xx... k .
X; 3w H,
_.

r‘vwxv . huh?

3
:73 . ..
‘ 14.; .1

, us,
1 r3511” 5. x
5.: $.22 ....,
5 am 1
.1
.23.“...

if. v.
<

5534.».

,3!{ n. I.
, 2,)

 

 

 

.rwww‘ A

7435!!!

 

.3 :n
55‘.”

a
. 5:6- :1. , .- ..
1917...?lz’r2-ka‘ .\ o , A .

abfihflfl. .wqxr‘, 5H1." ‘ mummmuw

.. 3
#1.”;
.. .1136!!le33. r:
23,..lfuiaiu Jar...

\ .13.

a i. ,

.
a

15:3; .1.
. a #4le
. I". (I

1.

1...!
.3

 

.. I i .J u. A
cum 1‘ tilt/lily.“
V , «Lani; .fi . «ls...
. , 1 .0‘ :4, .. ‘ V
1.3:} 1&3: ., é” .
2:3... . .
. ‘. “Khan...me
If 9...! t; i u.

4.). 1. 25..

» I11}- .0 ’1

fivmmmfln

 

 

 

 

 

 

 

 

 

Iiiiiiiiiiiiii‘"W

r ‘/ . ' \31293 01564 330

 

 

i \\\\\\\\\\\\1\\\\i

 

LIBRARY

Michigan State
University

 

 

 

This is to certify that the

dissertation entitled
Molecular Biochemistry of Thermoanaerobacter ethanolicus 39E
Amylopullulanase: Analysis of Substrate Cleavage
Specificity and Thermophilicity by Site-Directed and
Deletion Mutagenesis

 

presented by
Cynthia Ann Hollenbeck Petersen

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Biochemistry

 

 

M 4M2”

Major ofessor

Date //,/}dI/fé

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

 

PLACE ll RETURN BOX to romovo this checkout from your roootd.
TO AVOID FINES man on or before data duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU loAn Afflnnotlvo Action/Equal Opportunity Institution
may.”

MOLECULAR BIOCHEMISTRY OF Thermoanaerobacter ethanolicus 39E
AMYLOPULLULANASE: ANALYSIS OF SUBSTRATE CLEAVAGE
SPECIFICITY AND 'IHERMOPHIUCITY BY SITE-DIRECTED AND DELETION
MUTAGENESIS
By

Cynthia Ann Hollenbeck Petersen

A DISSERTATION
Submitted to
Michigan State University ‘
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Biochemistry

1996

ABSTRACT

MOLECULAR BIOCHEMISTRY OF Thermocnaerobacter ethanolicus 39E
AMYLOPULLULANASE: ANALYSIS OF SUBSTRATE CLEAVAGE
SPECIFICITY AND THERMOPHILICITY BY SITE-DIRECTED AND

DELETION MUTAGENESIS

by

Cynthia Ann Hollenbeck Petersen

The need to identify starch-hydrolyzing enzymes with increased thermal
stability has led researchers to screen thermophilic bacteria for enzymes with
these characterisfics. Amylopullulanase activity has been detected in various
thermophiles and hyperthermophiles. The enzyme's thermostability and ability
to attack both alpha-1,4 and alpha-1,6 glucosidic linkages found in starch may
lead to improvements in bioprocessing industrial starches. It has been
previously shown that four highly conserved regions exist in the active center of
various amylolytic enzymes. The active center of amylopullulanase from
Themaanaerobacter ethanolicus 39E was analyzed by means of site-directed
mutagenesis. A loss of catalytic activity was observed when each of the three
conserved catalytic residues in regions 11, III and IV was mutated. A change in
hydrolysis pattern was evident when mutating three of the four conserved
substrate binding residues in regions L II, and IV. Transglycosylation activity
was not detectable in the wild type or mutant constructs of amylopullulanase.
This is the first report of altered binding characteristics with amylopullulanase.
Native amylopullulanase and deletion mutants lacking the N or C-terminus have
been expressed in E. coli and analyzed for possible roles in enzyme
thermostability and thermophilicity. A putative thermophilicity region (TPR)

has been identified at the N-terminal end of the protein which is important for
maintaining the optimal activity of the enzyme. The TPR is shown to control the
temperature activity optimum without changing enzyme thermostability. The
Arrhenius plot for the wild type enzyme was linear, unlike the discontinuous
plot observed for the ApuN 324 deletion construct. The biphasic Arrhenius plot
appears to be due to a structural change in ApuN324 that does not alter the
binding characteristics of the enzyme relative to wild type. To our knowledge
this is the first report to show that enzyme stability and activity are controlled by
separate protein features. It is hypothesized that the flexibility of the mutant
protein was increased to account for a 20°C lower temperature optimum because
the Kmapp value was not altered relative to the wild type enzyme.

To my Grandma and Grandpa Hollenbeck

ACKNOWLEDGEMENTS

I owe many thanks to those who have helped me succeed in the
complefion of this dissertation. First, is my advisor, Dr. Greg Zeikus, who
accepted me into his lab at a time when I was searching for appreciation. He
believed in me every step of the way, was a great mentor, and helped complete
my development as a scientist. I never thought I would find the area of
biotechnology so interesu'ng, but the enthusiasm Dr. Zeikus had for the lab and
the research was contagious. I would also like to thank the members of my
guidance committee; Drs. Zach Burton, Lee Kroos, Loren Snyder and John Wang
for their continued support and encouragement over the years. My appreciation
also extends to everyone in the Macromolecular Structure Facility, the Computer
Facility, the administrative and secretarial staff for all of their help and friendly
smiles.

Graduate school did not only provide me with scientific development, but
rewarded me with many friendships that I will charish over my lifetime. I
especially thank Marty (Dr. Marty Regier) who is a friend like no other. Her
support and encouragement, which has carried me to the end, has been undying
and unconditional. I am glad that our 'post-graduate‘ endeavors have left us in
close proximity to continue many life experiences including Friday morning
coffee/ journal club. This brings me to Carla Margulies who, although sometimes
late, has also been involved in the coffee/ journal club experience. Her love for
science and worldly experiences is like no other I have seen. I was glad to have
had the opportunity to experience a worldly endeavor with Carla, even though
the hills were definitely much bigger than in Michigan! I have enjoyed listening

to Carla's views and opinions that have created great debate with many,
including two of my colleagues and friends Kevin Carr and Mark Sutton. I
would like to thank Kevin for always being there 24/ 7. Whether it was a trip to
The Peanut Barrel or computer assistance, I could always count on Kevin. Mark
taught me the importance of constructing a well balanced lunch (corn dogs, a
brownie and beer) and that the best football team is the Chicago Bears.
I also thank the Zeilcus lab members, past and present, for their friendship
and support over the years: Drs. Doug Burdette, Guoqiang Dong,
Iong-Hyun Park, Vladimir Tchemajenko, Claire Vieille, Mariet van der Werf, and
almost doctors Eric and Greg Zeikus and last but not least Maris Laivenieks.
Finally, my appreciation for the support and encouragement I received
from each and every member of my family, my husband, and Duke.

The potential of the Big M has yet to be utilized.

TABLE OF CONTENTS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

LIST OF TABLES - _ ix
LIST OF FIGURES ..... - ............ - . ....... - - . -- 3.xi
LIST OF ABBREVIATIONS - . ......... -..-xii
OVERVIEW ....................................................................................................................... 1
CHAPTER 1
Literature Review ............................................................................................................. 3
INTRODUCTION
Enzymes Involved 1n Conversion of Polysaccharides to Sugars - 4
Industrial Starch Bioprocessing - ....... -- 5
TI-IERMOPI-IILES,TI-IERMOZY1\4ES AND AMYLOSACCHARIDASES
General Description and Identification of Classes A 8
Amylosaccharidases - - _ - 10
Alpha-amylase ..... 21
Pullulanase 24
Amylopullulanase ................... 28
Neopullulanase - 30
ence Comparison 31
DISSERTATION OBIECTIVES AND SIGNIFICANCE - 33
REFERENCES - - ...... ............. . .......... 34
CHAPTER II
Biochemical and Enzymatic Characterization of Catalytic Activity and
alpha-1, 4 / alpha-1, 6 Cleavage Specificity of Recombinant
Amylopullulanase fiomThermoanaerobacter ethanolicus 39E
Abstract ...... - - -_ - -- -- -- -- _- _- __ _ - 44
Introduction -- ..... - - - 45
Experimental procedures -- 48
- - -- - .................... 55
Discussion - ........ ...... .......... . ..................... 81
Acknowledgements - - ........ 86
References ........... -- - . ........ 87

 

Vii

CHAPTER III
Molecular Analysis of Thermophilicity and Thermostability of Recombinant
Amylopullulanase fiom Thermoanaerobacter ethanolicus 39E.

 

 

 

 

 

 

 

 

 

 

 

Abstract _ ........ - - 90
Introduction 91
Experimentalprocedures - -- - - ---93
- - - - - - 99

Discussion - - 112
Acknowledgments ..... - - - 115
References - 116

CHAPTER IV

Conclusions and Directions for Future Research . - -_ - _- - 120

APPENDD( A

Construction and Characterization of Thennoanaerobacter ethanolicus 39E

Amylopullulanase Proline Deficient Mutants - 125

APPENDD( B

Attempts at Overexpression of Recombinant Amylopullulanase from

Thermoanaerobacter ethanolicus 39E - 144

APPENDD( C

Transglycosylation Activity Analysis of Recombinant Mutant and Wild Type

Amylopullulanase from Thermoamerobacter ethanolicus 39E - _ _ - ....... 153

viii

LIST OF TABLES

 

 

 

CHAPTER 1
Table 1 -Starch conversion process producing glucose and fructose ....................... 7
Table 2- Amylosaccharidases from thermophilic and hyperthermophilic
organisms ......... . - -- - 11
Table 3 -Amylosaccharidases: Cleavage action pattern and product
formafion - -- 13
Table 4 -General properties of alpha-amylases from bacteria and fungi ............... 23
Table 5 -General properties of bacterial pullulanases-- -- 27

Table 6 -General properties of bacterial enzymes with alpha-1, 4 and
alpha-1, 6 cleavage activity (Neopullulanases and

 

 

amylopullulanases) - - -- 29
CHAPTER II
Table 1 -Oligonucleotides used in this study - . - ........ - ..... 51
Table 2 Consensus sequences in the alpha-amylase family 57

 

Table 3 -Amylopullulanase mutants generated by site-directed
mutagenesis - -- ........ - 61

 

Table 4 -'I"1me course of reaction products produced from wild type
amylopullulanase on various low and high MW
oligosaccharides - -- 64

 

Table 5 Comparison of end products from enzymatic action of
His493G1n, Asn6OOSer/ Glu601Val, and wild type
enzymes on low and high MW oligosaccharides - 66

 

Table 6 -Kinetic analysis of wild type and mutant amylopullulanase on various
substratesa= .... 71

 

Table 7 -Activity comparisons of Asp597Leu, G1u626Asp, and Asp703Gly' ......... 73

Table 8 -Activity comparisons of His493Gln, Asn60(Ber/ G1u601Val,
His702Arg, His702Asp, and I-Iis702Lys -- - A -- - -- 76

 

 

Table 9 -Activity comparisons of Asp672Asn and Asp672Glu - ..... - 80
ix

Table 10 Summary of amino acid alteration in the four conserved regions
with amylopullulanase, neopullulanase, and alpha-amylase ......... 84

 

 

 

 

CHAPTER III

Table 1 -Kinetic analysis of wild type and ApuN 324 amylopullulanase on
pullulan at 40°C and 60°C. 108

Table 2 -Activation energy calculations for Arrhenius plots of wild type and
ApuN 324 amylopullulanase ......... _ 111

APPENDD( A

Table 1 -Amino acid alignment of the TPR of several fliermophilic
amylopullulanases - - - - ..... - . ..... 128

Table 2 -Oligonucleotides used in this study - -- _ - 130

Table 3 -Proline profiles of pullulan-digesting enzymes - - 140

 

LIST OF FIGURES

CHAPTER I
Figure 1 -Enzymatic hydrolysis of pullulan by enzymes with
alpha-1,4, alpha-1,6, or alpha-1,4 and alpha-1,6

 

 

hydrolytic activity - - - 15
Figure 2 -Proposed action pattern of neopullulanase from
B. steamthermophilus on pullulan - 18

Figure 3 -Proposed model of hydrolysis and transfer reactions of
alpha-amylase from B. licheniformis on pullulan and

 

 

 

 

 

maltooligosaccharides - - - -- - 20
Figure 4 -Proposed substrate binding and catalytic sites for Taka

amylase-A from A. oryzae - 26
CHAPTER II
Figure 1 -Physical map of the pUC18 clone (prAP164-UC) containing the

apu gene ...... - -- -- -50
Figure 2 -HPIC profile showing standard mixture of G1-G7 - -60
Figure 3 -Proposed action pattern of amylopullulanase -- - - ......... 68
CHAPTER III

Figure 1 -Effect of N and C-terminus deletions of the amylopullulanase gene
from Hermannaerobacter ethanolicus on optimum temperature for

 

enzyme activity - - 96
Figure 2 Optimal temperature profiles of wild type amylopullulanase and
deletion construct ApuN 324 - - 100

 

Figure 3 -Thermostability profiles of wild type and ApuN 324 deletion
construct of amylopullulanase after enzyme pre-incubation
at 85°C for 0-90 minutes 102

 

Figure 4 -Arrhenius plots for the recombinant T. ethanolicus 39E wild type
and ApuN 324 deletion construct of amylopullulanase between
25°C and 90°C ...... 107

 

APPENDD( A
Figure 1 - HCA comparison of amylopullulanase from T. ethanolicus 39E,
T. thermosulfurigenes EMl, and T. saccharolyticum B6A-RI ............... 133

Figure 2 -Thermostability and optimum temperature analysis after single
substitution of proline residues 1-5 from T. ethanolicus 39E with
the corresponding residues of the less thermophilic enzyme
from T. themosulfurigenes EMl -- -- _ ......................... 135

 

Figure 3 -Thermostability and optimum temperature profiles of double
proline mutants (Pro2105er/ Pr0213Gln and Pro24011e/
Pro244Leu) relative to wild type amylopullulanase - - 137

 

Adh

BSA
BLMA

G1-G7

ABBREVIATIONS

alcohol dehydrogenase

amylopullulanase

bovine serum albumin

Bacillus lidtenifonm's maltogenic alpha-amylase

base pair

glucose, maltose, maltoniose, maltotetraose, maltoheptaose
maltohexaose, and maltoheptaose

guanidine hydrochloride

hydrophobic cluster analysis

high fructose corn syrup

high performance ion chromatography

isopropyl fi-D-flfiosalacmpyranoside

molecular weight

pulsed amperometric detector

polyacrylamide gel electrophoresis

Polymerase chain reaction

Sodium dodecyl sulfate

Terrific broth

thermophilicity related region

OVERVIEW

This dissertation is divided into four chapters: a literature review, a
chapter on substrate cleavage characteristics and characterization of the dual
specificity (both alpha-1,4 and alpha-1,6 cleavage activity) of the enzyme by site-
directed mutagenesis, a chapter describing thermophilicity and identification of a
determinant responsible for maintenance of the enzyme's temperature activity
optimum by deletion mutagenesis, and a summary of the research and future

Chapter I reviews the literature on saccharolytic enzymes with an
emphasis on those enzymes isolated from thermophilic organisms. The
biochemical properties of the enzymes and sequence conservation as it relates to
substrate cleavage specificity will be examined, as well as thermophilic
organisms and their industrial application.

Chapter II, "Biochemical and Enzymatic Characterization of Catalytic
Activity and alpha-1,4/ alpha-1,6 Cleavage Specificity of Recombinant
Amylopullulanase from Thermoanaerobacter ethanolicus 39E," examines mutant
constructs created by single point mutations to study substrate cleavage
specificity and activity. Product analyses by HPIC are described that show the
catalytic specificity of this enzyme hydrolyzing both alpha-1,4 and alpha-1,6
linkages on branched and unbranched starch, related polysaccharides and linear
oligosaccharides.

Chapter 111," Molecular analysis of thermophilicity and thermostability of
recombinant amylopullulanase from Themoanaerobacter ethanolicus 39E,"
describes experiments on a series of nested deletion mutants of the
amylopullulanase gene. Gene products were isolated that possessed both

2

alpha-1,4 and alpha-1,6 activities. A mutant was identified having a lower
optimal temperature range relative to wild type and other deletion constructs.
The mutant also displays a discontinuous Arrhenius plot indicative of a
conformational change required for optimal activity. The final chapter
summarizes the findings of this study, conclusions, and directions for future
research.

CHAPTERI
LITERATURE REVIEW

I. Introduction

 

Commonly used enzymes in starch processing are alpha-amylase, beta-
amylase, glucoamylase, cyclodextrin glucotransferase and debranching enzymes.
These are collectively referred to as amylosaccharidases due to their ability to
cleave sugar units. Two general types of activities present in these enzymes are
alpha-1,4 and alpha-1,6 hydrolysis, and alpha-1,4 and alpha-1,6 transfer
reactions. The hydrolysis reaction results in the cleavage of glucosidic bonds in a
polysaccharide producing linear or branched low molecular weight
oligosaccharides. The transfer, or transglycosylation, reaction is the addition of
glucose to glucose or maltose produced from the hydrolysis reaction producing
small branched oligosaccharides like isopanose, isomaltose and panose.
Alpha-amylases cleave alpha-1,4 glucosidic linkages in an endo, or
random, fashion and are used as liquefying agents in the starch-processing
industry. Alpha-amylases have been identified in various bacteria, fungi, plants,
and animals. It is the only amylosaccharidase whose three-dimensional structure
(from Aspergillus oryzae and swine) has been determined (Matsuura et al., 1984
and Buisson et al., 1987). Alpha-amylase from A. oryzaz is widely used in starch
processing due to its thermal (90°C) and pH (up to pH 11) stability (Saito, 1973).
Beta-amylase, like alpha-amylase, cleaves alpha-1,4 glucosidic linkages. Its
pattern of action is an exotype, acting from the nonreducing end of starch,
producing maltose which is used in industry as a sweetner. Beta-amylases are
produced by plants and some bacteria. Thermophilic beta-amylase from
Clastridium thernwsulfurogenes cloned into Bacillus breois may be useful for starch
processes due to its activity and stability at high temperature (Mizukami et al.,
1992). Glucoamylase has been identified in yeast and fungi and functions as an
exotype enzyme removing glucose in an ordered fashion from the nonreducing

5

end of starch. In industry, this enzyme functions after liquifaction by alpha-
amylase and in conjuction with debranching enzyme to produce glucose from
starch. The enzyme from Aspergillus a'wamori (Nunberg et al., 1984) is most
widely used in industrial processes. The enzyme cyclodextrin glucotransferase
(CGTases) catalyzes both transglycosylation and hydrolysis reactions on starch to
produce alpha, beta, and gamma cyclodextrins. CGTases have been identified in
Bacillus species and Klebsiella pneumoniae (Fogarty et al., 1980). Debranching
enzymes include, but are not limited to, isoamylase, pullulanase,
amylopullulanase, and neopullulanase. All of these enzymes are capable of
cleaving alpha-1,6 glucosidic linkages as well as other hydrolytic and
transferring activities. This group of enzymes is present in higher plants and
microorganisms with isoamylase and pullulanase currently the most industrially
used. Isoamylase in combination with glucoamylase, and pullulanase in
conjunction with beta-amylase function to produce glucose and maltose,
respectively, from starch after action by alpha-amylase.

A more detailed review of amylosaccharidases whose substrate cleavage
specificity has been studied will be presented later as these enzymes are relevent
to this thesis project and the potential design of altered enzymes for use in
industrial starch degradation.

E SI 1 B. .

In the United States, 18 billion pounds of sweetners are produced per year
demonstrating the importance of the industrial starch conversion process
(Hebeda, 1987). Starch, a branched polymer containing alpha-1,4 and alpha-1,6
glucosidic linkages is used to manufacture sugars in a mulfistep process utilizing
different amylosaccharidases. Three classes of enzymes are involved in the
production of sugars from starch: 1) endo-amylase (alpha-amylase),

6

2) exo-amylases (beta-amylase, glucoamylase); and, 3) debranching enzymes
(pullulanase, isoamylase) (Ramesh et al., 1992). Starch bioprocessing normally
involves two steps, liquifaction and saccharification, which run at high
temperatures. Liquefaction is a process in which starch granules are gelatinized
in aqueous solution and partially hydrolyzed at alpha-1,4 branch points.
Saccharification results in conversion of liquefied starch to low molecular weight
saccharides by various debranching enzymes. Alpha-amylase is used in the
liquefaction process which occurs at pH 6-7 and 80°C-150°C over a period of up
to 3 hours. The pH is then adjusted to 4.0-4.5 and the temperature is lowered to
55-60°C for the saccharification step which occurs in the presence of
glucoamylase and the proper debranching enzyme over 24-90 hours

(Saha et al., 1989). Production of high fructose syrup is then achieved by the use
of glum isomerase (Table 1). Amylosaccharidases from thermoanaerobes such
as amylopullulanase may be suitable for application in starch conversion
biotechnologies because of their novel activity, extreme thermostability,
thermoactivity, and acidic pH compatibility. Industry is seeking to design
engineered enzymes with specificity for alpha-1,4 glucosidic linkages only,
specificity for alpha-1,6 glucosidic linkages only, and with various end-product
specificities. This would improve the starch degradation process by
development of an acid thermostable debranching enzyme or use it directly to

make conversion syrup for fermentation.

Table 1. Starch conversion process producing glucose and fructose
(Adapted from Zeikus, LG, 1990)

 

 

Substrate/
Product and
Process Stage Enzyme pH Temp. Metal
ST ARCH
Liquefaction Alpha- 6.0-7.0 80-120°C Ca'H’
l amylase
MALTODEXTRIN
(DE. 10-15)
[pH adjustment
with acid]
¢ Saccharifaction Glucoamylase 4.0-5.0 55-60°C
and
. Isoamylase
GLUCOSE
[filter,pH
adjustment,
addition of metal
cofactor]
Isomerization Glucose 7.0 58-60°C Mg-H'
l isomerase Mn”
Co++
GLUCOSE and

FRUCI‘OSE
mixture

 

 

8

Identification or design of a thermostable amylopullulanase, an
amylosaccharidase with alpha-1,4 and alpha-1,6 hydrolytic activity, would allow
a shift to a higher temperature in the saccharification process which would have
many industrial benefits: a) increased reaction rates with decreased operation
time, b) lower costs for enzyme purification, c) higher substrate concentrations
facilitating enzymatic degradation due to increased starch solubility, and
d) minimal risks of bacterial contamination (Antranikian, 1990).

II. Thermophiles and Thermozymes

Over 100 enzymes and proteins have, to date, been purified to
homogeneity from thermophilic micro-organisms. The discovery that many
enzymes from thermophilic organisms (thermozymes) have higher
thermostability than homologous proteins from mesophiles has inspired a search
for the molecular basis of this higher thermostability. Thermostability, measured
by the protein's ability to resist irreversible thermal inactivation, and is expressed
as the enzyme's half-life at a given temperature (Vieille et al , 1996). Enzyme
thermophilicity is the temperature at which the enzyme has its highest activity
(V ieille et al, 1996).

Thermophilic organisms can be classified as moderate thermophiles with a
growth optima of 60-80°C and hyperthermophiles with a growth optima above
80°C. Habitats suitable for growth of moderate thermophiles are widespread.
They include geothermally heated springs, ground water and sea water as well
as marine thermal vents and solar heated soils (Lowe et al., 1993). Environments
that exist at temperatures appropriate for the growth of hyperthermophiles are
less common. The most extremely thermophilic organisms have been isolated
from continental volcanic areas (Brock et a1 ., 1972).

9

Initial comparison of mesophilic and thermophilic proteins has identified
features unique to thermophilic proteins. It has been observed that thermophilic
enzymes are smaller, have less ordered structure, contain more hydrophobic
interactions, and less beta-structure, among other features (Amelunxen and
Murdock, 1978). However, it appears that the key to thermostability will be
determined by comparing thermodynamic properties, amino acid composition
and sequence of homologous mesophilic and thermophilic proteins; not by their
molecular architecture (Sundaram 1986). Site directed mutagenesis has become a
useful tool in the analysis of protein thermostabilization. Small changes in the
stabilizing forces caused by only one or two amino acid changes can raise the
relative stability of an enzyme by several degrees centigrade
(Coolbear et al., 1992). Other factors leading to increased thermostability include:
increased protein rigidity at mesophilic temperatures, location of proline
residues in the loop regions of thermophilic proteins, and interaction of the
protein with its surroundings (Vihinen, 1987 and Watanabe et al., 1991).

Enzymes from hyperthermophiles have a higher optimum temperature
(more thermophilic) and increased thermostability in comparison to moderate
thermophiles or mesophiles (growth optimum of 30-45°C), which is an
important rationale for use of thermostable enzymes in industrial starch
processing applications. A number of extremely thermostable enzymes of
potential industrial utility have been purified and/ or cloned from anaerobic
thermophiles. A number of hyperthermophiles are characterized by their ability
to ufilize complex saccharides which are metabolized to meet carbon and energy

requirements.

10

III. Amylosaccharidases

Amylosaccharidases have been identified in various mesophilic,
thermophilic and hyperthermophilic bacteria (T able 2), and Archae. They are
classified based on their enzymatic activity, substrate specificity, and products of
hydrolysis (Table 3). The two main catagories are endo-acting enzymes and exo-
acting enzymes which cleave internally and from the end of the polysaccharide,
respectively. Endo-acting amylosaccharidases produce a mixture of malto-
oligosaccharidases due to their random cleavage pattern. The exo-acting
enzymes cleave sugar units from the non-reducing end to the reducing end of the
substrate in an ordered, processive fashion (Fogarty et aL, 1979).

Amylosaccharidases are broadly grouped into amylases and pullulanases.
Pullulanases can be separated into four groups: pullulanase, isopullulanase,
amylopullulanase, and neopullulanase, depending on the cleavage products
produced from action on pullulan (Figure 1). Enzymes in this class have been
purified and characterized from a wide range of bacterial species. Pullulanase
from Thermoanaerobium Tok6-B1 (Plant et al., 1987), and amylopullulanase from
Closh‘idium thermohydrosulfuricum (Mathupala et al., 1990) hydrolyze the alpha-1,6
linkages in pullulan to produce maltotriose. Isopullulanase from Aspergillus niger
(Sakano et al., 1971) hydrolyzes alpha-1,4 linkages producing isopanose.
Neopullulanase from Bacillus stearothermophilus (Kuriki et al., 1988a) produces
panose by cleavage of alpha-1,4 linkages and a small amount of glucose and
maltose by the limited alpha-1,6 cleavage activity on pullulan.
Amylopullulanases and neopullulanases are also capable of cleaving alpha-1,4
linkages in starch producing low molecular weight oligosaccharides.
Isopullulanase and pullulanase can not hydrolyze starch.

11

Table 2. Amylosaccharidases from thermophilic and

 

 

hyperthermophilic organisms
Enzyme Organism Refemis)
WW

THERMOPHILES

a-Amylase Bacillus caldavelox (90°C) Bealin-Kelly et al., 1991

a—Amylase Bacillus stearothermaplu'lus (90°C) Brosnan et al., 1991
Gray et al., 1986

o-Amylase (AmyA) Dictyoglomus thermoaphflus (90°C) Fukusumi et al., 1988

a—Amylase (AmyB)
a-Amylase (AmyC)
B-Amylase
Amylopullulanase
Amylopullulanase
Amylopullulanase
Amylopullulanase
Amylopullulanase
Cyclodextrinase
Cyclodextrin
glycosyltransferase
a—Glucosidase
a—Glucosidase
B-Glucosidase
B-Glucosidase
Neopullulanase
Pullulanase
Pullulanase

Pullulanase

 

Dictyoglomus thermaphilum (80°C)
Dictyoglonrus thermophilum (70°C)

Wobacterium
thermosulfurigenes 4B (75°C)
Closiridium thamohydrostdfrm'cum E101
(85°C)
Thennoanaerobacter ethanolicus 39E (85°C)

Horinouchi et al., 1988
Horinouchi at al., 1988
Hyun and Zeikus 1985
Melasniemi 1987

Mathupala and Zeikus 1993

'I'hermoanaerobacterium sp. Tok6-Bl (80°C)I Plant at al., 1987

Thmnoanaerobacterium
saccharolyticum (75°C)

T. thermosulfiaigcnes EMl (70-75°C)

T. ethanolicus 39E (65°C)

T. thermosulfurigenes EMl (90-95°C)

Bacillus Sp- (75°C)

T. ethanolicus 39E (75°C)

Clostridium saccharolyticus

C. thermoellum

B. stearothennophilus (60-65°C)

B. stearathennophilus (65°C)

Baallus sp. (75°C)

Thermus sp. AMD-33 (70°C)

 

Saha at al., 1990
Wind at al., 1990

Saba and Zeikus 1990
Wind at al., 1990
Nakao at al., 1994
Saba and Zeikus 1991
Coolbear et al., 1992
Margaritis and Merchant 1986
Kuriki et al., 1988
Kuriki et al., 1988
Shen et al., 1990

Aubert et al., 1993
Nashihara et al., 1988

 

 

12

 

 

HYPERTHERMOPHILE
a-Amylase
(extracellular)
a-Amylase
(intracellular)
(Jr-Amylase
a-Amylase
B—Amylase
Amylopullulanase
Amylopullulanase
Amylopullulanase
B-Glucosidase

a-Glucosidase

 

Pyromesfiuiosus (100°C)
p. furiosus (100°C)

p. woesi (100°C)

Thermococcus profundus (80°C)
1. maritime (95°C)

9. furiosus (125°C)

1154 (no-125°C)

r. litoralis

Tlmmotoga 8p.

p. furiosus (110°C)

 

Koch et al., 1990

Lade et a1.., 1993

Koch et al., 1991

Chung et al., 1995
Schumann et al., 1991

Brown and Kelly 1993
Schuliger at al., 1993

Brown and Kelly 1993
Ruttersmith and Daniel 1993

Constantino at al., 1990

 

 

13

Table 3. Amylosaccharidases: Cleavage action pattern and product formation.

 

 

Enema:
alpha-amylase
beta-amylase
amyloglucosidase
alpha-glucosidase
pullulanase

isoamylase

cyclodextrin glucosyl
transferase

isopullulanase
oligo-1,6 glucosidase

 

endo
exo
exo
exo

endo

endo

endo

endo
exo
endo

endo

 

W
maltosaccharides, maltose
maltose

glucose

glucose

maltotriose from pullulan

maltodextrin from starch
and glycogen

maltosaccharides,
cyclodextrins

isopanose from pullulan
glucose from maltose
panose from pullulan

maltose, maltotriose, and

maltotetrose from starch

 

W
alpha—1,4

alpha-1,4
alpha-1,4] alpha-1,6
alpha-1,4/ alpha-1,6
alpha-1,6

alpha-1,6

alpha-1,4

alpha-1,4
alpha-1,6
alpha-1,4/alpha-1,6
alpha-LU alpha-1,6

 

 

14

Figure 1 Enzymatic hydrolysis of pullulan by enzymes having alpha-
1,4, alpha-1,6, or alpha-1,4 and alpha-1,6 hydrolytic activity. Open circles
and slashed circles denote non-reducing glucopyranosyl residues and
reducing glucose residues, respectively. Horizontal lines and vertical lines
connecting circles indicate alpha-1,4 and alpha-1,6 glucosidic linkages

respectively.

15

Figure 1. Enzymatic hydrolysis of pullulan by enzymes with alpha-1,4, alpha-
1,6, or alpha-1,4 and alpha-1,6 hydrolytic activity.

 

Pullulan

Q

Glucose

O-O-O

Maltotriose

0%

lsopanose

Panose

Enzyme

Glucoamylase

Pullulanase
and
Amylopullulanase

Isopullulanase

Neopullulanase

l6

Classically, amylases are separated into three groups: alpha—amylases,
beta-amylases, and glucoamylases and catalyze the cleavage of starch, glycogen,
and related glucans. Alpha-amylase is an endo-acting enzyme responsible for
the random cleavage of alpha-1,4 glucosidic linkages, whereas, beta-amylases
and glucoamylases are exo-acting enzymes that hydrolyze alpha-1,4 linkages, or
alpha-1,4 and alpha-1,6 linkages, respectively. The most widely studied alpha-
amylase is from Aspergillus oryzae (T aka amylase-A) (Matsuura et al., 1983).

Cleavage action pattern has been most extensively studied in
neopullulanase from B. stearothermophilus (Kuriki et al., 1991; Imanaka et al., 1989;
and Takata et al., 1989), alpha-amylase from B. licheniformis (Kim et al., 1992;

Kim et al., 1994; and Lee ct al., 1995) (Figure 2,3), and amylase-pullulanase from
Bacillus circulans F-2 (Kim et al., 1995). Cleavage activity of amylopullulanase
from T. ethanolicus 39E will be investigated in this dissertation and compared to
the cleavage patterns seen with the above enzymes.

In contrast to hydrolysis (break down) reactions, transglycosylation (build
up) reactions have been identified and extensively studied in neopullulanase.
This enzyme can catalyze the transglycosylation reaction to form alpha-1,4 and
alpha-1,6 glucosidic linkages producing highly branched oligosaccharides
(Kuriki et al., 1991 and Takata et al., 1992). Hydrolysis and branching activities
are present in one active center of neopullulanase as evidenced by results of site-
directed mutagenesis of conserved amino acids in regions I-IV
(Kuriki ct a1 ., 1991). The presence of transglycosylation activity will be examined
with amylopullulanase. ’

17

Figure 2 Proposed action pattern of neopullulanase. Open circles and
slashed circles denote non-reducing glucopyranosyl residues and
reducing glucose residues, respectively. Horizontal lines and verfical lines
connecting circles indicate alpha-1,4 and alpha-1,6 glucosidic linkages
respectively.

18

Figure 2. Proposed action pattern of neopullulanase from
B. stearothermophilus on pullulan (Adapted from Iminaka, I. and T. Kuriki

1989).

 

19

Figure 3. Proposed model of hydrolysis and transfer reactions of alpha-
amylase from B. licheniformis on pullulan and maltooligosaccharide
(Adapted from Lee, SJ. et al., 1995). (A) Alpha-amylase hydrolyzed
pullulan producing panose which, in the presence of excess glucose,
results in the formation of an alpha-1,6 linkage producing
61-O-a-(62-O-a-glucosyl-maltosyl)-glucose. (B) Hydrolysis of
maltooligosaccharides produces mainly maltose (path a) but also to
glucose and maltohiose (paths b and c). These products undergo transfer
reactions using additional glucose, maltose and maltotriose to produce
isomaltose, panose, isopanose, 610-01-maltosyl-maltose, 63-O-cr-maltosyl-
maltotriose, and 62-0-Q-maltotriosyl-maltose. Longer branched products,
if produced, are likely to be hydrolyzed by the enzyme. Open circles and
slashed circles denote non-reducing glucopyranosyl residues and
reducing glucose residues, respectively. Horizontal lines and vertical lines
connecting circles indicate alpha-1,4 and alpha-1,6 glucosidic linkages
respectively.

20

Figure 3. Proposed model of hydrolysis and transfer reactions of alpha-
amylase from B. licheniformis on pullulan and maltooligosaccharide.

A. Pullulan Donor Acceptor

+g¢+gv

B. Maltooligosaccharide

 

i
Q §883°8 rm

21

B. Classes of Amylosaccharidases used to Study Catalytic Activity and
Substrate Cleavage Specicifity
MW

The hydrolysis of starch is a common first step in the conversion of starch
into a utilizable substrate for fermentafion or for the conversion to dextrose and
high fructose syrups. The enzyme alpha-amylase catalyzes the cleavage of the
alpha-1,4-glucosidic linkages in starch, glycogen and related glucans. This
enzyme was among the earliest to be purified and its mechanisms of action and
substrate specificities have been widely studied (Thoma et al., 1971). Although
amylases share high homology in the four conserved regions, their optimum
temperature, pH, thermostability, and major product formation from starch vary
significantly among species (Fogarty et al., 1980). Alpha-amylases in general
cleave only the alpha-1,4 linkages of starch and can hydrolyze past alpha-1,6
branch points due to their random attack pattern with the substrate.

All alpha-amylases are metalloenzymes and tightly bind one calcium ion
per enzyme molecule (Thoma et al., 1971). Almost all alpha-amylases contain
and require Ca2+ for optimal activity, and it is essential for the folding of the
enzyme in Aspergillus oryzae (Matsuura et al., 1984) and porcine pancreatic alpha-
amylase (Buisson et al., 1987). Alpha-amylases of different origins have similar
enzymatic properties and calcium requirements (Yang et al., 1993). Alpha-
amylase is most acfive in the lower pH range (pH 4.5-7.0) with an optimum pH
near 5.5. The optimal temperature for activity is dependent on the
thermophilicity of the organism from which the enzyme was isolated. The
lowest temperature optimum reported is 30°C, while temperatures of 95°C to
100°C have been reported from hyperthermophiles isolated from deep sea vents
and fumaroles (Brown et a1 ., 1990). Molecular weights of alpha-amylase ranges
from 10,000 to 140,000, while most microbial alpha-amylases have a molecular

22

weight between 50,000 to 60,000 (T able 4). On the basis of sequence comparisons
a calcium binding site is suggested to be a common structural feature of all
alpha-amylases (Matsuura et al., 1984).

Among the amylolytic enzymes, 3—dimensional structures have been
determined only for the alpha-amylases from Aspergillus oryzae
(Matsuura etal., 1984) and from porcine pancreas (Buisson et al., 1987) by X-ray
crystallographic studies. Initial identification of substrate binding site and
catalytic residues was by difference Fourier analysis with Taka amylase-A from
Aspergillus oryzae. A model fitting of an amylose chain in the catalytic site
showed a possible binding interaction between substrate and enzyme (Figure 4).
Glu230 and Asp297 were shown to be catalytic residues, acting as a general acid
and a general base, respectively; I-Ii5122, Ly8209, HileO, and His296 were
proposed as substrate-binding site residues (Matsuura et al., 1984).

Detailed investigation of the function of amino acids in four conserved
regions has been done on maltogenic alpha-amylase from Bacillus lidrcniformis
(BLMA). BLMA has both alpha-1,4 hydrolytic activity and alpha—1,6 transferring
activity of glucose or maltose to oligosaccharides (Cheong et al., 1995). Site-
directed mutagenesis of amino acid residues His250, AsnB30, Glu331, and
Asp422 revealed their importance as putative substrate-binding site residues
(Cheong et al, 1995). His250, when substituted with Gln, resulted in an increased
hydrolysis activity on soluble starch, slightly decreased activity on pullulan, and
increased alpha-1,6 transferring activity as compared with the wild type BLMA
enzyme. Substitution of Asn300 and G1u331 with Ser and Val respectively
resulted in slightly better hydrolysis activity on pullulan and soluble starch with
no significant difference in alpha-1,6 transferring activity. More interestingly,
this mutation endowed the enzyme with alpha-1,6 cleavage activity producing a
similiar enzyme cleavage activity on pullulan as neopullulanase. Wild type

23

Table 4. General properties of alpha-amylases from bacteria and fungi

 

 

Optimum
Organism 1MW pH Temp (°C) Reference
Acinetobacter sp. 55,000 7.0 55 Onishi et al., 1978
Aspergillus awamori 54,000 5.0 50 Bhella et al., 1985
Aspergillus oryzae 56,000 5.0 55 Kundu et al., 1970
Bacillus sp. 54,000 2.0 70 Uchino, 1982
Bacillus sp. ND. 10.5 50 Yamamoto et al., 1972
Bacillus acidocaldarius 66,000 5.0 70 Kanno, 1986
Bacillus amyloliquejhciens 50,000 5.9 65 Borgia et al., 1978
Bacillus caldolyticus 10,000 5.4 70 Heinen et al., 1972
Bacillus cereus 55,000 6.0 55 Yoshigi et al., 1985
Bacillus coagulcns 62,000 6.2 50 Kitahata et al., 1983
Bacillus licheniformis 62,600 8.0 90 Morgan et al, 1981
Bacillus macerans 140,000 6.3 ND. DePinto et al., 1964
Bacillus stearothermaphilus 44,000 5.5 70 Tsukagoshi et al., 1984
Bacillus subtilis 25,000 6.5 50 Takasaki, 1985
Fusarium oxysporum N.D. 4.0 25 Chary et al., 1985
Iactobacillus cellobiasus 22,500 7.3 50 Sen et al., 1984
Pyrococcusfariosus 76,300 5.6 98 Brown ct al., 1990
Streptomyces aureofaciens 40,000 5.0 40 Hostinova et al., 1978
Thermotoga maritima 60,000 5.5 95 Schumann et al., 1991

 

 

 

 

 

 

24

BLMA alpha-amylase from B. licheniformis does not hydrolyze alpha-1,6 linkages
of pullulan (Lee, 1993), while neopullulanase from B. stearothermaphilus
hydrolyzes both alpha-1,4 and alpha-1,6 glucosidic linkages (Kuriki et al., 1991).
The double mutant, Asn330Ser/ Glu331Val, hydrolyzed pullulan to glucose
(10%), maltose (28% ), and panose (62%), while the wild type BLMA produced
panose only from hydrolysis of pullulan (Cheong et al, 1995). These results
indicate potential importance of Asn330 and/ or Glu331 to alpha-1,6 hydrolytic
activity.
ZLPullulanass

The enzyme used following liquifaction by alpha-amylase in the
bioprocessing of starch is the debranching enzyme pullulanase, which is capable
of cleaving alpha-1,6 glucosidic linkages in starch and pullulan
(Fogarty et al., 1979; Fogarty et al., 1980; Lee et al., 1971; and Price, 1968). This
enzyme improves the saccharification rates and yields, and decreases reaction
times in the production of glucose or conversion syrups and is used in
combination with liquifying alpha-amylase (Norman, 1979; Reilly, 1979; and
Saha ct al, 1987). Compared to the alpha-amylases, pullulanases have been
isolated from relatively few microorganisms, mostly mesophiles. The pH optima
of pullulanases range from 5.0 to 9.0, with optimum temperatures ranging from
30°C to 85°C. Molecular weights reported are from 55,000 to 450,000 (Table 5).
Calcium is important for maintaining thermostability as with the alpha-amylase,
but there is no evidence to indicate a role for calcium in the folding of the protein
(Fogarty et al., 1979 and Fogarty et al., 1980). A calcium binding site has yet to be
identified. Detailed experiments have yet to be done to study substrate cleavage
specificity in this group of enzymes. It has been shown that the cleavage pattern
of recombinant pullulanase from P. woesei on pullulan produces 100%
conversion to maltotriose (Rudiger et al., 1995).

25

Figure 4 Proposed substrate binding and catalytic sites for alpha-
amylase from A. oryzae (Adapted from Matsuura et al., 1984). The
corresponding amino acid residues of amylopullulanase are indicated by
rectangles.

26

Figure 4. Proposed substrate binding and catalytic sites for Taka amylase-A
from A. oryzae.

/
° )J’GluISG
o ,v’
1 I
l
6 .
was OH?!” —
WN‘ marzz

‘ é» Trp83
LeuZQE‘xo r
T) r74
LAspIGS

Va1231 '

. Glu 230 ___
I,
I
“(1291)” ,c
N

l H15 ”‘51; “18295 01035
Arg344
ASp 340

Tyr79

 
 
   

 

27

Table 5. General properties of bacterial pullulanases.

 

 

Optimum

anism MW pH EPPC) Reference __
Bacillus sp. 92,000 9.0 55 Nakamura et al., 1975
Bacillus acidopullulyticus 100,000 5.0 60 Schulein et al., 1985
Bacillus cereus var. myc. 110,000 6.5 50 Takasaki, 1976
Bacillus subtilis 450,000 7.0 60 Takasaki, 1987
Bacteroides thetaiotamicron 77,000 6.5 37 Smith et al., 1989
Clostridium thermasulf. 130,000 7.0 70 Buchardt et al., 1991
Klebsiella pneumoniae 143,000 6.0 50 Ohba et al., 1975
Klebsiella pneumoniae 143,000 5.0 47 Eisele et al., 1972
Micrococcus sp. 120,000 10.0 50 Kimura et al., 1990
Streptomyces sp. ND. 5.5 50 Yagisawa, 1971
Thermus aquaticus 83,000 8.0 85 Plant et al., 1986

 

 

 

 

 

 

28

Wanna:
A new class of amylosaccharidase which cleaves both alpha-1,4 and alpha-

1,6 glucosidic linkages has been identified and termed amylopullulanase or
neopullulanase, depending on the products produced from action on pullulan.
Amylopullulanase, which produces maltotriose from pullulan, was initially
reported as an enzyme containing dual activities (both alpha-amylase and
pullulanase) (Mathupala et al., 1990). Analysis of the enzyme from Baallus
circulans F~2 showed the presence of separate active sites for the two activities
(Sata et al., 1989), and from Clostridium thermohydrosulfwicum E101 which
reported a cassette model, where half of the gene encodes alpha-amylase activity
while the other half encodes pullulanase activity (Melasniemi et aL, 1990). Site
directed mutagenesis results, from the monomeric amylopullulanase from
Thennoanaerobacter ethanolicus 39E, identified a single active site involved in the
dual activity (Mathupala et al., 1993). The pH optima of these enzymes varies
from 5.5 to 6.0 with a temperature optimum range of 70°C to 100°C. The
molecular weight varies from 110,000 to 162,000 (Table 6), with a calcium
requirement for thermostability but not enzymatic activity (Mathupala and
Zeikus 1993). Hydrolysis products produced from cleavage of starch include
maltotetraose, maltotrime, maltose and glucose; a single cleavage product,
maltotriose, is produced from action of amylopullulanase on pullulan
(Mathupala et al., 1990). Amylopullulanase has a higher activity on pullulan than
starch (3:1) and is incapable of further hydrolysis of maltotriose

(Mathupala et al., 1990). This report will be the first to conduct a detailed
investigation into the cleavage pattern profile of wild type and site-directed
mutants of T. ethanolicus 39E amylopullulanase.

29

Table 6. General properties of bacterial enzymes with alpha-1,4 and alpha-1,6
cleavage activity (neopullulanases and amylopullulanases).

 

Optimum
Organism MW pH Temp (°C) Reference
FPi/rococcusfuriosus 110,000 5.5 100 Brown et al. 1993
Pyrococus litoralis 119,000 5.5 90 Brown et al. 1993
Thermoanaerobacter 162,000 6.0 90 Mathupala et al., 1993
ethanolicus 39E
Thermoanaerobacterium 142,000 6.0 70 Ramesh et al., 1994
saccharolyticum B6A-RI
Bacillus stearothermophilus 62,000 6.0 65 Kuriki et al., 1988

 

 

 

 

 

 

 

30

WW
Neopullulanase, which produces panose from pullulan, was identified

and shown to contain one active center that is involved in the dual amylase
(alpha-1,4 cleavage) and pullulanase (alpha-1,6 cleavage) activities as well as
alpha-1,4 and alpha-1,6 transferring activity (Kuriki et al., 1991; Imanaka and
Kuriki, 1989). This enzyme from Bacillus stearothermaphilus has a molecular
weight of 62,000 and is fairly thermostable. The optimum temperature is 65°C
with an optimum pH for the enzyme of 6.0 (Table 6) (Kuriki et al., 1988). Detailed
biochemical characterization with regard to gene organization, subunit structure,
catalytic and substrate binding sites, and mechanism of action is necessary to
obtain a better understanding of these enzymes with dual activities. The dual
activity of neopullulanase results in producfion of panose, maltose, and glucose
(3:1:1) from hydrolysis of pullulan (Imanaka and Kuriki, 1989). Similiar to
amylopullulanase, neopullulanase hydrolyzes pullulan more efficiently than
starch. This enzyme is also capable of transglycosylation which was analyzed
using maltotriose. The by-products of the reaction were glucose and maltose
from the hydrolysis reaction, in addition to the formation of branched
oligosaccharides (isomaltose, isopanose, and panose) by the transfer reaction
(Kuriki et al., 1993). One active site with dual activities was identified, and the
enzyme, catalytic activity, cleavage specificity, and transferring activity can be
altered by mutating residues in the four conserved regions (Kuriki et al., 1993).
Replacement of amino acid residues in neopullulanase corresponding to
the putative catalytic sites resulted in loss of enzyme activity toward alpha-1,4
and alpha-1,6 glucosidic linkages. Asp328, Glu357, and Asp424 of
neopullulanase correspond to catalytic residues, Asp206, Glu230 and Asp297, of
Taka amylase-A. Substitution of these amino acids in Taka amylase-A with His,

31

His and G111, respectively, resulted in complete loss of catalytic activity
(Kuriki et al., 1991).

When amino acid residues corresponding to putative substrate binding
sites (His247, His423, Asn331, and Glu332 of neopullulanase correspond to
Hi8122, His296, Ly3209, and Hi8210 of Taka amylase-A ) were replaced, a change
in alpha-1,4 and alpha-1,6 hydrolytic and transferring specificity was observed
(Kuriki et al., 1991). The production ratio of panose from pullulan was used as an
indicator for changed specificity in alpha-1,4 and alpha-1,6 cleavage activity. The
production of panose is increased when the specificity toward alpha-1,4 linkages
increases. In contrast, the production of panose decreases when the specificity
toward alpha-1,6 linkages is higher. When Hi5247 was mutated to Glu and
His423 to Glu the mutant enzymes exhibited a higher specificity toward alpha-
1,4 glucosidic linkages producing more panose from pullulan than the wild type
enzyme (Kuriki et al., 1992). The double mutant Asn331$erl Glu332Val had an
opposite effect, producing less panose from pullulan due to decreased specificity
toward alpha-1,4 glucosidic linkages (Kuriki et al., 1992). These results suggest
that amino acids Hi8247, His423, Asn331, and Glu332 are involved in substrate

cleavage specificity.

Lficquenceficnmaficn

Amylosaccharidases from different organisms generally display low
homology, but share significant homology in four conserved regions (Kuriki and
Imanaka 1989; Kuriki et al., 1990; Nakajima et al., 1986; and
Mathupala et al., 1993). Alpha-amylase, pullulanase, neopullulanase and _
amylopullulanase sequences contain these four conserved regions, and these
regions have been proposed to be essential for the commonality of enzymatic

function by amylosaccharidases. The conserved regions form an active center,

32

and the substrate binding site by comparison to the refined 3-dimensional
structure of alpha-amylase from A. oryzae Taka amylase-A

(Matsuura et al., 1984). The dual activity of amylopullulanase, which cleaves
both alpha-1,4 and alpha-1,6 linkages, functions by Similiar mechanisms. In
enzymes that cleave both alpha-1,4 and alpha-1,6 glucosidic linkages there has
been identified a duplicated region 11 (II') which is located between regions III
and IV (Matuschek et al ., 1994). It has not yet been determined whether this is
significant for the specificity of dual activity enzymes. A catalytic triad
comprised of two aspartate residues and one glutamate residue, in regions 11, III,
and IV, is required for both catalytic activities. Substrate binding is governed by
four amino acids present in regions I, II, and IV; three histidines, and a lysine
residue (Mathupala et al., 1993).

Amino acid substitutions in neopullulanase and Taka amylase-A by site-
directed mutagenesis confirmed one active center involved in substrate cleavage
specificity (Kuriki 1992 and Matsuura et aL, 1984). The specificity for alpha-1,4
versus alpha-1,6 cleavage action was altered by replacing amino acids that are
involved in substrate recognition. For example, a mutated neopullulanase was
obtained which exhibited higher cleavage activity for alpha-1,4 glucosidic
linkages producing a higher yield of panose from pullulan (Kuriki et al., 1991).
The difference in cleavage specificity may be due to a difference in the binding
interaction between the substrate and the enzyme. The enzyme activity can be
altered by manipulating the binding specificity for the substrate. Similiar
experiments are proposed for amylopullulanase from a hyperthermophile in an
effort to design an enzyme with altered cleavage activity for biotechnological

production of high fructose syrups.

33

IV. Dissertation objectives and significance

This disssertation focuses on characterization of amylopullulanase from
the thermophile Thermoanaerobacter ethanolicus 39E with regard to catalytic
activity, substrate cleavage specificity, and thermophilicity. Structure-function
relationships were analyzed in relation to catalysis and thermophilicity of this
enzyme. Catalysis was analyzed in four ways. First, the protein sequence was
compared with homologous mesophilic and thermophilic amylases,
pullulanases, amylopullulanases, and neopullulanases. Second, deletion
mutagenesis experiments were performed to determine regions of the gene
required for the maintenance of thermostability and thermophilicity. Finally,
mutations at amino acid residues important for substrate binding specificity and
catalytic activity were constructed to analyze their affect on enzyme activity and
the regulation of alpha-1,4 and alpha-1,6 glucosidic cleavages. Studies into the
molecular basis for enzyme activity, substrate specificity, and thermostability
will provide insight for designing enzymes of industrial importance. Specific

aims in this project included:
(i) Design of mutated amylopullulanase with altered catalytic activity
and specificity.

(ii) Identification of a specific protein region responsible for

thermophilic characteristics of amylopullulanase.

To better control the various steps of starch conversion, industry needs a
variety of monospecific enzymes which are optimally active at high temperatures
(90°C-100°C) and low pH (5.5). Biochemical and structural analysis of regions
related to thermostability and thermophilicity, as well as defining the amino acid
residues responsible for cleavage specificity, may allow design of mutated
amylopullulanase engineered to act as a true alpha-amylase or pullulanase in

starch bioprocessing.

REFERENCES

Amelunxen, RE. and A.L. Murdock. 1978. Mechanisms of thermophily. Critical
Review of Micro. 6: 343-393.

Antranikian, G. 1990. Physiology and enzymology of thermophilic anaerobic
bacteria degrading starch. FEMS Microbial. Rev. 75: 201-218.

Aubert, I.P., P. Beguin, and I. Millet. 1993. Genetics and Molecular Biology of
Anaerobic Bacteria 1993, (Sebald, M., ed.). Springer-Verlag, NY

Bealin-Kelly, F., C.T. Kelly, and W.M. Fogarty. 1990. The alpha-amylase of the
caldoactive bacterium Bacillus caldavelax. Biochem Soc Trans 18: 310-1.

Bhella, R.S., and I. Altosaar. 1985. Purification and some properties of the
extracellular alpha-amylase from Aspergillus awamori. Can. I. Microbial.
31: 149-153.

Borgia, P.T. and LL. Campbell. 1978. Alpha-amylase from five strains of
Bacillus amylaliquefiiciens: Evidence for identical primary structures. I. Bacterial.
134: 389-393.

Brock, T.D., KM. Brock, R.T. Belley, and R.L. Weiss. 1972. Sulfirlabus: A new
genus of sulfur oxidizing bacteria living at low pH and high temperature.
Archs. Microbiol. 84: 54-68.

Brosnan, M.P., C.T. Kelly, and W.M. Fogarty. 1991. Investigation of the
mechanisms of irreversible thermoinactivation of Bacillus stearathermaphilus

alpha-amylase. Eur. I. Biachem.. 203: 225-231.

Brown, W.C. and LL. Campbell. 1993. A new cloning vector and expression
strategy for genes encoding proteins toxic to Escherichia calii. Gene pp. 99-103.

Brown, SH. and RM. Kelly. 1993. Characterization of amylolytic enzymes,
having both alpha-1,4 and alpha-1,6 hydrolytic activity. Appl. and Environ.
Microbial. 59: 2614-2621.

Brown, S.H., I-LR. Costantino, and RM. Kelly. 1990. Characterization of
amylolytic enzyme activities associated with the hyperthermophilic
archaebacterium Pyracaccus furiasus. Appl. and Environ. Microbial. 56: 1985-1991.

Buisson, G., E. Duee, R. Haser, and F. Payan. 1987. Three dimensional

structure of porcine pancreatic alpha-amylase at 2.9A resolution. EMBO.
6: 3909-3916.

34

35

Burchardt, G., Wienecke, A., and H. Bahl. 1991. Isolation of the pullulanase
gene from Clastridium thermasulfuragenes (DSM 3896) and its expression in
Escherichia coli. Curr. Microbial. 22: 91-95.

Chary, SJ. and SM. Roddy. 1985. Folia Microbiol. (Prague) 30: 452.

Cheong, T.K. 1995. Modulation of Bacillus lichenifarmis amylases by site-directed
and deletion mutagenesis, Ph.D. Thesis, Seoul National University.

Coolbear, T., R.M. Daniel, and FUN. Morgan. 1992. The enzymes from extreme
thermophiles: bacterial sources, thermostabilities and industrial relevance. Adv.
Biochem. Eng/Biatechnal. 45: 57-98.

DePinto, I.A. and LL. Campbell. 1964. Formation and degradation of cyclic
dextrins by intracellular enzymes of Bacillus macerans. Science. 146: 1064-1065.

Eisele, B., LR. Rashed, and K. Wallenfels. 1972. Molecular characterization of
pullulanase from Aerabacter aeragenes. Eur. I. Biochem. 26: 62-67.

Fogarty, W.M., and CT. Kelly. 1979. Starch-degrading enzymes of microbial
origin. Prag. Ind. Microbial. 15: 87-150.

' Fogarty, W.M., and CT. Kelly. 1980. Amylases, amyloglucosidases and related
glucanases, p. 115-169. In A.H. Rose (ed.), Microbial enzymes and
bioconversions. Academic Press, Ind., New York.

Fukusumi, S., A. Kamizono, S. Horinouchi, and T. Beppu. 1988. Cloning and
nucleotide sequence of a heat-stable amylase gene from an anaerobic
thermophile, Diclyaglomus thermaphilum. Eur. I. Biochem. 174: 15-21.

Gray, G.L., S.E. Mainzer, M.W. Rey, M.H. Lamsa, K.L. Kindle, C. Cannona, and
C. Requadt. 1986. I. Bacteriol. 166: 635-643.

Hebeda, RE: (1987) 'Corn Sweeteners', In Corn Chemistry and Technology
(Watson, SA. and Ramstad, P.E., Eds), pp. 501-534. AACC, Ind. St. Paul,
Minnesota.

Heinen, U.I., and W. Heinen. 1972. Characteristics and properties of a caldo-
active bacterium producing extracellular enzymes and two related strains. Arch.
Microbial. 82: 1-23.

Horinouchi, S., S. Fukusumi, T. Ohshima, and T. Beppu. 1988. Cloning and
expression in Escherichia coli of two additional amylase genes of a strictly
anaerobic thermophile, Dictyaglamus thermaphilum, and their nucleotide
sequences with extremely low guanine-plus-cytosine contents. Eur. I. Biochem.
176: 243-253.

36

Hostinova, B., and I. Zelinka. 1978. Starch. 30: 338.

Hyun, H.H., and I.G. Zeikus. 1985. General characterization of thermostable
extracellular beta-amylase from Clastridium thermasulfuragenes. Appl. Environ.
Microbial. 49: 1162-1167.

Imanaka, T., and T. Kuriki. 1989. Pattern of action of Bacillus stearathennaphilus
neopullulanase on pullulan. I. Bacterial. 171: 369-374.

Kammo, M. 1986. A Bacillus acidacaldarius alpha-amylase that is highly stable to
heat under acidic conditions. Agric. Biol. Chem. 50: 23-31.

Kim, I.-C., Cha, I.-H., Kim. I.-R., Iang, S.-Y., Seo, B.-C., Cheong, T.-K., Lee, D.S.,
Choi, Y.D., and K.-H. Park. 1992. Catalytic properties of the cloned amylase
from Bacillus licheniformis. I. Biol. Chem. 267: 22108-22114.

Kim, C.-H., and Y.S. Kim. 1995. Substrate specificity and detailed
characterization of a bifunctional amylase-pullulanase enzyme from Bacillus
circulans F-2 having two different active sites on one polypepfide. Eur. I. Biochem.
227: 687-693.

Kim, I.-C., Yoo, S.-H., Lee, S.-I., Oh, B.-H., Kim, I.-W., and K.-H. Park.
Synthesis of branched oligosaccharides from starch by two amylases cloned from
Bacillus licheniformis. 1994. Biosci. Biatech. Biochem. 58: 416-418.

Kimura, T., and K. Horikoshi. 1990. Characterization of pullulan-hydrolysing
enzyme from an alkalopsychrotrophicMicraaaccus sp. Appl. Microbial. Biotechnol.
34: 52-56.

Kitahata, S., M. Taniguchi, SD. Beltran, T. Sugimoto, and S. Okada. 1983.
Purification and some properties of cyclodextrinase from iBacillus coagulans.
Agric. Biol. Chem. 47: 1441-1447.

Kundu, A.K., and S. Das. 1970. Production of amylase in liquid cluture by a
strain of Aspergillus aryzae. Appl. Microbial. 19: 598-603.

Kunkel, TA. 1985. Rapid and efficient site-specific mutagenesis without
phenotypic selecfion. Prac. Natl. Acad. Sci. USA. 82: 488-492.

Kuriki, T. 1992. Can a protein engineering interconvert
glucanhydrolases/ glucanotransferases, and their specificities? Trends in
Glycascience and Glycatech. 4: 567-572.

Kuriki, T., and T. Imanaka. 1989. Nucleotide sequence of the neopullulanase
gene from Bacillus stearathennaphilus. I. Gen. Microbial. 135:1521-1559.

37

Kuriki, T., I.-H. Park, and T. Imanaka. 1990. Characteristics of thermostable
pullulanase from Bacillus stearathermaphilus and nucleotide sequence of the gene.
I. Ferment. Biaeng. 69: 204-210.

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

Kuriki, T., I.H. Park, S. Okada, and T. Imanaka. 1988. Purification and
characterization of thermostable pullulanase from Bacillus stearathermaphilus and
molecular cloning and expression of the gene in Bacillus subtilis. Appl. environ.
Microbial. 54: 2881-2883.

Kuriki, T., H. Takata, S. Okada, and T. Imanaka. 1991. Analysis of the active
center of Bacillus stearathermaphilus neopullulanase. I. Bacterial. 173: 6147-6152.

Kuriki, T., H. Takata, S. Okada, and T. Imanaka. 1992. Alteration of the
specificity of neopullulanase by protein engineering. Denpun Kagaku. 39: 75-84.

Kuriki, T., M. Yanase, H. Takata, Y. Takesada, T. Imanaka, and S. Okada. 1993.
A new way of producing isomalto-oligosaccharide syrup by using the
transglycosylation reaction of neopullulanase. Appl. and Environ. Microbial.

59: 953-959.

Lamed, R., E.A. Bayer, B.C. Saha, I.G. Zeikus. 1988. Biotechnological potential
of enzymes from unique thermophiles. 8th Intermational Biotechnology
Symposium. 1: 371-383.

Lee, SJ. 1993. Action pattern of Bacillus licheniformis maltogenic amylase and
production of branched oligosaccharide. MS. thesis, Seoul National University.

Lee, E.Y.C., and WJ. Whelan. 1971. Glycogen and starch debranching enzymes,
p. 191-234. In PD. Boyer (ed.), The enzymes, vol. 5, 3rd ed. Academic Press, Inc.,
New York.

Lee, s.-]., Yoo, S.-H., Kim, M.-I., Kim, ].-w., Seok, H.-M., and K.-H. Park. 1994.
Production and characterization of branched oligosaccharides form liquefied
starch by the action of B. lidzenifvnnis amylase. Starch 47: 127-134.

Lowe, S.E., M.K. Iain, I.G. Zeikus. 1993. Biology, ecology, and biotechnoligical
applications of anaerobic bacteria adapted to environmental stresses in
temperature, pH, salinity, or substrates. Micrab. Rev. 57: 451-509.

38

Madi, B., G. Antranikian, K. Ohmiya, and G. Gottschalk. 1987. Thermostable
amylolytic enzymes from a new Clashidium isolate. Appl. Environ. Microbial.
53: 1661-1667.

Margaritis, A., and RE]. Merchant. 1986. Principles and bioteclmological

applications of bacterial ice nucleation. CRC Crit. Rev. Biatechnal.
4: 327-366.

Mathupala, S.P., and I.G. Zeikus. 1993. Improved purification and biochemical
characterization of extracellular amylopullulanase from Thermoanaerabacter
ethanalicus 39E. Appl. Microbial. Biotechnol. 39:487-493.

Mathupala, S., B.C. Saba, and I.G. Zeikus. 1990. Substrate competition and
specificity at the active site of amylopullulanase from Clastridium
thermahydrasulfuricum. Biochem. Biophys. Res. Commun. 166:126-132.

Mathupala, S.P., S.E. Lowe, S.M. Podkovyrov, and I.G. Zeikus. 1993.
Sequencing of the amylopullulanase (apu) gene of Thermaanaerobacter ethanalicus
39E, and identification of the active site by site-directed mutagenesis. I. Biol.
Chem. 268: 16332-16344.

Matsuura, Y., M. Kusunoki, W. Harada, and M. Kakudo. 1984. Structure and
possible catalytic residues of Taka-amylase A. I. Biochem. 95: 697-702.

Matuschek, M., G. Burchhardt, K. Sahm, and H. Bahl. 1994. Pullulanase of
Thermaanaerabacterium thermasulfurigenes EM] (Clastridium thermosulfuragenes):
Molecular analysis of the gene, composite structure of the enzyme, and a
common model for its attachment to the cell surface. I. Bacterial. 176: 3295-3302.

Melasniemi, H. 1987. Characterization of alpha-amylase and pullulanase
activities of Clashidium thermahydrasulfuricum. Biochem. I. 246: 193-197.

Melasniemi, H., M. Paloheimo, and L. Hernia. 1990. Nucleotide sequence of
the alpha-amylase-pullulanase gene from clastridium themahydrasulfitricum.
I. Gen. Microbial. 136: 447-454.

Mizukami, M., H. Yamagata, K. Sakaguchi, S. Udaka. 1992. Efficient
production of thermostable Clastridium thermasulfiiragenes beta-amylase by
Bacillus brevis. I. Ferment. Biaeng. 73: 112-115.

Moore, CA. and P. Dwoskin. 1970. Economics, production, and marketing. In
"Corn: Culture, Processing, Products," ed. C.P. Inglett, p. 84. Avi Pub. Co., Inc.,

Westport, Conn.

Morgan, EL, and F.G. Priest. 1981. Characterization of a thermostable alpha
amylase from Bacillus licheniformis NCIB6346. I. Appl. Bacterial. 50: 107-114.

39

Nakajima, R., T. Imanaka, and S. Aiba. 1986. Comparison of amino acid
sequences of eleven different alpha-amylases. Appl. Microbial. Biotechnol.
23: 355-360.

N akarnura, N ., K. Watanabe, and K. Horikoshi. 1975. Purification and some
properties of alkaline pullulanase from a strain of Bacillus No. 202-1, an
alkalophilic microorganism. Biachim. Biophys. Acta. 397: 188-193.

Nakao, M., T. N akayama, M. Harada, A. Kakudo, H. Ikemoto, S. Kobayashi,
and Y. Shibano. 1994. Appl. Microbiol. Biotechnol. 41: 337-343.

Nashihara, N., N. Nakamura, H. Nagayama, and K. Horikoshi. 1988. Cloning
and expression of the firermostable pullulanase gene from Thermus sp. strain
AMD-33 in Escherichia coli. FEMS Microbial. Lett. 49: 385-388.

Norman, B.E. 1979. The application of polysaccharide degrading enzymes in the
starch industry, in R.C.W. Berkeley, G.W. Gooday, and DC. Ellwood (ed.),
Microbial polysaccharides and polysaccharases. p. 339-376. Academic Press,
Inc., New York.

N unberg, I.I-L, I.H. Meade, G. Cole, F.C. Lawyer, P. McCabe, V. Schweickart, R.
Tal, V.P. Wittman, F.E. Flatgaard, and MA. Innis. 1984. Molecular cloning and
characterization of fire glucoamylase EC 3.2.1.3 gene of Aspergillus awari. Mal.
Cell. Biol. 4: 2306-2309.

Ohba, R., and S. Ueda. 1973. Purification, crystallization and some properties of
intracellular pullulanase from Aerabacter aeragenes . Agric. Biol. Chem.
37: 2821-2826.

Onishi, H. and O. Hidaka. 1978. Purification and properties of amylase
produced by a moderately halophilic Acinetabacter sp. Can. I. Microbial.
24: 1017-1023.

Plant, A.R., R.M., Clemens, H.W. Morgan, and RM. Daniel. 1987. Active-site
and substrate-specificity of Thermaanaerabium Tok6-B1 pullulanase. Biochem I.,
246, 537-541.

Plant, A.R., I-LW. Morgan, and RM. Daniel. 1986. A highly stable pullulanase
from Thermus aquaticus YT-1. Enzyme Micrab. Technol. 8: 668-672.

Price, I.C. 1968. Pullulanase, In I.A. Radley (ed.), Starch and its derivatives, 4th
ed. p. 509-527. Chapman and Hall, Ltd.

4O

Ramesh, M.V., B.C. Saha, S.P. Mathupala, S. Podkovrov, and I.G. Zeikus. 1994.
Biodegradation of starch and a-glycan polymers. In "Biochemistry of Microbial
Degradation," ed. C. Ratledge, p. 313. Kluwer Academic Publishers, N orwell,
MA.

Rudiger, A., P.L. Iorgensen, and G. Antranikian. 1995. Isolation and
characterization of a heat-stable pullulanase from the hyperthermophilic
archaeon Pyraaaccus waesei after cloning and expression of its gene in Escherichia
coli. Appl. Environ. Microbial. 61: 567-575.

Saha, B.C., R. Lamed, C.Y. Lee, S.P. Mathupala, and I.G. Zeikus. 1990.
Characterization of an endo-acting amylopullulanase from Thermaanaerabacter
strain B6A. Appl. Environ. Microbial. 56: 881-886.

Saha B.C. and I.G. Zeikus. 1987. Biotechnology of maltose syrup production.
Process Biochem 22: 78-82.

Saha, B.C. and I.G. Zeikus. 1989. Novel highly firermostable pullulanase from
firermophiles. Trends in Biotechnol. 7: 234-239.

Saha, B.C. and I.G. Zeikus. 1990. Characterization of firermostable
cyclodextrinase from Clastridium thermahydrasulfuricum 39E. Appl. Environ.
Microbial. 56: 2941-2943.

Saha, B.C. and I.G. Zeikus. 1991. Characterization of firermostable alpha-
glucosidase from Clastridium themwsulfiiragenes 39E. Appl. Microbial. Biotechnol.
35: 568-571.

Saito, N. 1973. A firermophilic extracellular alpha-amylase from Bacillus
lichenifbrmis. Arch. Biochem. Biophys. 155: 290-294.

Sakano, Y., N. Masuda, and T. Kobayashi. 1971. Hydrolysis of pullulan by a
novel enzyme from Aspergillus niger. Agric. Biol. Chem. 35: 971-973.

Sambrook, I., E.F. Fritsch, and T. Maniatis. 1986. Molecular cloning: a
laboratory manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

Sata, 11., M. Umeda, C.I-L Kinr, H. Taniguchi, and Y. Maruyama. 1989.
Amylase-pullulanase enzyme produced by B. circulansF2. Biochem. Biophys. Acta.
991: 388-394.

Schulein, M., and 3N. Hojer-Pedersen. 1984. Characterization of a new class of
firermophilic pullulanases from Bacillus acidapullulyticus. Ann. N. Y. Acud Sci. 434:
271

41

Schumann, I., A. Wrba, R. Iaenicke, and KO. Stetter. 1991. Topographical and
enzymatic characterization of amylases from the extremely firermophilic
eubacterium Thermataga maritima. FEBS Lett. 282: 122-126.

Sen, S., and S.L. Chakrabarty. 1984. I. Ferment. Technol. 62: 407.

Shen, GJ., K.C. Srivastava, B.C. Saba, and I.G. Zeikus. 1990. Appl. Microbiol.
Biotechnol. 33: 340-344.

Smith, K.A., and AA. Salyers. 1989. Cell associated pullulanase from
Bacteraides thetaiatamicran: cloning, characterization, and insertional mutagenesis
to determine role in pullulan utilization. I. Bacterial. 171: 2116-2123.

Sundaram, V.A. 1986. Physiology and growth of thermophilic bacteria. In
Brock, TD (ed) "'Ihermophiles: General, Molecular and Applied Microbiology”, p.
75 Wiley and Sons, New York.

Takata, H., T. Kuriki, S. Okada, Y. Takesada, M. Iizuka, N. Minamiura, and T.
Imanaka. 1992. Action of neopullulanase: neopullulanase catalyzes bofir
hydrolysis and transglycosylation of alpha-(1,4) and alpha-(1,6) blucosidic
linkages. I. Biol. Chem. 267: 18447-18452.

Takasaki, Y. 1976. Purification and enzymatic properties of beta-amylase and
pullulanase from Bacillus cereus var. myaaides. Agric. Biol. Chem. 40: 1523-1530.

Takasaki, Y. 1987. Pullulanase-amylase complex enzyme from Bacillus subtilus.
Agric. Biol. Chem. 51: 9-16.

Thoma, I.A.,I.E. Spradlin, and S. Dygert. 1971. Meth. Enzymol. 5: 115-189.

Trevelyan, W.E., D.P. Procter, and I.S. Harrison. 1950. Detection of sugars on
paper chromatograms. Nature. 166 pp. 444-445.

Tsukagoshi, N., H. Ihara, H. Yamagata, and S. Udaka. 1984. Cloning and
expression of a firermophilic alpha-amylase gene from Bacillus stearathermophilus
in Escherichia coli. Mal. Gen. Genet. 193: 58-63.

Uchino, F. 1982. A thermophilic and unusually acidophilic amylase produced
by a thermophilic acidophilic Bacillus sp. Agric. Biol. Chem. 46: 7-13.

Vieille, C., Burdette, D.S., and Zeikus, I.G. 1996. in Biotech. Ann. Rev., vol. 2,
(M. R. E1 Geweley, ed.) pp. 1-83. Elsevier, Amsterdam, Neth.

Vihinen, M. 1987. Relationship of protein flexibility to thermostability. Prat.
Engin. 1: 477-480.

42

Watanabe, K., K. Chishiro, K. Kitamura, Y. Suzuki. 1991. Praline residues
responsible for firermostability occur with high frequency in fire loop regions of
an extremely firermostable oligo-1,6-glucosidase from Bacillus thermaglucasidasius
KPlOO6. I. Biol. Chem. 266: 24287-24294.

Wind, RD., W. Liebl, R.M. Buitelaar, D. Penninga, A. Spreinat, L. Dijkhuizen,
and H. Bahl. 1995. Appl. Environ. Microbiol. 61: 1257-1265.

Yagisawa, M., K. Kata, Y. Koba, and S. Ueda. 1972. Pullulanase of Streptomyces
sp. No. 280. I. Ferment. Technol. 50: 572-579.

Yamamoto, M., T. Tanaka, and K. Horikoshi. 1972. Alkaline amylases of
alkalophilic bacteria. Agric. Biol. Chem. 36: 1819-1823.

Yang, M., A. Galizzi, and D. Henner. 1983. Nucleotide sequence of fire amylase
gene from Bacillus subtilis. Nucl. Acids Res. 11: 237-249.

Yoshigi, N., T. Chikano, and M. Kamirnura. 1985. Purification and properties
of amylase from Bacillus cereus NY -14. Agric. Biol. Chem. 49: 3369-3376.

Zeikus, I.G. 1990. In Biotechnology: Science, Education and Commercialization;
Vasil, I.K., Ed.; Elsevier Science Publishing Co., Inc.: New York, New York, pp
23-39.

CHAPTERII

Biochemical and Enzymatic Characterization of Catalytic Activity and alpha-
1,4 / alpha-1,6 Cleavage Specificity of Recombinant Amylopullulanase
fromThermaanaerobacter ethanolicus 39E

Prepared for submission to the Journal of Biological Chemistry

43

ABSTRACT

The active center of amylopullulanase from Thermaanaerabacter ethanalicus 39E
was analyzed by means of sitedirected mutagenesis using the crystal structure
of Taka-amylase A as a model. It was previously shown first four highly
conserved regions exist in fire active center of various amylolytic enzymes. When
fire active site residues Asp597, Glu626, and Asp703 were mutated, hydrolysis of
alpha-1,4 and alpha-1,6 glucosidic linkages were not detectable. Mutation of fire
individual substrate binding site residues altered the specificity as either an
increased or decreased alpha-1,4 cleavage activity, or resulted in complete loss of
bofir alpha-1,4 and alpha-1,6 cleavage activity. Pullulan, amylase, soluble starch,
and beta-limit dextrin were used as substrates to analyze substrate cleavage
specificity and the rate of catalytic activity. The His493Gln and

Asn600Ser/ G1u601Val mutants increased fire rate of alpha-1,4 and alpha-1,6
cleavage activity relative to fire wild type amylopullulanase. The His493Gln
mutant showed increased cleavage specificity for alpha-1,4 linkages on soluble
starch; while, the Asn6005er/ Glu601Val mutant resulted in decreased alpha-1,4
cleavage specificity on soluble starch when compared to the wild type enzyme.
Here we show firat the rate of hydrolysis and end product formation can be
altered by manipulating the putative substrate binding residues of
amylopullulanase without significantly changing the kinetic parameters. This is
the first report of changing fire binding characteristics of amylopullulanase.

INTRODUCTION

Amylopullulanase is an enzyme capable of hydrolyzing both alpha-1,4
and alpha-1,6 glucosidic linkages. It is distinguished from its counterparts alpha-
amylase or pullulanase, which have fire ability to hydrolyze only alpha-1,4 or
alpha-1,6 linkages, respectively.

Amylosaccharidases, broadly grouped into amylases and pullulanases, are
classified based on substrate cleavage specificity and have been identified in
various mesophilic, thermophilic, hyperfirermophilic bacteria, and Archae.
Pullulanases can be separated into four groups: pullulanase, isopullulanase,
amylopullulanase, and neopullulanase, depending on the products of cleavage
on pullulan. Examples include, but are not limited to, the following: pullulanase
from Thermaanaerabium Tok6-B1 (Plant at al., 1987), and amylopullulanase from
Thermaanaerabacter ethanalicus , formerlyClastridium thermahydrasulfuricum
(Mafirupala et al., 1990) which hydrolyze fire alpha-1,6 linkages in pullulan to
produce maltotriose; isopullulanase from Aspergillus niger (Sakano et al., 1971)
hydrolyzes alpha-1,4 linkages producing isopanose; and neopullulanase from
Bacillus stearathennaphilus (Kuriki et al., 1988a) produces panose by cleavage of
alpha-1,4 linkages and a small amount of glucose and maltose by fire limited
alpha-1,6 cleavage activity on pullulan. Amylopullulanases and neopullulanases
are also capable of cleaving alpha-1,4 linkages in starch, producing low
molecular weight oligosaccharides (Mafirupala et al., 1990; Kuriki et al., 1988a).
Isopullulanase and pullulanase can not hydrolyze starch.

Amylases are separated into fiuee groups: alpha-amylases, beta-amylases,
and glucoamylases. These enzymes catalyze fire cleavage of starch, glycogen,
and related glucans. Alpha-amylase is an endo-acting enzyme responsible for
the random cleavage of alpha-1,4 glucosidic linkages (Matsuura et al ., 1983),

4S

46

whereas, beta-amylases and glucoamylases are exo-acting enzymes firat
hydrolyze alpha-1,4 linkages, or alpha-1,4 and alpha-1,6 linkages, respectively
(Fogarty and Kelly 1979). The most widely studied alpha-amylase is from
Aspergillus aryzae (T aka amylase-A) (Matsuura et al., 1983). While no crystal
structure is available for an enzyme wifir bath alpha-1,4 and alpha-1,6 activity,
Taka amylase-A has been crystallized and fire alignment of amino acids in
homologous regions has been used for analysis of other alpha-amylases,
neopullulanases, isoamylases, pullulanases, cyclodextrin glucatransferases, and
amylopullulanases.

Cleavage action pattern has been most extensively studied in
neopullulanase from B. stearathermaphilus (Kuriki et al., 1991; Imanaka and Kuriki
1989; Takata et al., 1992), alpha-amylase from B..lichenifvrmis
(Kim et al., 1994; Lee et al., 1995; Kim et al., 1995), and amylase-pullulanase from
Bacillus circulans F-2 (Kim and Kim 1995). Cleavage activity of amylopullulanase
from T. ethanalicus 39E will be compared to fire cleavage patterns seen wifir fire
above enzymes. Amylopullulanase described in firis report showed alpha-1,4
cleavage activity against soluble starch, amylase, beta-limit dextrirr, and low
molecular weight oligosaccharides; and alpha-1,6 cleavage activity against
pullulan and beta limit dextrin. The catalytic activity an pullulan was 2-3 fold
higher then firat detected on amylase, soluble starch, and beta-limit dextrin.
Kinetic analysis of wild type amylopullulanase established similiar Kmapp
values for soluble starch, amylose, and beta-limit dextrin (0.87 mg / ml), while
Kmapp for pullulan was 2-fald lower (0.36 mg/ rrrl). The substrate affinity for the
mutant enzymes was similiar to firat reported above for wild type .
amylopullulanase. A model has been proposed for enzymatic activity an the
above menfioned substrates. This paper describes characterization of specific

47

amino acid residues important for catalytic activity and cleavage properties of
wild type and mutant amylopullulanase.

EXPERIMENTAL PROCEDURES

Chemicals and reagents - All chemicals were of molecular biology or
analytical grade and were obtained from Aldrich Chemical Co., or Sigma.

Bacterial strains, plasmids, and culture conditions - E. coli strain TG-l
{F‘traDB6 1.:ch [lacZ]M15 proA+B+/sup£ [hde-mch]5[rk'mk'Mch"] thi lac-
praAB]} was obtained from Amersham Corp. E. coli DHSaF' {F'hst17 (rk'mk‘l')
supE44 thi-I recAlgyrA (NaF) relAI (lacZ-argF)U169 deaR (80dlac (lacZ)M15}
was obtained from Befiresda Research Laboratories (Gaifirersburg, MD). E. coli
CIZ36 {dut, ung, thi, relA;pC1105 [Cmrli was obtained from BIO-RAD (Richmond,
CA). Plasmids M13mp19 (New England Biolabs, Beverly, MA) and pUC18
(Gibco BRL, Grand Island, NY) were used as cloning vectors. The plasmid
constructed in firis study is shown in Fig. 1.

Media and growth conditions - E. coli cultures were grown in LB
(Sambrook et al., 1989) medium (10 g of tryptone, 5 g of yeast extract, and 10 g of
NaCl per liter). Ampicillin was used at 100 ug/ ml for strains harboring plasmid.

Manipulation of DNA - Plasmid DNA purification, restriction analysis,
ligatians, PCR, and transformations were performed by conventional techniques
(Sambrooket al., 1989).

Oligonucleotides - Oligonucleotides were synfiresized in an Applied
Biosystems model 380A DNA synthesizer at fire Macromolecular Structure,
Synthesis and Sequencing Facility, Department of Biochemistry, Michigan State
University. The Oligonucleotides were subsequently 5'-phasphorylated using T4
polynucleotide kinase, for use in site directed mutagenesis (Table 1). 1

Enzyme purification - E. coli DHSoF' harboring the prAP164-UC wild type

or recombinant plasmids containing fire specific amino acid mutations were

48

49

Figure 1 Physical map of the pUC18 clone (prAP164-UC) containing the
apu gene. A fragment containing the apu gene with Eco RI and Bam HI
ends engineered by PCR was inserted into fire multiple cloning site of the
vector pUC18. Other loci are bla, B-lactamase gene conferring resistance to
ampicillin; lacl, fire lacI‘l repressor protein which binds downstream of fire
promoter to dawnregulate expression in the absence of induction;

ari, pUC18 origin of replication.

50

Figure 1. Physical map of the pUC18 clone (prAP164-UC) containing the apu
gene.

BamI-I I 429

 
 
   
 

PvuII918

Ball 1713
Beck 14809
Pst I 4635 Bgl 11 2448
”M m 4558 Pst I 2453
Apa 14294
Ban 11 4295 Kpn 12713

Hind III 3908

51

Table 1. Oligonucleotides used in this study

 

 

MIEAIIQN W
His49361n 5' CTATCATCACTTGTQIQAITGAAGACGCC 3'
Asp597Leu 5' GAGCAATTTCATTTGCAACAAQCAATCTCCAG 3'
AsnGOOSer/ 5' CGTGAGCAAEIAQAQITGCAACATCC 3'

Glu6 0 1Va1

Glu626A8p 5' GAAGCATCTCCCCAAAGQIQTGCAATCATTGG 3'
BiBTOZArg 5' CTCATGGTGTCACQACAACCTAAAAGG 3'
His702ABp 5' CTCATGGTGTCAIQACAACCTAAAAGG 3'
BiS7OZGlu 5' CTCATGGTGTCIICACAACCTAAAAGG 3'
His702Ly8 5' CTCATGGTGTCIIIACAACCTAAAAGG 3'
A3p703Gly 5' CTCATGGTQQQATGAGAACCTAAAAGG 3'
Asp672Asn 5' GTCAAGTTTTGCTGCAIITATAGGATTGTGAAC 3'
Asp67ZGlu 5' GTCAAGTTTTGCTGCIIQTAEAGGATTGTGAAC 3'
.A39672Phe 5' GTCAAGTTTTGCTGCAAAIATAGGATTGTGAAC 3'
Ly8675Arg 5' GCCTTTGGTCAAGIQITGCTGCAICTAEAG 3'
Ly8675A8p 5' GCCTTTGGTCAAGAIQTGCTGCAICTATAG 3'
Ly8675Ala 5' GCCTTTGGTCAAGIEQTGCTGCATCTAIAG 3'
Leu676818 5' CAIAAGCCTTTGGTCAIETTTTGCTGCATC 3'
PheBBZABn 5' GAIAITCAACAGGATCAIIACCTCTACTGTATTC 3'
Ly8401Ala 5' GTATATCCAGGTTTATCIEQATCAITCGGATTG 3'
A3p40261y 5' GTATATCCAGGTTTAQQTTTATCATTCGGATTG 3'
LysSZSLeu 5' CACCGTATGGAGAEAQTGACTGAICTCCCTG 3'

 

 

 

 

underlined; altered nucleotides in the primer to produce the desired
amino acid change.

52

grown at 37°C in 5x 10.0 ml LB medium containing ampicillin (100 ug/ ml). Cells
were harvested by centrifugation in a microcenfiifuge (12,000 rpm x 5 min) and
fire cell pellet resuspended in 0.5 ml 50 mM Na-acetate buffer (pH 6.0) containing
5 mM CaClz and 150 ug lysozyme. The suspension was incubated on ice for

30 minutes followed by fi'eeze-firaw. The cell lysate was centrifuged (14,000 rpm
x 5 min) and fire supernatant heat treated at 85°C for 5 min. After centrifugation
(14,000 rpm x 5 min), the supernatant was recovered and tested for amylase and
pullulanase activity. The purity of the various samples of Apu protein, judged
by coomassie stained SDS-PAGE, was ~15%. The host cell contained no
amylosaccharidase activity.

Protein determination and gel electrophoresis - Protein concentrations were
determined using bicinchroninic acid (BCA Assay Kit, Pierce Chemical Co.),
using bovine serum albumin as the standard. SDS-PAGE was performed
according to the method of Laemmli (Laemmli 1970) using 10.0% polyacrylamide
gels in a Mini-Protean II apparatus (Bio-Rad), and proteins were visualized by
staining wifir Coomassie Brilliant Blue R-250. The molecular weights of fire
recombinant proteins were estimated by comparison to high range molecular
weight standards (Bio-Rad).

Activity staining by native PAGE - Partially purified enzyme was
electrophoresed as for SDS-PAGE, on a Bio-Rad Mini Protean II electrophoresis
apparatus, except for fire absence of SDS in the buffer systems and in fire gel.
Prior to polymerization of fire gels, soluble starch was added to a final
concentration of 1.0% (w/ v). After electrophoresis, the gels were washed wifir
acetate buffer (pH 6.0), and incubated at 60°C for 5 to 10 min. For activity
staining of soluble starch embedded gels, a solution of 0.15% iodine: 1.5% K1 was
added as an overlay. Alpha-amylase activity could be detected as a clear band,
indicating absence of starch, against fire dark black backgrormd.

53

Enzyme assays - Cell extracts prepared as described above were used as
enzyme sources. Alpha-amylase activity was determined in a reaction mixture
firat contained 1.0% soluble starch (w/ v) in 50 mM Na-acetate buffer (pH 6.0)
with 5 mM CaClz, and enzyme. After incubation at 60°C for 30 min., fire reaction
was stopped by adding 0.8 ml of dinitrosalicylate solution [0.25 M NaOH,

71.0 mM sodium potassium tartrate, 4.0 mM N a2503, 5.0 mM phenol, and

44.0 mM 3,5 dinitrosalicylic acid] (Bernfeld 1955) and heated in a 100°C oil bafir
for 15 min. The samples were cooled on ice and fire absorbance at 640 nm was
measured. One unit of amylase activity was defined as fire amount of enzyme
which produces 1 umol of reducing sugar/ min wifir glucose as fire standard
(Miller 1959). Pullulanase activity was determined as above except that

1.0% pullulan was used in place of starch.

Kinetic analysis - Assays to determine Kmapp were conducted at 60°C wifir
substrate concentrations between 20 x Kmapp and 0.2 x Kmapp. Kinefic
parameters were determined using nonlinear curve fits of fire Michaelis-Menterr
equation to the data. Calculations were done on an IBM personal computer
using Kinzyme.

Oligonucleotide-directed mutagenesis of the amylopullulanase gene and
sequencing - Mutagenesis was performed using an oligonucleotide-directed in
vitro mutagenesis system (BIO-RAD, Richmond, CA). Chemically synthesized
oligonudeotides (18-mer to 28-mer) were used; fire sequences are shown in
Table 1. The Apa I-Bgl II segment of the prAP164-UC was ligated into M13mp19
usingXba I and Barn HIrestriction sites. (The Xbalsitewas end filledwifir
Klenow, producing a blunt end to complement fire blunt end produced byApa I,
and the Barn HI overhang is compatible with Bgl II). Uracil-containing single
strand DNA was synfiresized and used as fire template. An oligonucleotide
complementary to fire region to be altered, except for fire mismatch, was

54

hydridized to fire single-strand uracil DNA. The complementary strand was
firer synfiresized by T4 DNA polymerase using the oligonucleotide as primer.
Ligase was used to seal fire new strand. The double-stranded DNA containing
fire mutation of interest was transformed into E. coli TG-l. The mutations were
confirmed by DNA sequercing, by fire dideoxy chain termination technique of
Sange' et al. (Sanger et al., 1977), using fire Sequerase Version 2.0 kit (U .8.
Biochemical Corp., OH). The sequencing reaction was primed by annealing
17-mer synfiretic Oligonucleotides. The double stranded mutant DNA was firer
introduced back into prAP164-UC using Apa I and Bgl II restriction erzyme sites.
Analysis of hydrolysis products - Enzyme samples (0.05 Units) were
incubated in fire presence of various oligo- or polysaccharide substrates
(1.0% w/v) in 50 mM acetate buffe' (pH 6.0) wifir 5 mM CaClz at 60°C for
16 hours (unless otherwise indicated), and the products were analyzed by High
Performance Ion Chromatography (HPIC) using a CarboPac PA1 column
(Dionex BioLC4500i) and a pulsed amperome'ic detector (PAD, Dionex).
Solution A (100 mM N aOH in water), solution B (1.0 M N a-acetate in solution A),
and solution C (water) were used for elution. All fire solverts wee prepared
with Milli-Q wate', and filtered firrough a polyvinylidene difluoride membrane
filter (0.22um) (Gelman Sciences, Inc.). The samples wee eluted at a flow rate of
1.0 ml/ min wifir solution A (30%-10%), solution B (040%). and solution C (70%)
ove' 25 minutes. Twerty five microlites of 0.02% sample solution was injected
into fire column for analysis.

RESULTS

Subcloning and reconstruction of recombinant amylopullulanase - In order to
preserve stability of the erzyme during purification, a new subdone
(prAP164-UC) was constructed from fire original clone (pAPZ72)

(Mafirupala et al., 1993). Eliminafion of the 331 nt upstream from the GTG start
codon and fire 31 amino acid signal sequerce presert in pAPZ72 resulted in
increased stability of fire erzyme (data not shown). An Eco RI site was
introduced immediately preceeding fire first coding amino acid, and a Bam HI
site immediately following the TGA stop codon by PCR amplificafion from
Thermoanaerobacter ethanolicus 39E chromosomal DNA. The PCR product was
introduced into fire polylinker of pUC18 and the validity of fire consfi'uct
confirmed by sequerce analysis and restriction erzyme mapping.

Identification of conserved amino acid residues constituting the active center of
amylopullulanase - Four highly conserved regions presert in amylosaccharidases,
which are essertial for fire function of firese enzymes, have beer idertified
Oesperser et al., 1993;1(uriki and Imanaka 1989; Matuschek 1994;

Nakajma et al., 1986). Alignmert of amino acid residues in fire highly conserved
regions of Taka amylase-A (Matsurra et al., 1983) and other saccharolytic
erzymes (Table 2) was used for analysis of amylopullulanase. Asp206, Glu230
and Asp297, located in consersus regions II, III, and IV respectively, were
proposed as catalytic residues of Taka amylase-A. His-122, Asp206, Ly3209,
HileO, and His296, located in consensus regions I, H, and IV respectively, were
proposed as substrate-binding residues (Matsurra et al., 1983). g

It has been suggested previously (Mafirupala et al., 1993) and reported
here that Asp597, Glu626, and Asp703 of amylopullulanase, which correspond to
Asp206, Glu230, and Asp297 of Taka amylase-A, might act as catalytic sites.

55

56

Table 2 Consensus sequences in the alpha-amylase family. italicized: catalytic
residues, bold: substrate specificity residues. N umbeing of fire amino acid
residues starts at the N -te'minus of the mature protein. Undefined residues
correspond to fire split consersus region 11, in erzymes wifir dual activity,
relative to fire consensus for alpha-amylase.

57

Table 2. Consensus sequences in the alpha-amylase family (adapted from

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Matusclreketal., 1994)
CONSENSUS SEQUENCE
Region I Region 11/ Region III Region IV
RegionII'
AMYLASE
car-amylase consensus DAVINI Gm MD PVDNID
TAKAa-amylase 117 DWANI 202 GLRID'I'VKI 230 MD 292 mmw
B. lichemformrs BLMA 245 DAWN! 323 GWRLDVAII 356 m 418 LLDSID
_] i I' 'l
PULLULANASE
Kaerogene 620 DVVYNI 691 6mm 723 m 846 WSKID
K.pneumoniae 610 DWYNI 681 ammo! 713 m 836 WSKID
MAMA-33 339 DAWN! 406 6mm 439 m 522 MCID
ISOAMYLASE
P. anryloderamosa 291 DVVYNI 370 6mm 416 MW 502 F'IDVID
AMYLOPULLULANASE
39B 488 DGVFNI 593 GWRLDVAII/ 626 ELWND 698 LLGSID
668 “PIN
3101 488 DGVFNI 595 GWRLDVAII/ 62'] MD 699 LLGSID
669 “Plum
B6A-RI 487 DGVFNI 590 GWRLDVBII/ 623 ENWGD 695 LLGSID
665 HNPIM
EMl 488 DGVPNI 589 Mill 622 MD 694 1.1.6880
664 HNPIm .
NEOPULLULANASE
B. stearothermophflus 242 DAVPNI 324 GWRLDVAII/ 357 EIWHBD 419 LLGSID
382 W

 

 

 

 

 

 

58

Identification of His493, Asn600, Glu601, and His702 of amylopullulanase,
corresponding to Hi5122, Ly3209, HileO, and Hi5296 of Taka amylase-A, may be
responsible for substrate binding specificity.

Identification of a duplicated region 11 (region 11') in those erzymes wifir
bofir alpha-1,4 and alpha 1-6 hydrolytic activity (Matuschek et aL, 1994) may be
significant for fire dual specificity of firese erzymes. Mutational expeimerts
wifir Asp672 (region II') of amylopullulanase, corresponding to Asp206
(region 11) of Taka amylase-A, and Asp675 and Lys676, corresponding to Ly8209,
HileO of Taka amylase-A will be conducted to analyze fireir effect on catalytic
activity and substrate binding specificity respectively.

Based on amino acid comparison, we idertified amino acid residues
Asp597, Glu626, Asp703, I-Iis493, Asn600, Glu601, His702, Asp672, Lys675, and
Leu676 as targets for mutageresis to analyze fire active center of
amylopullulanase (Table 3).

Time course of substrate hydrolysis by wild type amylopullulanase - The
hydrolysis pattern of amylopullulanase from Thermoanaerobacter ethanolicus 39E
was tested on many low and high molecular weight oligosaccharides.
Maltotriose, maltotetraose, maltopertaose, maltohexaose, maltoheptaose,
pullulan, and soluble starch were digested wifir amylopullulanase, and fire
products analyzed by HPIC (Dionex). A 200 ul reaction mixture containing
1.0% (w/ v) fire indicated substrate in 50 mM N a-Acetate buffer (pH 6.0) wifir
5 mM CaClz and 0.05 Units erzyme was incubated at 60°C for 30 minutes. The
reaction was stopped by incubating at 100°C for 15 minutes and diluted 1:1500
before HPIC analysis. HPIC profile of glucose, maltose, maltotriose,
maltotetraose, maltopertaose, maltohexaose, and maltoheptaose (GI-G7)
standard mixture is shown in Figure 2.

59

Table 3. Amylopullulanase mutants generated by site directed mutagenesis

 

CONSERVED REGIONS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I II III IV
Wild Type 488 DGVFNI 593 GWRLQVAII 626 ELWND 698 men
81349361n —Q
Mp597Leu ....._;,__
MnGOOSer/ —-—8V
GluGOIVal
Glu626up l2—
Hie‘lOZArg —I—
813702Asp ___o.
8187026111 _'.
813702Lye —-G-
Alp703G1y __9
11'
Wild Type 488 DGVFNI 668 HNIPDAAKL 626 ELWND 698 LLGSID
MpG'IZAsn -—n—
up67261u —3—
Asp672Phe ——n—
Lys675Arg ——n—
Ly8675Aap _—_n_
Leu676A1a ——l

 

 

 

 

 

 

 

underlined: Active site amino acid residues

bold : Substrate binding site amino acid residues

 

60

Figure 2 HPIC profile showing standard mixutre of Gl—G7. A carbopak-
column (0.4 x 25 cm, 10mm of particle size, Dionex) and a PAD detector wee
used. A buffer (100 mM NaOH in water) and B We (1.0 M Na-acetate in A
buffer) were used for elution as described in Mateials and Mefirods.

61

Figure 2. HPIC profile showing standard mixture of Gl-G7.

 

 

 

 

 

 

 

 

 

-20.000mV 200.000mv
1.591 55.05 N
k
r
L
9.450 Glucose 1089.30 H
11.73} harm. 372.00 :1
12.700 haltotriose 344.20 M
13.841' paltotetraose 355.75 I
14.883> haltopentaose 320.75 I
15.825L Haltohexaose 275.20 H
16.633 Faltoheptaose 257.68 M
L
L—
g
l i
i 5
l i
I
l :
i \
!

 

62

Degradation of 1.0% soluble starch by wild type Apu was monitored over
a peiod of 96 hrs (Table 4). The major hydrolysis products were maltose and
maltotriose. As fire reaction proceeded maltotetraose was completely degraded,
maltose and maltotriose remained unchanged, and fire levels of glucose
accumulated. This erd product profile resembles that exhibited by thermophilic
alpha-amylases. The percert of maltotriose remained constant firroughout fire
time course which is consistent with fire hydrolysis results of maltotriose,
maltohexaose, and pullulan producing maltotriose as fire only product by action
of alpha-1,4 (maltotriose and maltohexaose as substrates), and alpha-1,6
(pullulan as substrate) hydrolytic activity.

Time course of maltoheptaose hydrolysis revealed maltotriose and glucose
as fire only hydrolysis products present at 80% and 20% respectively firroughout
fire reaction (Table 4). Maltoheptaose is hydrolyzed producing two maltotriose
and one glucose. Anofirer posibility for cleavage activity on maltoheptaose
would be production of maltotetraose and maltotriose . This is not likely due to
fire abserce of maltotefi'aose as an end product after the 16 hr reaction.

Ofirer substrates examined include maltotriose, maltohexaose, and
pullulan, all of which produced maltotriose as fire only hydrolysis product
(Table 5). Amylopullulanase has no cleavage activity on maltotriose. The
activity on maltohexaose is amylase like, while firat on pullulan is characteristic
of a pullulanase, bofir yielding maltotriose as fire sole product which can not be
furfirer hydrolyzed. Maltopertaose produced maltotriose and maltose each
present at roughly 50% after a 16 hr. reaction (Table 5). Hydrolysis of beta-limit
dextrin produced maltose, maltotriose, maltotetraose, maltopertaose, and ,
maltoheptaose as end products (Table 5). The maltopertaose and maltoheptaose

63

Table 4 Time course of reaction products produced from wild type
amylopullulanase on various low and high MW oligosaccharides.
Solutions of 1.0% pullulan, soluble starch, maltotriose (G3), maltohexaose
(G6), and maltoheptaose (G7) were incubated at 60°C wifir 0.05 U/ ml
partially purified erzyme. Samples were wifirdrawn at various time
points and heated at 100°C for 15 minutes for enzyme inactivation. The
reaction products were analyzed by HPIC for sugars (Dionex).
Abbreviations; glucose (G1), maltose (G2), and maltotetraose (G4).
Numbers indicate percert molar equivalerts relative to standards and
results after triplicate analysis.

 

64

Table 4. Time course of reaction products produced from wild type
amylopullulanase on various low and high MW oligosaccharides.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SUBSTRATE PRODUCT PERCENT OF TOTAL
1 hr 75: 16hr 24hr 96hr

Soluble CI 2.7 6.4 10.1 15.6 17.6
Starch 62 53.0 57.5 56.4 52.7 51.0
G3 332 28.1 29.2 28.7 30.7

G4 11.1 7.9 4.3 3.0 0
Maltoheptaose CI 20.3 18.7 18.5 17.4 20.2
(33 79.7 81.3 81.5 82.6 79.8
Pullulan 63 100 100 100 100 100
Maltotriose G3 100 100 100 100 100
Maltohexaose G3 100 100 100 100 100

 

 

 

 

 

 

 

 

65

Table 5 Comparison of end products from enzymatic action of
His493Gln, Asn600$erlGlu601Val, and wild type enzymes on low and
high MW oligosaccharides. Solutions of 1% pullulan, soluble starch,
beta-limit dextrin, maltotriose (G3), maltotetraose (G4), maltopentaose
(G5), maltohexaose (G6), and maltoheptaose (G7) were incubated at 60°C
wifir 0.05 U / ml parfially purified erzyme. Samples were wifirdrawn after
16 hours and heated at 100°C for 15 minutes for erzyme inactivation. The
reaction products were analyzed by HPIC for sugars (Dionex). Numbers
indicate percert molar equivalents relative to standards and results after
triplicate analysis.

66

Table 5. Comparison of end products from enzymatic action of His493Gln,
AsnGOOSer/GluGOIVal, and wild type enzymes on low and high MW
oligosaccharides.

A. WildType

Substrate End Products

 

B. Hiu93GIn
Substrate End Products

 

C. AsnGOOSer/GluflnVal
Substrate End Products

 

67

produced is indicative of debranching at alpha-1,6 branch points. A model of
Apu action pattern is proposed in Figure 3.

Amino acid replacement by site-directed mutagenesis and construction of mutated
amylopullulanase - The Apa I-Bgl II fragment of prAP164-UC (Figure 1) contains
fire four highly conserved regions which most likely constitute the active certer
of amylopullulanase. Subcloning into fire M13mp18 multiple-cloning site,
propagation of single stranded DNA, and subcloning back into pUC18 was
performed out as described in fire expeimertal procedures. The
1900 bp fragmert was sequerced to veify fire preserce of fire desired amino acid
change and confirm firat no second-site mutations were presert.

E. coli TG-l carrying fire wild type or mutant plasmids was grown in LB
and fire partially purified erzyme sample was prepared. Kinetic expeimerts
were pe-formed using fire standard assay at a variety of substrate (pullulan,
soluble starch, amylose, and beta-limit dextrin) concertrafions. Values for
Kmapp and Vmaxapp dete'mined from Lineweaver-Burk plots showed no
significant diffeerce between firose mutants which were active when compared
to fire wild type erzyme (Table 6).

Activity of mutated amylopullulanases - In a previous study, firree anrino
acids important for catalytic activity were idertified by sequerce aligrrmert wifir
alpha-amylase (T aka-amylase) from Aspergillus oryzae (Mafirupala et a1 ., 1993;
Matsurra et al., 1983). Due to fire reconstruction of recombinant
amylopullulanase mertioned earlier, mutational analysis of each of fire firree
catalyfic amino acid residues was repeated. Single mutants of Asp597 (in
conse'ved region II), G1u626 (in conserved region III), and Asp703 (in conserved
region IV), resulted in complete loss of erzymatic activity as seer previously
(Table 7).

68

Figure 3. Proposed action pattern of amylopullulanase. Oper circles
and slashed circles denote non-reducing glucopyranosyl residues and
reducing glum residues, respecfively. Horizontal lines and vertical lines
connecting circles indicate alpha-1,4 and alpha-1,6 glucosidic linkages
respectively.

69

Figure 3. Proposed action pattern of amylopullulanase

A. prim

W802... _.

rm to exact. ——’
maltotriose 0.0.“ —-—

W" 03-“ 0%

 

 

c m «can , a. Q
—->°' 0e
‘——>°' Doe

d.

 

 

 

E mdoportaoae a.
00-0sz ‘4. 06-5.?
F. mlotetraoae a, a
oooe a .. ooe

 

G. beta-llmlt dextrin

0000003830024? W
“0000003

 

70

Table 6 Kinetic analysis of wild type and mutant amylopullulanase on
various substrates. Assays to determine Kmapp and Vmaxapp were
conducted at 60°C wifir substrate concertrations betweer 20 x Kmapp and
0.2 x Kmapp. All assays wee done in triplicate.

71

Table 6. Kinetic analysis of wild type and mutant amylopullulanase on

 

 

 

 

 

 

 

various substrates.

Enzyme Kmapp (mg/ ml) / Vmaxapp (U nits/ mg)
Wild Type 0.36/ 32.76 0.87/ 25.20 0.90/ 28.14 0.92/ 27.98
His493Gln 0.32/ 35.96 0.93/ 25.26 0.88/ 36.14 1.05/ 32.17
Asn6005er/ GAO/38.74 101/2702 097/2128 LOO/30.61

Glu601Val

Asp672Asn 0.30/ 33.63 0.92/ 13.07 0.91/ 14.43 1.20/ 19.31
Asp672Glu 0.33/ 25.04 0.90 / 14.19 0.97 I 16.21 1.00/ 19.03

 

 

 

 

 

 

72

Table 7 Activity comparisons of Asp597Leu, Glu626Asp, and
Asp703Gly.

Rate of product formation represerts fire Units/ mg on pullulan, amylase,
soluble starch, and B—limit dextrin unde' standard assay conditions. One
unit is defined as fire amount of erzyme which produces 1.0 umol of
reducing sugar / min wifir glucose as fire standard. All assays wee done
in triplicate. P/ A rafio, pullulanase] amylase activity ratio.

73

Table 7. Activity comparisons of Asp597Leu, Glu626Asp, and Asp703Gly.

 

 

 

 

 

Rate of Product Formation
Substrate:
Wild Type 28.83 13.17 13.68 12.53 2.54
Asp597Leu < 0.001 < 0.1111 < 0.011 < 0.001 -
Glu626Asp < 0.001 < 01D] < 0.001 < 0.001 -
Asp703Gly < 0.11” < 0.001 < 0.001 < 0.001 -

 

 

 

 

 

 

 

74

Wher the putative substrate binding site residues of amylopullulanase
were mutated, the rate of product formation increased relative to wild type wifir
mutants I-Iis493Gln and Asn6005er/ Glu601Val. By contrast the mutants
His702Arg, His702Asp, His702Lys were void of cleavage activity on pullulan,
amylose, soluble starch, and beta-limit dextrin (Table 8). Mutafing His702 (in
conse'ved region IV) to Arg, Asp, and Lys resulted in complete loss of activity.
The His493G1n mutant and fire Asn6005e'l Glu601Val (corresponding to putative
substrate binding residues His247 and Asn331/Glu332 respectively in
neopullulanase), bofir resulted in an increased rate of alpha-1,4 and alpha-1,6
cleavage acfivity.

End product formation from low and high MW oligosaccharides using wild type
and mutated amylopullulanase - Incubation of wild type and mutant enzymes
wifir low-molecular weight oligosaccharides, pullulan and soluble starch resulted
in producfion of various oligosaccharide erd products, depending upon fire
substrate used. Amylopullulanase does not hydrolyze maltotriose. The only
product obtained upon incubation wifir pullulan was G3, demonstrating firat fire
erzyme shows typical pullulanase activity. Activity toward alpha-1,4 linkages is
seer as well, demonstrating amylase activity. The pe'cert of each erd product
produced diffes wher comparing wild type and mutant enzymes (Table 5).

Twerty-five percert of fire maltotetraose substrate remained
unhydrolyzed after 16 hour irrcubafion wifir wild type Apu, by contrast wifir fire
I-Iis493Gln mutant only 10.2% maltotetraose remaining. The furfirer cleavage to
produce G1/ G3 products are 20.3% /54.3% and 49.4% / 40.5%, respectively, for
wild type and mutant erzymes. Maltoheptaose as fire substrate produces 18.5%
G1 and 81.5% G3 for wild type erzymatic activity and 20.4% G1 and 79.6% G3
for fire His493G1n erzyme. The I-Iis493G1n erzyme has a higher breakdown to
G1 and G3 when maltotetraose is used as fire substrate indicating an increased

75

Table 8 Activity comparisons of I-Iis493Gln, Asn600$erlGlu601Val,
I-Iis702Arg, His702Asp, and I-Iis702Lys. Rate of product formation
represents fire Units/ mg on pullulan, amylose, soluble starch, and B-limit
dextrin under standard assay conditions. One unit is defined as fire
amount of erzyme which produces 1.0 umol of reducing sugar/ min wifir
glucose as fire standard. All assays wee done in triplicate. PIA ratio,
pullulanase/ amylase activity ratio.

76

Table 8. Activity comparisons of I-Iis493G1u, AsnGOOSer/GluGOIVal,
His702Arg, His702Asp, and His702Lys.

 

 

 

 

 

 

 

Rate of product formation
Substrate:
Wild type 28.83 13.17 13.68 12.53 2.54
His493Gln 39.99 20.61 20.36 17.41 2.10
Asn6005er/ 50.90 15.61 13.30 16.64 3.81
Glu601Val
His702Arg < 01111 < 0.001 < 0.001 < 0.001 -
His702Asp < 0.001 < 0.001 < 0.001 < 0.001 -
His702Lys < O.(XJ1 < 0.001 < 0.1111 < 0.1111 -

 

 

 

 

 

 

 

77

alpha-1,4 cleavage specificity. In contrast, fire Asn6005er/ Glu601Val enzyme
revealed a decreased alpha-1,4 cleavage specificity on maltoheptaose due to fire
lower production of G1 and G3 from hydrolysis activity. G1 and G3 produced
from action of Asn6005er/ Glu601Val on maltoheptaose are 14.1% and 85.9%,
respectively. Activity on maltotetraose is slighfiy higher firan fire wild type
erzyme but lower firer fire His493Gln erzyme, yielding 30.0% G1, 46.7% G3, and
23.3% remaining maltotetraose firat was not broker down after fire 16 hour
reaction (Table 5).

End product formation from cleavage activity on soluble starch provided
fire most interesting results. Wild type amylopullulanase produced 10.1% G1,
56.4% G2, 292% G3, and 4.3% G4 yielding a higher conversion to smaller
oligosaccharides firan fire His493G1n or Asn6OOSer/ Glu601Val mutant erzymes.
His493Gln produced 4.0% G1, 23.1% 62, 58.2% G3, and 14.7% G4; while fire
double mutant produced 46.8% G2, 35.2% G3, and 18.0% G4 wifir no detectable
production of G1 before fire completion of fire reaction (Table 5).

Action of fire wild type erzyme, fire His493Gln erzyme, and fire
Asn6005er/ Glu601Val erzyme produced an equal distribution of G2 (50%) and
G3 (50%) from hydrolysis of maltopentaose (Table 5). Maltotriose was produced
from activity of fire above erzymes on maltohexaose and pullulan wifir no
furfirer breakdown of maltotriose to maltose or glucose (Table 5).

The use of beta-limit dextrin as substrate produced no significant change
in hydrolysis products wher comparing His493Gln and Asn600$erl Glu601Val
mutants to wild type amylopullulanase. All produced statistically equivalert
amounts of maltose, maltotriose, maltotetraose, maltopertaose, and
maltoheptaose (Table 5). The rate of activity of I-Iis493Gln,

Asn6OOSer/ Glu601Val and wild type erzyme on beta-limit dextrins was sinriliar
to that observed for amylase or soluble starch (Table 8).

78

Analysis of region H ’ for altered catalytic or cleavage activity - The mutafions
we have constructed in fire duplicated region H (region H' - Table 2) have yet to
be analyzed in any amylolytic erzyme system. In enzymes wifir dual specificity,
the separation of region H into two discrete regions (H and H') might allow for
hydrolysis of bofir alpha-1,4 and alpha-1,6 linkages due to increased flexibility of
fire active certer. Site-directed mutants were constructed to determine if fire
separation of region II is important for catalytic activity and substrate cleavage
specificity.

We have shown previously (Table 7), firat Asp597 (region H) was esserfial
for catalytic activity of amylopullulanase. Asp672 (region H') in
amylopullulanase corresponding to catalytic residue Asp206 (region H) of Taka
alpha-amylase was mutated to Asn and Glu wifir sirrriliar reductions in rate of
activity on pullulan, amylose, soluble starch, and beta-limit dextrin; 5.7x, 7.9x,
7.5x, and 7.4x, respectively (Table 9).

We furfirer examined fire role of Lys675 and Leu676 of amylopullulanase
in region H' to determine if firese residues provide fire same function in substrate
binding as Ly3209 and HileO in region H of Taka alpha-amylase. Lys675 has
beer substituted wifir Arg, Asp and Ala; and Leu 676 was mutated to His,
similiar to firat presert in position 210 of Taka alpha-amylase. Thee were no
differerces observed in product formation or rate of activity wifir various
substrates wher compared to wild type amylopullulanase (data not shown).

79

Table 9 Activity comparisons of Asp672Asn and Asp672Glu. Rate of
product formation represents fire Units/ mg on pullulan, amylose, soluble
starch, and B-limit dextrin under standard assay conditions. One unit is
defined as fire amount of erzyme which produces 1.0 umol of reducing
sugar/ min wifir glucose as fire standard. AH assays wee done in
triplicate. P/ A ratio, pullulanase] amylase activity ratio.

 

80

Table 9. Activity comparisons of Asp672Asn and Asp672Glu.

 

 

 

 

Rate of product formation
Substrate:
Wild type 28.83 13.17 13.68 12.53 2.54
Asp672Asn 7.52 4.32 4.57 4.32 3.68
Asp67ZGlu 7.52 4.32 4.44 4.70 3.85

 

 

 

 

 

 

 

DISCUSSION

We tested fire effect on activity, subsfi'ate preference, and end product
formation by site—directed mutagenesis of conserved amino acid residues in
consersus regions I - N and duplicated region H' of amylopullulanase from
T. ethanolicus 39E. Catalytic activity and erd product formation by alpha-1,4
and/ or alpha-1,6 hydrolyfic activity of various mutants were alteed when
compared to the wild type amylopullulanase. Substrates used to analyze
hydrolytic activity included: pullulan, amylose, soluble starch, and beta-limit
dextrin.

Glu626, Asp703, and Asp597 of amylopullulanase correspond to fire
catalytic residues Glu230, Asp297, and Asp206 of Taka amylase-A
(Matsuura et al., 1983). Wher firese three residues were singly replaced wifir
Asp, Gly, and Leu, respectively, catalytic activity on pullulan, amylose, starch
and beta-limit dextrin was not detectable (Table 6). Therefore, amino acid
residues Asp597, Glu626, and Asp703 are absolutely required for catalytic
activity.

Several mutated amylopullulanases were ergineered by substituting
amino acid residues corresponding to fire substrate-binding residues of Taka
amylase-A. Some of the mutants showed altered alpha-1,4 and/ or alpha-1,6
cleavage specificity wher compared to the wild type amylopullulanase.
Mutating I-Iis702 (in conse'ved region IV) to Arg, Asp, and Lys resulted in
complete loss of activity. The same result was seen wher fire corresponding
amino acid (His423) from neopullulanase was mutated to Arg, Asp, and Lys
(Kuriki et a1 ., 1991). Mutating His702 to Arg, Asp, and Lys may abolish catalytic
activity due to its potential inte'fererce in the acid-base mechanism of one of fire
anrino acid residues (Asp703) absolutely required for catalysis. The His493Gln

81

82

mutant and fire Asn6005er/ Glu601Val (corresponding to His247 and
Asn331/Glu332 respectively in neopullulanase), bofir resulted in an increased
rate of alpha-1,4 cleavage activity. Similiar results were seen wifir fire above
neopullulanase mutants (Kuriki et al., 1991) and fire HisZSOGln mutant in

B. licheniformis alpha-amylase (Cheong 1995).

The I-Iis493G1n Apu shows more efficiert cleavage activity wher G4 was
used as fire substrate, yielding higher percertages of G1 / G3 after a 16 hr reaction
relative to wild type. The efficiercy is reduced wher soluble starch is used as fire
substrate yielding lower percertages of G1, G2,and G3 wifir more G4 remaining
after fire reaction was terminated. The neopullulanase mutant (His423Gln)
showed an increased production of panose (Kuriki et al., 1991), while fire alpha-
amylase mutant (His250Gln) showed similiar percertages of panose production
(Cheong 1995) relative to the wild type erzymes wifir pullulan as fire substrate.

The amylopullulanase double mutant Asn6OOSer/Glu601Val erzyme had
decreased cleavage specificity for alpha-1,4 linkages on soluble starch as
eviderced by fire lack of G1 producfion and more G4 remaining after fire
hydrolysis reaction (Table 5). The active certer of amylopullulanase is similiar to
that of neopullulanase, but their cleavage patte'rr on pullulan is completely
diffe'ert. Mutational analysis of Asn331$erl Glu332Val from neopullulanase
indicated a decrease in alpha-1,4 cleavage specificity and an increase in alpha-1,6
specificity as eviderced by fire producfion of less panose from pullulan firan fire
wild type erzyme (Kuriki et al., 1991). The mutant Asn33tBer/ Glu331Val from
alpha-amylase increased the alpha-1,4 hydrolytic specificity and erdowed fire
amylase wifir alpha-1,6 activity similar to firat of neopullulanase on pullulan.
These two amino acids wee firought to be related to alpha-1,6 hydrolytic activity
and specificity. A dramatic increase in the rate of alpha-1,4 and alpha-1,6
catalytic activity was seer when mutating amino acid residues 493 and 600/ 601

83

of amylopullulanase. It is difficult to examine fire specificity wifir
amylopullulanase using pullulan as fire substrate due to maltotriose being fire
only product produced from hydrolysis which can not be furfirer cleaved.

Mutations constructed from amino acid residues presert in the duplicated
region H (H') showed no differerce in fireir erd product formafian on low and
high molecular weight oligosaccharides. Thee was a significant decrease in fire
rate of product formation using pullulan, amylase, soluble starch, and beta-limit
dextrin as fire substrates wifir fire Asp672Asn and Asp672Glu erzymes wher
compared to wild type amylopullulanase. Bafir mutants had a 6x and 8x
reducfion in activity on pullulan and amylase, respectively. The rate of product
formation on soluble starch and beta-limit dextrins decreased similiarly to firat
seen wifir amylase for bofir Asp672Asn and Asp672Glu.

In summary (Table 10), our results show: 1) a loss of catalytic activity
wher each of fire firree conse'ved catalyfic residues are mutated, and
2) a change in hydrolysis pattern Wher mutating firree of fire four conserved
subsfi'ate binding residues. We took firis analysis one step furfirer and idertified
a duplicated region H (H‘) in dual activity erzymes which is sinriliar to region H
in alpha-amylase. Our hypofiresis was firat duplication of region H may be
important for catalytic activity and substrate specificity. Preliminary research
preserted here indicates firat individual amino acids in region H' may be
important for catalytic activity due to the decrease in rate of activity on all
substrates tested, but no difference was seen in substrate cleavage specificity.

84

Table 10. Summary of amino acid alteration in the four conserved regions
with amylopullulanase, neopullulanase, and alpha-amylase.

 

 

 

DAVINB GPRLDAAKB HNPIDm E'VID FVDNBD
£11m AmhraAddAlteredLReult
Apu fi/tnaease alpha-1,4

D/cntalytlcdly inactive
KB lincrease alpha-1,4

n/reducum in catalytic activity
Elsamerate and deavage activity

E/catalytically inactive

D/catalytically
inactive
Neo 8/ increase alpha-1,4
D/catalytically inactive
m / increase alpha-1,6
E/ catalytically inactive
I / increase
alpha-1,4
D/catalytically
inactive
BLMA 8/ increase alpha-1,4
D/catalytically inactive
Kl / increase alpha-1,4
endowed alpha-1,6
D/ catalytically

inactive

85

Future expeimerts will focus on a strategy to conve't the dual activity erzyme,
amylopullulanase, to an erzyme wifir only alpha-1,4 or alpha-1,6 hydrolytic
activity. The main limitation in fire design of an industrially significant erzyme
is fire inability to oveexpress fire recombinant protein. Thus far, our attempts at
oveexpression of amylopullulanase resulted in cell toxicity and loss of the inset
from fire cloning vector. A yeast oveexpression system is currerfiy being
pursued and, if successful, may allow us to oveexpress the protein and obtain
crystals for three-dimersional structure analysis. We will firer confirm fire
results preserted in firis report by model fitting oligosaccharides into the
catalytic site of the erzyme. This will provide detailed information of fire
binding mode betweer substrate and erzyme, and aid in fire production of
mutant monospecific, firermophilic amylopullulanases for industrial starch
degradation.

ACKNOWLEDGEMENTS
I gratefully acknowledge Dr. Friedman for providing beta-limit dextrins for use
in hydrolysis studies. I also gratefully acknowledge Kevin Carr and Dr. Marty
Regier for assistance in reviewing firis manuscript. I also gratefully acknowledge
Dr. Claire Vieille for helpful discussions. This research was supported by a grant
from fire US. Department of Agriculture, under the agreement 89-34189-4299,
and fire Research Excellence Fund from Michigan State University.

86

REFERENCES

Cheang T.K. (1995) Modulation of B. licheniformis amylase by site-directed and
deletion mutagenesis. Ph.D. Thesis, Seoul National University.

Fogarty, W.M., and CT. Kelly (1979) Starch degrading enzymes of microbial
origin. Prog. Ind. Microbial. 15: 87-150.

Imanaka, T., and Kuriki, T. (1989) Pattern of action of Bacillus stearothermophilus
neopullulanase on pullulan. 1. Bacteriol. 171: 369-374.

lespersen, I-LM., MacGregor, E.A., Sierks, M.R., and Svenssan, B. (1993)
Comparison of fire domain-leval organization of starch hydrolases and related
erzymes. I. Prat. Chem. 12: 791-805.

Kim, CH, and Kim, Y.S. (1995) Substrate specificity and detailed
characteization of a bifunctional amylase-pullulanase erzyme from Bacillus
circulans F-2 having two different active sites on one polypepfide. Eur. I. Biochem.
227: 687-693.

Kim, I.C., Cha, I.H., Kim, ].R., Iang, S.Y., Sea, B.C., Cheong, T.K., Lee, D.S.,
Choi, Y.D., and Park, K11. (1992) Catalytic properties of fire cloned amylase
from Bacillus lichenifarmis. 1. Biol. Chem. 267: 22108-22114.

Kim, I.C., Yaa, S.H., Lee, 5.1., Oh, B.H., Kim, ].W., and Park, K11. (1994)
Synfiresis of branched oligosaccharides from starch by two amylases cloned from
Bacillus lichenijbrmis. Biosci. Biotech. Biochem. 58: 416-418.

Kuriki, T., and Imanaka, T. (1989) N ucleofide sequerce of fire neopullulanase
gere from Bacillus stearothermophilus. 1. Gen. Microbiol. 135: 1521-1528.

Kuriki, T., Takata, H., Okada, S., and Imanaka, T. (1991) Analysis of the active
center of Bacillus stearothermophilus neopullulanase. I. Bacteriol. 173: 6147-6152

Kuriki, T., Okada, S., and Imanaka, T. (1988a) New type of pullulanase from
Bacillus stearothermophilus and molecular cloning and expression of fire gere in
Bacillus subtilis. I. Bacterial. 170:1554-1559.

Kuriki, T., Yanase,M., Takata, I-I., Imanaka, T. andS. Okada. (1993)]. Fermen.
Bioerginer. 76: 184-190.

Laemnrli, U. K. (1970). Cleavage of structural proteins during fire assembly of fire
head of bacteiophage T4. Nature (London) 227: 680-685.

87

88

Lee, S.]., Yaa, S.H., Kim, M.I., Kim. I.W., Seok. R.M., and Park, K.H. (1995)
Production and characteization of branched oligosaccharides form liquified
starch by fire action of B. licheniformis amylase. Starch . 47: 127-134.

Mathupala, S.P., Lowe, S.E., Padkavyrov, S.M., and Zeikus, I.G. (1993)
Sequencing of fire amylopullulanase (apu) gene of Thermoanaerobacter ethanolicus
39B, and iderfification of the active site by site-directed mutageresis. I. Biol.
Chem. 268: 16332-16344.

Mathupala, S., Saha, B.C., Zeikus, I.G. (1990) Substrate competifion and
specificity at fire active site of amylopullulanase from Clostridium
thermohydrosulfuricum. Biochem. Biophys. Res. Comm. 166, 126-132.

Matsuura, Y., Kusunaki, M., Harada, W., and Kakuda, M. (1983) Structure and
possible catalytic residues of Taka-amylase A. I. Biochem. 95: 697-702.

Matuschek, M., Burchhardt, G., Sahm, K., and Bahl, H. (1994) Pullulanase of
Thermaanaerobacterium thermosulfurigenes Eml (Clostridium thermosulfurogenes):
molecular analysis of fire gene, composite structure of fire enzyme, and a
common model for its attachmert to fire cell surface. I. Bacteriol. 176: 3295-3302.

Miller, G.L. 1959. Use of dinitrosalicyclic acid reagert for determination of
reducing sugar. Anal. Chem. 31: 426.

Nakajima, R., Imanaka, T., and Aiba, S. (1986) Comparison of arrrino acid
sequerces of elever diffe'ert alpha-amylases. Appl. Microbiol. Biotechnol.
23: 355-360.

Plant, A.R., Clemens, R.M., Morgan, H.W., Daniel, R.M. (1987) Active-site and
substrate-specificity of Thermoanaerobium Tok6-Bl pullulanase. Biochem I.,
247: 537-541

Sakana, Y., Masuda, N., and Kabayashi, T. (1971) Hydrolysis of pullulan by a
novel erzyme from Aspergillus niger. Agric. Biol. Chem, 35: 971-973.

Sambrook, I., Fritsch, E.F., Maniatis, T. (1989) Molecular Cloning: A Laboratory
Manual (Second Edition).

Sanger, F., Nicklen, S., and Caulsan, A.R. (1977) DNA sequencing wifir chain-
ternrinating inhibitors. Proc. Natl. Acad. Sci. U.S.A. 74: 5463.

Takata, H., Kuriki, T., Okada, S., Takesada, Y., Iizuka, M., Minamiura, N ., and
Imanaka, T. (1992) Action of neopullulanase. Neopullulanase catalyzes bofir
hydrolysis and transglycosylation of alpha-(1,4) and alpha-(1,6) glucosidic
linkages. I. Biol. Chem. 267: 18447-18452.

Chapter HI

Molecular analysis of thermophilicity and thermostability of recombinant
amylopullulanase from Thermoanaerobacter ethanolicus 39E.

Prepared for submission to Biachirrrica et Biophysica Acta

89

ABSTRACT

Thermophilicity and firermostability of amylopullulanase from
Thermaanaerobacter ethanolicus 39E were analyzed by N- and C-te'minal deletion
mutageresis. Sequence and HCA comparison of amylopullulanase from
T. ethanolicus 39E to organisms wifir lower firermostability showed high sequerce
homology and secondary structure similarity. Deletion mutant construction has
idertified a region related to firermophilicity at fire N-terrninus of T. ethanolicus
39E amylopullulanase. The region corresponding to anrino acid residues 107-324
is important for maintaining activity at fire enzyme's optimal temperature. The
wild type has an optimum tempeature of about 85°C while fire ApuN324
deletion construct has an optimum temperature of about 65°C wifir a broad
temperature range from 45°C to 93°C. An Arrherius plat for wild type
amylopullulanase was linear; however, fire observed plot for ApuN324 was
discontinuous, probably due to a temperature dependert structural change
occurring at 50°C. The Kmapp values of fire wild type and ApuN324 erzymes
were similar at bofir 40°C and 60°C; while fire Vmaxapp was lower for bofir
enzymes at 40°C relative to firat at 60°C. To our knowledge, firis report is fire
first to identify a change in themophilic characteistics of an enzyme. It is
hypofiresized firat fire flexibility of fire mutant protein was increased to account
for a 20°C lower temperature optimum because fire Krrrapp value was not alteed
relative to fire wild type erzyme.

9O

INTRODUCTION

Amylopullulanases from firermophiles are as active and stable above 90°C
as commercial Aspergillus oryzae alpha-amylase, thus having great patertial for a
single step liquifaction and saccharifaction process producing high
oligosaccharide syrups from starch. Attempts have beer made to understand fire
property of erhanced firermal stability inherert in the structures of firermaphilic
enzymes in comparison to homologous mesophilic erzymes.

Initial studies on increased temperature optima for erzyme activity from
firemaphilic organisms, wher compared to similar mesophilic erzymes, focused
on diffe'erces in fire arrrino acid composition. A recert report by Iaericke and
Bohn (1994) showed firat a shift in fire amino acid composition was not erough to
explain fire mechanism of themophilic erzyme adaptation. Extersive
comparison of the tertiary structures (Blake et al., 1993 and Bradley et al., 1993) of
mesophilic and firermaphilic erzymes indicated firat firee is no large detectable
structural diffeerce (T omazic and Klivanov 1988). It has also beer shown firat
hydrophobicity is fire dominant force of protein folding, whereas ion pairing,
hydrogen bonding, and Van der Waals interactions are of less importance but
may affect stability (Dill et al., 1990).

Hydrophobicity and increased packing of residues in fire protein core are
gererally accepted to be a major stabilizing factor in fire'mophilic proteins
(V ieille et al., 1996). For increased stability, fire positioning of fire residues wifirin
the inte'ior of fire protein is more important, due to geometrical considerations
(Sandberg and Terwilliger 1989) yielding increased protein rigidity
(Iaenicke 1991; Eijsink et al., 1992), firan fire total content of hydrophobic residues.

Many properties of proteins arise from two extremes in fireir folded
conformation, flexibility (increased molecular motion) and rigidity (resistance to

91

92

unfolding). The firermophilicity of erzymes may be closely related to flexibility
and rigidity of fire protein's folded conformation first is essential for preserving
their catalytically active structure. Flexibility is essertial for erzyme function
while rigidity is necessary for maintaining fire globular structure. Thermostable
erzymes are firought to have a more rigid structure firan firat of mesophilic
erzymes. This has beer addressed by utilizing hydrogen (deuteium) exchange
expeimerts. Expeimerrts done by Tsuboi et a1 (1978) wifir elongation factor Tu
showed a reduced rate of exchange wifir fire erzyme from Thermus thermophilus
relative to fire erzyme from E. coli (Fontana 1990). The higher fire exchange rate
fire higher fire flexibility of fire protein. The flexibility of firermophilic and
mesophilic erzymes was equivalert at fireir respective optimum temperatures
(Tsuboi et al., 1978). The rigidity of firermozymes has also beer demonstrated by
lowe' susceptibility to proteolytic degradation and chemical or firemal
denaturant unfolding (Wrba et al., 1990; Kanaya and Itaya 1992).

To examine features responsible for firermophilicity, we constructed and
analyzed deletion mutants of T. ethanolicus 39E amylopullulanase expressed in
E. coli . Irrfornration preserted here provides eviderce to support fire hypofiresis
firat the temperature activity optimum is related to optimal erzyme flexibility.

EXPERIMENTAL PROCEDURES

Reagents, Enzymes, and Oligonucleotides - All chemicals were of molecular
biology or analytical grade and obtained from Aldrich Chemical Co., or Sigma.
Restriction erzymes and ligase were obtained from Befiresda Research
Laboratories, United States Biochemical Co., or Boehringer Manheim.

Bacterial strains, plasmids, and transformations - E. coli strain SURE {e14-
(mcrA) (mchB-hstMR-mrr)171 endAI supE44thi-1 gyrA96 relA1 lac recB recI
sbcC umuC:Tn5(kan1’)uvrC [F' proAB lacI‘lZ M15Tn10(tet")]} from Strategere Co.
and DHScr-F' {F' 80dlac (lacZYA-argP)U169 deaR recAI endAI hstI 7(rk-
mk+)supE44 I- thi-l gyrA96 relA1}, from Befiresda Research Labs.) were used as
hosts for cloning, and E. coli BMH 71-18 mutS, IM109 and E31301 mutS from
Promega Corporation were used. E. coli strains were made competert by fire
Hanahan mefirod as described by Perbal (Pebal 1988), while recombinant
vectors were introduced in E. coli strains by heat-shock treatment
(Hanahan 1983).

Enzyme assays - For determination of amylopullulanase activity, 160 ul of
1.25% (w/v) pullulan in 50 mM acetate buffer (pH 6.0) containing 5 mM CaClz
and erzyme (heat treated culture supernatant) to a total volume of 200 pl were
mixed and incubated at 60°C for 30 min. The reaction was stopped by adding
0.8 ml of dinitrosalicylate solution [0.25 M NaOH, 71.0 mM sodium potassium
tartrate, 4.0 mM Na2903, 5.0 mM phenol, and 44.0 mM 3,5 dinitrosalicylic acid]
(Miller 1959) and heated in an oil bafir (100°C) for 15 min. The samples wee
cooled on ice and the absorbance of fire solution measured at 640 nm. One unit
of amylopullulanase activity was defined as fire amount of erzyme which
produced 1 umol of reducing sugar/ min (with glucose as fire standard) under
standard assay conditions.

93

94

Sequence analysis and DNA sequencing - The arrrino acid sequence deduced
from fire amylopullulanase DNA sequerce was compared wifir fire primary
structures of less firermostable amylopullulanases, alpha-amylases and
pullulanases available firrough GerBank (IntelliGenetics Inc., Mountain View,
CA). GCG Sequerce Analysis Software Package ver. 8.0 (1994) was used in fire
analysis and multiple sequence aligrrmerts and subsequert data manipulations.
Hydrophobic cluster analysis (HCA) of fire anrino acid sequerces was pe'formed
as described by Lemesle-Varloot et al (1990). Sequerase V.2.0 T7 DNA
polymerase and Sequerase V.2.0 sequercing kit from United States Biochemicals
were used for verifying fire mutated gere fragmert according to Sanger's
dideoxy chain temination mefirod (Sanger et al., 1977). For double-stranded
DNA sequercing, deraturation of double-stranded plasmid DNA was
pe'formed as described by Zrang et al (1988). For sequercing reactions where
lacZ fusion constructs wee involved, universal M13/ pUC forward and reverse
sequercing primers wee used. The synfiretic Oligonucleotides listed in Table 1
were also used as sequercing primes.

Deletion mutant construction - Several mutants wee constructed by
deletion from 5'- and/ or 3'-erds of fire apu gere and expressed in E. coli (Fig. 1).
Recombinant plasmid pAPZ72 (Mafirupala et al ., 1993) was digested wifir Aat H
and Nde 1, creating an exonuclease HI sersitive resfiiction site (Nde 1) towards fire
DNA insert, which was used to construct deletion mutants of fire DNA inset
from fire 3' to 5' direction (Deletion kit of New England BioLabs Inc.). Six
mutants, which were deleted at fire C-te'minal erd, wee constructed. To
construct deletion mutants in fire 5' to 3' direction, restriction sites Hind HI, Ban H,
HindIH, and KpnI at fire 5' erd of fire inset wee used. Ban H-Hind [H (5.4 kb),
Hind HI-Hind [H (4.8 kb) and Kpn I-Hind [H (4.2 kb ) fragmerts from pAPZ72
were obtained and ligated into pUC vectors. Four deletion

95

Figure 1 Effect of N and C-terminus deletions of amylopullulanase
gene from Thermoanaerobacter ethanolicus 39E an optimum
temperature for enzyme activity. The Apu deletion constructs wee
expressed in E. coli and partially purified by heat treatnrert at 85°C for

5 min. Optimum temperature analysis was done in triplicate unde'
standard assay conditions (similiar specific activities) wifir pullulan as fire
substrate at temperatures ranging from 45-98°C. The figure shows fire
structural gere of each construct and the number of deleted residues. The
dotted box depicts fire proposed catalytic domain of Apu. NA: not active;
T'PR (proposed firemophilicity region) spans amino acids 195-324.

 

96

Figure 1. Effect of N and C-terminus deletions of amylopullulanase gene from
Thermoanaerobacter ethanolicus 39E an optimum temperature for enzyme
activity.

 

 

 

 

 

 

 

 

 
 
 

 

 

Optimum
Temp. (°C)
1 386 829 1481
l
85
1481
l
| 85
1481
I
j 65
107 1224
ApuN106/ C257 ‘-.‘ i 85
107 1102
ApuN 106 / C379 - 80
107 827
ApuN106/ C654 NA

97

mutants at fire N-terminal erd wee constructed and characteized. Each
deletion mutant was transformed into E. coli DHSa, and tested for activity and
firermaclraracteistics. The deletion constructs were grown at 37°C in 5.0 ml of
LB media (Sambrook et al., 1989) containing ampicillin (50 ug/ ml). Cells were
harvested from 1.0 ml of fire culture by certrifugatian (14,000 rpm x 1.0 min) and
fire cell pellet lysted in 50 mM Na-acetate buffer (pH 6.0) containing 5 mM CaClz
and lysozyme. To test for firemal characteistics, the cell lysate was certrifuged
and fire supe'natant heat-treated at each temperature and certrifuged. The
supe'natant was used for activity analysis.

Thermal denaturation and temperature optimum - Samples of
amylopullulanase and mutant constructs were incubated in 50 mM Na-acetate/
5 mM CaClz in a firermally controlled oil bafir at«$°C for 0-210 minutes. The
time course of inactivation was followed by wifirdrawing samples every
10 min. The samples were firer cooled on ice. Substrate was added, and fire
samples processed under standard assay conditions. Optimum temperature
analysis was done by incubating fire erzyme reaction at various temperatures
uncle standard assay conditions.

Kinetic analysis - Assays to determine Kmapp were conducted at 60°C with
substrate concertrations betweer 20 x Kmapp and 0.2 x Kmapp. Kinetic
parameters wee determined using nonlinear curve fits of fire Michaelis-Merter
equafian to fire data. Calculations were done on an IBM pesonal computer
using Kinzynre.

Protein determination and gel electrophoresis - Protein concentrations were
determined using bicinchroninic acid (BCA Assay Kit, Pierce Chemical Co.),
using bovine serum albumin as fire standard. SDS-PAGE was performed 7
according to the mefirod of Laemmli (Laemmli 1970) using 10.0% polyacrylanride
gels in a Mini-Protean H apparatus (Bio-Rad), and protein bands were visualized

98

by staining wifir Coomassie Brilliant Blue R-250. The molecular weights of fire
recombinant proteins wee estimated by comparison to high range molecular
weight standards (Bio-Rad). The purity of the various samples of Apu protein,
judged by coomassie stained SDS-PAGE, was ~15%. The host cell contained no
detectable amylosaccharidase activity.

RESULTS

Identification of a thermophilicity region in T. ethanolicus 39E amylopullulanase
(Apu). - The secondary structures (16) of T. ethanolicus 39E Apu and those of less
firemophilic amylopullulanases from Thermaanaerobacter thermohydrosulfuricus
E101 (optimal temp. 80°C), Thermaanaerobacterium thermosulfurigenes EM1
(opfimal temp. 60°C), and Thermaanaerobacterium saccharolyticum B6A-RI (optimal
tenp. 70°C) were compared by Hydrophobic Cluster Analysis (HCA) and reveal
high secondary structure similarity.

Five deletion mutants were constructed from bofir the C-ternrinal and N-
ternrinal erds to idertify fireir potertial confiibution to firemophilicity and
fire'mostability. The constructs were expressed in Escherichia coli and
characteized after partial purification. ApuN 106 showed a similar optimum
tenperature wifir respect to wild type Apu (Figure 1). ApuN324 showed similar
half-life to wild type, however, fire optimal temperature for erzyme activity was
65°C versus 85°C for fire wild type enzyme (Figure 2). Due to fire shift in
optimum temperature seer with fire ApuN324 mutant, we have designated fire
region betweer amino acid 194 to 324 inrportant for mainterance of the
optimum temperature for erzyme activity. There was no diffeerce in fire
thermostability of fire ApuN324 deletion construct relafive to wild type
(Figure 3). Delefion constructs ApuN106/ C257 and ApuN106/ C379 maintained
an optimum temperature profile similar to firat of the wild type erzyme.
Deleting 654 amino acids from fire C-terminus (ApuN106/ C654) resulted in

complete loss of activity.

99

100

Figure 2. Optimal temperature profiles of wild type amylopullulanase
and deletion construct ApuN324. Optimum temperature analysis was
done in triplicate under standard assay conditions wifir pullulan as fire
substrate in 50 mM Na-Acetate buffer (pH 6.0) containing 5 mM CaClz at
temperatures ranging from 40-100°C.

 

101

Figure 2. Optimal temperature profiles of wild type amylopullulanase
and deletion construct ApuN324.

 

120
I
a 100 *-
E r
E
'c 80 r-
3
g 60 —
E’
3.2 40 —
O
0)
(3'
20 -
O m l r l r I 1 l r l r l

 

 

 

35 45 55 65 75 85 95 105

Temperature (°C)

 

-D— Wild Type

 

 

+ ApuN324

 

102

Figure 3. Thermostability profiles of wild type and ApuN324 deletion
construct of amylopullulanase after enzyme pre-incubatian at 85°C for
0-90 minutes. Aliquots of amylopullulanase and fire mutant construct
were incubated in 50 mM Na-acetate/ 5 mM CaClz in a firermally
controlled oil bafir at 85°C for 0-90 rrrin. The time course of inactivation
was followed by wifirdrawing samples every 10 min. The samples wee
firen cooled on ice. Substrate was added, and fire samples processed
under standard assay conditions. All expeimerts were done in triplicate.

 

 

 

 

103

Figure 3. Thermostability profiles of wild type and ApuN324 deletion
construct of amylopullulanase after enzyme pre-incubatran at 85°C for 0-90
min.

 

   
 

4 L1
39 t1 I2 (min)
WT = 97
3'8 N324 = 130

 

 

 

Ln (residual activity)

 

 

 

o 20 4o 60 80 100
Time (min.)

 

I D Wild Type
O ApuN324

 

 

 

104

Comparison of thermal and kinetic properties of the wild type and ApuN324
deletion construct of amylopullulanase. - The kinetic properties of wild type
amylopullulanase and ApuN324 were determined under standard assay
conditions at 40°C and 60°C. The Kmapp value for fire recombinant erzyme
(0.36 mg/ ml) is similar to that reported for native amylopullulanase from T.
ethanolicus 39E (0.35 mg/ ml); Kmapp for ApuN324 was 0.39 mg/ ml. Vmaxapp
were 32.7 Units / mg and 78.7 Units/ mg for fire wild type and ApuN324 enzymes
at 60°C, respectively (Table 1). Kmapp and Vmaxapp calculations at 40°C wee
0.36 mg/ ml and 19.3 Units/ mg for fire wild type enzyme, and 0.46 mg/ ml and
32.2 Units/ mg for ApuN324 (Table 1)

The Arrherius plot showing temperature-activity data for pullulan
hydrolysis was linear from 50°C to 85°C for fire wild type erzyme, by contrast
fire analysis shows a discontinuity for ApuN324 from 40°C to 65°C (Figure 5).
The discontinuity was observed at 50°C for the deletion construct. The slopes of
fire best fit regression line wee differert above (10 k] mol'l) and below
(56 k] mol'l) fire discontinuity for ApuN324 (Table 2). The slope of Arrherius
analysis for the wild type enzyme revealed an activation erergy of 39 k] mol'1
(Table 2). The activafian erergies above and below fire discontinuity for
ApuN324 wee significanfiy differert from firat observed for fire wild type

erzyme.

105

Table 1 Kinetic analysis of wild type and ApuN324 amylopullulanase
on pullulan at 40°C and 60°C. Kmapp and Vmaxapp determinations in

triplicate on pullulan at various substrate concentrations assayed at 40°C
and 60°C for
30 minutes.

 

 

106

Table 1. Kinetic analysis of wild type and ApuN324 amylopuHulanase on
pullulan at 40°C and 60°C.

Wild Type ApuN324
Kmapp Vmaxapp Kmapp Vmaxapp

 

40°C 0.36 mg/ ml 19.3 Units/ mg 0.46 mg/ ml 32.2 Units/ mg

 

 

60°C 0.35 mg / mg 32.7 Units/ mg 0.39 mg/ ml 78.7 Units/ mg

 

 

 

 

 

 

107

Figure 4. Arrhenius plots for the recombinant T. ethanolicus 39E wild
type and ApuN324 deletion construct of amylopuHulanase between
40°C and 85°C. Tenrperature-activity data done in triplicate for pullulan
hydrolysis. (A) Wild Type (B) ApuN324 deletion construct.

 

 

108

Figure 4. Arrhenius plots for the recoombinant T. ethanolicus 39E wild type
and ApuN324 deletion construct of amylopullulanase between 40°C and 85°C.

 

 

 

 

 

A.
5
E
E
E 4.5 P
C
3
.é‘
E 4 —
O
(0
§
§_ 3.5 -
.32.
E L
3 r l . l . l . l .
2.7 2.8 2.9 3 3.1 3.2
1000 x 1/Temp (K)
B.

 

 

In [specific activity (units/mg)]

 

 

3.5 . 1 . 4 .
2.9 3 3.1 3.2
1000 x 1/Temp (K)

 

109

Table 2. Activation energy calculations from Arrhenius plots of wild type and
ApuN324 amylopullulanase.

 

 

 

 

 

Agfinfim Bum (kl mglzlL
Discontinuity above below
Enzyme Temp. (°C) discontinuity discontinuity bD
Wild Type 3ND 39 39 0
ApuN324 50°C 10 56 46

 

 

 

 

 

 

aND: not detectable
bD = (activation erergy below fire discontinuity) - (activation erergy
above fire discontinuity)

DISCUSSION

Amylopullulanase deletion mutant constructs were made and a region
corresponding to fire first 324 amino acid residues was found to be important for
mainterance of enzyme activity at its optimal temperature (~85°C). The first 106
amino acids are not necessary for optimal activity at 85°C, but deletion of an
additional 218 arrrino acids (to amino acid 324) resulted in an altered optimum
temperature for enzyme activity of about 65°C. Deleting 379 amino acids
(ApuN106/ C379) from fire C-termirral erd had no effect on erzyme activity,
while deleting 654 amino acids (ApuN106/ C654) from fire C-teminus produced
an inactive erzyme.

The structural sinriliarity of mesophilic and firermophilic erzymes and
their enzymatic activity can be shown by fire appearance of linear Arrherius
plots. If fire structural integrity of an erzyme changes to maintain optimal
activity at an elevated temperature it would result in atypical Arrherius
behavior. Biphasic Arrherius plots would be seer and have beer reported for
bofir mesophilic and firermophilic enzymes
(Hartog and Daniel 1992; Hansel et al., 1987; Steigerwald 1990). The
discontinuities reported for thermozymes may indicate a structural change in fire
enzyme from a highly rigid to a less rigid structure needed for optimal erzymatic
activity (Wrba et al., 1990).

Arrhenius plots for mesophilic and firemophilic erzymes are typically
linear revealing fire mainterance of fire molecular architecture of fire erzyme in
fire temperature range for optimal acfivity. Biphasic Arrherius plots have beer
observed for some mesozymes and firermozymes due to a significant structural
change to accomodate fire change in environmental temperature. For example,

110

111

Glycealdehyde-3-phosphate dehydrogerase from Thermoproteus tenax has a
biphasic Arrherius plot (Hansel et al., 1987). Similar observations were also seer
in our lab wifir fire T. ethanolicus 39E 2° Adh when efiranol was used as fire
substrate (Burdette and Zeikus 1996).

The Arrherius plat for the wild type amylopullulanase was linear, unlike
fire discontinuous plot observed for fire ApuN324 deletion construct. The
discontinuity is due to fire preserce of two differert erzyme farms wifir differert
activation energies, probably due to a temperature deperdert structural change.
This observation is more of a destabilization evert over a wide range of
temperatures firan a catastrophic falling apart of fire erzyme. This is a unique
conformational change which is tenrperature deperdert. Work by Burdette and
Zeilcus (1996) wifir 2°Adh also showed firat if fire tenrperature for Arrherius
analysis was increased firere was no discontinuity. The conformational change
was not detectable due to fire appearance of one erzyme form at fire higher
tenperature.

The Kmapp for pullulan was obtained over a wide range of substate
cancertrations at bofir 40°C and 60°C. The temperatures chaser for analysis span
bafir sides of fire transition presert in fire Arrherius plot for ApuN324. The
deperderce of fire rate of pullulan hydrolysis on fire substrate concertration
followed Michaelis-Merter kinetics. The deletion construct and wild type
amylopullulanase bofir have similar affinity toward pullulan at bofir
temperatures; wifir ApuN324 having a higher maximal velocity at 40°C and 60°C
relative to fire wild type erzyme. The optimum temperature for ApuN324 was
about 65°C, versus 85°C for fire wild type erzyme, and accounts for fire higher
maximal velocity observed at 60°C. The sinriliar Kmapp values for fire mutant
and wild type erzymes suggest firat increased flexibility accounts for fire

112

differerce in optimum temperature. This obse'vation supports fire work by
Iaericke (1991) using deuteium exchange studies to examine protein flexibility.

We believe we have provided eviderce supporting fire hypofiresis firat
protein flexibility controls fire temperature opfimum of erzymes. Previously,
Iaericke (1991) has provided eviderce firat fire flexibility of mesophilic and
firemophilic erzymes is equivalert at fireir respective tempe'atures for optimal
activity. We show by nested deletions firat we can lowe' fire optimum
temperature of a fire-mozyme by 20°C and not effect fire catalytic site or
firemostability.

To our knowledge, firis report is fire first to idertify a region important for
fire temperature activity characteistics of an erzyme firat is distinct for what
controls catalysis or firermal stability. Flexibility measurenrerts wee limited by
fire inability to oveexpress and purify fire recombinant protein. Currert
attempts at oveexpression of amylopullulanase resulted in cell toxicity and loss
in fire inset from fire cloning vector. A yeast oveexpression system is currerfiy
being pursued and, if successful, will allow us to oveexpress fire protein and
obtain crystals for firree-dimersional structure analysis. Furfire studies into
flexibility measuremerts, fire firree dimersional structure, physical biochemical
properfies, and folding propeties of erzymes will help elucidate fire
characteistics unique to firemazymes.

ACKNOWLEDGEMENTS

I gratequy aclorowledge Kevin Carr and Dr. Marty Regier for assistance in
reviewing firis manuseipt. I also gratefully acknowledge Dr. Doug Burdette and
Dr. Claire Vieille for helpful discussions. This research was supported by a grant
from fire US. Departmert of Agriculture, under fire agreenrert 89-34189-4299,
and fire Research Excellerce Fund from Michigan State University.

113

REFERENCES

Ahern TJ. and A.M. Klivanov. 1985. The mechanism of irrevesible erzyme
inactivation at 100°C. Science 228: 1280-1284.

Blake, P.R., M.W. Day, B.T. Hsu, L. Joshua-Tor, and I.B. Park. 1992.
Comparison of fire X-ray structure of native rubredoxin from Pyraoaccus furiosus
with the NMR structure of fire zinc-substituted protein. Protein Sci. 1: 1522-1525.

Bahm, G., and R. Iaenicke. 1994. Relevance of sequence statistics for fire
properties of extremophilic proteins. Int. I. Peptide Protein Res. 43: 97-106.

Bradley, E.A., D.E. Stewart, M.W. Adams, and I.E. Wampler. 1993.
Investigations of fire thermostability of rubredoxin models using molecular
dynamics simulations. Protein Sci. 2: 650-665.

Burdette, D.S., C. VieiHe, and I.G. Zeikus. 1996. Cloning and expression of fire
gere ercoding the Thermoanaerobacter ethanolicus 39E secondary-alcohol
dehydrogerase and erzyme biochemical characteization. Biochem. I.

316: 115-122.

Dill, K.A. 1990. Dominant forces in protein folding. Biochemistry 29: 7133-7155.

Eijsink, V.G.H., B.W. Dijkstra, G. Vriend, LR. van der Zee, O.R. veltman, B.
van der Vienne, B. Van den Burg, S. Kempe, and G. Venema. 1992. The effect
of cavity-filling mutations on fire firermostability of Bacillus stearothermophilus
neutral protease. Protein Engng. 5: 421-426.

Fontana, A. 1990. How nature erginees protein (firermo) stability. in Life
uncle Extreme Conditions: Biochemical adaptation G. di Prisca, ed., Springer
Verlag, Heidelberg, pp. 89-113.

Genetics Computer Group. 1994. Program Manual for the Wisconson Package,
Version 8, Madison, Wisconson.

Hanahan, D. 1983. Studies of transformation of Escherichia coli with plasmid.
I. Mol. Biol. 166; 557-580.

Hartog, AT and R.M. Daniel. 1992. An alkaline phosphatase from Thermus Sp
strain Rt41A. Int. 1. Biochem. 24: 1657-1660.

114

115

Hansel, R., Laumann, S., Lang, I., Heumann, H. and F. Lattspeich. 1987.
Characteization of two D—glyceraldehyde-3-phasphate dehydrogerases from fire
extremely themophilic archaebacteium Thermoproteus tenax Eur. I. Biochem. 170:
325-333.

Iaenicke, R. 1991. Protein stabith and molecular adaptation to extreme
conditions. Eur. I. Biochem. 202: 715-728.

Kanaya, S., and M. Itaya. 1992. Expression, purification, and characteization of
a recombinant ribonuclease H from Thermus thermophilus HB8 I. Biol. Chem. 267:
10184-10192.

Laemmli, U.I(. 1970. Cleavage of structural proteins during fire assembly of fire
head of bacteiophage T4. Nature (London) 227: 680-685.

Lemesle, L., B. Henrissat, C. Gaboriaud, V. Bissery, A. Morgat, and I.P.
Maman. 1990. Hydrophobic cluster analysis: procedures to deive structural
and functional information from 2-D-represertation of protein sequerces.
Biachimie 72:555-574.

Mathupala, S.P., S.E. Lowe, S.M. Padkovyrov, and I.G. Zeikus. 1993.
Sequencing of fire amylopullulanase (apu) gere of Thermaanaerobacter ethanolicus
39E, and identification of fire active site by site-directed mutageresis. I. Biol.
Chem. 268: 16332-16344.

MiHer, G.L. 1959. Use of dinitrosalicyclic acid reagert for determination of
reducing sugar. Anal. Chem. 31: 426-428.

Perbal, B. 1988. in A practical guide to molecular cloning, 2nd ed. John Wiley
and Sons, New York.

Sambrook, I., Fritsch, ER, and Maniatis, T. 1989. Molecular cloning: A
Laboratory Manual (Second Edition).

Sandberg, W.S. and T.C. Terwilliger. 1989. Influence of inteior packing and
hydrophobicity on fire stability of a protein. Science 245: 54-57.

Sanger, F., S. Nicklerr, and A.R. Coulson. 1977. DNA sequencing wifir chairr-
te'minating inhibitors. Proc. Natl. Acad. Sci. USA 74: 5463-5467.

Segel, I.H. 1975. in Enzyme Kinetics: Behavior and Analysis of Rapid
Equilibrium and Steady-State Enzyme Systems, pp. 929-934, John Wiley &: Sons,
NY

Steigerwald, V.I., Beckler, G.S., and LN. Reeve. 1990. I. Bacteriol. 172: 4715-
4718.

116

Tomazic, 5.1., and A.M. Klivanov. 1988. Why is one Bacillus alpha-amylase
more resistant against irrevesible firermainactivation firan anofirer? I .Biol. Chem.
263: 3092-3096.

Vieille, C., I.M. Hess, R.M. Kelly, and I.G. Zeikus. 1995. xylA cloning and
sequercing and biacherrrical characteization of xylose isomerase from
Thermotaga neapolitana. Appl. Environ. Microbiol. 61: 1867-1875.

Wrba, A., Schweiger, A., Schultes, V. and R. Iaenicke. 1990. Extremely
firermastable D-glyceraldehyde-3-phosphate dehydrogerase from fire
eubacteium Thermotoga maritima. Biochemistry 29: 7584-7592.

Zhang, I-L, R. Schall, I. Browse, and C. Somerville. 1988. Double stranded DNA
sequercing as a choice for DNA sequencing. Nucleic Acids Res. 16: 1220-1223.

Chapter IV
Summary and Directions for Future Research

117

118

The-mophilic micro-organisms can be classified eifirer as moderate
firermophiles wifir growth optima of 60-80°C or hyperfire'mophiles having
growfir optima above 80°C. Habitats suitable for growth of firermaphiles are
widespread. Enzymes studied from hyperfirermophiles have a highe-
temperature optimum and increased firermastability ove' mode'ate
fire-mophiles or mesophiles (growfir optima of 25-45°C) (V ieille et al., 1996). This
is an irrrpartant rationale for use of firemophilic erzymes in industrial starch
processing applications firat require high temperature operation. A number of
extrenrely fire-mostable erzymes of potential industrial utility have beer purified
or fireir geres cloned from anaerobic thermophiles. A characteistic of a numbe'
of hyperfirermophiles is fire capability to utilize complex saccharides which are
metabolized to meet carbon and erergy requiremerts. Starch, which is used in
many industrial processes (e.g. production of fructose fermertation syrup), is
composed of glucose linked by bofir alpha-1,4 and alpha-1,6 linkages which can
be hydrolyzed into sugar syrups by various solubilizing and debranching
erzymes; known collectively as amylosaccharidases.

An interesting and novel erzyme, amylopullulanase, has beer idertified
in many firermophiles ( Hyun and Zeikus 1985; Ramesh et al., 1994; Melasniemi
1988; Saha et a1 ., 1988; Coleman et al., 1987; Antranikian 1990; Spreinat and
Antranikian 1990). Amylopullulanase is a debranching and solubilizing erzyme
capable of hydrolyzing bofir alpha-1,4 (alpha-amylase activity) and alpha-1,6
(pullulanase activity) linkages in starch and related sugars
( Hyun and Zeikus 1985; Ramesh et al., 1994; Melasnierrri 1988; Saha et al., 1988;
Coleman et al., 1987; Antrarrikian 1990; Spreinat and Antranikian 1990). Cloning
and purification of amylopullulanase (a maltotriose producing erzyme) and
neopullulanase (a panose producing erzyme) from modeate firermophilic

119

organisms has revealed four highly conserved regions important for erzymatic
activity (T akata et al., 1992). Hydrolysis of alpha-1,4 and alpha-1,6 linkages are
catalyzed by fire same active site (Sversson 1991).

Idertification of an acid stable amylopullulanase in a themophile will
have important biotechnological application because bofir elevated temperature
and low pH facilitate fireir use in industrial starch processing. Currerfiy, starch
is solubilized and saccharified at neutral pH and high temperature due to
limitations of current alpha-amylases and pullulanases. Amylopullulanase offers
fire capacity for a one-step process to produce new sugar syrups firat can be used
as feedstock for fire production of high-value products for yeast, fungal, or
bacteial fe'mertations. This dissetation describes expeimerts on fire
biochenical, biophysical, and molecular biological characteization of
‘ amylopullulanase from Thennoanaerobacter ethanolicus 39E wifir respect to
substrate cleavage characteistics and firermophilicity.

The recombinant gere was expressed in E. coli and partially purified by
cell lysis and heat treatrnert. Analysis of fire deduced amino acid sequerce of
many amylosaccharidases in relation to amylopullulanase idertified four highly
conserved regions wifirin firese erzymes important for catalysis and substrate
cleavage specificity. Geretic ergineeing techniques were used to construct
alteed apu gere products wifir diffeent catalytic activities and substrate
cleavage specificities. This suggests fire possibility for protein-ergineeing of fire
substrate binding sites in amylopullulanase to alter fire function of firis dual
activity erzyme to firat of eifirer alpha-amylase (alpha-1,4) or pullulanase
(alpha-1,6).

A fire'mophilicity region (T'PR) was also idertified by deletion
mutageresis at fire N-teminal erd of fire protein resulting in a shift of fire
optimum temperature for erzyme activity. Arrherius plot for wild type

 

120

amylopullulanase was linear, however, the observed plot for fire ApuN324
deletion was discontinuous due to a temperature deperdert structural change
necessary for optimal catalytic activity. Furfirer expeimerts into fire flexibility
and rigidity of fire protein (deuteium exchange, proteolysis analysis, and
exposure to deraturants) may iderfify a correlation betweer fire observed
optimal temperature shift and erzyme flexibility.

More detailed analysis of cleavage specificity and protein flexibility will
be aided by fire ability to oveexpress amylopullulanase. Currerfiy,
oveexpression has beer inhibited by cell toxicity and plasmid instability. A
yeast oveexpression system is currerfiy being pursued and, if successful, may
allow us to oveexpress fire protein and obtain crystals for firree-dimersional
structure analysis. This will allow us to confirm fire catalytic activity and
cleavage specificity results preserted in Chapter 2 by model fitfing starch, and
ofirer substrates, into fire catalytic site of fire erzyme. Detailed information into
fire binding mode betweer substrate and erzyme will be obtained, and aid in fire
production of a monospecific, firermophilic amylopullulanase for industrial
starch degradation. Examination of protein flexibility characteristics of the
ApuN324 nested deletion of amylopullulanase will also be possible; as well as
pullulan binding to fire active site of wild type amylopullulanase versus
ApuN324.

REFERENCES

Antranikian, G. 1990. Physiology and erzymology of firermophilic anaerobic
bacteia degrading starch. FEMS Microbiol. Rev. 75: 154-174.

Coleman, R.D., S.S. Yang, and MP. McAlister. 1987. Cloning of fire
debranclring-erzyme gene from Thermoanaerobium brackii into Escherichia coli and
Bacillus subtilis. I. Bacteriol. 169: 107-115.

Hyun, H.H., and I.G. Zeikus. 1985. General biochenrical characteizafion of
firermostable pullulanase and glucoamylase from Clostridium
thermohydrosulfuricum. Appl. Environ. Microbiol. 49:1168-1173.

Mathupala, S.P., S.E. Lowe, S.M. Podkovyrov, and I.G. Zeikus. 1993.
Sequercing of fire amylopullulanase (apu) gere of thermoanaerobacter ethanolicus
39E, and idertification of fire active site by site-directed mutageresis. I. Biol.
Chem. 268: 16332-16344.

Melasniemi, H. 1988. Characteization of alpha-amylase and pullulanase
acfivities of Clostridium thermohydrosulfuricum. Biochem. I. 250: 813-818.

Ramesh, M.V., S.M. Padkovyrav, S.E. Lowe, and I.G. Zeikus. 1994. Cloning
and sequencing of fire Thermaanaerobacterium saccharolyticum B6A-RI apu gere
and purification and clraracteizafion of fire amylopullulanase from Escherichia
coli. Appl. Environ. Microbiol. 60: 94-101.

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

Spreinat, A., and G. Antranikian. 1990. Analysis of fire amylolytic erzyme
system of Clostridium thermosulfurogenes EM1. Appl. Microbiol. Biotechnol. 33: 511-
518.

Takata, H., T. Kuriki, S. Okada, Y. Takesadam, M. Iizuka, N. Minamiura, and
T. Imanaka. 1992. Action of neopullulanase: neopuHulanase catalyzes bofir
hydrolysis and transglycosylation of alpha-(1,4) and alpha-(1,6) glucosidic
linkages. 1. Biol. Chem. 267: 18447-18452. '

Vieille, C., Burdette, D.S., and Zeikus, I.G. 1996. in Biotech. Ann. Rev., vol. 2,
(M. R. E1 Geweley, ed.) pp. 1-83. Elsevier, Amsterdam, Neth.

121

 

APPENDIX A

122

 

Construction and Characterization of Thermoanaerobacter ethanolicus 39E
AmylopuHulanase Praline Deficient Mutants

INTRODUCTTON

Initial comparison of mesophilic and fire'mophilic erzymes has identified
features unique to firermozymes. It has beer observed firat fire'mophilic
erzymes are smaller, have less ordeed structure, contain more hydrophobic
interactions, and less beta-structure, among ofirer features (V ieille et al., 1996).
However, it appears firat fire key to thermostability will be determined by
comparing fire'modynamic properties, amino acid composition and sequerce of
homologous mesophilic and firermophilic proteins; not by fireir molecular
architecture (Sundaram 1986). Site directed mutageresis has become a useful
tool in fire analysis of protein thermostabilization. Small changes in the
stabilizing forces caused by only one or two amino acid changes can raise fire
relative stability of an erzyme by several degrees certigrade
(Coolbear et al ., 1992). Ofirer factors leading to increased firermostability include:
increased protein rigidity at mesophilic temperatures, location of proline
residues in fire loop regions of firermophilic proteins, and interaction of fire
protein with its surroundings (V ihinen 1987).

Themophilicity and firermostability of erzymes are closely related to
flexibility and rigidity of the proteins' folded conformation. Praline residues may
play an important role in increasing fire rigidity of the erzyme, and it has beer
reported firat proline residues, presert at a high frequercy in beta-turns and
wifirin loop regions and binding adjacert secondary structures, are responsible

 

123

124

for firermastability ( Watanabe et al., 1991; Suzuki et al., 1987). The data
preserted here examines fire effect of mutating fire 5 individual T. ethanolicus
amylopullulanase proline residues (Prol95, Pro210, Pr0213, Pro240, Pr0244)
(Mafirupala et al., 1993) not presert in fire less firermoactiveThennaanaerobacterium
saccharolyticum B6A-RI and Clostridium thermosulfurogenes EMl amylopullulanase
enzymes based on sequerce aligrrmerts (Table 1). The firermophilicities and
firermostabilities of fire wild type and proline mutant erzymes are reported.

 

 

125

Table 1 Amino acid alignment of the T'PR of several thermophilic
amylopullulanases. The legerd at fire bottom of fire figure idertifies fire
organisms from which fire sequences are represented. The five unique proline
residues within T'PR of 39B are deroted as P1-P5. Praline residues presert in
TPR and depicted in bold.

126

Table 1. Amino acid alignment of the TPR of several thermophilic
amylopullulanases.

39E amino acid 148

l

 

 

B6A rosarcaaxo nearer-lull momma masrrxcar
m1 rasarcaclm marsraru museums rrasvpxsar
EM101 IQPAIGAGDD msrsraru momma summit!
393 rasararraaa msrsraru sanctum YTMWPKRY!
86A ammcva assnronsva amour-m murmur
m1 ammavx assuraruv'r amrrma murmur
EM101 armament: septum men-rm svssurwrar
393 ammo somrrmva rmrrma svsrmrrrl'ar
86A IYYDDLKHDT sarrrmrs arxvsarv'rr. moment:
arm Imam HDSH'RNPPG avxvoa-rv'rr. aromas
EM101 IYYDDLRHD‘I‘ warmers arrroarv'rr. stomachs
393 Imam warrants arm-carver. aromas
B6A mm RIGBSPDGNY smrnsra erratum:
EM1 mmm massacre smrnsra WI
34101 marsvm arcosraaxr mm mm:
395 mm xraasrasar mm mm:

QYKVTLGNTW

BFKVTLG'SW
BFKVTLGPSW

8PTLTGLDNN
SPILTGLDNN
NPPLTGPDNN
NP'LTGPDNN

SARISYIDDI
BARISYIDDI
SAKISYWDDI
SAKISYIDDI

LKDGTRIAIY
LKDGTKTAIY
LKDGTKTAXY
LKDGTKTAIY

 

39E :- Thennoanaerobacter ethanolicus 39E

B6A - Tlrennoanaerabacterium sacdrarolytiaim B6A-RI
EM1 - Clostridium themrosulfluogenes EM1 ’
EM101 - Clostridium thermosrdfurogenes EMlOl

 

MATERIALS AND METHODS

Reagents, Enzymes, and Oligonucleotides - All chemicals were of molecular biology
or analytical grade and obtained from Aldrich Chemical Co., or Sigma.
Restriction erzymes and ligase were obtained from Befiresda Research
Laboratories, United States Biochemical Co., or Boehringer Manheim.
Oligonucleotides wee synfiresized in an Applied Biosystenrs model 380A DNA
synthesizer at fire Macromolecular Structure Facility, Departrnert of
Biochemistry, Michigan State University. The Oligonucleotides were
subsequerfiy 5'-phosphorylated using T4 polynucleotide kinase, for use in site
directed mutagenesis (Table 2).

Bacterial strains, plasmids, and transformations - E. coli strain SURE
{e14-(mcrA) (mchB-hstMR-mrr)171 endAl supE44thi-1 gyrA96 relA1 lac recB
rec] sbcC umuC:Tn5(kan1')uvrC [F' proAB lacI‘lZ M15Tn10(tetr)]} from Strategere
Co. and DHScr-F' {F' 80nd (lacZYA-argF)U169 deaR recAl endAl hst17
(rk‘mk+)supE44 I- thi-l gyrA96 relA1}, from Befiresda Research Labs.) were used
as a host for transformation, and E. coli BMH 71-18 mutS, IM109, IA221 and
E51301 mutS from Promega Corporation were used. E. coli strains were made
competent by fire Harman mefirad as described by Pebal (Perbal 1988), while
recombinant vectors were introduced in E. coli strains by heat-shock treatnrent
(Harman 1983). '

Enzyme assays - For determination of amylopullulanase activity, 160 111 of
1.25% (w/ v) pullulan in 50 mM acetate buffe (pH 6.0) containing 5 mM CaClz
and erzyme (heat treated culture supenatant) to a total volume of 200 111 were
mixed and incubated at 60°C for 30 min. The reaction was stopped by adding
0.8 ml of dirritrasalicylate solution (Miller 1959) and heated in an oil bafir (100°C)
for 15 min. The samples were cooled on ice and the absorbance at 640 nm was

127

 

128

Table 2. Oligonucleotides used in this study.

 

 

Mutation Oligonucleotide synfiresized

pm195Am (1:1) 5'- CCCATGAGIICCC'I'AAAAG'ITACI'ITAAACI'C -3'
prongs“ (pg) 5'- CAAAGGAATA’ITTGAACCATI'ITGTI‘CAC .3'
pm213c1n (p3) 5'-CATAGGCTACATICAAITGAATATTI'GGAC-3'
Pro240He (P4) 5'- GCCCCTGTGAGIA’I'I'GGATI‘GTAATCI'G -3'
pm244Leu (p5) s- CATAATATATG'I'I‘ATI‘ATCAAQCCCI‘GTGAGAG -3'
Pmpzrloofgéln (132/ p3) 5'- CTACA'ITCAAIIQAATATIIQAACCATI'ITGHC .3'
Pmngliiluu (P4 / P5) 5'- GTTATI‘ATCAAGCCCI‘GTGAGIAII‘GGATI‘GTAATC .3'

 

 

 

Table 2. The alanine mutant Oligonucleotides for P1, P2, P3, P4, and P5 wee
synfiresized substituting fire underlined residues wifir
SEEK-3'. underlined: alteed nucleotides in fire primer to
produce fire desired amino acid change.

129

measured. One unit of amylopullulanase activity was defined as fire amount of
enzyme which produced 1 umol of reducing sugar (wifir glucose as fire
standard) / min under standard assay conditions.

Sequence analysis and DNA sequencing - The amino acid sequerce deduced
from fire amylopullulanase DNA sequerce was compared wifir fire primary
structures of less firermastable amylopullulanases, alpha-amylases and
pullulanases available firrough GerBank (IntelliGeretics Inc., Mountain View,
CA). GCG Sequerce Analysis Software Package ver. 8.0 (1994) was used in fire
analysis and multiple sequerce aligrrmerts and subsequert data manipulations.
Hydrophobic cluster analysis (HCA) of fire amino acid sequerces was pe'formed
as described by L. Lemesle-Varlaot et al (Lemesle et aL, 1990). Sequerase V2.0 T7
DNA polymerase and Sequanase V2.0 sequercing kit from United States
‘iochemicals were used for veifying fire mutated gere fragnrert according to
Sanger's dideoxy chain te'mination mefirod (Sanger et al., 1977). For double-
stranded DNA sequercing, deraturation of double-stranded plasmid DNA was
pe'forrned as described by Zhang et al (Zhang, et al., 1988).

Site-directed mutant construction - Mutageresis was done wifir an
oligonucleotide-directed in vitro mutageresis system (BIO-RAD, Richmond, CA).
Chemically synfiresized Oligonucleotides (18-mer to 28-mer) were used and fire
sequerces are shown in Table 1. The Apa I -Bgl H segmert of fire prAP164-UC
was ligated into M13mp19 using Xba I and Bam HI restriction sites. (The Xba I
site was erd filled wifir Klerow producing a blunt erd to complemert the blunt
end produced byApa I, and fire Bam I-H overhang is compatible with Bgl II).
Uracil-containing single strand DNA was synfiresized and used as fire template.
An oligonucleotide complenrertary to the region to be altered, except for fire
mismatch, was hydridized to fire single-strand uracil DNA. The complenentary
strand was firer synthesized by T4 DNA polymerase using fire oligo as primer.

130

Ligase was used to seal fire new strand to fire 5' erd of fire oligo. The double-
stranded DNA containing fire mutation of inteest was transformed into E. coli
TG-l. The mutations were confirmed by DNA sequercing, using fire Sequerase
Version 2.0 kit (U .S. Biochemical Corp., OH). The sequercing reaction was
primed by intenally annealing 17-mer synfiretic Oligonucleotides.

Thermal denaturation and temperature optimum - Aliquots of
amylopullulanase and mutant constructs wee incubated in 50 mM Na-acetate/
5 mM CaClz in a firemally controlled oil bafir at 85°C for 0-90 min. The time
course of inactivation was followed by wifirdrawing samples every
10 min. The samples wee firer cooled on ice. Substrate was added, and fire
samples processed under standard assay conditions. Optimum temperature
analysis was done by incubafing fire erzyme reaction at various tempe-atures
unde' standard assay condifions.

Protein determination and gel electrophoresis - Amylopullulanase
concertration in partially purified fractions was dete'mined by scanning
dersitometry of Coomassie Blue stained SDS-PAGE (Laemmli 1970) gels. BSA
served as a standard.

 

RESULTS

A firermophilicity region (TPR) was idertified in T. ethanolicus and was
compared by amino acid alignment and Hydrophobic Cluster Analysis (HCA) to
amylopullulanases from T. thermosulfurigenes EM1 and T. saccharolyticum B6A-RI
(Figure 1). In firis region, firere were five additional proline residues at positions
195, 210, 213, 240, and 244 (designated as P1, P2, P3, P4, and P5 respectively) in
fire T. ethanolicus erzyme (Table 1). The five proline residues were singly
substituted wifir alanine, by site-directed mutagenesis. No change in optimum
temperature or firermostability was detected (data not shown).

P1-P5 in T. ethanolicus were firen substituted singly with fire corresponding
residues in fire less the'mophilic erzyme T. thermosulfurigenes EMl (Asn, Ser,
Glrr, He, and Leu) (Figure 2). Similiar results were observed. Double mutants
ProZlOSer/ Pr0213Gln and Pr024OIle/ Pr0244Leu were constructed and again, no
change in optimum temperature or thermostability (Figure 3). The mntribution
of additional proline residues in fire more thermophilic erzymes does not seem
to give an individual positive effect on firermophilicity.

131

132

Figure 1 HCA comparison of thermophilic amylopullulanase at T'PR.
The legerd at fire bottom of fire figure idertifies fire organisms
represerted in the HCA analysis.

133

Figure 1. HCA comparison of thermophilic amylopullulanases at T'PR

 

| l
l I IISW‘L‘fiaQr
39E 8% 1901.
(80°C) $3; .

 

 

EM1 “1's 0 * ° ‘t
60°C "cl; 0. A 92‘ K|.|.‘”I.g
( ) 183%??? (girth/a";

|

(70°C) 32:1" 1.,

I * u ' '3 0'0 D r \
B6A-RI '1 q. 3 75‘ DIIOE‘DQ.MD N. "
‘ 9 fiasifig‘fi‘if ° ' a

‘47

 

 

39E = Thermaanaerobacter ethanolicus39E
EM1 = Clostridium thermosulfurogenes EMl
B6A-RI= Thermaanaerobacterium saccharolyticum B6A-RI

134

Figure 2. Thermostability and optimum temperature analysis after
single substitution of proline residues 1-5 from T. ethanolicus 39E with
the corresponding residues of the less thermophilic enzyme from T.
thermosulfurigenes EM1. Aliquots of amylopullulanase and fire mutant
constructs (similiar specific activities) were incubated in

50 mM N a-acetate / 5 mM CaClz in a firemally confi'olled oil bafir at 85°C
for 0-90 minutes. The time course of inactivation was followed by
withdrawing samples eve-y 10 minutes. The samples wee firen cooled on
ice. Substrate was added, and the samples processed under standard
assay conditions. Optimum temperature analysis was done in triplicate
under standard assay conditions with pullulan as fire substrate in

50 mM N a-Acetate buffer (pH 6.0) containing 5 mM CaClz at
temperatures ranging from 40-100°C.

 

Figure 2. Thermostability and optimum tempera

135

ture analysis after single

substitution of proline residues 1-5 from T. ethanolicus 39E with the
corresponding residues of the less thermophilic enzyme from T.

thermosulfurigenes EM1.

A. Thermostability Profiles

In (residual activity)

4.2
4
3.8

3.6 ’
3.4 '
3.2 h

3 .

 

 

 

 

l . l m l . l

 

0

20 40 60 80 100

Time (min.)

B. Optimum Temperature Profiles

Relative Activity (%)

100

80

60

4O

20

0

 

 

 

l r l r l

 

 

40

60 80 1 00 120

Temperature (°C)

 

 

iiiii

Wild Type
Prat 95Asn
Pr021 OSer
Pr021 SGln
Pr0240|le
Pr0244Leu

 

 

136

Figure 3. Thermostability and optimum temperature profiles of double
proline mutants (Pm210Ser/Pr0213GIn and Pr0240He/Pro244Leu)
relative to wild type amylopullulanase. Aliquots of amylopullulanase
and the mutant constructs (similiar specific activities) were incubated in
50 mM Na-acetate/ 5 mM CaC12 in a firermally controlled oil bafir at 85°C
for 0-90 nrinutes. The time course of inactivation was followed by
wifirdrawing samples eve'y 10 minutes. The samples wee firen cooled on
ice. Substrate was added, and fire samples processed under standard
assay conditions. Optimum temperature analysis was done in triplicate
unde' standard assay conditions wifir pullulan as the substrate in

50 mM Na-Acetate buffer (pH 6.0) containing 5 mM CaClz at
temperatures ranging from 40-100°C.

 

137

Figure 3. Thermostability and optimum temperature profiles of double
proline mutants (ProZlOSer/Pro213Gln and Pro240He/Pr0244Leu) relative to
wild type amylopuHulanase.

A. Thermostability Profiles

4.2

 

4
3.8 .
3.6 P
3.4 b

 

In (specific activity)

3.2

 

 

3..I.I.I.L.
o 20 40 60 80100

Time (min.)

 

B. Optimum Temperature Profiles

 

 

 

 

100—

? r

2280—

g

3360—

< .

$40-

5

£20-

Ob 1.1.1.
40 60 80100120

Temperature (°C)

 

+ Wild Type

+ Pr021OSer/Pr021SGIn

 

 

+ Pr024OIIeIPr0244Lsu

 

DISCUSSION

Research in fire area of fire dete-minants of protein stability is still in an
early stage. Site-directed mutants of T4 lysozyme in fire form Xaa to Pro
indicated no change in erzymatic activity or its three-dimersional structure
relative to wild type (Matfirews et al ., 1987). Specifically, wher Ala82, presert in
fire turn of an alpha-helix, was substituted with proline fire stability of fire T4
lysozyme was increased by 0.8 kcal/ mol at 64.7°C (Matfirews et al., 1987). This
initiated prospect firat amino acid subsfitutions might be a gereral strategy for
erhancing fire stability of proteins.

An increase in proline contert is seer in fire more firermophilic erzymes
(Table 3). The effect of proline residues on protein stabilization was studied by
site-directed mutagenesis of fire five proline residues (Pl-P5) in fire T. ethanolicus
39E amylopullulanase. Single substitutions to alanine, or fire corresponding
amino acid in fire less firermophilic organism, T. thermosulfurigenes EMl,
produced no significant change in activity, optimum temperature or
firermostability. While none of fire proline deficiert mutants demonstrated a loss
of fire'mostability relative to fire wild type, a more detailed mutageresis
approach needs to be conducted. Multiple mutations up to all five proline
residues should be done to confirm fire lack of effect of prolines on protein
stabilization.

138

' 139

Table 3 Praline profiles of amylopullulanase. Apu 39B is
amylopullulanase from Thermaanaerobacter ethanolicus 39E; Apu B6A-RI is
amylopullulanase from Thermaanaerobacterium saccharolyticum B6A-RI

 

 

140

Table 3. Praline profiles of amylopullulanase

Enzyme Optimum Temperature Praline Contert

 

Apu 39E 90°C 524%

 

 

Apu B6A-RI 75°C 4.04%

 

 

 

 

REFERENCES

Amelunxen, R.E. and A.L. Murdock. 1978. Mechanisms of thermophily. Critical
Review ofMicro 6: 343393.

Coolbear, T., RM Daniel, and H.W. Morgan. 1992. The enzymes from extreme
firemophiles: bacteial sources, firemostabilities and industrial relevance. Adv.
Biochem. Eng/Biotechnol. 45: 57-98.

Genetics Computer Group. 1994. Program Manual for fire Wisconson Package,
Vesion 8, Madison, Wisconson.

Hanahan, D. 1983. Studies of transformation of Escherichia coli wifir plasmid.
I. Mol. Biol. 166: 557-580.

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

Lemesle, L., B. Henrissat, C. Gabariaud, V. Bissery, A. Margat, and I.P.
Mornon. 1990. Hydrophobic cluster analysis: procedures to deive structural
and functional information from 2-D-represertation of protein sequerces.
Biachimie 72:555-574.

Mathupala, S.P., S.E. Lowe, S.M. Podkovyrov, and I.G. Zeikus. 1993.
Sequencing of fire amylopullulanase (apu) gere of Thermaanaerobacter ethanolicus
39E, and identification of fire active site by site-directed mutageresis. I. Biol.
Chem. 268: 16332-16344.

Matthews, B.W., H. Nicholson, and WJ. Becktel. 1987. Enhanced protein
the-mostability from site-directed mutations firat decrease fire entropy of folding.
Prac. Nafi. Acad. Sci. USA 84: 6663-6667.

Miller, G.L. 1959. Use of dinitrosalicyclic acid reagert for determination of
reducing sugar. Anal. Chem. 31: 426-428.

Perbal, B. 1988. in A practical guide to molecular cloning, 2nd ed. John Wiley
and Sons, New York.

Sanger, F., S. Nicklen, and A.R. Coulson. 1977 . DNA sequercing wifir chairr-
terminating inhibitors. Proc. Natl. Acad. Sci. USA 74: 5463-5467.

Sundararrr, V.A. 1986. Physiology and growth of firermophilic bacteia. in

Thermophiles: Gereral, Molecular and Applied Microbiology, p. 75 Wiley and
Sons, New York.

141

142

Suzuki, Y., K. Oishi, H. Nakano, and T. Nagayama. 1987. Appl. Microbiol.
Biotechnol. 26:546-551.

Vielle, C., Burdette, D.S. and Zeikus, I.G. 1996. in Biotech. Ann. Rev., vol. 2,
(M. R. E1 Geweley, ed.) pp. 1-83. Elsevie, Amsterdam, Nefir.

Vihinen, M. 1987. Relationship of protein flexibility to firemastability. Prat.
Engin. 1: 477-480.

Watanabe, K., K. Chishiro, K. Kitamura, and Y. Suzuki. 1991. Praline residues
responsible for fire'mostability occur wifir high frequercy in fire loop regions of
an extremely firermostable aligo-1,6- glucosidase from Bacillus thermoglucvsidasius
KPlOO6. I. Biol. Chem. 266: 24287-24292.

Zhang, H., R. Schall, I. Browse, and C. SamerviHe. 1988. Double stranded DNA
sequercing as a choice for DNA sequercing. Nucleic Acids Res. 16: 1220-1223.

Appendix B

143

Attempts at Overexpression of Recombinant Amylopullulanase from
Thermaanaerobacter ethanolicus 39E

General Procedure for Overexpression

Recombinant plasmids (pPROEX-l, pT7SC, and plN-lH-ompA) carrying
the apu gere wee h'ansformed into E. coli sh'ains TG-l, B121, MC1061,
I-MSI74(DE3)(pLysS), or IA221 for expression of fire target protein. Cultures
were grown at room temperature, 30°C, or 37°C in 20 x 10 ml cultures of
LB + 100 ug/ml ampicillin, and 25 ug/ ml clrloramphericol if a pLysS sh'ain was
being used. Cell growfir was monitored by measuring fire OD at 595 nm. Wher
the OD595 was betweer 0.6 and 0.8, expression was induced. IPT G was added to
a final concertration of 0.4 mM (unless ofirerwise indicated); expression of
protein varied from 1-3 hours at fire designated temperature. The cultures were
chilled by placing the tubes in ice-water baths, and cells collected by
cerhifugation in an Sorvall GS-3 rotor at 5,000 rpm for 10 minutes at 4°C. The
cleared supernatant was poured off and fire cell pellet was resusperded in
5 ml of 50 mM Na-Acetate (pH 6.0) / 5 mM NaCl and 0.3 mg / ml lysozyme. The
resusperded cells were incubated on ice for 30 min, firel frozer in liquid
nih'oger. Lysed cells wee firawed, cerhifuged in a Sorval SS-34 rotor at
12,000 rpm for 30 min and fire supenatant recovered. Protein was heat heated at
85°C for 15 min, chilled on ice, and fire supernatant recovered by certrifugation
in an SS-34 rotor at 8,000 rpm for 15 min. The protein was quantitated by SDS-
PAGE. The activity of fire protein was stable at 4°C for 6
2-3 monfirs.

144

RESULTS

 

This cloning vector (Polayes and Hughes 1994) is inducible by addition of
IPTG and contains fire 6xHis sequence allowing for ease of purification by
Ni-NTA resin (Hochul et al., 1987). E. coli host cells containing the verified
recombinant plasmid were grown and induced under fire various conditions
listed below with fire accompanying results.
1. Standard induction using E. coli TG—l with 0.6 mM IPTG at
37°C removing aliquots at 1, 2 and 3 hrs post induction.
Result: No overexpression and lower activity after
induction indicating a stability or toxicity problem.
2. Introduced a Kant (Vieira and Messing 1982) cartridge in prAP164-EX
to increase stability. Standard induction using E. coli TG-l wifir
0.6 mM IPTG at 37°C removing aliquots at 1, 2 and 3 hrs post
induction.
Result: No overexpression and lower activity after
induction.
3. Standard induction using strains void of some proteases
(E. coli BL21 and E. coli MC1061) wifir 0.6 mM IPTG at
37°C removing aliquots at 1, 2 and 3 hrs post induction.
Result: N o overexpression and lower activity after
induction. Analysis of fire starter culture and subcultures
after induction indicate firat the activity is higher in the
starter culture and decreases after subculturing.
4. Induction of starter culture (E. coli TG-l) wifir 0.6 mM IPTG at
37°C removing aliquots at 1, 2 and 3 hrs post induction.

14S

Result: Activity is maintained but no overexpression.

5. Induction of starter culture (E. coli TG-l) fitrating IPTG from 0.05 mM
to 4.0 mM at 37°C removing aliquots at 1, 2 and 3 hrs post
induction.

Result: Activity is maintained but firere is no overexpression.
There is no relevant difference in activity at the different
IPTG concentrations.

6. Induction of starter culture (E. coli TG-l) titrating IPTG from 0.05 mM
to 4.0 mM at different temperatures (RT, 30°C, and 37°C) removing
aliquots at 1, 2 and 3 hrs post induction.

Result: N o overexpression and no difference in activity at the

different temperatures.

 

This vector (Brown and Campbell 1993) suppresses readfirrough
transcription from cryptic promoters and start points on fire plasmid, in order to
reduce expression in fire absence of T7 RNA polymerase and improve use in the
expression of highly toxic gene products. Protein expression was inducible by
infection wifir lambda phage containaing fire T7 RNA polymerase gene
(lambdaCB6). Induction wifir IPTG and E. coli strain HMSl 74(DE3)pLysS was
also used. In firis strain, T71ysozyme was provided by pLysS which down
regulates activity of T7 RNA polymerase in fire absence of expression.

E. coli host cells containing the verified recombinant plasmid were grown
and induced under the various conditions listed below wifir fire accompanying
results.

1. Induction from starter culture wifir 0.4 mM IPTG in

HMSI 74(DE3)pLysS [source of T7 RNA polymerase] at 37°C
removing aliquots at 1, 2, and 3 hrs post induction.

146

147

Result: Active but no overexpression.

2. Induction from starter culture titrating IPTG from 0.05 mM to 4.0 mM
in HMS174(DE3)pLysS [source of T7 RNA polymerase] at 37°C
removing aliquots at 1, 2, and 3 hrs post induction.

Result: Active but no overexpression. N 0 difference in activity at
fire different IPTG concentrations.

3. Induction from starter culture wifir 0.4 mM IPTG in
HMS174(DE3)pLysS [source of T7 RNA polymerase] at various
temperatures (RT, 30°C, and 37°C) removing aliquots at 1, 2, and
3 hrs post induction.

Result: Active but no overexpression. No difference in activity at
fire different temperatures.

4. Induction from starter culture using lambdaCE6 andE. coli TG-l as
hoststrainat37°Cremovingaliquotsat0.5, 1,2, and3hrspost
induction.

Result: Active but no overexpression.

5. Induction from starter culture wifir using lambda CE6 and E. coli host
strains (MC1061 and 81.21) at 37°C removing aliquots at 0.5, 1, 2,
and 3 hrs post induction.

Result: Active but no overexpression.

6. Induction from starter culture wifir using‘lambda CE6 andE. coli TG-l
as host strain at various temperatures (RT, 30°C, and 37°C)
removing aliquots at 0.5, 1, 2, and 3 hrs post induction.

Result: Active but no overexpression.

 

This cloning vector (Ghrayeb et al., 1984)is inducible by addition of IPTG
and results in fire expressed protein being secreted into fire periplasmic space

148

using fire signal sequence of ampA (an E. coli outer membrane protein) in fire
cloning vector. E. coli host cells OA221) containing fire verified recombinant
plasmid were grown and induced under fire various conditions listed below wifir
fire accompanying results.
1. Induction at various temperatures (RT, 30°C, and 37°C wifir 0.05, 0.1,
0.05, 1.0, 2.0, and 4.0 mM IPTG removing samples at 1, 2, and
3 hrs post induction.
Result: No activity and no overexpression.
2. Induction as indicated above using TB media instead of LB media.
Result: No activity and no overexpression.

DISCUSSION

Our initial attempts using the pPROEX—l vector were to optimize for
purification following overexpression using fire 6x His sequence added to fire
N-terminus of the protein. Lack of overexpression halted any subsequent
attempts at purification. We then decided to try the pT7SC vector due to its
ability to suppresss readfirrough transcription from cryptic promoters on fire
plasmid and aid in fire expression of highly toxic gene products. Overexpression
was again not detected in firis system. The last attempt at protein expression was
done utilizing fire pIN-III-ompA secretion vector. Upon induction of gene
expression with IPTG, fire gene product is secreted into fire periplasmic space
using the signal sequence of ompA in the cloning vector. The advantages of firis
system include: stabith of fire gene product from protease activity due to
secretion into fire periplasrrric space; and maintenance of toxic enzymes due to
secretion of fire protein while being synfiresized. The consistent result in fire
pIN-III-ompA vector was complete loss of activity. The validity of fire construct
was confirmed by DNA and protein analysis. The presence of fire protein was
verified by activity staining using native PAGE containing soluble starch (see
experimental procedures Chapter 2).

Extensive examination of overexpression of amylopullulanase has been
problematic in all systems tried in firis report. Drs. Soroj Mafirupala, Sue Lowe,
and long-Hyun Park also tried overexpression of recombinant amylopullulanase
with sirrriliar results. Reasons for lack of overexpression may include, but are not
limited to: 1) fire large size of the protein (163,000 MW), 2) toxicity to cells, 3)
protein degradation, and 4) presence of AGG and ASA Arg codons which may
be problematic in the expression of large proteins in E. cali.. Personal

149

150

communication wifir ofirer investigators attenrpting to overexpress and purify
some thermozymes have encountered similiar problems. pEI' vector expression
wifir overlapping lacO/ P17 may give better repression as a modified E. coli
expression system. If overexpression is still an issue, a yeast secretion system
may be useful in overcoming fire toxicity problem inhibiting overexpression.
The analysis of protein production by SDS-PAGE would be aided by fire use of
antibodies to amylopullulanase from fire native organism. The current system,
native starch PAGE, for analysis of protein abundance requires large amounts of
protein for detection (1.0 ug) and is not quantitative.

REFERENCES

Brown, W.C. and ].L. Campbell. 1993. A new cloning vector and expression
strategy for genes encoding proteins toxic to Escherichia coli. Gene 127: 99-103.

Ghrayeb, I., H. Kimura, M. Takalrara, H. Hsiung, Y. Masue, and M. Inouye.
1984. Secretion cloning vectors in E. coli. Embo I. 3: 2437-2442.

Hochuli, E., H. Dobeli, and A. Schacher. 1987. I. Chromatography 411: 177-184.
Polayes, DA and AJ. Hughes. 1994. Focus

Vieira, J. and I. Messing. 1982. Gene 19: 259-262.

151

Appendix C

152

Transglycosylation Activity Analysis of Recombinant Mutant and Wild Type
Amylopullulanase from Thermaanaerobacter ethanolicus 39E

N o amylopullulanase has been studied for its transglycosylation activity.
The presence of transglycosylation activity in ofirer amylosaccharidases tempted
us to study firis enzymatic reaction wifir amylopullulanase. Transglycosylation
activity is present and has been studied in B. lichenifarmis alpha-amylase (BLMA)
and B. stearothermophilus neopullulanase. BLMA has alpha-1,6 transglycosylation
activity (Kim et al ., 1992) while neopullulanase has bofir alpha-1,4 and alpha-1,6
transglycosylation activity (Kuriki et al., 1993). Studies carried out wifir
neopullulanase showed a single catalytic site (regions I-IV) is responsible for bofir
hydrolysis and transfer activities. Results of experiments wifir amylopullulanase
reveal an inability to catalyze alpha-1,4 or alpha-1,6 transfer activities.

General Procedure for Testing Transglycosylation Activity

Enzyme samples (0.05 Units) were incubated in fire presence of 24.0%
liquefied corn starch in 50 mM acetate buffer (pH 6.0) wifir 5 mM CaClz at 60°C
for 24 hours, and fire products were analyzed by High Performance Ion
Chromatography (HPIC) using a CarboPac PA1 column (Dionex BioLC4500i)
and a pulsed amperomeric detector (PAD, Dionex). Buffer A (150 mM NaOH in
water), buffer B (600 mM Na-acetate in buffer A), and buffer C (water) were used
for elution. All fire solvents were prepared wifir Milli-Q water, and filtered
through a polyvinylidene difluoride menrbrane filter (0.22am) (Gelman Sciences,
Inc.). The samples were eluted at a flow rate of 1.0 ml/ min wifir 0-40% solvent B

153

154

for 40 min. Twenty five microliters of 0.1% sample solution was injected into fire

column for analysis.

Results of Transferase Analysis

Transferase analysis was done in collaboration wifir Dr. Kwan Hwa Park
(Seoul National University - Seoul, Korea) using partially purified extracts of
E. coli expressing fire mutant recombinant enzymes and wild type
amylopullulanase. The time course reaction wifir wild type amylopullulanase
resulted in hydrolysis of large oligosaccharides, accumulation of maltose, and fire
lack of increase in peak area of branched oligosaccharides, all evidence for lack of
transglycosylation activity. No detectable transglycosylation activity was present
in any of fire active mutants constructed in firis study.

REFERENCES

Kim, I.C., I.H. Cha, LR. Kim, S.Y. Iang, B.C. Seo, T.K. Cheong, D.S. Lee, Y.D.
Choi, and KR Park. 1992. Catalytic properties of fire cloned amylase from
Bacillus licheniformis. I. Biol. Chem. 267: 22108-22114.

Kuriki, T., M. Yanase, H. Takata, T. Imanaka, and S. Okada. 1993. A new way

of producing isomalto-aligosaccharide syrup by using fire transglycosylation
reaction of neopullulanase. I. Permen. Bioenginer. 76: 184-190.

155