V x V..V. . V . .. . ..., .V .... V. . . V .V. . V. . I .. . . W. «I. w A ‘
“Wm? .va V V T . . ., j. H . ....V V . .V .. VFW a

t ...h... GWOWLg.
. V .x .u
VVm 3....3mmﬁ

5.1

”6.2.51, wngﬁf. H. V . V. V. V . ,V V , V. V V _ ..V.‘...1V.VV..
«1. L .m.... a.» aura .,.V..V. V .. . .. V . .

.... .13 uerJV£Rr v.,.r¢..0. .

.9, .3... EV? ..

.1. ~
1.7.9? Marv... .
3...? E V.
...».
. as»...
. 16..

 

 

 

 

 

I, \p. .l

... V. $.H v uﬁm V . V .V
.. . I» v V, V . V .. V .. mum. 1.1.. fl
\mﬁ. ... “mgumnzﬁ ﬁawipfw ..utnmwﬁz : .....V.........V.. . V . . . . , V V 1 V . . V _ V . V . $_H.v.a..hknwﬂn:..h... 4‘
..I. a I... ... ..Uﬁ ..quV v . . V «4%? ..vurn a}. V V . V . V. . V . . .V. . V . lﬂPﬂctvd. Run“ “......”
..Wwﬁykuﬁ qia,{r..me :aymar .. V. V .V . .. . .. V .V _ V. V. .V V ..V -..-.wﬂﬁﬁmmm

Pa. . . .. a L . . . V . V . V V V . . . V V V
.. $3 Law. V .

.

 

v Avluo .1, «05¢...V 01h.
‘. a, o y
:I Mr,

I
. V. V . . .VIVVaﬁQ... .. V ..VVV V .
. V V V. V .. V V V V V V V. VllVlvf.-!‘§-IAI§1
. V . . . V . V. .V V . V . . ., V V V V .V V, . . .V . 1.1(.V..u.....a«I,(..1$.
magma... . ctrmusﬁaulbﬂ. V . . .. V . V . . .V V . . . V V. V

V 13m... C. V V .

V. _ . 3...: .. .V . .
”u Knowlyoaﬁﬂ ...: ...r {Iniivvﬁ . .. .. . . . V V . .V
‘vau.... . 2 3:. ....o ..... S. V

. . c.,HV»41,VVMVHN. udfnv:

 

--.:V ...)” !V
.V 1. “vs-Hun.

I
.OII ,ll.'h

.. V V . JMUMW 11% V V . . VV .....IVD H..ol.u.eroV!¢t.V‘¥bVHPnh‘
.l. . 3.13.1.1?! not. V. . V ..wﬁ....1-.v.. . . . . V . . . V
V . 7...}...Svlﬁtl . ,. .. Av. .. V V V V V V .V . .
. V 215 ”£60..qu V .21”! .. .MV‘W u .
bid :1. u.nr.1.V...m:2:....bl .
Ya I. 5. \ ,
hr. -..-hunk .
...aTﬁvﬂ.

. .V .. ...»).

. V . V, in V 2V V ,.
V ,.V . .V ....yI :1V3cl.un.;....: _

“.....H . V V . eff}. .8

~|o| .1 .[ vol...
.IV ... 'A u
V lnnl‘llt alt!
l’l?¢a|.1lt.l V.

 

11V” . t ..
. I .. Judah..- Huh} .

LEI-tr 1’...)
’vlu‘.'¢.»

.A
. .cl
DI! Vlvlll‘
31“;

p v ' t V!!! .' ...
b. .0 turn} ..I l
iVJ.¢.|.\IIA.

 

..irch' .... ‘2 vi?!-

. ..v..1.u.v.l. .
..1V . H ( ..V!» TﬂoEDH JEJ. 1

V . (V \V n )u.
.E‘. :3. b I osLlr Vin» . I! v v .5: ..r. , ...aiiluih‘. L .7... a ..lVLVu VV k tsz V; a». ‘ L: V- _ V r t: fl! .1 r L» n d H I |'

 

lHESlS

l

ANESTAT UN IV

lllllllllllllllll lllllllllllllllllllllllllll

3 1293 01563 39093

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

The Allosteric Regulation Sites
of ADP-Glucose Pyrophosphorylase
from Cyanobacterium Anabaena PCC 7120

presented by

Jun Sheng

has been accepted towards fulﬁllment
of the requirements for

Ph.D. degree in Biochemistry

Major professor

Date 8/7 /77Z
/

MSU is an Afﬁrmatiw Action/Equal Opportunity Institution 0-12771

 

LIBRARY
Michigan State
University

PLACE N RETURN BOX to roman this checkout from your rocord.
TO AVOID FINES Mum on or odor. duo duo.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

D

 

 

 

 

 

 

MSU In An Afﬂnnotlvo Adlai/EM Opportunity Institution
WM!

THE ALLOSTERIC REGULATION SITES OF ADP-GLUCOSE
PYROPHOSPHORYLASE FROM CYANOBACTERIUM ANABAENA PCC 7120

By

Jun Sheng

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

DOCTOR OF PHILOSOPHY

Department of Biochemistry

1997

ABSTRACT

THE ALLOSTERIC REGULATION SITES OF ADP-GLUCOSE
PYROPHOSPHORYLASE FROM CYANOBACTERIUM ANABAENA PCC 7120

By

Jun Sheng

ADP-glucose pyrophosphorylase is a key enzyme in the biosynthesis of glycogen
in bacteria and starch in plants. This enzyme catalyzes the conversion of glucose-l-P and
ATP to ADP-glucose and pyrophosphate. It is subject to allosteric regulation. ADP-
glucose pyrophosphorylase from cyanobacterium Anabaena PCC 7120 is mainly activated
by 3-P—glycerate and inhibited by phosphate.

Previous chemical modiﬁcation studies with pyridoxal-P suggested that Lys382 of
Anabaena ADP-glucose pyrophosphorylase is involved in the binding of 3-P-glycerate, the
allosteric activator. Site-directed mutagenesis of Lys382 of the Anabaena enzyme was
performed to determine the role of this residue. Replacing Lys382 with either arginine,
alanine, or glutamine produced mutant enzymes that were still activated by 3-P-glycerate
but with apparent afﬁnities 10- to 160-fold lower than that of the wild-type enzyme. In
contrast, the glutamic acid mutant enzyme was inhibited by 3-P—glycerate. These mutations
had lesser impact on the kinetic constants for the substrates and inhibitor, phosphate, and
on the thermal stability. These results indicate that both the charge and size of the residue
at position 382 inﬂuence the binding of 3-P-glycerate. Site-directed mutagenesis was also
performed to obtain a K382R-K419R double mutant. The apparent afﬁnity for 3-P-

glycerate of this double-mutant enzyme was 104-fold lower than that of the wild-type

enzyme, and the speciﬁcity for the activator of this mutant enzyme was altered.

Treatment of Anabaena ADP-glucose pyrophosphorylase with phenylglyoxal
resulted in a time- and concentration-dependent loss of enzyme activity. Phosphate, the
inhibitor, protected the enzyme from inactivation most effectively, while 3-P-glycerate,
fructose-1,6-P2, pyridoxal-P, and ATP plus magnesium were also good protectors. After
incubation with 2 mM phenylglyoxal for 1 hour, the modiﬁed enzyme had a 10-fold lower
apparent afﬁnity for phosphate in the absence of the activator, 3-P—glycerate, than that of
the wild—type enzyme. This result has implicated the involvement of an arginine residue
at the allosteric sites, most probably the inhibitor-binding site.

In order to identify the arginine residue, ﬁve arginine residues, which are
conserved in all higher-plant and cyanobacterial enzymes but not in enteric bacterial
enzymes, were individually converted to alanine by site-directed mutagenesis. The mutant
enzymes were puriﬁed and the properties of these mutants were compared with the wild-
type enzyme. Substitution of only Arg294 with alanine resulted in an enzyme with more
than lOO-fold or 40-fold lower afﬁnity for the inhibitor, phosphate, in the absence or
presence of 3-P-glycerate, respectively. This mutation had no or lesser impact on the
kinetic constants for the substrates and the activator, 3-P-glycerate. It is concluded that

Arg294 is a critical residue in phosphate binding and inhibition.

To my family

iv

ACKNOWLEDGEMENTS

I would like to thank my supervisor, Dr. Jack Preiss, for his direction and valuable
support. During the past ﬁve years, I have gained signiﬁcant experience in biochemistry
and molecular biology and learned how to conduct scientiﬁc research under Dr. Preiss'
guidance. I also thank my committee members, Drs. Michael Bagdasarian, Shelagh
Ferguson-Miller, Lee Kroos, and Honggao Yan, for their help through my graduate study.
Dr. David G. McConnell gave me tremendous help when I came to the United States and
did my ﬁrst rotation in his laboratory and he also gave me spiritual support during these
years.

Drs. Yee-yung Charng and Chris Meyer, with whom I began my research in this
laboratory, are not only my mentors but also my friends. They gave me many ideas and
technical help and encouraged me whenever I had a problem. Dr. Miguel A. Ballicora,
Dr. Minxian Wu, Brian Smith-White, and Yingbin Fu contributed their time in discussion
with me and provided many intriguing ideas. I am also grateful to other members in the
laboratory for their friendship which made me feel like I had a big family at Michigan
State University.

My parents and my brothers give me support through my education and without
their love I would not have a chance to gain my doctoral degree. I would like to thank my
newly-wedded wife, Xin, for her love and encouragement which made the ﬁnal stage of
my graduate study easier than otherwise it would be. I also want to thank the friends I met
here during the past several years, Bing Ren, Bo Zhang, Qing Yang, Xiaoyu Wu, Jing
Wang, Lei Lei, Bing Lian, Bing Li, Jie Qian, and many others. They gave me their hands

whenever I needed.

TABLE OF CONTENTS

Page
LIST OF TABLES .................................................................................... x
LIST OF FIGURES ................................................................................... xi
ABBREVIATIONS ................................................................................... xiii
INTRODUCTION ..................................................................................... 1
CHAPTER I LITERATURE REVIEW ........................................................... 6
1. Starch and Glycogen ...................................................................... 7
2. Synthesis of Starch and Glycogen ...................................................... 8

3. The Role of ADP-Glucose Pyrophosphorylase in Biosynthesis
of Starch and Glycogen .................................................................... 8
3.1. ADP-Glucose Pyrophosphorylase in Starch Synthesis .................... 9
3.2. ADP-Glucose Pyrophosphorylase in Glycogen Synthesis ............... ll
4. Regulation of the ADP-Glucose pathway ............................................. 11
4.1. Regulation of the ADP-Glucose Pathway in Starch Synthesis ......... 12

4.2. Regulation of the ADP-Glucose Pathway in Bacterial
Glycogen Synthesis ............................................................. 13
5. Subunit and Primary Structure of ADP-Glucose Pyrophosphorylase ............ 17

6. Structures of the Substrate and Regulator Binding Sites of

ADP-Glucose Pyrophosphorylase ....................................................... 24
6.1. The Primary Structure of the Substrate Binding Sites ................... 26
6.2. The Primary Structure of the Regulator Binding Sites .................. 30

7. Prediction of the Secondary Structure of ADP-Glucose Pyrophosphorylase. . . .39

References ...................................................................................... 44

vi

CHAPTER II SITE-DIRECTED MUTAGENESIS OF LYSINE382,
THE ACTIVATOR-BINDING SITE, OF ADP-GLUCOSE

PYROPHOSPHORYLASE FROM ANABAENA PCC 7120 ................. 51
Abstract ......................................................................................... 52
Introduction .................................................................................... 53
Material and Methods ........................................................................ 55

Reagents ............................................................................... 55

Bacterial Strains and Media ........................................................ 55

Site-Directed Mutagenesis .......................................................... 55

Expression and Puriﬁcation of the Wild-type

and Mutant Enzymes .............................................................. 56

Enzyme Assay ........................................................................ 56

Kinetic Characterization ............................................................ 59

Protein Assay ......................................................................... 60

Protein Electrophoresis and Immunoblotting .................................... 60

Molecular Mass Determination .................................................... 61

Thermal Stability ..................................................................... 61

Reductive Phosphopyridoxylation ................................................. 61
Results .......................................................................................... 62

Expression and Puriﬁcation of Lys382 Mutant Enzymes ....................... 62

Kinetic Characterization of Lys382 Mutant Enzymes ........................... 62

Kinetic Characterization of the K382R-K419R Enzyme ...................... 66

Effector Speciﬁcity .................................................................. 66

Molecular Masses of the Wild-Type and Mutant Enzymes .................. 67

Thermal Stability of Lys382 Mutant Enzymes ................................... 71

Reductive Phosphopyridoxylation of the

K382R-K419R Mutant Enzyme ................................................. 71
Discussion ...................................................................................... 72
References ...................................................................................... 76

CHAPTER III INVOLVEMENT OF ARGININE AT THE ALLOSTERIC
SITES OF ADP-GLUCOSE PYROPHOSPHORYLASE FROM
ANABAENA PCC 7120 .......................................................... 78

Abstract ......................................................................................... 79

vii

Introduction .................................................................................... 80

Material and Methods ........................................................................ 82
Reagents ............................................................................... 82
Expression and Puriﬁcation of ADP-glucose

Pyrophosphorylase ................................................................. 82
Enzyme Assay ........................................................................ 83
Kinetic Characterization ............................................................ 84
Modiﬁcation of ADP-glucose Pyrophosphorylase

with Phenylglyoxal ................................................................ 84
Incorporation of [“C]phenylglyoxal to

ADP-glucose Pyrophosphorylase ................................................ 84

Results .......................................................................................... 86
Inactivation of ADP-glucose Pyrophosphorylase

by Phenylglyoxal ................................................................... 86
Protection of Enzyme Activity by Substrates

and Allosteric Effectors ........................................................... 86
Effect of Modiﬁcation with Phenylglyoxal on

Enzyme Kinetics ................................................................... 86
Stoichiometry of Modiﬁcation ..................................................... 94

Discussion ...................................................................................... 99

References .................................................................................... 101

CHAPTER IV ARGININE294 IS ESSENTIAL FOR THE INHIBITION
OF ANABAENA PCC 7120 ADP-GLUCOSE

PYROPHOSPHORYLASE BY PHOSPHATE .............................. 103
Abstract ....................................................................................... 104
Introduction ................................................................................... 105
MATERIALS AND METHODS ......................................................... 107

Reagents .............................................................................. 107
Bacterial Strains and Media ....................................................... 107
Site-Directed Mutagenesis ......................................................... 107
Expression and Puriﬁcation of the Mutant Enzymes ......................... 109
Enzyme Assay ....................................................................... 109
Kinetic Characterization ........................................................... 110
Protein Assay ........................................................................ 1 10
Protein Electrophoresis and Immunoblotting .................................. 110

viii

Thermal Stability ................................................................... 111

Results ......................................................................................... 112
Expression of the Mutant Enzymes ............................................. 112

Screen for Alanine-Substituted Mutant Enzymes
with altered Pi Inhibition ........................................................ 112
Puriﬁcation of the Mutant Enzymes ............................................. 112
Kinetic Characterization of the R294A Enzyme .............................. 115

Kinetic Characterization of the R66A,

the R105A, and the R385A Enzymes ......................................... 121
Thermal Stability of the Wild-type and the Mutant Enzymes .............. 121
Discussion ..................................................................................... 122
References .................................................................................... 125
CHAPTER V SUMMARY AND PERSPECTIVES .......................................... 127

l. Lys382 of the cyanobacterium Anabaena
ADP-glucose pyrophosphorylase is involved
in the binding of the activator, 3PGA ................................................ 128

2. An arginine residue(s) is involved in
the allosteric regulation of Anabaena
ADP-glucose pyrophosphorylase ....................................................... 129

3. Arg294 of the Anabaena ADP-glucose
pyrophosphorylase is important for the

inhibition by Pi ............................................................................ 130
4. Future Research .......................................................................... 13 1
References .................................................................................... 133

ix

LIST OF TABLES

Page
Chapter I
1. Activator speciﬁcity of ADP-glucose pyrophosphorylases
from various bacterial species ............................................................ 15
Chapter II

1. Kinetic Parameters for Activation of the Wild-Type
and Mutant ADP-Glucose Pyrophosphorylases ........................................ 63

2. Kinetic Parameters of the Wild-Type and Mutant
ADP-Glucose Pyrophosphorylases ....................................................... 65

3. Speciﬁcity of Allosteric Effectors of the Wild-Type
and Mutant ADP-Glucose Pyrophosphorylases ........................................ 68
Chapter III
1. Effect of substrates and allosteric effectors on modiﬁcation
of Anabaena ADP-glucose pyrophosphorylase by phenylglyoxal .................. 93
Chapter IV
1. Conserved arginine residues in Anabaena ADP-glucose
pyrophosphorylase and theoligonecleotides used for

conversion to alanine ...................................................................... 108

2. Kinetic parameters of the Anabaena wild-type, R66A,
R105A, R294A, and R385A enzymes ................................................. 116

LIST OF FIGURES

Chapter I

1. Alignment of the primary structure of

ADP-glucose perphosphorylases ...............................................

2. Amino acid sequence of the phosphopyridoxylated
peptide, which was protected by ADP—glucose plus
Mg“, of E. coli ADP-glucose pyrophosphorylase

and the alignment with other known sequences ..............................

3. Amino acid sequence of the major azido-ADP-[l‘C] glucose
incorporated peptide of E. coli ADP-glucose pyrophosphorylase

and the alignment with other known sequences ..............................

4. Amino acid sequence of the phosphopyridoxylated
peptide, which was protected by fructose-1,6-P2,
of E. coli ADP-glucose pyrophosphorylase and the

alignment with other known sequences ........................................

5. Amino acid sequence of the phosphopyridoxylated peptides
of spinach leaf and Anabaena ADP-glucose pyrophosphorylases
and the alignment with other known sequences of higher-plant

and cyanobacterial enzymes ......................................................

6. Proﬁle Neural Network (PHD) prediction of the secondary

structure of ADP-glucose pyrophosphorylase .................................

Chapter II

1. Nucleotide sequence and encoded protein sequence
of the Anabaena ADP-glucose pyrophosphorylase gene
in the region of Lys382 and the synthetic oligonucleotides

used for site-directed mutagenesis at position 382 ...........................

Page

......... 19

......... 28

......... 31

......... 33

......... 36

......... 41

......... 57

2. Activation of the wild-type and K382R-K419R
enzymes by 3PGA, FBP, and AMP ..................................................... 69

Chapter 111

1. Inhibition of Anabaena ADP-glucose
pyrophosphorylase by phenylglyoxal .................................................... 87

2. Kinetics of inhibition of ADP-glucose pyrophosphorylase
activity by phenylglyoxal (enzyme activity was measured
under activated conditions) ................................................................ 89

3. Kinetics of inhibition of ADP-glucose pyrophosphorylase
activity by phenylglyoxal (enzyme activity was measured

under unactivated conditions) ............................................................. 91
4. Inhibition of ADP-glucose pyrophosphorylase by Pi ................................ 95
5. Incorporation of [”C]phenylglyoxal into
ADP-glucose pyrophosphorylase ......................................................... 97
Chapter IV

1. Pi inhibition of the wild-type, the R66A,
and the R294A enzymes in the crude extract ......................................... 113

2. The effect of Pi on the enzyme activity of
the wild-type and the R294A enzymes ................................................. 117

3. The effect of Pi and KC] on the enzyme activity
of the wild-type and the R294A enzyme .............................................. 119

xii

ABBREVIATIONS

3PGA, 3-P—glycerate

cpm, counts per minute

EDTA, ethylenediamine tetraacetic acid
FBP, fructose-1,6-P2

FPLC , fast performance liquid chromatography
kb, kilo-base pair(s)

kDa, kilo-dalton

LB medium, Luria-Bertani medium
PAGE, polyacrylamide gel electrophoresis
Pi, orthophosphate

PLP, pyridoxal—P

PPi, pyrophosphate

SDS, sodium dodecyl sulfate

xiii

INTRODUCTION

Starch is the major reserve carbohydrate of algae and higher plants, deposited as
a normal consequence of active photosynthesis. Glycogen is an energy reserve compound
in many bacteria, accumulated particularly when growth is limited and an excess carbon
source is available. It is believed that the biosynthesis of starch in algae and higher plants
and of bacterial glycogen occurs predominantly via the ADP-glucose pathway (Preiss &
Romeo, 1989; Preiss, 1991; Preiss, 1996; Preiss & Sivak, 1996).

In this pathway, ADP-glucose is synthesized from ATP and glucose-l-P in a
reaction catalyzed by ADP-glucose pyrophosphorylase (ATPza-D-glucose-l-P
adenylyltransferase, EC. 2.7.7.27). The glucosyl moiety of the sugar nucleotide is then
transferred to a maltodextrin, glycogen or starch in an a-1,4-glucosidic linkage by
glycogen synthase or starch synthase (EC. 2.4.1.21). Branching enzyme (BE; EC.
2.4.1.18) catalyzes the synthesis of the a-l ,6-glucosidic linkages found in amylopectin and
in glycogen.

ADP-glucose pyrophosphorylase is known as an allosterically regulated enzyme.
The enzyme from higher plants is mainly activated by 3-P—glycerate (3PGA) and inhibited
by phosphate (Pi) (Ghosh & Preiss, 1965; Sanwal et a1. , 1968), whereas the enzyme from
enteric bacteria is activated by fructose-1,6-P2 (FBP) and inhibited by AMP (Preiss et a1. ,

1966). The bacterial enzyme is homotetrarneric in structure (I-Iaugen et a1. , 1976), while

2
the higher-plant enzyme is more complex, being heterotetrameric with two different
subunits (Preiss, 1991).

Cyanobacteria are prokaryotic organisms that have the chloroplast type of
photosynthesis (Aitken, 1988). However, cyanobacteria synthesize glycogen as the major
carbon and energy reserve (Shively, 1988), in a similar manner to that in other bacteria.
The cyanobacterial ADP-glucose pyrophosphorylase is homotetrameric similar to the
enteric bacterial enzyme, but is regulated by 3PGA and Pi like the higher-plant enzyme
(Levi & Preiss, 1976; Iglesias et al., 1991). To understand the structure—function
relationships of the allosteric sites of the cyanobacterial ADP-glucose pyrophosphorylase
would be of great interest.

Previous studies have shown that the activator-binding site of the E. coli enzyme,
Lys”, is close to the N-terminus (Parsons & Preiss, 1978; Gardiol & Preiss, 1990), while
that of the spinach leaf enzyme is near the C-terminus (Morell et al., 1988; Ball & Preiss,
1994). Chemical modiﬁcation and site-directed mutagenesis studies on the cyanobacterial

“9 is involved in

ADP-glucose pyrophosphorylase from Anabaena have shown that Lys
binding of 3PGA (Charng et al., 1994). Chemical modiﬁcation studies of the K419R
enzyme with pyridoxal-P (PLP), where residue 419 is no longer reactive, have shown that
the phosphopyridoxylated K419R enzyme is less dependent on the presence of 3PGA for
activation and less sensitive to Pi inhibition. Lys382 of the mutant enzyme was found to be
labeled by PLP. Kinetic and protection studies suggested that Lys382 was also involved in
binding of the allosteric activator (Charng et al., 1994).

Site-directed mutagenesis of Lys382 of Anabaena ADP- glucose pyrophosphorylase

3

was performed to determine the role of this residue. Lys382 was replaced with either
arginine, alanine, glutamine, or glutamic acid. The mutant enzymes were puriﬁed and
characterized by kinetic studies and compared with the wild-type enzyme. The results
indicate that Lys382 is involved in the binding of the activator, 3PGA. This work was
published in Biochemistry (1996) 9;, 3115-3121. Chapter H of this dissertation is derived
from this publication.

Previous studies on E. coli, synechocystis, and spinach leaf enzymes using an
arginine-speciﬁc reagent, phenylglyoxal, suggested that there is also an essential arginine
residue involved in the allosteric activation and inhibition (Carlson & Preiss, 1982;
Iglesias et al., 1992; Ball & Preiss, 1992). Treatment of Anabaena ADP-glucose
pyrophosphorylase with phenylglyoxal resulted in a time- and concentration-dependent loss
of enzyme activity. Pi, the inhibitor, protected the enzyme from inactivation most
effectively, while 3PGA, FBP, PLP, and ATP plus magnesium were also good protectors.
The modiﬁed enzyme had lower apparent afﬁnities for Pi and 3PGA than those of the
wild-type enzyme. Thus, phenylglyoxal appears to interfere with the allosteric regulation,
as well as the catalysis, of Anabaena ADP-glucose pyrophosphorylase. This work is
presented as Chapter III of this dissertation.

In order to identify the arginine residue implicated by chemical modiﬁcation
(Chapter III), ﬁve conserved arginine residues were individually converted to alanine by
site-directed mutagenesis. The mutant enzymes were puriﬁed and the properties of these
mutants were compared with the wild-type enzyme. Substitution of arginine294 with alanine

resulted in an enzyme with more than 100-fold or 40-fold lower afﬁnity for the inhibitor,

4
Pi, in the absence or presence of 3PGA, respectively. This mutation had no or lesser
impact on the kinetic constants for the substrates and the activator, 3PGA. These results
suggest that Arg294 is involved in the binding of the inhibitor, P,. This work is included in

Chapter IV of this dissertation.

REFERENCES
Aitken, A. (1988) Methods Enzymol. 167, 145-154.
Ball, K., & Preiss, J. (1992) J. Protein Chem. 11, 231-238.
Ball, K., & Preiss, J. (1994) J. Biol. Chem. 269, 24706-24711.
Carlson, C.A., & Preiss, J. (1982) Biochemistry 21, 1929-1934.
Charng, Y., Iglesias, A. A., & Preiss, J. (1994)]. Biol. Chem. 269, 24107-24113.
Gardiol, A., & Preiss, J. (1990) Arch. Biochem. Biophys. 280, 175-180.
Ghosh, H. P., & Preiss, J. (1965) J. Biol. Chem. 240, 960-961.
Haugen, T. H., Ishaque, A., & Preiss, J. (1976) J. Biol. Chem. 251, 7880-7885.
Iglesias, A. A., Kakefuda, G., & Preiss, J. (1991) Plant Physiol. 97, 1187-1195.
Iglesias, A. A., Kakefuda, G., & Preiss, J. (1992) J. Protein Chem. 11, 119-128.
Levi, C., & Preiss, J. (1976) Plant Physiol. 58, 753-756.
Morel], M., Bloom, M., & Preiss, J. (1988) J. Biol. Chem. 263, 633-637.
Parsons, T., & Preiss, J. (1978) J. Biol. Chem. 253, 6197-6202.
Preiss, J ., Shen, L., Greenberg, E., & Gentner, N. (1966) Biochemistry 5, 1833-1845.
Preiss, J., & Romeo, T. (1989) Adv. Microb. Physiol. 30, 183-238.

Preiss, J. (1991) in Plant Molecular and Cell Biology (Mifﬂin, B., ed.) Vol. 7, pp 59-114,
Oxford University, Oxford.

Preiss, J. (1996) Biotech. Annu. Rev. 2, 259-279.

Preiss, J, & Sivak, M. (1996) in Photoassimilate Distribution in Plants and Crops
(Zamski, E., Schaffer, A. A., ed.) pp 63-96, Marcel Dekker, Inc., New York.

Sanwal, G. G., Greenberg, E., Hardie, J., Cameron, E. C., & Preiss, J. (1968) Plant
Physiol. 43, 417-427.

Shively, J. M. (1988) Methods Enzymol. 167, 195-203.

5

CHAPTER I

LITERATURE REVIEW

LITERATURE REVIEW

1 . Starch and Glycogen:

Starch, a polysaccharide, is the major carbon and energy reserve of most green
plants. It is the end-product of carbon ﬁxation by photosynthesis and is a water-insoluble
granule present in every type of tissue: leaves, fruit, pollen grains, roots, shoots, and
stems. The size and shape are varied depending on the plant source and the stage of
development. The starch granule contains two types of polysaccharides, amylose and
amylopectin. Amylose is mainly found as a linear chain of a-1,4 linked a-D-
glucopyranosyl residues of about 1000 units long, with occasional secondary chains
attached by OH ,6-glucosidic linkage. In contrast, amylopectin is more highly branched
with an average of one branch point, a-l,6—glucosidic linkage, every 20 to 26 units
(Manners, 1985; Morrison & Karkalas, 1990). Because of the use of starch in both food
and non-food industries, there is a desire to increase the content of starch in some plants
as well as manipulate the quality of starch.

Glycogen has been found as one of the major carbon and energy reserves in many
bacteria as well as in eukaryotic organisms. It is a branched polysaccharide that contains
about 90 % of its glucose in a-1,4-glucosidic linkages and the rest in a-1,6-linkages.
Glycogen usually accumulates in the bacterial cell when there is excess of carbon in the
media and when the nutrition supply for growth is limited (Preiss & Romeo, 1989). It is
considered as a storage compound which can be degraded to an energy form that is utilized

by the cell, enabling it to survive better under certain conditions (Preiss, 1984). The

 

8

observations that glycogen-containing bacterial cells survived longer than glycogen-less
bacterial cells under starvation conditions (Strange etal., 1961; Strange, 1968; Van Houte
& Jansen, 1970), and the utilization of glycogen during spore formation (Strasdine, 1968,
1972; Mackey & Morris, 1971; Slock & Stahly, 1974; Brana etal., 1980), are consistent

with the role of glycogen as an energy reserve.

2. Synthesis of Starch and Glycogen:

The biosynthesis of starch in algae and higher plants and of glycogen in bacteria
share common reactions. The ADP-glucose pathway is the predominant pathway (Preiss
& Romeo, 1989; Preiss, 1991, 1996; Preiss & Sivak, 1996) which is shown below.

1. ATP + oc-glucose-l-P ~ ADP-glucose + pyrophosphate

2. ADP-glucose + a-1,4-glucan ~ ADP + a-1,4-glucosyl-a-1,4-glucan

3. Elongated a-1,4-glucan chain - Branched a-1,6- a-1,4-glucan

First, the sugar nucleotide, ADP-glucose, is synthesized from ATP and glucose-l-P
in a reaction catalyzed by ADP-glucose pyrophosphorylase (ATPza-D-glucose-l-P
adenylyltransferase, EC. 2.7.7.27). In the next reaction, the glucosyl moiety of the sugar
nucleotide is transferred to a maltodextrin, glycogen or starch in an a-1,4-glucosidic
linkage by glycogen synthase or starchsynthase (BC. 2.4. 1.21). In the third reaction of
the pathway, branching enzyme (BE; EC. 2.4.1.18) catalyzes the synthesis of the a-1,6-

glucosidic linkages found in amylopectin and in glycogen.

9
3. The Role of ADP-Glucose Pyrophosphorylase in Biosynthesis of Starch and
Glycogen:

ADP-glucose pyrophosphorylase catalyzes the conversion of glucose-l-P and ATP
to ADP-glucose and pyrophosphate (PPi), the ﬁrst step in the biosynthesis of starch in
plants (Preiss, 1991, 1996; Preiss & Sivak, 1996) and glycogen in bacteria (Preiss &
Romeo, 1989, Preiss, 1996).

3.1. ADP-Glucose Pyrophosphorylase in Starch Synthesis:

In 1962, Espada ﬁrst described the enzymatic synthesis of ADP-glucose by ADP-
glucose pyrophosphorylase in soybean. ADP-glucose pyrophosphorylase was later found
distributed in chloroplasts and amyloplasts in which starch synthesis occurs (Okita et a1. ,
1979; ap Rees et al., 1984; Lin etal., 1988). The activity of this enzyme is sufﬁcient to
account for the rates of starch synthesis according to the studies on various plant tissues
(Ozbun et al., 1973; Heldt et al., 1977; Okita etal., 1979; ap Rees et al., 1984; Edwards
et al., 1988).

It is believed that ADP-glucose pathway is the main pathway leading to starch
synthesis. Data obtained from a number of biochemical and genetic studies support this
view. Mutants of maize endosperm having deﬁcient levels of ADP-glucose
pyrophosphorylase (5-10 % of wild-type) also have deﬁcient level of starch content (25-30
% of wild-type) (Tsai & Nelson, 1966; Dickinson & Preiss, 1969). Smith et a1. (1989)
have shown that one pea line with a mutation at the rb locus, which controls the level of
ADP-glucose pyrophosphorylase activity in developing pea embryos, contained 3-5 % of

the ADP-glucose pyrophosphorylase activity and 38-72 % of the starch found in the pea

10

line having the dominant genes. Lin and colleagues (1989) ﬁrst showed that starch
biosynthesis in the chloroplast is entirely dependent on the pathway involving ADP-
glucose pyrophosphorylase. A mutant of Arabidopsis thaliana containing less than 2 %
of the leaf starch content found in the normal strain had only 2 % of the normal leaf ADP-
glucose perphosphorylase activity (Lin et a1. , 1988a). A chimeric gene encoding
antisense ribonucleic acid for the ADP-glucose pyrophosphorylase was expressed in potato
tuber, causing a decrease in enzymatic activity to 2-5 % of the wild-type level and a
commensurate decrease in starch content (Mﬁller—Rober et al. , 1992). Stark et a1. (1992)
provided persuasive evidence for the role of ADP-glucose pyrophosphorylase in starch
synthesis. They increased ADP-glucose pyrophosphorylase activity by transformation of
the potato plant with a mutant Escherichia coli (E. coli) ADP-glucose pyrophosphorylase
gene insensitive to the regulators that affect the plant enzyme and found that the starch
content of potato tubers was increased by 30-60 %. Other data demonstrating a direct
relationship between activity of the ADP-glucose pyrophosphorylase and starch
accumulation in several plant species (Preiss & Levi, 1980; Preiss, 1988; Okita, 1992)
also support the hypothesis that the ﬁrst reaction catalyzed by ADP-glucose
pyrophosphorylase is the major regulatory step in starch synthesis.

There has been no relationship noted between starch synthesis and phosphorylase
or UDP-glucose pyrophosphorylase activity. Many studies have shown that the kinetic
properties of the enzymes in the ADP-glucose pathway (80.5 and Vm values) are consistent
with the known concentrations of substrates and effector metabolites present in plant cells.

In contrast, for the UDP—glucose-speciﬁc starch synthases and plant phosphorylases, the

11

high $0.5 values for the respective substrates (U DP-glucose and glucose-l-P) compared to
the actual cellular levels, strongly argues that these two enzymes do not play a signiﬁcant
role in starch biosynthesis.

3.2. ADP-Glucose Pyrophosphorylase in Glycogen Synthesis:

In mammalian tissues and eukaryotic micro-organisms, UDP—glucose was
considered to be the glycosyl donor for glycogen synthesis (Leloir et al., 1959; Stalmans
& Hers, 1973; Krebs & Preiss, 1975). In 1964, Sigal et al., reported that many UDP-
glucose pyrophosphorylase—deﬁcient mutants of E. coli were. still able to produce normal
amounts of glycogen. It was shown that all bacterial glycogen syntheses that have been
tested utilize ADP-glucose as a substrate much better than UDP-glucose (Greenberg &
Preiss, 1964). It was also shown that extracts of several bacteria contained both an ADP-
glucose pyrophosphorylase (Shen & Preiss, 1964) and an ADP-glucose-speciﬁc glycogen
synthase (Greenberg & Preiss, 1964). Mutants of E. coli and Salmonella typhimurium (S.
typhimurium), affected in ADP-glucose pyrophosphorylase and/or glycogen synthase, are
either glycogen-deﬁcient or glycogen-hyperproducing (Preiss, 1984; Preiss & Romeo,
1989). These studies suggested that bacterial glycogen synthesis utilizes ADP-glucose,
instead of UDP-glucose, as the glycosyl donor. ADP-glucose is produced from glucose-1-

P and ATP in the reaction catalyzed by bacterial ADP-glucose pyrophosphorylase.

4. Regulation of the ADP-Glucose pathway:
Regulation of the ADP-glucose pathway is believed to occur at the ADP-glucose

synthetic step in both starch synthesis (Preiss, 1984; Kruger, 1990; Preiss & Sivak, 1996)

12

and bacterial glycogen synthesis (Preiss & Romeo, 1989; Preiss, 1991 , 1996). The activity
of ADP-glucose pyrophosphorylase is allosterically regulated by metabolites. This is in
agreement with the concept that regulation of a biosynthetic pathway occurs at its ﬁrst
committed step.

4.1. Regulation of the ADP-Glucose Pathway in Starch Synthesis:

There is much evidence suggesting the major activator of ADP-glucose
pyrophosphorylase is 3PGA and the major inhibitor is Pi. It was ﬁrst reported by Ghosh
and Preiss (1965) that the spinach leaf enzyme is activated by 3PGA and inhibited by Pi.
This was also seen for ADP-glucose pyrophosphorylases from other higher plants (Sanwal
et al., 1968), green algae (Sanwal & Preiss, 1967; Ball et al., 1991), and blue-green
bacteria (Levi & Preiss, 1976; Iglesias et al., 1991). ADP-glucose pyrophosphorylases
from other species (Preiss, J. , 1982) had been shown to be activated by 3PGA. 3PGA also
caused the enzyme to have higher afﬁnity for the substrates, ATP and glucose-l-P, and
lower sensitivity to the inhibitor, Pi.

Both in vitro and in vivo experiments suggested that the activation of ADP-glucose
pyrophosphorylase by 3PGA and inhibition by Pi are physiologically important. Many of
these experiments (Preiss 1982, 1988, 1991; Preiss & Levi, 1980) showed a direct
correlation between the concentration of 3PGA and starch production, and an inverse one
between the concentration of Pi and starch content.

Pettersson and Pettersson (1989) applied modern control theory (Kacser & Burns,
1973; Kacser, 1987) to develop a kinetic model to determine the extent that metabolites,

which affect in vitro leaf ADP-glucose pyrophosphorylase activity, controlled the rate of

13

starch production under conditions of light and CO2 saturation. They reached the
conclusion that 3PGA and Pi play an important role in the regulation of starch synthesis.
In another Kacser-Burns analysis, it was shown that in Arabidopsis thaliana ADP-glucose
pyrophosphorylase is a major site of regulation for starch synthesis (Neuhaus & Stitt,
1990) and regulation of ADP-glucose pyrophosphorylase by 3PGA concentration and
3PGA/P, ratio is an important determinant of the rate of starch synthesis in vivo (Neuhaus
et a1. , 1989). i

In 1991, Ball et a1. isolated a starch-deﬁcient mutant of Chlamydomonas reinhardii
in which ADP-glucose pyrophosphorylase could not be effectively activated by 3PGA.
This work shows that activation of ADP-glucose pyrophosphorylase by 3PGA is important
physiologically.

4.2. Regulation of the ADP-Glucose Pathway in Bacterial Glycogen Synthesis:

Regulation of the ADP-glucose pathway occurs at the ﬁrst step, the ADP-glucose
synthetic step, in bacterial glycogen synthesis (Preiss & Romeo, 1989; Preiss, 1991, 1996)
as in starch synthesis. The activity of bacterial ADP-glucose pyrophosphorylase is also
allosterically regulated by metabolites (Preiss, 1984; Preiss & Romeo, 1989).

For most bacterial ADP-glucose pyrophosphorylases tested, glycolytic
intermediates, FBP, fructose-6-P, pyruvate, or 3PGA are the major activators while AMP,
ADP, or P, are the major inhibitors (Preiss & Romeo, 1989; Preiss, 1996). The activators
increase the apparent afﬁnity of the enzyme for its substrates, ATP and glucose-l-P. In
addition, the activators may reverse the inhibition of the enzyme by AMP, ADP, or P,.

The activators are required for optimal activity, and may be considered as

14

indicators of carbon in excess of energy needs (Preiss, 1984). The bacterial ADP-glucose
pyrophosphorylase is more diverse than the plant enzyme with respect to the speciﬁcity
for activators. The activator speciﬁcity of ADP-glucose pyrophosphorylases from various
bacterial species has been classiﬁed into seven groups (Preiss, 1984; Preiss & Romeo,
1989). It was postulated that the activator speciﬁcity seen in Table 1 is correlated with the
type of carbon degradation or assimilation pathway in the organisms (Preiss, 1984; Preiss
& Romeo, 1989).

For example, when bacteria, such as E. coli and other enterics, obtain their energy
mainly through glycolysis, ADP-glucose pyrophosphorylases are usually activated by FBP
(Preiss, 1984, 1996; Preiss & Romeo, 1989). Regulation of the glycolytic pathway is at
the phosphofructokinase step, the synthesis of FBP. For those organisms with Entner-
Doudoroff pathway as their predominant pathway, the major activators for ADP-glucose
pyrophosphorylases are fructose-6-P and pyruvate (Preiss, 1969, 1984, 1996; Preiss &
Romeo, 1989). These organisms have no phosphofructokinase but phosphoglucoisomerase.
For bacteria such as Rhodospirillum rubrum, pyruvate is the only activator of the ADP-
glucose pyrophosphorylases (Furlong & Preiss, 1969). These bacteria grow anaerobically
on pyruvate or lactate or on C02, and pyruvate is a product of CO2 ﬁxation under
anaerobic photosynthesis. For the photosynthetic cyanobacterial ADP-glucose
pyrophosphorylases, 3PGA, the initial CO2 ﬁxation product of oxygenic photosynthesis,
is the major activator (Preiss & Levi, 1980; Preiss, 1988, 1989, 1996). Thus, ADP-
glucose pyrophosphorylases, which are involved in synthesizing one of the ﬁnal products

of photosynthesis, glycogen, are stimulated by the ﬁrst product of CO2 ﬁxation, 3PGA.

15

 

Table 1: Activator speciﬁcity of ADP-glucose pyrophosphorylases from various bacterial

species“

 

bacteria

activator(s)

 

Clostridium pasteurianum, Enterobacter hafniae,
Serratia spp. (Ser. liquifaciens, Ser. marcescens)

Citrobacter freundii, Edwardsiella tarda,
Enterobacter aerogenes, Enterobacter cloacae,
Escherichia aurescens, Escherichia coli,
Klebsiella pneumoniae, Salmonella enteriditis,
Salmonella typhimurium, Shigella dysenteriae

Rhodospirillum spp. (Rh. ﬁtlvum,
Rh. molischianum, Rh. photometricum,
Rh. rubrum, Rh. tenue), Rhodocyclus purpureus

Anabaena 7120, Aphanocapsa 6308,
synechococcus 6301 , Synechocystis 6803

Aeromonas hydrophila, Micrococcus luteus,
Mycobacterium smegmatis, Rhodobacter viridis

Agrobacterium tumefaciens, Arthrobacter viscosus,
Chlorobium limicola, Chromatium vinosum,
Rhodobacter spp. (R. acidophila, R. blastica,

R. capsulata, R. palustris), Rhodomicrobium vannielii

Rhodobacter spp. (R. gelatinosa,
R. globiformis, R. sphaeroides)

none

ﬁ'uctose-l ,6-P2

pyruvate

3-P-glycerate

W'l ,6'P2
fructose-6-P

pyruvate
fructose—6-P

fructose-l ,6-P2
pyruvate
fructose-6-P

 

“ This table is based on Table 2 in a review (Preiss & Romeo, 1989).

 

16

ADP-glucose pyrophosphorylase from Rhodobacter spheroides is effectively activated by
either FBP, fructose-6-P, or pyruvate (Y ung & Preiss, 1982). This is consistent with the
view that there seems to be a correlation between the nature of the activator and the major
assimilation pathway, since Rhodobacter spheroides can metabolize glucose by both
glycolytic pathway and Entner-Doudoroff pathway and can assimilate CO2 during
anaerobic photosynthesis.

It is also shown in Table 1 that various groups may have an overlapping activator
speciﬁcity. This suggests that the structures of the activator sites for different groups are
related. It would be of interest to determine the natures of the activator sites from different
classes and compare them with each other.

AMP, ADP, and Pi are the major inhibitors for many bacterial ADP-glucose
pyrophosphorylases (Preiss, 1973, 1978; Preiss & Walsh, 1981). For the bacteria which
undergo glycolysis or the Entner-Doudoroff pathway, ADP-glucose pyrophosphorylases
are mainly inhibited by AMP and ADP or AMP alone. For cyanobacteria, the enzymes
are mainly inhibited by P,. Increasing concentrations of activator either reverse or prevent
the inhibition caused by the inhibitors (Preiss, 1969, 1973, 1978; Preiss & Walsh, 1981).
The activators may be considered as signals of carbon excess in the cell while the
inhibitors as indications of the low energy charge. This suggests that optimal ADP-glucose
and glycogen synthesis only occur with excess carbon and at high energy charge (Preiss,
1984).

It has been shown that allosteric regulation of bacterial ADP-glucose

pyrophosphorylase is important for regulation of glycogen synthesis in vivo. The ability

17
to accumulate glycogen in the mutant cells of E. coli (Cattaneo et a1. , 1969; Govons et al. ,
1969, 1973; Preiss et al., 1971, 1975, 1976) and S. typhimurium LT-2 (Steiner & Preiss,
1977) can be correlated to the altered regulatory properties of the mutant ADP-glucose
pyrophosphorylases. The mutants containing ADP-glucose pyrophosphorylases with higher
afﬁnity for the activator, FBP, and/or lower afﬁnity for the inhibitor, AMP, accumulate
more glycogen than the wild-type strain, while those containing ADP-glucose
pyrophosphorylases with lower afﬁnity for the activator accumulate less glycogen (Preiss

& Romeo, 1989).

5. Subunit and Primary Structure of ADP-Glucose Pyrophosphorylase:

ADP-glucose pyrophosphorylase from all sources is tetrameric in structure. For
all the bacterial enzymes studied to date, the enzyme is a homotetramer, with a molecular
mass of about 50 to 55 kDa for each subunit (Preiss, 1984). The higher-plant enzyme is
more complex, being heterotetrameric with two related but different subunits having
molecular masses in the 50 to 60 kDa range (Preiss, 1996).

The most studied higher-plant ADP-glucose pyrophosphorylase with respect to
structural properties is the spinach leaf enzyme (Morell et al., 1987; Preiss et al.,
1987a,b). This enzyme is composed of two different subunits with molecular masses of
51 and 54 kDa. The subunits can also be distinguished by amino acid composition, N-
terminal amino acid sequences, tryptic peptide mapping on high-performance liquid
chromatography (HPLC), and antigenic properties. The 51 kDa and 54 kDa subunits are
usually referred to as the small and large subunits, respectively. In irnmunoblotting

experiments, the polyclonal antibody raised against the small subunit react strongly with

18

the small subunit but weakly with the large subunit, and vice versa. These studies suggest
that the small and large subunits of the spinach leaf enzyme are structurally different
(Morell etal., 1987) and coded by two different genes. ADP-glucose pyrophosphorylases
from Arabidopsis thaliana leaf (Lin et al., 1988a,b), maize endosperm (Preiss et al.,
1990), potato tuber (Okita et al., 1990), and Chlamydomonas reinhardtii (Iglesias et al.,
1994) have been studied and they have been shown to be composed of two different
subunits.

Cyanobacteria are prokaryotic organisms that have metabolic properties similar to
chloroplasts of higher plants (Aitken, 1988). Mereschowsky originally proposed that the
chloroplast of eukaryotic cells originated by endosymbiosis of a cyanobacterium
(Mereschowsky, 1905). However, the former organisms synthesize glycogen as the major
carbohydrate reserve (Shively, 1988), in a similar manner to what is observed in other
bacteria. The cyanobacterial ADP-glucose pyrophosphorylase is homotetrameric similar
to the E. coli enzyme, but is regulated by 3PGA and P, like the higher-plant enzyme (Levi
& Preiss, 1976; Iglesias et al., 1991; Charng et al., 1992).

The structural genes of ADP-glucose pyrophosphorylases (glgC genes) of E. coli
and S. typhimurium have been cloned (Okita, etal., 1981; Leung & Preiss, 1987a) and
sequenced (Baecker et a1. , 1983; Leung & Preiss, 1987b). Amino acid sequences of these
two glgC genes have been deduced from the nucleotide sequences and compared (Figure
1). The total amino acid compositions of the deduced amino acid sequences are in
accordance with those determined by acid hydrolysis (Haugen et al., 1976; Lehmann &

Preiss, 1980) and the deduced amino acid sequences are identical to the partial sequences

19

Figure 1: Alignment of the primary structure of ADP-glucose pyrophosphorylase
from spinach leaf large subunit (31-54) and small subunit (sl-51), wheat leaf (wl), maize
sh-2 and bt-2 loci, wheat endosperm (we7 and we3), potato tuber large subunit (pot-l) and
small subunit (pot-s), Anabaena (ana), Synechocystis (syn), rice seed (rice), Arabidopsis
thaliana (arabid), E. coli (ecoli) and Salmonella typhimurium (s.t.). The number indicated
is corresponding to the sequence of the Anabaena enzyme.

.n unshah

35E. 4E8 4% M2. a. ”Ma. 30% KOHﬁg§0ﬂ09ﬂb0§00§h§4§
(“490493. .g004ae. "B. a. E. 3040HNO>E0 09 09H>0§0QZHUZQA4hQHHEO§4§g
4n.—4. 900.5». .Eh #8252039. :H2m094mmg mﬁgégggHa§g4§ﬂh

...» .:ﬁZﬂ magi. .0022049 agaﬁggggHﬂH§g4§A
5900.5»: 05mm 000E039. .ghagaﬁgggﬁanHgg4gh
4n4900h.».. 3:0004EOE. .gﬁagaﬁgégngﬂHgg4gh
404900;. .Eﬁm004g0g. .sgﬂggaﬁggngagg4gﬂh
5900.3». .Gggﬂ0. h. ..00h 2. NggﬁggﬁéHa§g4§
4949009.». .Egﬂh.h. .005. 2. 93§E09§H HEZHUZNEHGHEE4 E
(049005;. a. >0 §§a§§§a§H3§4 4949 “mg
40490.95». 04593094405589. ha. H0. ENNH :ggEEHKBH mgHﬂHgHAbagng
094900;. 04a§09§>00ﬁ4. hz. H.0 0§§90E0§00§Ha§4ﬁ9§§

34900.0». .0 ....... §.§ME.H4. . . . .EODZHOEHQ.H400H§4Q. . . . . .
. .

Egg . E . A30§4§0>2

ENQUSH: . >04 . A30§§0§

N§§H Hgaéma S§0§§

”Egan HOE4h0002> .........

949040003" Hangman—Mamas» . . .

HEEHHéﬁgmz . .

94904083” H§§m§9§m§a§g§

9&904000ﬂHH4AE

”Egan—"g

§0§§§§§0§E9§

Hugging . 6003059904439 . 04Eg§m00m40m04940 . ﬁ0§§ﬂmm§ . . .
300090§HH40>§§0090§§9QH0009§§.KHOMKO0HOEHEHODOHUWNHON . . . .

Egg. . . . . . .mangHEngE .......
_

éséiiiS-{i‘ééiéi‘i

«««««

HHHHH

vvvvv

A . unoov .n Gun—bah

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

{In} I o o 5: 4.2Idd. Hag 00000000000000 E
I I o I gag“ 43:3 00000000000000 Ema
hangar ............ on» . . §¢o>§oumagmgmohn . . 43:3]. . ASHE
ESE ............ 8N . . ﬁ‘GNrﬁSNB‘UWEMUhQ . . . plll 43:3 0 ASHE
533E» ............ on». . 82053395 .3393 . . . ii .. > Saga. .ASHE
Egan—Hz» ............ c8— . . Higgmagnnbumohn . . > .333. . 39.3
5.5585 ............ on» . . ﬁdgggaomgmumn . . tag . . ASHES
ESHEH ............ Ge . . .HH4OR . EQEQWHHHNUhQ . . . I iaiuh .§.§§s
. EHUHE ............ nae. ..HN40>.Z.N§AHHHMO&Q. . . . 435.03.0Hgg
EBGHE ............ E . . HN40> .ggﬁHHﬂmohﬂ . . . III 0 . Nun . Eonaﬁmﬂgﬂg
ESHg ............ E. ogg>.§H§a. . Ila. .ﬁﬁ.§§s
Managua» ............ an... . .>h¢o>.§mguum0hnm . . -ltnn. .58 . Egan
ESE ............ 2E. .Huédmnnpdmmqnnmounm. . - Eéasdngmgu
.553ng ............ BC. .qhgiugmuunuohn. . .....- .Eu.»>aa§.§.nu .........
hm. .SHE ............ a. .g‘. 0 I I . .‘ﬁHHgn. C .ﬁag. . . . DE
. _
O m N o o N
0.54 .§.a>h§.n I19): 133%. .4“. . .HNKHHEE
2% NE. Elsa 4. I." I ~ ...—333% . .4K . . .Nan-Hzg
I .5333. ....r E. .ZN . .ggag
liviU-n - III-H”
I- -u .55. a E. E. . Eggnog
-I:u .. .133 ..... r 5.5. an . Enron»
I. nan-”In L ngu E Iggog
bag—nu... .53! ...... E. Q.» . HEHHOE
I .. . :55... a £5.39. .8». .nadﬁauog
... E. g
lip LL- :u-Iul I II. M“:
I F Egg
g i I. Jan F .E ........
Ila-r 1.3.3.... '1 ganﬂHaH" ...GEEOE

 

 

H>BO.QQAA¢QKH>B>.HO..OKHhHUQBﬂZ‘dﬂObZQ‘ZﬂHM>ZDOH¢£ZKQHH‘MKHUMZMUHGHQHOQHH‘AAK

HbBO.MﬁHA¢GKH>B>.HO..MKHhHUQBMN¢¢HO>ZGm2HHMbZﬁUHMdZHGHH‘dHHMMZOOHOﬂ>>MOM¢¢AAK
HHHO.MQHA¢QKH>B?.HO..mKHhMDQBHM14HO>ZQMZHHK>ZHUHM¢ZKGHH‘ﬁﬂHﬂDZKUHOHﬁbMQM‘dAAK
ddﬁﬂhﬂaﬂ.¢ddd¢ﬂKH>B?.HO..MKHhHOQBﬂ¢¢4HO>ZD>ZHHK>ZDOHN¢2KDHH‘ﬂﬂHﬂDZKOHGHmHOQKHOﬂAM
.Q‘HBbZKH>>>.HO..2“HHhUAHHMZdHObZHMZbHEﬁZKOHM‘ZMGHH‘ﬂﬂHBBOmOHaddHﬂaﬁﬂﬂﬂBD
.QBHbi2Kﬂ>>?.HO.cMMHNhOOmHNQ4MO>ZDKZHHM>DNOHM¢ZMQHH‘ﬂﬂHHBGNOHDbNHQUKGHMUO

H C
o B .HO..NKHMHUHHmmndﬁOHOMMZBbeZMUHK‘ZSHHHUBﬂHKHZMUHOHQbKQ‘AAHKm
.ndem2KA>>B.HO..QNHNhOHm¢MG¢HOZUﬂ¢ZdHBbZKUHM‘ZKQHHDZOHNHZHUHOthKUH‘ﬂOQG
.DKH d2ﬂhH>B.HO..m..H>DG...ndﬂabbﬂm2MHHbzo. M‘ZKDHH‘GKH lllll
.

Fl IFI-rl ﬂ
?hna aavgvaa
cetzaiﬁuaw £2;

.o8>Hﬁﬂd>lﬂﬂH20hﬂBHﬂ>Mhhﬂ>mO>>>M ..... OMHHOOOh>HMKdBZUNmUNKNOth‘MﬁAMHSMBMHﬁBBODHBQﬂOﬁB?.NdﬂKNZdMR .0..
..BhﬂhﬂﬂhlﬂﬂHﬂUth>ﬂ>ﬂuhﬂbﬂﬂ>>>h ..... OnH>000ubﬂmzdhzummauﬂﬁabhﬂdﬁﬁﬂmHBNBKHExagnhxndﬂﬁ>>.uddDQZdKS HHOO.
MH‘HHHHHD‘UZHQMQHHHddﬂmHUm¢ﬂ0>>h2ﬂHHUBKH>UOHDHbmnﬁbnd ..... Qﬂ>uKMQQHMKAOBHHﬁdﬁﬂnhhthmHmMKBHOQZGBH NIUA
. . . .hﬂEHg .0.—"Al“!
MH‘GBMMHD‘USHABQHHH‘GﬂmHOMMAUHbMNNHKUZKHbUOHUHbmnBHnﬂ.....ndm.HummﬂﬂﬂﬁaﬁhHﬁﬂﬂﬂﬂﬂhﬂhﬂﬁbﬁﬂﬂﬁHOﬂZ‘Zﬁ HnIHI
ﬂﬂddﬂﬂNNDRUZHQDQHHH‘UﬂuHUNKAD>>MNNHKUBKH>UOMOH>hﬂB>Dﬂ ..... OAS.KQQQHHMAOHMHh‘ﬂﬂﬁhhhhﬂhbhﬂﬂﬂHOﬂZ‘Zﬂ IIHOQ
KH‘HBHHHD‘GSRAQQHHH‘UﬂmHOmﬂAD>>NNNHKUZKH>UOﬂDH>hnb>nﬂ ..... nﬁbuKMﬂﬁQNKQOBNHﬁdﬂﬂﬂhhhhﬂﬁbhﬂkﬁHOﬂBﬂZﬂ .Oﬂu
“HumMHHMDZDS>QBQHHBUQMHHMMGHOQ>MNNHMUOKHZUONUHEMHBme...oo2H3oKﬁhﬁﬂhmOMBHH“(KHZHthQmmmoﬂaﬂdﬂzﬂﬂﬂ ah.
Mﬂ>m¢dhhnddzuﬂmHHHSUOBHHNMM>OA>MMOHKUZMAHUDHUHHMHBbMU ..... nagoKBmmﬂNMdMBHHﬁdﬂﬂnﬂhmhmm2h008ﬂ4HZRZH lﬂl
HHMHBONHD£UZZhBQMdﬂ>OUQAMNMHD>HMNH>QUQKHMUDNMHH‘GKHKU ..... ZDH.Kﬁmmﬂhﬂmmﬂhhmﬁﬂmnhhahﬂm.hﬂOﬁﬂdﬂMﬂZﬂ HIOOQ
HDHQHHMD‘USZS‘ZKAHM mZAKmm>UHmBﬁHHMUHMAhUONAHHdﬂuHMU..... Gm.KBNhAHMﬁWthABKQQNhﬂhKQ.mOHUﬂm221QH MI?
SMHHBHHMQ‘USSZ‘ZKAHMUMZANQo..htﬁﬂﬂHKUﬂKﬂhUOﬂAHH‘HKHMU ..... Mam Kammﬂﬂmmmﬁhhmﬁﬂmnhhﬂhﬂm.AOHUQGM21ﬂh DI! .
dﬂﬂﬂBﬂMHﬂdUSSPMQKQHUOWN>EMD>UH>MNHHZUMMQAUOGMthﬁMEKU ..... OH OH hﬂUﬂﬁ‘ HhﬁBKhﬂHhﬂhK.thHBﬂdﬂzdﬁﬁ Nlﬁl
KHSEHHHh946Q3?BQKAN>ZmOHMNMHOEhNNHbMUKDAhUOMmHHMGBHKM ..... 2m ﬁdZﬂﬂﬁHHﬂhKﬂdﬂhhﬂhM.mmaﬂdﬂ¢A2dﬂh #3
dﬂHﬂMBDdG AABQKHN>BQB?..... Mm mazmmmahﬂﬁﬂﬁdﬁhhuhﬂoQQMHHﬂdAZdﬂh ind.

23

determined by Edman degradation (Baecker et al., 1983; Leung & Preiss, 1987b). There
is 90 % identity in the deduced amino acid sequences of the E. coli and S. typhimurium
glgC genes, and most of the changes are conservative (Preiss & Romeo, 1989).

The structural genes of the higher-plant ADP-glucose pyrophosphorylases have
been cloned from several sources and tissues (Figure 1), such as maize endosperm (Barton
et al., 1986), rice seed (Preiss et al., 1987a; Anderson et al., 1989), wheat leaf and
endosperm (Olive et al. , 1989), potato tuber (Miiller-Rober et al. , 1990; Anderson et al.,
1990; Nakata et al., 1991), and spinach leaf (Smith-White & Preiss, 1992). A systematic
comparison of the primary structures of ADP- glucose pyrophosphorylases from different
sources has been done by Smith-White and Preiss (1992). The small subunit is highly
conserved with an amino acid sequence identity varying from 85 to 95 % when the
enzymes from different plant sources and tissues are compared. Both the identity among
large subunits from different sources and that between the large and small subunits are
about 50 to 60 %. The identity of the E. coli enzyme with the higher-plant enzymes is
about 30 %.

Analysis of the deduced amino acid sequences (Figure 1) indicated that the
cyanobacterial enzyme is more similar to the higher-plant enzyme than to the other
bacterial enzymes in primary structure (Kakefuda et al., 1992; Charng et al., 1992).
Furthermore, the cyanobacterial enzyme is more similar to the small than to the large
subunit of higher-plant ADP-glucose pyrophosphorylases (Kakefuda et al. , 1992; Charng
et al. , 1992). This is in agreement with the fact that the cross-reaction of cyanobacterial

enzyme with the antibody prepared with the spinach leaf large subunit was much weaker

24

than with the antibody prepared with the small subunit (Iglesias et al. , 1991).

Recent studies on potato tuber ADP-glucose pyrophosphorylase have shown that
the small subunit expressed by itself exhibited high enzyme activity , with lower afﬁnity
for activator 3PGA and higher sensitivity toward Pi inhibition (Ballicora et al., 1995). It
is suggested that the small subunit of higher-plant ADP-glucose pyrophosphorylase
functions as the catalytic subunit while the large subunit is involved in the allosteric
regulation of the activity (Ballicora et al., 1995). Based on the above observations, it has
been postulated that the higher-plant ADP-glucose pyrophosphorylase genes evolved from
the bacterial enzyme gene by gene duplication and diversion (Smith-White & Preiss,

1992).

6. Structures of the Substrate and Regulator Binding Sites of ADP-Glucose
Pyrophosphorylase:

From steady-state kinetics (Paule & Preiss, 1971; Kleczkowski, et al., 1993) and
equilibrium binding (Haugen & Preiss, 1979) studies of the reaction catalyzed by ADP-
glucose pyrophosphorylase, substrates are known to bind in an ordered mechanism. ATP
binds ﬁrst, followed by the binding of glucose-l-P (Haugen & Preiss, 1979). Product
release is ordered with PPi released ﬁrst, followed by ADP-glucose (Paule & Preiss,
1971).

According to Haugen and Preiss (1979), there are four binding sites for substrates,
activators, and inhibitor per tetrameric E. coli enzyme. In binding experiments, CrATP,

ATP, and FBP appear to bind to only half of the expected sites in the tetrameric enzyme,

25

while ADP-glucose, the activators, pyridoxal-P (PLP) and 1,6-hexanediol-P2, and the
inhibitor, AMP, bind to four sites per tetrameric enzyme. Glucose-l-P does not bind to
the enzyme unless CrATP and MgCl2 are present and binds to four sites per tetrameric
enzyme. In the presence of glucose-l-P, CrATP binds to four sites per tetrameric enzyme.
It was suggested that the enzyme ﬁrst binds two ATP molecules per tetramer which allows
the binding of glucose-l-P, then binds two more ATP molecules. The binding of ATP and
FBP is synergistic. ATP and FBP must be present together to effectively inhibit AMP
binding. Thus, there are binding sites for substrates, activators, and inhibitor located on
each subunit and the binding sites can interact.

Structure—function studies to determine the location of substrate and regulator
binding sites of ADP-glucose pyrophosphorylase have been mainly performed by chemical
modiﬁcation using substrate or regulator analogs and site—directed mutagenesis of cloned
ADP-glucose pyrophosphorylases. Characterization of the mutant E. coli enzyme with
altered regulatory properties also provides very useful information.

In chemical modiﬁcation studies, several compounds have been used as the afﬁnity
labels for ADP-glucose pyrophosphorylase, PLP, 8—azido—ATP, 8-azido-ADP-glucose, 8-
azido-AMP, and phenylglyoxal. They all have the following properties. First, the afﬁnity
label has to be shown as an analog of the substrate, activator, or inhibitor. Second, the
compound can bind the enzyme covalently in chosen conditions. Then, the labeled enzyme
is subjected to proteolysis and the labeled peptide(s) is isolated and sequenced (Preiss &
Sivak, 1996).

PLP is an activator of many bacterial and higher-plant ADP-glucose

26

pyrophosphorylases (Parsons & Preiss, 1978a; Preiss et al., 1987b; Iglesias et al., 1991)
and is considered as an analog of 3PGA. It has been used as a site speciﬁc probe to
identify lysyl residues in enzymes that have binding sites for sugar phosphates. A Schiff
base is formed between PLP and the e-amino group of lysine and can be converted to a
covalent bond by reduction with sodium borohydride (NaBH4).

The photoafﬁnity labeling agents 8-azido-ATP and 8-azido-ADP-glucose are
considered as substrate analogs (Lee et al., 1986) and 8-azido-AMP as an inhibitor analog
(Larsen et al. , 1986). In the presence of UV light, these 8-azido compounds form a nitrene
radical which can react with electron-rich residues, and therefore, become covalently
incorporated into the enzyme.

Phenylglyoxal is an arginine-speciﬁc reagent (Takahashi, 1968). It reacts with the
guanido group of the arginine residue under mild conditions, pH 7 to 8 and 25 °C. The
derivative contains two phenylglyoxal moieties per guanido group.

6.1. The Primary Structure of the Substrate Binding Sites:

Reductive phosphopyridoxylation of E. coli ADP-glucose pyrophosphorylase at
high concentrations of PLP (Parsons & Preiss, 1978a) cause a decrease in enzyme activity,
and this inactivation could be prevented by the presence of ADP-glucose plus Mg“. After
reductive phosphopyridoxylation in the presence of FBP or hexanediol-l ,6-P2, the labeled
enzyme was cleaved with CNBr. The labeled peptide was isolated and sequenced. Lys‘”
was found to be labeled with [’11] PLP. These experiments suggested that Lys‘” is required
for either substrate binding or for catalytic activity.

Site-directed mutagenesis experiments have been performed to further study the

27
role of Lys‘” (Hill et al., 1991). The apparent afﬁnity for glucose-l-P of the mutant
enzyme, which was generated by substitution of Lys‘” with glutamate, is 12,000-fold
lower than that of the wild-type enzyme. The kinetic constants for ATP, Mg“, ADP-
glucose, and FBP are relatively unchanged. The catalytic efﬁciency of the mutant enzyme
is similar to that measured for the wild-type enzyme. According to the kinetic studies on
a series of Lys‘95 mutants, both the charge and size of the this lysine are believed to be
important for proper binding of glucose-l-P.

Figure 2 shows the alignment of the E. coli amino acid sequence near Lys195 with
known sequences from other sources. Lys195 is conserved in all the sequences. Recently,
site-directed mutagenesis was performed on potato tuber ADP-glucose pyrophosphorylase
and the corresponding lysine was shown to be involved in the glucose-l-P binding of the
potato tuber enzyme (Fu et a1. , unpublished data).

To obtain further information on the substrate binding sites of the E. coli ADP-
glucose pyrophosphorylase, the photoafﬁnity labeling agent 8-azido-ATP and 8-azido-
ADP-glucose were used (Lee et al., 1986; Lee & Preiss, 1986). The kinetic experiments
indicated that these two agents speciﬁcally interact at the substrate site of ADP-glucose
pyrophosphorylase. Photo-inactivation of the E. coli enzyme with either 8-azido-ATP or
8-azido-ADP—glucose, in the presence of FBP and Mg2+ , was effectively prevented by the
presence of either ATP or ADP-glucose. The incorporation sites of [“C]8-azido-ADP-

glucose into the E. coli enzyme were determined by tryptic digestion of the modiﬁed

28

Figure 2: Amino acid sequence of the phosphopyridoxylated peptide, which was
protected by ADP-glucose plus Mg“, of E. coli ADP-glucose pyrophosphorylase and the
alignment with other known sequences. The labeled lysine residue, Lys‘”, is in bold. The
numbers indicated correspond to the sequence of the E. coli enzyme. Nomenclature is the
same as in Figure 1.

51-54
wl
sh—2
we7
we3
pot-l
ana
syn
rice
pot—s
sl-Sl
arabid
bt-Z
ecoli
S.t.

29

182 195 200

I I I
K ...... VLSFSEKPKGDD
KIDDTGRVISFSEKPRGAD
KIDHTGRVLQFFEKPKGAD
KFDSSGRVVQFSEQPKGDD
KFDSSGRVVQFSEKPKGDD
KIDSRGRVVQFAEKPKGFD
KIDNSGRVIDFSEKPKGEA
KIDAQGRITDFSEKPRGK.
KIDEEGRIVEFAEKPKGEQ
KIDEEGRIIEFAEKPQGEQ
KIDETGRIIEFAEKPKGEQ
KIDEEGRIIEFAEKPKGEH
KIDEEGRIIEFAEKPKGEQ
AVDENDKIIEFVEKP.ANP
AVDESDKIIDFVEKP.ANP

Figure 2

30
enzyme and isolation of the labeled peptides via HPLC (Lee & Preiss, 1986). Amino-
terminal sequence analysis showed that about 65 % of the label was in the region of 108
to 128 and Tyr“4 was believed to be the acceptor of the nitrene radical. This region is
highly conserved in all sequences known to date (Figure 3). About 20 % of the label was
incorporated between Cys163 and Lys208 and Lys195 is in that region. Another region with

15 % incorporation is between Arg381 and Mg”. Tyr114

was found in a predicted
Rossman—fold supersecondary structure, a structure consisting of three consecutive strands
of parallel B—pleated sheet and two joining a-helices (Larsen et al., 1986). Study of the

114

mutant enzyme with a Tyr to phenylalanine substitution via site-directed mutagenesis

indicated that Tyr‘“ is involved directly or indirectly in enzyme catalysis (Kumar et al. ,
1988).
6.2. The Primary Structure of the Regulator Binding Sites:

Previous studies have shown that the activator-binding site of the E. coli enzyme
is near the N-terminus (Parsons & Preiss, 1978a,b; Gardiol & Preiss, 1990). Reductive
phosphopyridoxylation of the E. coli ADP-glucose pyrophosphorylase in the presence of
ADP-glucose plus Mgl+ led to formation of an enzyme which no longer required FBP for
high activity (Parsons & Preiss, 1978a). The incorporation of PLP was inhibited by the
presence of other activators, such as FBP and hexanediol- l ,6-P2. These results suggest that
reductive phosphopyridoxylation occurred at the activator binding site (Parsons & Preiss,
1978a). The E. coli enzyme labeled with PLP was degraded with CNBr and the labeled
peptide was isolated and sequenced (Parsons & Preiss, 1978b). Lys39 was found to be the

labeled lysine residue. The region around Lys39 (Figure 4) is highly conserved in all the

31

Figure 3: Amino acid sequence of the major azido-ADP-P‘C] glucose incorporated
peptide of E. coli ADP-glucose pyrophosphorylase and the alignment with other known
sequences. Tyr‘“ is in bold. The numbers indicated correspond to the sequence of the E.
coli enzyme. Nomenclature is the same as in Figure 1.

81-54
sh-Z
we7
pot—l
ana
syn
rice
pot—s
sl-Sl
arabid
bt-2
ecoli
s.t.

108

....S.

32

129

.WFQGTADAVRQFGWLFED

PEEPAG.WFQGTADSIRKFIWVLED
PGEAAG.WFRGTADAWRK.IWVLED
PGEAGKKWFQGTADAVRKFIWVFED

PENPN.
KDNPD.
PDNPN.
PENPD.
PENPD.
PENPN.
PDNPN.
MKGEN.
MKGEN.

.WFQGTADAVRQYLWMLQE
.WFQGTADAVRQYLWLFRE
.WFQGTADAVRQYLWLFEE
.WFQGTADAVRQYLWLFEE
.WFQGTADAVRQYLWLFEE
.W .................
.WFQGTADAVRQYLWLFEE
.WYRGTADAVTQNLDIIRR
.WYRGTADAVTQNLDIIRR

Figure 3

33

Figure 4: Amino acid sequence of the phosphopyridoxylated peptide, which was
protected by fructose-1,6-P2, of E. coli ADP-glucose pyrophosphorylase and the alignment
with other known sequences. The labeled lysine residue, Lys”, is in bold. The numbers
indicated correspond to the sequence of the E. coli enzyme. Nomenclature is the same as
in Figure 1. .

sl—54
sh-2
we7
pot-l
ana
syn
rice
pot-s
sl-Sl
arabid
bt-Z
ecoli
s.t.

34

33

I

LFPL ...... PAV
LFPLTSTRATPAV
LFPLTSTRATPAV
LFPLTSRTATPAV
LYPLTKLRAKPAV
LYPLTKLRAKPAV
LYPLTKKRAKPAV
LYPLTKKRAKPAV
LYPLTKKRAKPAV
LYPLTKKRAKPAV
LYPLTKKRAKPAV
LKDLTNKRAKPAV
LKDLANKRAKPAV

Figure 4

35

ADP—glucose pyrophosphorylase sequences known to date. Substitution of Lys39 with
glutamic acid formed an enzyme with 23-fold lower apparent afﬁnity for the activator,
FBP (Gardiol & Preiss, 1990). Therefore, Lys39 is believed to be involved in the binding
of FBP.

Similar studies using PLP as the afﬁnity label agent was conducted for the spinach
leaf ADP-glucose pyrophosphorylase (Morell et al., 1988; Preiss et al., 1992; Ball &
Preiss, 1994). When PLP is covalently bound, the enzyme no longer requires 3PGA for
activation. It was shown that one lysine residue of the small subunit (Morell et al. , 1988)
and three lysine residues of the large subunit (Preiss et al., 1992; Ball & Preiss, 1994) are
involved in the binding of PLP (Figure 5). The modiﬁcation of these lysine residues by
PLP is prevented by the presence of 3PGA, while only the lysine residue of site 1 of the
small subunit (Lys‘w) and site 2 of the large subunit are prevented by Pi from being
modiﬁed. These observations indicate that the activator analog, PLP, binds at the activator
site and the most important sites involved in the binding of activator are site 1 and site 2.

Recent chemical modiﬁcation and site-directed mutagenesis studies on the
cyanobacterial ADP-glucose pyrophosphorylase from Anabaena have shown that Lys‘”,
corresponding to Lys440 of the small subunit of spinach leaf enzyme (Figure 5), is involved
in the binding of the activator, 3PGA (Charng et al. , 1994). Chemical modiﬁcation with
PLP resulted in an enzyme that no longer required activator for high activity and the
modiﬁcation was prevented by 3PGA and Pi. Lys“19 was found to be labeled by PLP.

419

Replacing Lys with either arginine, alanine, glutamine, or glutamic acid produced

36

Figure 5: Amino acid sequence of the phosphopyridoxylated peptides of spinach
leaf and Anabaena ADP-glucose pyrophosphorylases and the alignment with other known
sequences of higher-plant and cyanobacterial enzymes. The labeled lysine residues are in
bold. The number indicated corresponds to the sequence of the small subunit of the
spinach enzyme. Nomenclature is the same as in Figure 1.

81-54
wl
sh-2
we7
we3
pot—l
ana
syn
rice
pot—s
sl-Sl
bt-2

Site 3

KWFQGTADAVRQ
GWFQGTADSIRK
GWFRGTADAWRK
KWFQGTADAVRK
NWFQGTADAVRQ
DWFQGTADAVRQ
NWFQGTADAVRQ
DWFQGTADAVRQ
DWFQGTADAVRQ
NWFQGTADAVRQ

37

Site 2

IKDAIIDKNAR
IQNCIIDKNAR
IRNCIIDMNAR
ISNCIIDMNAR
ISNCIIDMNAR
IRKCIIDKNAK
IRRAIIDKNAR
IRRAIIDKNAR
IRRAIIDKNAR
IKRAIIDKNAR
IKRAIIDKNAR
IRRAIIDKNAR

Figure 5

Site 1

440
I

SGITVIFKNATIKDGVV

SGITVVLKNSVIADGLVI
SGIVVILKNATINECLVI
SGIVVIQKNATIKDGTVV
SGIVVIQKNATIKDGTVV
SGIIIILEKATIRDGTVI
SGIVVVLKNAVITDGTII
NGIVVVIKNVTIADGTVI
SGIVTVIKDALLLAEQLY
SGIVTVIKDALIPSGIII
SGIVTVIKDALIPSGTVI
GGIVTVIKDALLPSGTVI

38

mutant enzymes with apparent afﬁnities for 3PGA 25-, 50-, 140-, or 150-fold lower than
that of the wild-type enzyme, respectively. These mutant enzymes still could be activated
at higher concentrations of 3PGA, which suggested that other residue(s) might be involved
in the binding of 3PGA. Chemical modiﬁcation studies of the K419R enzyme have shown
that PLP can also be incorporated into this mutant enzyme. The phosphopyridoxylated
K419R enzyme is less dependent on the presence of 3PGA for activation and less sensitive
to Pi inhibition. Lys382 of the mutant enzyme was identiﬁed as the labeled residue by PLP
and corresponds to the PLP-labeled lysine residue of site. 2 of the spinach leaf large
subunit (Figure 5). Kinetic and protection studies suggested that Lys382 might be also
involved in binding of the allosteric activator (Charng et al., 1994).

The photoafﬁnity analog 8-azido-AMP was used as a site-speciﬁc probe for the
inhibitor-binding sites of the E. coli ADP-glucose pyrophosphorylase (Larsen et al. , 1986).
In the absence of light, the enzyme was inhibited by 8-azido—AMP and the inhibition is
similar to that observed with the natural inhibitor, AMP. Tyr‘“, which is located within
the major incorporation site of 8-azido-ADP-glucose, was identiﬁed as the amino acid
modiﬁed by 8-azido-AMP. This suggests that the binding site for AMP might be
overlapping with those for the substrates, ATP and ADP-glucose.

The arginine-speciﬁc reagent phenylglyoxal has been used to covalently modify
ADP-glucose pyrophosphorylases from E. coli (Carlson & Preiss, 1982), Synechocystis
(Iglesias et al., 1992), and spinach leaf (Ball & Preiss, 1992). The results suggested that
there was an essential arginine residue(s) involved in the allosteric regulation. However,

the arginine residue(s) labeled by phenylglyoxal has not been identiﬁed.

39

Mutant E. coli stains with altered glycogen accumulation have been produced by
chemical mutagenesis (Govons et al., 1969; Preiss et al., 1976; Creuzat-Sigal et al.,
1972). ADP-glucose pyrophosphorylases from E. coli B mutant strain CL1136 (Preiss et
al., 1969) and SGS (Govons et al., 1969, 1973) and E. coli K12 mutant strain 618
(Cattaneo et al. , 1969; Creuzat-Sigal et al. , 1972) have higher activity in the absence of
the activator, FBP, and are less sensitive to the inhibition by AMP than the wild-type
enzyme. It was found that Arg67 of mutant CL1136 (Ghosh et al. , 1992), Prom of mutant
sos (Meyer et al., 1992), and my336 of mutant 618 (Kumar et al., 1939) were changed
to cysteine, serine, and aspartate, respectively. The enzyme from E. coli B mutant strain
S614, in which Ala44 was replaced with threonine, has lower afﬁnities for the substrates,
the activator, FBP, and the inhibitor, AMP (Meyer et al., 1993). These residues might

also play important roles in the regulation of ADP-glucose pyrophosphorylase.

7. Prediction of the Secondary Structure of ADP-Glucose Pyrophosphorylase:

The structures of ADP-glucose pyrophosphorylases from different sources have
been predicted in the basis of their amino acid sequences by Ballicora et al. (1996) using
hydrophobic cluster analysis (Lemesle—Varloot et al. , 1990) and PHD program (Rost &
Sander, 1993). Hydrophobic analysis shows that ADP-glucose pyrophosphorylases are
identical in the position of many clusters and in some others the differences are small. The
small insertions seen among ADP-glucose pyrophosphorylases are not part of the "core"
of the protein because there are no buried amino acids in these insertions, indicating that

ADP-glucose pyrophosphorylases from different sources share a common folding pattern

40

though they have different quaternary structures (homotetrameric in bacteria and
heterotetrameric in higher plants) and different speciﬁcities for the regulators. A general
structure (Figure 6) that ﬁts all of these protein was predicted using the PHD program
(Rost & Sander, 1993). ADP-glucose pyrophosphorylase is an alpha/beta protein with
some parts of it mainly beta, for example, the C-terminus and the domain denoted as 3
(Figure 6).

Most of the conserved amino acids known to have roles in the binding of substrates

(T yr114 and Lys‘” in E. call) and activators (Lys39 in E. coli and Lys419

in Anabaena) are
located in 100ps, or very close to loops. The residues Prom and Gly336 that are important
for the regulation of the E. coli enzyme are also in loops. A supersecondary structure seen
in nucleotide-binding proteins (Rossmann et al. , 1974) is also present in this model. The
glycine loop in the domain 1, which is thought to bind the phosphates of the ATP, and the
region 2, with three beta sheets and alpha helices compatible with a Rossmann fold. The
predicted secondary structure in region 1 and 2 is identical to the structure of the
oncogenic protein H-Ras (p21) which is used as the folding model for nucleotide phosphate
binding of GTP (Tong et al., 1991).

The results from trypsin treatment of the E. coli and Anabaéna ADP-glucose
pyrophosphorylases (Charng & Preiss, unpublished data) are in good agreement with the
predicted model. Exposed loops would be more sensitive to proteolysis and the proteolysis
studies show that trypsin cuts at sites predicted as loops by the model. Recent proteolysis

studies of ADP-glucose pyrophosphorylases from E. coli (Wu et al. , unpublished data) and

Anabaena (Sheng et al. , unpublished data) with proteinase K also conﬁrmed that the

41

Figure 6: Proﬁle Neural Network (PHD) prediction of the secondary structure of
ADP-glucose pyrophosphorylase (Ballicora et al., 1996). Section 1 contains the FBP
activator site KRAKPAV (E. call) in a loop as well as Arg‘”. Section 2 has the putative
ATP binding site, Tyr‘“ in a loop area between a beta-strand and alpha-helix starting at
GTAD. The glucose-l-P binding site is also seen in a loop among a series of predicted
beta-strands. The topology between regions 1 and 2 cannot be ascertained and is drawn
as a dotted line.

 

 

GlPSite V

 

 

 

 

 

 

mzu ,.

O
..-...OOOOOIOOOOOOI......I.---O-"

 

.00....

 

......COOCCCCOU

N—T
(V'

 

 

 

42

 

 

 

 

 

 

 

 

Anahaennndplm

Helmmaloop

 

   
 

Figure 6

43
predicted model is valid, since all the cleavage sites are in the predicted loops. The
cleavage sites close to the putative glucose-l-P binding site (Lys‘95 in E. coll) are protected
by ADP-glucose, Mg“, and the activator. Interestingly, the E. coli enzyme treated with
proteinase K in the presence of ADP-glucose, Mg“, and FBP, with 11 amino acids from
the N -terminus and 2 amino acids from the C-terminus cleaved, retains maximal activity
but it is no longer sensitive to AMP inhibition and does not require FBP for activation
(Wu et al. , unpublished data). It was postulated that the N-terminal loop is involved in the

allosteric regulation of the E. coli enzyme.

REFERENCES

Aitken, A. (1988) Methods Enzymol. 167, 145-154.

Anderson, J. M., Hnilo, J., Larson, R., Okita, T. W., Morell, M., & Preiss, J. (1989)
J. Biol. Chem. 264, 12238-12242.

Anderson, J. M., Okita, T. W., & Preiss, J. (1990) in The Molecular Biology of the
Potato (Vayda, M. E., & Park, W. D., eds) pp. 159-180, International, Wallingford.

ap Rees, T. (1984) in Storage Carbohydrates in Vascular Plants (Lewis, D. H., ed.) pp
53-73, Cambridge University Press, Cambridge.

Baecker, P. A., Furlong, C. E., & Preiss, J. (1983) J. Biol. Chem. 258, 5084-5088.
Ball, K., & Preiss, J. (1992) J. Protein Chem. 11, 231-238.
Ball, K., & Preiss, J. (1994) J. Biol. Chem. 269, 24706-24711.

Ball, 8., Marianne, T., Dirik, L., Fresnoy, M., Delrue, B., & Decq, A. (1991) Planta
185, 17-26.

Ballicora, M. A., Laughlin, M. J., Fu, Y., Okita, T. W., Barry, G. F., & Preiss, J.
(1995) Plant Physiol. 109, 245-251.

Ballicora, M. A., Fu, Y., Wu, M-X., Sheng, J ., Nesbitt, & Preiss, J. (1996) inAbstracts
of the 5th NIAR/COE International Symposium (Nakamura, Y. ed.) pp 5-11, National
Institute of Agrobiological Resources, Tsukuba, Japan.

Barton, C., Yang, L., Galvin, M., Sengupa-Gopalan, C., & Borelli, T. (1986) in
Regulation of Carbon and Nitrogen Reduction and Utilization in Maize (Shannon, J. C. ,
Kinevel, D. P., & Boyer, C. D., eds.) pp 363-365, American Society of Plant
Physiologist, Rockville, MD.

Brana, A. F., Manzanal, M. B., Hardisson, C. (1980) J. Bacterial. 144, 1139-1142.
Carlson, C.A., & Preiss, J. (1982) Biochemistry 21, 1929-1934.

Cattaneo, J ., Damotte, M., Sigal, N., Sanchez-Medina, G., & Puig, J. (1969) Biochem.
Biophys. Res. Commun. 34, 694-701.

Charng, Y., Kakefuda, G., Iglesias, A. A., Buikema, W. J ., & Preiss, J. (1992) Plant
Mol. Biol. 20, 37-47.

44

45

Charng, Y., Iglesias, A. A., & Preiss, J. (1994) J. Biol. Chem. 269, 24107-24113.
Creuzat-Sigal, N., Latil-Damotte, M., Cattaneo, J ., & Puig, J. (1972) in Biochemistry of
the Glycosidic Linkage (Piras, R., & Pontis, H. G., eds.) pp 647-680, Academic Press,
New York.

Dickinson, D. B., & Preiss, J. (1969) Plant Physiol. 44, 1058-1062.

Edwards, J ., Green, J. H., & ap Rees, T. (1988) Phytochemistry 27, 1615-1620.
Espada, J. (1962) J. Biol. Chem. 237, 3577-3581.

Furlong, C. E., Preiss, J. (1969) J. Biol. Chem. 244, 2539-2548.

Gardiol, A., & Preiss, J. (1990) Arch. Biochem. Biophys. 280, 175-180.

Ghosh, H. P., & Preiss, J. (1965) J. Biol. Chem. 240, 960-961.

Ghosh, P., Meyer, C. R., Remy, E., Peterson, D., & Preiss, J. (1992) Arch. Biochem.
Biophys. 296, 122-128.

Govons, S., Vinopal, R., Ingraham, J ., & Preiss, J. (1969) J. Bacteriol. 97, 970-972.

Govons, S., Gentner, N., Greenberg, E., & Preiss, J. (1973) J. Biol. Chem. 248, 1731-
1740.

Greenberg, E., & Preiss, J. (1964) J. Biol. Chem. 239, 4314-4315.
Haugen, T. H., Ishaque, A., & Preiss, J. (1976) J. Biol. Chem. 251, 7880-7885.
Haugen, T. H., & Preiss, J. (1979) J. Biol. Chem. 254, 127-136.

Heldt, H. W., Chon, C. J., Maronde, D., Herold, A., Stankovic, Z. S., Walker, D. A.,
Kraminer, A., Kirk, M. R., & Herber, U. (1977) Plant Physiol. 59, 1146-1155.

Hill, M. A., Kaufmann, K., Okita, J., & Preiss, J. (1991) J. Biol. Chem. 266, 12455-
12460.

Iglesias, A. A., Kakefuda, G., & Preiss, J. (1991) Plant Physiol. 97, 1187-1195.
Iglesias, A. A., Kakefuda, G., & Preiss, J. (1992) J. Protein Chem. 11, 119-128.

Iglesias, A. A., Charng, Y., Ball, 8. & Preiss, J. (1994) Plant Physiol. 104, 1287-1294.

46
Kacser, H., & Burns, J. A. (1973) symposium Soc. Exp. Biol. 27, 65-107.

Kacser, H. (1987) in The Biochemistry of Plants (Davies, D. D., ed.) Vol 11, pp 39-67,
Academic Press, New York.

Kakefuda, G., Charng, Y., Iglesias, A. A., Mcintosh, L., & Preiss, J. (1992) Plant
Physiol. 99, 359-361.

Kleczkowski, L. A., Villand, P., Preiss, J., & Olsen, 0. -A. (1993) J. Biol. Chem. 268,
6228-6233.

Krebs, E. G., & Preiss, J. (1975) in International Review of Science, Biochemistry of
Carbohydrates (Whelan, W. J ., ed.) pp 337-389, University Park Press, Baltimore.

Kruger, N. J. (1990) in Plant Physiology, Biochemistry and. Molecular Biology (Dennis,
D. T., & Terpin, D. H., eds.) pp 59-76, John Wiley, New York.

Kumar, A., Tanaka, T., Lee, Y. M., & Preiss, J. (1988) J. Biol. Chem. 263, 14634-
14639.

Kumar, A., Ghosh, R, Lee, Y. M., Hill, M. A., & Preiss, J. (1989) J. Biol. Chem. 264,
10464-10471.

Larsen, C. E., Lee, Y. M., & Preiss, J. (1986) J. Biol. Chem. 261, 15402-15409.
Lee, Y. M., Mukherjee, S., & Preiss, J. (1986) Arch. Biochem. Biophys. 244, 585-595.
Lee, Y. M., & Preiss, J. (1986) J. Biol. Chem. 261, 1058-1064.

Lehmann, M., & Preiss, J. (1980) J. Bacteriol. 143, 120-127 .

Leloir, L. F., Olivarria, J. M., Goldemberg, S. H., & Carminatti, H. (1959) Arch.
Biochem. Biophys. 81 , 508.

Lemesle-Varloot, L., Henrissat, B., Gaboriaud, C., Bissery, V., Morgat, A., & Momon,
J. P. (1990) Biochimie. 72, 555-574.

Leung, P. S. C., & Preiss, J. (1987a) J. Bacteriol. 129, 4349-4354.
Leung, P. S. C., & Preiss, J. (1987b) J. Bacteriol. 129, 4355-4360.
Levi, C., & Preiss, J. (1976) Plant Physiol. 58, 753-756.

Lin, T. -P., Caspar, T., Somerville, C., & Preiss, J. (1988a) Plant Physiol. 86, 1131-

47
1135.

Lin, T. -P., Caspar, T., Somerville, C., & Preiss, J. (1988b) Plant Physiol. 88, 1175-
1181.

Mackey, B. M., Morris, J. G. (1971) J. Gen. Microbial. 66, 1-13.

Manners, D. J. (1985) in Biochemistry of Storage Carbohydrates in Green Plants (Dey, '
P. M., & Dixon, R. A., eds.) pp 149-203, Academic Press, New York.

Mereschowsky, C. (1905) Biol. Zentralbl. 25, 593.
Meyer, C. R., Ghosh, P., Remy, E., & Preiss, J. (1992) J. Bacterial. 174, 4509-4512.

Meyer, C. R., Ghosh, P., Nadler, S., & Preiss, J. (1993) Arch. Biochem. Biophys. 302,
64-71.

Morell, M., Bloom, M., Knowles, V., & Preiss, J. (1987) Plant Physiol. 85, 182-187.
Morell, M., Bloom, M., & Preiss, J. (1988) J. Biol. Chem. 263, 633-637.

Morrison, W. R., & Karkalas, J. (1990) in Methods in Plant Biochemistry (Dey, P. M.,
ed.) Vol 2, pp 323-353, Academic Press, New York.

Miiller-Rtiber, B. T., KoBmann, J ., Hannah, L. C., Wilmitzer, L., & Sonnewald, U.
(1990) Mol. Gen. Genet. 224, 136-146.

Muller-Robot, B. T., Sonnewald, U., & Wilmitzer, L. (1992) EMBO J. 11, 1229-1238.

Nakata, P. A., Greene, T. W., Anderson, J. M., Smith-White, B. J., Okita, T. W., &
Preiss, J. (1991) Plant Mol. Biol. 17, 1089-1093.

Neuhaus, H. E., Kruckeberg, A. L., Feil, R., & Stitt, M. (1989) Planta 178, 110-122.
Neuhaus, H. E., & Stitt M. (1990) Planta 182, 445-454.

Okita, T. W., Greenberg, E., Kuhn, D. N., & Preiss, J. (1979) Plant Physiol. 64, 187-
192.

Okita, T. W., Rodriguez, R. L., & Preiss, J. (1981) J. Biol. Chem. 256, 6944-6952.

Okita, T. W., Nakata, P. A., Anderson, J. M., Sowokinos, J., Morell, M., & Preiss, J.
(1990) Plant Physiol. 93, 785-790.

48
Olive, M. R., Ellis, R. J., & Schuch, W. W. (1989) Plant Mol. Biol. 12, 525-538.

Ozbun, J. L., Hawker, J. S., Greenberg, E., Lammel, C., & Preiss, J. (1973) Plant
Physiol. 51, 1-5.

Parsons, T., & Preiss, J. (1978a) J. Biol. Chem. 253, 6197-6202.

Parsons, T., & Preiss, J. (1978b) J. Biol. Chem. 253, 7638-7645.

Paule, M.R., and Preiss, J. (1971) J. Biol. Chem. 246, 4602-4609.
Pettersson, G., & Ryde-Pettersson, U. (1989) Eur. J. Biochem. 179, 169-172.
Preiss, J. (1969) Curr. Topics Cell Reg. 1, 125-161.

Preiss, J ., Sabraw, A., & Greenberg, E. (1971) Biochem. Biophys. Res. Commun. 42,
180-196.

Preiss, J. (1973) The Enzymes 8, 73-119.
Preiss, J ., Greenberg, E., & Sabraw, A. (1975) J. Biol. Chem. 250, 7631-7638.
Preiss, J ., Lammel, C., & Greenberg, E. (1976) Arch. Biochem. Biophys. 174, 105-119.

Preiss, J. (1978) in Advances in Enzymalagy and Related Areas of Molecular Biology
(Meister, A., ed.) Vol. 46, pp 317-381, Wiley-Interscience, New York.

Preiss, J., & Levi, C. (1980) in Biochemistry of Plant Carbohydrates (Preiss, J ., ed.) Vol
3, pp 371-423, Academic Press, New York.

Preiss, J ., & Walsh, D. A. (1981) in Biology of Carbohydrates (Ginsburg, V., ed.)
Chapter 5, pp 199-313, J. Wiley and Sons, New York.

Preiss, J. (1982) Annu. Rev. Plant Physiol. 54, 431-454.
Preiss, J. (1984) Annu. Rev. Microbial. 38, 419-458.

Preiss, J., Bloom, M., Morell, M., Knowles, V. L., Plaxon, W. C., Okita, T. W.,
Larson, R., Harmon, A. C., & Putnam-Evans, C. (1987a) in Tailoring Genes for Crop
Improvement (Bruening, G., Harada, J., Kosuge, T., & Hollaender, A., eds.) pp 132-152,
Plenum, New York.

Preiss, J ., Morell, M., Bloom, M., Knowles, V. L., & Lin, T. P. (1987b) in Proceeding
of the V11 International Congress on Photosynthesis (Biggins, J ., ed.) Vol. 3, pp 693-700,

49

Martinus Nijhoff, Dordrecht, the Netherlands.

Peiss, J. (1988) in Biochemistry of Plants (Preiss, J., ed.) Vol XIV, pp 181-254,
Academic Press, New York.

Preiss, J ., & Romeo, T. (1989) Adv. Microb. Physiol. 30, 183-238.

Preiss, J., Danner, S., Summers, P. S., Morell, M., Barton, C. R., Yang, L., & Nieder,
M. (1990) Plant Physiol. 92, 881-885.

Preiss, J. (1991) in Plant Molecularand Cell Biology (Mifflin, B., ed.) Vol. 7, pp 59-114,
Oxford University, Oxford.

Preiss, J ., Ball, K., Charng, Y., & Iglesias, A. A. (1992) in Research in Photosynthesis
(Murata, N., ed.) Vol. 3, pp 697-700, Kluwer Academic Publishers, Dordrecht,
Netherlands.

Preiss, J. (1996) Biotech. Annu. Rev. 2, 259-279.

Preiss, J, & Sivak, M. (1996) in Photaassimilate Distribution in Plants and Crops
(Zamski, E., & Schaffer, A. A., eds.) pp 63-96, Marcel Dekker, Inc., New York.

Rossmann, M. G., Moras, D., & Olsen, K. W. (1974) Nature 250, 194-199.
Rost & Sander (1993) Proc. Natl. Acad. Sci. 90, 7558-7562.
Sanwal, G. G., & Preiss, J. (1967) Arch. Biochem. Biophys. 119, 454-469.

Sanwal, G. G., Greenberg, E., Hardie, J., Cameron, E. C., & Preiss, J. (1968) Plant
Physiol. 43, 417-427.

Shen, L., & Preiss, J. (1964) Biochem. Biophys. Res. Commun. 17, 424-429.
Shively, J. M. (1988) Methods Enzymol. 167, 195-203.

Slock, J. A., & Stahly, D. P. (1974) J. Bacterial. 120, 399-406.

Smith, A. M., Bettey, M., & Bedford, I. D. (1989) Plant Physiol. 89, 1279-1284.
Smith-White, B. J ., & Preiss, J. (1992) J. Mol. Eval. 34, 449-464.

Stalmans, W. & Hers, H. G. (1973) in The Enzyme (Boyer, P. D., ed.) 3rd ed, vol. 8, pp
309-361. Academic Press, New York.

50

Stark, D. M., Timmerman, K. P., Barry, G. F., Preiss, J., & Kishore, G. M. (1992)
Science 258, 287-292.

Steiner, K. E., & Preiss, J. (1977) J. Bacterial. 129, 246-253.

Strange, R. E., Dark, F. A., & Ness, A. G. (1961) J. Gen. Microbial. 25, 61-76.
Strange, R. E. (1968) Nature 220, 606-607.

Strasdine, G. A. (1968) Can. J. Microbiol. 14, 1059-1062.

Strasdine, G. A. (1972) Can. J. Microbiol. 18, 211-217.

Takahashi, K. (1968) J. Biol. Chem. 243, 6171-6179.

Tong, L. A., de Vos, A. M., Milburn, M. V., Kim, S. H. (1991) J. Mol. Biol., 217, 503-
516.

Tsai, C. Y., & Nelson, 0. E. (1966) Science 151, 341-343.
Van Houte, J ., & Jansen, H. M. (1970) J. Bacterial. 101, 1083-1085.

Yung, S., & Preiss, J. (1982) J. Bacterial. 151, 742-749.

CHAPTER II
SITE-DIRECTED MUTAGENESIS OF LYSINEm, THE ACTIVATOR-
BINDING SITE, OF ADP-GLUCOSE PYROPHOSPHORYLASE FROM

ANABAENA PCC 7120

51

ABSTRACT

Previous studies have shown that a highly conserved lysyl residue (Lysm) near the
C-terminus of Anabaena ADP-glucose pyrophosphorylase is involved in the binding of
3PGA, the allosteric activator (Charng et al. , 1994). Phosphopyridoxylation of the K419R
mutant enzyme modiﬁed another conserved lysyl residue (Lysm), suggesting that this
residue might also be located within the activator-binding site (Charng et al. , 1994). Site-
directed mutagenesis of Lys382 of the Anabaena enzyme was performed to determine the
role of this residue. Replacing Lys382 with either arginine, alanine, or glutamine produced
mutant enzymes that were still activated by 3PGA but with apparent afﬁnities 10— to 160-
fold lower than that of the wild-type enzyme. In contrast, the glutamic acid mutant enzyme
was inhibited by 3PGA. These mutations had lesser impact on the kinetic constants for the
substrates and inhibitor, Pi, and on the thermal stability. These results indicate that both
the charge and size of the residue at position 382 inﬂuence the binding of 3PGA. Site-
directed mutagenesis was also performed to obtain a K382R-K419R double mutant. The
apparent afﬁnity for 3PGA of this double-mutant enzyme was 104-fold lower than that of
the wild-type enzyme, and the speciﬁcity for activator of this mutant enzyme was altered.
The K382R-K419R enzyme could not be phosphopyridoxylated, suggesting that other

lysine residues are not involved in the binding of 3PGA.

52

INTRODUCTION

ADP-glucose pyrophosphorylase catalyzes the conversion of glucose-l-P and ATP
to ADP-glucose and PP,. This allosterically regulated enzyme catalyzes the ﬁrst committed
step in the biosynthesis of glycogen in bacteria (Preiss & Romeo, 1989; Preiss, 1996) and
starch in plants (Preiss, 1991, Preiss, 1996; Preiss & Sivak, 1996). ADP-glucose
pyrophosphorylase from higher plants is mainly activated by 3PGA and inhibited by P,
(Ghosh & Preiss, 1965; Sanwal et al., 1968), whereas the enzyme from enteric bacteria
is activated by FBP and inhibited by AMP (Preiss et al., 1966). The bacterial enzyme is
homotetrameric in structure (Haugen et al., 1976), while. the higher-plant enzyme is
heterotetrameric with two different subunits (Preiss, 1991).

Cyanobacteria are prokaryotic organisms that have metabolic properties similar to
chloroplasts of higher plants (Aitken, 1988). However, the former organisms synthesize
glycogen as the major carbohydrate reserve (Shively, 1988), in a similar manner to what
is observed in bacteria. The cyanobacterial ADP-glucose pyrophosphorylase is
homotetrameric similar to the E. coli enzyme, but is regulated by 3PGA and P, like the
' higher-plant enzyme (Levi & Preiss, 1976; Iglesias et al., 1991). Analysis of the deduced
amino acid sequence indicated that the cyanobacterial enzyme is more similar to the
higher-plant enzyme than to the enteric bacterial enzyme in primary structure (Smith-
White & Preiss, 1992). It is of great interest to understand the structure—function
relationships of the allosteric site of the cyanobacterial ADP-glucose pyrophosphorylase
due to the properties of the cyanobacterial enzyme and the key position of these

photosynthetic prokaryotes during evolution (Aitken, 1988).

53

54

Previous studies have shown that the activator-binding site of the E. coli enzyme,
Lys”, is near the N-terminus (Parsons & Preiss, 1978; Gardiol & Preiss, 1990), while that
of the spinach leaf enzyme, Lys”, is situated toward the C-terminus (Morell et al., 1988;
Ball & Preiss, 1994). Chemical modiﬁcation and site-directed mutagenesis studies on the
cyanobacterial ADP-glucose pyrophosphorylase from Anabaena have shown that Lys”,
corresponding to Lys“10 in spinach leaf enzyme, is involved in binding of the activator,
3PGA (Charng et al., 1994). Lys‘”9 was labeled by PLP, an analogue of the activator,
3PGA. Replacing Lys419 with either arginine, alanine, glutamine, or glutamic acid
produced mutant enzymes with apparent affmities for 3PGA 25-, 50—, 140-, or lSO-fold
lower than that of wild-type enzyme, respectively. These mutant enzymes still could be
activated by 3PGA, which suggested that other residue(s) might be involved in the binding
of 3PGA. Chemical modiﬁcation studies of the K419R enzyme have shown that PLP can
also be incorporated into this mutant enzyme. The phosphopyridoxylated K419R enzyme
is less dependent on the presence of 3PGA for activation and less sensitive to P, inhibition.
Lys332 of the mutant enzyme was found to be labeled by PLP. Kinetic and protection
studies suggested that Lys382 was also involved in binding of the allosteric activator
(Charng et al., 1994).

Here, the results of further site-directed mutagenesis and chemical modiﬁcation of
the cyanobacterial ADP-glucose pyrophosphorylase from Anabaena are reported. Lys382
was replaced with either arginine, alanine, glutamine, or glutamic acid with site-directed
mutagenesis to determine the role of this residue. A double-mutant enzyme, K382R-

K419R, was also generated and characterized.

MATERIALS AND METHODS

Reagents

[”P]PP, was purchased from Du Pant-New England Nuclear. [“C]Glucose—1-P was
from ICN Pharmaceuticals Inc. [a-35S]dATP and the in vitro mutagenesis kit were from
Amersham Corp. [4-3H]PLP was synthesized and puriﬁed as described previously (Morell
et al. , 1988) according to the method of Stock et al. (1966) and Ahrens and Kortnyk
(1969). Oligonucleotides were synthesized and puriﬁed by the Macromolecular Facility
at Michigan State University. All other reagents were purchased at the highest quality
available. .
Bacterial Strains and Media

E. coli strain TGl [K12, A(lac-pro), supE, thi, hsdD5/F'traD36, proA+B+, lacl“,
lacZAMlS] was used for site-directed mutagenesis and grown in LB medium. E. coli
mutant strain AC70R1-504, which is deﬁcient in ADP-glucose pyrophosphorylase activity
(Carlson et al., 1976), was used for expression of the Anabaena ADP-glucose
pyrophosphorylase gene (Charng et al. , 1992) and grown in enriched medium containing
1.1% KzHPO4, 0.85% KHZPO4, 0.6% yeast extract, and 0.2% glucose, pH 7.0.
Site-Directed Mutagenesis

Plasmid pAnaE3a was used for both site-directed mutagenesis and gene expression.
In pAnaE3a, a 5.5 kb EcoRI fragment of Anabaena genomic DNA containing the
Anabaena ADP-glucose pyrophosphorylase gene and its putative promoter was ligated
onto the EcoRI site of pUC119 plasmid (Charng et al., 1992). The orientation of the gene

is opposite to the lac promoter of pUC119 and enables TGl cells to synthesize

55

56
single-stranded DNA containing the antisense strand of the gene using helper phage
M13KO7 (Sambrook et al., 1982). Site-directed mutagenesis was performed according to
the method of Sayers et al. (1988) using the in vitro mutagenesis kit from Amersham
Corp. The Lys382 mutant enzymes with substitution of arginine, alanine, glutamine, and
glutamate acid were designated as K382R, K382A, K382Q, and K382E, respectively. The

419

double mutant enzyme, in which both Lys382 and Lys were replaced with arginine, was
obtained by performing site-directed mutagenesis on the K419R mutant gene (Charng et
al., 1994) and designated as K382R-K419R. The oligonucleotides used to create the
mutants are shown in Figure 1. The plasmids recovered in the last step of the mutagenesis
were screened by dideoxy sequencing (Sanger etal., 1977) in the regions of the desired
mutations. Prior to expression of the mutant enzymes, the entire coding regions of these
mutant alleles were sequenced to verify that there were no unintended mutations.
Expression and Puriﬁcation of the Wild-Type and Mutant Enzymes

The wild-type and mutant genes were expressed in AC70R1-504 cells. Cell extracts
were prepared by sonication. The enzymes were puriﬁed by ion-exchange chromatography
on DEAE-Sepharose, FPLC chromatography on Mono Q, and Phenyl-Superose columns
as described previously (Iglesias et al., 1991; Charng et al., 1992).
Enzyme Assay

(A) Assay 1. Enzymatic activity was measured in the pyrophosphorylase direction

at 37 °C according to Preiss et al. (1966) during enzyme puriﬁcation. The reaction

mixtures contained 80 mM Hepes—NaOH buffer (pH 7.0), 2 mM ADP-glucose, 8 mM

57

Figure 1: Nucleotide sequence and encoded protein sequence of the Anabaena
ADP-glucose pyrophosphorylase gene in the region of Lys382 (upper) and the synthetic
oligonucleotides used for site-directed mutagenesis at position 382 (lower). The position
382 codons are underlined.

Arg-382
Ala-382
Gln—382

Glu-382

5'- ATC

3'- TAG

5'- ATC

5'- ATC

5'- ATC

5'- ATC

ATC

TAG

ATC

ATC

ATC

ATC

58

GAT

CTA

GAT

GAT

GAT

GAT

Figure 1

HEW:

O
C)
r-l

C)
O
'-1

O
>
C)

O
>
>

GCC

CGG

GCC

GCC

GCC

GCC

CGC -3'

GCG -5'

CGC -3‘

CGC -3'

CGC -3'

CGC -3'

59
MgClz, 2 mM [32P]PP, (about 1500—3000 cpm/nmol), 4 mM 3PGA, 4 mM NaF, 50 ug of
bovine serum albumin, and enzyme in a total volume of 250 uL. For the K382Q and
K382R-K419R enzymes, 16 mM 3PGA was included in the reaction.

(B) Assay II. Enzymatic activity in the ADP-glucose synthesis direction at 37 °C
was measured according to the method of Preiss et al. (1966) to determine kinetic
constants. For assay of the wild-type enzyme in the presence of activator, reaction
mixtures contained 100 mM Hepes—NaOH buffer (pH 8.0), 0.5 mM [”C]glucose-1-P
(about 1000 cpm/nmol), 2.5 mM ATP, 10 mM MgC12, 2.5 'mM 3PGA, 50 ng of bovine
serum albumin, 0.15 unit of inorganic perphosphatase, and enzyme in a ﬁnal volume of
200 ptL. For assay of the mutant enzymes, the reaction conditions were identical to wild-
type, except that the amount of 3PGA was altered to obtain maximal activity. For the
K382R, the K382A, the K382Q, and the K382R-K419R enzymes, 5, 20, 30, and 20 mM
3PGA were used, respectively. The assay of the wild-type enzyme in the absence of
activator was the same as described above except that 3PGA was omitted and the amount
of ATP was 5 mM in the reaction mixture. The amounts of ATP and [“C]glucose-l-P were
altered for some mutant enzymes to obtain maximal activity. The amount of ATP was
increased to 8 mM for K382Q and K382E enzymes. For the K382R, the K382A, and the
K382R-K419R enzymes, 3 mM [“C]glucose-l-P was used, while 2 mM was used for
K382E.

Kinetic Characterization
Kinetic data were plotted as initial velocity versus substrate or effector

concentration. Saturating concentrations of substrates and effectors were determined to

60

ensure that maximal velocity was attained. Data were replotted as double-reciprocal plots
to determine Vmax. Kinetic constants from hyperbolic plots were also determined by
double-reciprocal plots. Sigmoidal plots were replotted as Hill plots to obtain kinetic
constants. Interaction coefﬁcients, n,,, were also determined by Hill plots. Kinetic
constants were expressed as A0,, 80.5, and 10.5, which correspond to the concentration of
activator, substrate, or inhibitor giving 50% of maximal activation, velocity, and
inhibition, respectively. All the kinetic parameters calculated from double-reciprocal or
Hill plots were in good agreement with those obtained by using a computer program
(Canellas & Wedding, 1980) which performed nonlinear iterative least-squares ﬁtting to
the Hill equation.
Protein Assay

Protein concentration was determined by using bicinchoninic acid reagent (Smith
et al., 1985) with bovine serum albumin as the standard.
Protein Electrophoresis and Immunoblotting

Polyacrylamide gel electrophoresis with SDS (SDS—PAGE) was performed
according to Laemmli (1970). Following electrophoresis, proteins on the gel were
visualized by staining with Coomassie Brilliant Blue R-250 or electroblotted onto a
nitrocellulose membrane according to Bumette (1981). After electroblotting, nitrocellulose
membranes were treated with afﬁnity-puriﬁed anti-spinach leaf ADP-glucose
pyrophosphorylase IgG (Morell et al., 1987), and the antigen—antibody complex was
visualized via treatment with alkaline phosphatase linked goat anti-rabbit IgG followed by

staining with BM purple AP-substrate precipitating reagent (from Boehringer Mannheim

61

GmbH).
Molecular Mass Determination

The molecular masses of the puriﬁed wild-type and mutant enzymes were
determined on sucrose density gradients according to Martin and Ames (1961). The
markers used were rabbit muscle lactate dehydrogenase (MW 144,000) and rabbit muscle
pyruvate kinase (MW 237,000) (Worthington, 1988).
Thermal Stability

The puriﬁed enzymes were diluted to the same concentration, 0.2 mg/mL, in 50
mM Hepes—NaOH (pH 7.0) containing 1 mg/mL bovine serum albumin. The samples
were heated for 5 min in a 60 °C water bath and then immediately placed on ice. The
activities of these heated enzymes were assayed in the pyrophosphorolysis direction as
described above. The activity of the K382E enzyme was assayed in the absence of the
activator, 3PGA.
Reductive Phosphopyridoxylation

One hundred micrograms of the K382R-K419R enzyme in 50 mM Hepes—NaOH
(pH 8.0) was incubated with 0.1 mM or 0.5 mM [3H]PLP (about 38,000 cpm/nmol) in a
ﬁnal volume of 1 mL in the dark at room temperature for 30 min. Then 100 uL of NaBIL
was added to a ﬁnal concentration of '49 mM to reduce the Schiff base formed between
PLP and lysine residues. The reaction mixture was incubated in the dark at room
temperature for another 60 min. The incorporation of [3H]PLP into protein was measured
by determining trichloroacetic acid precipitable counts as described previously (Charng

et al., 1994).

RESULTS

Expression and Purification of Lys’” Mutant Enzymes

Normal expression of the mutant enzymes was conﬁrmed by resolving the crude
extract proteins with SDS—PAGE. Anabaena ADP-glucose pyrophosphorylase was
identiﬁed by immunoblotting with antibody prepared against the spinach leaf ADP-glucose
pyrophosphorylase that has been shown to be reactive with the Anabaena enzyme (Iglesias
et al., 1991; Charng et al., 1992). All the mutant enzymes were expressed at a level
similar to the wild-type enzyme based on the result of immunoblotting. The apparent sizes
of these mutant enzymes were the same as that of the wild-type enzyme. The mutant
enzymes were puriﬁed to greater than 90% homogeneity as estimated by SDS—PAGE of
about 5 pg of protein.
Kinetic Characterization of Lys382 Mutant Enzymes

The apparent afﬁnity for 3PGA decreased dramatically when Lys382 was replaced
with either arginine, alanine, or glutamine (Table 1). The A05 values for 3PGA of the
K382R, the K382A, and the K382Q enzymes were about 10-, 40-, and 160-fold higher
than that of wild-type enzyme, respectively. The interaction coefﬁcients were changed
from 1.0 for the wild-type to 1.7 for the mutant enzymes. This suggested that the binding
of the activator to the mutant enzymes was cooperative. The K382E enzyme was not
activated by 3PGA, but rather was inhibited with an 105 value of 5.9 mM. This is in
contrast with the K419E mutant enzyme (Charng et al., 1994) which is activated by 3PGA
with an A,5 about 6.0 mM.

In the absence of activator, the VM value of the K382A enzyme was slightly

62

63

 

Table 1: Kinetic Parameters for Activation of the Wild-Type and Mutant ADP-Glucose
Pyrophosphorylases“

 

 

Vmu(—-—3PGA) Vm(+3PGA)
A0_5(3PGA)(mM) (units/mg)” (units/mg)
wild-type 0.050 1 0.005 (1.0) 6.9 i 0.3 60 i 4
K382R 0.53 i 0.05 (1.7) 0.58 :1; 0.04 61 j; 6
K382A 1.9 i 0.2 (1.7) 8.8 i 0.7 57 j: 5
K382Q 7.8 i 0.5 (1.7) 0.87 :1; 0.03 5.9 :1: 0.3
K382E 5.9 :1: 0.2 (1.2)‘ 0.22 :1: 0.01 -
K419Rd 1.0 :t 0.1 (1.8) 0.8 :1; 0.1 103 i 11
K419Ed 6.0 :1: 0.8 (1.9) 0.4 :t 0.1 21 :1: 3
K382R-K419R 5.2 :I: 0.2 (2.1) 0.50 :1: 0.01 1.9 i 0.1

 

" Reactions were performed in the synthesis direction, assay 11 as described under
Materials and Methods. Data represent the mean :1: standard deviation of two independent
experiments. The values in parentheses are the Hill interaction coefﬁcients. ” One unit of
enzyme activity is expressed as the amount of enzyme required to form 1 pmol of ADP-
glucose/min at 37 °C assayed in the synthesis direction as described under Materials and
Methods. ‘ The K382E enzyme is inhibited by 3PGA. Therefore, the constant is 105
instead of A05. ‘ Cited from Charng et al. (1994).

 

64

higher, about 1.3-fold, than that of the wild-type enzyme, while the Vmax values of the
K382R, the K382Q, and the K382E enzymes were about 8, l3, and 3% of the wild-type
enzyme Vm, respectively. In the presence of saturating 3PGA, however, the Vm, values
of the K382R and the K382A enzymes were similar to that of wild-type, while the Vmax
value of the K382Q enzyme was about 10% of the wild-type V,,,,,. The degree of activation
increased from 8.7-fold for the wild-type enzyme to 105-fold for the K382R enzyme.
However, those for the K382A and the K382Q enzymes were about 6.5- and 6.8-fold,
respectively, similar to that of the wild-type enzyme (T able. 1).

The apparent afﬁnities for the substrates (ATP, glucose-l-P, Mg“) the and
inhibitor (P,) were all relatively less affected by the mutations at position 382 (Table 2),
indicating that the conformations of these ligand-binding sites were essentially unchanged.
The only signiﬁcant changes observed were the 4- or 5-fold increase in the 50.5 for ATP
of the K382Q enzyme in the absence or presence of activator, the 6- and 5-fold increases
in the S0, for glucose-l-P of the K382R and the K382E enzymes in the absence of
activator, respectively, and the 11-fold increase in the 105 value for P, of the K382E
enzyme in the absence of activator (Table 2). These changes are smaller than the 10-, 40-,
and 160-fold increases seen in the A05 value for the arginine, alanine, and glutamine
mutants, respectively. For the wild-type enzyme, 2.5 mM 3PGA increased the 105 value
for P, from 55 rm to 1.0 mM. With the apparent afﬁnity relatively unchanged for P, but
largely decreased for 3PGA, the K382R, the K382A, and the K382Q enzymes were more
sensitive to P, inhibition by having lower [05 in the presence of 2.5 mM 3PGA, the amount

which was saturating for the wild-type enzyme (Table 2).

65

 

A30: .3 8 med—U So 08 38.35 mavv— 05 can «gem 05 we 880888: 0.:qu 25. o .28 n. N 83 wow: 888.:wd-m mwo 85858088

08 8883wdtm Mo 3888 05 E

05 mg? 856 . 858508 88838 :5 05 88 8.85288 5 828.» 05. 558885 88:38qu 9.8

no .8890: 85.53» H 508 05 68882 San— .80502 :5 88582 Sun: 80288.. «a a .288 838:: $8588 08 8 8088.68 803 8088M ..

 

 

8.: 88 H 888 8.: 8 H 88 2.: 88 H :8 8.8 88 H 8.8 28 on
8.: 88 H 88 8.: 28 H 8.4 88: 88 H 28 8.: 38 H 88 88 882338802
8. : 88 H 2 8 8.: 2 H 2 8.: 888 H 888 8.: 88 H 88 28 on
2. :888H 888 88: 88 H S. 8.: 288 H 288 8.: 48 H N8 28 88:82
8 e 88 H 84.8 2...: 8 H 88 2.: 888 H 388 8.: 88 H R8 28 2
G: 88 H :8 8.: «8 H 28 8.: 88 H 28 8.: 28 H 5 88 .8232
2 :88 H 88 8.: 8 H «8 88: 88 H 88 88: 28 H 88 88 “.8802
8. : 88 H 2 8 8.8 28 H 3. 2.: 888 H 688 8.: 88 H 88 28 8m
8. :88 H a 8 88: 28 H 88 8.: 888 H $88 8.: 8 H 28 88 0882
8.8 so H 88 8.: 28 H 5.. 8.: 388 H ~88 8.: 88 H 88 :8 on
6:88 H 28 2.2: 28 H 88 88: 88 H 28 8.: 8 H 28 25: <89
8.8 88 H 9.8 8.: 28 H 88 2 : 388 H 388 8.: 88 H 88 28 n
2.: 88 H n 8 88: 28 H 88 8.: 88 H 38 8.: 28 H 3 88 8802
8988 H 8.2 8.3 8 H 88 88: 388 H 488 8.: 88 H :8 :8 an
8.:888H «88 88: 28 H to 8.8 288 H 988 8.: 28 H 4.2 88 25.83
28: .8 he .32 285 a»: 9:5 8-2-9.828 38: 8k <98

 

.8 3: 3m

 

.8388er 3820.83 8:2 ea 25.83 on. to 83.88 8252 8 2.8.

 

66

Kinetic Characterization of the K382R-K419R Enzyme

When both Lysm and Lys419 were replaced with arginine, the A0,, value for 3PGA
was about lOO-fold higher than that of the wild-type enzyme, while those of the K382R
and the K419R enzymes were 10- and 20-fold higher than that of the wild-type enzyme,
respectively (Fable 1). The interaction coefﬁcient was changed from 1.0 for the wild-type
to 2.1 for this double mutant enzyme. The Vm value was about 7% of the wild-type VM
in the absence of 3PGA and about 3% of that of the wild-type enzyme in the presence of
3PGA. Thus, catalytic efﬁciency was substantially decreased compared to the wild-type
enzyme. The mutations at residue 382 and 419 did not cause much alteration in the
apparent afﬁnities for the substrates, ATP, glucose-l-P, Mg“, and inhibitor, Pi (Table 2).
The only signiﬁcant changes observed were the 4-fold increase in the S05 for ATP in the
presence of 3PGA, the 7-fold increase in the So.5 for glucose-l-P in the absence of
activator, and the 4-fold increase in the 10.5 for Pi in the absence of activator (Table 2).
Effector Specificity

Previous studies have shown that a mutant Anabaena enzyme, in which Lys419 was
replaced by a glutamine, has an altered activator speciﬁcity (Charng et al. , 1995). The
mutant enzyme is activated more effectively by FBP than by 3PGA at lower concentrations
(Charng et al., 1995). It was of interest to examine whether the speciﬁcities for the
allosteric effectors of the Lys382 mutant enzymes had changed. Several compounds, some
of which are known as the major activators or inhibitors of other bacterial enzymes
(Preiss, 1991), were used to test their effects on the mutants (Table 3). Some compounds

have much different effects on the mutant enzymes than on the wild-type enzyme. But

67

3PGA still is the major activator for the K382R, the K382A, and the K382Q mutant
enzymes, while Pi still is the most effective inhibitor for all mutant enzymes. However,
the K382E enzyme cannot be activated by 3PGA, nor by any other effectors used. The
wild-type enzyme is activated by 3PGA more effectively than by either FBP or AMP
(Figure 2A). For the K382R-K419R enzyme, FBP and AMP activate the enzyme 5.4— and
8.3-fold at 5 mM, respectively (T able 3), and more effectively than 3PGA (Figure 2B),
which has only a 2.7-fold activation (Table 3). The A05 values for FBP and AMP of the
double-mutant enzyme are 0.35 and 2.0 mM, respectively, which are similar to those of
the wild-type enzyme, 0.11 and 1.8 mM.
Molecular Masses of the Wild-Type and Mutant Enzymes

The molecular mass of the wild-type enzyme determined on a sucrose density
gradient is 197 kDa. The calculated molecular mass based on the deduced amino acid
sequence is 193 kDa. This result is also in reasonable agreement with the molecular mass,
225 kDa, determined by gel ﬁltration chromatography (Iglesias et al., 1991) and veriﬁes
a tetrameric structure for ADP-glucose pyrophosphorylase from Anabaena. To ascertain
whether the loss of enzyme activity or the change of kinetic constants was perhaps due to
the instability of the tetrameric complex, each mutant was subjected to sucrose density
gradient sedimentation. The sedimentation patterns for the mutant enzymes were identical
to that of the wild-type enzyme with a single peak where the apparent molecular masses
ranged from 196 to 201 kDa, indicating that all the mutant enzymes retain the tetrameric

StﬂlCtlli'C .

68

 

Table 3: Speciﬁcity of Allosteric Effectors of the Wild-Type and Mutant ADP-Glucose Pyrophosphorylases”

 

ADP-glucose formed (nmol/10 min)

 

 

 

 

 

effector wild—type K382R K382A K382Q K382E. K382R-K419R
none 0.58 0.20 0.97 0.82 1.99 1.12
effector concn relative activity

effector (mM) wild-type K382R K382A K382Q K382E K382R-K419R
none 1.0 1.0 1.0 1.0 1.0 1.0
3PGA 2 9.8

5 10.7 89.5 6.4 3.1 0.6 2.7
fructose-6-P 2 5.7

5 6.5 7.5 1.9 1.7 1.0 1.1
fructose-1 ,6-P2 2 l .8

5 1.9 1.9 0.2 0.9 0.7 5.4
glucose—6-P 2 3.8

5 5.3 2.9 1.3 1.9 0.8 0.9
glucose-1,6-P2 2 0.4

5 0.4 1.8 0.2 0.2 0.9 4.6
pyruvate 2 1.0

5 1.1 1.3 0.9 2.3 0.9 1.0
PLP 2 2.0

5 2.0 16.8 0.3 0.2 0.1 1.9
P-enolpyruvate 2 6. 1

5 6.8 10.0 0.5 0.7 0.8 2.6
NADPH 0.5 0.8 1.0 1.0 0.9 0.8 1.0

2 0.4 0.9 0.6 0.7 0.7 0.9
hexane-1,6—diol-P2 0.5 0.6 0.5 0.2 0.5 0.5 0.7

2 0.5 0.3 0.1 0.2 0.3 0.6
ADP 2 0.9 1.8 0.6 1.6 0.8 2.1

5 0.5 0.3 0.3 0.4 0.2 1.2
AMP 2 2.5 4.3 0.7 2.0 0.9 5.9

5 3.3 1.6 0.7 1.3 0.5 8.3
Pi 2 0.1 0.3 0.1 0.2 0.1 0.3

 

‘ Reactions were performed in the synthesis direction, assay 11 as described under Materials and Methods, with the
presence of effectors as indicated. Data represent the average of two duplications with less than 10% deviation. The
amounts of the wild-type, the K382R, the K382A, the K382Q, the K3825, and the K382R-K419R enzymes used were
0.01, 0.04, 0.01, 0.15, 0.90, and 0.24 pg, respectively.

 

69

Figure 2: Activation of the wild-type (A) and K382R-K419R (B) enzymes by 3PGA
(O), FBP (O), and AMP (I). Initial velocities of the enzymes were determined in the
synthesis direction, assay 11, as described under Materials and Methods with the
concentration of activators being varied. The amounts of the wild-type and the K382R-
K419R enzymes used were about 0.02 and 0.48 pg, respectively.

70

 

 

A

 

A

 

A O

 

 

.
6 2 8 4 0
1. «I

EEO—Sec... €253 oaooaado<

[Activator], mM

 

 

 

    

 

 

 
 

 

. .
O 6 2 8 4 0
2 4| 4|

EEoEoEc 6252 33336.0(

20

16

12

[Activator], mM

Figure 2

71

Thermal Stability of Lys382 Mutant Enzymes

After heat treatment at 60 °C for 5 min, the activity of the wild-type enzyme
remained unchanged, while the K382R, the K382A, the K382Q, and the K382E enzymes
retained 94, 84, 84, and 85 % activity, respectively. The result indicates that Lys382 is not
required for the stability of the enzyme. The K382R-K419R enzyme retained about 78 %
activity after the same heat treatment.
Reductive Phosphopyridoxylation of the K382R-K419R Mutant Enzyme

Reductive phosphopyridoxylation of the K382R-K419R enzyme was performed to
determine if additional lysine residues could be labeled. The double-mutant enzyme was
activated about 2-fold by PLP, which was similar to that of the wild-type enzyme (Charng
et al., 1994). The A05 value of 0.20 mM was about 200-fold higher than that of the wild-
type enzyme (Charng et a1. , 1994). Unlike the wild-type and the K419R enzymes (Chamg
et al. , 1994), reductive phosphopyridoxylation had no effect on either enzyme activity or
3PGA activation of the K382R-K419R enzyme. In the presence of 0.1 or 0.5 mM
[3H]PLP, no incorporation of PLP into the mutant enzyme was found. Thus, no other

lysine residue appears to be involved in PLP binding.

DISCUSSION

Previous studies of Anabaena ADP-glucose pyrophosphorylase have shown that
LysM9 is located within the activator-binding site (Charng et al., 1994). Another lysyl
residue, Lysm, was modiﬁed by PLP when Lys419 was replaced by arginine (Charng et
al. , 1994), suggesting that Lys382 is also part of the activator-binding site. Alignment of
all the amino acid sequences of ADP-glucose pyrophosphorylases available has shown that
Lys382 is conserved in the cyanobacterial enzymes, in the small subunit of the higher-plant
enzymes, and in the large subunit of the spinach leaf, wheat leaf, and potato tuber enzymes
(Smith-White & Preiss, 1992). Site-directed mutagenesis experiments have been performed
to probe and verify the function of Lys382 of Anabaena ADP-glucose pyrophosphorylase.
The large effects on the A05 values for 3PGA when Lys382 was replaced by other amino
acids and the inability of 3PGA to activate the glutamic acid mutant enzyme are consistent
with the view that Lys382 is involved in 3PGA binding.

As substitutions of Lys382 go from basic to neutral, the apparent afﬁnities for 3PGA
decrease. When the substituted amino acid is acidic, the mutant enzyme cannot be
activated by 3PGA, suggesting that 3PGA cannot bind to the 3PGA-binding site. Thus,
the cationic property seems to be the most important factor in 3PGA binding at position
382. The arginine mutant enzyme has about a 10-fold lower apparent afﬁnity for 3PGA,
indicating that charge alone is insufﬁcient for the proper binding of 3PGA. The glutamine
mutant enzyme has a much lower apparent afﬁnity for 3PGA than the alanine mutant
enzyme though both glutamine and alanine are neutral amino acids. Therefore, the size of

the amino acid may also be important due to steric interference with proper binding of the

72

73
activator, 3PGA.

Interestingly, the K382E enzyme cannot be activated by 3PGA but rather is
inhibited. It seemed that 3PGA bound to the activator site induces an abnormal
conformational change when Lys382 was replaced by glutamic acid. But this is not the case
at position 419. The K419E enzyme still can be activated by 3PGA for about 50-fold with
an A05 value of 6.0 mM (Charng et al., 1994). Somehow, the enzyme can tolerate the
negative charge at position 419 for the binding of 3PGA.

All the mutant enzymes retain at least 80 % activities after heat treatment. Lys382
is obviously not critical to the stability of the native folded state. The kinetic constants for
ATP, Mg“, and glucose-l-P are relatively unaffected for the ﬁve mutants compared to
the dramatic changes in A05, showing the tolerance to amino acid substitution at position
382. The mutant enzymes are tetrameric in structure as is the wild-type enzyme, also
indicating that Lys382 is not essential for maintaining the quaternary structure of the
enzyme.

Based on previous studies (Charng et a1. , 1994) and our results, we can conclude
that both Lys382 and Lys419 are involved in the binding of 3PGA. Both charge and size of
these two lysyl residues are important for the binding of activator. However, the charge
seems to be more important, and this may be explained by the interactions between the two
positively charged e—amino groups and the negatively charged carboxyl and phosphate
groups of the activator, 3PGA. One possibility is that one of the two lysyl residues
interacts with the carboxyl group and the other one with the phosphate group. The two

negatively charged groups of 3PGA may interact with the two lysyl residues on the same

74

subunit or on two different subunits. In either case, the two lysyl residues that bind to the
same 3PGA molecule must be very close in three-dimensional structure.

The afﬁnities for P, of these mutant enzymes are relatively unaffected compared
to the dramatic changes of afﬁnities for 3PGA. This suggests that P, has a different binding
site, but the P, binding site may overlap with the 3PGA binding site and this may explain
why increasing concentrations of 3PGA reverses P, inhibition and that P, at higher
concentrations can reverse 3PGA activation (Charng et a1. , 1992, 1994).

One amino acid substitution at position 382 changes the effects of almost all
analogues of activator, 3PGA, on the Anabaena ADP-glucose pyrophosphorylase,
suggesting that these analogues bind at the same site as 3PGA to activate the enzyme.
Previous studies also have shown that FBP (Ghosh & Preiss, 1966; Charng et al., 1995)
and P-enolpyruvate (Ghosh & Preiss, 1966) bind to the same site as 3PGA. The fact that
no effectors activate the K382E enzyme can be explained by the electrostatic repulsion
between anionic ligands and enzyme.

FBP and AMP become more effective as activators than 3PGA for the K382R-
K419R enzyme. The alteration of the activator speciﬁcity was also observed for the
K419Q Anabaena mutant enzyme (Charng et al., 1995). For the K419Q enzyme, FBP is
about as effective as 3PGA and activates the enzyme more than 3PGA at lower
concentrations. For the K382R-K419R enzyme, FBP and AMP have much higher afﬁnities
than 3PGA and activate the enzyme more than 3PGA even at higher concentrations. It
should be noted that FBP and AMP are the activator and inhibitor, respectively, of the E.

coli ADP-glucose pyrophosphorylase. Both residues in the E. coli enzyme, corresponding

75
to Lys382 and Lys“19 in Anabaena enzyme, are arginine instead of lysine (Charng et al.,
1992). It is suggestive that the differences of activator speciﬁcity relate to the difference

of amino acids at the activator-binding site of the E. coli and Anabaena/higher-plant

enzymes.

REFERENCES
Ahrens, H., & Kortnyk, W. (1969) Anal. Biochem. 30, 413-420.
Aitken, A. (1988) Methods Enzymol. 167, 145-154.
Ball, K., & Preiss, J. (1994) J. Biol. Chem. 269, 24706-24711.
Bumette, W. W. (1981) Anal. Biochem. 112, 195-203.
Canellas, P. F., & Wedding, R. T. (1980) Arch. Biochem. Biophys. 199, 259-264.
Carlson, C. A., Parsons, T. F., & Preiss, J. (1976) J. Biol. Chem. 251, 7886-7892.

Charng, Y. Kakefuda, G., Iglesias, A. A, Buikema, W. J. & Preiss, J. (1992) Plant
Mol. Biol. 20, 37-47.

Charng, Y., Iglesias, A. A., & Preiss, J. (1994)]. Biol. Chem. 269, 24107-24113.
Charng, Y., Sheng, J ., & Preiss, J. (1995) Arch. Biochem. Biophys. 318, 476-480.
Gardiol, A., & Preiss, J. (1990) Arch. Biochem. Biophys. 280, 175-180.

Ghosh, H. P., & Preiss, J. (1965) J. Biol. Chem. 240, 960-961.

Ghosh, H. P., & Preiss, J. (1966) J. Biol. Chem. 241, 4491-4504.

Haugen, T. H., Ishaque, A., & Preiss, J. (1976) J. Biol. Chem. 251, 7880-7885.
Iglesias, A. A., Kakefuda, G., & Preiss, J. (1991) Plant Physiol. 97, 1187-1195.
Laemmli, U. K. (1970) Nature 227, 680-685.

Levi, C., & Preiss, J. (1976) Plant Physiol. 58, 753-756.

Martin, R. G., & Ames, B. N. (1961) J. Biol. Chem. 236, 1372-1379.

Morell, M., Bloom, M., Knowles, V., & Preiss, J. (1987) Plant Physiol. 85, 182-187.
Morell, M., Bloom, M., & Preiss, J. (1988) J. Biol. Chem. 263, 633-637.
Parsons, T., & Preiss, J. (1978) J. Biol. Chem. 253, 6197-6202.

Preiss, J ., Shen, L., Greenberg, E., & Gentner, N. (1966) Biochemistry 5, 1833-1845.

76

77

Preiss, J ., & Romeo, T. (1989) Adv. Microb. Physiol. 30, 183-238.

Preiss, J. (1991) in Plant Molecularand Cell Biology (Mifﬂin, B., ed.) Vol. 7, pp 59-114,
Oxford University, Oxford.

Preiss, J. (1996) Biotech. Annu. Rev. 2, 259-279.

Preiss, J, & Sivak, M. (1996) in Photoassimilate Distribution in Plants and Crops
(Zamski, E., & Schaffer, A. A., eds.) pp 63—96, Marcel Dekker, Inc., New York.

Sambrook, J ., Fritsch, E. F., & Maniatis, T. (1982) Molecular Cloning: A Laboratory
Manual, 2nd Ed., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

Sanger, F., Nicklen, S., & Coulson, A. R. (1977) Proc. Natl. Acad. Sci. U.S.A. 74,
5463-5467.

Sanwal, G. G., Greenberg, E., Hardie, J., Cameron, E. C., & Preiss, J. (1968) Plant
Physiol. 43, 417-427.

Sayers, J. R., Schmidt, W., & Eckstein, F. (1988) Nucleic. Acids Res. 16, 791-802.
Shively, J. M. (1988) Methods Enzymol. 167, 195-203.

Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H.,
Provenzano, M. D., Fujimoto, E. K., Geoke, N. M., Olson, B. J., & Klenk, D. C.
(1985) Anal. Biochem. 150, 76-85.

Smith-White, B. J ., & Preiss, J. (1992) J. Mol. Evol. 34, 449-464.

Stock, A., Ortanderl, F., & Pﬂeiderer, G. (1966) Biochem. Z. 344, 353-360.

Worthington, C. C. , (1988) Worthington Enzyme Manual: Enzymes and Related
Biochemicals, Worthington Biochemical Corporation, Freehold, NJ.

CHAPTER III

INVOLVEMENT OF ARGININE AT THE ALLOSTERIC SITES OF ADP-

GLUCOSE PYROPHOSPHORYLASE FROM ANABAENA PCC 7120

78

ABSTRACT

Treatment of ADP-glucose pyrophosphorylase from the cyanobacterium Anabaena
PCC 7120 with phenylglyoxal in 50 mM Hepes, pH 8.0, at 25 °C resulted in a time- and
concentration-dependent loss of enzyme activity. P,, the inhibitor, protected the enzyme
from inactivation most effectively, while 3PGA, FBP, PLP, and ATP plus magnesium
were also good protectors. After incubation with 2 mM phenylglyoxal for 1 hour, the
modiﬁed enzyme had a 10-fold lower apparent afﬁnity for P, in the absence of the
activator, 3PGA, than that of the wild-type enzyme. The apparent afﬁnity for 3PGA of the
modiﬁed enzyme was 4-fold lower than that of the wild-type enzyme. Thus, phenylglyoxal
appears to interfere with the allosteric regulation, as well as the catalysis, of ADP-glucose

pyrophosphorylase from Anabaena.

79

INTRODUCTION

ADP-glucose pyrophosphorylase is a key enzyme in the biosynthesis of glycogen
in bacteria (Preiss & Romeo, 1989; Preiss, 1996) and starch in plants (Preiss, 1991;
Preiss, 1996; Preiss & Sivak, 1996). This enzyme catalyzes the conversion of glucose-l-P
and ATP to ADP-glucose and PPi (Shen & Preiss, 1964).

ADP-glucose pyrophosphorylase is subject to allosteric regulation (Preiss &
Romeo, 1989; Preiss, 1996). There are signiﬁcant differences between the bacterial and
plant enzyme. The enzyme from enteric bacteria is activated by FBP and inhibited by
AMP (Preiss et al., 1966), whereas the enzyme from higher plants is mainly activated by
3PGA and inhibited by P, (Ghosh & Preiss, 1965; Sanwal et al., 1968). In addition, the
bacterial enzyme is homotetrameric in structure (Haugen et a1. , 1976), while the higher-
plant enzyme is heterotetrameric with two different subunits (Preiss, 1991).

Cyanobacteria have metabolic properties similar to chloroplasts of higher plants
(Aitken, 1988), but synthesize glycogen as the major carbohydrate reserve (Shively,
1988), in a similar manner to what is observed in bacteria. The cyanobacterial ADP-
glucose pyrophosphorylase is regulated by 3PGA and P, like the hi gher-plant enzyme (Levi
& Preiss, 1976; Iglesias et al., 1991), but is homotetrameric similar to the E. coli enzyme
(Iglesias et al., 1991). Due to the properties of the cyanobacterial enzyme and the key
position of these photosynthetic prokaryotes during evolution (Aitken, 1988), it is of great
interest to understand the structure—function relationships of the allosteric site of the
cyanobacterial ADP-glucose pyrophosphorylase.

Previous chemical modiﬁcation with PLP and site-directed mutagenesis studies

80

81
have shown that the activator-binding site of the E. coli ADP-glucose pyrophosphorylase,
Lys”, is near the N-terminus (Parsons & Preiss, 1978; Gardiol & Preiss, 1990), while

419

those of the cyanobacterial enzyme from Anabaena, Lys382 and Lys , are close to the C-

terrninus (Sheng et al., 1996; Charng et al., 1994). Studies on the spinach leaf enzyme

419 '

have shown that Lys“), which corresponds to Lys 1n Anabaena, is involved in binding
of the activator, 3PGA (Morell et al., 1988; Ball & Preiss, 1994).

Studies on E. coli, synechocystis, and spinach leaf enzymes using an arginine-
speciﬁc reagent, phenylglyoxal, suggested that there is also an essential arginine residue
involved in the allosteric activation and inhibition (Carlson & Preiss, 1982; Iglesias et al. ,
1992; Ball & Preiss, 1992).

ADP-glucose pyrophosphorylase from cyanobacterium Anabaena PCC 7120 has
been puriﬁed to homogeneity. The present study shows that the Anabaena enzyme is
inactivated by phenylglyoxal and that the inactivation can be prevented by the presence of
the activator 3PGA, the inhibitor P,, or the substrate ATP plus magnesium. The modiﬁed

enzyme is less sensitive to 3PGA activation and Pi inhibition. Stoichiometry studies show

that there is one arginine residue modiﬁed per enzyme subunit during inactivation.

MATERIALS AND METHODS

Reagents

Phenylglyoxal was purchased from Sigma. Stock solutions were prepared in 50 mM
Hepes-NaOH (pH 8.0). [7-”C]phenylglyoxal (27.0 mCi/mmol) was from Amersham.
[”C]Glucose-l-P was from ICN Pharmaceuticals Inc. All other reagents were of the
highest quality available.
Expression and Puriﬁcation of ADP-Glucose Pyrophosphorylase

E. coli mutant strain AC70R1-504, which is deﬁcient in ADP-glucose
pyrophosphorylase activity (Carlson et al., 1976), was used for expression of the
Anabaena ADP-glucose pyrophosphorylase gene (Charng et al., 1992) and grown in
enriched medium containing 1.1% KZHPO4, 0.85% KH2P04, 0.6% yeast extract, and
0.2% glucose, pH 7.0. The enzyme was puriﬁed as previously described (Iglesias et al.,
1991; Charng et al., 1992) with some modiﬁcation. Frozen cells were thawed into the
extraction buffer containing 20 mM K—P,, pH 7.5, 5 mM DTT, 1 mL EDTA (about 5 mL
of buffer/g of cells). The cell paste was sonicated for four 1 min intervals (1 min on ice
between two sonications). The homogenate was centrifuged at 10,000 rpm for 10 min. The
supernatant was absorbed onto a DEAE-sepharose Fast-Flow column (about 0.03 mL bed
volume/mg of protein) that had been equilibrated with 20 mM K—P, buffer, pH 7.5,
containing 2 mM DTT. After washing, the enzyme was eluted with a linear gradient
consisting of 5-bed volumes of the above buffer in the mixing chamber and S-bed volumes
of 50 mM K-P,, pH 6.0, containing 2 mM D'l'I‘ and 0.3 M KCl in the reservoir chamber.

Fractions of 9 mL were collected and those containing activity were pooled.

82

83

Ammonium sulfate was added to the sample with stirring to a ﬁnal concentration of 50%
saturation. After centrifugation at 8000 rpm for 20 min, the pellet was put on ice for 30
min and resuspended in 1.2 M ammonium sulfate, pH 7.5 (about 0.1 mL/mg of protein).
After centrifugation, the new pellet was dissolved in Buffer A (about 0.5 mL/mg of
protein) for mono Q (20 mM Bis-Tris-propane, 5 mM K-P,, 1 mM EDTA, 10% sucrose
(w/v), 2 mM DTT, pH 7.0.) and dialyzed against the same buffer. After dialysis, the
sample was applied to a Mono Q HR16/ 10 column equilibrated with buffer A. The column
was washed with 100 mL of Buffer A and eluted with two linear KC] gradients (50 mL,
0-0.1 M; 200 mL, 0.1-0.3 M) in Buffer A. Fractions of 5 mL were collected and those
containing activity were pooled. The pooled sample was concentrated to a ﬁnal
concentration of about 2 mg/mL in an Amicon concentrator ﬁtted with a YM-30
membrane.
Enzyme Assay

Enzymatic activity was measured in the ADP-glucose synthesis direction at 37 °C
according to the method of Preiss et al. (1966). (A) Activated conditions: The synthesis
of ADP-[“C]glucose from [”C]glucose-1-P and ATP was measured in the presence of
activator, 3PGA. The reaction mixture contained 100 mM Hepes—NaOH buffer (pH 8.0),
0.5 mM [”C]glucose-l-P (about 1000 cpm/nmol), 2.5 mM ATP, 10 mM MgC12, 2.5 mM
3PGA, 50 ug of bovine serum albumin, 0.15 unit of inorganic pyrophosphatase, and
enzyme in a ﬁnal volume of 200 pL. (B) Unactivated conditions: The assay in the absence
of activator was the same as described above except that 3PGA was omitted and the

amount of ATP was 5 mM in the reaction mixture.

84

Kinetic Characterization

Kinetic data were plotted as initial velocity versus substrate or effector
concentration. Saturating concentrations of substrates and effectors were determined to
ensure that maximal velocity was attained. Kinetic constants from hyperbolic plots were
determined by double-reciprocal plots. Sigmoidal plots were replotted as Hill plots to
obtain kinetic constants. Kinetic constants were expressed as A0, and 105, which
correspond to the concentration of activator and inhibitor giving 50 % of maximal
activation and inhibition, respectively. All the kinetic parameters calculated from double-
reciprocal or Hill plots were in good agreement with those obtained by using a computer
program (Canellas & Wedding, 1980) which performed nonlinear iterative least-squares
ﬁtting to the Hill equation.
Modification of ADP-Glucose Pyrophosphorylase with Phenylglyoxal

ADP-glucose pyrophosphorylase (0.5 uM) in 50 mM Hepes-NaOH (pH 8.0) was
incubated in the dark at room temperature with different concentrations of phenylglyoxal,
as indicated in the ﬁgure legends. The reactions were started by the addition of enzyme
and stopped by the addition of 10 mM arginine. In protection experiments, substrates or
allosteric effectors were added before the addition of enzyme. When necessary, the
reaction mixtures were passed through Sephadex G-50 before assaying for activity.
Incorporation of [“Clphenylglyoxal into ADP-Glucose Pyrophosphorylase

One hundred micrograms of ADP-glucose pyrophosphorylase in 50 mM
Hepes—NaOH (pH 8.0) was incubated with 4 mM [”C]phenylglyoxal (11,000 cpm/nmol)

in a volume of 100 uL. The incubation was carried out in the dark at room temperature.

85

At different times, aliquots of 10 al. were removed, mixed with 2 uL of 100 mM arginine
and after 5 min put on ice. The volume of each sample was brought to 100 uL with 50 mM
Hepes-NaOH (pH 8.0) and 10 aL was withdrawn to check the enzyme activity. One mL
of cold 10% (w/v) trichloroacetic acid was added to the remaining 90 uL of sample
followed by 1 h of incubation on ice. The precipitate was collected by centrifugation at
12,000 rpm for 15 min, washed twice with 1 mL of cold 10% trichloroacetic acid, and
dissolved in 400 aL of 3% NaZCO3 in 0.1 N NaOH. Two hundred pl. of this sample was
used to measure the protein concentration with a micro bicinchoninic acid reagent kit from
Pierce. The radioactivity was determined by counting the other 200 aL of the sample in

5 mL of Safety Solve.

RESULTS

Inactivation of ADP-Glucose Pyrophosphorylase by Phenylglyoxal

ADP-glucose pyrophosphorylase was inactivated by phenylglyoxal when the
incubation was carried out in Hepes buffer (pH 8.0) at room temperature over a 2 hours
period (Figure l). The 3PGA-activated and unactivated activity were affected by
phenylglyoxal in a similar pattern. Loss of enzyme activity with time was measured at
various concentrations of phenylglyoxal. The inactivations of both activated and
unactivated enzyme activity followed pseudo-ﬁrst-order kinetics (Figure 2A and 3A). The
orders of inactivation were determined (Figure 2B and 3B) according to Levy et al., 1963.
The n values were 0.95 and 0.90, respectively. This suggested that the loss of enzyme
activity resulted from at least one molecule of phenylglyoxal binding per enzyme subunit.
Protection of Enzyme Activity by Substrates and Allosteric Effectors

The effects of substrates and allosteric effectors on the inactivation by
phenylglyoxal are seen in Table 1. The protection against inactivation was indicated by the
increase in the time required for half-inactivation. For both 3PGA-activated and
unactivated activity, P, proved to be the most effective at protecting the enzyme from
inactivation by phenylglyoxal, while 3PGA and other activators, fructose-1,6-bisphosphate
and PLP, were also very good protectors. Substrate ATP plus Mg2+ had a similar
protection against inactivation as 3PGA, while PP, had a lesser effect. All the other

substrates had no or very little protection against the inactivation by phenylglyoxal.

86

87

Figure 1: Inhibition of Anabaena ADP-glucose pyrophosphorylase by
phenylglyoxal. Enzyme (0.5 aM) was incubated for 2 h with phenylglyoxal. Activity was
measured in the presence (O) and absence (O) of 2.5 mM 3PGA (Assay II) and 100 %
activity was 8.6 and 6.4 nmol of ADP-glucose formed per 10 min, respectively. The data
represents the mean of two separate experiments.

88

 

% Initial actuvnty
s. a)
. .° . ‘P

 

0 . ,

 

 

0.0 1.0 ' 2:0 370 4:0 7 5:0
[Phenylglyoxal], mM

Figure 1

89

Figure 2: Kinetics of inhibition of ADP-glucose pyrophosphorylase activity by
phenylglyoxal. (A) The loss of activity with time when enzyme (0.5 uM) was incubated
with the following concentrations of phenylglyoxal: 0, 0.4, 0.8, 1.5, and 2.4 mM. The
activity was measured under activated conditions (Assay 11A) and 100 % activity was 8.9
nmol of ADP-glucose formed per 10 min. The data represents the mean of two separate
experiments. (B) Determination of the order of the reaction.

N 2&5
SE .mEC.

ON? cor om 00 CV ON 0

— p p . _ s p . _ . -
S... .383: 3..

r E vow . do «we ova No. 1.9. I. OP

r'ﬁo

 

 

rﬁu.

. r v
q
N

and...

' 5
in m] 601

2.5...

 

 

m 1:.

 

E md

E Yo

 

 

l.. o o o « . 02
<

 

MARS? new %

91

Figure 3: Kinetics of inhibition of ADP-glucose pyrophosphorylase activity by
phenylglyoxal. (A) The loss of activity with time when enzyme (0.5 PM) was incubated
with the following concentrations of phenylglyoxal: 0, 0.4, 0.8, 1.5, and 2.4 mM. The
activity was measured under unactivated conditions (Assay HE) and 100 % activity was
6.9 nmol of ADP-glucose formed per 10 min. The data represents the mean of two
separate experiments. (B) Determination of the order of the reaction.

m 23E

SE .mEz.

 

92

 

 

 

 

cm? 2.: o.m o.m ow o.N o
n n F - vo—ZENNnx no. N 03v n
..... o - .9, q ... nor
E YN . a
sum
cad-... N
......n. ..
5:: m... ...-
. m 1' I
2E wd .
.28 v.0 .
o ”

 

 

 

o b I 2:

 

I‘M/“139 IBM“! %

93

 

Table 1: Effect of substrates and allosteric effectors on modiﬁcation of Anabaena ADP-
glucose pyrophosphorylase by phenylglyoxal

 

 

substrate/effector concentration (mM) to, (min)
+ 3PGA - 3PGA

none - 36 39
MgCl2 2 41 41
ATP + 10 mM MgCl2 2 240 375
ADP-glucose + 10 mM MgCl2 2 85 70
glucose-l-P 2 50 60
PP, 2 180 . 70
P, 2 370 700
3-P—glycerate 2 220 300
fructose-1 ,6-bisphosphate 2 220 220
pyridoxal—P 0.05 210 145

 

Enzyme was incubated with 2 mM phenylglyoxal in the presence of different ligands at the
stated concentrations. Aliquots were removed at different times and the modiﬁcations were
stopped with 10 mM arginine. Activity of the modiﬁed enzyme was assayed in the

presence and absence of 3PGA.

 

94
Effect of Modification with Phenylglyoxal on Enzyme Kinetics

Modiﬁcation with phenylglyoxal made the enzyme less sensitive to P, to inhibition
(Fig. 4). When the enzyme was assayed in the presence of 3PGA, the 10,, increased from
0.93 mM for the unmodiﬁed enzyme to 1.7 mM for the enzyme incubated with 2 mM
phenylglyoxal for 1 h (Fig. 4A). When the enzyme was assayed in the absence of
activator, the modiﬁed enzyme had an 10, about 0.27 mM, which was 10-fold higher than
that of the unmodiﬁed enzyme, 0.026 mM (Fig. 4B).

Inactivation of the enzyme with 1 mM and 2 mM phenylglyoxal for 1 h resulted in
an enzyme with A05 value for 3PGA of 0.12 mM and 0.22 mM, respectively, 2- and 4-fold
higher than that of the unmodiﬁed enzyme, 0.058 mM (data not shown).

Stoichiometry of Modification

Stoichiometry of the modiﬁcation by phenylglyoxal was examined by measuring
the [”C]phenylglyoxal incorporation during enzyme inactivation. Figure 5 shows the
incorporation of phenylglyoxal into the enzyme followed a linear correlation with the loss
of enzyme activity. At full inactivation, about 8 mol of phenylglyoxal was incorporated
per mol of enzyme. Since phenylglyoxal reacts with arginine in a 2:1 stoichiometry
(Takahashi, 1968), and the enzyme is a tetramer, this corresponds to 1 arginine residue

modiﬁed per subunit.

95

Figure 4: Inhibition of ADP-glucose pyrophosphorylase by P,. The enzyme was
incubated for l h in the absence (O) or presence (O) of 2 mM phenylglyoxal. After
incubation, enzyme activity was assayed under activated (A) or unactivated (B) conditions
(Assay II) and 100 % activity was 9.1 and 8.3 nmol of ADP-glucose formed per 10 min,
respectively. The data represents the mean of two separate experiments.

96

 

 

 

[P1] ,0 mM

 

. . . . .
0 0 0
8 6 4.

33:8 $25 ea

 

20-

 

 

 

 

 

3:8 325 a.

 

[Pl], mM

Figure 4

97

Figure 5: Incorporation of [”C]phenylglyoxal into ADP-glucose
pyrophosphorylase. Enzyme was incubated with 4 mM [”C]phenylglyoxal for different
times. Aliquots were assayed for incorporated radioactivity and 3PGA activated activity
(Assay 11A). The data represents the mean of two separate experiments and 100 % activity
was 14.3 nmol of ADP-glucose formed per 10 min.

% initial activity

98

 

 

 

 

. , . , . , . 1
0.0 2.0 4.0 6.0 8.0

14
mol [ Clphenylglyoxal I mol enzyme

Figure 5

 

DISCUSSION

Guanidinium groups are frequently involved in the binding of anionic ligands to
proteins (Takahashi, 1968; Riordan, 1979). They are suited to bind phosphate groups due
to their planar structure and hydrogen-bonding capability (Cotton et al., 1973; Riordan,
1979). The fact that 3PGA and P, are the physiological allosteric regulators of ADP-
glucose pyrophosphorylase and the previous chemical modiﬁcation studies with
phenylglyoxal (Carlson & Preiss, 1982; Iglesias et al., 1992; Ball & Preiss, 1992) suggest
that the allosteric region of the enzyme might contain arginine residue(s) involved in the
binding of the effectors. '

Anabaena ADP-glucose pyrophosphorylase was inactivated by phenylglyoxal. At
full inactivation, one arginine residue was modiﬁed per enzyme subunit. The activator
3PGA protected the enzyme from inactivation. However, the inhibitor P, proved to be
more effective at protecting the enzyme than 3PGA. In addition, substrate ATP plus Mg“
also protected the enzyme from inactivation, which was not seen in previous studies with
other ADP-glucose pyrophosphorylases (Carlson & Preiss, 1982; Ball & Preiss, 1992;
Iglesias et al., 1992).

Kinetic studies indicate that modiﬁcation of the Anabaena enzyme by phenylglyoxal
was not simply inactivating catalysis but also resulted from interference with normal
allosteric regulation. The modiﬁcation decreased the enzyme's sensitivity to inhibition by
P, and activation by 3PGA. These results suggest that guanidinium groups are in the
vicinity of the regulatory site and that the activator 3PGA and the inhibitor P, binding sites

are overlapping or they are connected by a conformational change, where binding at one

99

100
site affected binding at another site.

Although the enzymes from E. coli (Carlson & Preiss, 1982) and Synechocystis
(Iglesias et al., 1992) were also inactivated by phenylglyoxal, enzyme activity in the
absence of activator was not affected much by phenylglyoxal. The Anabaena and spinach
leaf enzymes showed loss of both activated and unactivated activity when modiﬁed by
phenylglyoxal.

These studies suggest that either modiﬁcation with phenylglyoxal affects the
enzyme's catalysis or phenylglyoxal could mimic P, as the allosteric inhibitor. Further site-
directed mutagenesis (random or alanine scanning) study might be useful to locate the

position(s) of the arginine residue(s) involved in the allosteric regulation.

REFERENCES
Aitken, A. (1988) Methods Enzymol. 167, 145-154.
Ball, K., & Preiss, J. (1992) J. Protein Chem. 11, 231-238.
Ball, K., & Preiss, J. (1994) J. Biol. Chem. 269, 24706-24711.
Canellas, P. F., & Wedding, R. T. (1980) Arch. Biochem. Biophys. 199, 259-264.
Carlson, C. A., Parsons, T. F., & Preiss, J. (1976) J. Biol. Chem. 251 , 7886-7892.
Carlson, C.A., & Preiss, J. (1982) Biochemistry 21, 1929-1934.

Charng, Y., Kakefuda, G, Iglesias, A. A., Buikema, W. J, & Preiss, J. (1992) Plant
Mol. Biol. 20, 37—47.

Charng, Y., Iglesias, A. A., & Preiss, J. (1994)]. Biol. Chem. 269, 24107-24113.

Cotton, F. A., Hazen, E. E., Day, V. W., Larsen, 8., Norman, J. G., Wong, S. T. K.,
& Johnson, K. H. (1973) J. Am. Chem. Soc. 95, 2367-2369.

Gardiol, A., & Preiss, J. (1990) Arch. Biochem. Biophys. 280, 175-180.

Ghosh, H. P., & Preiss, J. (1965) J. Biol. Chem. 240, 960-961.

Haugen, T. H., Ishaque, A., & Preiss, J. (1976) J. Biol. Chem. 251, 7880-7885.
Iglesias, A. A., Kakefuda, G., & Preiss, J. (1991) Plant Physiol. 97, 1187-1195.
Iglesias, A. A., Kakefuda, G., & Preiss, J. (1992) J. Protein Chem. .11, 119-128.
Levi, C., & Preiss, J. (1976) Plant Physiol. 58, 753-756.

Morell, M., Bloom, M., & Preiss, J. (1988) J. Biol. Chem. 263, 633-637.

Parsons, T., & Preiss, J. (1978) J. Biol. Chem. 253, 6197-6202.

Preiss, J ., Shen, L., Greenberg, E., & Gentner, N. (1966) Biochemistry 5, 1833-1845.
Preiss, J ., & Romeo, T. (1989) Adv. Microb. Physiol. 30, 183—238.

Preiss, J. (1991) in Plant Molecular and Cell Biology (Mifflin, B., ed.) Vol. 7, pp 59-114,
Oxford University, Oxford.

101

102
Preiss, J. (1996) Biotech. Annu. Rev. 2, 259-279.

Preiss, J, & Sivak, M. (1996) in Photoassimilate Distribution in Plants and Crops
(Zamski, E., Schaffer, A. A., ed.) pp 63-96, Marcel Dekker, Inc., New York.

Riordan, J. F. (1979) Mol. Cell. Biochem. 26, 71-92.

Sanwal, G. G., Greenberg, E., Hardie, J., Cameron, E. C., & Preiss, J. (1968) Plant
Physiol. 43, 417-427.

Shen, L., & Preiss, J. (1964) Biochem. Biophys. Res. Commun. 17, 424-429.
Sheng, J., Charng, Y., & Preiss, J. (1996) Biochemistry 35, 3115-3121.
Shively, J. M. (1988) Methods Enzymol. 167, 195-203.

Takahashi, K. (1968) J. Biol. Chem. 243, 6171-6179.

CHAPTER IV

ARGININE“ IS ESSENTIAL FOR THE INHIBITION OF ANABAENA PCC

7120 ADP-GLUCOSE PYROPHOSPHORYLASE BY PHOSPHATE

103

ABSTRACT

The chemical modiﬁcation study with phenylglyoxal has implicated the
involvement of an arginine residue at the allosteric sites, most probably the inhibitor-
binding site, of ADP-glucose pyrophosphorylase from cyanobacterium Anabaena PCC
7120 (Chapter III). In order to identify the arginine residue, ﬁve arginine residues, which
are conserved in all higher-plant and cyanobacterial enzymes but not in enteric bacterial
enzymes, were individually converted to alanine by site-directed mutagenesis. The mutant
enzymes were puriﬁed and the properties of these mutants were compared with the wild-
type enzyme. Substitution of arginine294 with alanine resulted in an enzyme with more than
100-fold or 40-fold lower afﬁnity for the inhibitor, P,, in the absence or presence of
3PGA, respectively. This mutation had no or lesser impact on the kinetic constants for the

substrates and the activator, 3PGA.

104

INTRODUCTION

ADP-glucose pyrophosphorylase catalyzes the reaction in which ADP-glucose, the
glucosyl donor of plant starch and bacterial glycogen synthesis, is synthesized from ATP
and glucose-l-P (Preiss & Romeo, 1989; Preiss, 1991; Preiss, 1996; Preiss & Sivak,
1996). The enzyme from higher plants and cyanobacteria is mainly activated by 3PGA and
inhibited by P, (Ghosh & Preiss, 1965; Sanwal et al., 1968), whereas the enzyme from
enteric bacteria is activated by FBP and inhibited by AMP (Preiss et al., 1966).

Previous chemical modiﬁcation with PLP and site-directed mutagenesis studies
have shown that the activator-binding site of the E. coli ADP-glucose pyrophosphorylase,
Lys”, is near the N-terminus (Parsons & Preiss, 1978; Gardiol & Preiss, 1990), while

419

those of the cyanobacterial enzyme from Anabaena, Lys382 and Lys , are close to the C-

terminus (Sheng et al., 1996; Charng et al., 1994). Studies on the spinach leaf enzyme

419

have shown that Lys‘”, which corresponds to Lys in Anabaena, is involved in binding
of the activator, 3PGA (Morell et al., 1988; Ball & Preiss, 1994).

Previous studies on E. coli, Synechocystis, and spinach leaf enzymes using an
arginine-speciﬁc reagent, phenylglyoxal, suggested that there was an essential arginine
residue involved in the allosteric activation and inhibition (Carlson & Preiss, 1982;
Iglesias et al., 1992; Ball & Preiss, 1992).

The present study (Chapter III) shows that the Anabaena enzyme is inactivated by
phenylglyoxal. The inactivation can be prevented by the presence of the activator 3PGA,

the inhibitor P,, or the substrate ATP plus magnesium, with P, as the best protector. The

apparent afﬁnity for P, of the modiﬁed enzyme is 10—fold lower than that

105

106

of the wild-type enzyme in the absence of 3PGA. The modiﬁed enzyme also has a 4-fold
lower afﬁnity for 3PGA. This suggested that either modiﬁcation affected the enzyme's
catalysis or phenylglyoxal could mimic P, as the allosteric inhibitor.

Alanine scanning of ﬁve conserved arginine residues which might be involved in
P, inhibition was performed. There are ﬁve arginine residues, Arg“, Argws, Arg‘",
Mg“, and Arg’“, conserved in the higher-plant and the cyanobacterial enzymes but not
in the E. coli enzyme. These ﬁve arginine residues were considered as the candidates for
P,-binding site, since P, inhibits the higher-plant and the cyanobacterial enzymes but it is
not the inhibitor of the E. coli enzyme. Replacing Ar 9“ with alanine resulted in an
enzyme with more than 100- or 40-fold lower apparent afﬁnity for P, in the absence or
presence of the activator, 3PGA, respectively. This result indicated that Arg294 was

involved in the binding of the inhibitor, P,.

MATERIALS AND METHODS

Reagents

[32P]PP, was purchased from Du Pont-New England Nuclear. [“C]Glucose-1-P was
from ICN Pharmaceuticals Inc. [oz-”SMATP and the in vitro mutagenesis kit were from
Amersham Corp. Oligonucleotides were synthesized and puriﬁed by the Macromolecular
Facility at Michigan State University. All other reagents were purchased at the highest
quality available.
Bacterial Strains and Media

E. coli strain TGl [K12, A(lac-pro), supE, thi, hsdD5/F‘traD36, proA+B+, lacl“,
lacZAMlS] was used for site-directed mutagenesis and grown in LB medium. E. coli
mutant strain AC70R1-504, which is deﬁcient in ADP-glucose pyrophosphorylase activity
(Carlson et al., 1976), was used for expression of the Anabaena ADP-glucose
pyrophosphorylase gene (Charng et al. , 1992) and grown in enriched medium containing
1.1% KZHP04, 0.85% KHZPO4, 0.6% yeast extract, and 0.2% glucose, pH 7.0.
Site-Directed Mutagenesis

Site-directed mutagenesis was performed according to the method of Sayers et al.
(1988) using the in vitro mutagenesis kit from Amersham Corp. Plasmid pAnaE3a was
used for both site-directed mutagenesis and gene expression (Charng et al. , 1992; Charng
et al., 1994; Sheng etal., 1996). The mutant enzymes with alanine substitution at position
66, 105, 171, 294, and 385 were designated as R66A, R105A, R171A, R294A, and
R385A, respectively. The oligonucleotides used for conversion to alanine are shown in

Table 1. The plasmids recovered in the last step of the mutagenesis were screened by

107

108

 

Table 1: Conserved arginine residues in Anabaena ADP-glucose pyrophosphorylase and
the oligonecleotides used for conversion to alanine

 

 

position amino acid sequence original sequence mutant sequence
66 SLNRHLS TCTCTCAATCGCCACATTGCC TCTCTCAATGCTCACATTGCC
105 DAVRQYL GATGCTGTACGTCAGTATCTC GATGCTGTAGCTCAGTATCTC
171 NSGRVID AACTCTGGACGAGTCATTGAT AACTCTGGAGCTGTCATTGAT
294 TRARYLP ACCCGCGCTCGTTACTTACCA ACCCGCGCTGCTTACTTACCA

3 8 5 KNARIGH AACAATGCCCGCATCGGTCAC AACAATGCCGCTATCGGTCAC

 

109

dideoxy sequencing (Sanger et al., 1977) in the regions of the desired mutations. The
entire coding regions of these mutant alleles were sequenced to verify that there were no
unintended mutations.

Expression and Purification of the Mutant Enzymes

The mutant genes were expressed in AC70Rl-504 cells. The R66A and the R294A
enzymes were puriﬁed to homogeneity as previously described (Chapter III). The
puriﬁcation procedure for the R105A and the R385A enzymes was the same as that for the
R66A and the R294A enzymes except that after the pellet was resuspended in 1.2 M
ammonium sulfate, pH 7.5 and centrifugation, the new pellet was dissolved in 50 mM
Hepes-NaOH, 2 mM DTT, pH 8.0 (about 0.5 mL/mg of protein).

Enzyme Assay

Assay I. Enzymatic activity was measured in the pyrophosphorolysis direction at
37 °C according to Preiss et al. (1966) during enzyme puriﬁcation. The reaction mixtures
contained 80 mM Hepes—NaOH buffer (pH 7.0), 2 mM ADP-glucose, 8 mM MgC12, 2
mM [32P]PP, (about 1500—3000 cpm/nmol), 4 mM 3PGA, 4 mM NaF, 50 pg of bovine
serum albumin, and enzyme in a total volume of 250 pL.

Assay II. Enzymatic activity was measured in the ADP-glucose synthesis direction
at 37 °C according to the method of Preiss et al. (1966). (A) Activated conditions: The
synthesis of ADP-[“C]glucose from [”C]glucose-1-P and ATP was measured in the
presence of activator, 3PGA. The reaction mixture contained 100 mM Hepes—NaOH
buffer (pH 8.0), 0.5 mM [“C]glucose-1-P (about 1000 cpm/nmol), 2.5 mM ATP, 10 mM

MgC12, 2.5 mM 3PGA, 50 pg of bovine serum albumin, 0.15 unit of inorganic

110

pyrophosphatase, and enzyme in a ﬁnal volume of 200 pL. (B) Unactivated conditions:
The assay in the absence of activator was the same as described above except that 3PGA
was omitted and the amount of ATP was 5 mM in the reaction mixture. For assay of the
R66A and the R294A enzymes, the reaction conditions were identical to wild-type, except
that 1.0 mM [“C]glucose-1-P was used to obtain maximal activity.
Kinetic Characterization

Kinetic data were plotted as initial velocity versus substrate or effector
concentration. Saturating concentrations of substrates and effectors were determined to
ensure that maximal velocity was attained. Data were replotted as double-reciprocal plots
to determine Vmu. Kinetic constants from hyperbolic plots were also determined by
double-reciprocal plots. Sigmoidal plots were replotted as Hill plots to obtain kinetic
constants. Interaction coefﬁcients, n", were also determined by Hill plots. Kinetic
constants were expressed as A0,, 80.5, and 10.5, which correspond to the concentration of
activator, substrate, or inhibitor giving 50% of maximal activation, velocity, and
inhibition, respectively. All the kinetic parameters calculated from double-reciprocal or
Hill plots were in good agreement with those obtained by using a computer program
(Canellas & Wedding, 1980) which performed nonlinear iterative least-squares ﬁtting to
the Hill equation.
Protein Assay

Protein concentration was determined by using bicinchoninic acid reagent (Smith

et al. , 1985) with bovine serum albumin as the standard.

111

Protein Electrophoresis and Immunoblotting

SDS—PAGE was performed according to Laemmli (1970). Following
electrophoresis, proteins on the gel were visualized by staining with Coomassie Brilliant
Blue R-250 or electroblotted onto a nitrocellulose membrane according to Bumette (1981).
After electroblotting, nitrocellulose membranes were treated with afﬁnity-puriﬁed anti-
spinach leaf ADP-glucose pyrophosphorylase IgG (Morell et al., 1987), and the
antigen—antibody complex was visualized via treatment with alkaline phOsphatase linked
goat anti-rabbit IgG followed by staining with BM purple AP-substrate precipitating
reagent (from Boehringer Mannheim GmbH).
Thermal Stability

The puriﬁed enzymes were diluted to the same concentration, 0.2 mg/mL, in 50
mM Hepes—NaOH (pH 7.0) containing 1 mg/mL bovine serum albumin. The samples
were heated for 5 min in a 60 °C water bath and then immediately placed on ice. The

activities of these heated enzymes were assayed in the pyrophosphorolysis direction as

described above.

RESULTS

Expression of the Mutant Enzymes

The expression of the mutant enzyme was conﬁrmed by resolving the crude extract
proteins with SDS—~PAGE. Anabaena ADP-glucose pyrophosphorylase was identiﬁed by
immunoblotting with antibody prepared against the spinach leaf ADP-glucose
pyrophosphorylase that has been shown to be reactive with the Anabaena enzyme (Iglesias
et al., 1991; Charng et al., 1992). The R66A, the R105A, and the R294A enzymes were
expressed at a level similar to the wild-type enzyme, while the R385A enzyme was
expressed less than the wild-type enzyme based on the result of immunoblotting. The
apparent size of the mutant enzymes were the same as that of the wild-type enzyme. The
R171A enzyme could not be identiﬁed by immunoblotting and the activity in the crude
extract could not be detected, while the amount of the expression plasmid in the cells was
similar to that of the wild-type enzyme.
Screen for Alanine-Substituted Mutant Enzymes with Altered P, Inhibition

Enzymatic activity in the crude extract was measured in the pyrophosphorolysis
direction (Assay I). The R66A and the R294A enzymes were less sensitive to P, inhibition
than the wild-type enzyme (Figure 1). The inhibition of the R105A enzyme and the R385A
enzyme by P, was similar to that of the wild-type enzyme (data not shown).
Purification of the Mutant Enzymes

The R66A and the R294A enzymes were puriﬁed to greater than 90%
homogeneity, while the R105A and the R385A enzymes were puriﬁed to greater than 50%

and 80% , respectively, as estimated by SDS—PAGE of about 5 pg of protein.

112

113

Figure 1: P, inhibition of the wild-type (O), the R66A (I), and the R294A (v)
enzymes in the crude extract. Enzyme activity was assayed in the pyrophosphorolysis
direction (Assay I) and 100 % activity was 20.6, 7.7, and 5.9 nmol of ATP formed per
10 min, respectively. The data represents the mean of two separate experiments.

114

 

 

 

 

 

33:8 .55 as

 

[Pi], mM

Figure 1

1 15
Kinetic Characterization of the R294A Enzyme

In the synthesis direction, the apparent afﬁnity for P, decreased dramatically when
Arg294 was replaced with alanine (Table 2). The 10.5 of the wild-type enzyme was 0.055mM
and the enzyme was inhibited 100 % by 0.04 mM P, in the absence of 3PGA, while the
remaining activity of the R294A enzyme was more than 60 % at 4 mM P, (Figure 2A). In
the presence of 3PGA, the 10.5 of the wild-type enzyme was 1.0 mM and the enzyme was
inhibited 100 % by 2 mM P,, while P, seemed to have no effect on the R294A enzyme up
to 4 mM (Figure 213). '

Up to 100 mM of P, was used to see the effect of P, at higher concentration. The
R294A enzyme was totally inhibited at 100 mM and the apparent afﬁnity for P, was 5.2
mM or 38 mM in the absence (Figure 2A) or presence (Figure 2B) of 3PGA, respectively.
Considering this effect may be due to the high ionic strength but not the speciﬁc inhibition
by P,, KCl with the ionic strength was used to see the effect on enzyme activity (Figure
3). KCl had very little effect on enzyme activity of the wild-type enzyme compared to P,
at lower ionic strength but inactivated the enzyme at high ionic strength (Figure 3A). The
effect of KCl on the R294A enzyme activity was less than P, at low ionic strength but
similar to P, at high ionic strength (Figure 3B). This suggested that the inhibition of P, at
high concentration was due to the ionic strength. Thus the actual 10,, for P, of the R294A
enzyme was more than 5.2 mM and 38 mM in the absence and presence of 3PGA,
respectively (Table 2).

In the presence of saturating 3PGA, the Vm value of the R294A enzyme was about

3-fold higher than that of wild-type, while in the absence of the activator, the VM value

116

 

.8528 $8588 05 S
enigma 0.. mm “a abacus—win? .«o 38.: u Show 9 8:38 0850 Mo 2508a ca 3 332:5 s bran“ 0% he «E: 25 .

 

 

.«2 «no: :2. «a: ..«8 +
86 a $5 _ a : 3 a 3. 3 a 3. no a me . cusses is
e: s... a «35 6.: ~86 a 98... e: 85 « Sod ca: Sod « god 8.: “86 « Bod :5 .agaﬁaa
8.: 85 « Ba ad _ a 8 A 8.: 8.0 a as 2.: ad a and 6.3 86 a 3 +
6.: «85 « ~85 6.: a... a an A e: «85 a E; a: 85 a go 8.: «85 a mad . 25 ..m
6.3 3 a 1.. . 9.3 3 « tn 2.: ..o a 3 a: 3 a 3 as: 3 a 9n +
G: 3 a 3 as: 3 a 3 ed 3 a 2. ed: 3 «.3 Ge 3 a to - as can:
8.: 89¢ a «85 8.: «85 a god 2.: Bod m E; 8.: 89¢ « ~36 ad: «86 a 235 +
3.8 nod « ”86 8.9 86 a :5 6.: ”85 a as... 8.: 485 a 485 8.9 8.:: a «86 - :5 disease
3.: ad a 26 a: 8... a n3 8.: So u :5 8.: 85 a $3 a: 86 a :6 +
a: 3 a 3 9.: 3 a S 8.: 3 a 3 e: 3 a 3 a: 3 a I - :a ..E.
3:21. .3— .am 3583.9...
<32 <32 <88 52 25.33

 

.acoatomxo 80:83qu 95 ...o dosage: 3%qu
a use 05 some? 35 sends <22 as 632 £82 .éem .2533 8% as .8 masseuse Boa: a 23

 

117

Figure 2: The effect of P, on the enzyme activity of the wild-type (O) and the
R294A (O) enzymes. One hundred percent activity was 6.5 and 12.0 nmol of ADP-glucose
formed per 10 min, respectively, when enzyme activity was assayed in the absence (A) of
3PGA (Assay IIB). When enzyme activity was assayed in the presence (B) of 3PGA (Assay
IIA), 100 % activity was 5.4 and 11.6 nmol of ADP-glucose formed per 10 min,
respectively. The data represents the mean of two separate experiments.

% initial activity

% initial activity

118

 

100

80

60

 

 

 

 

1.0

2.0 3.0 4.0 50.0 100.0
[Pi], mM

ll

 

20 -

 

 

0.0

II

 

119

Figure 3: The effect of P, (O) and KCl (I) on the enzyme activity of the wild-type
(A) and the R294A enzyme (B). Enzyme activity was assayed in the absence of 3PGA
(Assay IIB). One hundred percent activity was 6.5 and 12.2 nmol of ADP-glucose formed
per 10 min for the wild-type enzyme and 13.2 and 14.1 nmol of ADP-glucose formed per
10 min for the R294A enzyme. The data represents the mean of two separate experiments.

% initial activity

% initial activity

120

J]
II

 

 

Jl

 

  

 

 

' é ' 5 "'160'263300
Ionic strength

0
‘-

 

100

80

60-

40-

20-

 

 

 

 

0 so too ' 150 ' zoo ' 250 300
Ionic strength

Figure 3

121

was about 2-fold higher. The apparent afﬁnities for the substrates, ATP, glucose-l-P,
Mg“ , and activator, 3PGA were all relatively unaffected by the mutation at position 294
(Table 2), indicating that the conformation of this P,-binding site was relatively unchanged.
Kinetic Characterization of the R66A, the R105A, and the R385A Enzymes

The apparent afﬁnity for P, of the R66A enzyme decreased about 5-fold compared
to the wild-type enzyme, while those of the R105A and the R385A enzymes were similar
to that of the wild-type (Table 2). The afﬁnities for substrates, ATP, glucose-l-P, Mg“,
and activator, 3PGA were all relatively unaffected by the mutations (Table 2), indicating
that the conformations of these ligand-binding sites are relatively unchanged. The only
signiﬁcant changes observed were the 4- and 7-fold increase in the 80.5 for ATP of the
R66A and the R105A enzyme in the presence of activator, and the 10- and 5-fold
decreases in the Vm of the R385A enzyme in the absence and presence of activator,
respectively (Table 2).
Thermal Stability of the Wild-Type and the Mutant Enzymes

After heat treatment at 60 °C for 5 min, the activity of the wild—type, the R66A,
R105A, R294A, and the R385A enzymes retained 92, 82, 79, 88, and 75% activity,
respectively. The result indicates that these arginine residues are not required for the

stability of the enzyme.

DISCUSSION

Anabaena ADP-glucose pyrophosphorylase was inactivated by phenylglyoxal. At
full inactivation, one arginine residue was modiﬁed per enzyme subunit. The activator
3PGA protected the enzyme from inactivation. However, the inhibitor P, proved to be
more effective at protecting the enzyme than 3PGA. In addition, substrate ATP plus Mg“
also protected the enzyme from inactivation, which was not seen in previous studies with
other ADP—glucose pyrophosphorylases (Carlson & Preiss, 1982; Ball & Preiss, 1992;
Iglesias et al., 1992).

Alignment of all the amino acid sequences of ADP-glucose pyrophosphorylase
available has shown that these ﬁve arginine residues are conserved in all the cyanobacterial
and the higher-plant enzymes (Smith-White & Preiss, 1992). Considering the effect of the
ionic strength on enzyme activity, the actual apparent afﬁnity for P, of the R294A enzyme
should be more than 100- and 40-fold lower than those of the wild-type enzyme in the

absence and presence of 3PGA, respectively. This large effect on the 10,, value for P, when

294

Arg was replaced by alanine suggests that Arg“ is involved in P, binding.

Although the 105 for P, of the R66A enzyme in the absence of 3PGA was 5-fold
higher than that of the wild-type, this change was not signiﬁcant compared to the more
than 100-fold change for the R294A enzyme. It was also not speciﬁc considering the So,
for ATP in the presence of 3PGA and the afﬁnity for glucose-l-P in the absence of 3PGA
were changed 4- and 2-fold, respectively. Thus, Arg66 seems not to be directly involved

in P, inhibition.

The R105A enzyme had a 7-fold lower afﬁnity for ATP in the presence of 3PGA

122

123

while it had similar kinetic constants for other substrates, activator 3PGA, and inhibitor
P,. This suggested that Arg‘°5 might be close to ATP-binding site. It should be noted that
Arg105 is eight amino acids away in the alignment (Smith-White & Preiss, 1992) from
Tyr‘” in E. coli enzyme, which is the putative ATP-binding site (Lee & Preiss, 1986).

The apparent afﬁnities for all substrates, activator 3PGA, and inhibitor P, of the
R385A enzyme were similar to those of the wild-type enzyme. The Va,” of this mutant
enzyme was 10- or 5-fold lower than that of the wild-type enzyme in the absence or
presence of 3PGA, respectively, which suggests that mutation at Arg’” may result in a
conformational change and affect the enzyme's catalysis.

The R171A enzyme could not be identiﬁed by immunoblotting and no activity
could be detected in the crude extract, though the amount of the expression plasmid in the
cells was similar to that of the wild-type expression plasmid. Argm may be important for
maintaining the proper structure of ADP-glucose pyrophosphorylase. All the other mutant
enzymes were expressed at a level similar to the wild-type enzyme and retained more than
75% activities after heat treatment. Arg“, Argw’, Arg294, and Arg’“ are obviously not
critical to the stability of the native folded state. Evidence of this was also observed in the
relatively unchanged kinetic constants obtained for ATP, Mg“, glucose-l-P, and 3PGA
for the four mutants, showing the tolerance to amino acid substitution at these positions.

Due to the properties of the cyanobacterial ADP-glucose pyrophosphorylase and
the key position of these photosynthetic prokaryotes during evolution (Aitken, 1988), it
is of great interest to understand the structure—function relationships of the allosteric sites

of the cyanobacterial ADP-glucose pyrophosphorylase. Previous studies have shown that

124

Lys382 and Lys419 of Anabaena enzyme were involved in the binding of the activator,
3PGA. The present study clearly indicated that Arg294 of Anabaena enzyme is involved in
the binding of P,. The activator and the inhibitor obviously bind to different sites.

It would be interesting to replace Arg294 with another positively charged amino
acid, lysine, or other neutral amino acids with larger size than alanine, such as leucine or
glutamine, or even a negatively charged amino acid, like glutamic acid. This would show
if the charge is the most important factor for binding of P, and if the size of the amino acid
at this position is also important. The fact that the corresponding amino acid at this
position in the E. coli enzyme is glutamic acid makes it more interesting to see whether
or not the inhibitor speciﬁcity would change when substituting Arg294 in the Anabaena
enzyme with glutamic acid or replacing the corresponding glutamic acid in E. coli enzyme
with arginine.

Although more site-directed mutagenesis study at this P,-binding site would be very
interesting, the most reliable way to study the structure-function relationships of ADP-

glucose pyrophosphorylase is to crystallize the enzyme and determine the 3-dimensional

structure .

REFERENCES
Aitken, A. (1988) Methods Enzymol. 167, 145-154.
Ball, K., & Preiss, J. (1992) J. Protein Chem. 11, 231-238.
Ball, K., & Preiss, J. (1994) J. Biol. Chem. 269, 24706-24711.
Bumette, W. W. (1981) Anal. Biochem. 112, 195-203.
Canellas, P. F., & Wedding, R. T. (1980) Arch. Biochem. Biophys. 199, 259-264.
Carlson, C. A., Parsons, T. F., & Preiss, J. (1976) J. Biol. Chem. 251, 7886-7892.
Carlson, C.A., & Preiss, J. (1982) Biochemistry 21, 1929-1934. '

Charng, Y., Kakefuda, G., Iglesias, A. A., Buikema, W. J ., & Preiss, J. (1992) Plant
Mol. Biol. 20, 37-47.

Charng, Y., Iglesias, A. A., & Preiss, J. (1994) J. Biol. Chem. 269, 24107-24113.
Gardiol, A., & Preiss, J. (1990) Arch. Biochem. Biophys. 280, 175-180.

Ghosh, H. P., & Preiss, J. (1965) J. Biol. Chem. 240, 960-961.

Iglesias, A. A., Kakefuda, G., & Preiss, J. (1991) Plant Physiol. 97, 1187-1195.
Iglesias, A. A., Kakefuda, G., & Preiss, J. (1992) J. Protein Chem. 11, 119-128.
Laemmli, U. K. (1970) Nature 227, 680-685.

Lee, Y. M., & Preiss, J. (1986) J. Biol. Chem. 261, 1058-1064.

Morell, M., Bloom, M., Knowles, V., & Preiss, J. (1987) Plant Physiol. 85, 182-187.
Morell, M., Bloom, M., & Preiss, J. (1988) J. Biol. Chem. 263, 633-637.

Parsons, T., & Preiss, J. (1978) J. Biol. Chem. 253, 6197-6202.

Preiss, J ., Shen, L., Greenberg, E., & Gentner, N. (1966) Biochemistry 5, 1833-1845.
Preiss, J ., & Romeo, T. (1989) Adv. Microb. Physiol. 30, 183-238.

Preiss, J. (1991) in Plant Molecular and Cell Biology (Mifflin, B., ed.) Vol. 7, pp 59-

125

126
114, Oxford University, Oxford.
Preiss, J. (1996) Biotech. Annu. Rev. 2, 259-279.

Preiss, J, & Sivak, M. (1996) in Photoassimilate Distribution in Plants and Crops
(Zamski, E., Schaffer, A. A., ed.) pp 63-96, Marcel Dekker, Inc., New York.

Sanger, F., Nicklen, S., & Coulson, A. R. (1977) Proc. Natl. Acad. Sci. U.S.A. 74,
5463-5467.

Sanwal, G. G., Greenberg, E., Hardie, J., Cameron, E. C., & Preiss, J. (1968) Plant
Physiol. 43, 417-427.

Sayers, J. R., Schmidt, W., & Eckstein, F. (1988) Nucleic. Acids Res. 16, 791-802.
Sheng, J., Charng, Y., & Preiss, J. (1996) Biochemistry 35, 3115—3121.

Smith, P. K., Krohn, R. 1., Hermanson, G. T., Mallia, A. K., Gartner, F. H.,
Provenzano, M. D., Fujirnoto, E. K., Geoke, N. M., Olson, B. J., & Klenk, D. C.
(1985) Anal. Biochem. 150, 76-85.

Smith-White, B. J ., & Preiss, J. (1992) J. Mol. Evol. 34, 449-464.

CHAPTER V

SUMMARY AND PERSPECTIVES

127

SUMNIARY AND PERSPECTIVES

1 . Lys382 of the cyanobacterium Anabaena ADP-glucose pyrophosphorylase is involved
in the binding of the activator, 3PGA.

A site—directed mutagenesis study on the cyanobacterium Anabaena ADP-glucose
pyrophosphorylase was conducted to determine the role of Lys382 in binding of the
activator, 3PGA. This work was based on a previous chemical modiﬁcation study.
Phosphopyridoxylation of the K419R mutant enzyme modiﬁed Lysm, suggesting that this
residue might also be located within the activator-binding site (Charng et al., 1994). The
corresponding lysine residue on the large subunit of spinach leaf enzyme was labeled with
PLP in a similar phosphopyridoxylation experiment (Ball & Preiss, 1994). Alignment of
all the amino acid sequences of ADP-glucose pyrophosphorylases available has shown that
Lys3m is conserved in the cyanobacterial enzymes, in the small subunit of the higher-plant
enzymes, and in the large subunit of the spinach leaf, wheat leaf, and potato tuber enzymes
(Smith-White & Preiss, 1992).

Replacing Lys382 with either arginine, alanine, or glutamine produced mutant
enzymes with apparent afﬁnities for 3PGA 10—, 40-, or 160-fold lower than that of the
wild-type enzyme. The glutamic acid mutant enzyme was inhibited by 3PGA and it may
be due to an abnormal conformational change. The kinetic constants for ATP, Mg“,
glucose-l-P, and P, are relatively unchanged for the ﬁve mutants compared to the dramatic
changes in Am, suggesting that there are different binding sites for the substrates and the

inhibitor, P,.

128

129
Site-directed mutagenesis was also performed to obtain a K382R-K419R double
mutant. The apparent afﬁnity for 3PGA of this double-mutant enzyme was about 100-fold
lower than that of the wild-type enzyme, and the speciﬁcity for activator of this mutant
enzyme was altered. FBP and AMP, the major activator and inhibitor of the E. coli ADP-
glucose pyrophosphorylase, were more effective as activators than 3PGA. Both residues

419

in E. coli enzyme, corresponding to Lys382 and Lys in Anabaena enzyme, are arginine
instead of lysine (Charng et al. , 1992). The differences of activator speciﬁcity may relate
to the difference of amino acids at the activator-binding site of the E. coli and
Anabaena/higher-plant enzymes. Chemical modiﬁcation study on the K382R-K419R
enzyme with PLP suggested that other lysine residues are not involved in the binding of
3PGA.

Based on previous studies (Charng et al. , 1994) and our results, we can conclude
that both Lys382 and Lys419 are involved in the binding of 3PGA. Both charge and size of
these two lysine residues are important for the binding of the activator. The charge seems
to be more important, which may be explained by the interactions between the two
positively charged e-amino groups and the negatively charged carboxyl and phosphate

groups of 3PGA.

2. An arginine residue(s) is involved in the allosteric regulation of Anabaena ADP-
glucose pyrophosphorylase.
Previous chemical modiﬁcation studies with phenylglyoxal have shown that there

is an essential arginine residue(s) involved in the allosteric regulation (Carlson & Preiss,

130

1982; Iglesias et al. , 1992; Ball & Preiss, 1992). Treatment of ADP-glucose
pyrophosphorylase from the cyanobacterium Anabaena with phenylglyoxal resulted in a
time- and concentration-dependent loss of both activated and unactivated activity when
modiﬁed by phenylglyoxal. P,, the inhibitor, protected the enzyme from inactivation most
effectively, while 3PGA, FBP, PLP, and ATP plus magnesium were also good protectors.
The modiﬁed enzyme had lower apparent afﬁnity for P, in the absence of the activator,
3PGA, than that of the wild-type enzyme. At full inactivation, one arginine residue was
modiﬁed per enzyme subunit. These results suggested that either modiﬁcation affected the

enzyme's catalysis or phenylglyoxal could mimic P, as the allosteric inhibitor.

3. Argz“ of the Anabaena ADP-glucose pyrophosphorylase is important for the
inhibition by P,.

The chemical modiﬁcation study with phenylglyoxal has implicated the
involvement of an arginine residue(s) at the allosteric sites, most probably the inhibitor-
binding site, of ADP-glucose pyrophosphorylase from cyanobacterium Anabaena. In order
to identify the arginine residue(s), ﬁve conserved arginine residues were individually
converted to alanine by site-directed mutagenesis. The properties of these mutant enzymes
were compared with those of the wild-type enzyme. Substitution of arginine294 with alanine
produced an enzyme with more than 100-fold or 40-fold lower afﬁnity for the inhibitor,
P,, in the absence or presence of 3PGA, respectively. This mutation had no or lesser
impact on the kinetic constants for the substrates and the activator, 3PGA. This large

effect on the 1,,5 value for P, when Argz94 was replaced by alanine suggests that Arg“294 is

131
involved in P, binding.

The apparent afﬁnity for P, of the R66A enzyme decreased about 5-fold compared
to that of the wild-type enzyme, while those of the R105A and the R385A enzymes were
similar to that of the wild-type. The afﬁnities for the substrates, ATP, glucose-l-P, Mg“,
and the activator, 3PGA were all relatively unaffected by the mutations at position 66,
105, and 385, indicating that the conformations of these ligand-binding sites are relatively
unchanged. The only signiﬁcant changes observed were the 4- and 7 -fold increase in the
5,5 for ATP of the R66A and the R105A enzyme in the presence of activator, and the 10-
and 5-fold decreases in the Vm, of the R385A enzyme in the absence and presence of
activator, respectively.

The R171A enzyme could not be identiﬁed by immunoblotting and no activity
could be detected in the crude extract, though the amount of the expression plasmid in the
cells was similar to that of the wild-type expression plasmid. Arg171 may be important for

maintaining the proper structure of ADP-glucose pyrophosphorylase.

4. Future Research:

It is of great interest to understand the structure—function relationships of the
allosteric sites of the cyanobacterial ADP-glucose pyrophosphorylase due to the properties
of the cyanobacterial ADP-glucose pyrophosphorylase and the key position of these
photosynthetic prokaryotes during evolution (Aitken, 1988). We have shown that Lysm,

419

in addition to Lys , of Anabaena enzyme is involved in the binding of the activator,

3PGA. The present study also indicates that Arg294 of the Anabaena enzyme is involved

132

in the binding of P,. The activator and the inhibitor obviously bind to different sites.

It would be of interest to replace Arg294 with another positively charged amino acid,
lysine, or other neutral amino acids with larger size than alanine, such as leucine or
glutamine, or even a negatively charged amino acid, like glutamic acid. This would show
if the charge is the most important factor for binding of P, and if the size of the amino acid
at this position is also important. The fact that the corresponding amino acid at this
position in the E. coli enzyme is glutamic acid makes it more interesting to see if the
inhibitor speciﬁcity would change when substituting Arg294 in the Anabaena enzyme with
glutamic acid or replacing the corresponding glutamic acid in E. coli enzyme with
arginine. More site—directed mutagenesis studies on other cyanobacterial and higher-plant
ADP-glucose pyrophosphorylases would show if the arginines corresponding to Arg294 are
also important for P, inhibition.

The combination of chemical modiﬁcation and site-directed mutagenesis is very
useful to study the structure-function relationships of ADP-glucose pyrophosphorylase,
while determination of the 3-dimensional structure will be very helpful. Other biophysical
techniques, such as NMR and circular dichroism, also may be very useful to provide

information about the structure of ADP-glucose pyrophosphorylase.

REFERENCES

Aitken, A. (1988) Methods Enzymol. 167, 145-154.

Ball, K., & Preiss, J. (1992) J. Protein Chem. 11, 231-238.
Ball, K., & Preiss, J. (1994) J. Biol. Chem. 269, 24706-24711.
Carlson, C.A., & Preiss, J. (1982) Biochemistry 21, 1929-1934.

Charng, Y., Kakefuda, G., Iglesias, A. A., Buikema, W. J ., & Preiss, J. (1992) Plant
- Mol. Biol. 20, 37-47.

Charng, Y., Iglesias, A. A., & Preiss, J. (1994) J. Biol. Chem. 269, 24107-24113.
Iglesias, A. A., Kakefuda, G., & Preiss, J. (1992) J. Protein Chem. 11, 119-128.

Smith-White, B. J ., & Preiss, J. (1992) J. Mol. Evol. 34, 449-464.

133