect of metals on activity and thermal stability of
EDTA-treated glucose isomerase from C. thermosulfurogenes . . . 93

viii

Chapter IV
1. Comparison of glucose/xylose isomerase activities in
E. coli xyIA mutant strains carrying the recombinant
plasmid pCGI38 .............................. 117

2. Expression of cloned glucose isomerase by recombinant
shuttle plasmid pMLGl in E. coli and B. subtilis ......... 118

3. Effect of D-xylose and D-glucose on synthesis of recombinant
thennostable glucose isomerase in E. coli and B. subtilis . . . . 119

4. Purification scheme of cloned thermostable glucose isomerase
from E. coli and B. subtilis ....................... 126

Chapter V

1. Expression of C. thermosulfurogenes xylA gene by recombinant
plasmid in E. coli. ............................. 152

2. Comparison of site-directed mutant glucose isomerase activities
in E. coli carrying different alleles of C. thermasulfurogenes
xylA gene .................................. 169

3. Effect of diethylpyrocarbonate (DEPC) on glucose isomerase
activity of the wild type and site-directed mutant enzymes . . . 171

4. Comparison of kinetic properties of site-directed mutant and
wild type xylose isomerases ........ . ............... 172

ix

LIST OF FIGURES

page
Chapter I

1. Schematic illustration of the proposed catalytic mechanism

of glucose isomerase ........................... ll
2. Schematic diagram of the three dimensional structure of the

active site in Streptomyces rubiginosus xylose isomerase ...... 14
3. Schematic process biochemistry of enzymatic fructose production

from starch .................................. 16

Chapter II

1. Effect of glucose addition on glucose isomerase synthesis

during xylan fermentation ........................ 45
2. Effect of pH control on saccharidase production during xylose

fermentation ................................. 47
3. Comparison of temperature optima for activity (A) and

thermostability (B) of glucose isomerase, amylase, and

B—galactosidase activities ......................... 50
4. Comparison of pH optima for activity (A) and pH

stability (B) of glucose isomerase, amylase, and

B—galactosidase activities ......................... 52
5. Single step conversion of starch (A) and lactose (B)

into a fructose mixture by a T hermaanaerobacrer enzyme

preparation ................................. 55

Chapter III

1.

Electrophoretic analysis of glucose isomerases purified from

Thermoanaerobacter (A) and C. thermosulfurogenes (B) .....

Comparison of N-tenninal amino acid sequence of purified
glucose isomerases ...........................

Effect of temperature on glucose isomerase activity ........

Effect of pH on thermostable glucose isomerase activity (A)

and stability (B) ..............................

Thermostability of EDTA-treated glucose isomerase from
C. thernwsulfurogenes in the presence or absence of

metal ions ..................................

Crystals of glucose isomerase from C. thernwsulfurogenes

Chapter IV

1.

2.

Indicator plate for thermostable glucose isomerase activity . . .
Physical and genetic maps of plasmids pCGIBS and pMLGl,
which carry the C. thermosulfurogenes DNA insert expressing

thermostable glucose isomerase ....................

Comparison of native versus recombinant glucose isomerase
thermal stabilities ............................

SDS-PAGE analysis of glucose isomerase activity fractions . .
Time course of glucose conversion into fructose using

heat-treated glucose isomerase obtained from recombinant
B. subtilis .................................

xi

77

83

85

88

.91

108

115

Chapter V

Deletion mapping of the xylose isomerase gene (xylA)
of C. rhermosulfurogenes ........................ 142

Sequencing strategy for DNA encoding C. thermasulfurogenes
glucose isomerase ............................ 144

Synthetic oligonucleotide primers for site-directed
mutagenesis ................................ 147

Nucleotide and deduced amino acid sequence of
C. thermosulfurogenes xyIA gene .................. 153

Comparison of the amino acid sequences of different
xylose isomerases ............................ 158

Summary of amino acid sequence homology between different
xylose isomerases ............................ 160

Comparison of hydropathy profiles of xylose isomerases from
C. thermosulfurogenes, B. subtilis, and S. violaceoniger . . . . 163

SDS-PAGE comparison of xylose isomerases expressed by
different alleles of C. thermosulfurogenes xylA gene ...... 167

Plot of relative log of apparent Vmax versus pH for
Gln1m mutant and wild type xylose isomerases .......... 174

xii

EDTA
SDS
PAGE

MOPS
DEPC
PEG

KDa
D.E.
O.D.
FPLC

HPLC

ABBREVIATIONS

base pair

ethylenediamine tetraacetic acid
sodium dodecyl sulfate
polyacrylamide gel electrophoresis
isopropyl-B-D-thiogalactopyranoside
3-[N-morpholino]propanesulfonic acid
diethylpyrocarbonate
polyethyleneglycol

molecular weight

kilo-dalton

dextrose equivalence

optical density

fast performance liquid chromatography

high pressure liquid chromatography

xiii

INTRODUCTION

Use of microbial enzymes in human life is an ancient art. By experience and
empirical methods, it has developed to a highly sophisticated state; wine processing,
cheese manufacture, and preparation of fermented oriental food are good examples.
Recently, microbial and biochemical science with modern technology has made it
possible to develop the enzyme industry at a larger scale. At present, the two major
industrial enzymes are saccharidases and proteases which are used in starch processing
and detergent manufacturing industries, respectively. Glucose isomerase is the most
important commercial saccharidase and it is utilized in conversion of glucose into
fructose to produce high fructose corn syrup as a food sweetener. The industrial
process for sweetener production requires a glucose isomerase that is stable at high
temperature and low pH, and available at low cost. Many studies have been made
on glucose isomerase from meSOphilic microbial sources for enzyme production, its
physicochemical and enzymatic properties, and three dimensional structure.

Essentially, very little is known about the physiological, biochemical, and
molecular genetic features of glucose isomerases from thermophilic bacteria that are
known to produce intrinsically thermostable and therrnoactive enzymes (e.g., amylase,
cellulase, protease, or alcohol dehydrogenase). Therefore, an interest was generated
from the concept that certain thermoanaerobes may possess enzymatic features unique
for glucose isomerase activity. Furthermore, the molecular mechanism for high
thermostability of glucose isomerase remains to be identified and understood.

The research presented in this thesis was initiated to understand the biochemical

2
and physiological properties of glucose isomerase from thermoanaerobic bacteria; to
investigate the molecular mechanisms that account for enzymatic catalysis and high
thermostability of the enzyme in comparison with thennolabile glucose isomerases;
and to assess the biotechnological potential of the thermostable glucose isomerase
from therrnoanaerobes for novel sweetener production processes.

This thesis is composed of six chapters including a literature review, two
manuscripts that were submitted for publication in Applied and Environmental
Microbiology, a manuscript which was submitted for publication in the Journal of
Bacteriology, a paper which was submitted for publication in the Journal of Biological
Chemistry, and a final chapter that deals with concluding remarks and direction of
future research.

The literature review in Chapter 1 describes previous and present knowledge
about glucose isomerase in relation to: general biochemical properties, genetic
organization, structure-function relationships and the industrial process for fructose
production. In addition, the general properties and molecular mechanisms employed
by thermophilic enzymes; and, the general biochemical features of other microbial
saccharidases that are involved in industrial sweetener production are described. This ‘
chapter provides the rationale for detailed studies on both fundamental and applied
aspects of thermophilic glucose isomerase from thermoanaerobic bacteria.

Chapter 11, "Regulation and characterization of Thermoanaerobacter glucose
isomerase in relation to thermostable saccharidase synthesis and development of a
single step process for sweetener production", describes the general physicochemical

properties and regulation of glucose isomerase from Thermoanaerobacter strain B6A.

3

Furthermore, these studies represent the first demonstration that a saccharidase mixture
produced by a single microorganism can be used to directly process starch or lactose
into a fructose sweetener.

Chapter III, "Purification and characterization of thermostable glucose isomerase
from Clostridium thermosulfurogenes and Thermoanaerobacter", compares the
biochemical and physicochemical properties of glucose isomerases purified from two
different thermoanaerobic bacteria in detail. The results indicate that these
microorganisms produce highly thermophilic glucose isomerases which display very
similar enzymatic and physicochemical properties. Crystals were formed from C.
thermosulfurogenes glucose isomerase which can be used for future X—ray
crystallographic studies to identify the three dimensional structure of this enzyme.
These findings advance fundamental understanding of the biochemical properties of
thermostable glucose isomerases and of the evolutionary relationship between these
two different Species of thermoanaerobic bacteria.

Chapter IV, "Cloning and expression of the Clostridium thermosulfurogenes
glucose isomerase gene in Escherichia coli and Bacillus subtilis", demonsuates cloning
of the gene encoding thermophilic glucose isomerase from C. thermosulfurogenes.
This is the first glucose isomerase gene from a thermophilic bacterium that was
cloned and over-expressed in E. coli and B. subtilis. This chapter also describes a
new screening method for thermostable glucose isomerase, based on a specific assay
that detects conversion of ' fructose to glucose on agar plates. The recombinant
glucose isomerase expressed in the mesophilic hosts displayed identical

physicochemical (i.e., thermophilicity and molecular weight) and enzymatic properties

4
to those of the native enzyme of C. thermosulfurogenes, and this unique feature made
it possible to use a simple heat treatment of crude cell extract as one of the most
efficient purification steps when the thermophilic glucose isomerase was produced in
B. subtilis a mesophilic and food-safe host. Finally, the potential of the recombinant
glucose isomerase for industrial application was assessed.

Chapter V, "Molecular characterization of the xylose (glucose) isomerase gene
from Clostridium thermosulfurogenes: Role of His101 in enzymatic catalysis",
determined the nucleotide sequence of C. thermosulfurogenes xylA gene and compared
the deduced amino acid sequence with those of other thermostable and thermolabile
glucose isomerases. To understand the molecular mechanism of enzymatic catalysis
and thennophilicity of the enzyme, several key amino acids were changed by site
directed mutagenesis. The experimental results indicated that His101 is the catalytic
residue and functions as a base catalyst during glucose (xylose) isomerization; and that
cystine disulfide bonds are not used in the protein molecule to achieve enzyme
thermostability. In the final chapter the utility of these findings are discussed in
relation to future research aimed at understanding structure function relationships of
glucose isomerase, and at protein engineering to redesign the molecule in order to

'achieve desired properties (i.e., higher acid stability, turnover number, and substrate

specificity).

Chapter 1.

LITERATURE REVIEW

LITERATURE REVIEW

A. Glucose Isomerase

Xylose isomerase (D-xylose ketol isomerase, EC 5.3.1.5) is an intracellular
enzyme that catalyzes the isomerization reaction between D-xylose and D-xylulose
during the xylose metabolism in various microorganisms. Because this enzyme can
also convert D-glucose to D-fructose, the enzyme is often referred to as glucose

isomerase (1).

[1] Physiological and Biochemical Properties

Production of glucose isomerase has been reported from a large number of
bacteria and actinomycetes that can grow on xylose as an energy source. Among
these, Streptomyces species have been studied most extensively and used as a source
of industrial enzymes (1-5). Production and properties of glucose isomerases from
Actinoplanes missouriensis, Bacillus coagulans, Lactobacillus (an anaerobic species),
Norcardia, and Arthrobacter species are also well characterized and some of them are
used in the sweetener industry (6-9). However, due to the commercial importance of
this enzyme, most of the industrial strains used as glucose isomerase producers have
not been disclosed or the information available is brief and concentrated in the patent
literature. The regulation of glucose isomerase synthesis and the final yield of
enzyme depends on the selected microorganism and its chosen growth conditions (i.e.
temperature, pH, mineral salts, and nutrients) (9). Most microorganisms require

xylose to induce glucose isomerase production and enzyme synthesis is catabolite

7
repressed by the presence of glucose in the fermentation medium. However, because
of the relatively high cost of xylose in industrial fermentation processes, isolation of
constitutive mutants have been attempted and culture conditions which allow glucose
isomerase production on cheaper xylan or hemicellulose substrates have been
investigated (11,12). Most glucose isomerase producing organisms characterized are
mesophiles except for reports on two aerobic, moderate thermophilic, Bacillus
stearothermophilus and B. coagulans (2,90). Therefore, by further screening. of
thermophilic organisms more thermostable glucose isomerases may be identified. In
the last few years many new species of extremely thermophilic bacteria, predominately
anaerobic species have been isolated (65).

General biochemical properties of glucose isomerase from representative
microbial sources are summarized in Table 1. Although numerous glucose
isomerases have been characterized for their enzymatic properties, very few of them
have been purified to homogeneity. Most purified glucose isomerases have molecular
weights ranging from 80,000 to 190,000 and are composed of 2 or 4 identical
subunits (10,13). The pH optima for different glucose isomerases reported lie in the
range of 6.5 to 9.5 (10). Temperature optima for enzyme activity and stability vary
greatly depending on the source of glucose isomerase. Glucose isomerase is a
metalloenzyme, and it requires a divalent cation such as Mn“, Co”, or Mg++ for both
catalytic activity and enzyme stability. Su'uctural and catalytic roles of metal ions
have been proposed for enzymatic activity of various glucose isomerases (14-16).
The reaction catalyzed by glucose isomerase follows Michelis-Menten kinetic

mechanism over a wide range of substrate concentrations. This enzyme

Table 1. Some PrOpcrties of Glucose Isomerase

 

 

Temp. pH Metal Ion Molecular No. Reference
Microorganism Optimum Optimum Requirement Weight of
C C) Sub-
.. units
Escherichia coli 45 6.0 -- 96.000 2 23,91
Bacillus subtilis ~50 7.5 --- --- -- This study
Lacrobacillus brevis 50 6.0-7.0 Mn“,Co” 195,000 --- 15
Arthrobacter strain B3728 =80 8.0 Mg“ 180,000 4 29
Ampullariella Sp. strain 3876 ~75 170.000 4 V 28
Streptomyces violaceoniger >70 7.5 Mg“ 160.000 4 26
Streptomyces Strain YT-S 8O 8.0-8.5 Mg“,Co“ 157,000 4 l
Streptomyces olivochromogenes 80 8.0-10.0 Mg”,Co“ 120.000 2 2,37
Streptomyces flavogriseus 70 7.5 Mg“,Co°‘ 171.000 4 12
Strcptomyces griseofuscus 85 8.5 Mg”,Co” 180.000 4 93,94
Streptomyces violaceruber 80 7.5-9.5 Mg“,Co“ --- ——- 5
Bacillus coagulans 75 7.0 Co“ 160.000 4 9O
Bacillus stearothermopillus 80 7.5-8.0 Co“ 130.000 --— 2
Actinoplanes missouriensis 90 7.0-7.5 Mg“,Co“ 80,000 2 6.38.92

 

9
displays a lower Km and higher V“ for xylose than for glucose indicating that the
physiological function for this enzyme is in xylose isomerization. The concentration
that between D-glucose and D—fructose is about 50:50 at 60°C. The equilibrium

ratio is more favorable towards fructose at temperatures higher than 60°C (13).

[2] Genetic organization and Primary Structure of xylA gene

In spite of the industrial importance of the enzyme, few glucose isomerase
genes have been cloned and characterized for their primary structure and genetic
organization. Regulation of xylose catabolism in Salmonella typhimurium and E. coli
involves an operon consisting of three structural genes; xyfI‘, xle, xylA, responsible
for xylose transport, xylulose kinase, and xylose isomerase, respectively (17-19).
Expression of these structural genes are under the positive control of a regulatory
gene, xle. The structural gene coding for xylose isomerase from E. coli and B.
subtilis have been cloned and sequenced (20,21); and, a comparison of their nucleotide
sequence showed about 50% homology (22). Overexpression of E. “coli xylose
isomerase by using tac or lac promoters in E. coli strains has been reported (23-25).
The glucose isomerase gene of Streptomyces violaceoniger and Ampullariella species
has been cloned in a Streptomyces host and in E. coli, respectively, and their deduced
amino acid sequences have been reported (26-28). Although nucleotide sequences
are not available, the amino acid sequences of Arthrobacter species have been

published (29).

10
[3] Catalytic Mechanism and Three Dimensional Structure

The proposed catalytic mechanism for glucose isomerization involves formation
of a cis—enediol intermediate via an intramolecular hydrogen transfer without proton
exchange with water (30,31). A schematic diagram of the proposed mechanism is
illustrated in Figure 1 (32). In the first step a base group of the enzyme initiates
ring opening of the a-pyranose form of glucose which is known to be the preferred
substrate to the B—anomer (33,34). This process may be assisted by a neighboring
acidic group. In the second step, another basic group in the active site abstracts the
proton from the Q atom of the substrate to facilitate the formation of a cis-enediol
intermediate. In the final step, the proton is transferred back to the Cl atom of the
substrate , and the first base and acidic groups again cooperate in the ring closing of
the intermediate to form the furanose product. Although the exact amino acid
residues which take part in catalysis at the active site are not proven, inhibition
studies on Streptomyces glucose isomerase activity by chemical modification suggested
the involvement of a histidine residue for catalytic activity of the enzyme (35).

The three dimensional structure of xylose isomerases have been derived from
general enzyme sources including Streptomyces rubiginosus, S. olivochromogenes,
Actinoplanes missouriensis, S. violaceoniger, S. albus, and Arthrobacter by X-ray
crystallographic studies at different range of resolutions (3639,29). Structure analysis
of S. rubiginosus enzyme crystals at 4A resolution indicated that the polypeptide chain
consists of two structural domains containing an eight stranded B-sheet and Ot-helix
configuration for the larger domain and a loop structure for the smaller domain which

overlaps with a larger domain in another subunit to form a tightly bound enzyme

11

Figure 1. Schematic illustration of the proposed catalytic mechanism of glucose

isomerase (32).

12

I
O

 

l3
dimer. The final tetrameric structure of native enzyme consists of two such dimers
(36). Recent determination of the fine three dimensional structure of this enzyme
complex with substrate (xylose) or designed-inactivator at 1.9A (40) suggested specific
amino acid residues for the putative active site and metal ion binding site (Figure 2).
The histidine residue at position 54 was suggested as the catalytic residue and base
catalyst, and this histidine is correctly placed to abstract a proton from the C1 or Q
atom of the subsu'ate. The predicted structure of the active site in Arthrobacter
xylose isomerase also indicated that the histidine residue at the same position 54 of
the enzyme may be the catalytic residue (29). However, no biological or functional
evidence for the proposed mechanism has been established nor have site directed

mutants been used to provide proof that a specific histidine is involved in catalysis.

[4] Industrial applications
Currently, glucose isomerase is one of the most important industrial enzymes
used in the food industry, as it is employed to convert glucose to fructose for the
production of high fructose corn syrup (HFCS) as a nutritive sweetener (41). In
1982, more than 200 tons of glucose isomerase were used in the sweetener business '
and 8 billion pounds of HFCS were produced in the United States alone (42).

The overall process for HFCS production starts with starch processing steps,
including liquefaction and saccharification, and ends with the isomerization step (43)
(Figure 3). In the first stage, (rt-amylase converts raw starch into malto dexuin (DE.
1015) by catalyzing a random hydrolysis of the oc-1,4 glycosidic linkages in starch.

Then these D.E. 10-15 oligo-saccharides are hydrolyzed to produce glucose with 90%

14

Figure 2. Schematic diagram of the three dimensional structure of the the active

site in Streptomyces rubiginosus xylose isomerse (40).

15

/

Asp 245 HN ASp 255
O \-N
/ _ O _/
I C\ \\ I, I
G|u181 o / o---Mn:+.—o
\ IO Glu217 I, \O
A3925?

\
\ /0~ \ I H O
C \\ \ l '2 \
Mn++

16

Figure 3. Schematic process biochemistry of enzymatic fructose production from

starch.

l7

 

 

 

 

 

 

Process Stage Enzyme pH Temperature Metal ions
Liquefaction Bacterial 6.0 - 80 - 1 20°C Ca+ +
Alpha-Amylase 7.0
Malto
Dextrln
(D.E. 10
-15)
(pH Adjustment
with Acid)
Sacchari- Glucoamylase 4.0 _ 55-so°c
iication 5.0
(Filter, pH Adjustment
Addition oi Metal
Cotactor)
Isomerization Glucose 7.0 53-50%; Mg + +
isomerase Mn" +
C0 + 4-

 

Glucose
and
Fructose
Mixture
(58:42)

 

 

 

18

conversion from starch by glucoamylase. which liberates B-glucose consecutively
from the non-reducing ends of the oligo-saccharides. Prior to the saccharification
step, pH and temperature are adjusted to optimize the activity and stability of
glucoamylase. Glucose syrup produced in the second stage is filtered and refined
to remove Ca“ ions and other by-products which are inhibitory to glucose isomerase
activity. Isomerization of glucose syrup by glucose isomerase produces a
fructose/glucose mixture (42:58%), and this final step also requires a re-adjustment of
pH and the addition of metal cofactors. To generate the final sweetener product
containing 55% fructose, which was judged to be acceptable to replace the sucrose
used in food and beverage industry, the 42% fructose syrup is separated by large-
scale chromatographic columns, and then blended with an appropriate amount of
original feedstock (44). Industrial glucose isomerases are the most expensive starch
processing enzymes and they are used in the immobilized form to both prolong
enzyme half-life by enhancing thermostability and to allow use of a continuous flow
reactor system (45,46). Thus, the high process cost for glucose isomerase might be
lowered by using a new source of enzyme with high temperature stability and activity.
The current indusuial process is limited to 60°C because of undesirable by-products ‘
and color formation during the glucose isomerization reaction at high temperature and
the alkaline pH required for enzymatic stability. If the pH optimum of the enzyme
activity could be lowered below 6.5 and temperature optimum raised to 70°C, while
enzyme stability is retained, faster reaction rates during isomerization and higher final
fructose concenu'ations at equilibrium could be attained. Furthermore, these novel

enzymatic process improvements could reduce the viscosity of the glucose feedstock

19

and eliminate pH re-adjustments made after the saccharification step.

B. Other saccharidases used in food industry

An extensive variety of microbial enzymes are used in the food industry, and
the principal enzymes include amylase for starch processing, glucose isomerase for
sweetener production, B-galactosidase for milk produCIs and whey processing,
pectinase for wine or fruit juice clarification, and protease for meat processing and
cheese manufacturing (47). Saccharidases are the mosr important food-indusuial
enzymes employed to degrade polysaccharide to oligosaccharides or monosugars, and
to convert these sugars into the final desirable form of saccharides in food products.
This section will review only the enzymatic aspect of microbial amylases and B-

galactosidase.

[1] Amylase
Amylases are starch degrading enzymes that are widespread in microorganisms,
plants, and animals. The term amylase was used to designate enzymes that catalyze
the hydrolysis of a-l,4 glucosidic linkages of polysaccharides such as starch or
glycogen (48). Starch is the principal storage carbohydrate of plants and its major
commercial source in the US. is com. Starch consists of two high molecular weight
components: amylose, a linear glucose polymer which contains (rt-1,4 glucosidic

linkages; and, amylopectin, a branched polymer which contains a-l,6 glucosidic

20
linkages between linear (rt-1,4 glucose chains (49). Microbial amylases such as or-
amylase, glucoamylase, and starch debranching pullulanase have been widely used in

starch processing indusuies.

(a) (rt-amylase: a-amylases catalyze a random hydrolysis of the (rt-1,4
glucosidic linkages in amylose, amylopectin, and glycogen in an endo-fashion. The
end products of enzymatic hydrolysis are oligosaccharides of varying chain lengths
with or without branched points with the (rt-configuration. at the reducing glucose unit.

or-amylases are able to by-pass 0t-1,6 branch points in starch and the enzyme action
results in a rapid decrease in viscosity of starch solution and in the iodine staining
power (50). tat-amylases are generally stable in the pH range 5.5-8.0, and optimal
activity of the enzymes occurs between pH 5.5 and 6.5 (51). These enzymes have
been classified as metalloenzymes and they are stabilized by the presence of calcium
ions (48). There are two different types of enzymes, saccharifying or-amylase and
liquefying or-amylase which differ in their hydrolysis limits on soluble starch (52,53).
Thermostable liquefying (rt—amylases isolated from Bacillus species are used in the

first liquefaction step during starch processing for sweetener production (Figure 2).

(b) Glucoamylase: Glucoamylase is an exo-acting enzyme that yields B-D-
glucose from the non-reducing ends of amylose, amylopectin, and glycogen by
hydrolysis of (rt-1,4 glycosidic linkages in a consecutive manner. It also hydrolyzes
0t-l,6 and (1-1,3 linkages at a much slower rate than Ot-l,4 linkages. Glucoamylases

are found in fungi and the enzymes used commercially originate from either

21

Aspergillus niger or Rhizopus species (54), where they are used for the conversion of
malto-oligosaccharides into D-glucose (Figure 2). The pH and temperature optima
of most glucoamylases are in the range of 4.5-5.0 and 40-60°C, respectively (55).

All fungal glucoamylases are glycoproteins containing 5-20% carbohydrate in the
enzyme molecule. Glucoamylases differ from a—glucosidases in their subsuate
specificity and the stereo configuration of the glucose product (56). Ot-glucosidase
catalyzes the hydrolysis of or-l,4 linked or-D-glucose residues from the non-reducing
ends in short chain oligosaccharides or maltose to release ct-D-glucose as a final

product.

(c) Pullulanase: Pullulanase is a starch debranching enzyme that hydrolyzes
0t-1,6 glucosidic linkages in pullulan, amylopectin or glycogen (57). The enzyme
converts pullulan into maltotriose in an endo-fashion (58), and isomaltose, panose, or
amylose are not degraded by pullulanase (59). The pH optima for pullulanases were
reported to be 5.9-7.0, and the temperature optima of the enzyme ranges between 45-
60°C (55). When starch is hydrolyzed with or—amylase, the branch points of
amylopectin are resistant to attack, and the prolonged action of (rt-amylase on starch '
results in the formation of or-limit dextrin. Glucoamylases, on the other hand, are
able to hydrolyze the (rt-1,6 linkages of starch, but the reaction proceeds relatively
slowly due to their low affinity for the branch points. Therefore, pullulanases are
generally used in combination with the saccharifying or-amylase, or glucoamylase to
improve the saccharification and glucose yields from starch. It is interesting to note

that recent studies on new pullulanases purified from thermoanaerobic bacteria

22
indicated that both or-1,6 linkages in pullulan and (rt-1,4 linkages in amylose could be

cleaved by a new enzyme type, amylopullulanase (60).

[2] B-galactosidase

B-galactosidase hydrolyzes the B-D-galactosidic linkage of lactose to generate
an equimolar mixture of glucose and galactose. In practice, depending on the
conditions, the equimolar pattern is sometimes not followed because B-galactosidase
displays reversion activity and can form allolactose and other oligosaccharide from
galactose (61). Temperature and pH optima for this enzyme activity differs
according to the enzyme source. In general, fungal B-galactosidases have pH optima
in the acid range (2.5-4.5) and bacterial enzymes are in the neutral region (6.5-7.5)
(62). Although the temperature optima of most B—galactosidases examined range
between 50-60°C, one thermostable enzyme from Thermus aquaticus, an aerobic
species, was reported to have an optimal temperature at 80°C (63). Lactose is a
disaccharide found in milk products and by-products (whey), and is an important
dietary carbohydrate. Hydrolysis of lactose during the processing of milk based
products can potentially solve the digestion problems of lactase deficient adults, ’
improve the solubility and decrease the tendency of crystallization during ice cream
and yogurt production, and if further converted to fructose by other enzymes require

less additive sugars in these dairy products (64).

23

C. Thermostable Biocatalysts

The recent interest in biotechnology coupled with the discovery of new, novel
therrnophiles has prompted studies on the general properties of their enzymes,
understanding molecular mechanisms of enzyme thennophilicity, and the utilization
of them for indusuial purposes (65). As thermostable biocatalysts, thermophilic
bacteria and/or their enzymes have potential for industrial applications due to their
unique kinetic and stability properties, and because operation at high temperature often
minimizes the risk of microbial contamination and lowers capital cost for large scale
production with improved mass transfer rates for certain substrates and products
(66,67). In this section, the microbial and enzymatic aspects of thermophilicity will

be reviewed as it relates to the research presented in this thesis.

[1] Thermophilic bacteria

Extremely thermophilic bacteria display temperature optima for growth at 60°C
to above 100°C, but do not proliferate at temperature below 40°C. Extreme
therm0philes often possess faster metabolic rates and produce more active and stable
enzymes than those found in either mesophilic bacteria, yeasts, fungi, plants, or
animals (66). A variety of thermophilic saccharolytic anaerobes have been isolated
from self-heating (manure piles and wet soils) or volcanic features in nature
(68,69,70,71). These thermoanaerobic species display diverse properties in terms of
growth characteristics, substrates utilized, metabolic pathways, enzymes, and

fermentation products.

24

Clostridium thermosulfitrogenes is a Gram-negative, spore forming rod that has
a double-layered wall without an outer membrane layer (72). This novel species can
utilize pectin as an energy source and transforms thiosulfate into elemental sulfur
which deposits in the culture medium and on the cells. The organism ferments a
variety of carbohydrates at 60°C such as xylose, galactose, glucose, starch, maltose,
cellobiose, and sucrose; and it forms ethanol, acetate, lactate, and H,/CO2 as
fermentation products. It also produces a highly thermostable and active B—amylase
when grown on starch (73).

Thermoanaerobacter strain B6A is a Gram-negative, non-spore forming rod
that utilizes a very wide range of carbohydrates as a carbon source including xylan,
xylose, starch, and glucose, but not cellulose (74). It can grow in a chemically
defined medium at pH 3.5 and 60°C. Fermentation products from either glucose or
xylan include ethanol, acetate, lactate, and HJCOZ.

Clostridium thermohydrosulfuricum is a spore forming rod and its endospore
displays extreme heat resistance (75). This species grows optimally at temperature
of 65°C, and utilizes a broad range of subsu'ates as energy sources including xylose,
starch, galactose, and cellobiose. Starch grown cells produce an amylopullulanase '
that can cleave both or-l,4 and a—l,6 glucosidic linkages of amylose and pullulan,

respectively (60).

[2] Thermophilic enzymes
There has been increasing research interest in enzymes from thermophilic

microorganisms during the last decade. Thermophilic enzymes are very active and

25
diSplay long half-life at elevated temperature ( > 60°C), and are less active but more
resistant to most common protein denaturants than their counterparts from mesophilic
sources at lowered (< 40°C) temperatures (76,77). A variety of thermostable
enzymes have been studied and characterized for their physicochemical and catalytic
properties, and many of them are now used extensively in industrial processes
(78-80). In general, high thermostability of enzyme from extreme thermophiles is
an inuinsic property, specified by the amino acid sequence and certain key amino
acid residues in the protein secondary and tertiary structure (81-84). A major
molecular mechanism proposed for thennophilicity of enzymes is explained on the
basis of higher free energy of stabilization (A G) within the protein molecule, which
represents a delicate balance between the stabilizing forces and the conformational
enu'opy of the protein required to maintain the tertiary structure of the enzyme
(85,86). Small percentage changes in either the stabilizing or destabilizing forces can
result in large changes in the net free energy of stabilization, which explains how
high thermostability of a protein can be conferred by small differences in the amino
acid sequence that give rise to a few additional intramolecular interactions without any
"obvious" su'uctural alterations (82,87). Recently, protein engineering approaches '
have used site directed mutagenesis techniques to insert amino acid changes into
enzymes in order to demonstrate the effect of protein chemical structure on enzyme
thermostability (88,89). However, a limitation to the successful design of a
thennophilic-thennostable protein by engineering techniques is that the effects of
amino acid replacements are not easy to predict and thus it is difficult to decide

which amino acid substitutions should be made. Therefore, an understanding of the

26
exact structure-function mechanisms that account for activity and stability relationships
in thermophilic enzymes could be essential for engineering high thermostability and
activity in proteins. Unfortunately, no specific explanations on the molecular
mechanism of enzyme thennophilicity can be made at this time because only a
handful of reports exist on understanding how to make a thermolabile enzyme stable
at high temperature or why a thermophilic enzyme is active and stable at high

temperature.

10.

11.

12.

13.

14.

15.
16.

REFERENCES

Takasaki, Y., Kosugi, Y., and Kanbayashi, A. (1969) Agric. Biol. Chem.
33, 1527-1534

Suekane, M., Tamura, M., and Tomimura, C. (1978) Agric. Biol. Chem. 42,
909-917

Chen, W.P., and Anderson, A.W. (1979) Appl. Environ. Microbiol. 38, 1111-
1119

Kasumi, T., Hayashi, K., and Tsumura, N. (1981) Agric. Biol. Chem. 45, 619-
627

Callens, M., Kersters-Hilderson, H., Opstal, O.V., and De Bruyne, CK. (1986)
Enzyme Micro. Technol. 8, 696—700

Scallet, B.L., Shieh, K.,Ehrenthal, I., and Slapshak, L. (1974) Die Starke 26,
405-444

Danno, G. (1970) Agric. Biol. Chem. 34, 1795-1804

Yamanaka, K., and Takahara, N. (1977) Agric. Biol. Chem. 41, 1909-1915
Chen, W. (1980) Process Biochem. 15, 30-35

Chen, W. (1980) Process Biochem. 15, 36-41

Srinivasan, M.C., Vartak, H.G., Powar, V.K., and Khire, J.M. (1983) Biotech.
Lett. 5, 611-614

Chen, W.P., Anderson, A.W., and Han, Y.W. (1979) Appl. Environ. Microbiol.
37, 324-331

Anuim, R.L., Collila, W., and Schnyder, BJ. (1979) in Applied Biochemistry
and Bioengineering (Wingard, L.B., ed) vol. 2, pp. 97, Academic Press, New
York

Dano, G. (1971) Agric. Biol. Chem. 35, 997-1006

Yamanaka, K. (1969) Archiv. Biochem. Biophys. 131, 502-506

Young. J.M., Schray, KL, and Mildvan, A.S. (1975) J. Biol. Chem. 250, 9021-
9027

27

17.

18.
19.
20.

21.
22.
23.

24.

25.

26.

27.

28.

29.
30.

31.

32.

33.

28
Lawlis, V.B., Dennis, M.S., Chen, E.Y., Smith, DH, and Henner, DJ. (1984)
Appl. Environ. Microbiol. 47, 15-21
Shamanna, D.K., and Sanderson, KB. (1979) J. Bacteriol. 139, 71-79
Ghangas, GS, and Wilson, DB. (1984) J. Bacteriol. 157, 158-164
Schellenberg, G.D., Sarthy, A., Larson, A.E., Backer, M.P., Crabb, J.W.,
Lidstrom, M., Hall, B.D., and Furlong, CE. (1984) J. Biol. Chem. 259, 6826-
6832
Wilhelm, M., and Hollenberg, GP. (1984) EMBO Journal 3, 2555-2560
Wilhelm, M., and Hollenberg, GP. (1985) Nucleic Acids Res. 13, 5717-5723

Batt, C.A., O’Neil, E., Novak, S.R., Ko, J., and Sinskey, A. (1986)
Biotechnology Progress. 2, 140-144

Panagiotis, ES, and Ho, N.W.Y. (1985) Enzyme Micro. Technol. 7, 592-596

Wovcha, M.G., Steuerwald, D.L., and Brooks, K.F. (1983) Appl. Environ.
Microbiol. 45, 1402-1404

Marcel, T., Drocourt, D., and Tiraby, G. (1987) Mol. Gen. Genet. 208, 121-
126

Drocourt, D., Bejar, S., Calmels, T., Reynes, J.P., and Tiraby, G. (1988)
Nucleic Acids Res. 16, 9337

Saari, G.C., Kumar, A.A., Kawasaki, G.H., Insley, M.Y., and O’Hara, RI.
(1987) J. Bacteriol. 169, 612-618

Henrick, K., Collyer, C.A., and Blow, D.M. (1989) J. Mol. Biol. 208, 129-157

Rose, I.A., O’Connell, EL, and Mortlock, RP. (1969) Biochem. Biophysic.
Acta 178, 376-379

Ramchander, S., and Feather, MS. (1977) Arch. Biochem. Biophys. 178, 576-
580

Suckling, K.E. (1985) in Comprehensive Biotechnology (Moo-Young, M., Bull,
A., and Dalton, H. ed) vol.l, pp. 539-551, Pergarmon Press, Oxford

Schray, K.J., and Rose, LA. (1971) Biochemistry 10, 1058-1062

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

29

Makkee, M., Kieboom, A.P.G., and van Bekkum, H. (1984) Reel.: J. R. Neth.
Chem. Soc. 103, 361-164

Gaikwas, S.M., More, M.W., Vartak, H.G., and Deshpande, V.V. (1988)
Biochem. Biophys. Res. Commun. 155, 270-277

Carrell, H.L., Rubin, B.H., Hurley, T.J., and Glusker, LP. (1984) J. Biol. Chem.
259, 3230-3236

Farber, G.K., Petsko, GA, and Ringe, D. (1987) Protein Engineering 1, 459-
466

Rey, F., Jenkins, J., Janin, J., Lasters, 1., Alard, P., Claessens, M., Matthyssens,
G., and Wodak, S. (1988) Proteins 4, 165-172

Dauter, Z., Dauter, M., Hemker, J., Witzel, H., and Wilson, KS. (1989) FEBS
Letters 7, 1-8

Carrell, H.L., Glusker, J.P., Burger, V., Manfre, F., Tritsch, D., and
Biellmann, J.-F. (1989) Proc. Natl. Acad. Sci. USA 86, 4440-4444

Lambert, P.W., and Meers, LL. (1983) Phil. Trans. R. Soc. Lond. 300,
263-282

Coker, L.B., and Venkatasubramanian, K. (1985) in Biotechnology Applications
and Research (Cheremisinoff, RN, and Ovellette, R.P., eds) pp. 165-171,

, Technorrric Publishing Co., Lancaster, Basel

Reichelt, JR. (1983) in Industrial Enzymology (Godfrey, T., and Reichelt, J.,
eds) pp. 375-396, The Nature Press

Bucke, C. (1983) Trends in Biotechnol. 1, 67-69

Verhoff, F.H., Boguslawski, G., Lantero, O.J., Schlager, S.T., and Jao, Y.C.
(1985) in Comprehensive Biotechnology (Moo-Young, M., ed), vol.3, pp. 837-
859, Pergamon Press. Oxford, New York

Schnyder, 3.]. (1974) Die Starke 26, 409-412

Peppler, H.J., and Reed, G. (1987) in Biotechnology (Kennedy, J.F. ed), vol.
7a, pp. 547-603, VCH Publishers, NY

Fischer, E.H., and Stein, EA. (1960) in The Enzymes (Boyer, P.D., Lardy, L.,
and Myrback, K. eds), vol. 4, pp.3l3-343, Academic Press, NY

French, D. (1975) in Biochemistry of Carbohydrates (Kornberg, H.L., Phillips,

50.

51.
52.
53.
54.

55.

56.

57.

58.
59.

61.
62.
63.

65.

66.
67.

30
D.C., and Whelan, W.J. eds), vol. 5. PP. 267-335

Fogarty, W.M., and Kelly, CT. (1980) in Microbial Enzymes and
Bioconversions (Rose, A.H. ed), pp. 116-158, Academic Press

Manning, G.B., and Campbell, LL. (1961) J.Biol. Chem. 236, 2952-2957
Fukumoto, J., and Okada, S. (1963) J. Ferment. Technol. 41, 427-434

Pazur, J.H., and Okada, S. (1966) J. Biochem. 241, 4146-4151

Underkofler, L.A., Denault, L.J., and Hon, E.F. (1965) Die Starke 17, 179-184

Fogarty, W.M. (1982) in Microbial Enzymes and Biorechnology (Fogarty, W.M.
ed), pp. 1-92, Applied Sciece Publishers, NY

Reese, E.T., Maguire, A.H., and Parrish, F.W. (1968) Can. J. Biochem. 46, 25-
34

Lee, E.Y.C., and Whelan, W.J. (1971) in The Enzymes (Boyer, P.D. ed), vol.
5, pp. 191-234, Academic Press, NY

Drummond, G.S., Smith, E.E., and Whelan, W.J. (1969) FEBS Lett. 5, 85-88
Bender, H., and Wallenfels, K. (1966) Methods in Enzymology 8,555-559

Saha, B.C., Mathupala, SR, and Zeikus, JG. (1988) Biochem. J. 252, 343-
348

Shukla, T. (1975) CRC Critical Reviews in Food Technology 1, 325-347
Gekas, V., and Lopez-Leiva, M. (1985) Process Biochem. February, 2-12

Ulrich, J.T., McFeters, G.A., and Temple, KL. (1972) J. Bacteriol. 110, 691-
698

Pedersen, S., and Norman, BE. (1986) in Chemical Aspecs of Food Enzymes
(Andrews, A.T. ed), pp.156-187, Royal Society of Chemistry

Brock, TD. (1986) in Therrnophiles; General, Molecular, and Applied
Microbiology. (Brock, T.D. ed), pp. 1-16, John Wiley and Sons, New York

Zeikus, JG. (1979) Enzyme Microb. Technol. 1, 243-252

Zeikus, JG. (1980) Ann. Rev. Microbiol. 34, 423-464

68.
69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.
82.

83.

31
Ben-Bassat, A., and Zeikus, J.G. (1981) Arch. Microbiol. 128, 365-370

Weimer, P.J., Wagner, L.W., Knowlton, S., and Ng, T.K. (1984) Arch.
Microbiol. 138, 31-36

Ljungdahl, L.G., Bryant, F., Carreira, L., Saiki, T., and Weigel, T.
(1981) in Trends in the Biology of Fermentation for Fuels and
Chemicals. (Lollaender ed.), pp. 397, Plenum Press. New York

Zeikus, J.G., Ben-Bassat, A., and Hegge, P.W. ( 1980) J. Bacteriol. 143,
432-440

Schink, B., and Zeikus, J.G. (1983) J. Gen. Microbiol. 129, 1149-1158

Hyun, H.H., and Zeikus, J.G. (1985) Appl. Environ. Microbiol. 49, 1162-
1167

Weimer, RI. (1985) Arch. Microbiol. 143, 130-136

Hyun, H.H., Zeikus, J.G., Longin, R., Millet, J., and Ryten, A. (1983)
J. Bacteriol. 156, 1332-1337

Zuber, H. (1978) in Biochemistry of Thennophily (Friedman, S.M. ed),
pp. 267-285, Academic Press, New York.

Amelunxen, RE, and Murdock, AL. (1978) in Microbial Life in
Extreme Environments (Kushner, D.J. ed), pp. 215-278, Academic Press,
London.

Ng. T.K., Kenealy, W.R. (1986) in Thermophiles; General, Molecular,
and Applied Microbiology. (Brock, T.D. ed), pp. 197-215, John Wiley
and Sons, New York

Klibanov, A.M. (1983) in Applied Microbiology (Laskin, A.I. ed), vol.29,
pp. 1-28, Academic Press, New York.

Schmid, RD. (1979) in Advances in Biochemical Engineering (Fiechter,
A. et al. eds), vol.12, pp. 41-118, Spring-Verlag, Berlin.

Frank, G., and Tratschin, JD. (1979) Biochemistry 18, 5698-5703

Daniel, RM. (1986) in Protein Structure, Folding, and Design (Oxender,
D.L. ed), pp. 291-296, Alan R. Liss, Inc., New York.

Fontana, A. (1988) Biophys. Chem. 29, 181-193

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

32
Vihinen, M. (1987) Protein Engineering 1, 477-480

Creighton, TE. (1984) in Protein; Su'ucture and Molecular Properties, pp.
133-158, W.H. Freeman and Co., New York.

Dill, K.A. (1987) in Protein Engineering (Oxender, D.L., Fos, C.F. eds),
pp. 187-192, Alan R. Liss, Inc., New York.

Holmes, M.A., and Matthews, B.W. (1982) J. Mol. Biol. 160, 623-639

Fersht, A.R., Leatherbarrow, R.J., and Wells, TN. (1986) Nature 322,
284-286

Beasty, A.M., Hurle, M., Manz, J.T., Stackhouse, T., and Matthews, CR.
(1987) in Protein Engineering (Oxender, D.L., Fox, C.F. eds), pp. 91-
102, Alan R. Liss, Inc., New York.

Danno, G., Yoshimura, S., and Natake, M. (1967) Agric. Biol. Chem.
3, 284-292

Tucker, M.Y., Tucker, M.P., Himmel, M.E., Grohmann, K., and Lastick,
SM. (1988) Biotechnol. Lett. 10, 79-84

Gong, C.S., Chen, L., and Tsao, GT. (1980) Biotechnol. Bioeng. 22,
833-845

Kasumi, T., Hayashi, K., and Tsumura, N. (1981) Agric. Biol. Chem. 45,
619-627

Kasumi, T., Hayashi, K., Tsumura, N., and Takagi, T. (1981) Agric.
Biol. Chem. 45, 1087-1095

Chapter II

Regulation and Characterization of Thermoanaerobacter
Glucose Isomerase in Relation to Thermostable Saccharidase
Synthesis and Development of a Single Step Process

for Sweetener Production

33

ABSTRACT

Regulation of glucose isomerase synthesis was studied in Thermoanaerobacter
strain B6A which fermented a wide variety of carbohydrates including glucose, xylose,
lactose, starch, and xylan. Glucogenic amylase activities and B-galactosidase were
produced constitutively, whereas the synthesis of glucose isomerase was induced by
either xylose or xylan. Production of saccharidase was not significantly repressed
by the presence of glucose or 2-deoxy-glucose in the growth media. In order to
maximize glucose isomerase production, the culture pH was controlled at 5.5 during
xylose fermentation. The apparent temperature and pH optima for these cell-bound
saccharidase activities were: glucose isomerase 80°C, 7.0-7.5; glucogenic amylases
70°C, 5.0-5.5; and B-galactosidase 60°C, 6.0-6.5. Glucose isomerase, glucogenic
amylases, and B-galactosidase activities were produced in xylose grown cells that were
active and stable at 60°C and pH 6.0-6.5. Under single step process conditions,
these enzyme activities in whole cells or cell extracts converted starch or lactose
directly into fructose mixtures. Starch was 96% converted into a 49:51 mixture of
glucose and fructose; whereas lactose was 85% converted into a 40:31:29 mixture of '

galactose, glucose, and fructose.

34

INTRODUCTION

The production of sweetener from corn starch by microbial saccharidases is an
important application of enzyme technology in the food industry (7). The current
process for high fructose corn syrup production involves several separate enzymatic
steps including liquefaction by (rt-amylase, saccharification by glucoamylase, and
isomerization by glucose isomerase. These steps require different reaction conditions
(i.e. temperature, pH, and metal cofactors) (1,3). Most industrial saccharidases used
in sweetener production require high thermostability and low production costs, and
they are produced by mesophilic microbes. Many studies have been made on
searching for more thermostable saccharidases and on immobilization of glucose
isomerase to prolong enzyme half-life in the reactor (14,19,28). Also, many efforts
have been made on strain improvement for hyper-producers of saccharidases and on
selection for the use of catabolite resistant or constitutive mutants to lower enzyme
production costs (12,24).

Recently, our laboratory reported Clostridium thermohydrosulfuricum produce
highly thermostable amylolytic enzymes including glucogenic amylase, pullulanase,
and glucosidase activities (14-16). When purified to homogeneity, this pullulanase
was shown to cleave both a-1,6 bonds in pullulan and tit-1,4 bonds in amylose and
has been named amylopullulanase (23), and Thermoanaerobacter strain B6A has the
same enzyme activities (B.C. Saha, R. Lamed, C. Lee, S.P. Mathupala, J.G. Zeikus,
manuscript submitted to Appl. Environ. Microbiol.). Although glucose isomerase and

B-galactosidase have been detected in thermophilic microorganisms (6,11,20), little

35

36

is known about the biochemical properties and regulation of these enzymes in
thermoanaerobic bacteria. In this paper, we report on the regulation and general
biochemical properties of glucose isomerase in relation to other saccharidase activities
from T hermoanaerobacter strain B6A. Notably, growth conditions were discovered
that enabled glucose isomerase to be produced in conjunction with glucogenic
amylases and B-galactosidase activities. Thus, a single step process was developed
to produce fructose mixtures from starch or lactose with Thermoanaerobacter

saccharidases.

MATERIALS AND METHODS

Chemicals and Gases

Medium components and all chemicals were reagent grade. Larchwood xylan
(Lot 113F-0003) was purchased from Sigma Chemical Co. (St. Louis, Mo.). The
N,,/CO2 (95:5) gas was obtained from Linde, Union Carbide Corporation (East

Lansing, MI) and passed through heated copper columns to remove 02 prior to use.

Organism

Thermoanaerobacter strain B6A isolated from a volcanic hot spring in
Themopolis, Wyoming (27) was obtained from Dr. Paul Weimer (USDA Dairy Forage
Lab, Madison, WI) and was maintained by stringent anaerobic culture techniques (29)

in CM5 medium containing 0.5% xylan.

Culture Conditions, and Growth Measurement

Experimental cultures were grown at 60°C without shaking in either 125 ml
serum bottles or in 26 ml anaerobic pressure tubes that contained 50 ml or 10 ml of
TYE medium (29) supplemented with 1% carbon source and with a N,/CO2 (95:5) gas
headspace. Enzyme production time course studies during xylose fermentation were
conducted in a New Brunswick (Edison, NJ.) Multigen fermentor that contained 500
ml of TYE medium. The ferrnentor cultures were incubated at 60°C and gassed
continuously with N2/CO2 (95:5) gas and mixed at 100 rpm. For the constant pH

experiments, 0.5N NH.,OH was added during the fermentation by a feeding pump

37

38

which was connected to a pH controller. Cells used for crude enzyme extract
preparations were cultured in a 14 liter New Brunswick fermentor containing 10 L
of TYE medium at pH 6.8 with 1% xylose that was stirred at 60°C under a NleO2
(95:5) gas stream.

Cell growth in media containing soluble substrates was determined by measuring
the optical density of the culture broth at 660 nm. When xylan was present in
culture medium, ethanol concentration in culture supernatant was used to measure
growth. Ethanol was measured in acidified samples by gas chromatography using
a flame ionization detector with N2 as the carrier gas and methods described elsewhere
(16). Experimental ethanol production was related to growth by standard optical

density (O.D.) versus ethanol plots determined with xylose medium.

Enzyme Preparations and Assays

For the preparation of the washed cells and culture supernatants, cultures were
harvested during late exponential growth phase by cenuifugation at 8,000 x g for 15
min. The supernatant was decanted, and the cells were washed with double distilled
water and suspended in the appropriate amount of double distilled water.

Cell extract was prepared using fermentation grown cells recovered at the
exponential phase (11 h) with a Millipore pellicon cell harvester (Bedford, MA.) and
washed with double distilled water. Wet cell paste (1 g) was suspended in either
50 ml of 20 mM sodium phosphate buffer (pH 7.0) containing 10 mM MgCl2 for
glucose isomerase preparation, in 50 ml of 100 mM sodium acetate buffer (pH 5.5)

for glucogenic amylase, or in 50 ml of 100 mM sodium phosphate buffer (pH 6.0)

39

for B—galactosidase. The cells were disrupted by passage through a French pressure
cell (American Instrument Co. Inc., Silver Spring, MD) at 20,000 lbfrn2 and the
supernatant was collected after centrifugation at 15,000 x g for 30 min at 4°C.
Protein concentration was detennined by the Lowry assay method (18). The crude
saccharidase preparation was stable upon incubation indicating the lack of significant
protease activity. Xylanase activities were assayed by measuring the rate of reducing
sugar formation from xylan; whereas, amylase is reported as glucose or reducing
sugars produced from starch or pullulan (when indicated). The reaction mixture
contained 1% subsuate in 0.1 M sodium acetate buffer at pH 6.0, 5.5 or 5.0 with
pullulan, soluble starch or xylan as respective substrates. After 30 min incubation
at 65°C, the reaction mixtures were boiled in a steam bath for 5 min to stop the
reaction. The samples were cenuifuged and the amount of reducing sugars were
quantified by the diniu'osalicylic acid method (21). Alternatively, glucose in the
supernatant was determined by using either a glucose analyzer (Yellow Stone
Instrument, model 27) or by a Sigma enzymatic glucose diagnostic kit. One unit of
activity is defined as the amount of enzyme which released 1 moi of reducing sugar
per min under the assay conditions described above, with glucose as standard for
amylase activity and with xylose as standard for xylanase activity.

Glucose isomerase activity was measured by incubating a reaction mixture that
contained 0.8 M glucose in 0.1 M sodium phosphate or 0.1 M MOPS buffer (pH 7.0)
and 10 mM MgSO., 1 mM C00,, and the enzyme. After incubation at 65°C for
30 min, the amount of fructose formed was estimated by the cysteine carbazole

sulfuric acid method (8). B-galactosidase activity was assayed by measuring the

4O
amount of phenoxide ion liberated from ortho-nitrophenol-B-D-galactopyranoside
(ONPG) at 420 nm after 20 min incubation at 60°C. The reaction mixture contained
10 mM KCl, 1 mM MgSO., 5 mM 2-mercaptoethanol, 2.7 mM ONPG with the
enzyme in 100 mM sodium phosphate buffer (pH 6.0) (22).

One unit of amylase, glucose isomerase, and B—galactosidase activity is defined
as the amount of enzyme required to produce 1 umol of glucose, fructose, and
phenoxide ion, respectively, per min under the assay conditions described above. All
the enzyme activities were determined at points where product formation was linear

with time.

One Step Conversion of Starch or Lactose to Fructose

Starch (i.e., maltodextrin DE 10 and soluble starch) or lactose in 0.1 M sodium
phosphate buffer containing 10 mM MgSO4 and 1 mM CoCl, was incubated with cell
extracts prepared from xylose grown cells at various temperatures (60-70°C) and pH’s
(6.0-6.8). Samples were taken during time course experiments performed in 50 ml
serum vials containing 5 ml reaction mixtures that were sealed with rubber bungs, and
were shaken at 100 rpm in a New Brunswick water bath shaker. Samples withdrawn
from the reaCtion mixture were boiled in a steam bath for 5 min and centrifuged
before sugar analysis. Quantitative and qualitative analyses were performed by high
pressure liquid chromatography with saccharide analysis columns heated to 85°C.
An Aminex HPX-87C and HPX-87P column (Bio-Rad Laboratories; Richmond, CA)

were used for analysis of starch and lactose conversion, respectively.

RESULTS

Location and Types of Saccharidase Activities

Thermoanaerobacter strain B6A ferments starch and hemicellulose but not
cellulose(27). In preliminary experiments, halos appeared around colonies grown on
either Remazol Brilliant Blue-xylan agar plates or on starch agar plates stained with
iodine indicating that the organism produced amylase and xylanase activities.

Experiments were initiated to determine the kinds of saccharidases and their
cellular location when Thermoanaerobacter was grown on different saccharides as
carbon and energy sources for growth. Table 1 shows that Thermoanaerobacter
produces glucose isomerase, B-galactosidase as well as glucogenic amylase and
xylanase activities. Glucogenic amylase activity was extracellular and cell bound
when starch or pullulan was used as subsuate. It was not possible to discern
whether the glucogenic amylase represented a mixture of a-amylase and glucoamylase
or amylopullulanase and or-glucosidase by the assays used. However, About 70-
80% of xylanase activity was excreted into the medium and the remaining activity
was cell associated in xylan grown cultures. On the other hand, glucose isomerase,

and B-galactosidase activities were t0tally cell bound.

Regulation of Glucose Isomerase Production
Experiments were conducted to determine the regulation of glucose isomerase
synthesis in relation to amylase and B-galactosidase producrion. The results presented

in Table 2 illustrate that glucose isomerase was only produced when either xylose or

41

42

Table l. Cellularglocation of saccharidase activities in Thermoanaerobacter strain B6A.‘I

 

Activity (unit/ml)

 

 

Carbon Final Growth
Saccharidase Substrate (OD. 660 nm) Supernatant Washed Cell"
Glucose
Isomerase xylose l. 15 0.00 0.21
B-galacto-
sidase lactose 0.98 0.00 0. 14
Amylase
starch 1.07 0.04 0.19
pullulan 1.07 0.73 0.43
Xylanasec xylan 0.80 0.47 0. 18

 

‘Cultures were grown in TYE medium with 1.0% carbon substrate at 60°C.

”Cells were resuspended in 1/10 culture volume of double distilled water and activity was
converted to the original culture volume.

cDue to the interference of xylan optical absorbance, cell growth was converted from ethanol
production in relation to a standard growth curve.

43

Table 2. Effect of growth substrate on saccharidase synthesis in Thermoanaerobacter
strain B6A.

 

Specific Activity (U/mg cell protein)

 

 

' Final
Growth" Glucose Glucogenic B—galactosidase

Growth Substrate‘ (O.D.6«,) Isomerase Amylase

Starch 1.36 0.00 0.61 0.41
Lactose 0.83 0.00 0.54 0.46
Maltose 0.75 0.00 0.43 0.38
Cellobiose 1.23 0.00 0.59 0.44
Glucose 1.60 0.00 0.42 0.31
Xylose 1.47 0.62 0.60 0.47
Xylan 0.68 0.39 0.58 0.33
Xylose + Glucose 1.46 0.36 0.48 0.34
Starch + Glucose 1.24 0.00 0.46 0.35
Lactose + Glucose 1.00 0.00 0.50 0.40
Xylose + 2-Deoxy-Glc 0.80 0.40 0.38 0.33
Starch + 2-Deoxy-Glc 0.95 0.00 0.48 0.42
Lactose + 2-Deoxy-Glc 0.46 0.00 0.30 0.33
Glucose + 2-Deoxy-Glc 0.98 0.00 0.33 0.25
2-Deoxy-Glc 0.1 1 n.d. n.d. n.d.
None 0. 10 n.d n.d. n.d.

*

n.d. = Not Determined.
'Cultures were grown on TYE medium containing 0.5% main substrate and with or without
0.3% supplementing glucose or 0.1% 2-deoxy-glucose (2-Deoxy-Glc).

bOptical density was measured at the early stationary growth phase.

44

xylan was present as an inducer in the culture medium. Notably, glucogenic amylase
and B-galactosidase were produced constitutively on the wide range of growth
substrates tested. The specific activities of glucogenic amylases from starch grown
cultures and B-galactosidase from lactose grown cultures were equal to those obtained
from xylose grown culture. The presence of 0.1% 2-deoxy-glucose in culture media
containing either 0.5% starch, xylose, or lactose did not repress synthesis of either
glucogenic amylase, glucose isomerase, or [Ll-galactosidase. respectively.

In order to assess the mechanism of glucose isomerase synthesis, experiments
were performed where glucose was added during xylan fermentation time courses and
glucose isomerase activity was periodically assayed. Due to the turbidity of xylan,
cell growth was monitored by measuring the concentration of ethanol in the culture
broth. Ethanol was one of the major end products of Thermoanaerobacter
fermentations (27) and its exponential production during growth is proportional to cell
density measured by absorbance at 660 nm during xylose fermentation. These results
shown in Figure 1 indicate that synthesis of glucose isomerase during exponential
growth was not catabolite repressed by glucose, while xylanase production ceased after
glucose addition during exponential growth on xylan (data not shown).

Figure 2 compares glucose isomerase and amylase aetivity levels during xylose
fermentations in the absence (A) and presence (B) of pH control at 5.5. Synthesis
of both enzyme activities were tightly growth coupled under either condition.
Glucose isomerase activity, however, decreased rapidly in the stationary phase cultures
that were not maintained at pH 5.5. Amylase and B-galactosidase (data not shown)

were quite stable throughout the stationary phase. Consequently, pH control at 5.5

45

Figure 1. Effect of glucose addition on glucose isomerase synthesis during xylan
fermentation.

Cultures were grown on CM5 medium with 0.5% xylan at 60°C, and 0.3%
glucose was added during the middle of exponential growth phase (8.). Ethanol
concentration in culture broth was measured as a growth indicator and 25 mM ethanol
represents CD. 1.0 at 660 nm in a standard growth curve. The closed symbols
represent the culture growth whereas open symbols represent enzyme activity. The
culture conditions indicated by symbols are: triangles, glucose only (control); circles,

without glucose addition; squares, with glucose addition.

46

 

N
01

N
0

GROWTH (mM,Ethanol produced)
5

 

Growth (control)
N

 

 

 

_ 7
Growth
(+Glcl
Growth
Glc I-Glcl ..
" (Lil-(31c):
G.Ht-Glc)
G.HCOMtOll
A A A A A 1 A
4 6 8 IO 12 l4 IS 20

Time ( hours)

.0
N

0.15

.0

.0
O
or

'( 1mm) Aigxiiov asoratuosl asoonig

47

Figure 2. Effect of pH control on saccharidase production during xylose
fermentation.

Cultures were grown in 500 ml TYE medium containing 2% xylose in a
Multigen fermentor at 60°C. A: without pH control; B: pH was controlled at 5.5 by

feeding 0.5 N NI-LOH. Symbols represent glucose isomerase (G.I.) and amylase (A.).

48

maize >923. 2:3: maize >35: 2:3:

0000 7.6. 4.321.
7.54. 0050 0.0

 

20
1

16

14

J.

«(5.1.
r
12

 

 

 

20

8 10 12 14 16
Timelhoursl

6

 

 

 

 

 

m
U
m
l O I‘
v I e
m
.. 8 T
.. 6
i 4
2
B __ _ . _ _
A _ _ _ 1O 0 O. 0.
7654rc4. w m m m6.5.4 4. 3.. 2 l
Amnesia AEcommddv £265 3:: 31a 250309 5380

47

Figure 2. Effect of pH control on saccharidase production during xylose
fermentation.

Cultures were grown in 500 ml TYE medium containing 2% xylose in a
Multigen fermentor at 60°C. A: without pH control; B: pH was controlled at 5.5 by

feeding 0.5 N NI-LOH. Symbols represent glucose isomerase (G.I.) and amylase (A.).

48

 

 

 

 
 
 

(Iw/m Kiww owizua

 

 

 

 

 

 

0.4 '

 

Time (hours)

A
35 7. 77.0
55 .0
r: so 50
a 4 40
E 4. -
§ ‘00 0.7
‘0. 10- 0.6
3 K A. 05
3 <— 6.1. 03
65 tq- 0.2
0.1
1 1 l 1 l
2 4 6 8 IO 12 I4 16 20
Time (hours)
B
23: 7.0
g e.
I 5.
£14. ... 4.0
,1 T
E40—
8 0.7
g 3.0— 0.6
9' 0.5
£ 20-
E» 03
0 “1— 0.2
0.1
1 1 1 1 4 1 J _
2 4 6 e 10 12 I4 16 20

49
was used in xylose fermentations to obtain Thermoanaerobacter enzyme preparations

that contained active levels of amylase, glucose isomerase, and B-galactosidase.

Physicochemical Properties of Saccharidases

The temperature and pH activity and stability profiles for glucose isomerase
and B-galactosidase were compared to the amylase activities in Thermoanaerobacter
exuacts prepared from washed whole cells grown on xylose. Glucose isomerase,
amylase, and B-galactosidase displayed an apparent temperature optimum for activity
between 75-80°C, 70°C, and 65°C, respectively (Figure 3A). The effect of
temperature on the stability of these enzyme activities is shown in Figure 3B. Cell
exu'acts for enzyme assays were pre-incubated prior to measuring the residual a
activities in 100 mM sodium phosphate buffer (pH 7.0) containing 10 mM MgSO,
and 1 mM CoCl2 for glucose isomerase, in 100 mM sodium acetate buffer (pH 5.5)
for amylase, and in 100 mM sodium acetate buffer (pH 6.0) for B-galactosidase.
Under these conditions, glucose isomerase, amylase, and B-galactosidase were stable
up to 60 min at 85°C, 70°C, and 60°C, respectively.

Figure 4 illustrates the dependence of saccharidase activities on pH. Glucose
isomerase displayed a broad pH range for activity from 5.5 to 9.0 with an apparent
pH optimum of 7.0-7.5. During the enzyme assay at alkaline pH, chemical
isomerization of glucose to fructose by alkali caused a high background reading and
the amount of product formed by enzymatic isomerization was calculated by
subtracting the control value from the total experimental value. The amylase and B-

galactosidase activities displayed a narrower pH range for activity than that of glucose

50

Figure 3. Comparison of temperature optima for activities (A) and stabilities (B)
of glucose isomerase, amylase, and B-galactosidase.

Enzymes were assayed with cell extract from xylose grown cells. 100% activity
corresponds to 0.60 U/mg; 0.58 U/mg, 0.46 U/mg for glucose isomerase, amylase, B-
galactosidase, respectively. Cell extracts in 50 mM sodium phosphate buffer (pH 7.0)
for glucose isomerase, in 100 mM sodium acetate buffer (pH 5.5) for amylase, and
in 100 mM sodium phosphate buffer (pH 6.0) for B-galactosidase were preincubated

at the indicated temperatures prior to the assay for residual enzyme activities.

51

2;»: us: a; EDGEEE

 

        

 

 

 

 

 

 

 

 

 

 

 

O m CM 0% Din O.N O._ 00 _ 0.0 0.0 0.x. 0.0 0.0 0.? 0pm 0
0.2. . MN? .1 i cm
0 o 1 O».
0.8 e e e o e. . .4 0m

. .. u 8

O.\\1111O o o o o o o

0.3 Sov_mo.uo_oo-m o omenfiozgooi 1 oo_

1 ON

1 0*»

4 Om

1 Om
30:64 30164 r 00.

. u I m s a h a o

1 ON
l 04

l Om

J Om
0.3 39060.: 30020 3.89:3. 3320 109

 

 

 

 

.m .<

1%) AIIAIIOV BAIIV '138

52

Figure 4. Comparison of pH optima for activity (A) and stability (B) of glucose
isomerase, amylase, and B-galactosidase.

Enzyme activities were assayed with cell extracts in 100 mM glycylglycine
hydrochloride (H), sodium acetate (El—Cl), and sodium phosphate (O-O) buffers.

Residual activity was measured (B) after treatment at 60°C for l h.

53

:23 In

ox ow

0.0 or Own ON. O..O_ O.m

p
1

:23 :e

ow owe

0d Qv Oem

 

b
-

  

30336200 -Q

 

omolE<

h
d

 

  
  

omon_no.uo_oo-m

P

 

301:3

h
.-

10m
100_

ON
Ov
Ow
Om
OO_

 

 

 

 

 

 

 

$03602 3320

 

 

380602 .3820

1 OO_

 

 

. <

(°/e)AJ_1/\I_LDV BAIIVWBU

54
isomerase and the apparent pH optima were between 5.0-5.5 and 6.0-6.5, respectively.
The stability of these enzymes in relation to pH was examined by measuring residual
activities after incubation for 1 h at 60°C at different pH values (Figure 4B).
Under these assay conditions, glucose isomerase, amylase, and B-galactosidase
activities were stable at pH 6.0-7.5, pH 5.5-7.0, and pH 5.5-7.0, respectively. All
the saccharidase activities were stable and displayed at least 60% of maximal activity

at pH 6.0.

Single Step Conversion of Starch or Lactose into Fructose Syrups

The feasibility of using Thermoanaerobacter saccharidases in a single step
enzymatic process for producing fructose syrups from liquefied starch or whey-lactose
was examined using a crude extract of xylose grown Thermoanaerobacter cells which
contained environmentally compatible saccharidases. Figure 5 depicts a typical time
course for production of monosaccharides from starch or lactose using a single step
process with Thermoanaerobacter saccharidases at 60°C. More than 90% of the
starch or lactose was hydrolyzed into glucose mixtures within 4 h while the
isomerization of glucose into fructose approached equilibrium by the end of incubation '
period (20-48 h). An unknown oligosaccharide, putatively allolactose, was produced
at the early stages of lactose conversion and was slowly degraded into monomers.
Allolactose is a reversion product formed by B-galactosidase activity.

Table 3 compares the influences of specific reaction condition changes on the
final saccharide product ratio achieved during single step conversion of starch and

lactose by Thermoanaerobacter saccharidases. During enzymatic starch conversion,

55

Figure 5. Single step conversion of starch (A) and lactose (B) into a fructose
mixture by a Thermoanaerobacter enzyme preparation.
5% maltodexuin (A) and 5% lactose (B) were incubated with cell extract at 60°C

in 100 mM sodium phosphate buffer at pH 6.0 and 6.5, respectively.

Socchoride Conversion (°/e)

Soccnoride Conversion (°/e)

56

 

 

 

 

 

 

A. STARCH
100
80 o-Storch
1 GI
60‘: o . . '/ ”cf-3.:
'. ' V T"
l .
I ' .
4O . . a ‘ \Fructose
20", .
. \ Oligomers
IO 20 30 4O 50
Time (hrs)
8. LACTOSE
logr rL
501‘; ,1
‘,-—Loctose
40‘— \

Goloctosc
3. /
\
Q

30~ \

 

 

 

 

 

 

 

‘ / Glucose
~ G;
20 ‘ ‘ Fructose
/ Allolactose
i0 ‘~~~\-.T‘W -
Ks. _____
‘T‘N. ~~~~~~~~~~~ _.
o--- - 1— i 4 - - ---——t-~‘
IO 20 3O 4O 50

Time ( hrs )

 

57

Table 3. Effect of reaction conditions on the final monosaccharide product ratio during

a starch or lactose conversion process with a Thermoanaerobacter enzyme preparation.‘

 

 

Substrate Concenuation pH Temperature Conversion Final Product
(w/v) % (°C) Yield (%)" Ratio‘
(Glc:Frc)
Liquefied starch 5 6.0 60 92 53:47
6.0 65 93 5 1:49
6.0 70 96 49:5 1
20 6.0 60 . 84 58:42
20 6.0 70 9 1 5 1:49
(GalzGlczFrc)
Lactose 5 6.4 60 85 40:3 1 :29
6.8 60 81 40:31:29
6.8 65 76 35:33:32
20 6.4 60 75 36:33:31

 

‘Cell extract (5 mg and 10 mg/ml of reaction vol.) prepared from xylose grown cells was
added into the reaction mixture containing 5% and 20% substrate, respectively. Maltodextlin
(DE 10) was used as a liquefied starch.

”Total amount of monosaccharides hydrolyzed from the substrate.

cRelative ratio between monosaccharides was measured after 48 hours incubation under the

given reaction conditions.

58

a higher sweetener concentration ratio for fructose to glucose and a higher yield of
starch conversion into monosaccharides were achieved at 70°C than at 60°C.
Liquefied starch was used here but equivalent results were obtained with soluble
starch and the saccharidase preparation also hydrolyzed raw starch (data not shown).
The highest sweetener conversion (fructose-51 to glucose-49) was achieved at pH
6.0 and 70°C from 5% (w/v) maltodextrin with a final conversion yield of 96%.

Enzymatic hydrolysis of 5% lactose at pH 6.8 versus 6.4 lowered the total conversion
yield of lactose, but did not affect the final product ratio between galactose, glucose,
and fructose. Enzymatic hydrolysis of 20% lactose at 65°C versus 60°C lowered

both the lactose conversion yield and the galactose concentration in the final products.

DISCUSSION

To our best knowledge, these findings represent the first reported studies on
the general physicochemical properties and regulation of glucose isomerase from a
thermophilic microorganism. Although thermophilicity is required in industrial
glucose isomerase, thermostable enzymes produced by mesophilic bacteria have been
examined as principal indusuial sources (4). Furthermore, these studies represent the
first demonstration that a saccharidase mixture produced by a single microorganism
can be used to directly process starch or lactose into a fructose sweetener. Several
previous studies have demonstrated that glucoamylase and glucose isomerase mixtures
from different microbes can process oligodexuins into fructose syrups with marginal
success because of enzyme pH and thermal incompatibilities (10, 13).

Regulation of amylase and xylanase activities in Thermoanaerobacter strain
B6A is different from the inducible and glucose catabolite repressed synthesis of
amylase activities reported in C. thermosulfltrogenes and C. thermohydrosulfuricum
(14-16). Thermostable B-galactosidase has been previously reported in aerobic
Thermus species; however, enzyme synthesis is regulated by induction and glucose
catabolite repression. On the other hand, the B-galactosidase of Thermoanaerobacter
strain B6A is constitutive and non-catabolite repressible.

As expected of enzymes from thermophiles, the apparent optimum temperatures
for glucose isomerase (75-80°C), glucogenic amylases (70°C), and B—galactosidase
(65°C) in crude extracts were relatively high. The reported temperature optima of

glucose isomerases from other microorganisms vary with enzyme source, they range

59

60

from 45°C to 90°C (5). The temperature optima of glucogenic amylase activities
reported in Clostridium thermohydrosulfuricum (13) was similar to those in
Thermoanaerobacter. The temperature optima of B-galactosidases range from 35°C
to the high of 80°C reported in Thermus strains (6,9). Analysis of
Thermoanaerobacter saccharidase therrnostabilities showed that more than 95% of
glucose isomerase, glucogenic amylases, and B-galactosidase were retained after 60
min incubation at 85°C, 70°C, and 60°C.

The pH optima for glucose isomerase, glucoamylase, and B-galactosidase are
generally between 7.0 to 8.5, 4.5 to 5.5, and 4.5 to 7.5, respectively (5,14,9).
Although the apparent pH optima for Thermoanaerobacter glucose isomerase, B-
galactosidase, and glucogenic amylase activities fall in this range, all three enzymes
were stable and active at pH 6.0-6.5 and 60°C. The glucose isomerase displayed the
least acid stability and activity of the saccharidases studied. In a separate report
(manuscript submitted) we purified the Thermoanaerobacter glucose isomerase to
homogeneity and showed that the 200,000 MW tetrarner has pH and thermal
properties identical to those reported here. Thus, we conclude here that the rapid
destruction of glucose isomerase activity in non-pH controlled xylose fermentation is
due to enzyme instability at the low pH values and high temperature of stationary
phase cultures.

The difference in pH optima or temperature stability of these saccharidases has
been a major problem in demonstrating the feasibility of a single step process for
conversion of starch or lactose to fructose mixtures. The present study indicates that

saccharidases simultaneously produced by Thermoanaerobacter are environmentally

61
compatible at pH 6.0 and 60°C and can be used coordinately in a single step
conversion process for production of fructose sweetener from starch or milk derived
subsuates.

The final ratios between glucose and fructose during the single step conversion
process from starch at various temperatures were very similar to the theoretical values
at equilibrium of the glucose isomerization reaction (1,17). The maximum percent
of starch conversion from liquefied starch during this process was also very similar
to the values (94-96% of dextrose) after starch saccharification in the multistep
commercial processes (1). The Thermoanaerobacter saccharidase preparation could
be operated at 70°C to achieve the higher equilibrium concenu'ations of fructose and
faster reaction rates than those practiced in indusuial processes. If reduced to
practice, a single step starch hydrolysis process at high temperature (>60°C) and acid
pH (<pH 6.0) would lower costs for producing high fructose sweetener from corn
starch. The Thermoanaerobacter saccharidases provide a first stage feasibility model
for developments of such a process. Use of Thermoanaerobacter saccharidases can
also directly enhance sweetness in milk containing products by converting lactose to
fructose during the manufacturing process. Hydrolysis of lactose in these products
can potentially solve the digestion problems of lactase deficient adults, improve the
solubility during ice cream and yogurt production and require less additive sugars in
dairy products. Further studies on long term stability of the immobilized enzymes
and its food safety are required before any commercial value for Thermoanaerobacter

saccharidases can be proposed.

10.

ll.

12.

13.

LITERATURE CITED

Anuim, R. L., W. Colliala, and B. J. Schnyder. 1979. Glucose
isomeraseproduction of high-fructose syrups, p. 97-155. In L.B. Wingard
(ed.),Applied Biochemistry and Bioengineering, Academic Press, New York.

Aunstrup K., O. Andresen, E. A. Falch, and T. K Nielsen. 1979. Production
of microbial enzymes, p. 281-309. In Microbial Technology, 2nd ed., vol 1,
Academic Press.

Buke, C. 1980. Enzymes in fructose manufacture, p. 51-72. In G.G. Birch, N.
Blackebrough, and J.K. Parker (ed), Enzymes and Food Processing. Applied
Science Publishers, London.

Chen, W. 19803. Glucose isomerase (a review). Process Biochem. 15:30-35.
Chen, W. 1980b. Glucose isomerase (a review). Process Biochem. 15:36-41.

Cowan, D. A., R. M. Daniel, A. M. Martin, and H. W. Morgan. 1984. Some
properties of a B-galactosidase from an extremely thermophilic bacterium.
Biotechnol. Bioeng. 26:1141-1145.

Crueger, W., and A. Crueger. 1982. Enzymes, p. 161-174. In W. Crueger and
A. Crueger (ed.), Biotechnology: a textbook of indusuial microbiology, Science
Tech. Inc., Madison.

Dishe, Z., and E. Borenfreund. 1951. A new spectro-photometric method for
the detection and determination of keto sugars and trioses. J. Biol. Chem.
192:583-587.

Gekas, V., and M. Lopez-Leiva. 1985. Hydrolysis of lactose: a literature
review. Process Biochem. 20:2-12.

Guillaume, R., and P. Walon. 1977. Preparation of levulose from granular
starch. US Patent 4,009,074.

Goodman, R. E., and D. M. Pederson. 1976. B—galactosidase from Bacillus
stearothermophilus. Can. J. Microbiol. 22:817-825.

Hafner, E. W. 1985. Constitutive production of a thermostable glucose
isomerase. US. Patent 4,551,430.

Horwath, R. O., and R. M. Irbe. 1986. Process for preparing fructose from
starch. U. S. Patent 4,605,619.

62

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

63

Hyun, H. H., and J. G. Zeikus. 1985. General biochemical characterization of
thermostable pullulanase and glucoamylase from Clostridium
thermohydrosulfuricum. Appl. Environ. Microbiol. 49:1168-1173.

Hyun, H. H., and J. G. Zeikus. 1985. Regulation and genetic enhancement of
glucoamylase and pullulanase production in Clostridium thermohydrosulfuricum.
J. Bacteriol. 164:1146-1 152.

Hyun, H. H., G. J. Shen, and J. G. Zeikus. 1985. Differential amylosaccharide
metabolism of Clostridium thermosulfurogenes and Clostridium
thermohydrosulfuricum. J. Bacteriol. 164:1153-1161.

Jorgensen, O. B., L. G. Karlsen, N. B. Nielsen, S. Pedersen, S. Rugh, and
Bagsvaerd. 1988. A new immobilized glucose isomerase with high productivity
produced by a strain of Streptomyces murinus. Starch/Starke 40:307-313.

Lowry, O. H., J. Rosebrough, A. I. Farr, and R. J. Randall. 1951. Protein
measurement with the foline phenol reagent. J. Biol. Chem. 193:265-275.

Melasniemi, H. 1987. Characterization of (rt-amylase and pullulanase activities
of Clostridium thermohydrosulfuricum: Evidence for a novel thermostable
amylase. Biochem. J. 246:193-197.

Melasniemi, H. 1987. Effect of carbon source on production of thermostable
a-amylase, pullulanase and or-glucosidase by Clostridium
thermohydrosulfuricum. J. Gen. Microbiol. 133:883-890.

Miller, G. L. 1959. Dinitrosalicylic acid reagent for detection of reducing
sugar. Anal. Chem. 31:426-428.

Miller, J. H. 1972. Purification of B-galactosidase, p. 398-404. In J.H. Miller
(ed.), Experiments in Molecular Genetics. Cold Spring Harbor Laboratory,

Saha, B. C., S. P. Mathupala, and J. G. Zeikus. 1988. Purification and
characterization of a highly thermostable pullulanase from Clostridium
thermohydrosulfuricum. Biochem. J. 252:343-348.

Sanchez, 8., and C. M. Quinto. 1975. D-Glucose isomerase: constitutive and
catabolite repression-resistant mutants of Streptomyces phaeochromogenes. Appl.
Microbiol. 30:750-754.

Ulrich, J. T., G. A. McFeters, and K. L. Temple, 1972. Induction and
Characterization of B-galactosidase in an Extreme Therrnophile. J. Bacteriol.
1 10:691-698.

26.

27.

28.

29.

30.

64

Wiegel, J.W.K. 1986. Genus Thermoanaerobacter, p. 1379-1383. In J.G.
Holt, P.H.A. Sneath, N.S. Mair, and ME. Sharpe (ed), Bergy’s Manual
of Systematic Bacteriology, vol 2, 9th ed. Williams and Wilkins,
Baltimore.

Weimer, P. J., L. W. Wagner, S. Knowlton, and T. K. Ng. 1984.
Thermophilic anaerobic bacteria which ferment hemicellulose:
characterization of organisms and identification of plasmids. Arch.
Microbiol. 138:31-36.

Yokote, Y., K. Kimura, and H. Samejima. 1975. Production of high fructose
syrup by glucose isomerase immobilized on phenol-formaldehyde resine. Die
Starke. 27:302-306.

Zeikus, J. G., A. Ben-Bassat, and P. W. Hegge. 1980. Microbiology of
methanogenesis in thermal, volcanic environments. J. Bacteriol. 143:432-440.

Zeikus, J. G. 1979. Thermophilic bacteria: ecology, physiology, and
technology. 1979. Enzyme Microb. Technol. 1:243-252.

Chapter III

Purification and Characterization of Thermostable Glucose Isomerase

from Clostridium thermosulfurogenes and Thermoanaerobacter

65

ABSTRACT

Glucose isomerases produced by Thermoanaerobacter strain B6A and
Clostridium thermosulfurogenes strain 48 were purified 10- to 11-fold to homogeneity
and their physicochemical and catalytic properties were determined. Both purified
enzymes displayed very similar properties (native M: 200 kDa, tetrameric subunit
composition, and apparent temperature and pH optima at 80°C and pH 7.0-7.5). The
enzymes were stable at pH 5.5-12.0, and maintained more than 90% activity after
incubation at high temperature (85°C) for 1 hr in the presence of metal ions. The
N-terrninal amino acid sequence of both thermostable glucose isomerases were Met-
Asn-Lys-Tyr-Phe-Glu-Asn and were not homologous to that of the thermolabile
Bacillus subtilis enzyme. The glucose isomerase from C. thermosulfurogenes and
Thermoanaerobacter displayed isoelectric pH’s of 4.9 and 4.8, and their Km and

V“ values for D-glucose were 1040, 1260 rnin'1

and 140, 120 mM, respectively.
Both enzymes displayed higher K“ and lower K., values for D-xylose than those for
D-glucose. C. thermosulfurogenes enzyme required Co“ or Mg++ for thermal stability
and glucose isomerase activity, and Mn“ or these metals for xylose isomerase activity.

Crystals of C. thermosulfurogenes glucose isomerase were formed at room temperature

by the hanging drop method using 16%-18% PEG 4000 in 0.1 M citrate buffer.

66

INTRODUCTION

D-Glucose (D-xylose) isomerase (EC 5.3.1.5) is an intracellular enzyme found
in a number of bacteria which can utilize xylose as a carbon substrate for growth (8).
Although the physiological function of the enzyme in vivo is isomerization of D-
xylose to D-xylulose, this enzyme also converts D-glucose to D-fructose in vitro (32).
The latter activity of the enzyme is being used in industry for production of high
fructose corn syrup, and glucose isomerase is one of the largest volume commercial
enzymes used today (2,6).) Due to the indusuial significance of the enzyme, glucose
isomerases have been studied from various microorganisms and their catalytic and
physicochemical properties have been reviewed (9). Immobilization techniques and
a continuous isomerization process with the enzyme have also been described (18,29).

Most commercially available glucose isomerases are isolated from mesophilic
microorganisms including those from Streptomyces, Actinoplanes, F lavobacterium, and
Bacillus species. These enzymes are generally thermosrable, and are utilized in the
immobilized form to enhance enzyme half-life (35). These enzymes require metal
ions for their activity and stability, and the pH Optima for enzyme activity fall in the
range of 7.5-9.0. The reaction temperature used in the current industrial process for
sweetener production is limited to 60°C because of by-product and color formation
during reaction at high temperature and alkaline pH (7). However, operation of the
reaction at higher temperature than 60°C has several advantages to achieve faster
reaction rates, higher equilibrium concentrations of fructose, and reduced viscosity of

the substrate and product. Therefore, a novel glucose isomerase with a lower

67

68
optimum pH (neutral or slightly acidic) and high thermal stability might have practical
value for indusuial application.

Thermophilic microorganisms are known to produce intrinsically thermostable
enzymes which have been evolved and adapted to the extreme environment of their
natural habitat (1). Advantages of using these thermostable enzymes in industrial
processes were proposed (24, 40). Our laboratory has recently reported on
purification and biochemical characterization of extremely thermostable B-arnylase and
pullulanase from Clostridium thermosulfurogenes and C. thermohydrosulfuricum.
respectively (25, 30). In spite of the large amount of studies on glucose isomerases
from various enzyme sources, nothing is known about the detailed biochemical or
molecular genetic properties of glucose isomerases from thermoanaerobic bacteria.

Recently we demonstrated that a new Thermoanaerobacter species contained
an inducible xylose/glucose isomerase activity (C. Lee, B. Saba, and J. G. Zeikus,
manuscript submitted to Appl. Environ. Microbiol.). We demonstrate here that
thermostable glucose isomerases are also present in C. thermohydrosulfuricum and
C. thermosulfurogenes; and, report on the biochemical characterization of similar
glucose isomerases purified from C. thermosulfurogenes strain 43 and

Thermoanaerobacter strain B6A.

MATERIALS AND METHODS

Chemicals, Organisms, and Growth Conditions

Medium components and all other chemicals were reagent grade. C.
thermosulfurogenes strain 48 (28) and Thermoanaerobacter strain B6A (36) were
routinely grown at 60°C, and C. thermohydrosulfuricum strain 39E (39) was grown
at 65°C in anaerobic 26 ml pressure tubes or in 1 liter round bottom flasks that
contained TYE medium (39) supplemented with 1% xylose. For large-scale enzyme
preparation, cultures were grown at 60°C in a 14 liter Braun fermentor (West
Germany) containing 10 liter TYE medium with 2% xylose, and the culture pH was
maintained at 5.5 on line by control with l N NaOH. Cells were harvested at the
late exponential growth phase using a Pellicon cell harvester (Millipore Corp.,
Bedford, MA) and washed with 50 mM MOPS buffer (pH 7.0) containing 10 mM

MgSO, and 1 mM CoClz prior to freezing.

Enzyme Assays

Glucose isomerase activity was measured by incubating a 1 ml reaction mixture
that contained 0.8 M glucose, 10 mM MgSO,, 1 mM C00,, and an enzyme
preparation in 100 mM MOPS buffer (pH 7.0). For the assay of xylose isomerase
activity, the reaction mixture (1 ml) contained 70 mM xylose, 10 mM MnSO,, and
the enzyme preparation in 100 mM MOPS buffer (pH 7.0). After 30 min incubation
at 65°C, 1 ml of 0.5 M perchloric acid was added to stop the reaction and the

mixture was further diluted 50- and lO-fold with double disrilled water before the

69

70

quantitative analysis for fructose and xylulose, respectively. The amount of fructose
and xylulose was estimated by the cysteine-carbazole sulfuric acid method (14).

Fructose isomerizing activity was assayed under the same reaction conditions as the
glucose isomerase assay, except 0.4 M fructose instead of 0.8 M glucose was used
in the reaction mixture. After 30 min incubation, the reaction was terminated by
placing the assay tubes on ice, and the amount of glucose formed was estimated with
a glucose analyzer (Yellowstone Instrument, model 27). One unit of glucose
isomerase, xylose isomerase, and fructose isomerase activity is defined as the amount
of enzyme that produced 1 umol of fructose, xylulose, and glucose, respectively, per
min under the assay condition. Protein concentration was determined by the method

of Lowry (23) with bovine serum albumin used as the standard.

Purification of Glucose Isomerase
All the procedures were performed under aerobic conditions at 4°C.

(1) Preparation of cell extract: 50 g of cells of Thermoanaerobacter or C.
thermosulfurogenes were suspended in 200 ml of 50 mM MOPS buffer (pH 7.0)
containing 10 mM MgSO, and 1 mM C00,. The cells were disrupted by two
passages through a French pressure cell at 18,000 p.s.i. The cell debris was removed
by centrifugation at 12,000 x g for 20 min, and the supernatant was used as the
crude enzyme preparation.

(ii) Heat treatment: The cell extracts of Thermoanaerobacter and C.
thermosulfurogenes were heated for 15 min at 85°C and 80°C, respectively, and

cooled to 4°C. The soluble fractions were recovered after centrifugation at 12,000

71
x g for 20 min.

(iii) Ammonium sulfate fractionation: Solid ammonium sulfate was added to
the heat-treated extracts to give 65% saturation and the precipitate was removed by
centrifugation. More ammonium sulfate was added to the supernatant to give 85%
saturation. This precipitate was collected, and dissolved in 50 mM MOPS buffer
(pH 7.0) containing 5 mM MgSO, and 0.5 mM C00,, and dialyzed overnight against
the same buffer.

(iv) Column chromatography: The above enzyme preparations were loaded onto
DEAE-Sepharose CL-6B columns (4.0 cm x 32 cm), previously equilibrated with 50
mM MOPS buffer (pH 7.0) containing 5 mM MgSO4 and 0.5 mM C00,. The
columns were washed with same buffer and then eluted with NaCl salt gradient (0.0
to 0.5 M) in the same buffer. The activity peak fractions were pooled and
concentrated by ultrafiltration (YM 30 membrane, Amicon Co.). The enzyme
solutions were divided into three portions and each portion was applied onto a
Superose-12HR (Pharmacia, Inc.) gel filtration column, and were eluted with the same

buffer with a Pharmacia FPLC System.

Electrophoresis and molecular weight determination
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was
performed as described by Laemmli (22). Native-PAGE was performed without
SDS, and Tris-HCl buffer (pH 8.6) was used during polyacrylamide gel preparation.
Protein bands were visualized by Coomassie Brilliant Blue G-250 staining.

Molecular weights of purified glucose isomerases were determined by gel

72

filtration on a Superose-IZHR column in a FPLC system, and blue dextran (Mr
2,000,000); apoferritin (443,000); alcohol dehydrogenase (150,000); and bovine serum
albumin (66,000) were used as molecular weight standards. The subunit molecular
weights were estimated by SDS-PAGE with low range protein molecular weight
standards (Bio Rad Laboratories) : phosphorylase (Mr 97,400); bovine serum albumin
(66,200); ovalbumin (42,700); carbonic anhydrase (31,000); and soybean trypsin
inhibitor (21,500).

Two different pH ranges of Servalyt-Precotes isoelecuic focusing gels (Serva
Co.; pH 3-10 and pH 3-6) were used for isoelecuic pH determination. An LKB
ultrapore isoelectric focusing apparatus was used and the gels were stained with Serva

Blue W.

Amino acid compositions and sequence determination

Samples were prepared by washing the purified glucose isomerase preparations
with double distilled water five times with a Cenuicon-30 (Amicon Co.) filtration
device to remove metal salts in the enzyme solution. Amino acid composition
analysis was performed with Water’s Pico-Tag amino acid analyzer, and the N-
terminal amino acid sequences were identified with a Beckman Model 890M phase
sequencer in the Macromolecular Structure Facility, Department of Biochemistry,

Michigan State University.

Metal ion effects on enzyme activity and stability

To prepare a metal ion-free enzyme solution, glucose isomerase purified from C.

73
thermosulfurogenes was treated with 5 mM EDTA at 60°C for 1 hr, and then washed

with double distilled water five times by an Amicon Centricon-30 filtration device.
The effect of metal ions on enzyme thermal stability was determined by measuring
residual glucose isomerase activity under optimum assay conditions after a 15 min
preincubation at 80°C in the presence of various metal ions (1 mM). The effect of
metal ions on enzyme activity was performed at 55°C where the enzyme was Stable

without metal cofactors during the reaction period.

Protein crystallization

The hanging drop vapor diffusion method was used to crystallize the purified
glucose isomerase from C. thermosulfurogenes in Linbro tissue culture multi-well
plates (Flow Laboratories, Inc., Virginia). The protein solution used to prepare
crystals contained glucose isomerase (15 mg/ml) in 50 mM MOPS buffer (pH 7.0)
with 10 mM MgSO, and 1 mM C00,. The reservoir solution (1 ml) in each well
contained a series of different PEG 4000 concentrations (5%-30%) in 0.1M citrate
buffer (pH 5.5) containing 0.2 M LiZSO,. Equal volume (6 pl) of protein and
reservoir solutions were mixed to make a droplet on a siliconized glass microscope '
clover slip and then the cover slips were sealed onto the corresponding well with high
vacuum grease. Duplicated samples of the plates were incubated at 4°C and 20°C

until crystals appeared.

RESULTS

Comparison of glucose isomerase activity in thermoanaerobes

The general biochemical properties of glucose isomerase in crude cell extracts
of three different therrnoanaerobes were compared (T able 1). Glucose isomerase
activity from C. thermohydrosulfuricum differed from the other species by displaying
a slightly higher thermostability and apparent temperature optimum and a basic (8.5)
pH optimum for enzyme activity. All three glucose isomerases required metal ions
(CoH, Mg“) for their activity and stability. Glucose isomerase was purified from C.
thermosulfurogenes and Thermoanaerobacter because of their similar neutral pH

optima for activity and thermal properties.

Purification and molecular properties

The protocols used for glucose isomerase purification from C. thermosulfurogenes
and Thermoanaerobacter strain B6A achieved a 11- to 10-fold purification and 19-
13% yield from cell extracts (see Table 2). An ammonium sulfate fractionation was
used to purify the enzyme from Thermoanaerobacter because it removed a certain
protein that was not well separated by the following steps. In the final purification
step, only a single major protein peak was detected by Superose-12 gel filtration with
fast performance liquid chromatography. The purified enzymes were considered to
be homogeneous by the detection of single bands on SDS-PAGE, native PAGE, and
Serva Precoat isoelecuic focusing gels (Figure l).

The molecular weight of purified glucose isomerase was determined by

74

75

Table 1. Biochemical properties of glucose isomerase activity in crude cell extracts of

 

 

thermoanaerobes.

Property C. thermosulfurogenes Thermoanaerobacter C. thermohydrosulfuricum
strain 43 strain B6A strain 39B

Optimum pH 7.0-7.5 7.0-7.5 8.5

Optimum temperature 75-80°C 80°C 80-85°C

Temperature stability‘ 80°C 85°C 90°C

 

'Indicates the highest temperature where more then 90% of original aetivity was retained after

a 1 h incubation.

76

Table 2. Summary of glucose isomerase purification steps

A. T hermoanaerobacter

 

 

 

 

 

Total Total Specific
Step Protein Activity Activity Purification

(mg) (unit) (unit/mg)
Cell Free Extract 2820 1410 0.5 1.0
Heat Treatment
(85°C, 15 nrin) 725 1310 1.8 3.6
Ammonium Sulfate
Fractionation 254 585 2.3 4.6
DEAE-Sepharose
Anion Exchange 94 385 4.1 8.2
Superose-12
Gel Filtration 38 181 4.8 9.6
B. C. thennosulfurogenes

Total Total Specific

Step Protein Activity Activity Purification

(mg) (unit) (unit/mg)
Cell Free Extract 2640 820 0.3 1.0
Heat Treatment
(80°C, 10 min) 920 690 0.8 2.7
DEAE-Sepharose
Anion Exchange 118 342 2.9 9.7
Superose-12
Gel Filtration 46 156 3.4 11.3

 

77

Figure 1. Electrophoretic analysis of glucose isomerases purified from
Thermoanaerobacter (A) and C. thermosulfurogenes (B).

pl: isoelectric focusing gel electrophoresis with Servalyt-Precotes gel (pH range;
3-10), Native: native-PAGE with Tris-HCl buffer (pH 8.6), SDS: SDS-PAGE with a
12% gel, pH represents pH separations of protein standards, Mr represents MW

separation of protein standards.

78

wow

3. V—t‘

1v. «at I.

3. _.m.. In.

3. met all...

1v. 00..-
1x50: 11.!

:2 m <

e232 3
r.‘ low.
.. E a... l 3
. . E I3.
.44 ind

m.-. . .<- . m < :a

79
Superose—lZ gel filtration with FPLC. Glucose isomerase from C. thermosulfurogenes
and Thermoanaerobacter strain B6A displayed an identical molecular weight of
200,000. SDS-PAGE analysis displayed a single band for both enzymes with a
molecular weight of 50,000 indicating that both enzymes were composed of homo-
tetrameric subunits. Isoelectric points of C. thermosulfurogenes and
Thermoanaerobacter glucose isomerase were pH 4.9 and 4.8, respectively.

The amino acid composition of the purified glucose isomerases are compared
in Table 3. These enzymes contained similar amino acid molar ratios except for
glutamine and glutamate. Alanine was the most abundant residue and histidine, the
proposed catalytic residue (15) was present in both enzyme molecules. The total
amount of hydrophobic residues in the enzyme molecule was slightly higher in
Thermoanaerobacter than in C. thermosulfurogenes.

Comparison of N-terminal sequences of the first seven amino acid residues in
the two thermostable glucose isomerases with those of the thermal labile enzymes
from Escherichia coli (27) and Bacillus subtilis (37) indicated that the N-terminal
sequences of the two thermophile glucose isomerases were identical but they displayed

lower homology with mesophile enzymes (Figure 2).

Physicochemical properties

The purified glucose isomerases diSplayed very similar apparent temperature and
pH optima for enzyme activity at 80°C and pH 7.0-7.5 (Figure 3,4). Both glucose
isomerases were stable in the broad pH range of 5.5 to 12.0, and were readily

denatured at pH values lower than 5 (Figure 4).

80
Table 3. Amino acid compositions of thermostable glucose isomerase from C.

thermosulfurogenes and Thermoanaerobacter.

 

Molar Ratio (%)

 

 

Amino Acid C. thermosulfurogenes Thermoanaerobacter
Asx 9.5 7.9
Glx 7.3 3.6
Ser 4.0 3.6
Gly 9.1 8.5
His 2.0 2.3
Arg 5.9 6.0
Thr 4.4 4.9
Ala 13.9 15.4
Pro 5.3 4.7
Tyr 4.3 4.9
Val 4.3 3.7
Met 1.7 2.7
111 4.3 3.9
Leu 7.1 7.9
Phe 8.8 11.3
Lys 7.6 8.1
Trp n.d. n.d.
Cys n.d. n.d.
Total 100 100
Hydrophilic 50.3 45.5
Hydrophobic 49.7 54.5

 

Asx: Aspartic acid and asparagine; Glx: Glutamic acid and glutamine; n.d.: Not determined.

81

Figure 2. Comparison of N-terminal amino acid sequence of purified glucose
isomerases.

Abbreviations are: 4B, C. thermosulfurogenes; B6A, Thermoanaerobacter; B.C.,
E. coli; B.S., Bacillus subtilis. Experimentally determined sequence of 4B and B6A

enzymes were aligned with published data for E. coli (25) and B. subtilis (35).

4B

B6A

E.C.

B.S.

Met

Met

Met

Met

Asn

Asn

Gln

Ala

82

Lys
Lys
Ala

Gln

Tyr
Tyr
Tyr

Ser

Phe

Phe

Phe

His

Gln
Gln
Asp

Ser

Asn

Asn

Gln

Ser

83

Figure 3. Effect of temperature on glucose isomerase activity.
100% activity corresponds to 7.1 unit/mg, 8.8 unit/mg for C. thermosulfllrogenes

(H) and Thermoanaerobacter (H) glucose isomerases. respectively.

Relative Activity (%)

IOO

50

84

 

 

 

 

30 40 so so 70 so 90

Temperature (C°)

IOO

 

85

Figure 4. Effect of pH on thermostable glucose isomerase activity (A) and
stability (B).

The symbols represent: H , C. thermosulfurogenes; H ,
T hermoanaerobacter. Enzyme activities were assayed in either glycylglycine buffer
at pH 3.0-4.5 and 8.5-13.0, sodium acetate buffer at pH 4.5-6.5 or MOPS buffer at
pH 6.5-8.5. Enzyme stabilities were detected by assaying the residual activities
after preincubation of the enzyme in these buffer solutions with 5 mM MgSO4 and
0.5 mM CoCl2 at 60°C for 1 h. 100% activity corresponds to 4.5 unit/mg and 5.3

unit/mg, for C. thermosulfurogenes and Thermoanaerobacter, respectively.

86

A . Activity

 

 

l
9 IO ll

1
8

 

 

@L 3284 3:23“.

pH

8. Stability

 

 

 

 

3 3354 .8235.

 

87

The effect of temperature on the stability of C. thermosulfurogenes glucose
isomerase is shown in Figure 5. EDTA treated enzyme was stable for 1 hr at 60°C
but it was readily denatured at 70°C. In the presence of metal ions (5 mM MgSO,
and 0.5 mM CoClz), the enzyme was stable for 1 hr at 85°C. The half-life of the
glucose isomerase in 50 mM phosphate buffer (pH 7.0) containing 5 mM MgSO4 and
0.5 mM CoCl2 was 198 hr at 60°C and 42 hr at 70°C (data not shown). The effect
of various metals on thermal stability of EDTA treated glucose isomerase from
C. thermosulfurogenes is shown in Table 4. The enzyme required Co++ and/or Mg++
for high thermal stability, and Mn” was less effective to protect the enzyme from
heat denaturation at 80°C. Other metal ions examined in this study did not enhance
thermostability of the enzyme, and Cu“ and Zn”+ showed an inhibitory effect.

Crystals of glucose isomerase from C. thermosulfurogenes usually appeared on
the next day in a hanging drop in the well with the reservoir solution containing
16%-18% PEG 4000 at room temperature. Figure 6 shows crystals of glucose
isomerase which exhibited a long and flat shape with a size of 0.7 x 0.1 mm in
length and width. The preliminary results of X-ray diffraction analysis confirmed
that the crystal is a protein crystal and is large enough for detailed analysis for future

studies on the three dimensional structure of the enzyme.

Catalytic properties
The kinetic features of the purified glucose isomerases on three different
substrates are compared in Table 5. K.,, and Vm, values were obtained from

Lineweaver—Burk plots of specific activities at various substrate concentrations.

88

Figure 5. Thermostability of EDTA-treated glucose isomerase of C.
thermosulfurogenes in the presence (0—0) or absence (H) of metal ions.

Residual activities were assayed after preincubation of the DEAE-treated enzyme
in 50 mM MOPS buffer (pH 7.0) with or without 5 mM MgSO‘ and 0.5 mM CoCl2
at indicated temperatures for various time periods. 100% activity corresponds to 4.9

unit/mg of the pure enzyme.

89

cm

Om

0v

2:: we:

Om

ON

0_

 

 

—

coon

ooOm

comm
comm

ooow

_

_

q

_

_

 

tom

59

 

(%) anav nonmsaa

Table 4. Comparison of thermostable glucose isomerase kinetic properties from

 

 

thermoanaerobes.

K... (mM) V... (unit/mg) k... (llmin)
Substrate 4B B6A 4B B6A 4B B6A
Glucose 140 120 5.2 6.3 1040 1260
Fructose 60 50 2.5 2.8 500 560
Xylose 20 16 15.7 17.6 3140 3520

 

43: C. thermosulfurogenes; B6A: Thermoanaerobacter.

91

Figure 6. Crystals of glucose isomerase from C. thermosulfurogenes.

92

 

93

Table 5. Effect of metals on activity and themal stability of EDTA-treated glucose

isomerase from C. thermosulfurogenes.

 

Isomerase Activity'

Thermal Stability”

 

Metal Glucose Xylose (% Residual)
(% maximum)
None 2 13 14
MgSO4 (1 mM) 21 89 94
MgSO. (5 mM) 97 90 94
C002 (1 mM) 100 98 100
MnSO. (1 mM) 6 100 61
FeSO. (1 mM) 4 15 15
NiSO. (1 mM) 2 8 20
BaClz (1 mM) 2 14 16
CaCl2 (1 mM) 2 16 22
ZnSO. (1 mM) 0 0 0
CuSO4 (1 mM) 0 0 0
Mgso, + CoCl, (1 mM) 98 117 "111
MgSO4 + MnSO, (1 mM) 16 113 86
MnSO4 + CoCl2 (1 mM) 35 120 78

 

'Enzyme activity was assayed at 55°C where the enzyme was stable during the reaction

period without metal cofactors.

I’Residual activity was measured at 65°C after incubation of these enzymes in the solution

containing metal cofactor at 80°C for 15 min.

94

Glucose isomerases from two different thermoanaerobic bacteria had very similar
apparent K.,, V.,“, and K, for glucose, fructose, and xylose. Both enzymes
displayed lower K., values for xylose than for glucose or fructose, and the Kll values
for fructose were about 2-fold lower than those for glucose. The apparent V.,, and
k“, values of these enzymes with xylose as substrate were approximately 3-fold higher
than those with glucose, and were approximately 6-fold higher than those with
fructose.

The effect of various metal ions on the activity of EDTA treated enzymes from
C. thermosulfurogenes was investigated (see Table 4). Co“ or Mg“ was required
for glucose isomerase activity, whereas Mn“ did not enhance enzyme activity. The
addition of Co’+ and Mg“ did not show a synergistic effect on glucose isomerase
activity and, Mn“ reduced the activation effect of Co” or Mg“. A minimum
concentration of 5 mM MgSO. or 0.5 mM CoCl2 was required to achieve maximum
glucose isomerase activity (data not shown). For xylose isomerase activity, Mn++ (1
mM), Co” (1 mM), or Mg“ (1 mM) were required and any combination of these
metal ions showed a synergistic activation effect. Other metal ions had very little
effect on both glucose isomerase and xylose isomerase activities, and Zn“ (1 mM)

and Cu++ (1 mM) totally inhibited enzyme activity on glucose and xylose.

DISCUSSION

To our best knowledge, this study represents the first detailed characterization
of glucose isomerases purified from thermoanaerobic bacteria. As expected, the
glucose isomerases in these microbes were very thermostable and two different types
of activity were identified based on their pH optima. The glucose isomerases with
a neutral pH optimum were purified from C. thermosulfurogenes and
Thermoanaerobacter because of enzyme requirements in industry (2,7) and the desire
to compare similar enzyme activities from different species.

The enzymes from both microorganisms were very stable under the conditions
used during purification. Glucose isomerase was one of the major proteins
(approximately 10% of total protein) in the cell extract, and purification of the
enzymes was relatively simple. The abundance of this enzyme in the cell extracts
may be due to the fact that the enzyme had a relatively low It, and high K.,, which
is typical of xylose isomerase (2,31), and perhaps the organism needed to overproduce
this enzyme which may be rate limiting for growth on xylose. The growth rate of
these bacteria is faster on glucose than xylose (28,36).

The overall biochemical and physicochemical properties of the glucose
isomerases purified from these two thermoanaerobic bacteria were similar yet these
species differ in sporulation and in saccharidase activities (unpublished results).
Different microbial glucose/xylose isomerases characterized previously vary in
molecular weight from 80-195 KDa and are composed of two or four identical

subunits. Molecular weights (200 KDa) and tetrameric subunit composition of both

95

96

thermoanaerobic glucose isomerases are similar to the 195 KDa enzyme present in
Lactobacillus brevis (38). Glucose isomerases characterized from Streptomyces
species (32,20), Arthrobacter (l7), Bacillus coagulans (l2), and Flavobacterium
arborescens (5) display smaller molecular weights (157-183 KDa) with tetrameric
subunit compositions. The size and shape of C. thermosulfurogenes glucose
isomerase crystals was very different from those reported previously from orher
enzyme sources (12,31,32,38). Glucose isomerases from E. coli (34), alkalophilic
Bacillus (21), Actinoplanes missouriensis (16), and Streptomyces olivochromogenes
(31) are dimers with molecular weights of 80-120 KDa.

The optimum pH for the glucose isomerizing activity of C. thermosulfurogenes
and Thermoanaerobacter was 7.0-7.5, which was similar to those of the enzymes
from B. coagulans (13) and A. missouriensis (26), and was lower than those (pH 8.0-
10.0) of enzymes from S. phaechromogenes (33), S. griseofuscus (l9), and S.
olivochromogens (31). The enzymes from both thennoanaerobes were stable within
the pH range of 5.5-11.0 and displayed isoelecuic points of pH 4.9 and 4.8.

Glucose isomerases are generally classified into two types of enzymes
according to their thermal stability and temperature optima for enzyme activity (8,9).
Lactobacillus brevis glucose isomerase (38), B. subtilis (C. Lee, et al. manuscript
submitted to Appl. Environ. Microbio.), and E. coli xylose isomerase (4) are
thermolabile and active at 37°~50°C. Glucose isomerases produced from C.
thermosulfurogenes and Thermoanaerobacter were highly thermostable, and displayed
one of the highest optimum temperature for activity among reported glucose

isomerases (8,9). Glucose isomerases from certain mesophilic microbial sources (e.g.,

97
S. phaeochromogenes (32) and A. missouriensis (25)], however, can display quite high
temperature optima and stability (9).

The requirement of Co“ or Mg” for glucose isomerizing activity and the
requirement of Mn“ for xylose isomerizing activity of C. thermosulfurogenes enzyme
was similar to that reported for B. coagulans enzyme (11). Although C.
thermosulfurogenes glucose isomerase required Co“ or Mg‘+ for its optimal
thermostability, EDTA treated enzyme was more thermostable than the enzymes from
E. coli or B. subtilis in the presence of these metals (4, unpublished result).

These findings demonstrate that the two distinct thermoanaerobic bacteria
produce highly thermostable glucose isomerases with close similarity in
physicochemical and catalytic properties. Because these bacteria have evolved in
thermal hot spring ecosystems (28,36) and are thought to have a common phylogenic
origin (3), one might expect that their glucose isomerases have similar properties yet
represent distinct species. At present, the molecular mechanism of thermostability
in these enzymes is not clear and remains to be solved. The gene coding for
thermostable glucose isomerase of C. thermosulfurogenes has been cloned and its
amino acid sequence displays higher homology to the thermolabile enzyme of B.
subtilis than the thermostable enzyme of Streptomyces (C. Lee et al., manuscript
submitted to J. Biol. Chem). Further studies on the three dimensional structure of
the enzyme and protein modification via site directed mutagenesis are required to

explain the molecular mechanism of high thermostability in this glucose isomerase.

10.

11.

12.

LITERATURE CITED

Amelunxen, R. E., A. L. Murdock. 1978. Microbial life at high temperatures:
Mechanisms and molecular aspects, p. 217-277. In DJ. Kushner (ed.) Microbial
life in extreme environments, Academic Press, San Francisco.

Anuim, R. L., W. Colliala, and B. Schnyder. 1979. Glucose isomerase
production of high fructose syrups, P. 97-155. In L.B. Wingard (ed.) Applied
Biochemistry and Bioengineering, Academic Press, New York.

Bateson, M. M., Wiegel, J., and D. M. Ward. (1989) Comparative analysis
of 16S ribosomal RNA sequences of thermophilic fermentative bacteria isolated
from hot spring cyanobacterial mats. Syst. Appl. Microbiol. 12:1-7.

Batt, C.A., e. O’Neil, S.R. Novak, J. Ko, and A. Sinskey. 1986.
Hyperexpression of Escherichia coli xylose isomerase. Biotechnol. Prog.
2:140-144.

Boguslawski, G. 1983. Purification and characterization of glucose isomerase
from Flavobacterium arborescens. J. Appl. Biochem. 5:186-196.

Bucke, C. 1980. Enzymes in fructose manufacture, p. 51-72. In G.G. Birch,
N. Blackebrough, and J.K. Parker (ed.), Enzymes and Food Processing.
Applied Science Publishers, London.

Bucke, C. 1977. Industrial glucose isomerase, p. 147-171. In A. Wiseman (ed.)
Topics in Enzymes and Fermentation Biotechnology, part b, Halstead Press,
Chichester.

Chen, W. 1980a. Glucose isomerase (a review). Process Biochem. 15:30-35.
Chen, W. 1980b. Glucose isomerase (a review). Process Biochem. 15:36-41.

Chen, W. P., A. W. Anderson, and Y. W. Han. 1979. Extraction of
glucose isomerase from Streptomyces flavogrisius. Appl. Environ.
Microbiol. 37:785-787.

Danno, G., S. Yoshimura, and M. Natake. 1967. Studies on D-glucose
isomerizing activity of D-xylose-grown cells from Bacillus coagulans, strain
HN-68. Agric. Biol. Chem. 31:284-292.

Danno, G. 1970. Studies on D-glucose-isomerizing enzyme from Bacillus

coagulans, strain HN-68 Part IV. Purification, crystallization and some
physicochemical properties. Agr. Biol. Chem. 34:1795-1804.

98

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

99

Danno, G. 1970. Studies on D-glucose-isomerizing enzyme from Bacillus
coagulans, strain HN-68 Part V. Comparative study on the three activities of
D-glucose, D-xylose, and D-ribose isomerization of the crystalline enzyme.
Agric. Biol. Chem. 34:1805-1814.

Dishe, Z., and E. Borenfreund. 1951. A new spectrophotometric methods for
the detection and determination of keto sugars and trioses. J. Biol. Chem.
192:583-587.

Gaikwad, S. M., M. W. More, H. G. Vartak, and V. V. Deshpande. 1988.
Evidence for the essential histidine residue at the active site of glucose/xylose
isomerase from Streptomyces. Biochem. Biophys. Res. Commun. 155:270-277.

Gong, C.-S., L. F. Chen, and G. T. Tsao. 1980. Purification and properties of
glucose isomerase of Actinoplanes missouriensis. Biotechnol. Bioeng. 12:833-
845.

Henrik, K., C. A. Collyer, and D. M. Blow. 1989. Structures of D-xylose
isomerase from Arthrobacter strain B3728 containing the inhibitors xylitol and
D-sorbitol at 2.5 A and 2.3A resolution, respectively. J. Mol. Biol. 208:129-
157.

Jorgensen, O. B., L. G. Karlsen, N. B. Nielsen, S. Pedersen, and S. Rugh,
Bagsvaerd (Denmark). 1988. A new immobilized glucose isomerase with high
productivity produced by a strain of Streptomyces murinus. Starch/Starke.
40:307-313.

Kasumi, T., K. Hayashi, and N. Tsumura. 1981. Purification and enzymatic
properties of glucose isomerase from Streptomyces griseofuscus, S-4l. Agric.
Biol. Chem. 45:619-627.

Kasumi, T., K. Hayashi, N. Tsumura, and T. Takagi. 1981. Physicochemical
characterization of glucose isomerase from Streptomyces griseofuscus, S-41.
Agric. Biol. Chem. 45:1087-1095.

Kwon, H., M. Kitada, and K. Horikoshi. 1987. Purification and properties of
D-xylose isomerase from alkalophilic Bacillus No. KX-6. Agric. Biol. Chem.
51:1983-1989.

Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of
the head of bacteriophage T4. Nature (London). 227:680-685.

Lowry, O. H., J. Rosebrough, A. I. Farr, and R. J. Randall. 1951. Protein
measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

100

Ng, T. K., W. R. Kenealy. 1987. Industrial applications of thermostable
enzymes, p. 197-215. In T.D. Brock (ed.) Thermophiles; General, molecular
and applied microbiology. John Wiley and Sons, New York.

Saha, B. C., S. Mathupala, and J. G. Zeikus. 1988. Purification and
characterization of a highly thermostable pullulanase from Clostridium
thermohydrosulfuricum. Biochem. J. 252:343-348.

Scallet, B. L., K. Shieh, I. Ehrenthal, and L. Slapshak. 1974. Studies in the
isomerization of D-glucose. Die Starke. 26:405-408.

Schellenberg, G. D., A. Sarthy, A. E. Larson, M. P. Backer, J. W. Crabb, M.
Lindstrom, B. D. Hall, and C. E. Furlong. 1984. Xylose isomerase from
Escherichia coli: characterization of the protein and the structural gene. J. Biol.
Chem. 259:6826-6832.

Schink, B., and J. G. Zeikus. 1983. Clostridium thernwsulfurogenes sp. nov.,
a new thermophile that produces elemental sulphur from thiosulfate. J. Gen.
Microbiol. 129: 1 149-1 158.

Schnyder, B. J., 1974. Continuous isomerization of glucose to fructose on a
commercial basis. Die Starke. 26:409-412.

Shen, G., B. C. Saha, Y. E. Lee, L. Bhatnagar, and J. G. Zeikus. 1988.
Purification and characterization of a novel thermostable B-amylase from
Clostridium thermosulfurogenes. Biochem. J. 254:835-840.

Suekane, M., M. Tamura, and C. Tomimura. 1978. Physico-chemical and
enzymatic properties of purified glucose isomerases from Streptomyces
olivochromogenes and Bacillus stearothermophilus. Agric. Biol. Chem. 42:909-
917.

Takasaki, Y., Y. Kosugi, and A. Kanbayashi. 1969. Studies on sugar-
isomerizing enzyme purification, crystallization and some properties of glucose
isomerase from Streptomyces sp. Agric. Biol. Chem. 33:1527-1534.

Tsumura, N. and T. Sato. 1965. Enzymatic conversion of D-glucose to D-
fructose Part VI. Properties of the enzyme from Streptomyces
phaeochromogenus. Agric. Biol. Chem. 29:1129-1134.

Tucker, M. Y., M. P. Tucker, M. E. Himmel, K. Grohmann, and S. M.
Lastick. 1988. Properties of genetically overproduced E. coli xylose isomerase.
Biotechnol. Lett. 10:79-84.

35.

36.

37.

38.

39.

40.

101

Verhoff, F. H., G. Boguslawski, O. J. Lantero, S. T. Schlager, and Y. C. Jao.
1985. Glucose isomerase, p. 837-859. In H.W. Balanch, S. Drew, and D.I.C.
Wang (eds.), Comprehensive Biotechnology , vol. 3, Pergamon Press, Oxford.

Weimer, P. J., L. W. Wagner, S. Knowlton, and T. K. N g. 1984. Thermophilic
anaerobic bacteria which ferment hemicellulose: Characterization of organisms
and identification of plasmids. Arch. Microbiol. 138:31-36.

Wilhelm, M., and C. P. Hollenberg. 1985. Nucleotide sequence of the
Bacillus subtilis xylose isomerase gene: extensive homology between the
Bacillus and Escherichia coli enzyme. Nucleic Acids Res. 13:5717-5723.

Yamanaka, K. 1968. Purification, crystallization and properties of the D-xylose
isomerase from Lactobacillus brevis. Biochim. Biophys. Acta. 151:670-680.

Zeikus, J. G., A. Ben-Bassat, and P. W. Hegge. 1980. Microbiology of
methanogenesis in thermal, volcanic environments. J. Bacteriol. 143:432-440.

Zeikus, J. G. 1979. Thermophilic bacteria: ecology, physiology, and
technology. Enzyme Microb. Technol. 1:243-252.

Chapter IV

Cloning and Expression of the Clostridium thermosulfurogenes

Glucose Isomerase Gene in Escherichia coli and Bacillus subtilis

102

ABSTRACT

The gene encoding thermostable glucose isomerase in Clostridium
thernwsulfilrogenes was cloned by complementation of glucose isomerase activity in
a xylA mutant of Escherichia coli. A new assay method for thermostable glucose
isomerase activity on agar plates, using a top agar mixture containing fructose, glucose
oxidase, peroxidase and benzidine, was developed. One positive clone, carrying
plasmid pCGI38 was isolated this from a cosmid library of C. thermsulfurogenes
DNA. The plasmid was further subcloned into a Bacillus cloning vector, pTB522,
to generate a shuttle plasmid pMLGl, able to replicate in both E. coli and Bacillus
subtilis. Expression of thermostable glucose isomerase gene in both species was
constitutive whereas synthesis of the enzyme in C. thermosulfurogenes was inducible
by D-xylose. B. subtilis and E. coli produced higher levels of thermostable glucose
isomerase (1.54 and 0.46 units/mg protein, respectively) than C. thermosulfurogenes
(0.29 units/mg protein). The glucose isomerase synthesized in E. coli and B. subtilis
was purified to homogeneity and displayed identical properties (subunit M,=50 kDa,
tetrameric molecular structure, thermostability, metal ion requirement, apparent
temperature and pH optima) to those of the native enzyme purified from C.
thermosulfurogenes. Simple heat treaunent of crude extracts from E. coli and B.
subtilis cells, carrying the recombinant plasmids, at 85°C for 15 min generated 80%
pure glucose isomerase free of protease activity. The maximum conversion yield of
glucose (35%, wt/wt) to fructose with the thermostable glucose isomerase (10.8 units/g

dry substrate) was 52% at pH 7.0 and 70°C.

103

INTRODUCTION

Xylose isomerase (D-xylose ketol isomerase, EC 5.3.1.5) which reversibly
catalyzes the isomerization reaction between D-xylose and D-xylulose can also convert
D-glucose into D-fructose, hence the enzyme is often referred to as glucose isomerase
(27). Glucose isomerase is one of the largest volume commercial enzymes used
today, and it is employed for the industrial production of high fructose corn sweetener
(9). Improvement in the enzyme process used by the sweetener industry requires
higher thermal and acid stability of glucose isomerase, higher specific activity and
over-production of the enzyme, and a lower cost enzyme recovery process from the
host cell after fermentation (9,19).

In spite of the industrial importance of microbial glucose isomerase, few reports
exist on molecular comparisons of the enzyme in different species (especially in
therrnophiles) or on molecular understanding of enzyme thermal and acid stability.
The D-xylose isomerase genes from both E. coli and B. subtilis have been cloned and
sequenced (25,28), and over-expression of E. coli xylose isomerase by using tac or
lac promoters in E. coli strains has been reported (2,24,29). Cloning of the glucose
isomerase gene from Streptomyces violaceoniger in a Streptomyces host has been
reported (22).

Recently, we have purified and characterized thermostable glucose isomerases
from thermoanaerobic-bacteria (C. Lee and J.G. Zeikus, manuscript submitted to J.
Bacteriol.) At present, nothing is known about the molecular structure or

organization of glucose isomerase genes from thermophilic microorganisms.

104

105

Therefore, we undertook Studies aimed at cloning of a gene coding for thermostable
glucose isomerase from an anaerobic thermophile and its over-expression in a food-
safe B. subtilis host. In this paper we report on a new plate assay method for
detection of thermostable glucose isomerase, on cloning and over-expression of the
thermophile gene and enzyme, and on biochemical and biotechnological features of

the recombinant enzyme.

MATERIALS AND METHODS

Bacterial strains and plasmids

C. thermosulfurogenes strain 4B was used as the source of the glucose isomerase
gene (26). E. coli strain W595 [F malAl xyl—7 ara-13 mtl-2 tonA2 galTl] (1),
obtained from Dr. Bachmann, E. coli Genetic Stock Center, Yale University, and
HBlOl [F hsd820 ara-14 recAl3 proA12 lach galK2 vpoL20 mtl-l xyl-S] (21)
were used for DNA manipulation and complementation test. B. subtilis strain NAl
[org-15 hst hde amy‘ npr‘] (18) which does not exhibit any glucose isomerase
activity at 65°C was used as a host strain for subcloning of the glucose isomerase
gene. Cosrnid vector pHSG262(4) was used for the construction of a genomic
library of C. thermosulfurogenes in E. coli HBlOl. Bacillus cloning vector pTB522

(16) was used during the subcloning procedure.

Chemicals, media, and growth conditions

All chemicals were of reagent grade. C. thermosulfurogenes strain 43 was
grown under anaerobic conditions at 60°C in TYE medium containing 1% xylose or
maltose as a carbon source (31). For growth of E. coli and B. subtilis strains, Luria
broth (10 g tryptone, 5 g yeast extract, and 5 g NaCl per liter at pH 7.5) was used
for liquid culture and solid media contained 1.5% agar (Difco Co., Detroit, MI).
Kanamycin (50 ttg/ml) or tetracycline (25 ug/ml) were used for selection of pHSG262

or pTB522 carrying strains, respectively.

106

107

Plate assays for detection of thermostable glucose isomerase

Fructose (2%), MgSO, (5 mM), CoCl2 (0.5 mM), glucose oxidase (20 unit/ml),
peroxidase (4 unit/ml), and benzidine (0.4 mg/ml) in 100 mM MOPS buffer (pH 7.0)
were mixed with top agar (0.7%) at 50°C, and poured on the colonies grown
overnight on LB agar plates with 1% xylose. After soft agar had solidified the
plates were incubated overnight at 50°C. Positive clones showed a dark brown halo
around the colonies (see Figure 1). One positive colony was found after screening
1000 transformants by this method. MacConkey agar plates containing 1% xylose
were also used to detect complementation of xylose utilization in E. coli xyl-S mutants

at 37°C.

Extraction and manipulation of DNA

Chromosomal DNA was isolated from C. thermosulfurogenes cells harvested at
mid-logarithmic growth phase by a modification of the Marmur procedure (23).
Cells suspended in 0.25 M sucrose were treated with lysozyme (0.5 mg/ml) and
EDTA (50 mM, pH 8.0) for 5 min followed by SDS (0.2%) at 60°C. The lysate
was treated with proteinase K (0.1 mg/ml) and further deproteinized with repeated
phenol/chloroform extraction. DNA was precipitated with 0.6 volumes of
isopropanol, collected by "spooling" on a glass rod, and dissolved in 20 mM Tris, 1
mM EDTA. Plasmid DNA was purified by the procedure described by Clewell (6)
followed by centrifugation in a cesium chloride-ethidium bromide density gradient.
The alkaline denaturation method was used for rapid extraction of plasmid DNA (3).

Digestion of DNA with restriction endonucleases, separation of fragments by

108

Figure l. Indicator plate for thermostable glucose isomerase activity.
1: E. coli HBlOl/pCGI38, 2: E. coli HBlOl/pHSG262, 3: E. coli JM107,

4: B. subtilis NAl.

1 ()9

 

110
electrophoresis, recovery of DNA from agarose gels, dephosphorylation with calf
intestine phosphatase, ligation of DNA fragments, and transformation of E. coli with
plasmid DNA were performed as described by Maniatis et a1. (21) or as recommended

by the manufacturers of enzymes.

Construction of genomic library and cloning procedure in E. coli

Left and right arms of the cosmid vector, pHSG262, dephosphorylated at HincII
and EcoRI sites were prepared as described by Brady et al. (4). HincII-BamI-II and
EcoRI-BamHI fragments were then ligated with DNA fragments (30-40 kb in length)
prepared by partial digestion of chromosomal DNA of C. thermosulfurogenes with
Sau3AI. The ligated mixture was packaged in vitro into bacteriophage lambda heads
(10), and the recombinant cosmids were introduced into cells of 118101 strain by
transfection (12). Approximately 1,000 colonies containing recombinant plasmids
were replicated onto LB agar plates with kanamycin. After growth at 37°C
overnight, colonies were screened by the direct plate assay for thermostable glucose
isomerase activity. The recombinant plasmid pCGI38 was isolated from one positive
clone and introduced by transformation into different xyl mutants of E. coli.
Transformants were selected on MacConkey agar plates supplemented with 1% D-

xylose and kanamycin.

Subcloning of a DNA fragment containing glucose isomerase gene into B. subtilis
EcoRI digested pCGI38 was ligated with pTBSZZ that had been cut with EcoRI

and dephosphorylated. The ligation mixture was used to transform competent cells

111

of B. subtilis as described earlier (15). Transformants were selected on LB agar
plates with tetracycline and recombinant clones were identified by the restriction

pattern of the plasmid DNA isolated by rapid alkaline extraction.

Enzyme assays and protein determination

Washed whole cells, cell-free extracts or purified preparations were used as
enzyme sources. Cells grown to late-logarithmic phase in LB media with or without
appropriate antibiotics and 1% xylose were harvested by centrifugation at 5,000 x g
for 10 min, and washed with 50 mM MOPS buffer (pH 7.0) once and resuspended
in 1/10 of the original volume of the same buffer. Cell-free extract was prepared
by sonication of the washed cell suspensions on ice. The activity of glucose
isomerase was measured by incubating a reaction mixture that contained 10 mM
MgSO,, 1 mM CoClz, 0.8 M glucose, and the enzyme in 100 mM MOPS buffer (pH
7.0) at 65°C. For the assay of xylose isomerase activity, the reaction mixture was
incubated at 37°C and contained 70 mM xylose, 10 mM MnSO,, and the enzyme
preparation in the same buffer. The amount of fructose or xylulose formed after the
enzyme reaction was estimated by the cysteine-carbazole sulfuric acid method (11).
One unit of activity is defined as the amount of enzyme which released lttmol of
ketose per min under the assay conditions described above. Protein concentration
was determined by the method of Lowry et a1. (20). Bovine serum albumin was

used as a standard.

112
Purification of glucose isomerase

Cell extract preparation: 10 g of wet weight E. coli W595/pCGI38 cells or 3
g of wet weight B. subtilis NAl/pMLGl cells were suspended in 4 volume of 50 mM
MOPS buffer (pH 7.0) containing MgSO4 (10 mM) and CoCl2 (1 mM). The cells
were broken by two passages through a French pressure cell (American Instrument
Co. Inc., Silver Spring, MD) at 18,000 p.s.i. The debris was removed by
centrifugation at 12,000 x g for 20 min.

Heat treatment step: The cell extracts obtained from the broken cells were
heated at 85°C for 15 min, and cooled to 4°C. The precipitated protein was
removed by centrifugation at 12,000 x g for 20 min.

Anion exchange column chromatography: DEAE-sepharose CL-6B columns were
equilibrated with 50 mM MOPS buffer (pH 6.8) containing 5 mM MgSO4 and 0.5
mM C00,. The soluble fractions from the heated cell extracts were loaded onto the
columns and eluted with a NaCl salt gradient (0.0-0.5M) in the same buffer. The
fractions containing significant glucose isomerase activity were pooled.

Gel filtration chromatography: A prepacked gel filtration column, Superose 12
HR 10/30 (Pharmacia, Inc., Picataway, NJ.) was used with the Pharnacia FPLC
System. The pooled fractions from the previous step were concentrated into 0.2 ml
of enzyme solution by using ultrafiltration membrane YM30 with 30 KDa molecular
weight cut-off (Amicon; Division of W.R. Grace and Co., Danverse, MA) prior to the
loading. Blue dextran (2000 KDa), alcohol dehydrogenase (150 KDa), and bovine

serum albumin (66 KDa) were used as molecular weight standards.

113
Time course of glucose conversion into fructose
Glucose solution (35%, wt/wt) was incubated with the heat u'eated cell extract
from B. subtilis NAl/pMLGl as a glucose isomerase source (10.8 unit/g of dry
substrate) in 50 mM MOPS buffer (pH 7.0) containing MgSO. (5 mM) and CoCl2
(0.5 mM) at 70°C. Quantitative and qualitative analyses for glucose and fructose
were performed by HPLC as described elsewhere (C. Lee, B.C. Saba, and J .G. Zeikus,

manuscript submitted to Appl. Environ. Microbiol.).

RESULTS

Expression of recombinant glucose isomerase

The recombinant plasmid pCGI38 was able to complement the xyl-5 mutation in
E. coli strain HB101. It contained the vector pHSG262 and an insert fragment of
4.0 kb (Figure 2). Since pCG38 originated from a cosmid library we must conclude
that the insert had undergone a partial deletion during propagation in the E. coli host.
Upon introduction into different xylA mutants of E. coli, pCGIB8 conferred the xyl‘
phenotype to the hosts and the ability to express a thermostable xylose and glucose
isomerase (Table 1).

A new recombinant plasmid, pMLGl, was constructed by ligation of pCGI38
with a Bacillus vector pTB522 at the EcoRI site. This double-replicon could be
maintained in both E. coli and B. subtilis. As shown in Table 2 this plasmid
conferred the ability to express thermostable glucose isomerase to both organisms.
The orientation of the insert DNA fragment in the vector did not affect the level of
glucose isomerase in either of the two hosts (results nor shown). We conclude,
therefore, that the cloned xylA gene from C. thermosulfurogenes was still under the
control of its own promoter and this promoter (or promoter region) was expressed
well in each of the three bacterial hosts tested.

In C. thermosulfurogenes, the glucose isomerase gene was not expressed unless
D-xylose was present in the medium (Table 3). The gene present on the cloned
DNA fragment was expressed constitutively in both E. coli and B. subtilis and the

addition of o-xylose to the medium did not increase its production (Table 3).

114

115

Figure 2. Physical and genetic maps of plasmids pCGI38 and pMLGl, which
carry the C. thennosulfurogenes DNA insert expressing thermostable glucose
isomerase.

Derivation of DNA fragments represent :3 , pHSG262; —---- , pTBSZ3;
— , C. thermosulfurogenes. Abbreviations for the restriction sites are: A, Ach;
B, BalI; Bg, BglII; C, ClaI; E, EcoRI; M, MluI; P, PstI; S, SmaI; X, XbaI.
Kanamycin and tetracycline resistance genes are indicated by [IIIII and g ,

respectively.

116

 

117

Table 1. Comparison of glucose/xylose isomerase activities in E. coli xyl‘ mutant strains

carrying the recombinant plasmid pCGI38.‘

 

Specific activity (unit/mg protein)

E. coli Glucose Xylose

 

Strain Plasmid (65°C) (37°C)
C. thermosulfurogenes -- 0.29 0.07
(control)

JM107 (xyl‘) -- 0.00 0.08
HB101 pHSG262 0.00 0.00
HB 101 pCG138 0.21 0.05
W595 pHSG262 0.00 0.00
W595 pCG138 0.46 n.d.

 

‘E. coli transformants carrying pCGI38 or pHSG262 were cultured in LB media with 1%
xylose and kanamycin. Strain JM107 was grown in LB medium with 1% xylose. C.
thermosulfurogenes was grown anaerobically in TYE medium with 1% xylose at 60°C.

n.d.: not determined.

118

Table 2. Expression of cloned glucose isomerase by recombinant shuttle plasmid pMLGl

in E. coli and B. subtilis‘

 

 

Antibiotics Specific activity
Host Plasmid resistance (unit/mg protein)
E. coli W595 pHSG262 Kan' 0.00
E. coli W595 pMLGl Kan’, Tet‘ 0.50
B. subtilis NA1 pTB522 Tet' 0.00
B. subtilis NA1 pMLGl Tet', Kan‘ 1.54

 

' Cells were cultured in LB medium with 1% xylose and apprOpriate antibiotics. Cells were
harvested and washed in 50 mM MOPS buffer (pH 7.0) before enzyme assay at 65°C and

protein determination.

119

Table 3. Effect of D-xylose and D-glucose on synthesis of recombinant thermostable

glucose isomerase in E. coli and B. subtilis.‘

 

Strain'

Plasmid

Specific Activity (unit/mg protein)

Carbohydrates added

 

xylose glucose none
C. thermosulfurogenes -- 0.29 0.00 --
(control)
E. coli W595 pHSG262 0.00 0.00 0.00
E. coli W595 pCGI38 0.46 0.24 0.43
B. subtilis NA1 pTB522 0.00 0.00 0.00
B. subtilis NA1 pMLGl 1.54 1.29 1.47

 

' E. coli and B. subtilis were grown in LB media with 1% carbohydrate as indicated and '

C. thermosulfurogenes was grown in TYE medium with 1% carbohydrate prior to assay of

activity at 65°C.

120
A reduction of specific activity upon addition of glucose might be a result of

catabolite repression.

Properties of recombinant glucose isomerase

The general biochemical properties of glucose isomerase, synthesized in E. coli
and B. subtilis cells, containing the gene from C. thermosulfurogenes, were very
similar to those determined for the enzyme produced by its original host. The
enzyme maintained 80 - 95% of the original activity after preincubation of cell
extracts at 80°C (Figure 3). The apparent pH and temperature optima for the activity
of the recombinant glucose isomerase isolated from the mesophilic bacterial hosts
were 80°C and pH 7.0.

The usefulness of the heat treatment step in purification of thermostable
glucose isomerase from either E. coli or B. subtilis is shown in Table 4. After the
heat treatment of crude cell extracts the enzyme was obtained at 70 - 80% purity.
After additional steps of purification by anion exchange chromatography and gel
filtration, preparations of homogeneous glucose isomerase were obtained on SDS-
polyacrylamide gel electrophoresis. The enzyme isolated from either C. '
thermosulfurogenes or the mesophilic organisms consisted of one type of subunit with
a molecular weight of 50 kDa (Figure 4). Upon gel filtration of the native enzyme
on Superose-12 a molecular weight of 200 kDa was observed indicating that the
native enzyme molecule is a tetramer of identical subunits.

As indicated by the results in Figure 4, the C. thermosulfurogenes glucose

isomerase gene was expressed at high levels from the recombinant plasmids in both

121

Figure 3. Comparison of native versus cloned glucose isomerase thermal
stabilities.

- Symbols represent: H , C. thermosulfurogenes (CT); H, E. coli (EC)
W595/pCGI38; O-—Ot (BS) B. subtilis NA1/pMLGl; H, E. coli JM107; and,
H, B. subtilis NA1. Residual activities were assayed at 65°C (except controls,
E. coli JM107 and B. subtilis NA1, which were assayed at 37°C) after preincubation

of cell extracts at 80°C for various time periods.

Residual Activity (%)

 

122

 

 

 

 

100 .w ' -' E.C.
‘fl . . .
O 8.5
0 e
80-
CIR
60-4
4o-
, 20" 8.8. Control
/ . E.C. Control
{—1 (L
M; 4 .1 1.? 1: fl?

 

 

 

S 10 20 30 40 50

Time (min)

 

123

Figure 4. SDS-PAGE analysis of glucose isomerase activity fractions.

Various fractions obtained during purification of native and recombinant glucose
isomerase were analyzed by 12% SDS-PAGE, and visualized by Coomassie blue
staining. Lane C.t., glucose isomerase purified from C. thermosulfurogenes; 1, whole
cell preparation from E. coli W595/pHSG262 or B. subtilis NA1/pTB522 that did not
carry the DNA insert; 2, whole cell preparation from E. coli W595/pCGI38 or B.
subtilis NA1/pMLGl; 3, soluble fraction from heat-treated cell extract of E. coli
W595/pCGI38 or B. subtilis NA1/pMLGl; 4, glucose isomerase purified from E. coli
W595/pCGI38 or B. subtilis NA1/pMLGl; S.t., molecular weight standards (97,400.
phosphorylase; 66,200. bovine serum albumin; 42,700. ovalbumin: 31,000. carbonic

anhydrase; 21,500. soybean trypsin inhibitor).

124

E.coli I B. subtilis1
C.t.1 2 34' 1 2 34 S.t.

 

125

E. coli and B. subtilis. This was confirmed by the fact that 7-fold purification was
required to generate a homogeneous enzyme preparation from crude extracts of E. coli
and only 3-fold purification from the extracts of B. subtilis (Table 4).

To test the practical utility of recombinant thermostable glucose isomerase
produced by food-safe strains of B. subtilis, the time course of fructose formation
from concentrated glucose syrups were tested at temperatures higher than those used
in current industrial practice. As shown in Figure 5, the isomerization reaction
reached equilibrium and achieved a yield of 52% fructose at 70°C and pH 7.0 after

8 hours of incubation.

126

Table 4. Purification scheme of cloned thermostable glucose isomerase from E. coli and B.

 

 

 

subtilis"
Total Total Specific Purifi-
protein activity activity cation
Step (mg) (units) (units/mg) (fold)
E.C. B.S. E.C. B.S. E.C. B.S. E.C. B.S.
Crude cell 800 122 400 159 0.5 1.3 1 l
extract
Heat treatment 144 37 331 96 2.4 2.6 5 2
(85°C, 15 min)
DEAE-sepharose 88 28 282 84 3.2 3.5 6.4 2.8
anion exchange
FPLC-Superose 12 39 15 136 56 3.5 3.7 7 2.9

gel filtration

 

Abbreviations: E.C., E. coli W595 containing the plasmid pCG138; and, B.S., B. subtilis NA1

containing the plasmid pMLGl.

127

Figure 5. Time course of glucose conversion into fructose using heat-treated
glucose isomerase obtained from recombinant B. subtilis.
Glucose isomerization was performed at 70°C under conditions described in

Methods.

100
o\° 80
E
.9.
g 60
0
26
U)
a 40
(I)

20

128

 

 

 

Time ( hours)

ll—o— Fructose
u—O— Glucose

~—
-—
_

24

 

DISCUSSION

This work demonstrates cloning of the gene for thermostable xylose and glucose
isomerase from C. thermosulfurogenes. As far as we are aware, this is the first
glucose isomerase gene from a thermophilic bacterium that has been cloned and over-
expressed in E. coli and B. subtilis. A new screening method, based on a specific
assay that detected conversion of fructose to glucose on agar plates facilitated the
isolation of positive clones. In E. coli xylA mutants, detection of over-produced
xylose isomerase was also possible on McConkey/xylose agar plates. However, the
direct assay of fructose to glucose conversion is not only more specific, it is also
applicable to non-fermenting, mesophilic or thermophilic hosts.

When the cloned xylA gene of E. coli, containing its own promoter, was
intoduced into B. subtilis, it was not expressed unless a B. licheniformis promoter was
inserted upstream (13). However, the deduced amino acid sequence of xylose
isomerase gene from C. thermosulfurogenes exhibited a high degree of homology to
those of both E. coli and B. subtilis xylA genes, and it was expressed well from a
putative promoter region located upstream of the structural gene, in each of the three ~
organisms used in this study (Table 2, 3, and C. Lee, et al., manuscript submitted to
J. Biol. Chem).

It is known that single amino acid substitutions in the primary structure of an
enzyme may alter dramatically its biochemical properties including thermostability,
turnover number or substrate specificity (8,30). The identical biochemical properties

of thermostable glucose isomerase produced in its native host and in E. coli and B.

129

130
subtilis indicates that the cloned gene did not undergo significant mutational changes
upon manipulation and propagation in the new hosts.

High thermostability of C. thermosulfurogenes glucose isomerase made it
possible to use heat treatment of crude cell extracts as one of the most efficient
purification steps when the enzyme was produced in mesophilic hosts (Figure 4).
The presence of metal ions (Mg“ and Co“) and high protein concentration in cell
extracts were essential for optimal recovery of the enzyme during the heat treatment
step. The Bacillus subtilis host strain used here lacked neutral protease but contained
alkaline protease. The heat treated recombinant glucose isomerase preparation
recovered from B. subtilis lacked detectable protease activity.

Analysis of crude extracts and of the purified preparations by SDS-PAGE and
by specific activity determinations indicated that the thermostable glucose isomerase
specified by the recombinant plasmids used in this study was one of the most
abundant proteins in the cell extracts and acounted for approximately 14% of total cell
extract protein in E. coli and 34% in B. subtilis. Heterologous proteins over-
produced in E. coli often form insoluble "inclusion bodies" (7). Whether a protein
precipitates in the cytoplasm or remains soluble, however, depends also on the nature
of the protein being over-produced (5). In the case of the recombinant thermostable
glucose isomerase, the enzyme remained soluble which may be an important factor
in attempts to scale-up its production.

Findings of this study also suggest potential of utilizing the recombinant
thermostable glucose isomerase over-expressed in B. subtilis for sweetener production.

The neutral pH optimum for the activity of the cloned glucose isomerase is suitable

131
for the neutral or slightly acidic reaction conditions which reduce by-products and
color formation during fructose production. The high thermostability of the enzyme
allowed us to operate the isomerization reaction at high temperature and hence to shift
the equilibrium towards higher fructose (17). Since, B. subtilis is a meSOphilic and
non-pathogenic host, it is feasible to produce a food-safe and thermostable glucose
isomerase with a practical grade of purity (80%) by a simple heat treatment step.
Also, constitutive production of the enzyme in B. subtilis will reduce the enzyme cost
by saving on expensive inducers. Further studies on the long-term thermostability,
metal ion requirements, and safety of the enzyme are needed before practical
industrial utility of the recombinant thermostable glucose isomerase in B. subtilis can

be assessed.

10.

11.

12.

LITERATURE CITED

Bachmann, B. J. 1972. Pedigrees of some mutant strains of Escherichia coli
K-12. Bacteriol. Rev. 36:525-557.

Batt, C. A., E. O’Neill, S. R. Novak, J. Ko, and A. Sinskey. 1986. Hyper-
expression of Escherichia coli xylose isomerase. BiOtechnology Progress.
2:140-144.

Birnboim, H. C., and J. Doly. 1979. A rapid alkaline extraction procedure for
screening recombinant plasmid DNA. Nucleic Acids Res. 7:1513-1523.

Brady, G., H. M. Jantzen, H. U. Bernard, R. Brown, T. Hashimoto—Gotoh,
and G. Schiietz. 1984. New cosmid vectors developed for eukaryotic DNA
cloning. Gene 27:223-232.

Buell, G., and N. Panayotatos. 1986. Mechanism and practice, p. 345-363.
In W. Reznikoff and L. Gold (ed.), Maximizing gene expression. Butterworths
Publishers. Boston.

Clewell, D. B. 1972. Nature of Col E1 plasmid replication in Escherichia
coli in the presence of chlorarnphenicol. J. Bacteriol. 110:667-676.

Craik, C. S. 1987. Expression and overexpression of proteins discussion group
summary, p. 467-476. In D. L. Oxender (ed.), Protein structure, folding, and
design 2. Alan R. Liss, Inc. New York.

Craik, C. S., C. Largman, T. Fletcher, S. Roczniak, P J. Barr, R. Fletterik,
and W. J. Rutter. 1985. Redesigning trypsin: Alteration of substrate specificity.
Science. 228:291-297.

Crueger, W., and A. Crueger. 1982. Enzymes, p. 161-174. In W. Crueger and
A. Crueger (ed.), Biotechnology: a textbook of industrial microbiology, Science
Tech. Inc., Madison.

Dilella, A. G., and S. L. C. Woo. 1987. Cloning large segments of
genomic DNA using cosmid vectors. Methods Enzymol. 152:199-212.

Dishe, Z., and E. Borenfreund. 1951. A new spectrophotometric
methods for the detection and determination of keto sugars and trioses.
J. Biol. Chem. 192:583-587.

Frischauf, A.-M. 1987. Construction and characterization of a genomic
library in lambda vectors. Methods Enzymol. 152:190—199.

132

13.

14.

15.

l6.

17.

18.

19.

20.

21.

22.

23.

24.

133

Huang, J. J., and N. W. Y. Ho. 1985. Cloning and expression of the
Escherichia coli D-xylose isomerase gene in Bacillus subtilis. Biochem.
Biophys. Res. Commun. 126:1154-1160.

Hyun, H. H., and J. G. Zeikus. 1985. General biochemical
characterization of thermostable extracellular B-amylase from Clostridium
thermosulfurogenes. Appl. Environ. Microbiol. 49:1162-1167.

Imanaka, T., M. Fujii, and S. Aiba. 1981. Isolation and characterization
of antibiotic resistance plasmids from thermophilic bacilli and
construction of deletion plasmids. J. Bacteriol. 146:1091-1097.

Imanaka, T., T. Himeno, and S. Aiba. 1985. Effect of in vitro DNA
rearrangement in the NHz-terminal region of the penicillinase gene from
Bacillus licheniformis on the mode of expression in Bacillus subtilis.
J. Gen. Microbiol. 131:1753.1763.

Jorgenson, O. B., L. G. Karlsen, N. B. Nielson, S. Pedersen, and S.
Rugh. 1988. A new immobilized glucose isomerase with high
productivity produced by a strain of Streptomyces murinus. Starch.
40:307-313.

Kuriki, T., S. Okada, and T. Imanaka. 1988. New type of pullulanase
from Bacillus stearothermophilus and molecular cloning and expression
of the gene in Bacillus subtilis. J. Bacteriol. 170:1554-1559.

Lambert, P. W., and J. L. Meers. 1983. The production of industrial
enzymes. Phil. Trans. R. Soc. Lond. B300:263-282.

Lowry, O. H., N. J. Rosebrough, A. I. Farr, and R. J. Randall. 1951.
Protein measurement with the Folin phenol reagent. J. Biol. Chem.
193:265-275.

Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning:
a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring
Harbor, N.Y.

Marcel, T., D. Drocourt, and G. Tiraby. 1987. Cloning of the glucose
isomerase (D-xylose isomerase) and xylulose kinase genes of
Streptomyces violaceoniger. Mol. Gen. Genet. 208:121-126.

Marmur, J. 1961. A procedure for the isolation of deoxyribonucleic acid
from microorganisms. J. Mol. Biol. 3:208-218.

Panagiotis, E. S., and N. W. Y. Ho. 1985. Over-production of D-xylose

25.

26.

27.

28.

29.

30.

31.

134

isomerase in Escherichia coli by cloning the D-xylose isomerase gene.
Enzyme Micro. Technol. 7:592-596.

Schellenberg, G. D., A. Sarthy, A. E. Larson, M. P. Backer, J. W.
Crabb, M. Lidstrom, B. D. Hall, and C. E. Furlong. 1984. Xylose
isomerase from Escherichia coli: characterization of the protein and the
structural gene. J. Biol. Chem. 259:6826-6832.

Schink, B., and J. G. Zeikus. 1983. Clostridium thermosulfurogenes
sp. nov., a new thermophile that produces elemental sulphur from
thiosulphate. J. Gen. Microbiol. 129:1149-1158.

Takasaki, Y., Y. Kosugi, and A. Kanbayashi. 1969. Studies on sugar-
isomerizing enzyme; purification, crystallization and some properties of
glucose isomerase. Arg. Biol. Chem. 33:1527-1534.

Wilhelm, M., and C. P. Hollenberg. 1985. Nucleotide sequence of the
Bacillus subtilis xylose isomerase gene: extensive homology between
the Bacillus and Escherichia coli enzyme. Nucleic Acids Res.

13:5717-5723.

Wovcha, M. G., D. L. Steuerwald, and K. E. Brooks. 1983.
Amplification of D-xylose and D-glucose isomerase activities in
Escherichia coli by gene cloning. Appl. Environ. Microbiol. 45:1402-
1404.

Yutani, K., K. Ogasahara, Y. Sugino, and A. Matsushiro. 1977. Effect
of a single amino acid substitution on stability of conformation of a
protein. Nature. 267:274-275.

Zeikus, J. G., A. Ben-Bassat, and P. Hegge. 1980. Microbiology of
methanogenesis in thermal, volcanic environments. J. Bacteriol.
143:432-440.

 

Chapter V

Molecular Characterization of the Xylose (Glucose) Isomerase Gene from

Clostridium thermosulfurogenes: Role of Hism in Enzymatic Catalysis

135

ABSTRACT

The gene coding for thermophilic xylose (glucose) isomerase of Clostridium
thermosulfurogenes was isolated and its complete nucleotide sequence was determined.
The structural gene (xylA) of the xylose isomerase encodes a polypeptide of 439
amino acids with an estimated molecular weight of 50,474. The deduced amino acid
sequence of thermostable C. thermosulfurogenes xylose isomerase showed 50% and
70% homology to those of thermolabile xylose isomerases from Escherichia coli and
Bacillus subtilis, respectively, whereas, it displayed much lower homology (22-24%)
to those of thermostable xylose isomerases from Ampullariella, Arthrobacter, and
Streptomyces violaceoniger. However, several discrete regions which are highly
conserved throughout amino acid sequences of all six xylose isomerases could be
detected. To elucidate the catalytic mechanism and the role of histidine in the active
site of xylose isomerase, four histidine residues at different positions in the C.
thermosulfurogenes enzyme were individually changed to phenylalanine by site-
directed mutagenesis. Substitution of His101 by phenylalanine completely abolished
enzyme activity whereas the other histidine substitutions had no effect on enzyme
activity. When His101 was changed to glutamine, approximately 10% of wild-type
enzyme activity was retained without appreciable change in the apparent K., for
glucose or xylose. The Gln101 mutant enzyme was not inactivated by treatment with
10 mM diethylpyrocarbonate while the xylose isomerases containing histidine at
position 101 were totally inactivated. In addition, the log Vmax“, values of the

Gln101 mutant enzyme at different pH values over the range of 5.0-8.0 remained

136

137
constant, whereas those of the wild-type (llism) enzyme decreased markedly at pH
6.5-5.0. These findings are consistent with the proposed catalytic mechanism for
glucose or xylose isomerization that is accomplished by hisridine in the active site

acting as a general base catalyst.

Introduction

Xylose isomerase (D-xylose ketol isomerase, EC 5.3.1.5) is an intracellular
enzyme that catalyzes the isomerization reaction of D-xylose to D-xylulose. Its
practical significance and commercial importance stems from its ability to isomerize
D-glucose and convert it to o-fructose. It is, therefore, often referred to as glucose
isomerase and is widely used in industry for production of high fructose syrups (1-
3). Practical use of the enzyme has stimulated research on the physicochemical
properties, function and structure of the enzyme from different sources and on the
organization and regulation of its gene in different microorganisms (for review see 3-
9). Thus, organization of the xyl operon in Escherichia coli (10, 13), Bacillus
subtilis (11,14) and Salmonella typhimurium (12) has been established by genetic and
nucleotide sequence analysis. In addition, the complete nucleotide sequence of the
xylA gene from Ampullariella sp (15) and Streptomyces violaceoniger (16) were
reported

It has been proposed that enzymatic isomerization of xylose to xylulose and
glucose to fructose involves a cis-enediol intermediate via intramolecular hydrogen -
transfer and is accomplished by general base catalysis (17,18). Inhibition studies
on Streptomyces glucose isomerase by chemical modification suggest the involvement
of a histidine residue as the general base in the catalytic center of the enzyme (19).
Recent crystallographic studies on xylose isomerases from S. rubiginosus, as well as
on the enzyme from Arthrobacter, have indicated a putative structure of the active site

and suggested that histidine residue at position 54 is involved in initial abstraction of

138

139

the proton from the first carbon of the substrate and thus assisted the formation of
enediol intermediate (20,21). However, no biochemical evidence in support of the
proposed mechanism and the involvement of Hiss4 in the catalysis has been published
to date.

In our laboratory, xylose isomerases from thermoanaerobic bacteria have been
characterized for their biochemical properties (Lee, C. et a1. manuscript submitted to
J. Bacteriol.), and the gene coding for xylose isomerase of Clostridium
thermosulfurogenes has been cloned and over-expressed in both E. coli and B. subtilis
(Lee, C. et a1. manuscript submitted to Appl. Environ. Microbiol.). Xylose isomerase
of C. thermosulfurogenes purified from its original host or from either of the two
recombinant mes0philic organisms exhibited high thermosrability up to 85°C and was
crystallized by the vapor diffusion method.

In this report we have isolated the gene coding for the Clostridium
thermosulfurogenes xylose isomerase, which we propose to call xylA in accordance
with the nomenclature used for E. coli. We have determined the complete nucleotide
sequence of this gene and demonstrated, by site-directed mutagenesis, the role of
histidine1m in catalysis and a glutamine1m mutant enzyme that displays acid stable

isomerase activity.

MATERIALS AND METHODS

Bacterial strains, enzymes, and chemicals

E. coli strain HBlOl [Fhst20 ara-l recA13 proA12 lach galK2 VpoL20 mtl-
1 xyl-5] (22) was used for expression of the cloned C. thermosulfurogenes xylose
isomerase gene and E. coli strain JM107 [endAl gyrA96 thi hst17 supE44 relAl
A(lac-proAB) / F’ traD36 proAB lacIq ZAM15] (23) was used as a host cell and lawn
cells during nucleotide sequence determination and site-directed mutagenesis. E. coli
strain GM33 [dam-3] (24) was used to confirm a methylated ClaI restriction site in
the cloned DNA. Restriction endonucleases and enzymes for subcloning experiments
were obtained from Bethesda Research Laboratories, Gaithersburg, MD. and New
England Biolabs, Inc., Beverly, MA.. [or-”S]dATP (500 Ci/mmol) was obtained from
Du Pout-New England Nuclear, Wilmington, DE.. All chemicals were of reagent

grade and of the highest purity available.

Subcloning and manipulation of DNA

A 5.2 kilobase EcoRI/Smal DNA fragment containing the C. thermosulfurogenes
xylose isomerase gene was isolated from plasmid pCGI38, and subcloned between the
EcoRI and SmaI sites of expression vectors pMMB67EH and pMMB67HE (25). The
resulting recombinant plasmids contained the xylA gene in both orientations with
respect to the inducible tac promoter of each vector. Determination of glucose
isomerase activity, after induction with IPTG, in E. coli cells containing these

recombinant plasmids, indicated that the transcriptional orientation of the xylA gene

140

141

is in the direction from the Ach site towards the BglII site as shown in the physical
map of pCMG1-2 plasmid in Figure 1. A series of derivatives, containing deletions
at each end of the insert DNA, were generated by Bal3l nuclease treatment of
pCMGl-2, after linearization with either BamHI or EcoRI. Shortened DNA
fragments were blunt ended by treatment with T4 DNA polymerase and reinserted as
EcoRI-blunt-end or BamHI-blunt-end fragments into the vector pMMB67EH (26).
Complementation of xylose isomerase activity in the strain HB1010 by the deletion
plasmids was scored on MacConkey agar plates containing 1% xylose in the presence
or absence of IPTG (0.2 mM).

The alkaline denaturation method was used for rapid extraction of plasmid
DNA (27). Recovery of DNA from agarose gels was performed by using an
electroelution apparatus, Elutrap (Schleicher and Schuell, Inc., Keene, NH).
Digestion of DNA with restriction endonucleases, separation of fragments by
electrophoresis, fill-in reaction of recessed ends in DNA fragments, ligation of DNA
fragments, and transformation of E. coli with plasmid DNA were performed as
described by Maniatis et a1. (28) or as recommended by the manufacturers of

enzymes.

Nucleotide sequence determination

DNA inserts (see Figure 1) of pCMG11-3 and pCMG11-9 were digested with
restriction endonucleases, and the resulting DNA fragments were isolated and
subcloned into M13mp18 or M13mp19 bacteriophage vectors. Nucleotide sequences

of the subcloned inserts (see Figure 2) were determined by the dideoxy chain

142

Figure 1. Deletion mapping of the xylose isomerase gene (xylA) of C.
thermosulfurogenes.

The partial restriction map around the xylA gene is shown at the top. Open
bars (:3) represent C. thermosulfurogenes DNA, and the closed bar (—) and

) represent the DNA of pHSG262 (41) and pMMB67EH vector

 

the lines (
plasmids, respectively. The triangles indicate an inducible tac promoter of
pMMB67EH. Gene expression by each deletion plasmid were detected by
complementation of xylose isomerase activity of E. coli HBlOI on MacConkey agar
plates containing 1% xylose. The closed bar marked xylA denOtes the approximate
position of the structural gene xylA. Restriction sites: E, EcoRI; P, PstI; C, ClaI;

A, Ach; Bg, BglII; Hc, HincII; S, SmaI; B, BamHI; Ha, HaeIII.

143

 

 

 

 

 

 

 

 

 

 

 

Enzyme
Plasmid Expression
+ IPTG -IPTG
tac E PC A C P 89 He S B
pCMG1-2 W 11 l I l IJ + +
E e
pCMGZ-S W l__ + 4.
E e
pCMG2-3 W |L_ _ _
E e
pCMG10-4 W 11— + +
E B
pCMG11-3 W:— + +
E SB
pCMGll-Q WET—i; + _
xylA

144

Figure 2. Sequencing strategy for DNA encoding C. thermosulfurogenes glucose
isomerase.

The open bar (1:!) represents the coding region. The solid lines (_)
represent the EcoRI/BamHI fragments from the recombinant plasmids, pCMG11-9
and pCMG11-3 (Figure 1). The arrowes indicate the direction and extent of the
nucleotide sequence determined by the dideoxy chain termination method (29). For
the abbreviation of restriction sites see the legend to Figure 1 except Ha for HaeIII.

bp, base pairs

145

 

_Llll ll ”11"—

 

 

 

 

 

 

pCMG11-9:
pCMG‘l1-3:
‘— fi 100 bp

146
termination method (29), using a Sequenase Version 2.0 Kit (United States
Biochemical Co.). The sequence information was analyzed using the Sequence
Analysis Software Package of the Genetics Computer Group, version 5, (University

of Wisconsin) (30).

Site-directed mutagenesis

Plasmid pCMG11-3 was digested with EcoRI and BamHI. The 1.4 kilobase
DNA fragment, containing the entire C. thermosulfurogenes xylA gene and its
promoter region was isolated and subcloned into the EcoRI/BamHI sites of M13mp19.
The oligonucleotide primers, obtained from Genetic Designs, Inc., (Houston, TX),
were designed to be complementary to the single-strand template DNA and to contain
appropriate mismatches as indicated in Figure 3. Synthesis of mutant genes and
selection were performed by the method of Eckstein and co-workers (31), using a kit
of Oligonucleotide-directed Mutagenesis System, version 2 (Amersham Co., Arlington
Heights IL). Nucleotide sequences of the mutant genes obtained were confirmed by
the di-deoxy sequencing method as described above. The 1.4 kilobase EcoRI/BamHI
fragments, containing the mutant genes were subcloned from each M13mp19
recombinant into the EcoRI and BamHI sites of pMMB67EH vectors, and introduced

into E. coli strain HBlOl.

Enzyme purification
E. coli cells, carrying appropriate recombinant plasmids that specified mutant or

wild-type xylose isomerase were cultivated overnight in LB media (28), containing

147

Figure 3. Synthetic oligonucleotide primers for site-directed mutagenesis.
The template DNA represents the coding strand of the xylA gene inserted in

M13mp19.

148

40 4S
Template DNA : 3' - TTC TGC TAC CTC CTC GTA GAA GCG AAA AGA TAT CGA - 5'
(Wild type) Lys Thr Met Gln Glu His Len Arg Phe Ser Ile Ala
Primer I : 5' -- AG AC6 ATG GAG GAG TTT CTT CGC TTT TCT ATA GC -- 3'
(for Phe4l) Phe

 

. 65 70 75
Template DNA : 3' - TAC GTT TCC GGT ACC TTA GTG ATA TGT CTA GGA TAC CTG - 5'
(Wild type) Met Gln Arg Pro Trp Asn His Tyr Th: Asp Pro Met Asp
Primer II : 5' --- G CAA AGG CCA TGG AAT TTC TAT ACA GAT CCT ATG G --- 3'
(for Phe71) . Phe
95 100 105
Template DNA : 3’ - TAT TTA CGT GGC ATA AAG ACG AAG GTA CTA TCT CTA TAA CGG GGA - 5’
(wild type) Ila Asn Ala Pro Tyr Phe Cy: Phe His Asp Arg Asp Ile Ala Pro
Primer III : S' --------- GCA COG TAT TTC TGC TTC TTT GAT AGA GAT ATT GCC CC -- 3'
(for PhelOl) Phe
Primer IV : S' ----------- A CCG TAT TTC TGC TTC CAA GAT AGA GAT ATT GCC ----- 3'
(for GlnlOl) Gln
Primer v : 5' -- TA rut'r GCA occ rxr r'rc GCC rrc CA‘.‘ GAT AGA CAT A ----------- 3'
(for Ala99) Ala
150 155
Template DNA : 3' - AGG TTA GOT TCT AAA CAC GTA CCA CG? AGT TGC AGA - 5'
(wild type) Ser Asn Pro Arg Phe Val 8i: Gly Ale Ser Thr Ser
Primer VI : S' -- CC AAT CCA AGA TTT GTG TTT GGT GCA TCA ACG TC -- 3'

(for PhelSZ) Phe

149
100 ug/ml Ampicillin, at 37°C with vigorous shaking. Cells harvested by
centrifugation at 5,000 x g were washed and suspended in 50 mM MOPS buffer (pH
7.0) containing 10 mM MgSO. and 1 mM CoC12. For preparation of cell extracts,
cells were broken by two passages through a French pressure cell at 18,000 p.s.i., and
the debris was removed by centrifugation at 12,000 x g for 20 min. The cell
extracts were stirred at 85°C for 20 min and centrifuged at 12,000 x g for 30 min.
In the case of the His-Phe”2 mutant enzyme, 60°C was used during the heat treatment
because this mutant enzyme exhibited lower thermostability than the wild-type
enzyme. The soluble fractions from the heated cell extracts were loaded onto a
DEAE-Sepharose CL—6B columns (4.0 cm x 32 cm), pre-equilibrated with 50 mM
MOPS buffer pH 7.0, and proteins were eluted with linear NaCl salt gradients (0.0-
0.5M) in the same buffer. Fractions containing significant glucose isomerase activity
were pooled. For the His-Phe“,I mutant protein, which lacks enzyme activity, the

protein was identified by SDS-PAGE analysis from the fractions.

Enzyme assays

Reaction mixtures (1 ml) contained: 0.8 M glucose, 10 mM MgSO,, 1 mM
C00,, and the enzyme in 100 mM MOPS buffer (pH 7.0). For the assay of xylose
isomerase activity, the reaction mixture (1 ml) contained: 70 mM xylose, 10 mM
MnSO,, and the enzyme in 100 mM MOPS buffer (pH 7.0). After 30 min
incubation at 65°C, 1 ml of 0.5 M perchloric acid was added to stop the reaction and
the mixture was further diluted 10- and 50-fold with double distilled water before the

addition of cysteine-carbazole-sulfuric acid reagent (32). Fructose isomerizing

150

activity was assayed under the same reaction conditions as in the glucose isomerase
assay except for the substitution of 0.4 M fructose as a substrate in the reaction
mixture. After 30 min incubation at 65°C, the reaction was terminated by placing
the reaction tubes in ice water, and the amount of glucose formed was determined
with a glucose analyzer model 27 (Yellowstone Instrument Co. Inc., Yellow Springs,
OH). One unit of activity is defined as the amount of enzyme that produced 1 umol
of product per min under the assay conditions described above. Apparent K., and
V.,, values were obtained from Lineweaver-Burk plots of specific activities at various
substrate concentrations. Protein concentration was determined by the method of

Lowry et a1. (33) with bovine serum albumin as standard.

RESULTS AND DISCUSSION

Expression of C. thermosulfurogenes xylA gene

Physical maps of the DNA fragment containing the C. thermosulfurogenes xylA
gene and the deletion fragments generated from it are shown in Figure 1.
Expression of glucose isomerase by E. coli HB101, harboring recombinant plasmids
containing these DNA fragments is presented in Table 1. The xylA gene in plasmid
pCMG11-3 (and in all recombinant plasmids carrying bigger DNA inserts; results not
shown) was expressed constitutively, whereas the expression from the smaller DNA
fragment, present in the plasmid pCMG11-9 occurred only in the presence of IPTG.
These results indicate that the region of approximately 0.1 kilobase upstream of the
xylA gene, present in the plasmid pGM11-3, contains promoter sequences functional
in E. coli, presumably the original promoter of the C. thermosulfurogenes xylA gene,
whereas the 1.2 kilobase insert of the plasmid pCMGl 1-9 contains the structural gene

of xylose isomerase that is expressed from the me promoter of the vector.

Nucleotide sequence of C. thermosulfurogenes xylA gene

The nucleotide sequence of the insert in the recombinant plasmid pCMG11-3
was 1434 base pairs in length and contained an open reading frame Starting with the
ATG codon at nucleotides 79-81 and ending with two termination codons TAA and
TGA at nucleotides 1396-1401 (Figure 4). It was confirmed as the correct reading
frame for the C. thermosulfurogenes xylA gene by comparison of the N-terminal

amino acid sequence (7 amino acids underlined in Figure 4), determined on the xylose

151

152

Table 1. Expression of Clostridium xylA gene by recombinant plasmids in E. coli
Specific activity of glucose isomerase were determined as described under Methods
and Materials from cell extracts which contained a recombinant plasmid with different size
of insert DNA (see, Figure 2). Cultures of E. coli HB101, containing recombinant plasmids
described in Fig. 2, were grown in LB media containing no inducers, 1% D-xylose, 0.2 mM
IPTG, or 1% xylose and 0.2 mM IPTG, at 37°C in the presence of ampicillin(100 ug/ml).
Cells were harvested at late exponential growth phase, washed with 50 mM MOPS buffer,
resuspended in the same buffer and disrupted by sonication. Cell debris was removed by
centrifugation at 12,000 x g and the extracts were used for activity determination as

described in Materials and Methods.

 

Glucose isomerase activity (unit/mg protein)

 

 

 

Plasmids Inducer
None D-xylose IPTG D-xylose + IPTG
pMMB67EH 0.00 0.00 0.00 0.00

pCMG11-3 0.29 0.26 0.56 0.55

pCMG11-9 0.00 0.00 0.12 0.10

 

153

Figure 4. Nucleotide and deduced amino acid sequence of C. thermosulfurogenes
xylA gene.

The numbers on the left refer to the”. nucleotide sequence starting at the first
base of the insert DNA of pCMG11-3. The deduced amino acid sequence is shown
below the nucleotide strand. The ATG initiation codon of the correct open reading
frame was identified by comparison of the deduced amino acid sequence with the
seven NHz-terminal amino acid sequence (underlined) determined from the purified
C. thermosulfurogenes xylose isomerase. The putative -35 and -10 sequences in the
promoter region and the putative ribosome-binding site (RBS) are underlined. The
first base of the insert DNA of pCMG11-9 begins at HindIII site (AGCI'I‘)
immediately downstream from the ribosome-binding site. The restriction sites used
in this study are shown above the nucleotide sequence. The ClaI site in parenthesis
could be detected only in a methylation deficient (dam) E. coli strain (24) (see

Methods and Materials).

154

1 . . . . . .
CGGATTTTTTAAATTTGTGTAGAATATATAATATATAATGTTTGTTGGACAGACAAACGA

 

-35 ~10
61 . HindIII . . . . .
ATAGAAGGAGGAAGCTTTATGAATAAATATTTTGAGAACGTATCTAAAATAAAATATGAA
R B s MetAsnLysm-PhecluAanaISerLysIleLyalyrGlu
1 5
121 . ClaI .

GGACCAAAATCAAACAATCCTTATTCTTTTAAATTTTACAATCCTGAGGAAGTAATCGAT
GlyProLysSerAsnAsnProTyrSerPheLysPheTyrAsnProGluGluValI1eA3p
25

181 . . . . . .
GGTAAGACGATGGAGGAGCATCTTCGCTTTTCTATAGCTTATTGGCACACTTTTACTGCT
GlyLysThrMetGluGluHisLeuArgPheSerIleAlaTerrpHisThrPheThrAla

45 .-

241 . . . HaeIII . .
GATGGAACAGATCAATTTGGCAAAGCTACCATGCAAAGGCCATGGAATCACTATACAGAT
AspGIyThrAspGlnPheGlyLysAlaThrMetGlnArgProTrpAanisTyrrhrAsp

65

301 . . . . . .
CCTATGGACATAGCTAAAGCAAGGGTAGAGGCAGCATTTGAGTTTTTTGATAAGATAAAT
ProMetAspIleAlaLysAlaArgValGluAlaAlaPheG1uPhePheAspLysIleAsn

85

361 . . . . . .
GCACCGTATTTCTGCTTCCATGATAGAGATATTGCCCCTGAAGGAGACACTCTTAGAGAG
AlaProTyrPheCysPheHisAspArgAspIleAlaProGluGlyAspThrLeuArgGlu

105

421 . . . . . .

ACGAACAAAAATTTAGATACAATAGTTGCTATGATAAAGGATTACTTGAAGACCAGCAAG

ThrAsnLysAanLeuAapThrIleValAlaMetIleLysAspTyrLeuLysTtrSerLys
125

481 . . . . . .
ACGAAAGTTTTGTGGGGTACTGCGAATCTTTTCTCCAATCCAAGATTTGTGCATGGTGCA
ThrLysValLeuTrpGlyThrAlaAsnLeuPheSerAsnProArgPheValfiisGlyAla

145

541 . . . Pat! . . .
TCAACGTCTTGCAATGCCGATGTTTTCGCATATTCTGCAGCGCAAGTCAAAAAAGCACTT
SerThrSerCysAanAlaAspValPheAlaTyrSerAlaAlaGanalLyaLysAlaLeu

165

601 . . . . . .
GAGATTACTAAGGAGCTTGGTGGCGAAAACTACGTATTCTGGGGTGGAAGAGAAGGATAT
GluIlerhrLysGluLeuGIyGlyGIuAsnTyrValpherrpGlyGlyArgGluGlyTyr

185

661 . . . . . .
GAGACACTTCTCAATACAGATATGGAGTTTGAGCTTGATAATTTTGCAAGATTTTTGCAC
GluThrLeuLeuAsnThrAspMetGluPheGluLeuAspA:nPheAlaArQPheLeuHi3

205

155

721 . . . HaeIII . .
ATGGCTGTTGATTATGCAAAGGAAATCGGCTTTGAAGGCCAGTTCTTGATTGAGCCGAAG
MetAlaValAspTyrAlaLysGluIleclyPheGluGIyGInPheLeuIlecluProLys

225

781 . . . . . .
CCAAAGGAGCCTACAAAGCATCAATACGACTTTGACGTGGCAAATGTATTGGCATTCTTG
ProLysGluProThrLysHiaGlnTyrAspPheAspValAlaAanalLeuAlaPheLeu

245

841 . . . . . .
AGAAAATACGATCTTGACAAATATTTCAAAGTTAATATCGAAGCAAATCATGCAACATTA
ArgLyaTyrAspLeuAspLysTerheLyaValAsnIleGluAlaAanisAlaThrLeu

265

901 . . . . . (ClaI)
GCATTCCATGATTTCCAGCATGAGCTAAGATACGCCAGAATAAACGGTGTATTAGGATCG
AlaPheHisAspPheGlnHisGluLeuArgTyrAlaArgIleAsnGlyValLeuGlySer

285

961 . . . . . .
ATTGACGCAAATACGGGTGATATGCTATTAGGATGGGATACAGATCAGTTCCCTACAGAT
I1eA3pA1aAsnThrGlyAspMetLeuLeuGlyTrpAspThrAspGInPheProThrAsp

305

1021 . . . . . .
ATACGCATGACAACACTTGCTATGTATGAAGTTATAAAGATGGGTGGATTTGACAAAGGC
IleArgMetThrThrLeuAlaMetTyrGluValI1eLysxetGlyGlyPheAspLysGly

325

1081 . . . . 139111 .
GGACTCAACTTCGATGCGAAAGTAAGACGTGCTTCATTTGAGCCAGAAGATCTTTTCTTG
GlyLeuAsnPheAspAlaLysVa1ArgArgA1aSerPhe61uProGluAspLeuPheLeu

345

1141 . . . . . HindIII .
GGTCACATAGCAGGAATGGATGCTTTTGCAAAAGGCTTCAAAGTGGCTTACAAGCTTGTA
GlyHisIleAlaGlyHetAspAlaPheAlaLysGlyPheLy3ValAlaTerysLeuVal

365

1201 . . . . . .
AAAGATAGGGTATTTGACAAGTTCATCGAAGAAAGATATGCAAGCTACAAAGATGGCATA
LysAspArgVa1PheA3pLysPheIleGluGluArgTyrAlaSerTerysAapGlyIle

385

1261 . . . . . .
GGTGCAGATATTGTAAGTGGGAAAGCTGATTTTAGAAGTCTTGAAAAGTATGCATTAGAG
GlyAlaAspIIeVaISerGlyLysAlaAapPheArgSerLeuGluLysTyrAlaLeuGlu

405

1321 . HincII . . . . .
CGCAGCCAGATTGTCAACAAATCAGGAAGACAAGAGCTATTAGAGTCAATCCTAAATCAG
ArgSerGlnI1eVa1AsnLyaSerGlyArgGlnGluLeuLeuGluSerIleLeuAanGln

425

1381 . . . . .
TATTTGTTTGCAGAATAATGAAACATGAGGGCAGCTTCATGCTTCATTAAAGCG
TereuPheAIaGlu"""

439

156

isomerase purified from C. thermosulfurogenes, to the amino acid sequence deduced
from the nucleotide sequence presented in Figure 3. Thus, the predicted polypeptide
encoded by this reading frame contains 439 amino acids and has a calculated
molecular weight of 50,474, a value that is in close agreement with the subunit
molecular weight estimated by SDS-polyacrylamide gel electrophoresis on xylose
isomerase purified from the original host. The ATG initiation codon is preceded,
with a spacing of 7 base pairs, by a sequence AAGGAGG which is complementary
to the 3’ end of both the E. coli and B. subtilis 16S ribosomal RNA (34,35). This
sequence presumably acts as ribosome-binding site during the expression of the C.
thermosulfurogenes xylA gene in these organisms. Upstream of the ribosome binding
site a sequence CGGA'I'I‘, at nucleotides 1-6 and TATAATATATAAT, at nucleotides
27-39, are homologous to -35 and -10 region, respectively, of promoters found in E.
coli and B. subtilis (36,37).

It has been reported that xylA gene of E. coli could not be expressed from its
own promoter in B. subtilis cells unless a promoter from B. licheniformis was inserted
upstream of the structural gene (38). In the case of the C. thermosulfurogenes xylA
gene the region of the first 81 base pairs, shown in Figure 4, was sufficient for high '
level expression in both E. coli and B. subtilis (Table 1 and unpublished data). It is
not clear at present whether the same sequences are functional as promoters and
ribosome binding sites in each of three organisms (C. thermosulfurogenes, E. coli, and
B. subtilis) or whether different sequences present in the upstream region function in

different hosts.

157

Comparison of amino acid sequences in different xylose isomerases

Comparison of the amino acid sequence, deduced from the nucleotide sequence
of C. thermosulfurogenes xylA gene, to amino acid sequences of xylose isomerases
from E. coli (13), B. subtilis (14), S. violaceoniger (16), Ampullariella sp. 3876 (15)
and Arthrobacter strain 3728 (21) are presented in Figure 5. The length of the
polypeptide chain encoded by the C. thermosulfurogenes xylA gene (439 amino acid
residues) is similar to those of E. coli and B. subtilis enzymes (440 residues), whereas
the enzymes from S. violaceoniger, Ampullariella, and Arthrobacter have polypeptides
shorter by 47 amino acid residues at the N-terrninal end. The deduced amino acid
sequence of thermostable C. thermosulfurogenes xylose isomerase exhibits a higher
degree of homology to the sequences of thermolabile xylose isomerases from E. coli
and B. subtilis (SO-70%) than to those of thermostable enzymes from S. violaceoniger,
Ampullariella and Arthrobacter (22-24%; see Figure 6). On the other hand, the
sequences of thermostable xylose isomerases of the mesophilic microorganisms, S.
violaceoniger, Ampullariella, and Arthrobacter, display a high degree of homology
among themselves (65-68%; Figure 5 and 6). These comparisons indicate that there
are three distinct classes of xylose isomerases, thermolabile of the E. coli and B.
subtilis type, thermostable of the C. thermosulfuricum type, and thermostable of the
S. violaceoniger type.

To understand the molecular mechanism of thermostability in these enzymes,
the possibility that disulfide bond formation and composition of hydrOphobic regions
within each protein molecule could contribute to the compactness of the secondary

and tertiary structure was considered. The sequence of C. thermosulfurogenes xylose

158

Figure 5. Comparison of the amino acid sequences of different xylose isomerases.

The abbreviations are: E.c., Escherichia coli (13); B.S., Bacillus subtilis (14);
Ct, Clostridium thermosulfurogenes (this study); S.v., Streptomyces violaceoniger
(16); Amp., Ampullariella sp. strain 3876 (15); Art., Arthrobacter strain 83728 (21).
The sequences of thermolabile enzymes are placed above, and the sequences of
thermostable enzymes below that of C. thermosulfurogenes enzyme. The sequences
are aligned to give a maximum homology by using the computer program, University
of Wisconsin-Genetics Computer Group, version 5 (30). Dots in the amino acid
sequences represent the gaps introduced to obtain the best fit. Identical residues
between C. thermosulfurogenes and other species amino acid sequences are denoted
by asterisks, and highly conserved regions in all xylose isomerases are boxed. The
arrow indicates the position of the His101 residue in C. thermosulfurogenes xylose

isomerase.

(5 CD!“
n O

H'
e

S.v.
Amp.
Art.

E.c.
8.5.

C.t.
S.v.

Amp.
Art.

E.c.
B.S.

C.t.
S.v.

Amp.
Art.

 

   

159

....... HQAYFDQLDRVRYEGSKSSNPLAFRHYNPDELVLGKRMEEHLRFAACYHHIFCHNGA GAFNRP. HQQPGEALALAKRKADVAFEFFHK
HAQSHSSSVNYFGSVNKVVI[GKASINPLAFKYYNPQEVIGGKIMKIHLRFSIAYHHIFIADGID IHORP. HDHYKG. MOLARARVEAAFEMFEK

l. O 1. 1.. l. IOIQIOO IO... 0 0 if. Q Q

....... HNKYFLNVSKIKYEGPKSNNPYSFKFYNPLEVIOGKIHEEHLRFSIAYHHIFIADGI THORP.HNNYIDPHDIAKARVEAAFEFFDK
O O Q

...................................................... MSFQPIPEDKF HIVGHQGRDPFGDAIRPALO.PV[IVQRLAE
...................................................... HSLQATPODKF HTVGHQAROAIGDAIRPVLO.PVEAVHKLAE
...................................................... HSVQPIPADHF HTVGHIGADPFGVAIRKNLD.PVEAVHKLAE

 

 

LHGTANCFTNPRYGAG INPDPEVFSHAAIOVVIAHEATHKLGGENYVLHGGRE
LHNTANMFINPRFVHGAATSCKADVFAYAAAOVKKGLEIAKELGAENYVFHGGRE

0. fit. it... it .0 Q 0. Q. Q... 0....

LHGJANLFSNPRFVHGASISCNADVFAYSAAOVKKALEITKEWGGENYVFHGGRE
O .

IGASLKEYINNFAQMVDVLAGKQEESGV
EGSILKEINONLOIIVGMIKDYMRDSNV

it it. t O

{GDILREINKNLOIIVAHIKDYLKISKI
.

  
     
  

e Me n o .1 e 10.0.1

PMAIINLFTHPVFKDGSFTANSRSVRRYALRKIIRNIDLAAEL KTYVAHGGRd
PMVITNLFIHPVFKD SFTSNDRSVRRYAIRKVLROMDLGAELGAKTLVLHGGRE
PHVITNLFSHPVFKD GFISNORSIRRFALAKVLHNIDLAAE GAETFVMHGGR

fGSSDTER....ESHIKRFRQALDAIGM
FGADAAIR....DGIVAGFSKALDEIGL

 

 

 

 

 

FDAIEAER....EKILGDFNQALKDTGL

 

 

 

ILAGHSFHHEIATAIALGLF
IJHQYOIDAAIIIAFLKQYGLDNHFK LEAN TLAGHTFEHELRMARVHGLL

F iii. if. t. 0 it. i 9 it t

TKHQYDFDVANVLAFLRXYDLDKYF NIEAN ILAfflOFQHELRYARINGVL
‘0 0 Q I *

LLNJOLRQEREQLGRFHONVVLHKHKIGFQGILL
LLNIDLKFELDNLARFMHMAVDYAKEIEYTGQFL

fitttt I it i O t O

LLNIDHEEELONEARFLHHAVOYAKEIGFEGQFL
9 ea 0 e

  
  
   

RGDILLPTVGHALAFIERLERPELYG NPEVG
R60ILLPTAGHAIAFVQELERPELFG NPETG
RGDIFLPTVGHGLAFIEOLEHGOIVG PEIG

QMAGLNFPHGIAQALHAGKL
QHSNLNFTOGIAQALHHKKL
OHAGLNFIHGIAOALHAEKL

GGAKDVROALORMKEAFDLLGEYVIAQGYOLRFA
OSAKDVGAALORYREALNLLAOYSEDQGYGLPFA
DGSKDLAAALDRMREGVOIAAGYIKDKGYNLRIA

 

 

  
  
 

= G ..... DAQLGNDIOQFPNSVEENALVHYEILKAGGFTT LNFDAKVR STDKYDLFYGHIGAHDTHALALKIAARHIEUGELDKRIA..
G ..... HPLLGHDTDEFPIDLYSTILAHYEILQNGGLGSGGLNFDAKVR SFEPDDLVYAHIAGHDAFARGLKVAHKLIEORVFEDVIQ...

t iii... .0 i Ct. A. Ifiiiiilllti t .0 it 0. t t t

G ..... DHLLGHUIDQFPIDIRMTTLAHYEVIKHGGFDKGGLNFDAKVR SFEPEDLFLGHIAGHDAFAKGFKVAYKLVKDRVFOKFIE...
I it t
QSGIKYDQOLRFGAGDLRAAFHLVDLLESA ..... GYE.GPRHFOFKPP E.DFDGVHASAEGCHRNYLILKERAAAFRADPEVQEALRAAR
- QHGPKFDQDLVFGHGOLLNAFSLVDLLENGPOGGPAYO.GPRHFDYKPS E.DFDGVHISAKDNIRHYLLLKERAKAFRADPEVQAALAESK
ORGIKYDQOLVFGHGDLISAFFIVDLLENGFPNGGPKY PRHFDYKPS 0.GYDGVHDSAKANHSHYLLLKERALAFRADPECQEAMKISG

 

.QRYSGHNSELGQQILKGOMSLADLAKYAQEHHLSPVHQSGRQEQLENLVNHYLFOK 440
.HRYRSFJEGIGLEIIEGRARFHILEQYALNNKIIK. NESGROERLKPILNQ 440

it 0 t t t. it... i .

.ERYASYKDGIGADIVSGKADFRSLEKYALERSQI.VNKSGRQELLESILNQYLFAE 439
O 0 O

LDOLAQPT..AADGLEALLADRIAIIOFDVEAA.AARAAHPFERLOQLAHDHLLGARG 389
VDLLRJPTLNPGLIYAOLLADRSAFLDYDADAVGAKG..YGFVKLNQLAIDHLLGAR 394
VFELGEIILNAGESAADLMNDSASFAGFDAEAA.AIRNFA.fIRLNQLAIEHLLGSR 395

92
98

92
45
45

192
198

192
141

I41
141

292
298

292
241

241
241

384
390

384
334

339
340

160

Figure 6. Summary of amino acid sequence homology between different xylose
isomerases.

The percent of homology was calculated for the number of amino acid residues
that match exactly between amino acid sequences from different sources by using the

University of Wisconsin-Genetics Computer Group, version 5 program (30).

161

 

B. subtilis

 

 

 

 

_ 50%

 

l

Thermolabile 70%
enzymes

E. coli

 

 

 

50%

Thermostable
enzymes

 

 

 

24°/o

 

Arthrobacter

22%

 

 

6 5%

/

 

 

S. violaceoniger

 

6 6 °/o

\

 

>-— 68% —-J

 

 

 

A mpullariella

 

 

162

isomerase contains two cysteine residues, at positions of 99 and 158, whereas the
thermolabile enzyme of B. subtilis type contains only one cysteine residue at amino
acid position 158 of the enzyme (Figure 5). The Cys,9 of C. thermosulfurogenes
enzyme was, therefore, changed, by site-directed mutagenesis, to alanine which is
present in at the corresponding site in the B. subtilis enzyme. This substitution,
however, did not affect either the thermostability or specific activity of the C.
thermosulfurogenes enzyme. Moreover, treatment with a reducing agent (25 mM
dithiothreitol) did not affect the specific activity of the wild-type enzyme (Table 2,
and results not shown). If we consider, in addition, the fact that thermoresistant
enzymes from ' S. violaceoniger and Arthrobacter each contain only one cysteine
residue and the Ampullariella enzyme contains no cysteine we have to conclude that,
in these xylose isomerase molecules, disulfide bonds do not contribute significantly
to the compactness of the molecule which is generally considered to be a requirement
for protein thermostability.

Another feature believed to contribute significantly to thermostability of
proteins is the number and distribution of hydrophobic regions in the molecule (40).
We have, therefore, compared the predicted hydrophobicity profiles of the Clostridium,
B. subtilis and S. violaceoniger enzymes using the method of Kyte and Doolittle (39).
The results, presented in Figure 7, indicated that the thermostable enzyme from S.
violaceoniger contained fewer hydrophobic regions than did the thermostable enzyme
from Clostridium, which in turn, exhibited a very similar hydrophobicity profile to
that of the thermolabile enzyme from B. subtilis. These results argue against the

general idea that enzyme thermostability relates to the compactness of the native

163

Figure 7. Comparison of hydropathy profiles of xylose isomerases from C.
thermosulfurogenes (C.t.), B. subtilis (B.S.), and S. violaceoniger (S.v.).

The deduced amino acid sequences were analyzed by the method of Kyte and
Doolittle (39) with a computer program MSEQ, University of Michigan (42) using a
window size of 7-residues. The hydrophobic regions have a positive hydropathy
index score and appear above the central dotted line. Amino acid numbers refer to the
sequence of C. thermosulfurogenes enzyme. Asterisks indicate the regions where the
hydropathy profile is significantly different between C. thermosulfurogenes and S.

violaceoniger enzymes.

164

$253: Eon oEE<

ovv oov omn oon omu cow om? cop om

 

4

9?
\, ., 22,. ... s I. _
lflfibflbk§§$WWfi7>Pg>%.J/Ipf:w>/\\52<RI :75)k ?

 

 

9o
mm
C

 

 

 

 

. . . .---. t t .. t . t -.
. ..- - ... .. ._
- e.* «t w .- - to «t .06

 

xepu! AqiedorpAH

165

molecule which can be predicted alone from the number of disulfide and hydrophobic
bonds that may be formed by the polypeptide chain. It is concluded here that
regions actually involved in the formation and maintenance of the three dimensional
structure are the key factors that determine the general resistance of protein molecule
to thermal changes. In support of this conclusion, the site-directed mutant Phe,,2
enzyme, described below, was less thermal stable than the wild-type His,,2 enzyme
by 20°C (data not shown). Xylose isomerases that share extensive homology, but
differ widely in thermostability, such as that of B. subtilis and C. thermosulfurogenes
seem to be excellent models for further investigation of this problem.

Despite the existence of profound differences in the amino acid sequences
between the individual classes of xylose isomerase, outlined in Figures 5 and 6,
several regions are highly conserved throughout all of the six xylose isomerase
polypeptides (Figure 5). Amino acid residues suggested to be parts of the catalytic
site, metal ion binding site and substrate binding site, by X-ray crystallographic
studies reported (20) on the S. rubiginosus xylose isomerase are found as the most

strictly conserved residues in these regions.

Modification of the xylose isomerase catalytic site

A complete amino acid sequence of S. rubiginosus xylose isomerase has not
been published. However, a recent X-ray crystallographic study (20) of the enzyme
proposed that His“ is an incipient proton abstracting residue acting as the general base
during the isomerization reaction. Comparison of amino acid sequences in Figure

5 indicated that His101 of C. thermosulfurogenes xylose isomerase is part of a highly

166

conserved region and it corresponds by homology analysis to His,4 of the enzyme
from S. violaceoniger (16), an organism closely related to S. rubiginosus. To provide
biochemical evidence for the involvement of histidine in the mechanism of catalytic
isomerization of xylose and glucose we have changed several histidine residues in the
C. thermosulfurogenes enzyme to phenylalanine by oligonucleotide-directed
mutagenesis of the cloned C. thermosulfurogenes xylA gene. Phenylalanine was
chosen because it has a side chain comparable in size to histidine, but can not
function as a base.

Mutant enzymes were expressed well in E. coli HB101 cells as indicated by
the results of SDS-polyacrylamide gel electrophoresis of the cell lysates (Figure 8A).
They could also be purified to near-homogeneity by the same procedures used for the
wild-type enzyme (Figure 8B). Substitution of histidine residues at positions 41, 71
or 152 of the polypeptide chain did not affect enzymatic activity (measured as glucose
isomerase) of the protein, whereas substitution of His101 completely destroyed both
xylose and glucose isomerase activity of the xylA product (Table 2 and 4).

To confirm the role of His1m as the hydrogen abstracting residue in the active
site of the enzyme, two further experiments were performed. In the first
experiment, the Hism, residue was substituted by glutamine for several reasons. First,
the resonance form of the oxygen atom in the carbonyl group of glutamine side chain
might have some nucleophilic properties and abstract a hydrogen from C, atom of
xylose or glucose. Second, the distance between the carbonyl group of the side
chain to the appropriate atoms of the substrate may be similar to that of the imidazole

of histidine. The results showed that Gln101 mutant enzyme was active as xylose and

167

Figure 8. SDS-polyacrylamide gel electrophoresis comparison of mutant xylose
isomerases expressed by different alleles of C. thermosulfurogenes xylA gene.
(A) Total protein (approximately 100-200 rig/lane) of crude cell lysates of E.
coli HB101 carrying chimeric plasmids expressing different mutant xylA genes and
(B) purified mutant xylose isomerases (approximately 4-10 ug/lane), isolated from the
E. coli HB101 as described in Materials and Methods, analyzed by 12% SDS-PAGE
(43) at constant current of 25 mA using a 16 x 20 cm Protein H Multi-Cell (Bio-
Rad) and visualized by Coomassie blue staining. Lanes : l, His—Who“; 2, His—)Phem;
3, His—iPhem; 4, His—)Phem; 5, His—>Glnm; wt, wild-type; Ctr, E.coli HB101 carrying
the vector plasmid without inserts; Mr, molecular weight standards (97,400,
phosphorylase; 66,200, bovine serum albumin; 42,700, ovalbumin; 31,000, carbonic
anhydrase; 21,500, soybean trypsin inhibitor). The arrow indicates the polypeptide

band corresponding to the wild-type xylose isomerase.

(B)

(A)

5 Mr wt

4

3

5Mr

2

wt Ctr 1

97.000 —

168

\ 112,700-—
— 51 000\
- 21.500 \_

/

”ill t- at

"Ill'l' .1 ‘
Jame:
HUI-ea: I
; WWII: 4
118.18! 4

 

169

Table 2. Comparison of site-directed mutant glucose isomerase activities in E. coli
carrying different alleles of C. thermosulfurogenes xylA gene.
Preparation of cell extracts, enzyme purification, and glucose isomerase assay for each

mutant and wild-type enzyme was performed as described in Materials and Methods.

 

Specific Activity (unit/mg)'

 

Enzyme Cell Extract Purified Enzyme
Wild-type 0.32 (1.00) 3.7 (1.00)
His - Phe41 0.33 (1.03) 3.6 (0.97)
His - Phe“ 0.36 (1.12) 4.0 (1.08)
His - Phe,,2 0.31 (0.97) 3.2 (0.86)
His - Phew, 0.00 (0.00) ' 0.0 (0.00)
His - Glnw, 0.03 (0.09) 0.4 (0.11)
Cys - Ala,9 0.34 (1.06) 3.7 (1.00)

 

‘ Fraction of specific activity relative to wild-type is given in parenthesis.

170
glucose isomerase, although only a fraction of specific activity of the wild-type
enzyme was retained (Table 2 and 4).

In the second experiment, the effect of diethylpyrocarbonate, known to inhibit
enzymes containing histidine as a catalytic residue by covalent modification of the
imidazole moiety (19), on specific activities of wild-type and mutant enzymes was
determined. As shown in Table 3, the enzymes containing histidine at the position
101 were strongly inhibited by 1 mM DEPC, whereas the Gln1m mutant enzyme was
not inhibited even by 10 mM DEPC. The results of these experiments indicate that
His1m is essential for the catalytic activity of C. thermosulfurogenes xylose isomerase.
Moreover, they are consistent with the proposed catalytic mechanism of the
isomerization reaction that involves a general base catalysis mediated by imidazole
moiety of a specific histidine of xylose isomerase (Hisml in this enzyme).

Kinetic properties of the wild-type and mutant enzymes are compared in Table
4. The apparent K. values for glucose did not change significantly upon substitution
of His1m by glutamine. This indicated that this histidine residue is not essential for
substrate binding. The apparent V“, of the Gln101 mutant enzyme was changed to
a different extent for xylose, glucose, and fructose. However, the Phe“n mutant '
enzyme showed no activity with either of these substrates (Table 4). We conclude,
therefore, that the same amino acid residue (Hism) is involved in the catalytic reaction
for each of these three substrates. It should be noted that the ratio of apparent
maximal velocities for the fructose to glucose conversion was 2.4-fold higher than that
obtained with the wild-type enzyme. It is possible that the position of the functional

group of glutamine, in the folded molecule of the enzyme, is closer to the Cl atom

171

Table 3. Effect of diethylpyrocarbonate (DEPC) on glucose isomerase activity of the
wild-type and site-directed mutant enzymes.

Purified enzymes (80 11g for wild-type and His-Phe” mutant enzyme and 800 11g for
His-Gln101 mutant enzyme) in 50 mM sodium phosphate buffer (pH 7.0) were incubated
with DEPC at room temperature for 30 min. The residual glucose isomerase activity was
assayed as described in Materials and Methods and expressed as percentage of specific

activity found in the control without DEPC.

 

Residual Activity (%)

 

Enzyme 0 mM DEPC 1 mM DEPC 10 mM DEPC
Wild-type 100 8 1
His - Phe“ 100 10 3

His - Gln1m 100 100 100

 

172

Table 4. Comparison of kinetic properties of site-directed mutant and wild-type xylose
isomerases

Enzyme activities with each substrate were determined with purified enzymes at various
substrate concentrations as described under Materials and Methods. Apparent Km and V“

values were obtained from Lineweaver-Burk plot.

 

 

K,I (mM) Vmax (unigmg) Ratio of Vm
Enzyme Glucose Xylose Glucose Fructose (Frc/Glc)
Wild-type 142 14.0 5.3 2.9 0.55
His - Phen 130 14.9 5.5 3.3 0.60
His - Phe”.2 152 16.2 5.2 2.9 0.56
His - Pheml n.d. 0.0 0.0 0.0 n.d.
His - Gln101 138 1.6 0.6 0.8 1.33

 

n.d. Not Deterrninable

173

of the substrate than the imidazole residue of the Hism. This assumption may be
verified when the information on three-dimensional structure of the Gln1m enzyme
becomes available.

Dependence of kinetic parameters on pH for glucose isomerization of the wild-
type and Gln101 mutant enzymes were determined over the pH range of 5.0 to 10.0.
The plot of log Vmax“, versus pH indicated that the apparent pKa values of the
catalytic residues in the wild-type enzyme were approximately 6.0 and 9.4, whereas
the Gln,“ mutant enzyme displayed only one detectible pKa value at 9.4 (Figure 9).

Although the apparent pKa value of an amino acid residue in an enzyme may be
affected by its micro-environment (e.g. neighboring amino acid residues or ions), the
estimated pKa value of 6.0 may represent the apparent pKa of the imidazole moiety
of the His1m of the wild-type enzyme. Therefore, the observed changes of the enzyme
activity at pH below 7.0 may be due to the different status of protonation of the
imidazole moiety in the Hism, which would affect the nucleophilicity of this
functional group. On the other hand, the amino and carbonyl groups of the
glutamine side chain are not protonable over the physiological pH range, and the
potential nucleophilic property of these groups is not expected to be altered by the
change of pH from 8.0 to 5.0.

We have not yet definitely ruled out the possibility that the changes in
enzymatic activity, reported in this paper, might be due to structural changes in the
catalytic site, resulting from the introduction of strongly hydrophobic residue,
phenylalanine, in place of Hislol and that glutamine simply played a role of a more

hydrophilic substitution than phenylalanine. However, based on the existing

174

, Figure 9. Plot of relative log of apparent Vm versus pH for Glnm mutant
(H) and wild type (O—-0) xylose isomerases.

Apparent V“, values at different pH were determined from Lineweaver-Burk
plots. The scale of relative log Vmax“. indicates the fraction of each experimental

value at different pH relative to the maximal value.

175

 

 

I

10.0

8.0

6.0

5.0

 

 

1
O.

‘0.
O

‘ddOXDulA 50'] engtolaa

0.
O

176
crystallographic data and on results on comparing the amino acid sequence and
properties of sixmutant xylose isomerases including their inhibition by DEPC and the
change of activity at different pH values, we favor the interpretation forwarded in this
paper that Hism, is the amino acid that acts as a general base catalysts in the active
site of the C. thermosulfurogenes xylose isomerase. Further studies are required to
understand the mechanism of isomerase activity at acidic pH values including
additional site specific amino acid changes and three-dimensional structure analysis

of the modified molecules.

10.

11.

12.

13.

14.

15.

REFERENCES

Takasaki, Y., Kosugi, Y., and Kanbayashi, A. (1969) Arg. Biol. Chem.
33, 1527-1534

Antrim, R.L., Colliala, W., and Schnyder, B. (1979) in Applied
Biochemistry and Bioengineering (Wingard, L.B., ed) pp. 97-155,
Academic Press, New York

Bucke, C. (1980) in Enzymes and Food Processing (Birch, G.G.,
Blackebrough, N., and Parker, J.K., ed) pp. 51-72, Applied Science
Publishers, London

Chen, W. (1980) Process Biochem. 15, 30-35

Chen, W. (1980) Process Biochem. 15, 36—41

Carrell, H.L., Rubin, B.H., Hurley, T.J., and Glusker, JP. (1984) J.
Biol. Chem. 259, 3230-3236

Farber, G.K., Petsko, G.A., and Ringe, D. (1987) Protein Engineering
1, 459-466

Rey, F., Jenkins, J., Janin, J., Lasters, I., Alard, P., Claessens, M.,
Matthyssens, G., Wodak, S. (1988) Proteins 4, 165-172

Dauter, Z., Dauter, M., Hemker, J., Witzel, H., and Wilson, K. (1989)
FEBS Letters 7, 1-8

Batt, C.A., Claps, M.C., Bodis, M.S., Jamas, S., and Sinskey, AJ.
(1985) Can. J. Microbiol. 31, 930-933

Wilhelm, M., and Hollenberg, GP. (1984) EMBO Journal 3, 2555-2560
Shamanna, D.K., and Sanderson, KB. (1979) J. Bacteriol. 139, 71-79
Schellenberg, G.D., Sarthy, A., Larson, A.E., Backer, M.P., Crabb, J .W.,
Lidstrom, M., Hall, B.D., and Furlong, CE. (1984) J. Biol. Chem. 259,
6826-6832

Wilhelm, M., and Hollenberg, GP. (1985) Nucleic Acids Res. 13, 5717-
5723

Saari, G.C., Kumar, A.A., Kawasaki, G.H., Insley, M.Y., and O’Hara,

177

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

178
R]. (1987) J. Bacteriol. 169, 612-618

Drocourt, D., Bejar, S., Calmels, T., Reynes, J.P., and Tiraby, G. (1988)
Nucleic Acids Res. 16, 9337

Rose, I.A., O’Connell, E.L., and Mortlock, RP. (1969) Biochem.
Biophys. Acta 178, 376-379

Suckling, KB. (1985) in Comprehensive Biotechnology (Moo-Young,
M., Bull, A., and Dalton, H. ed) vol.1, pp. 539-551, Pergamon Press,
Oxford

Gaikwad, S.M., More, M.W., Vartak, H.G., and Deshpande, V.V. (1988)
Biochem. Biophys. Res. Commun. 155, 270-277

Carrell, H.L., Glusker, J.P., Burger, V., Manfre, F., Tritsch, D., and
Biellmann, J.-F. (1989) Proc. Natl. Acad. Sci. USA 86, 4440-4444

Henrick, K., Collyer, C.A., and Blow D.M. (1989) J. Mol. Biol. 208,
129-157

Boyer, H.W., and Roulland-Dussoix, D. (1969) J. Mol. Biol. 41, 459-
472

Yanisch-Perron, C., Vieira, J., and Messing, J. (1985) Gene 33, 103-
119. Marinus, M.G., and Konrad, EB. (1976) Mol. Gen. Genet. 149,
273-277

Marinus, M.G., and Konrad, EB. (1976) Mol. Gen. Genet. 149, 273-
277

Fiirste, J.P., Pansegrau, W., Frank, R., Blocker, H., Scholz, P.,
Bagdasarian, M., and Lanka E. (1986) Gene 48, 119-131

Poncs, M., Solowiejczyk, D., Ballantine, M., Schwartz, E., and Surry,
S. (1982) Proc. Natl. Acad. Sci. USA 79 4298-4302

Birnboim, H.C., and Doly, J. (1979) Nucleic Acids Res. 7, 1513-1523
Maniatis, T., Fritsch, E.E., and Sambrook, J. (1982) in Molecular
cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold
Spring Harbor, N.Y.

Sanger, F., Nicklen, S., and Coulson, AR. (1977) Proc. Natl. Acad. Sci.
USA 74, 5463-5467

30.

31.

32.
33.

34.

35.

36.

37.

38.

39.
40.

41.

42.
43.

179

Devereux, I., Haeberli, P., and Smithies, O. (1984) Nucleic Acids Res.
12, 387-395

Sayers, J.R., Schmidt, W., and Eckstein, F., (1988) Nucleic Acids Res.
16, 791-802

Dishe, Z., and Borenfreund, E. (1951) J. Biol. Chem. 192, 583-587

Lowry, O.H., Rosebrough, N.J., Farr, A.I., and Randall, R.J. (1951) J.
Biol. Chem. 193, 265-275

Stormo, G.D., Schneider, T.D., and Gold, L. (1982) Nucleic Acids Res.
10, 2971-2996

McLaughlin, J.R., Murray, C.L., and Rabinowitz, J.C. (1981) J. Biol.
Chem. 256, 11283-11291

Harley, CB, and Reynolds, RP. (1987) Nucleic Acids Res. 15, 2343-
2361

Moran, C.P., Lang, N., LeGrice, S.F., Lee, G., Stephens, M.,
Sonenshein, A.L., Pero, J., and Losick, R. (1982) Mol. Gen. Genet. 186,
339-346

Huang, J.J., and Ho, W..Y (1985) Biochem. Biophys. Res. Commun.
126, 1154- 1160

Kyte, J., and Doolittle, R.F. (1982) J. Mol. Biol. 157, 105-132

Kitamoto, N., Yamagata, H., Kato, T., Tsukagoshi, N., and Udaka, S.
(1988) J. Bacteriol. 170, 5848-5854

Brady, G., Jantzen, H.M., Bernard, H.U., Brown, R., Hashimoto-Gotoh,
T., and Schlitz, G. (1984) Gene 27, 223-232

Black, S.D., and Glorioso, J.C. (1986) Biotechniques 4, 448-460

Laemmli, U.K. (1970) Nature (London) 227, 680-685

Chapter VI

CONCLUSIONS AND PERSPECTIVES

180

CONCLUSIONS

The studies described in this thesis on thermostable glucose isomerases of the
thermoanaerobes Thermoanaerobacter strain B6A and Clostridium thermosulfurogenes
broaden knowledge and information on the following topics: physicochemical
properties and regulation of glucose isomerase in thermoanaerobes; potential
applications of these thermostable glucose isomerases in novel sweetener production
processes; molecular and genetic features of the glucose isomerase gene from
thermoanaerobic bacteria; catalytic function of histidine in the active site during
glucose isomerization; and site directed alteration of kinetic and catalytic properties
of glucose isomerase by protein engineering.

Most investigations on the production and biochemical characterization of glucose
isomerase have concentrated on mesophilic, aerobic microorganisms. Glucose
isomerase is the most expensive commercial saccharidase used in sweetener production
from starch, and the current industrial process can be improved with an enzyme that
possess higher thermostability and higher activity at acidic pH. Therefore, the
present studies were initiated to advance fundamental understanding about the -
biochemical and physiological properties of glucose isomerase of thermoanaerobic
bacteria and the molecular mechanisms that account for enzymatic catalysis and
thermostability. ‘

Initially, two different types of glucose isomerase activities were identified in
diverse thermoanaerobic species based on differences in their apparent pH Optima.

Thermoanaerobacter and C. thermosulfurogenes which produced cell-bound,

181

182

thermostable glucose isomerases with neutral pH optima for activity were selected
as model microorganisms for further studies. Thermoanaerobacter constitutively
produced glucogenic amylase and B-galactosidase activities which were
environmentally compatible (i.e. active and stable at low pH and high temperature)
with glucose isomerase activity. These findings made it possible to design a new
sweetener production process that directly converted starch or lactose into a fructose
sweetener by using a saccharidase mixture produced simultaneously by
Thermoanaerobacter. This novel single step process represents a very efficient
sweetener production system because it can eliminate the need for multiple steps and
enzymes in the conventional starch process, and solve the digestion problems of
lactase deficient individuals by hydrolysis of lactose into sweetener when used in
dairy products. The utility of this single step process for industrial sweetener
production requires further investigation on enzyme immobilization, continuous
reaction system development, improvement of enzyme activity and stability at low
pH, and other studies on enzyme production cost and safety.

Biochemical properties of glucose isomerases purified from Thermoanaerobacter
and C. thermosulfurogenes were characterized in detail. The two distinct
thermoanaerobic bacterial species produced highly thermostable glucose isomerases
with close similarity in physicochemical and catalytic properties. These findings
suggest that glucose isomerases from species with a common phylogenic origin have
not diversified dramatically during their evolution under the similar conditions in
thermal hot spring ecosystems. At present, the molecular mechanism of

thermophilicity in their glucose isomerase is not clear and remains to be solved.

183

However, further studies on three dimensional structure of the enzyme and protein
modification via site directed mutagenesis may be able to answer the questions.
The gene coding for thermostable glucose isomerase of C. thermosulfurogenes
was cloned and overproduced in Escherichia coli and Bacillus subtilis using its
original promoter. It is worth noting that a hyper-expression vector system by using
the promoter of the C. thermosulfurogenes was deve10ped in mesophilic B. subtilis or
E. coli hosts. Production of thermostable glucose isomerase in a meSOphilic host
was of particular interest in simplifying large scale purification of the enzyme by a
simple heat treatment procedure. Thus, these findings suggest the potential of
utilizing the recombinant thermostable glucose isomerase over-expressed in a food
microorganism B. subtilis as an industrial enzyme for sweetener production.
Nucleotide and deduced amino acid sequence analysis of the cloned glucose
isomerase gene advanced insights on the primary structure of thermophilic glucose
isomerases. Comparison of amino acid composition of glucose isomerases and
substitution of several amino acid residues in the C. thermosulfurogenes enzyme
molecule indicated that the factors which determine protein thermostability are the
three dimensional coordination of key amino acids involved in the formation and ~
maintenance of the protein structure; and, not simply the presence of disulfide bonds
or the number of hydrophobic bonds which can be predicted from the primary
structure of the native molecule. Glucose isomerases that share extensive homology,
but differ significantly in thermostability, such as those from B. subtilis and C.
thermosulfurogenes seem to be excellent models for further investigation in

determining the molecular mechanisms of protein thermophilicity.

184

Moreover, the catalytic residue in the active site of C. thermosulfurogenes
glucose isomerase was identified by homology analysis with other glucose isomerase
amino acid sequences in relation to their three dimensional structure predicted from
X-ray crystallographic studies. Histidine was demonstrated to function in catalysis
by site directed mutagenesis, and by redesigning the catalytic site of the enzyme
molecule to contain glutamine in lieu of histidine, a thermostable glucose isomerase
with acid stable activity was obtained. Further studies including additional site
specific amino acid changes and three-dimensional structure analysis of the modified
molecules are required to understand the exact chemical mechanism of isomerase
activity at acidic pH values or to further improve enzyme substrate specificity and
activity.

In ending, the results on biochemical and molecular characterization of glucose
isomerase completed in this thesis will serve as a basis to start assessing the industrial
applications of both the native and site-specifically engineered glucose isomerases
developed here. In addition, the glucose isomerase gene and the engineered enzyme
production system can serve as a model system for future research aiming at
understanding the molecular mechanism of glucose isomerase catalysis and
thermostability, and also for providing a rationale basis to redesign the molecule for
property improvements by protein engineering. Hopefully, the present report on
cloning, crystallization, and determination of the primary structure of a thermostable
glucose isomerase will also be useful for the continuation of basic studies on

understanding enzyme therrnophilicity in relation to its three dimensional structure.