Production, purification, characterization and cloning of the cyclomaltodextrinase from Bacillus subtilis high temperatu...

MISSING IMAGE

Material Information

Title:
Production, purification, characterization and cloning of the cyclomaltodextrinase from Bacillus subtilis high temperature growth transformant H-17 comparison to the parent enzymes from Bacillus subtilis 25S and Bacillus caldolyticus C2
Physical Description:
xiv, 160 leaves : ill., photos ; 29 cm.
Language:
English
Creator:
Krohn, Bradley Martin, 1959-
Publication Date:

Subjects

Subjects / Keywords:
Bacillus subtilis   ( lcsh )
Molecular cloning   ( lcsh )
Thermophilic bacteria   ( lcsh )
Soil microbiology   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1991.
Bibliography:
Includes bibliographical references (leaves 146-159).
Statement of Responsibility:
by Bradley Martin Krohn.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001745460
notis - AJF8241
oclc - 26272930
System ID:
AA00003726:00001

Full Text












PRODUCTION, PURIFICATION, CHARACTERIZATION AND CLONING
OF THE CYCLOMALTODEXTRINASE FROM BACILLUS SUBTILIS
HIGH TEMPERATURE GROWTH TRANSFORMANT H-17:
COMPARISON TO THE PARENT ENZYMES FROM
BACILLUS SUBTILIS 25S AND BACILLUS CALDOLYTICUS C2










BY

BRADLEY MARTIN KROHN


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

UNIVERSITY OF FLORIDA

1991


UNIVERSITY OF FLORIDA LIBRARIES

























This dissertation is dedicated to my parents, Lou and Doc,
for their support, enthusiasm, patience, and understanding during
the course of my doctoral program at the University of Florida.
















ACKNOWLEDGEMENTS

I would like to thank Dr. James Lindsay, my major advisor, for his guidance,

support, motivation, patience, commitment, and never-ending enthusiasm for my

doctoral research and graduate program. I am extremely grateful to the other

members of my doctoral committee, Dr. Dennis Duggan, Dr. Lonnie Ingram, Dr.

Ronald Schmidt, Dr. Maurice Marshall, and Dr. Jesse Gregory, for their interest,

suggestions, review of manuscripts, and use of their equipment. I am deeply

appreciative to the Food Science and Human Nutrition Department for providing

me with a graduate assistantship for the duration of my research.

I wish to thank the students and postdocs in the lab of Dr. L.O. Ingram for

their assistance and interaction while I worked in their lab for three months.

Thanks go to Thom, Dirk, Morgan, Bee, Melissa and Larry, for making our lab an

enjoyable, fun, and efficient environment to work in. I am also grateful to Walter

Jones for his availability and professional photography, and to Edward Jaggers for

washing my glassware. Finally, my sincerest thanks go to my family, Carole, Frank,

Jeff, Frank, my grandparents, and aunts and uncles for all their love, advice, and

support.















TABLE OF CONTENTS

page

ACKNOWLEDGEMENTS. iii

LIST OF TABLES vi

LIST OF FIGURES viii

ABSTRACT xii

CHAPTERS

1 INTRODUCTION 1

2 REVIEW OF LITERATURE 7

Thermophiles and Thermophily 7
The Origin and Genetics of Thermophiles 10
Thermostability of Cell Components 14
Protein Engineering of Industrially
Important Enzymes. 29
Microbial Amylases in Starch Bioprocessing 36
Bacterial Cyclomaltodextrinases 39


3 PURIFICATION, CHARACTERIZATION,
AND COMPARISON OF THE CYCLO-
MALTODEXTRINASE FROM B. SUBTILIS
25S, B. CALDOLYTICUS C2, AND
B. SUBTILIS HIGH TEMPERATURE
GROWTH TRANSFORMANT H-17 45

Materials and Methods 46
Results and Discussion 53















4 SUBSTRATE SPECIFITIES, AFFINITIES, AND
CLEAVAGE PATTERNS OF THE CYCLO-
MALTODEXTRINASE FROM B. SUBTILIS
25S, B. CALDOLYTICUS C2, AND THE
B. SUBTILIS HIGH TEMPERATURE
GROWTH TRANSFORMANT H-17 74

Materials and Methods 74
Results and Discussion 77


5 CLONING OF THE CYCLO-
MALTODEXTRINASE GENE FROM
B. SUBTILIS HIGH TEMPERATURE
GROWTH TRANSFORMANT H-17 89

Materials and Methods 90
Results and Discussion 110


6 SUMMARY AND CONCLUSIONS 142


REFERENCES 146


BIOGRAPHICAL SKETCH 160















LIST OF TABLES


Table page


1 Purification of B. subtilis 25S cyclomaltodextrinase 54

2 Purification of B. caldolyticus C2 cyclomaltodextrinase 55

3 Purification of B. subtilis H-17 cyclomaltodextrinase 56

4 Biochemical and biophysical comparison of
B. subtilis 25S, B. caldolyticus C2, and
B. subtilis H-17 cyclomaltodextrinase 60

5 Cationic inhibition of B. subtilis 25S,
B. caldolyticus C2, and B. subtilis
H-17 cyclomaltodextrinase 64

6 Amino acid compositions of B. subtilis 25S,
B. caldolyticus C2, and B. subtilis
H-17 cyclomaltodextrinase 71

7 Experimental relative rates of hydrolysis of
malto-oligosaccharides, cyclodextrins, and
polysaccharides by B. subtilis 25S, B. caldolyticus
C2, and B. subtilis H-17 cyclomaltodextrinase 79















Table


page


8 Km and Vmax values of B. subtilis 25S and
B. subtilis H-17 cyclomaltodextrinase 83

9 Comparison of cyclomatodextrinase activity between
B. subtilis YB886 carrying pPL708 with no insert,
and B. subtilis YB886 carrying pPL708 with a
probe-positive 3 kb Eco RI fragment 139















LIST OF FIGURES


Figure page


1 SDS-PAGE of purified B. subtilis 25S
cyclomaltodextrinase on a 7.5% polyacrylamide gel 57

2 SDS-PAGE of purified B. caldolyticus C2
cyclomaltodextrinase on a 7.5% polyacrylamide gel 58

3 SDS-PAGE of purified B. subtilis H-17
cyclomaltodextrinase on a 7.5% polyacrylamide gel 59

4 Isoelectric focusing gel of purified B. subtilis
25S, B. caldolyticus C2, or B. subtilis
H-17 cyclomaltodextrinase 61

5 Effect of pH on the activity of (A) B. subtilis
25S cyclomaltodextrinase, (B) B. caldolyticus
C2 cyclomaltodextrinase, and (C) B. subtilis
H-17 cyclomaltodextrinase 63

6 Effect of temperature on the activity of
(A) B. subtilis 25S cyclomaltodextrinase,
(B) B. caldolyticus C2 cyclomaltodextrinase, and
(C) B. subtilis H-17 cyclomaltodextrinase. 66















Figure


7 (A) Effect of temperature and incubation time on the
thermostability of B. subtilis 25S cyclomaltodextrinase.
(B) Effect of temperature, incubation time, and 0.02%
2-Me or 0.01 mM EDTA on the thermostability of
B. caldolyticus C2 cyclomaltodextrinase. (C) Effect
of temperature, incubation time, and 0.02% 2-Me on
the thermostability of the B. subtilis H-17
cyclomaltodextrinase. (D) Effect of temperature,
incubation time, and 0.005 mM EDTA on the
thermostability of B. subtilis H-17 cyclomaltodextrinase 68

8 TLC of 10 min (A), 1 h (B), and 3 h (C) products
of linear malto-oligosaccharide hydrolysis by
B. subtilis 25S, B. caldolyticus C2, or
B. subtilis H-17 cyclomaltodextrinase 81

9 TLC of 20 min hydrolysis products of
CDs and 1.5 h hydrolysis products of
polysaccharides by B. subtilis 25S
or B. subtilis H-17 cyclomaltodextrinase 85

10 TLC of products of pullulan hydrolysis
by B. subtilis 25S and B. subtilis
H-17 cyclomaltodextrinase 85

11 Genetic map of E. coli plasmid cloning vector
pUC18 (top) and B. subtilis plasmid cloning
vector pPL708 (bottom) 96

12 General strategy for cloning Sau 3A fragments
into the Bam HI site of plasmid vector pUC18. 112

13 Starch-LA agar plate stained with iodine crystal vapors 113


page















Figure


14 Pullulan-Reactive Red-LA agar plate which had
been previously exposed to chloroform vapors 114

15 N-terminal amino acid sequence of the B. subtilis
H-17 thermostable cyclomaltodextrinase 116

16 Southern blot hybridization of DNA probe to complete
restriction enzyme digests of B. subtilis H-17 DNA 117

17 Southern blot hybridization of DNA probe to amplified
gene libraries constructed in pUC18 119

18 Hybridization of DNA probe to colony lift of E. coli
DH5a transformed with Pst I, Bam HI, or Eco RI library 120

19 Immunologically-screened colony lifts of E. coli DH5a
transformed with gene libraries constructed in pUC18. Top
membranes: E. coli transformed with amplified
Pst I library. Bottom membranes: E. coli transformed
with amplified Bam HI library 122

20 Agarose gel electrophoresis of digested plasmid DNA
from several immuno-positive E. coli DH5a transformants
identified on the corresponding colony lift 124

21 Southern blot analysis of Pst I digested plasmid
DNA from several immuno-positive E. coli DH5a
transformants carrying pUC18 with a
Pst I insert (top). Southern blot analysis of
Bam HI digested plasmid DNA from several
immuno-positive E. coli DH5a transformants
carrying pUC18 with a Bam HI insert (bottom) 126


page















Figure


22 Colony lift of subcultured immuno-positive
E. coli DH5a transformants. Colonies carrying
pUC18 with a Pst I or Barn HI insert show
loss of immunological signal 127

23 General strategy for cloning DNA probe-
positive Eco RI fragments into the Eco
RI site of plasmid vector pPL708. 132

24 Immunologically-screened colony lifts
of B. subtilis YB886 transformed with
Eco RI library constructed in pPL708 133

25 Southern blot hybridization of DNA probe to
Eco RI digested plasmid DNA from four
B. subtilis YB886 immuno-positive transformants 134

26 Southern blot hybridization of DNA probe to
Eco RI digest of cesium chloride density
gradient-purified pPL708 carrying the
probe-positive Eco RI insert 135

27 Immunologically-screened colony lifts of
E. coli XL1-Blue transformed with pUC18
carrying the DNA probe-positive
Eco RI fragment purified from pPL708 140

28 Ochterlony gel diffusion of purified
rabbit polyclonal antisera against
late log phase cell-free extracts 141


page















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


PRODUCTION, PURIFICATION, CHARACTERIZATION AND CLONING
OF THE CYCLOMALTODEXTRINASE FROM BACILLUS SUBTILIS
HIGH TEMPERATURE GROWTH TRANSFORMANT H-17:
COMPARISON TO THE PARENT ENZYMES FROM
BACILLUS SUBTILIS 25S AND BACILLUS CALDOLYTICUS C2


BY

BRADLEY MARTIN KROHN

December 1991



Chairman: James A. Lindsay
Major Department: Food Science and Human Nutrition

The p-nitrophenyl-a-D-maltoside hydrolyzing cyclomaltodextrinase (EC

3.2.1.54) from the mesophile Bacillus subtilis 25S, the obligate thermophile Bacillus

caldolyticus C2, and the Bacillus subtilis high-temperature growth transformant

H-17 were purified, characterized and compared. All three cyclomaltodextrinases

displayed maximal rates of hydrolysis and identical hydrolysis patterns for linear

malto-oligosaccharides and e- and 13- cyclodextrin, with maltose and glucose as the

final products. Starch, amylose, and amylopectin were degraded slowly to maltose,

xii









in an exo-fashion by preferential cleavage of maltose units from the nonreducing

ends. Neither enzyme showed activity against p-nitrophenyl-a-D-glucopyranoside,

maltose, isomaltose, isomaltotriose or panose. The enzymes demonstrated pullulan

hydrolase activity due to their hydrolysis of pullulan to glucose, maltose, and

(iso)panose.

The 25S, C2, and H-17 enzymes were composed of two identical subunits of

Mr 55,000, 60,000, and 55,000, respectively. The 25S, C2, and H-17 enzymes had

a pi of 4.85, pH optimum of 7.5, 7.0, and 7.5, and Km values for the chromogenic

substrate p-nitrophenyl-a-D-maltoside of 2.96 mM, 1.31 mM, and 1.46 mM,

respectively. The 25S enzyme exhibited optimal activity between 35-370C, and

complete inactivation after 10 min at 450C. This contrasts with the C2 enzyme

which showed optimal activity at 600C, and retained 100% of initial activity at 60C

for 2 h, and with the H-17 enzyme which showed optimal activity between 650C and

680C, and retained 100% of initial activity at 65oC for 1 h. Both the C2

and H-17 enzymes required 2-mercaptoethanol or EDTA for thermostability. A

comparison of the amino acid compositions showed an increase in proline, alanine,

and leucine residues for the C2 enzyme, and an increase in proline, alanine, and

glycine residues for the%4-17 enzyme.

The H-17 cyclomaltodextrinase gene was cloned on separate Pst 1, Bam H1,

and Eco R1 fragments in the plasmid vector pUC18, but was expressed in an

inactive form in the host, E. coli DH5a. High level constitutive expression of the

xiii









gene product was also detrimental to the E. coli host, which led to structural

instability of the recombinant plasmid. The cyclomaltodextrinase gene was then

cloned on a 3 kb Eco RI fragment in the plasmid vector pPL708, and the fragment

was structurally maintained in the host B. subtilis YB886. The cloned gene product

appeared to be in an enzymatically active form in the B. subtilis host; however,

expression was at a low level.















CHAPTER 1
INTRODUCTION


One aspect in the evolution of biotechnology is the use of enzymes to

replace chemical catalysts as the agents of chemical processes. The industrial

hydrolytic depolymerisation of starch, once achieved by an acid-catalyzed process,

is now achieved by starch-degrading enzymes, the amylases. Ironically, the first

enzymes to be recognized as specific biocatalysts were those which could hydrolyze

starch. Diastase, the active component of malt, was first isolated in France in 1833,

and later shown to be a combination of a- and B-amylase [24]. In 1894, a mixture

of amylases from Aspergillus oryzae was used by the Japanese as a digestive aid for

the consumption of rice starch [24]. Historically, acid-catalyzed hydrolysis of starch

to glucose was first conducted on an industrial scale in Germany, in 1850 [24]. It

was not until the late 1960s that the traditional acid-catalyzed production of glucose

was redesigned to an enzyme hydrolysis procedure on a commercial scale. This

change was prompted by the breakthrough in the development of amyloglucosidase,

which enabled starch to be completely degraded enzymatically to glucose. The

advantages to an enzyme hydrolysis were higher yield, higher purity, and easier

crystallization. The process was further developed in 1973 by introduction of a

thermostable a-amylase from Bacillus licheniformis. Starch liquefaction could now









2
be rapidly achieved at 95C. followed by saccharification with amyloglucosidase,

which enabled production of syrups containing up to 98% glucose. Today, the

precise end-product composition can be controlled with different amylases, to give

corn syrups having various levels of oligosaccharides, maltose, and glucose, with the

desired chemical and physical properties. Furthermore, high glucose corn syrup

may now be converted to the food system sweetener, high fructose corn syrup, by

glucose isomerase [20,24].

The use of the thermostable B. licheniformis a-amylase at 95C in the

starch conversion process has several advantages, most notably a greater reaction

rate, decreased risk of microbial contamination, lowered viscosity, and increased

solubility of substrate and product. While liquefaction rapidly occurs within two

hours at pH 6.5, the conversion of maltodextrins to glucose or maltose syrups

currently suffers several drawbacks. Saccharification by amyloglucosidase or

B-(maltogenic) amylase requires a drop in temperature to 55-600C, and pH

to 4.0-4.5 or 5.0-5.5, respectively. Furthermore, unacceptably long holding times for

saccharification (48-96 h) are required to achieve maximal yields of glucose or

maltose [20,24,72]. Therefore a new amyloglucosidase, a-glucosidase, or maltogenic

amylase with higher thermostability and pH optimum may have commercial

applications following the partial hydrolysis of starch.

Since most commercial enzymes are of microbial origin, the development of

industrially important thermostable enzymes has predominantly involved three









3

approaches. One system employs screening multitudes of bacteria, usually

thermophilic, for the production of a naturally occurring enzyme with high

thermostability and pertinent biochemical characteristics. A second approach

involves mutagenesis of the microorganism that produces minute quantities of a

desired enzyme, such that expression of the enzyme's gene is derepressed and the

enzyme is overproduced. The third technique is currently the most actively pursued

among enzymologists and molecular biologists. The gene for a thermolabile enzyme

is cloned, and specific amino acid substitutions are introduced into the enzyme by

cassette mutagenesis of the gene. Theoretically, these amino acid alterations should

produce additional intramolecular non-covalent or covalent interactions that will

increase the enzyme's resistance to denaturation during elevated temperatures.

Unfortunately, all three approaches are tedious, time consuming, and cost

inefficient. In particular, site-directed mutagenesis requires knowledge of the

enzyme's three dimensional structure through x-ray crystallography studies. The

strategy then relies upon knowledge of the primary amino acid sequence, and the

ability to predict which of 20 possible amino acid substitutions might result in an

increase in enzyme thermostability, without sacrifice of catalytic efficiency.

A previously developed, unique procedure [39] may prove to be a viable

alternative to achieving thermostability in commercially important, thermolabile

enzymes. The approach involves the genetic conversion of a mesophilic strain to

a thermophilic strain, by transforming the mesophile with a small fragment of









4
genomic DNA from a thermophilic donor of the same genus. It should be stressed

that only a small, but critical fragment of DNA is transferred, which induces wide

pleiomorphic effects. The high temperature growth (HTG) transformants resulting

from this genetic conversion are then evaluated for their ability to produce proteins

and enzymes that are thermostable in vivo. Therefore, this method does not rely

upon gene cloning to obtain knowledge of the primary amino acid sequence of the

thermolabile enzyme; nor does it rely upon one's ability to predict amino acid

substitutions that may result in increased enzyme thermostability. The strategy is

to simply allow the intracellular biochemical processes of the HTG transformant to

perform the subtle "work" necessary to convert all the cellular components,

including enzymes, into thermostable macromolecules.

The overall aim of this dissertation was to screen a wide range of HTG

transformants, generated from the mesophile Bacillus subtilis 25S, for their ability

to produce various thermostable starch degrading enzymes. This evaluation

involved searching for a possible glucogenic amylase that may have industrial value.

After studying several glucan hydrolases produced by one HTG transformant,

B. subtilis H-17, an enzyme was chosen based on its ability to degrade

maltodextrins. During this investigation, the enzyme was classified as a

cyclomaltodextrinase. Within the overall aim was the biochemical, biophysical and

genetic characterization of the cyclomaltodextrinase, and its comparison to the

analogous enzymes from the mesophilic and thermophilic parents.









5

The specific objectives to this study were as follows:

(a) To determine optimal growth conditions for maximal production of the

cyclomaltodextrinase from B. subtilis 25S, B. caldolyticus C2, and B. subtilis

H-17.

(b) To purify, characterize, and compare the thermostable cyclomaltodextrinase

from H-17 and the thermolabile cyclomaltodextrinase from 25S.

(c) To determine the possible molecular mechanisms) which confer the

thermostable properties of the H-17 cyclomaltodextrinase, based on structure-

stability relationships

(d) To determine if the donor, B. caldolyticus C2, also produces a thermostable

cyclomaltodextrinase.

(e) To show that the H-17 cyclomaltodexrinase is not a product of a specific

thermophilic gene transferred from donor to recipient during transformation

of 25S to thermophily.

(f) To determine if the H-17 cyclomaltodextrinase has advantages over currently

used fungal amyloglucosidase or B-amylase for the industrial conversion of

starch to glucose or maltose, respectively.

(g) To clone the H-17 cyclomaltodextrinase gene and express its product in a

suitable host.

(h) To determine the thermostability of the cloned H-17 cyclomaltodextrinase

gene product in a mesophilic host.










6

The results would hopefully demonstrate the potential for development of

thermostable starch-processing enzymes by the generation of HTG transformants

from generally recognized as safe (GRAS) status Bacillus mesophiles. Furthermore,

cloning the H-17 cyclomaltodextrinase gene may allow a determination of its origin,

and the molecular mechanisms) which confer enzyme thermostability within the

HTG transformants.













CHAPTER 2
REVIEW OF LITERATURE



Thermophiles and Thermophily

Temperature is one of the most important environmental factors that affects

biological processes, the structure and metabolic function of cellular components,

and the evolution of life. The majority of living organisms have adapted to a

moderate environmental temperature near that of the Earth's surface average of

12C [116,117]. While the upper temperature limit for all living organisms is still

unknown, it is assumed that all forms of higher organisms do not survive above

500C. In addition, no eukaryotic microorganisms are known to exist above 620C.

Molds and yeasts are generally considered thermophilic if they grow as high as 40-

50C [11,83,116]. However, in high temperature environments, many prokaryotic

microorganisms have upper temperature limits greater than 60-70C. Thus far, only

bacteria are able to adapt optimally to elevated temperatures. Bacterial growth in

nature has been shown to occur from -50C to 1000C [83,116]. The bacterial species

diversity within a high temperature region is further influenced by other

environmental parameters such as pH, available energy sources, osmolarity, mineral

content, and toxic metals. Examples of extreme thermal environments include

terrestrial hot springs, deepsea hydrothermal vents, continental and submarine









8

volcanic areas, small bays warmed by the sun, solar heated soil, thermally polluted

streams, geothermal power plants, hot water storage tanks, and mechanical heating

systems [11,44,88,116]. For bacteria living in regions of high environmental

temperatures, molecular mechanisms of thermophily, and how such organisms

evolved, have become a focal point of investigation.

Bacteria are usually classified into arbitrary groups based upon their

preference to live within a limited temperature optimum for growth. Psychrophilic

bacteria are capable of growth between -5-250C. Mesophiles have temperature

optima between 25-450C. Thermophilic bacteria show optimal growth between

45-100C [115,116]. Thermophiles may be further divided into (a) facultative

thermophiles, which have an optimal growth temperature of 45-550C but may grow

at 25-300C, (b) obligate thermophiles, which have an optimal growth temperature

of approximately 55-65C but cannot grow below 400C, and (c) extreme

thermophiles, which grow optimally above 70C but cannot grow below 50C

[28,79,108]. However, one should be cautious when classifying according to growth

temperatures since many microorganisms are borderline and could be assigned to

either category depending upon growth conditions and the investigator's point of

view.

In order to live optimally within their normal environmental temperature

range, microorganisms must structurally adapt their proteins, nucleic acids, and

lipids to function efficiently [115,116]. Since bacterial cells are essentially aqueous









9

chemical systems, their viability may only be limited to temperatures at which water

exists in a liquid state [81,116]. Indeed, microbial growth is thought to occur at

temperatures greater than 100C, in which water remains in a liquid state under

high pressure, for example, near deepsea hydrothermal vents [83]. However, every

bacterial species has optimal growth within a relatively limited temperature range

that rarely exceeds 30C. This environmental temperature regulates in the cell, not

only the rates of enzyme catalyzed reactions, but also the structural state of water

and of active biopolymers such as proteins, nucleic acids, and lipid membranes

[115,116]. The structure and function of cellular biopolymers is based on the

number and distribution of noncovalent bonds. Since these relatively weak bonds

are essential for stability and function, then even slight inputs of thermal energy can

disrupt noncovalent interactions which may lead to drastic structural and functional

alterations [114]. To achieve thermoadaptation, it is reasonable to assume that the

functional molecules of metabolism, and the structural molecules of cellular

components have been adapted such that bacteria are designed to attain maximal

metabolic efficiency at their optimal growth temperature [115,116]. Not only do

cellular components of thermophiles become structurally thermostable, but

thermophilic metabolism is based upon enzymes that have adapted to a high

temperature optimum for biological activity [114].










10

The Origin and Genetics of Thermophiles

From an evolutionary standpoint, arguments regarding the origin of

thermophiles have developed along two philosophies. The first, that thermophiles

evolved from mesophiles by either adaptation or mutation is based upon the

ubiquitous occurrence of thermophilic species in non-thermophilic environments

[82]. If the thermostabilities of proteins and enzymes are encoded in their

structural genes, then thermostability may be altered if a mutation occurs.

Therefore, one could hypothesize that any type of microorganism could develop

thermophilic properties if given sufficient time and opportunity. However, the

conversion of a mesophile to a thermophile would require so many mutations that

the probability of all these occurring in one cell is extremely small. Since the

phylogenetic transition from a mesophile to a thermophile cannot be due to a

spontaneous mutation of a single protein, it is improbable that acceptable mutations

of most, if not all, enzymes and proteins could occur that would render a

corresponding variation in thermostability and biological activity [114].

A more credible argument, that mesophiles originated from thermophiles, is

based on speculation that the earliest cellular forms evolved in primordial waters

considerably warmer than contemporary oceans or lakes [82]. The spontaneous

transition between thermophilic and mesophilic bacilli, as observed in recent years,

not only supports this second argument, but suggests that thermophilic properties










11

are encoded in a small number of adaptor genes that control structure genes

[17,85,116].

Several attempts to identify the genetic determinants) of bacterial growth

temperatures have implied that the transfer of a small number of genes may

transform a mesophile into a thermophile, or vice versa. Cotransformation and

cotransduction experiments in Bacillus subtilis [47,48] demonstrated that the

temperature sensitive locus (tms-26) is closely linked to the streptomycin region.

The function of the tms-26 marker in a high temperature growth B. subtilis mutant

was to confer growth at 550C. Since genetic modification of the S-12 protein within

the 30s ribosomal subunit confers streptomycin resistance, the study suggests the

tms-26 gene's involvement with ribosomal structure.

DNA studies [12,18] of several related Bacillus species show that the region

surrounding the streptomycin marker is highly conserved in base sequence. This

conserved core of genetic material also includes genes which encode for ribosomal

and transfer RNA's, and resistance to other antibiotics such as erythromycin and

micrococcin. While there appears to be a gradient of conservatism, the results

suggest the region to be relatively resistant to evolutionary change.

This high base sequence homology may have allowed recombination of a

segments) of Bacillus caldolyticus DNA with the B. subtilis chromosome [39]. In

a unique study, mesophilic B. subtilis (strs, purA [ade 16], growth optimum 370C)

was transformed with purified genomic DNA from the obligate thermophile









12

B. caldolyticus (strr, purA+ [ade 16], growth optimum 720C). Strr, purA B. subtilis

transformants were isolated that grew at 700C and 550C but not at 370C. Since the

majority of genes encoding for antibiotic resistance, ribosomal proteins, rRNA,

tRNA, and protein synthesizing components are located immediately following the

purA gene in the early replicating region of the B. subtilis chromosome [64], the str

and purA+ genes were assumed to be cotransformed with those encoding for

ribosomal and tRNA functions. The study suggested that recombination of these

genes into the host would induce wide pleiomorphic effects in which microbial

stability is achieved by converting the entire cell of the mesophile into that of a

thermophile. Specifically, alteration of the protein-synthesizing machinery at the

tRNA or ribosomal level would produce translationally-modified enzymes and

proteins with increased thermostability. To support their assertion, the researchers

isolated L-histidinol dehydrogenase (HDH) from the thermophilic transformants.

Compared to inactivation at 700C of HDH from the recipient B. subtilis, the HDH's

from several high temperature growth transformants were thermostable at 100C,

whether assayed as crude extracts or as purified enzymes.

In a similar study [22], B. subtilis was transformed with DNA from Bacillus

stearothermophilus or B. caldolyticus. High temperature growth transformants that

grew at 650C were isolated. The ribosomal proteins from the thermostable

ribosomes of the transformants were analyzed using two-dimensional polyacrylamide

gel electrophoresis. The gel patterns suggested that essentially all the genes that









13

encode for ribosomal proteins were transferred from donor to recipient. However,

the mechanism by which these and any other cotransferred genes exert microbial

thermostability in the host has yet to be determined.

An alternative approach [17] screened for spontaneous thermotolerant

mutant derivatives of mesophilic bacteria at 10C above their upper growth limit.

When prototrophic strains of B. subtilis and Bacillus pumilis were plated out in large

numbers, thermophilic mutants that were able to grow between 50C and 70C were

isolated at a frequency of 10 DNA from one thermophilic B. subtilis mutant was

used to transform a mesophilic lys', trp" B. subtilis auxotroph. Doubly-prototrophic

(lys trp ) B. subtilis transformants were thermophilic, ie. grew between 50C and

700C. The transformation frequency of the thermophilic trait was about 10"7, while

the cotransformation frequency of the two unlinked prototrophic markers (lys ,

trp ) was also about 10 Since the transformation frequency of the thermophilic

trait is similar to that for the transformation of two unlinked genes, the authors

suggest that thermophily is the result of mutations in two unlinked genes. However,

they do not speculate as to what those two genes are, and how close they are to

the lys and trp markers. In a follow-up study [23], ribosomes from one of the

spontaneous thermophilic B. subtilis mutants were thermostable at 60C for 30

minutes, while ribosomes from the mesophilic parent were completely inactivated.

Furthermore, the addition of polyamines to the cell-free extract from the

thermophilic mutant stimulated polyphenylalanine synthesis at both 550C and 650C,









14
but inhibited protein synthesis with the cell-free extract from the mesophilic parent.

The authors suggest, in the transition from mesophile to thermophile or vice versa,

global mechanisms are in operation and that prime candidates may include genes

encoding for polyamines, protein methylases, and DNA topoisomerases.



Thermostability of Cell Components

While the origin of thermophiles and their evolutionary relationship to

mesophilic species within the same genus is by no means established, thermophilic

and mesophilic microorganisms do seem to have a common origin [44,114,115].

Not only are thermophilic species found in most bacterial genera, but they resemble

their mesophilic counterparts in that they ferment similar carbohydrates, utilize

similar nitrogen sources, and have similar oxidative pathways. Cellular structures

and sequence homologies of proteins, enzymes, and nucleic acids of thermophiles

are also very much alike if not nearly identical to that of their mesophilic

counterparts except that the former usually have much higher thermostabilities

[41,82,114].

With respect to mesophiles, several mechanisms have been proposed to

explain the optimal growth and reproduction of thermophiles at high temperatures.

Early theories attribute heat stability of thermolabile components to a) low cellular

water content, b) transport of protective factors such as calcium from the

environment into the cell, c) alteration in the nature of cell membranes, and d)









15

rapid resynthesis of heat-denatured cell components [43,114]. Newer theories

include a) synthesis of organic polymers, such as polyamines, that act as protector

molecules, b) lipid interaction such that cell membranes stabilize heat labile

macromolecules, and c) site-specific biochemical modification of macromolecules

[58]. Perhaps there is no single mechanism that confers thermophily. Rather, a

combination of molecular alterations allow optimal growth at elevated

environmental temperatures. However, contrary to earlier theories, a massive

turnover of cell components does not occur. Rather, thermophily is based on the

thermostability of individual cellular components [90]. It now appears evident that

structural comparisons of thermophilic and mesophilic nucleic acids, proteins, and

lipids from closely related organisms can best provide a molecular explanation for

the exceptional ability of thermophilic bacteria to live at high temperatures.



Deoxyribonucleic Acid

Studies [79,87,104] concerned with the thermostability of DNA have

compared the guanine plus cytosine (G+C) content and melting temperature of

DNA isolated from thermophilic strains with that of mesophilic strains of the same

genus. The DNA of thermophiles showed a consistently higher G+C content than

that of DNA from mesophiles. Thermal melting profiles demonstrated that

thermophilic DNA had higher melting temperatures than DNA from mesophiles.

The greater stability to thermal-induced strand separation is attributable to a more









16

extensive hydrogen bonding that occurs with a higher G+C content. Therefore, the

G+C content often correlates with maximum growth temperature, although

Clostridium species are an exception. However, one cannot conclude that the

thermostability of DNA, due to a higher G+C content, has any relationship with

the ability of a thermophile to grow at high temperatures.



Ribonucleic Acid and Ribosomes

The base composition of messenger RNA (mRNA), isolated from the

obligate thermophile B. stearothermophilus, was, as expected, nearly identical to the

values obtained for DNA from the same organism [77]. As with DNA, it was

concluded that the thermostability of mRNA probably does not play an important

role in the ability of thermophiles to grow at high temperatures. Since active

mRNA exists in a linear single stranded form, elevated growth temperatures should

also have little or no effect on the secondary structure of mRNA, regardless of

G+C content.

The base compositions of transfer RNA (tRNA) from several strains of

B. stearothermophilus have been shown to be very similar to that reported for the

mesophile Escherichia coli. In addition, the thermal melting profiles of tRNA from

B. stearothermophilus and E. coli were nearly identical [19,43,104]. The tRNA(Val)

and tRNA(Phe) from B. stearothermophilus also reacted with the respective valine

and phenylalanine-amino-acyl-tRNA synthetases from E. coli and yeast [43].









17
Consequently, it appears that tRNA structures are very similar, regardless of

microbial source. However, the G+C content of tRNA from the extreme

thermophile Thermus aquaticus (63.5%) was higher than that of either E. coli

(59.5%) or B. stearothermophilus (58%) [113]. Furthermore, tRNA from T.

aquaticus and E. coli had melting points of 860C and 80&C, respectively. Since T.

aquaticus has a maximum growth temperature 11-170C higher than that reported

for strains of B. stearothermophilus, it is possible that the enhanced thermostability

of T. aquaticus tRNA is a reflection of the greater G+C content.

Related studies [101-103] have also shown that tRNA from the extreme

thermophile Thermus thermophilus not only has a melting point higher than that

from mesophiles, but has a high G+C content (90%) in the base paired region and

a modified base, 5-methyl-2-thiouridine (or 2-thio-ribothymidine), instead of

unmodified ribothymidine in the T-loop. Furthermore, when they compared the

sequence of tRNA specific for formyl-methionine from T. thermophilus with that of

E. coli, they found not only a 90% G+C content in the base-paired region of the

thermophilic tRNA, but that a G-U pair in E. coli tRNA was replaced by a G-C

pair in the T. thermophilus tRNA. The results indicate that thermophilic

tRNA(fMet) is thermally stabilized by formation of an extra intramolecular

hydrogen bond when a uridyl residue is replaced with a cytydyl residue, and by

increased stacking interaction due to thiolation of a ribothymidine residue.

Although the changes are quite subtle, they are located in a region quite distant









18

from the anticodon loop and amino acid accepting terminal, both of which are

important for tRNA function.

Reviews [19,43,90,104] indicate a variety of conflicting results regarding the

effect of ribosomal RNA (rRNA) G+C content and melting points on the

thermostability of ribosomes. The thermal melting profiles of ribosomes have

shown that ribosomes from thermophiles undergo thermal denaturation at higher

temperatures than ribosomes from mesophiles, perhaps due to the increase in G+C

content of rRNA from thermophiles. Elucidation of the molecular basis of

ribosomal thermostability remains unclear. Although the rRNA nucleotide

composition may play a significant role, the same authors [19,43,90,104] agree that

other factors such as primary structure of ribosomal proteins, the stacking

arrangement of ribosomal protein and rRNA, and the association of polyamines

with ribosomes, are probably more important in the stability of ribosomes in

thermophiles. Specific polyamines, such as the tetramine spermine and the triamine

spermidine, both synthesized by B. stearothermophilus, have been shown to aid in

the association of ribosomal subunits at high temperatures [43,56]. More than 12

distinct polyamines, NH,(CH,)3[NH(CH2)3],NH(CH,)4NH,, were also isolated from

T. thermophilus [56,59]. The tetramines, thermine and thermospermine were the

main forms, while the pentamines, caldopentamine and homocaldopentamine, and

hexamines, caldohexamine and homocaldohexamine were also present. Thermine

or thermospermine were shown to initiate and maintain polyphenylalanine synthesis










19

directed by polyuridylic acid at high temperatures (50-80C). When the tRNA(Phe),

polyuridylic acid, and ribosomes were incubated at high temperature in the absence

of the polyamine, a ribosome-mRNA-amino-acyl-tRNA ternary complex also formed

but was inactive. The results indicate that during protein synthesis, thermine and

thermospermine may play an in vivo role in the initial formation of the active

ternary complex between ribosomes, mRNA, and amino-acyl-tRNA at high

temperatures.

Thermine was also found in the extreme thermophiles T. aquaticus and

Thermus flavus, while no detectable amount of thermine occurred in the obligate

thermophile B. stearothermophilus [56]. It appears that, not only thermine and

thermospermine, but other novel polyamines such as caldine and sym-homo-

spermidine occur only in extreme thermophiles, which implies their involvement in

extreme thermophily [57]. Although not yet understood, these unusual polyamines

may also play important roles in other biochemical reactions such as DNA

replication, transcription, and cell division, besides just translation.



Lipids and Cell Membranes

The central structural feature of microbial membranes is the phospholipid

bilayer, which consists of peripheral proteins bound to the polar portion of the lipid

bilayer, and integral proteins embedded within the nonpolar portion of the lipid

bilayer [43,49,105]. A unique feature of the phospholipid bilayer is its ability to









20

undergo a reversible phase transition from a thermotropic gel (solid) to a liquid-

crystalline state [49]. In the gel or solid state, the fatty acyl chains form a close

hexagonal packing, which results in a restricted inter and intramolecular motion,

and a rigid, somewhat impermeable bilayer structure. During the gel to liquid-

crystalline transition, selective melting of the phospholipid hydrocarbon chains in the

interior of the bilayer occurs. However, the transition is not sharp but broad due

to heterogeneity of fatty acyl chains, which results in simultaneous domains of gel

and liquid-crystalline phases. In the liquid-crystalline state, although the

hydrocarbon chains are in a partially melted condition, the bilayer structure is

maintained by electrostatic interactions between polar head groups and hydrophobic

forces, which allows for a loosely packed, fluid, and somewhat permeable bilayer

structure [49]. Consequently, cell membranes must be in the liquid-crystalline state

such that transport functions, and activity of membrane-associated enzymes ensure

cell growth [43].

The temperature range at which the gel to liquid-crystalline transition occurs

depends upon the nature of the fatty acyl content of membrane lipids. It is

generally accepted that most microorganisms alter their membrane lipid composition

in response to a change in environmental temperature. The membranes of

thermophiles have shown a higher content of apolar fatty acyl residues with higher

melting temperatures than the residues of psychrophilic and mesophilic

microorganisms. This would minimize the effect of high environmental temperature









21

on the physical state of the membrane lipids by raising the temperature of the gel

to liquid-crystalline phase transition. Therefore, when thermophilic bacteria adapt

to increases in cultivation temperature, their membranes shift to a higher

proportion of longer, saturated fatty acids, and monomethyl-branched chains, and

to a decreased proportion of unsaturated fatty acids [36,43,49,67,70,104,105]. As

a general trend, the iso- and anteiso- monomethyl-branched chains tend to be the

predominant form of apolar fatty acyl residues in thermophilic bacteria. However,

in nearly all thermophiles examined, the majority, if not all, of the complex (polar)

lipids contain a carbohydrate residue. It appears that lipids enriched in

carbohydrates usually form the major lipid class, which implies a significant role in

membrane thermostability [36,90]. Although not well understood, analysis of lipids

from thermophiles suggest general trends in which there may be a combined variety

of strategies for the molecular basis of thermostable membranes.



Proteins and Enzymes

Proteins have a limited temperature range within which structural integrity

is maintained. All known proteins (enzymes) unfold, denature and thereby lose

biological activity upon a certain intensity of thermoexposure. A thermophilic

enzyme is more thermostable at higher temperatures, but less active at lower

temperatures than a mesophilic enzyme. Thermostable enzymes from thermophiles,

compared to thermolabile enzymes from mesophilic species of the same genus,










22

resist thermal unfolding at elevated temperatures due to differences in the number,

strength, and distribution of their intramolecular noncovalent forces [32,50,99,117].

Therefore, the primary structure of an enzyme ultimately determines the number

and type of noncovalent interactions. Amino acid sequence analysis of several

mesophilic enzymes and their thermophilic counterparts has shown that a few

specific amino acid substitutions at critical regions may account for very subtle

structural and conformational differences that lead to large alterations in

thermostability [2,32,99,117]. That is, a gain in thermostability of an enzyme usually

does not require a drastic or extreme structural rearrangement in its conformation,

but does require structural maintenance of its catalytic site. An enzyme from a

thermophile consistently shows the same structural characteristics such as

polypeptide chain length, secondary structure, catalytic site, globular domains,

subunit structure, and modulation of activity by metal ions and effectors, as does

the same enzyme isolated from a mesophile of the same genus [52,86].

Probably the most important, if not decisive, mechanism that confers protein

thermostability is hydrophobic interaction [3,50,99]. The total hydrophobic amino

acid content of a thermophilic enzyme, and that of its mesophilic counterpart, may

or may not greatly differ. Rather, thermostability depends upon an increased

proportion of hydrophobic residues around the active site. The overall effect is a

greater internal and decreased external hydrophobicity. Therefore, strengthening

of the internal nonpolarity is achieved by substitution of specific amino acid









23

residues, characteristic of mesophilic enzymes, with aliphatic amino acids, with

retention of the overall structure and catalytic properties. As a result, internal

hydrophobic domains contribute to compact packing of amino acid residues. A

compact globular structure excludes water from internal cavities, which allows for

improved enzyme stability.

A second factor that confers protein thermostability is the formation of

additional electrostatic interactions such as divalent cation salt bridges and ion pairs

[50,62,63,99]. Since electrostatic interactions are formed between negatively-charged

aspartate and glutamate residues and positively-charged lysine, histidine, and

arginine residues, then a greater proportion of these residues in a thermophilic

enzyme may account for a higher thermostability than that of its mesophilic

counterpart. Furthermore, thermophilic enzymes frequently have higher levels of

arginine relative to lysine and histidine. Because the pKa's for lysine and histidine

are lower than the pKa for arginine, at alkaline pH values, lysine and histidine side

chains will dissociate and electrostatic interactions will be disrupted. Greater

thermostability results from stronger electrostatic interactions formed by arginine

rather than lysine or histidine [32].

A third molecular mechanism that enhances enzyme thermostability is an

increase in intramolecular hydrogen bonds [50,99]. A comparison of thermophilic

ferredoxins and hemoglobins with mesophilic counterparts concluded that

thermostability is increased by the formation of a few new hydrogen bonds and salt









24

bridges [63]. In addition, 19 additional hydrogen bonds were detected in

thermophilic protease, which were absent in the mesophilic enzyme [54]. It appears

that an alteration in the number of hydrogen bonds sometimes induces a change

in the secondary structure of thermophilic enzymes compared with their mesophilic

counterparts. However, some thermophilic proteins lack polar amino acids, mostly

serine and threonine [50]. Assuming a high degree of internal hydrophobicity,

internal localization of serine and threonine is thermodynamically unfavorable. The

substitution of internal series and threonines with nonpolar residues theoretically

increases thermostability. Therefore, in some thermophilic proteins, intramolecular

hydrogen-bonding may be assumed to be localized at the surface of the molecule.

Similarly, in an enzyme's compact interior, all polar groups are hydrogen-bonded

and that alteration of size, shape, or polarity of a single side chain could destabilize

enzyme structure [86]. Since water mobility presents little barrier to intramolecular

hydrogen bonding alterations between surface residues, an increase in polar and

a decrease in nonpolar surface residues could enhance thermostability.

Other factors that enhance protein thermostability include the formation of

intramolecular disulfide bonds, substrate binding, and post-translational modification

such as glycosylation or chemical modification of key surface groups [50,52,99].

However, one enzyme's mechanism of thermostability may be different from

another's. Thermostabilization may also be attributable to the simultaneous

contribution of several different mechanisms. Consequently, interpretation often









25
becomes complex and even contradictory when comparing structural mechanisms

of thermostability among different thermostable enzymes. Rather than present a

comprehensive review of all comparisons between thermostable enzymes and their

mesophilic counterparts, only a few significant studies will be presented in which

the factors that confer thermostability are somewhat understood.

The complete lactate dehydrogenase (LDH) primary sequences from the

thermophiles Bacillus caldotenax, B. caldolyticus, and B. stearothermophilus, the

mesophiles Bacillus megaterium, and B. subtilis, and the psychrophile Bacillus

psychrosaccharolyticus underwent an extensive comparative structural analysis

[25,115,117]. The sequence homology between the LDH variants was between 60%

and 70%. Specific temperature-related amino acid substitutions occurred in which

polar amino acids, particularly serine and threonine, in the mesophilic

(psychrophilic) enzymes are exchanged for hydrophobic and charged residues,

particularly alanine, arginine, and aspartate in the thermophilic enzymes. These

substitutions did not occur among thermophilic or among mesophilic LDH variants,

which contrasts to the 12-13 substitutions between thermophilic and mesophilic

LDH. Three-dimensional structure analysis of all the LDH variants indicated that

most of the substitutions were buried in the same strategically important regions,

particularly near the active site and in contact regions of the subunits. This

indicates that there are apparent regions of significant importance in temperature

adaptation. Furthermore, in the transition of mesophilic (psychrophilic) to









26

thermophilic LDH, a dynamic hydrogen bond system is converted to hydrophobic

interactions or ion pairs in which the free energy of the enzyme is increased.

Conversely, in the transition of thermophilic to mesophilic (psychrophilic) LDH, a

reduction in energy-yielding hydrophobic interaction and ion pairs occurs in favor

of hydrogen bonding. The author concluded that, at high temperatures, the

preferred thermophilic residues of thermophilic LDH should increase

thermostability, but produce a more rigid and less active structure at low

temperatures. However, polar residues, perhaps hydrated, should lead to a more

dynamic, flexible, and active structure at low temperatures, but a labile structure

at high temperatures. These temperature-related amino acid substitutions are

stored in the structure of the genetic code, and hence are based on evolutionary

temperature-related base substitutions.

The genes for alanine dehydrogenase were cloned and sequenced from the

mesophile Bacillus sphaericus and B. stearothermophilus [34]. A comparison of the

primary amino acid sequences showed a 73% homology, with the non-identical

residues clustered in a few regions of relatively short length. The residues involved

in catalysis and coenzyme binding were conserved in a sequence of high homology

(>80%), while the short sequences of low similarity were believed to contribute to

the thermostability of the B. stearothermophilus enzyme. An additional cysteine in

the B. stearothermophilus alanine dehydrogenase may form an interior disulfide bond

with one of two nearby cysteines, which may enhance thermostability.









27

The oligo-1,6-glucosidases from Bacillus cereus (mesophile), Bacillus coagulans

(facultative thermophile), Bacillus sp. KP1071 (facultative thermophile), Bacillus

thermoglucosidasius (obligate thermophile), and Bacillus flavocaldarius (extreme

thermophile) were compared for amino acid composition, structural parameters, and

thermostability [95]. Results showed that the proline content greatly increased in

a linear fashion with the increase in thermostability among all five enzymes. The

hydrophobic residues, in particular alanine and leucine, also showed upward

tendencies in parallel with the increase in thermostability, while polar residues

showed downward trends. Although the enzymes were quite similar in terms of

structural parameters, the content of a-helix former decreased with an increase in

thermostability, while B-sheet former remained nearly constant. The increase in

proline, an a-helix breaker, could contribute to improved turn stabilization. While

this may increase the disordered regions at the expense of helix formations, it could

produce a greater potential for close packing of hydrophobic regions. The

strengthened hydrophobic interactions would tighten the molecule as a whole and

thereby enhance thermostability.

The thermostable a-amylase from B. stearothermophilus was compared to the

thermolabile a-amylase from B. subtilis [111]. Results indicated that binding of

calcium ions to both mesophilic and thermophilic a-amylase considerably enhances

the stability of the native conformation of both enzymes. However, calcium-free

thermophilic a-amylase was very susceptible to thermal denaturation, but was more









28
resistant to heat than calcium-free mesophilic a-amylase. This suggests differences

in the amino acid sequence between the two proteins. It was concluded that the

difference in thermostability of a-amylase from B. stearothermophilus and B. subtilis

is caused by the difference in the enzyme's affinity to the calcium ion at elevated

temperatures, which may be a function of amino acid sequence.

The liquefying a-amylases produced by Bacillus licheniformis, Bacillus

amyloliquefaciens, and B. stearothermophilus contained highly homologous amino acid

sequences, with a 64% homology between the B. stearothermophilus and

B. amyloliquefaciens enzymes, a 67% homology between the B. stearothermophilus

and B. licheniformis enzymes, and an 80% homology between the B. licheniformis

and B. amyloliquefaciens enzymes [26]. Because the regions of extensive homology

include active sites, they probably are required for maintaining protein conformation

and enzymatic activity. Furthermore, the thermostable B. stearothermophilus

a-amylase contained two cysteine residues located near the enzymatically functional

region, while the thermolabile B. amyloliquefaciens enzyme contained no cysteine

residue. The formation of a disulfide bridge between the two cysteine residues

could account for thermostability in the B. stearothermophilus a-amylase. However,

the structural differences in non-homologous regions of the three a-amylases are

presumed to be responsible for differences in thermostability [112]. Because the

hydropathy profiles of all three enzymes were significantly hydrophilic, the authors









29

suggested that salt bridges between charged and polar amino acids within non-

homologous regions may account for thermostability of the a-amylases.



Protein Engineering of Industrially Important Enzymes

The current total world market for industrial enzymes is over 500 million

dollars per year. About 20 microbial enzymes account for the majority of this

market. Microbial enzymes applied to food processing and industrial operations

include protease, a-, B-, and gluco-amylase, pullulanase, glucose isomerase, cellulase,

hemicellulase, lipase, pectinase, lactase, and alcohol dehydrogenase [14,52,83].

Where possible, thermostable enzymes are now utilized extensively for industrial

processing. Adequate thermostability of commercial enzymes may be defined as

retention of enzyme activity upon exposure to temperatures of 50C or above, for

prolonged periods [50,99]. Their industrial application may be advantageous for

several reasons [32,52,84,99]: (a) Higher reaction rates can be obtained, since for

every 10C increase in temperature, reaction rates approximately double.

Consequently, for each 10C increase in operating temperature, the holding times

can be shortened, or the amount of enzyme required for a given conversion can be

theoretically halved. (b) Microbial contamination of enzyme reactions lasting

several days is less likely to occur at operation temperatures of 60C or greater.

(c) At elevated temperatures, higher reactor productivity may be achieved due to

greater solubility of reactants, reduced viscosity, and improved mass transfer rate.









30

(d) Enzymes of high thermostability often show increased resistance to chemical

denaturation and thus longer shelf lives.

One factor that limits commercial enzyme application is the high cost of

isolating and purifying sufficient amounts of the enzyme. A second limitation is

that, while enzymes have evolved to function optimally under normal physiological

conditions, they may not function under nonphysiological industrial conditions that

include extremes of pH, ionic strength, oxidation, and temperature [27,52,97].

Consequently, robust enzymes with longer half-lives under process conditions are

required for industrial applications. As a solution, genetic engineering techniques

have attempted to the improve production of commercial enzymes by

(a) amplification of the production of specific enzymes by mutation of

microorganisms, (b) cloning and synthesis of enzymes in Generally Recognized As

Safe (GRAS) organisms, and (c) protein engineering by genetic modification

(mutagenesis) in which enzyme structure is altered such that one or more functional

properties are improved under nonphysiological, extreme conditions, that is, high

temperatures.

The major cost savings resulting from the benefits of thermostable enzymes

currently prompt the development of thermostable properties to be engineered into

enzyme molecules based on structure-stability relationships. Specifically, one goal

of protein engineering is to enhance the thermostability by genetically introducing

new noncovalent and/or covalent bonds within the enzyme. Genetic modification









31

may involve several techniques [65,99,107]. One approach relies upon knowledge

of the enzyme's three-dimensional structure by high resolution x-ray crystallography

and computer analysis, which delineates possible amino acids responsible for

thermal sensitivity [1]. Variant amino acid sequences may then be designed by site-

directed mutagenesis of the cloned gene encoding for the enzyme, so that specific

nucleotide substitutions are created. Alternatively, cassette mutagenesis may

achieve the same substitutions utilizing synthetic oligonucleotides. Restriction and

ligation enzymes are employed to replace any sequence of the cloned gene with the

synthetic fragment carrying one or more specific nucleotide substitutions (or

deletions/insertions). The recombinant plasmid is then transformed into the

appropriate host and transformants are screened for production of thermostable

enzyme. However, either approach may (a) be time consuming due to the

substitution of many possible amino acids at each site of interest, and (b) cause

destabilization due disruption of pre-existing stabilizing interactions.

At Genex Corp., site-directed mutagenesis has successfully introduced a

disulfide bond into subtilisin BPN, the alkaline protease secreted by B. subtilis [60].

While this enzyme has no pre-existing cysteine residues in the wild type structure,

cysteine substitutions at Thr-22 and Ser-87 positions generated a Cys-22/Cys-87

disulfide subtilisin variant that had catalytic activity essentially equivalent to that of

the wild-type enzyme. The disulfide variant, expressed in subtilisin-negative

B. subtilis, had a melting temperature of 3.10C higher than that of the wild type









32
protein and 5.80C higher than that of the reduced form (-SH HS-) of the variant.

Furthermore, under a variety of kinetic conditions, the disulfide variant underwent

thermal inactivation at half the rate of that of the wild-type enzyme.

Conversely, a related approach, which employs random mutagenesis of the

cloned gene, can unpredictably create new and interesting enzymes that may have

novel properties, including increased or decreased thermostability. The key element

to this approach is the ability to screen large numbers of variants for increased

thermostability. The same researchers at Genex Corp. [69] used chemical

mutagenesis to introduce random mutations into the cloned subtilisin gene, and

transformed the recombinant plasmids into subtilisin-negative B. subtilis. A

chromogenic substrate activity stain of a nitrocellulose bacterial colony lift was

employed to isolated ten enzyme variants with increased resistance to thermal

inactivation. All the variants were the result of a single amino acid substitution due

to a point mutation. One variant enzyme had more than a four-fold thermal

resistance at 650C. A single amino acid substitution of serine for asparagine at

position 218 was present in the variant protease. The mutation, which occurred at

one end of a B-hairpin structure, caused shortening of hydrogen bonds across the

chains of the hairpin [106]. A triple combination mutant was then constructed from

this Asn-218 to Ser variant by using oligonucleotide-directed cassette mutagenesis.

The addition of Gly-131 to Asp, and Thr-254 to Ala mutations increased the

variant's ti2 of thermal inactivation at 650C to 11.6-fold over that of the wild-type,









33

without alteration of catalytic properties. Therefore, minor independent alterations

in amino acid sequence dramatically increased thermostability without radical

changes in the tertiary protein structure.

The genes for the neutral protease from B. stearothermophilus and

B. caldolyticus have been cloned and sequenced [98]. A comparison of the derived

primary amino acid sequences showed both enzymes to differ at only three amino

acid positions, 4, 59, and 66. Furthermore, the B. caldolyticus enzyme had a

thermostability and temperature optimum of 7 to 80C higher than that of the

B. stearothermophilus enzyme. Using cassette mutagenesis, the substitutions Ala-4

to Thr, Thr-59 to Ala, and Thr-66 to Phe were introduced into the

B. stearothermophilus enzyme. The mutation Thr-66 to Phe increased

thermostability by 6.20C, while the mutations Ala-4 to Thr and Thr-59 to Ala

increased thermostability by 1.75 and 1.50C, respectively. While the thermostability

of the triple mutant theoretically should have been 9.450C higher, it was only

equivalent to that of the wild-type B. caldolyticus neutral protease. A three-

dimensional model of the variant enzyme showed the substituted residues to be

surface located. The results indicated that solvent-exposed residues may be

important in conferring thermostability to neutral proteases, even though two

hydrophilic residues were replaced with hydrophobic residues.

Oligonucleotide-directed mutagenesis of the a-amylase from

B. amyloliquefaciens deleted the Arg-176 and Gly-177 residues, and substituted Gin









34
for Glu-178 and Ala for Lys-269 [109]. Results showed the variant enzyme to be

as thermostable as the a-amylase from B. licheniformis, that is, more than 80%

retention of activity after 30 min at 900C. However, the variant enzyme

demonstrated a kinetic temperature optimum of 65C, which suggested reversible

inactivation at temperatures above 650C.

Oligonucleotide-directed mutagenesis was also employed to introduce single

point mutations in the cloned lactate dehydrogenase (LDH) genes from

B. megaterium and B. stearothermophilus [118]. The substitutions of Thr-29 or

Ser-39 to Ala residues in the mesophilic LDH from B. megaterium increased the

enzyme's thermostability by 150C. When alanine was simultaneously introduced at

both positions, a 200C increase in thermostability was observed. The authors

suggest that the more helix-forming alanine residues stabilize the a-B helix of LDH,

and serve to exclude water molecules across the Q-axis between the subunit

a-helices. Unfortunately, an increase in Km for pyruvate resulted, which led to a

three-fold reduction in activity when compared to the wild type enzyme. The

reverse double substitutions, Ala-29 to Thr and Ala-39 to Ser, in thermophilic LDH

from B. stearothermophilus, did not alter the high thermostability. However, the

LDH activity of this variant was increased two-fold. The results indicate the

stability and activity of the B. stearothermophilus and B. megaterium LDH to be

based on a highly cooperative system of noncovalent bonds which is influenced

differentially by amino acid substitutions.









35

A third approach to genetic modification of enzymes involves isolation of

enzyme variants without structural information on the wild type protein. A gene

encoding an enzyme from a mesophile is cloned, introduced into a thermophilic

host, and enzyme activity is selected at the higher growth temperature of the host.

The cloned gene for mesophilic kanamycin nucleotidyltransferase (KNTase) was

introduced into B. stearothermophilus, and kanamycin resistant transformants were

selected at 630C [37]. All the purified KNTase variants were more thermostable

than the wild type enzyme, and all had the same single amino acid replacement of

Asp-80 to Tyr. Variants even more thermostable were obtained from the first

variant by selecting for B. stearothermophilus kanamycin resistance at 700C. All

these Tyr-80 variants carried the same additional substitution of Thr-130 to Lys.

The authors suggest that the alterations at positions 80 and 130 act independently

and additively to thermostabilize KNTase. Furthermore, all the KNTase variants

had specific activities equivalent to that of the wild type enzyme. The advantage

to this technique is that, by selecting for enzyme activity, thermostable variants are

generated in which enhancement of stability was not made at the expense of

catalytic efficiency. This biological selection strategy accounts for all variables of

activity and structural stability simultaneously. As thermophilic host-vector cloning

systems are further developed, the strategy can be readily extended to other

mesophilic genes of greater industrial value.









36

Another approach to achieving enzyme thermostability involves conversion

of the entire cell of a mesophile into that of a thermophile [39]. Although less well

studied, potentially any mesophile, producing a thermolabile enzyme(s) of

commercial interest, could be converted to a thermophile by transforming it with

DNA from a related thermophilic species. This method does not rely upon

structural knowledge of the mesophilic enzyme, nor does it rely upon one's ability

to predict appropriate amino acid substitutions. The high temperature transformant

carries out all the necessary changes to convert a thermolabile enzyme to a

thermostable form. Another advantage is that it does not rely on gene cloning,

mutagenesis, transformation, and screening for expression of the variant cloned

gene. In theory, the transfer of a small number of genes to the mesophile would

exert wide pleiomorphic effects in which translationally modified proteins are

produced with increased thermostability.



Microbial Amylases in Starch Bioprocessing

Amylases are starch degrading enzymes that have several industrial

applications in the production of corn syrups that contain varied amounts of malto-

oligosaccharides, maltose, and glucose. Although amylases are widely distributed

in nature and are produced by a variety of microorganisms, most commercial

amylases are produced from Bacillus species. The composition profile of the corn

syrup produced from starch hydrolysis depends upon the nature of the amylase, and









37

the reaction conditions employed. The three basic types of amylolytic enzymes used

in starch conversion are (a) endo-amylase (a-amylase) (b) exo-amylase (1-amylase

and glucoamylase) and (c) debranching enzyme (pullulanase and isoamylase).

a-Amylases hydrolyze the internal a-1,4-glucosidic bonds in amylose and amylopectin

to produce short-chained maltodextrins, but the a-1,6-glucosidic branches in

amylopectin are not attacked. Depending on the length of malto-oligosaccharides

produced and the source of enzyme, a-amylases are further classified as

saccharifying or liquefying. For example, the B. subtilis saccharifying enzyme

produces large quantities of maltotriose, a slow decrease in starch viscosity, and a

rapid increase in reducing power, while the B. amyloliquefaciens liquefying enzyme

produces mainly maltohexaose, with a slower increase in reducing power but a rapid

decrease in starch viscosity. Because of their extremely high thermostability (850C-

100C), the a-amylases from B. licheniformis and B. amyloliquefaciens are the most

commercially significant and most widely employed [9,20,21,72].

Glucoamylase and B-amylase act in exo-fashion by consecutively cleaving

glucose and maltose units, respectively, from the non-reducing ends of amylose,

amylopectin, and maltodextrins. Although all commercial glucoamylases are of

fungal origin and have low thermostability (40C-60C), they are used for the

production of high glucose syrups subsequent to starch liquefaction. B-amylase,

which produces maltose in the B-anomeric form, is found widely in higher plants.

However, several bacteria, including B. polymyxa, B. megaterium, and B. circulans,









38

produce B-amylases similar to those of plant origin. Although bacterial

B-amylases are also of low thermostability (40C-60C), their industrial application

is in the production of high maltose syrups in excess of 80% maltose [9,20].

The industrially-employed Klebsiella pneumoniae or Bacillus acidopullulyticus

pullulanase catalyses the hydrolysis of a-1,6-glucosidic linkages in amylopectin and

pullulan to produce linear maltodextrins and maltotriose, respectively. Because it

also has low thermostability (45C-60C), it generally is used in combination with

glucoamylase or 1-amylase to improve the yields of syrups high in glucose or

maltose, respectively [9,20,72]. Recently, novel highly-thermostable (>900C)

pullulanases have been isolated from thermophilic anaerobic bacteria. Although

these enzymes cleave the a-1,4-glucosidic linkages of starch, they are classified as

either isopullulanase or neopullulanase based on their hydrolysis of a-1,4 linkages

of pullulan to produce isopanose or panose, respectively [72,81].

The industrial conversion of starch to glucose or maltose syrups currently

suffers several drawbacks [20,21,72]. While liquefaction rapidly occurs within

2 hours at 950C, pH 6.5, saccharification requires a drop in temperature

to 550C-60C, and pH to 4.0-4.5 (glucoamylase) or 5.0-5.5 (B-amylase).

Furthermore, the saccharification step requires a 48-96 hour holding time in order

to achieve maximal levels of glucose or maltose. Other problems include the

formation of reversion products and the possibility of microbial growth at the lower

saccharification temperature [20,72]. Consequently, the development of glucogenic,










39

maltogenic, and debranching enzymes with exceptionally high thermostabilities and

more neutral pH optima would make industrial starch processing more cost-

efficient. Ideally, the production of corn syrups would be a single step process in

which liquefaction, debranching, and saccharification occur simultaneously.

Furthermore, these enzymes must be produced from microorganisms that have

GRAS status by the Food and Drug Administration. To this end, protein

engineering techniques of starch degrading enzymes may eventually produce variant

enzymes with the desired catalytic properties that will optimize any type of starch

conversion process.



Bacterial Cyclomaltodextrinases

Many amylolytic microorganisms capable of catalyzing the hydrolysis of starch

by the production of a-amylases, also produce intracellular and extracellular

maltodextrinases and a-glucosidases. These latter enzymes play an essential role

in the microbial conversion of starch to glucose, by hydrolyzing maltodextrins and

maltose produced from amylolytic hydrolysis of starch [4,20,91,96]. Several bacterial

species also produce cyclomaltodextrinases (CDase) (EC 3.2.1.54), which rapidly

cleave cyclodextrins and linear maltodextrins, but hydrolyze starch at significantly

slower rates [30,31,53]. Cyclodextrins (CDs) are cyclic oligosaccharides composed

of six (a-CD), seven (B-CD), eight (i--CD), or more a-linked glucose units. Because

CDs lack terminal non-reducing glucose residues, they resist the hydrolytic action









40
of exo-amylases. They may competitively inhibit many B-amylases and pullulanases.

Due to their cyclic nature, they are very slowly cleaved, if at all, by endo-amylases

[20,51,71,89]. Consequently, CDases appear to be a separate, special amylase class

in which they often share common biochemical characteristics, substrate specificities,

and end-product profiles.

The purified cyclodextrinase from B. coagulans had a molecular weight, as

determined by SDS-PAGE, of 62,000, and an isoelectric point of 5.0 [31]. The

enzyme optimal activity at pH 6.2 and 500C, and was thermostable at 45"C, pH 7.0

for two hours. The enzyme hydrolyzed maltotetraose, maltopentaose, maltohexaose,

and a-, B-, and y, CDs faster than maltotriose and short chain amylose, but did not

cleave maltose. The enzyme recognized and cleaved the a-maltosyl group in the

non-reducing end [30]. The hydrolysis products had the a-configuration and were

mainly maltose. Starch, amylose, and amylopectin were hydrolyzed at rates 1% that

for B-CD.

The intracellular cyclodextrin-hydrolyzing enzyme from B. sphaericus was

purified and estimated to be a homodimer having a native molecular weight of

144,000, and subunit molecular weight of 72,000 [53]. The enzyme had a pH

optimum of 8.0, was stable at 250C, pH 5.5-9.5 for 24 hours, and was inactivated

at 500C for 10 minutes. The enzyme most rapidly hydrolyzed B-CD, followed by

maltoheptaose, maltopentaose, a-CD, and maltohexaose. Starch, amylopectin,

amylose, and pullulan were degraded at less than 4% the rate of B-CD cleavage.










41

The purified cyclodextrinase from B. macerans had a pH and temperature

optimum of 6.2-6.4 and 30-400C, respectively [15,16]. The enzyme cleaved a-, B-,

and -y CDs, maltoheptaose, maltohexaose, maltoheptaose, maltotetraose, and

maltotriose to mainly maltose. Amylose, amylopectin, glycogen, or starch were

negligibly degraded. For CDs, the enzyme initially opened the ring to give a linear

molecule with the corresponding number of glucose units. Linear malto-

oligosaccharides were then degraded by removal of maltose units from the non-

reducing end.

A pullulan hydrolase from B. stearothermophilus KP1064 rapidly cleaved a-

and B-CD, a-limit dextrins, amylose, and the maltotriose analogue, p-phenyl-a-D-

maltoside [92]. Amylopectin, starch, and B-limit dextrins were hydrolyzed

significantly slower. Maltose was the main product from these substrates, while

pullulan was slowly cleaved to mainly panose. The ability to split CD rings and

cleave pullulan indicated that hydrolytic action could be of the endo type. The

enzyme had a native molecular weight of 115,000 and consisted of two identical

subunits. The enzyme had an isoelectric point of 4.4, pH optimum of 5.8,

temperature optimum of 550C, and was thermostable at 650C for ten minutes.

Sulfhydryl reagents (p-chloromercuribenzoate, 5,5'-dithio-bis(2-nitrobenzoate)

strongly inhibited activity, which indicated that cysteine was required for catalysis.









42

The authors suggested assignment of the pullulan hydrolase to a unique type of

maltogenic a-amylase.

The intracellular cyclomaltodextrinase was purified from an alkalophilic

Bacillus species that was identified as closely relating B. circulans [110]. The

enzyme had an isoelectric point of 4.2, a native molecular weight of 126,000, and

consisted of two subunits of 67,000. The pH and temperature optima were 6.0 and

500C, respectively, but heating at 600C for 10 min led to inactivation. The enzyme

rapidly hydrolyzed a-CD, maltotriose, and maltotetraose, while B- and --CD,

maltopentaose, maltohexaose, and maltoheptaose were cleaved 2-3 times slower.

For all substrates, maltose was the main product. Maltose and starch were not

degraded. Thiol reagents inhibited the enzyme, which suggested that sulfhydryl

groups may exist in the active site.

The purified maltogenic a-amylase from Bacillus thermoamyloliquefaciens KP

1071 had a native and subunit molecular weight of 115,000 and 67,000, respectively

[94]. The pi, pH optimum, and temperature optimum were 4.7, 6.2, and 630C,

respectively. Maltogenic a-amylase was thermostable for 10 minutes at 650C,

pH 6.8, in the presence of 5 mM EDTA. However, a decrease in thermostability

resulted when EDTA was absent or calcium was present. The enzyme rapidly

hydrolyzed a- and B-CD, maltotriose, maltotetraose, maltopentaose, p-phenyl-a-

maltoside, a-limit dextrin, and short chained amylose, while amylopectin, starch, and

glycogen were slowly degraded. Maltose was the major product for all substrates,









43

except maltotriose. Amylopectin was degraded in exo-fashion by preferential

cleavage of maltose units from the nonreducing ends, and hydrolysis of it's a-1,6-

branch points. While the enzyme slowly cleaved pullulan at the a-1,4-bonds to give

mainly panose, small amounts of glucose and maltose indicated the a-1,6-bonds

were cleaved at a lower frequency. Although activity was completely inhibited by

p-chloromercuribenzoate, the enzyme did not contain a cysteine or a cystine

residue. Comparison of amino acid compositions indicated the maltogenic

a-amylase to be homologous to the B. stearothermophilus KP 1064 maltogenic

a-amylase [92]. The authors suggest the B. thermoamyloliquefaciens to be the first

maltogenic exo-acting a-amylase able to cleave a-1,6-bonds in amylopectin.

The thermostable cyclodextrinase from Clostridium thermohydrosulfuricum 39E

[73] is the first and, thus far, only CDase to be isolated from a thermophilic

anaerobe. The enzyme had optimal activity at pH 6.0 and 650C, and had a half-

life of three hours at 650C. a-CD was rapidly hydrolyzed, while B-CD, starch, and

amylose were degraded at rates 67%, 50%, and 53%, respectively, to that of

a-CD. p-Chloromercuribenzoate inhibited activity, which suggested sulfhydryl groups

are involved in activity.

An intracellular amylase from Pseudomonas MSI had a molecular weight of

96,000, and optimal activity at pH 5.5 and 500C [29]. The enzyme rapidly

hydrolyzed a-, B-, and -y-CD, linear maltodextrins, and amylose, while amylopectin,

B-limit dextrin, and glycogen were hydrolyzed approximately 20 times slower. The









44

amylase had endo-type activity in which equimolar amounts of glucose and maltose

were produced as final products for all substrates.

A CD-degrading, cell-bound glucoamylase was isolated from a Flavobacterium

species [6]. a-, 1-, and 7-CD, maltotriose, and amylose were rapidly degraded, with

the final product exclusively glucose. Amylopectin and glycogen were poor

substrates. The enzyme had a pH optimum of 5.5-6.5, required calcium for activity,

and was inactive at 550C. The enzyme resembled human intestinal glucoamylase.

Although there is a limited amount of published research on microbial

cyclomaltodextrinases, the characterization of the CDases from the various Bacillus

species indicates several common features shared among the enzymes: (a) They

are dimers that have identical subunits. (b) They rapidly cleave CDs and linear

malto-oligosaccharides at varying rates, while starch polymers are degraded much

slower. (c) They produce mainly maltose with small quantities of glucose as the

final degradation products, by apparent exo-cleavage of maltose units from the

nonreducing ends. (d) In most cases, thiol reagents inhibit activity. Due to the

high relatedness among Bacillus species, it is very probable that other species of

bacilli also produce their own forms of CDase. To this date, there are no reports

of purifications and characterizations of a CDase from either B. subtilis or

B. caldolyticus. Consequently, the information presented in the following chapters

should be considered new and unique with respect to bacterial CDases.















CHAPTER 3
PURIFICATION, CHARACTERIZATION, AND COMPARISON
OF THE CYCLOMALTODEXTRINASE FROM B. SUBTILIS 25S,
B. CALDOLYTICUS C2, AND B. SUBTILIS HIGH TEMPERATURE
GROWTH TRANSFORMANT H-17


Enzymology is currently experiencing unprecedented growth and expansion

in the development of thermostable industrial enzymes, particularly amylolytic

enzymes for commercial starch processing [100]. Comparisons between

thermolabile and thermostable enzymes may help to elucidate the molecular basis

of enzyme thermostability, thus facilitating future protein engineering. The ideal

enzymes for comparison should be derived from the same genus, and have similar

physical, chemical, and structural properties but differ only in thermostability.

B. subtilis high temperature growth (HTG) transformants were previously

generated by transformation of mesophilic, amylolytic B. subtilis 25S with DNA from

the obligate thermophile B. caldolyticus C2 [39]. This chapter describes the

purification, characterization, and comparison of a p-nitrophenyl-a-D-maltoside-

hydrolyzing cyclomaltodextrinase (CDase) from B. subtilis 25S, B. caldolyticus C2,

and B. subtilis HTG transformant H-17.









46

Materials and Methods



Organisms and Growth Conditions

The methodology for generation of B. subtilis 25S, B. caldolyticus C2, and

B. subtilis H-17 has been described previously [38,39]. All strains were grown

aerobically, to late log phase, in 4 L batches (250 rpm, airflow 0.5 L/min) at 37C

(25S) and 600C (C2, H-17) in a Queue Systems Mouse Fermenter (Parkersburg,

WV). The growth medium consisted of 0.5% maltose (autoclaved separately), 0.5%

starch, 2.5% Bacto peptone, 0.3% Bacto yeast extract, 0.2% Bacto meat extract,

0.3% dipotassium hydrogen phosphate, and 0.1% potassium dihydrogen phosphate,

pH 7.2.



Enzyme Purification

Four liters of 25S, C2, or H-17 cells were harvested by centrifugation at

9,000 x g in a Sorvall RC-5B GSA rotor, at 40C for 20 min. The pellet was

suspended in 20 mM, pH 8.0 potassium phosphate buffer, and lysozyme was added

to a final concentration of 100 /g/ml. The suspension was stirred at 370C for 3 h,

after which the cell debris was removed by centrifugation at 27,000 x g in a Sorvall

RC-5B SS-34 rotor, at 40C for 15 min. Nucleic acids were precipitated by slow

addition of a neutralized 10% (wt/vol) solution of streptomycin sulfate (1/10 volume

of the supernatant) [74], and after 1 h of stirring at 40C, the suspension was









47

centrifuged at 27,000 x g at 40C for 15 min. Solid ammonium sulfate was added

to the supernatant and the 40-80% fraction was retained, and redissolved in

20 mM, pH 7.0 potassium phosphate buffer containing 0.05% 2-mercaptoethanol

(2-Me) (buffer A). Phenylmethylsulfonyl fluoride was added to a final

concentration of 2mM, and the solution was dialyzed for 24 h against 4 x 2 L

volumes of buffer A. Precipitated protein was removed from the dialysate by

centrifugation at 27,000 x g at 40C for 15 min. The dialysate was applied to a

2.5 x 35 cm DEAE-cellulose (Pharmacia, Uppsala, Sweden) column previously

equilibrated with buffer A, after which the column was washed with 500 ml of

buffer A at 30 ml/h (250C) to remove unretained protein. The enzyme was then

eluted at 15 ml/h with a 1 L linear 0-0.25 M, pH 7.0 NaCI gradient, in buffer A.

Peak active fractions were pooled, concentrated (to approximately 10 ml) by

ultrafiltration with a YM30 Amicon membrane, and applied to a 2.75 x 75 cm

Sephadex G-75 (Pharmacia) column previously equilibrated with buffer A. The

enzyme was eluted at a rate of 10 ml/h (25C) with buffer A. Peak active fractions

were pooled, concentrated (to approximately 6 ml) by ultrafiltration with a YM30

membrane, and then equilibrated with 5 mM, pH 7.0 phosphate buffer containing

0.05% (2-Me) (buffer B) by three ultra-filtrations with the same YM30 membrane.

The retentate was applied to a 2.5 x 20 cm hydroxyapatite (HA-Ultrogel) column

previously equilibrated with buffer B. The column was then washed with 200 ml

of buffer B at 16 ml/h (250C) to remove unretained protein. The enzyme eluted









48

immediately behind the void volume which eliminated the need for a phosphate

gradient. Peak active fractions were pooled, concentrated (to approximately 6 ml)

by ultrafiltration with a YM30 membrane, and applied to a 2.75 x 50 cm Sephacryl

S-200 (Pharmacia) column previously equilibrated with buffer A. The enzyme was

eluted at a rate of 12 ml/h (250C) with buffer A, and peak active fractions were

pooled and concentrated (to approximately 2 ml) by ultrafiltration with an Amicon

Centriprep-30. The enzyme sample was sterile-filtered, and stored aseptically with

0.05% 2-Me or 0.01 mM EDTA.



Enzyme Assay

Cyclomaltodextrinase (CDase) activity was determined by the release of

p-nitrophenol from p-nitrophenyl-a-D-maltoside (PNM) using a Beckman DU-7

spectrophotometer with an electrically-heated (pelltier) temperature control

(0-99C +/- 0.02"C). The standard reaction mixture (1.0 ml) contained 33.3 mM,

potassium phosphate buffer, 1.98 /mol PNM, and 0-0.1 ml of enzyme preparation.

25S and H-17 CDase were assayed at pH 7.5 in the presence of 0.02% 2-Me, while

C2 CDase was assayed at pH 7.0 in the presence of 0.01 mM EDTA. The reaction

mixture was incubated at 350C, 600C, or 650C for the 25S, C2, or H-17 enzyme,

respectively. The complete mixture (0.9 ml) without substrate was allowed to

thermally equilibrate for 1 min after which 0.1 ml of warmed substrate was added.

The increase in absorbance/min at 400 nm was automatically calculated at 25 sec









49

intervals for approximately 5 min. A molar extinction coefficient of 9,600

M' cm' was used to calculate the amount of p-nitrophenol released/min [93]. One

unit of enzyme activity was defined as the amount of enzyme required for the

release of 1 jmol of p-nitrophenol/ ml/min. Protein was estimated by the method

of Lowry et al. [45], with bovine serum albumin as the standard.

To perform activity stains on polyacrylamide and isoelectric focusing gels, the

gel was soaked in 2 x 200 ml volumes of 100 mM, pH 7.5 potassium phosphate

buffer for 10 min, and then incubated at 350C or 600C for 10-20 min in 33 mM, pH

7.5 potassium phosphate buffer containing 3 mg/ml PNM and 0.02% 2-Me. Bands

that displayed enzymatic activity stained yellow, while the gel remained clear. The

gel was then rinsed in deionized water and stained for protein with Coomassie

Brilliant Blue R-250 and Amidoschwartz 10B.



Relative Molecular Mass (Mr)

The Mr of the enzymes was determined by polyacrylamide gel electrophoresis

(PAGE) with and without sodium dodecyl sulfate (SDS) and 2-Me, using Bio-Rad

Mini-Protean II slab gels and the buffer system of Laemmli [35]. Spacing and

separating gels were 4.0% and 7.5% polyacrylamide, respectively. Molecular weight

standards (26,600-180,000) were obtained from Sigma Chemical Co., St. Louis, Mo.









50

Gels were stained with 0.15% Coomassie Brilliant Blue R-250 and 0.05%

Amidoschwartz 10B in. 40% ethanol/10% glacial acetic acid for 2 h.



Isoelectric Point (pl)

The pi values were determined according to manufacture's instructions, using

Isogel Agarose IEF Plates (FMC BioProducts) over a pH range of 3-10. IEF

standards (pI 3.55-9.9) were purchased from Sigma. Isoelectric focusing was

performed with a Hoeffer Isobox cooling chamber at 120C. After staining for

activity (see Enzyme Assay), the gel was then rinsed in deionized water and stained

for protein with Coomassie Brilliant Blue R-250 as described by the manufacturer.



Effect of Temperature and pH

Temperature profiles for the 25S, C2, and H-17 enzymes were determined

under standard conditions over the range 10-52C (25S) and 25-800C (C2, H-17).

The pH profiles were determined under standard assay conditions in 33 mM

phosphate, over the pH range 4-10 for the 25S, C2, and H-17 enzymes. Initial

velocities within the first 1.5 min were used to calculate relative activities (% of

maximum).











Effect of Chemicals

To determine the effect of various cations, all solutions used were prepared

from triple-deionized water. The 25S, C2, or H-17 enzyme was dialyzed against

2 x 0.5 L volumes of pH 7.5, 7.0, or 7.5, respectively, 50 mM sodium barbital/ 10

mM EDTA/0.05% 2-Me, for 8 h at 40C. The enzymes were then dialyzed against

3 x 0.5 L volumes of the same buffer without EDTA, for 16 h at 40C. Activity was

determined under standard conditions, except that the assay buffer was 50 mM

sodium barbital/0.02% 2-Me, and contained a final concentration of 5 mM cation

(chloride form). The reaction mixture without substrate was pre-incubated with the

selected cation for 10 min at 350C (25S) or 45C (C2, H-17). Residual activity was

then determined at standard incubation temperatures. The effect of 2-Me, EDTA

or Tris-HCl on 25S, C2, and H-17 activity was examined under standard assay

conditions. The final concentration ranges were 0.005-0.5% (2-Me), 0.005-1.0 mM

(EDTA) or 5mM (Tris HC1).



Thermal Stability

The effect of 2-Me (0.005-0.5%) or EDTA (0.005-1.0 mM) on thermostability

was examined. Under standard assay conditions, the enzyme was incubated in

closed cuvettes for 1 h, at a single temperature maintained by the

spectrophotometer heating unit. Incubation temperatures ranged from 35-45C

(25S) and 60-75C (C2, H-17). Assays were performed at 5-15 min intervals during









52

the incubation period, by the addition of PNM. Initial velocities were calculated

and related to those at time zero.



K Value Determination

The initial rates of hydrolysis of PNM were performed under standard assay

conditions. The Km values were determined by plotting 1/V vs 1/S according to the

Lineweaver-Burk method [42].



Amino Acid Analysis

Amino acid analyses were performed by the Protein Chemistry Core Facility

in the Department of Biochemistry, University of Florida. Protein samples were

prepared by hydrolysis in 6N HC1, or DMSO/HCI hydrolysis for cysteic acid

determination. Amino acid analyses were performed in duplicate on a Beckman

6300 instrument, using a cationic exchange resin and ninhydrin-based quantification.

Standards were run before and after each analysis, and internal standards were

included for every analysis.













Results and Discussion

The purifications of the 25S, C2, and H-17 CDase are summarized in Tables

1-3, respectively. Initial studies indicated that the 25S enzyme required a minimum

of 0.02% 2-Me in order to maintain activity, while the C2 or H-17 enzyme required

a minimum of 0.02% 2-Me or 0.01 mM EDTA for thermostability. Therefore,

0.05% 2-Me was included in all column chromatography buffers.

All three enzymes migrated as single bands during SDS-PAGE (Figures

1-3). The Mr of the 25S, C2, and H-17 CDase was approximately 55,000, 60,000,

and 55,000 as determined by SDS-PAGE, and, 110,000, 120,000, and 110,000 as

determined by PAGE, respectively. This indicates that the native structure of all

three enzymes is a dimer composed of two subunits of equivalent Mr. Table 4

summarizes the biochemical and biophysical characterization of the 25S, C2, and

H-17 CDase. All three enzymes had the same pi (Figure 4), and similar pH

optima (Figure 5), Km values for PNM, and Tris-HCl inhibition. Neither enzyme

had a cation requirement for activity or thermostability, while each enzyme

exhibited similar cation inhibition (Table 5). However, the 25S, C2, and H-17

CDase exhibited strikingly different thermostabilities. The purified 25S enzyme

showed optimal activity between 35-370C (Figure 6), and complete inactivation after

incubation at 450C for 10 min, when assayed at pH 7.5 (Figure 7). Although 0.02%
















Purification of B. subtilis 25S cyclomaltodextrinase.


Enzyme Activity


Purification
Step


Total
Protein
(mg)


Total
Activity
(Units)


Specific
Activity
(U/mg Prot.)


Purification
(Fold)


Cell-free extract

40-80% (NH4)2SO,
fraction

DEAE-cellulose
chromatography

Sephadex G-75
chromatography

Hydroxyapatite
chromatography

Sephacryl S-200
chromatography


650


217


28.5


15.4


2.5


42.0


28.6


0.065


0.12


0.34


0.44


2.52


100


38.8


154 12


Table 1.


Yield
(%)


4.9 10.0
















Purification of B. caldolyticus C2 cyclomaltodextrinase.


Enzyme Activity


Purification
Step


Total
Protein
(mg)


Total
Activity
(Units)


Specific
Activity
(U/mg Prot.)


Purification
(Fold)


Cell-free extract

40-80% (NH,)2SO,
fraction

DEAE-cellulose
chromatography

Sephadex G-75
chromatography

Hydroxyapatite
chromatography

Sephacryl S-200
chromatography


28.0


15.0


22.1


11.4


2.0


0.051


0.09


0.37


0.70


3.7


100


14.0


74.0


120 19


Table 2.


Yield
(%)


5.3 6.0
















Purification of B. subtilis H-17 cyclomaltodextrinase.


Enzyme Activity


Purification
Step


Total
Protein
(mg)


Total
Activity
(Units)


Specific
Activity
(U/mg Prot.)


Purification
(Fold)


Cell-free extract

40-80% (NH,)ZSO,
fraction

DEAE-cellulose
chromatography

Sephadex G-75
chromatography

Hydroxyapatite
chromatography

Sephacryl S-200
chromatography


520


136


34.0


14.4


31.7


15.6


0.061


0.11


0.27


0.53


1.8


100


29.5


103 18


Table 3.


Yield
(%)


5.7 6.3












































Figure 1. SDS-PAGE of purified B. subtilis 25S cyclomaltodextrinase on a 7.5%
polyacrylamide gel. Lane 1: SDS molecular weight standards (1)
a,-macroglobulin; (2) 13-galactosidase; (3) fructose-6-phosphate kinase;
(4) pyruvate kinase; (5) fumarase; (6) lactate dehydrogenase. Lane
2: 10-20 Ag of purified enzyme.













































Figure 2. SDS-PAGE of purified B. caldolyticus C2 cyclomaltodextrinase on a
7.5% polyacrylamide gel. Lane 1: SDS molecular weight standards
(1) a2-macroglobulin; (2) 1-galactosidase; (3) fructose-6-phosphate
kinase; (4) pyruvate kinase; (5) fumarase. Lane 2: 10-20 ug of
purified enzyme.













































Figure 3. SDS-PAGE of purified B. subtilis H-17 cyclomaltodextrinase on a 7.5%
polyacrylamide gel. Lane 1: SDS molecular weight standards (1) az-
macroglobulin; (2) B-galactosidase; (3) fructose-6-phosphate kinase;
(4) pyruvate kinase; (5) fumarase; (6) lactate dehydrogenase; (7)
triosephosphate isomerase. Lane 2: 10-20 ig of purified enzyme.
















Table 4.


Biochemical and biophysical
B. caldolyticus C2, and B. subtilis


Temperature Optimum

% Activity Remaining After 1 hr at 65C

pH Optimum

Isoelectric Point

Km for PNM (mM)

Relative Molecular Mass (Mr) (SDS-PAGE)

0.02% 2-Mercaptoethanol Requirement

0.005 mM EDTA Requirement

Inhibition by 5 mM Tris


comparison of B. subtilis
H-17 cyclomaltodextrinase.

H-17 25S

65-680C 350C

100 0 (

7.5 7.5

4.8 4.8 4

1.46 2.96 1

55,000 55,000 60,

+ +

+


25S,


C2

iO0C

56

..0

.8

.31

00

+

+


+ +









61










1 2 "






JJ-1













Figure 4. Isoelectric focusing gel of purified B. subtilis 25S, B. caldolyticus C2,
or B. subtilis H-17 cyclomaltodextrinase. Lane 1: pi standards (1)
amyloglucosidase; (2) ovalbumin; (3) carbonic anhydrase; (4) horse
myoglobin (major and minor); (5) cytochrome C. Lane 2: 10-20 Ag
purified enzyme.





























Figure 5. Effect of pH on the activity of (A) B. subtilis 25S cyclo-
maltodextrinase, (B) B. caldolyticus C2 cyclomaltodextrinase, and (C)
B. subtilis H-17 cyclomaltodextrinase. The pH profiles were
determined under standard assay conditions in 33 mM phosphate, over
the pH range 4-10 for each enzyme. Initial velocities within the first
1.5 min were used to calculate relative activities (% of maximum).





















100





S70



50





20

10


3 4 5 6 7 8 9 10 11 12
pH


0 i I I a I I 1 I
3 4 3 7 G l 10 11 12
pH


100

C




6 30 C
so.









7 20

10
0

3 4 S 6 7 8 9 10 11 12
pH


100



2 80

so
E


S60



340
0I 30

S20

10
















Cationic inhibition of B. subtilis 25S, B. caldolyticus C2, and B. subtilis
H-17 cyclomaltodextrinase.


(5 mM final conc.)


MgClz
CaC1,
SrC1,
BaCl
MnC1,
CoC1,
ZnC1,

LiCI
NaCI
KC1
CsCI


% Inhibition 25S


69
90
78
87
100
100
100

14
0
0
0


% Inhibition H-17


42
72
38
50
100
100
82

18
0
0
0


% Inhibition C2


41
66
47
40
100
100
100

23
0
0
0


Table 5.

































Figure 6. Effect of temperature on the activity of (A) B. subtilis 25S
cyclomaltodextrinase, (B) B. caldolyticus C2 cyclomaltodextrinase, and
(C) B. subtilis H-17 cyclomaltodextrinase. Temperature profiles were
determined under standard conditions over the range 10-520C (25S)
and 25-800C (C2, H-17).

























so.
E

570
`F60
E so

40

i 30-

20
10


0 10 20 30 40 5S
Tempwratur (Deg. C)


0.

0

0




;o


10
10


20 30 40 50 60 70 B0 90
Temperature (Do. C)


0 1 1 1 I I I I
20 30 40 50 60 70 80 91
Temperature (Deg. C)


























Figure 7. (A) Effect of temperature and incubation time on the thermostability
of B. subtilis 25S cyclomaltodextrinase. (B) Effect of temperature,
incubation time, and 0.02% 2-Me or 0.01 mM EDTA on the
thermostability of B. caldolyticus C2 cyclomaltodextrinase. (C) Effect
of temperature, incubation time, and 0.02% 2-Me on the
thermostability of the B. subtilis H-17 cyclomaltodextrinase. (D) Effect
of temperature, incubation time, and 0.005 mM EDTA on the
thermostability of B. subtilis H-17 cyclomaltodextrinase. Under
standard assay conditions, the enzyme was incubated in closed cuvettes
for 1 h, at a single temperature maintained by the spectrophotometer
heating unit. Incubation temperatures ranged from 35-450C (25S) and
60-75C (C2, H-17). Assays were performed at 5-15 min intervals
during the incubation period, by the addition of PNM. Initial
velocities within the first 0.5 min were calculated and related to those
at time zero.











68








A B
to* -- -- -- -- 100l1 -- --- -- --
35 C 60 C0 W-A/2--M






s so 5
1070 S 70

5so so.
a
\S6 C: I-ME
40 40o

30 u 30 7o Q WTA
< 20 20 -
S10 43 c 40 C 10
S 4 C 0 Ci 2-MI-
c I I AI II 0
0 15 30 43 6o 0 15 30 45 60
Time minutesm) Time (minute.)



C D



?. \', C
0 065 c


70 a c g 70,
60 C G0

SS 507
lt4 40c


20 20 68 C
72 C
S10 10 70 C
8 ~ -- 75 C
1 o o ._ I ___ I ___ I I II
0 15 30 45 60 0 15 30 45 60
Time minutes ) Tim. (minutes)












2-Me was required to maintain 25S enzyme activity, neither 2-Me or EDTA, at all

levels tested, enhanced the thermostability of 25S enzyme.

This contrasts with the purified C2 enzyme which showed optimal activity at

600C (Figure 6), and retained 100% of initial activity after incubation at 600C for

2 h (Figure 7), when assayed at pH 7.0. However, a minimum concentration of

0.02% 2-Me or 0.01 mM EDTA was required for thermostability of C2 CDase,

although EDTA more effectively stabilized the C2 enzyme, than did 2-Me.

Furthermore, the H-17 enzyme showed optimal activity between 65-680C (Figure 6),

and retained 100% of initial activity after incubation at 650C for 1 h (Figure 7),

when assayed at pH 7.5. A minimum concentration of 0.02% 2-Me or 0.005 mM

EDTA was required for thermostability of the H-17 enzyme. In contrast with the

C2 enzyme, 2-Me more effectively stabilized the H-17 enzyme, than did EDTA.

For both the C2 and H-17 CDase, higher levels of either compound did not

enhance thermostability, and combinations of 2-Me and EDTA did not produce a

synergistic effect whereby thermostability was increased.

The maintenance of 25S enzyme activity by 2-Me probably occurs by

reduction of an oxidized sulfhydryl residue(s) (cysteine) at the active site, or by

reduction of a disulfide bond that forms at the active site due to oxidation.

However, the mechanism by which 2-Me confers C2 and H-17 enzyme

thermostability is unknown. The low level of 2-Me may be sufficient to reduce

oxidized sulfhydryl groups to allow them to undergo disulfide bond formation.









70

Another, but less likely, mechanism could be 2-Me's intramolecular bifunctional

hydrogen bonding with polar side groups. However, the use of an extrinsic factor

such as 2-Me for stabilization may simply reflect the nature of the reducing

atmosphere and presence of other sulfhydryl-containing molecules inside the cell

which protect cysteine sulfhydryl groups. It is possible that EDTA confers H-17

enzyme thermostability by either forming an ion pair bridge between its negatively

charged carboxyl groups and positively charged amino side groups, or by chelating

cations that interfere with intramolecular ion-pair formation between charged amino

acid side groups.

Table 6 shows the amino acid compositions of the 25S, C2, and H-17 CDase.

The most significant difference between the 25S and C2 enzyme is the increase in

the hydrophobic residues alanine, leucine, and proline. One of the most important

mechanisms that confers enzyme thermostability is hydrophobic interaction [3,50,99].

Our results concur with a previous observation which reported a strong correlation

between an increase in alanine, leucine and particularly proline, and the rise in

thermostability of five Bacillus oligo-1,6-glucosidases [95]. Proline, due to

its a-helix breaking ability, can contribute to improved turn stabilization and a

greater potential for close packing of protein domains. Thus, by increasing the

number of proline residues in conjunction with other hydrophobic residues such as

leucine and alanine, thermostability may be enhanced by strengthened hydrophobic

interactions, thereby tightening the molecule as a whole.
















Table 6. Amino acid compositions of B. subtilis 25S, B. caldolyticus C2,
and B. subtilis H-17 cyclomaltodextrinase.


25S


H-17


Amino Acid Mol%


Cys
Asx
Thr
Ser
Glx
Pro
Gly
Ala
Val
Met
Ile
Leu
Tyr
Phe
His
Lys
Arg


Res/Subunit Mol%


1.51
11.94
3.82
4.20
10.16
3.79
18.31
6.58
5.32
0.22
4.68
7.13
3.34
5.63
3.12
5.02
5.23


1.60
7.30
3.27
4.33
9.66
4.53
22.65
8.75
5.96
0.23
4.15
7.59
2.19
3.41
2.55
6.58
5.36


Res/Subunit Mol% Res/Subunit


2.05
9.36
5.00
4.72
11.05
4.86
9.63
12.02
5.68
0.15
4.75
9.73
3.28
4.11
2.72
5.55
5.34









72

Similarly, the H-17 enzyme shows a significant decrease in

(aspartate/asparagine), tyrosine and phenylalanine residues, accompanied by

significant increases in glycine, alanine, proline and lysine residues, when compared

to the thermolabile 25S enzyme. The increase in alanine, glycine and proline may

increase the H-17 enzyme's internal hydrophobicity, which in turn might influence

the enzyme's thermostability. Interestingly, although there were changes in the

amino acid composition between the two enzymes, there was no change in the pi

value. For the H-17 enzyme, the increase in lysine residues may be offset by the

decrease in aspartate/asparagine, especially if the decrease was in asparagine

residues.

The results indicate that protein thermostability may be achieved via a

unique approach in which the entire cell of a mesophile is converted into that of

a thermophile. Previous studies have proposed that transformation to thermophily

involves a limited number of genes which exert wide pleiomorphic effects [39].

These effects manifest themselves in the production of thermostable cellular

components and products, possibly by specific mistranslation of messenger RNA

[39,40]. Thus far, none of the enzymes purified from the B. subtilis 25S HTG

transformants have been shown to be products of specific thermophilic genes

transferred from donor to recipient during the transformation of 25S to thermophily

[13]. Because the donor, B. caldolyticus C2, also produces a thermostable CDase,

it was necessary to exclude the possibility that the thermostable H-17 enzyme was









73

actually encoded by a cointegrated gene from B. caldolyticus C2. This enzyme has

a different Mr, pH optimum, temperature optimum, and effect of 2-Me or EDTA

on thermostability, when compared to the H-17 enzyme. The C2 enzyme appears

to achieve thermostability by a mechanism different than H-17. The H-17 enzyme

may, in part, derive its greatest thermostability through disulfide bond formation,

whereas C2 may, in part, derive its greatest thermostability by undergoing an

altered conformation in the presence of EDTA. It is therefore believed that the

H-17 and C2 enzymes are not products of the same gene, and that the

B. caldolyticus C2 enzyme is not produced by the HTG transformant H-17.

In conclusion, amino acid sequence analyses of several mesophilic enzymes

and their thermophilic counterparts have suggested that a few specific amino acid

substitutions at critical regions may account for very subtle structural and

conformational differences, that lead to large alterations in thermostability

[32,50,99,115]. Cloning and nucleic acid sequencing of the 25S, C2, and H-17

CDase genes will allow determination of the primary amino acid sequences. This

will aid in elucidating the structural mechanisms for increased thermostability in

these enzymes, which is a fundamental requirement for future protein engineering.















CHAPTER 4
SUBSTRATE SPECIFITIES, AFFINITIES, AND
CLEAVAGE PATTERNS OF THE CYCLOMALTODEXTRINASE
FROM B. SUBTILIS 25S, B. CALDOLYTICUS C2,
AND THE B. SUBTILIS HIGH TEMPERATURE
GROWTH TRANSFORMANT H-17



Cyclomaltodextrinase (EC 3.2.1.54) (CDase) hydrolyses cyclodextrins (CDs)

and linear maltodextrins much more rapidly than starch and other related polymers

[30,31,53]. This chapter describes how the 25S, C2, and H-17 enzymes described

in Chapter 1, were further characterized by their classification as CDases.

Hydrolytic activity was also determined to indicate whether the H-17 CDase has

commercial application as a saccharifying and/or debranching enzyme.



Materials and Methods



Organism and Growth Conditions

B. subtilis 25S, B. caldolyticus C2, and B. subtilis H-17 were grown aerobically

to late log phase as described previously.









75

Enzyme Purification and Standard Assay

The 25S, C2 and H-17 enzymes were purified as described previously. To

eliminate any possible low levels of contaminating enzyme activity during substrate

hydrolysis assays, the 25S, C2, and H-17 enzymes were further purified by

preparative native PAGE. Two Bio-Rad Mini-Protean II gels were stained for

activity, as described previously, and the active band was precisely cut from each

gel with a razor blade. The polyacrylamide strips were pulverized in 2-3 ml of

standard assay buffer in a hand-held, glass tissue homogenizer. The polyacrylamide

suspension was centrifuged at 6,000 x g for 5 min at 250C, and the supernatant

removed. The polyacrylamide was washed with 2 ml of the same buffer and

recentrifuged. The supernatants were combined, sterile-filtered, and concentrated

and equilibrated in standard assay buffer by ultrafiltration with an Amicon

Centricon-30. Enzyme activity was determined by following the release of

p-nitrophenol from p-nitrophenyl-a-D-maltoside under previously described standard

assay conditions.



Substrate Specificity

Enzyme activity was determined for p-nitrophenyl-a-D-glucopyranoside

(PNG), maltose (G2), maltotriose (G3), maltotetraose (G4), maltopentaose (G5),

maltohexaose (G6), maltoheptaose (G7), isomaltose (IG2), isomaltotriose (IG3),

panose (Pan), a-cyclodextrin (a-CD), B-cyclodextrin B-(CD), pullulan, soluble potato









76

starch, potato amylose, and potato amylopectin. All malto-oligosaccharide and CD

solutions were sterile-filtered and stored at 40C. All substrates were purchased

from Sigma, and all substrate solutions were screened for purity using thin-layer

chromatography.

Hydrolysis assays employed 10 mM malto-oligosaccharides, 10 mM CDs, or

2% polysaccharides. The reaction mixture (500 /l) contained 470 p1l of substrate

in 0.02% 2-mercaptoethanol, 33 mM phosphate buffer (pH 7.5), plus 30 tl

(0.05-0.1 units) of purified enzyme. Samples were incubated at 37C, 600C, or 65C

for 25S, C2, or H-17, respectively. The reaction was terminated at various intervals

(5-120 min) by boiling 100 pj aliquots of the mixture for 3 min. and then assayed

for glucose and reducing power. Reducing sugars were determined by the

dinitrosalicylic acid method [7], and glucose was determined with a peroxidase/

glucose oxidase colorimetric diagnostic kit (Sigma No. 510-A). Initial rates of

hydrolysis were used to determine relative substrate specificities. Enzyme activity

for PNG was determined under standard assay conditions.



Km and VmaxDetermination

Initial velocities were determined under the above assay conditions and were

expressed as /moles of reducing groups/min/mg protein, with glucose as the

standard. Protein was determined by the method of Lowry et al. [45] with bovine

serum albumin as the standard. Km and Vmax values were determined according









77

to the Lineweaver-Burk method [42]. Soluble potato starch and potato amylopectin

concentrations ranged over 0.2-4.0%, while maltoheptaose and a-cyclodextrin ranged

over 0.2-10 mM.



Thin-Layer Chromatography (TLC)

The products of malto-oligosaccharide, CD, and polysaccharide hydrolysis

were analyzed by TLC using precoated Silica Gel G Redi/Plates (Fisher Scientific,

Orlando). The plates were developed by a single ascent at 250C, with the solvent

system n-butanol-ethanol-water (3:4:1.25, vol/vol) in a pre-equilibrated 10 x 25 x 29

cm developing tank. Dried plates were stained for carbohydrate by spraying with

a 6.5 mM solution of N-(1-naphthyl) ethylenediamine dihydrochloride in methanol

containing 3% sulfuric acid [8], followed by heating at 1100C for 15 min. Reference

standards were commercial malto-oligosaccharides obtained from Sigma.



Results and Discussion

The 25S, C2, or H-17 CDase hydrolyzed p-nitrophenol-a-D-maltoside in a

linear fashion, regardless of enzyme or substrate concentration. This suggested that

maltose and p-nitrophenol were released from PNM. If glucose and PNG were

initially released from PNM, then the assay would show a delayed color formation

due to the eventual hydrolysis of PNG. Neither enzyme was active against

p-nitrophenyl-a-D-glucopyranoside, maltose, isomaltose, isomaltotriose, or panose.









78

Table 7 shows the experimental initial relative rates of hydrolysis of linear

malto-oligosaccharides, a- and B-CD, and polysaccharides, as determined by

reducing power and glucose liberation. The 25S, C2, and H-17 enzymes

demonstrated similar substrate specificities as the greatest increase in reducing

power was produced with maltohexaose, maltoheptaose, and a- and B- CD, while

starch, amylose, amylopectin, and pullulan were hydrolyzed significantly less (some

data for C2 not shown). For either enzyme, hydrolysis of G3 produced the greatest

quantity of glucose, while G7 hydrolysis produced the greatest increase in reducing

power. It appeared that the enzyme preferred to sequentially cleave a maltose unit

from the nonreducing end of the oligosaccharide, since less glucose was produced

from the longer-chained substrates. However, there was, in part, random endo-

activity since G4 and G6 hydrolysis also produced glucose, instead of only G2.

Figure 8 shows the TLC analyses of the 10 min, 1 h, and 3 h products of linear

malto-oligosaccharide hydrolysis by either the 25S, C2, or H-17 enzyme. The TLC

results confirmed that the endo-hydrolysis by either enzyme was also, in part,

nonspecific, since G4 was a low level intermediate produced from both G5 and G7

hydrolysis. Furthermore, in addition to G2 and G4, G3 was an intermediate

produced from G6 hydrolysis. The subsequent hydrolysis of the G3 intermediate

would account for the production of glucose from G6 cleavage. The enzymes also

exhibited slight transglucosylase activity as seen with G4 hydrolysis.
















Table 7. Experimental relative rates of hydrolysis of malto-oligosaccharides,
cyclodextrins, and polysaccharides by B. subtilis 25S, B. caldolyticus C2,
and B. subtilis H-17 cyclomaltodextrinase.

(expressed as % maximum)


Substrate


Glucose Assay

25S H-17 C2


G2
G3
G4
G5
G6
G7
IG2, IG3, Panose
a-Cyclodextrin
B-Cyclodextrin
Starch
Amylopectin
Amylose
Pullulan


0
100
11
45
29
32
0


Reducing Sugar Assay

25S H-17 C2


0
100
36
52
39
45
0


0 0
20 16
35 43
67 63
81 68
100 100
0 0
85 90
95 70
10 11
22 25
12 10
7 7


0
24
45
85
98
100
0













2 -M +



0 C 00





to
UU
.8 J0








re
U U U
oo *t


















0r +
0 1 0






cU-o
.^ on s



03v















8

**-*~


I-81* u rrcr U


.* .. ... .





.r



0 4


,Cem e in r
090099 IS'9









82

The 25S and H-17 Km and Vmax values for maltoheptaose, a-CD, starch, and

amylopectin are presented in Table 8. Both enzymes showed greater affinities for

maltoheptaose and a-CD than for starch and amylopectin. Consequently, the 25S

and H-17 enzymes were classified as cyclomaltodextrinases, based on substrate

specificity and affinity. It may be inferred that the C2 enzyme is also a

cyclomaltodextrinase based upon the enzyme's similar characteristics. One may

note that the H-17 CDase attacked a-CD faster than B-CD, while the 25S CDase

cleaved B-CD faster than a-CD. Perhaps the increase in the H-17 enzyme's

internal hydrophobicity over its 25S counterpart tightens the molecule as a whole

through strengthened hydrophobic interaction. While this may enhance

thermostability, the larger ring structure of B-CD may become less accessible than

a-CD to the active site of the H-17 enzyme.

Figure 9 shows the TLC analysis of the 20 min hydrolysis products of a- and

B-CD, and 1.5 h cleavage products of amylose, amylopectin, and starch by either

the 25S or H-17 enzyme. TLC indicates that the enzymes initially open the ring

of a- or B-CD forming a linear molecule with the corresponding number of glucose

units. Degradation of the linear molecule to shorter malto-oligosaccharides follows,

with the accumulation of maltose as the final main product. However, it is unclear

whether the hydrolysis of CDs by 25S or H-17 CDase proceeds via a multiple

attack, or a multiple site attack model [20]. The enzymes appear to exhibit a

preferred attack model [20] for starch, amylopectin, and amylose, in which maltose
















Table 8. Km and Vmax values of B. subtilis 25S and B. subtilis H-17
cyclomaltodextrinase.



25S H-17

Substrate Km Vmaxa Km Vmaa


Maltoheptaose
a-Cyclodextrin
Starch
Amylopectin


a


1.1 mg/ml (0.96 mM)
1.1 mg/ml (1.10 mM)
20 mg/ml
20 mg/ml


1.0 mg/ml (0.91 mM)
0.9 mg/ml (0.93 mM)
22 mg/ml
20 mg/ml


moles ot reducing groups/mn/mg ot protein


I


















Figure 9. TLC of 20 min hydrolysis products of CDs and 1.5 h hydrolysis
products of polysaccharides by B. subtilis 25S or B. subtilis H-17
cyclomaltodextrinase. Lane 1: reference standards G1 (glucose), G2
(maltose), G3 (maltotriose), G4 (maltotetraose), G5 (maltopentaose),
G6 (maltohexaose). Lane 2: a-cyclodextrin + CDase. Lane 3: 13-
cyclodextrin + CDase. Lane 4: amylose + CDase. Lane 5:
amylopectin + CDase. Lane 6: starch + CDase. Lane 7: reference
standards G1 and IG2 (isomaltose).


Figure 10.


TLC of products of pullulan hydrolysis by B. subtilis 25S and B. subtilis
H-17 cyclomaltodextrinase. Lane 1: Reference standards G1
(glucose), G2 (maltose), and G3 (maltotriose). Lane 2: Pullulan (no
enzyme). Lanes 3 and 4: 30 min and 2 h hydrolysis of pullulan by
25S CDase, respectively. Lanes 5 and 6: 30 min and 2 h hydrolysis
of pullulan by H-17 CDase, respectively. Lane 7: Reference standards
IG2 (isomaltose) and PAN (panose).








85






8j t T H
62j

















I .3.m
1 2 3 4.. .









a:a.."

It r f

K";...f..

.: : E.. .. .
...... ,


d n


7 1










86

is the main end product. TLC indicates these polysaccharides were degraded in

exo-fashion by preferential cleavage of maltose units from non-reducing ends,

although small amounts of glucose were detectable.

Both 25S and H-17 CDase's ability to cleave a-1,4 bonds of pullulan, together

with their rapid attack of a- and 1-CD rings indicate that the enzymes do not

necessarily require a non-reducing terminal end and that their action can be of the

endo-type. Although the hydrolysis of pullulan was slow, it's 1,6-branch structure

suggests that the presence of amylopectin's 1,6-branch points, may be a factor in

both 25S and H-17 CDase's slightly faster cleavage of amylopectin over starch or

amylose. Figure 10 shows TLC analysis of the 30 min and 2 h products of pullulan

hydrolysis by 25S and H-17 CDase. It appears both enzymes produced glucose,

maltose, and panose from pullulan. Identical cleavage patterns were observed for

the C2 CDase. Traditional pullulanases hydrolyze only the a-1,6-glycosidic linkages

of pullulan, releasing maltotriose. Consequently, the 25S, H-17, and C2 CDases

demonstrated either isopullulanase hydrolysiss of a-1,4 linkages to produce

isopanose), or neopullulanase hydrolysiss of a-1,4 linkages to produce panose)

activity [72].

The C2 and H-17 thermostable CDase are very similar to the

B. thermoamyloliquefaciens KP 1071 thermostable a-amylase II [94], and the

B. stearothermophilus KP 1064 thermostable maltogenic a-amylase [92] in that they

all, (a) behave as low active, exo-acting, maltogenic enzymes against starch, amylose,