Investigation of the kinetic mechanism of glutamine- and ammonia-dependent reactions of E. coli asparagine synthetase B ...

MISSING IMAGE

Material Information

Title:
Investigation of the kinetic mechanism of glutamine- and ammonia-dependent reactions of E. coli asparagine synthetase B using isotope partitioning and steady-state kinetics
Physical Description:
Book
Creator:
Habibzadegah-Tari, Pouran, 1960-
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 028194357
oclc - 48660237
System ID:
AA00022852:00001

Full Text












INVESTIGATION OF THE KINETIC MECHANISM OF GLUTAMINE- AND
AMMONIA-DEPENDENT REACTIONS OF E. coli ASPARAGINE SYNTHETASE
B USING ISOTOPE PARTITIONING AND STEADY-STATE KINETICS















By


POURAN HABIBZADEGAH-TARI


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



UNIVERSITY OF FLORIDA


1996















ACKNOWLEDGMENTS


I would like to thank the members of my committee, Dr

cain, Dr Cohen, Dr Dunn, and Dr Richards for their guidance

and suggestions. I would like to thank my advisor Dr Schuster

for giving me the chance to do research in his laboratory.

Although he was always busy with a lot of other things, the

very limited time I had with him was filled with new ideas to

solve problems. I would very much like to thank Dr Alison

helping me with my research, and for being there for me

always.

I would like to thank all the friends I met in and out

of Dr Schuster's laboratory for their support and

encouragement.

I thank my husband, Esfandiar, for being my friend for

the past eighteen years, specially during this period of my

life. I am grateful for his ongoing support and for lifting

my spirit when it was down.

I am specially grateful to my mom for coming to this

country and taking care of my daughter while I was busy with

school. I know how hard it was for her to stay in this

country for ten months and not be able to communicate with

anyone. I want thank her with all my heart and say how sorry

I am for not being able to spend enough time with her.










Finally, I want thank my daughter, Maryam, for being who

she is, my love. I hope I can make it up to her after I

finish school.
















TABLE OF CONTENTS


AKNOWLEDGEMENTS ............................................ ii

LIST OF TABLES ............................................. vi

LIST OF FIGURES ........................................... vii

ABSTRACT .................................................. xii

CHAPTERS

1 INTRODUCTION................................................ 1

2 INVESTIGATION OF THE MECHANISM OF E. COLI ASPARAGINE
SYNTHETASE USING ISOTOPE PARTITIONING..................... 26
Introduction .......................................... 26
Materials and Methods ................................ 29
Chemicals and Reagents ............................. 29
Expression of the Protein and Purification ......... 30
Protein Concentration Determination ................ 31
Isotope Partitioning Experiments with
Radioactive L-Aspartate, Glutamine- and
Ammonia-Dependent Reactions ...................... 31
Determination of Kd for L-Aspartate ................ 34
Isotope Partitioning Experiments with
Radioactive ATP, Glutamine- and Ammonia-
Dependent Reactions ............. ...................34
Aspartate-Dependent ATP Hydrolysis ................ 36
Theory ................................................ 37
Results ................................................... 39
Isotope Partitioning Experiments with
Radioactive L-Aspartate, Glutamine- and
Ammonia-Dependent Reactions ...................... 39
Isotope Partitioning Experiments with
Radioactive ATP, Glutamine- and Ammonia-
Dependent Reactions .............................. 40
Aspartate-Dependent ATP Hydrolysis ................. 41
Discussion ............................................ 47

3 SUBSTRATE BINDING AND PRODUCT RELEASE OF ASPARAGINE
SYNTHETASE B STUDIED BY STEADY STATE KINETICS ............ 55
Introduction .......................................... 55









Materials and Methods ................................. 56
Chemicals and Reagents ............................. 56
Expression of the Protein and Purification ......... 57
Protein Concentration Determination ................ 57
Enzyme Assays ...................................... 57
Stoichiometry of PPi and L-Glutamate ............... 59
Results ............................................... 60
Initial Rate Studies ............................... 60
Inhibition by Substrate Analogs .................... 65
Stoichiometry of Glutamine-Dependent Reaction ...... 67
Product Inhibition Studies ......................... 68
Discussion ............................................ 74

4 EFFECT OF TEMPERATURE ON THE ASPARAGINE SYNTHETASE B.... 146
Introduction ......................................... .146
Materials and Methods ................................ 146
Chemicals and Reagents ............................ 146
Expression of the Protein and Purification ........ 147
Protein Concentration Determination ............... 147
Enzyme Assays ..................................... 147
Thermodynamic of Activation ....................... 148
Results and Discussion ............................... 149

5 SUMMARY AND CONCLUSIONS ................................. 157

LIST OF REFERENCES ........................................ 166

BIOGRAPHICAL SKETCH ....................................... 172















LIST OF TABLES


2.1 Trapping of L-aspartate from Complexes in the Steady
State, Using (14C) L-aspartate, Ammonia- and
Glutamine- Dependent Reactions ..................... 43

2.2 Trapping of ATP from Complexes in the Steady State,
Using (3H) ATP, Ammonia- and Glutamine- Dependent
Reactions .......................................... 44

3.1 Inhibition patterns for ASB obtained with P-methyl
aspartate, AMP-PNP and L-glutamic acid y-methyl ester
with respect to L-aspartate, ATP and L-glutamine. .142

3.2 Product inhibition data for ammonia-dependent
reaction of ASB ................................... .143

3.3 Product inhibition data for glutamine-dependent
reaction of ASB ................................... 144

4.1 Thermodynamic properties for ammonia- and glutamine-
dependent reactions of ASB. ....................... 156















LIST OF FIGURES


1.1 (a) Currently accepted mechanism for the hydrolysis
of L-glutamine to yield ammonia and an acylenzyme 1
by analogy with purF enzyme, GPA. (b) Synthesis of L-
asparagine by reaction of ammonia with activated
aspartyl derivative 2. (c) Hydrolysis reaction to
yield L-glutamate from the acylenzyme 1. ............ 23

1.2 Sequence alignment of the N-terminal domains of E.
coli ASB and human AS as deduced from oligonucleotide
sequencing ......................................... 24

1.3 Proposed mechanism for the synthesis of L-asparagine
by E. coli ASB, via an imide intermediate 3. ........ 25

2.1 Determination of Kd for Aspartate. ................. 45

2.2 Rate of AMP formation as a function of time. ........ 46

3.1A Double-reciprocal plot of initial velocity versus L-
aspartate concentration at various fixed
concentrations of ATP ............................. 96

3.1B Replots of the reciprocal-fixed variable substrate
ATP vs slope (plus) and intercept (square) from
Fig.3.1A ........................................... 97

3.2A Double-reciprocal plot of initial velocity versus L-
glutamine concentration at various concentrations of
ATP ................................................ 98

3.2B Replots of the reciprocal-fixed variable substrate
ATP vs slope (plus) and intercept (cross) from
Fig.3.2A ........................................... 99

3.3A Double-reciprocal plot of initial velocity versus L-
aspartate concentration at various fixed
concentrations of L-glutamine. ..................... 100

3.3B Replots of the reciprocal-fixed variable substrate L-
glutamine vs slope (plus) and intercept (cross) from
Fig. 3.3A ......................................... 101


vii









3.4A Double-reciprocal plot of initial velocity versus ATP
concentration at various fixed concentrations of L-
aspartate ......................................... 102

3.4B Replots of the reciprocal-fixed variable substrate L-
aspartate vs slope (plus) and intercept (cross) from
Fig. 3.4A ......................................... 103

3.5A Double-reciprocal plot of initial velocity versus ATP
concentration at various fixed concentrations of
NH3 ............................................. 104

3.5B Replots of the reciprocal-fixed variable substrate
NH3 vs slope (plus) and intercept (cross) from Fig.
3.5A .............................................. 105

3.6A Double-reciprocal plot of initial velocity versus L-
aspartate concentration at various fixed
concentrations of NH3 ............................. 106

3.6B Replots of the reciprocal-fixed variable substrate
NH3 vs slope (plus) and intercept (cross) from Fig.
3.6A .............................................. 107

3.7 Double-reciprocal plot of initial velocity versus L-
glutamine concentration at fixed concentration of
LGH ............................................... 108

3.8 Double-reciprocal plot of initial velocity versus ATP
concentration at fixed concentration of L-glutamine
(plus) and LGH (cross) (0.2 mM) .................. 109

3.9 Double-reciprocal plot of initial velocity versus L-
aspartate concentration at various fixed
concentrations of ATP. ............................ .110

3.10 Double-reciprocal plot of initial velocity versus L-
glutamine concentration at various concentrations of
ATP ............................................... 111

3.11 The ratio of L-glutamate produced/PPi produced versus
concentration of L-glutamine. ..................... 112

3.12 Double-reciprocal plot of initial velocity versus NH3
concentration at various fixed concentrations of L-
asparagine ........................................ 113

3.13 Double-reciprocal plot of initial velocity versus ATP
concentration at various fixed concentrations of L-
asparagine ........................................ 114


viii









3.14 Double-reciprocal plot of initial velocity versus L-
aspartate concentration at various fixed
concentrations of L-asparagine. ................... 115

3.15 Double-reciprocal plot of initial velocity versus ATP
concentration at various fixed concentrations of
AMP ............................................... 116

3.16 Double-reciprocal plot of initial velocity versus L-
aspartate concentration at various fixed
concentrations of AMP ............................. 117

3.17 Double-reciprocal plot of initial velocity versus NH3
concentration at various fixed concentrations of
AMP. ............................................... 118

3.18 Double-reciprocal plot of initial velocity versus L-
glutamine concentration at various fixed
concentrations of L-asparagine. ................... 119

3.19 Double-reciprocal plot of initial velocity versus ATP
at various fixed concentrations of L-asparagine. ..120

3.20 Double-reciprocal plot of initial velocity versus L-
aspartate concentration at various fixed
concentrations of L-asparagine. ................... 121

3.21 Double-reciprocal plot of initial velocity versus ATP
concentration at various fixed concentrations of
PPi .............................................. 122

3.22 Double-reciprocal plot of initial velocity versus L-
aspartate concentration at various fixed
concentrations of PPi. ............................. 123

3.23 Double-reciprocal plot of initial velocity versus ATP
concentration at various fixed concentrations of L-
glutamate ......................................... 124

3.24 Double-reciprocal plot of initial velocity versus L-
aspartate concentration at various fixed
concentrations of L-glutamate. .................... 125

3.25 Double-reciprocal plot of initial velocity versus L-
glutamine concentration at various fixed
concentrations of L-glutamate. .................... 126

3.26 Double-reciprocal plot of initial velocity versus L-
glutamine concentration at various fixed
concentrations of AMP. ............................ 127









3.27 Double-reciprocal plot of initial velocity versus L-
aspartate concentration at various fixed
concentrations of AMP ............................. 128

3.28 Double-reciprocal plot of initial velocity versus ATP
concentration at various fixed concentrations of
AMP ............................................... 129

3.29 Double-reciprocal plot of initial velocity versus L-
aspartate concentration at various fixed
concentrations of L-glutamine in the presence of L-
glutamate (50 mM) ................................. 130

3.30 Double-reciprocal plot of initial velocity versus L-
aspartate concentration at various fixed
concentrations of L-glutamine in the presence of PPi
(0.4 mM) ......................................... 131

3.31 Double inhibition studies plot of L-asparagine
concentration versus reciprocal initial velocity at
various fixed concentrations of AMP. ............... 132

3.32 Double inhibition studies plot of L-glutamate
concentration versus reciprocal initial velocity at
various fixed concentrations of AMP ............... 133

3.33 Double inhibition studies plot of L-glutamate
concentration versus reciprocal initial velocity at
various fixed concentrations of L-asparagine. ......134

3.34 Double-reciprocal plot of initial velocity versus L-
aspartate concentration at fixed varied
concentrations of L-glutamate and AMP in a constant
ratio ([L-glutamate] = 5 [AMP]) ................... 135

3.35 Double-reciprocal plot of initial velocity versus ATP
concentration at fixed varied concentrations of L-
aspartate ......................................... 136

3.36 Double-reciprocal plot of initial velocity versus ATP
concentration at fixed varied concentrations of L-
aspartate......................................... .137

3.37 Double-reciprocal plot of initial velocity versus ATP
concentration at fixed varied concentrations of L-
aspartate ......................................... 138

3.38 Double-reciprocal plot of initial velocity versus ATP
concentration at fixed varied concentrations of L-
aspartate ......................................... 139









3.39 Double-reciprocal plot of initial velocity versus ATP
concentration at fixed varied concentrations of L-
aspartate ......................................... 140

3.40 Double-reciprocal plot of initial velocity versus ATP
concentration at fixed varied concentrations of L-
aspartate ......................................... 141

3.41 Double-reciprocal plot of initial velocity versus ATP
concentration at fixed varied concentrations of L-
aspartate ......................................... 142

4.1 Arrhenius plot of the Vmax values for the ammonia-
dependent reaction, varying NH3 concentration. .....152

4.2 Arrhenius plot of the Vmax values for the ammonia-
dependent reaction, varying ATP concentration. ....153

4.3 Arrhenius plot of the Vmax values for the ammonia-
dependent reaction, varying L-aspartate
concentration ..................................... 154

4.4 Arrhenius plot of the Vmax values for the glutamine-
dependent reaction, varying L-glutamine
concentration ..................................... 155















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


INVESTIGATION OF THE KINETIC MECHANISM OF GLUTAMINE- AND
AMMONIA-DEPENDENT REACTIONS OF E. coli ASPARAGINE SYNTHETASE
B USING ISOTOPE PARTITIONING AND STEADY-STATE KINETICS

By

POURAN HABIBZADEGAH-TARI

August, 1996




Chairman: Dr. Sheldon M. Schuster
Major Department: Biochemistry and Molecular Biology


The kinetic mechanism of the Escherichia coli asparagine

synthetase B was deduced from initial velocity studies. In

addition to varying substrate concentrations in a variety of

ratios, products and substrate analogs were used as

inhibitors in additional studies. While these studies

provided limitations to the possible kinetic mechanisms, the

results were equivocal. The data were consistent with a ping

pong mechanism with either inorganic pyrophosphate or L-

glutamate released prior to addition of other substrates.

Therefore, from these data, two ter-quad mechanisms, bi-uni-

uni-ter ping-pong and uni-uni-bi-ter ping-pong, were

possible. In order to resolve this dilema, a series of

isotope partitioning studies of both the glutamine- and


xii










ammonia-dependent reactions of ASB were carried out as well.

Major conclusions derived from the isotope partitioning

experiments regarding the kinetic mechanism of ASB are as

follows: 1) The enzyme catalyzes aspartate-dependent ATP

hydrolysis, which requires no nitrogen source. 2) The binding

and hydrolysis of L-glutamine or the binding of ammonia is

not required prior to ATP and L-aspartate binding for the

synthesis reaction. These results together clearly support an

ordered bi-uni-uni-ter ping-pong mechanism for the ammonia-

and the glutamine-dependent reactions of E. coli ASB, with

ATP binding first and L-aspartate second. This is followed by

release of PPi and subsequent addition of L-glutamine or

ammonia.

The information obtained from product inhibition studies

was also consistent with the existence of an isomerization

step following the release of the last product. Another very

important observation made in this work was that the

glutaminase reaction was shown to be occurring at the same

time as the synthetase reaction, and in fact increasing with

increasing concentrations of L-glutamine

These observations resulted in a complete model for the

glutamine-dependent AS reaction, and computer modeling was

used to stimulate the proposed mechanism. According to the

proposed model the reaction mechanism is quite complex, and

is an ordered bi-uni-uni-ter ping-pong mechanism with a side

reaction, glutaminase, for the E. coli ASB enzyme.


xiii















CHAPTER 1
INTRODUCTION

Vauqulien and Robiquot (1806) were the first to report

the isolation of L-asparagine, which was the first amino acid

ever identified. The metabolic importance of this amino acid

has emerged from numerous studies that have appeared

describing both the anabolic and catabolic pathways from

numerous organisms. Early on it was shown that L-asparagine

can be hydrolyzed by L-asparaginase to yield L-aspartate and

ammonia (Broome, 1963), and for many years this was thought

to be the major use of L-asparagine. However, a study by

Meister et al. (1952) had suggested that L-asparagine can

also be degraded via asparagine transamination followed by

amide hydrolysis. Later, Moraga et al. (1989) described the

mode of L-asparagine catabolism via asparagine transaminase

and w-amidase in rat liver mitochondria. In addition,

evidence was presented that suggested the existence of an as

yet uncharacterized pathway for asparagine catabolism in

mitochondria. L-asparagine has also been shown to be involved

in providing the residue necessary for linkage of

oligosaccharides to glycoproteins (Spiro, 1969), so that

possibly the most important fate of L-asparagine is its

incorporation into protein (Coony and Handschumacher, 1970).

The interest in asparagine metabolism was intensified

from the finding by Kidd (1953) that a factor in guinea pig










serum had antitumor activity. Later, Broome related the

antitumor properties of guinea pig serum to its L-

asparaginase activity (1963 and 1968). Since that time, L-

asparaginase, which catalyzes the hydrolysis of L-asparagine,

has been used clinically to treat patients with lymphomas

(Oettgen et al., 1970 and Ertel et al., 1979). When used on

patients with acute lymphoblastic leukemia (ALL), L-

asparaginase resulted in complete remission for 40% of the

patients (Uren et al., 1977). When L-asparaginase was used in

combination with prednisone and vincristine, it resulted in

95% complete remission for ALL previously untreated patients

(Uren et al., 1977). Although L-asparaginase is a potent

chemotherapeutic agent, several side effects have been

reported that would limit its general utility. These side

effects include chills, fever, nausea, life-threatening serum

ammonia concentrations, liver dysfunction (Oettgen et al.,

1970 and Terebello et al., 1986), central nervous dysfunction

(Land et al., 1972),and tumor resistance (Uren et al., 1977).

The effectiveness of L-asparaginase is due to its

ability to lower the circulating level of asparagine (Broome,

1968). This suggests the possibility that a highly specific

and potent inhibitor of the enzyme responsible for synthesis

of L-asparagine, namely asparagine synthetase (AS), might be

effective in treating tumors. A great deal of the early work

on AS was directed at this goal which involved screening of a

broad range of inhibitors. Several hundreds of randomly

selected compounds and a host of available and newly










synthesized substrate analogs have been tested as AS

inhibitors. Few compounds inhibited AS, and those that did,

exhibited weak inhibition. The most promising inhibitor of

asparagine synthetase, O-aspartyl methylamide, increased life

span from 31 to 79 percent when administered to mice bearing

L-asparaginase resistant tumors (Uren et al., 1977). However,

this compound was proven to be a better inhibitor of the L-

asparaginase than of AS. No other compounds have been found

to possess sufficient potency to warrant further study.

Therefore, detailed structural, chemical and mechanistic

information is essential in order to design effective

inhibitors of AS. For example, given the success in the

discovery of drugs based upon transition state analogs

(Barlett and Marlowe, 1983), elucidation of the mechanistic

details underlying the AS reaction mechanism may prove an

alternate approach to obtaining potent AS inhibitors.

Therefore, we have chosen to study in detail the mechanism of

AS.

Asparagine synthetase was demonstrated for the first

time in Lactobacillis arabinosus (Ravel et al., 1962). The

bacterial enzyme was shown to catalyze conversion of L-

aspartate to 0-aspartylhydroxamate in the presence of

hydroxylamine and ATP, and alternatively to L-asparagine in

the presence of ammonia and ATP.


L-Asp + ATP + NH N L-Asn + AMP + PP. (1)
3 1










Along with the adenosine triphosphate (ATP) a divalent metal

ion, either Mn2+ or Mg2+, was required. Adenosine 5'-

phosphate and inorganic pyrophosphate (PPi) were shown to be

products along with L-asparagine. The production of

asparagine synthetase decreased more than 10-fold when the

cells were grown in the presence of L-asparagine. In

addition, the activity in vitro of the enzyme was inhibited

by L-asparagine (Ravel et al., 1962).

Following the partial purification (10-fold), the

initial velocity and other properties were determined. The Km

for ATP, L-aspartate and Mg2+ were 0.2 mM, 4.2 mM and 3.5 mM,

respectively. The pH optimum for the formation of 3-

aspartylhydroxamate was between 6.0 and 6.5, and for the

synthesis of L-asparagine was 8.2. The enzyme was also shown

to catalyze an aspartate-dependent exchange of ATP and PPi

which was inhibited by L-asparagine. L-glutamine did not

serve as a nitrogen source.

Asparagine synthetase from Streptococcus bovis was

studied by Burchall et al. in 1964. The bacterial extract was

shown to catalyze the formation of a hydroxamate of L-

aspartate in the presence of ATP and Mg+2. The AS from

Streptococcus bovis was partially purified from the extract

(21-fold) and characterized. The bacterial enzyme was shown

to catalyze the conversion of L-aspartate to 3-

aspartylhydroxamate in the presence of hydroxylamine, ATP and
Mg+2- Ammonia could also be used to replace hydroxylamine,

forming L-asparagine, but L-glutamine could not. ATP was






5


converted to AMP and PPi, and no ADP was detected. The Km's

for ATP, L-aspartate, NH4+ and Mg2+ were 4 mM, 26 mM and 4 mM

and 45 mM, respectively. Fifty percent of the enzyme activity

was lost when AS was incubated with iodoacetate (10 mM), p-

hydroxymercuribenzoate (10 mM) and silver nitrate (10 mM). L-

asparagine, which inhibited the enzyme activity at a

concentration of 2 mM, was found to be a competitive

inhibitor with respect to L-aspartate. Curiously, however,

the enzyme synthesis was not repressed when bacterial cells

were grown in the presence of L-asparagine.

The presence of AS in E. coli was demonstrated by Ceder

and Schwartz (1969a). AS was purified 370-fold from a mutant

of E.coli that was either deficient in L-asparaginase II or

else its activity was inhibited by 5-diazo-4-oxo-l-norvaline

(DONV). The molecular weight of the enzyme was determined to

be 80,000 by gel filtration and was shown to be stabilized by

2-mercaptoethanol and by 10% glycerol. The asparagine

synthesis reaction required L-aspartate, an ATP-Mg2+ complex,

and ammonia, with stoichiometric production of PPi and AMP.

L-glutamine did not serve as a nitrogen source, but

hydroxylamine could be substituted for ammonia, forming 3-

aspartylhydroxamate instead of L-asparagine. The pH optimum

was found to be 8.4. The enzyme, in the absence of ammonia,

was shown to catalyze an aspartate-dependent exchange of ATP

and PPi, which was inhibited by L-asparagine.

Kinetic studies were performed to obtain the mechanism

of the E. coli AS enzyme (Ceder and Schwartz, 1969b). Initial










velocities of AS activity were determined when the

concentration of two of the three substrates, L-aspartate,

ATP, and ammonia were varied, keeping the third substrate

constant. When either L-aspartate and ammonia, or ATP and

ammonia was varied, a parallel initial velocity pattern was

seen, characteristic of a ping-pong mechanism where a product

(PPi) is released before ammonia binds the enzyme. The

initial velocity pattern was intersecting when L-aspartate

and ATP were the variable substrates, suggesting that these

two substrates are added in a sequential manner to the

enzyme. The enzyme was also shown to catalyze the transfer of
180 from the 0-carbonyl group of aspartate to the phosphate

of AMP, suggesting a 3 aspartyl-adenylate intermediate. The

product inhibition studies indicated that the order of

addition of ATP and L-aspartate is random, which was

confirmed by kinetic studies of the PPi-ATP exchange. Initial

rates of the 32P-PPi exchange reaction were determined with

varying concentrations of ATP and substrate, while

maintaining the concentration of PPi constant. An

intersecting pattern was observed. In both, the point of

intersection was situated to the left of the ordinate. These

results were consistent with the rapid equilibrium random

mechanism, for which a general rate equation was presented.

The mechanism of E. coli AS was reported to be bi-uni-uni-bi

Ping-Pong, with random ATP and L-aspartate addition and

random AMP and L-asparagine release as shown below.








ATP

iy
/E-ATP



E-Asp

ATP A


Asp Asn AMP

A n< AMP\

E-ATP-Asp4-* E-AMP-Asp41- E-AMP-Asn
/------\/


sp


E-Asn /

AMP Asn


Later, Felton et al. (1980) showed that E. coli contains two
genes coding for asparagine synthetase, which were renamed
asnA and asnB. Their studies showed that these genes are
located at two points in the chromosome and that both must be
mutated to produce an auxotroph.
Biochemical and genetic studies were performed by
Humbert and Simoni (1980) to determine if the E. coli genes
were distinct or the if they were products of gene
duplication. Two strains, asnA+ asnB and asnA asnB+, were
constructed, and the asparagine synthetic reaction of their
extracts was characterized. Their studies showed that asnA
gene codes for the enzyme previously characterized by Ceder
and Schwartz (1969). The asnB gene coded for an enzyme (ASB)
that was different from ASA. Dialyzed extracts containing ASB
enzyme had a lower specific activity for asparagine
synthesis. ASB was able to use L-glutamine or ammonia as the
nitrogen source. ASB was also distinguished from ASA by its










greater liability at low temperatures and greater stability at

high temperatures.

The E. coli ASA, an ammonia-dependent AS, is composed of

330 amino acids (Nakamura et al., 1981), while the E.coli

ASB, a glutamine-dependent AS, is composed of 554 amino acids

(Scofield et al., 1990). No similarity can be detected

between the amino acid sequence of E. coil ASA and E. coli

ASB, indicating that although the two AS synthesize the same

product, they have evolved from different sources. A highly

conserved protein motif characteristic of Class II aminoacyl

tRNA synthetase was found to align with a region of E. coli

ASA. Site directed mutagenesis of some of the conserved

residues in the motif resulted in an inactive enzyme,

suggesting the possibility that E. coli ASA evolved from an

ancestral aminoacyl tRNA synthetase (Hinchman et al., 1992).

The E. coli ASB was aligned with the protein sequence of

human asparagine synthetase that can use both glutamine and

ammonia as substrates. Interestingly, the two proteins were

shown to be quite homologous (37%) with regard to their amino

acid sequence. This indicated that the two genes may have

evolved from a common ancestral gene (Scofield et al. 1990).

The significant amount of similarity between the human and E.

coli glutamine-dependent gene products might suggest that

these highly conserved regions are critical for the enzymatic

activity or structure in both enzymes.

Cloning, overexpression and characterization of ASB was

reported by Scofield et al. (1990) and Boehlein et al.










(1994). The recombinant ASB possessed a molecular mass of

about 63 KDa. Addition of chloride (10 mM) to the assay

medium increased glutamine-dependent activity ASB by a factor

of 2 but did not affect the ammonia-dependent reaction.

Magnesium ion (Mg2+) gave the highest activity. Co2+, the

only ion that could replace Mg2'+, supported 80 and 50% of the

Mg2+-dependent ASB activity when either L-glutamine or

ammonia was the nitrogen source, respectively. A number of

nucleotides were also assayed as substrates for ASB. dATP and

ATP were utilized in both reactions at a similar rate,

whereas GTP utilization was only 15% that of ATP. The pH

optima were determined to be 6.5-8 for the glutamine- and

ammonia-dependent activities of ASB.

Asparagine synthetase was also purified and

characterized from Klebsiela aerogenes (Reitzer and

Magasanik, 1982) and Saccharomyces cerevisiae (Ramos and

Wiame, 1979 & 1980). Asparagine synthetase has also been

isolated and characterized from higher organisms such as

chick embryo liver (Arfin, 1967), beef pancreas (Luehr and

Schuster, 1985), and rat liver (Hongo et al., 1978).

Asparagine synthetase from mammalian sources was shown to

catalyze the conversion of L-aspartate to L-asparagine in the

presence of L-glutamine with concomitant cleavage of ATP to

AMP and PPi as shown below.


L-Asp + ATP + L-GIn L-Asn + L-Glu + AMP + PP (2)










Ammonia was also shown to serve as a nitrogen source in the

absence of glutamine. When glutamine is used, a glutaminase

activity is associated with the asparagine synthetase

activity. This glutaminase activity was shown to occur in the


L-GIn + H20 -- L-Glu + NH3 (3)

absence and in the presence of all substrates of the

asparagine synthetase.

Underlying all of these previous studies of AS is the

concept that any potent and specific inhibitor of AS could

become a very useful chemotherapeutic agent. The failure of

previous attempts to inhibit AS specifically is due to a

deficiency of our understanding of the chemical and kinetic

mechanism of the enzyme. Although very little information is

available regarding the chemical mechanism of AS, there is a

general agreement that aspartyl-AMP, an activated form of L-

aspartate, is the reaction intermediate (Luehr and Schuster,

1985, Horowitz and Meister, 1972). Nevertheless, the pathway

by which the amide nitrogen is transferred from the L-

glutamine to this activated complex in AS needs further

characterization.

The ability to use ammonia or L-glutamine as the

nitrogen source is also found in other amidotransferases such

as such as glutamine phosphoribosylpyrophosphate

amidotransferase, GPA, (Mei and Zalkin, 1989), CTP synthetase

(Weng et al., 1986), and GMP synthetase (Zalkin and Truit,










1977). Two types of glutamine amide transfer domains (GAT

domains) have been identified in glutamine amidotransferase

enzymes. The first type shows homology to the GAT domain

derived from the purF gene and is called a purF-type GAT

domain. This amidotransferase subfamily includes glutamine

phosphoribosylpyrophosphate amidotransferase, GPA, (Mei and

Zalkin, 1982), AS (Andrulis et al., 1987 & Scofield et al.,

1990) and glucosamine-6-P synthase (Walker et al., 1984). The

N-terminal four amino acids in these enzymes are highly

conserved. The second type shows homology to the GAT domain

derived from the trpG gene which includes CTP synthetase

(Weng et al., 1986), carbamyl-phosphate synthetase (Piette et

al., 1984) and GMP synthetase (Zalkin and Truit, 1977). A

trpG-type GAT domain is characterized by three internal

regions of conserved amino acids (Mei and Zalkin, 1987).

In 1989, Mei and Zalkin reported the chemical pathway by

which nitrogen is transferred from L-glutamine in related

glutamine-dependent enzymes. In purF-type enzymes, most

notably GPA (Mei and Zalkin, 1989), the N-terminal cysteine

(Cys-1) residue appears critical for the formation of a

covalent glutaminyl intermediate. In this family, the Cys-1

was shown to be critical for glutamine-dependent, but not

ammonia-dependent, activity because covalent modification of

this residue with diazo-oxo-norleucine (DON) resulted only in

the elimination of glutamine-dependent activity. In GPA a

'catalytic triad' composed of Cys-l, His-101 and Asp-29 was

proposed. Mutagenesis of these residues in GPA resulted in










the loss of glutamine-, but not ammonia-dependent activity.

The cysteine-histidine interaction of amidotransferases

resembles that in papain. In papain, His-159 functions as a

general base to increase the nucleophilicity of Cys-25; an

acidic residue is not involved in the proton shuttle (see

Zalkin et al., 1989). Rather, the side chain of Asn-175 is

hydrogen bonded with imidazole N-3 of His-159 to fix its

position, whereas imidazole N-l accepts a proton from Cys-25.

(see Zalkin et al., 1989). In GPA, in analogy to papain, it

was proposed that Cys-1 participates in an amide hydrolysis

reaction to release ammonia, and His and Asp act as general

bases in the reaction. On the basis of the chemical

modification and site directed mutagenesis, it was proposed

that glutamine is converted to an acylenzyme intermediate by

the nucleophilic attack of the Cys-1 thiolate anion on the

primary amide. Such a reaction would release ammonia which,

if protected from protonation, can undergo nucleophilic

reaction.

The human AS, also a member of the "purF" type of

glutamine amide transfer enzyme, is characterized by the

presence of an N-terminal cysteine followed by conserved

glycine and isoleucine residues (See Mei and Zalkin, 1989 and

Richards and Schuster, 1992). Site-specific mutagenesis was

used to replace the N-terminal cysteine (Cys-l) by alanine in

human AS. The mutation resulted in the loss of the glutamine-

dependent AS activity, while the ammonia-dependent activity

remained unaffected (Van Heeke and Schuster, 1989). Because










of the fact that human AS has the conserved residues defining

the proposed catalytic triad, it was suggested (Mei and

Zalkin, 1989) that the hydrolysis of L-glutamine to yield

free ammonia is the basis of the glutamine-dependent activity

in this enzyme (Fig. 1.1). Due to the lack of structural

information concerning the location of the L-glutamine and

ammonia binding pockets in human AS, the support for

enzymatic generation of 'free' ammonia as a reaction

intermediate is circumstantial. Further, the molecular

mechanism by which ammonia is sequestered from solvent and

retained in its unprotonated form remains an open question

(Richards and Schuster, 1992). On the other hand, it is true

that all asparagine synthetases, and other purF enzymes as

well, can utilize ammonia as a nitrogen source in the absence

of L-glutamine, showing that free ammonia not only binds

within the enzyme active site, but also remains sufficiently

nucleophilic to release L-asparagine from aspartyl-AMP

(Richards and Schuster, 1992). By contrast, the E.coli ASB

lacks the conserved histidine residue, necessary for the

nitrogen transfer if the reaction proceeds by the accepted

pathway in other glutamine amidotranferases (Fig. 1.2), but

still retains the ability to synthesize L-asparagine and

exhibits similar specificity to human AS (Richards and

Schuster, 1992). Based on these findings it was suggested

that at least for the ASB enzyme, the proposed mechanism for

nitrogen transfer in the glutamine-dependent synthesis of

asparagine, may not occur (Richards and Schuster, 1992). This










led to an alternative chemical pathway for the nitrogen

transfer reaction, proposed by Richards and Schuster (1992),

which does not require the generation of free ammonia (Fig.

1.3). According to this mechanism L-glutamine reacts directly

with aspartyl-AMP to form an imide intermediate. Cys-1 is

still essential for the glutamine-dependent activity because

hydrolysis of the imide is required to produce L-asparagine

and the acylenzyme derivative of L-glutamate. One of the

important features of the intermediate is that the nitrogen

retains amide character and therefore possesses little

basicity. The imide can interact with the protein through the

same functional groups used for interaction with L-glutamine

and L-aspartate. In addition, the anionic form of L-

asparagine is the leaving group which relieves the need for

general acid catalysis by a histidine residue. Finally, such

a reaction mechanism eliminates the possibility of diffusion

of ammonia from the active site during the reaction and

removes the need for additional structural features for

maintaining ammonia in its unprotonated form (Richards and

Schuster, 1992). The two catalytic mechanisms described here

are fundamentally different, and efforts have been underway

to address the differences posed by the two proposed

mechanisms.

Studying the kinetic mechanism, which is the focus of

this dissertation, would allow us to draw certain conclusions

about the order of the binding of the substrates and release

of products. For example, if L-glutamate release occurs prior










to the binding of either ATP or L-aspartate, it would be very

difficult to justify the mechanistic hypothesis involving the

imide intermediate (Richards and Schuster, 1992). Based on

their proposal for the nitrogen transfer mechanism, L-

glutamine reacts directly with aspartyl-AMP to form an

asymmetric imide intermediate. Therefore, L-glutamate release

can not happen prior to ATP and L-aspartate binding. The

kinetic mechanism of AS has been proposed for several forms

of AS: e.g., ASA from E. coli (Ceder and Schwartz,

1969,1969), 6C3HED-RG1 tumor (Chou, 1970), mouse pancrease

(Milman et al., 1980), beef pancrease (Markin et al., 1981),

and rat liver (Hongo and Sato, 1985), and as discussed below,

considerable controversy exists.

On the basis of initial velocity and product inhibition

studies, Chou (1970) suggested that the reaction catalyzed by

AS from 6C3HED-RG1 tumor is ping-pong. He observed double

reciprocal plots which showed parallel lines when either L-

glutamine was varied versus ATP or when L-glutamine was

varied versus L-aspartate. Double reciprocal plots of

parallel lines were also observed when L-aspartate was varied

versus ATP. Inhibition studies showed L-asparagine was a

competitive inhibitor with respect to L-glutamine, and PPi

was a competitive inhibitor with respect to ATP. Based on

these observations, the following penta-uni-bi ping-pong

mechanism was proposed for the 6C3HED-RG1 tumor line AS.









Gin Glu ATP ppi Asp AMP Asn

t ,t t E





Using steady-state kinetic methods, Milman et al. (1980)
proposed a uni-uni-bi-ter ping-pong Theorell-Chance mechanism

for the glutamine-dependent reaction of mouse AS. Initial

velocity and product inhibition studies were conducted with

the glutamine-dependent reaction of the AS from mouse
pancreas. Parallel double reciprocal plots were observed with
L-glutamine versus either L-aspartate or ATP, while
intersecting patterns were observed with L-aspartate versus
ATP. These patterns were reported to be indicative of a

hybrid ping-pong mechanism consisting of a glutaminase
partial reaction and a sequential catalysis involving L-

aspartate and ATP. Product inhibition studies involving the
four products, L-glutamate, AMP, PPi and L-asparagine, were
carried out to delineate further the kinetic mechanism. The

patterns from these experiments were reported to be
consistent with a hybrid uni-uni-bi-ter ping-pong Theroll-

Chance mechanism where the glutaminase reaction occurs first.
In other words, L-glutamine binds first followed by the









Gin Glu Asp ATP pi AMP Asn


E-NH3
Asn AP






release of L-glutamate. L-aspartate is the second substrate

to bind followed by a Theorell-Chance step, (ATP on/PPi off).

AMP and L-asparagine are subsequently released in an ordered

fashion. However, there was disagreement between the

predicted patterns and the experimental results for AMP

versus ATP and L-asparagine versus all three substrates.

These discrepancies were rationalized to suggest the

formation E.NH3.Asn and E.NH3.AMP abortive complexes.

The kinetic mechanism of bovine pancreatic AS was

deduced from initial velocity studies and product inhibition

studies, of both the ammonia- and glutamine-dependent

reactions (Markin et al., 1981). For the glutamine-dependent

reaction, parallel lines were observed in the double

reciprocal plots of 1/V versus 1/[L-glutamine] at varied L-

aspartate concentrations, and in the plot of 1/V versus

I/[ATP] at varied L-aspartate concentrations. Intersecting

lines were found for the plot of 1/V versus I/[ATP] at varied

glutamine concentrations. For further clarification of the

order of substrate addition and product release, the product

inhibition of the initial velocity, including a dual










inhibition study, was done for the glutamine-dependent AS.

The results supported an ordered bi-uni-uni-ter ping-pong

mechanism. According to the proposed mechanism, L-glutamine

and ATP sequentially bind followed by the release of L-

glutamate and the addition of L-aspartate and the release of

PPi, AMP, and L-asparagine. The mechanism was found to be

significantly different for the ammonia-dependent reaction.

NH3 bound first followed by a random addition of ATP and L-

aspartate. PPi, AMP, and L-asparagine are then sequentially

released. From these studies, a comprehensive mechanism has

been proposed through which either glutamine or NH3 can

provide nitrogen for L-asparagine production from L-aspartate

as shown below.




Glu and NH3















The kinetic mechanism of rat liver AS was studied by Hongo
Giand Sato (1985). The initial velocity studies, with L-









glutamine as a varied substrate and L-aspartate as a varied
t ^AT P G~u


E ATF^-- PPI AMP Asn
\ H ^ __ E-ATP-NK,_______+_____+f
\J/ E-ATP-Asp-NH 3- E-Asn-AMP-PP.
\Asp ATP






The kinetic mechanism of rat liver AS was studied by Hongo

and Sato (1985). The initial velocity studies, with L-

glutamine as a varied substrate and L-aspartate as a varied









changing fixed substrate, showed a series of parallel lines
in the form of a double reciprocal plot. A parallel double
reciprocal plot was also observed between L-glutamine and
ATP. On the other hand, intersecting lines were obtained

between L-aspartate and ATP. Product inhibition studies were
also conducted. These studies were performed with a high
concentration of Mg2+. The mechanism of the reaction was
suggested to be uni-uni-bi-ter ping-pong Theorell-Chance.
According to their mechanism, L-glutamine binds first

followed by L-glutamate release, and L-aspartate and ATP bind
in an ordered manner followed by ordered release of PPi, AMP,

L-asparagine.




Gin Glu Asp ATP PPi AMP Asn

E t \E t t




When the Mg2+ concentration was kept 0.5-2.0 mM over the ATP
concentration, the binding of substrates after the release of
L-glutamate was in rapid equilibrium with ordered Mg2+ and
random L-aspartate-ATP.









AATP

Gin Glu Ma 7 --A ppi Mg, AMP, Asn

E E




MgATP Asp





Clearly there are dramatic differences in the mechanisms

presented, and the sources of these differences are unclear.

The differences regarding the order of substrate binding and

product release could be due to the different enzyme sources

used. Another potential cause for differing results may be

the presence of certain contaminants. While it is accepted

that the degree of enzymatic purity will not alter a kinetic

mechanism, the presence of any contaminant that breaks down

the substrates or the measured product could have

substantially contributed to the discrepancies. This is

especially important because the presence of a contaminating

glutaminase or asparaginase activity could potentially alter

the observations. Further, the differences between the

results could be due to the single experimental approach

utilized. All studies in the past relied solely on steady

state kinetic methods, and no other technique, such as

isotope partitioning, was employed.










The kinetic mechanism of E. coli ASB, for both the L-

glutamine-dependent and ammonia-dependent reactions, using an

overexpressed, stable and pure enzyme has been investigated.

Although the techniques employed are basically those used in

the past, the enzyme is far more pure, and highly stable.

These studies provide us with information regarding the

kinetic mechanism and relevant rate constants. In addition,

the ability to use techniques such as isotope partitioning

(trapping), pioneered by Meister and Rose and their co-

workers (Krishnaswamy et al., 1962; Rose et al., 1974),

allows us to address problems such as substrate inhibition

and/or abortive complex formation observed with steady state

analyses, and will provide us with information about

catalytic competency of the enzyme-substrate complexes. The

data presented will show that no decision could be made on

any one mechanism proposed for glutamine-dependent reaction

of ASB, if we were to rely only on steady state kinetics. But

together with the help of isotope partitioning experiments,

support is provided for an ordered bi-uni-uni-ter ping-pong

mechanism for the glutamine-dependent reaction of E. coli

ASB. ATP binds first and L-aspartate second. This is followed

by release of PPi and subsequent addition of L-glutamine. The

L-glutamate release is not required prior to the binding of

ATP and L-aspartate. However, further studies suggest that

the simple ordered bi-uni-uni-ter ping-pong mechanism is not

applicable to E. coli ASB enzyme. From these data a

comprehensive mechanism was proposed for the glutamine-






22



dependent AS reaction of ASB, which was examined through

computer modeling.












I"~ H
NN,

0*
001)
AWn


%Nf LC l N Hi,
.4. ^CA 41 r
SC0- N


|j Asp


0
S
H"3N"I LCAlr N H'
C02" N
0 H
COj- N-

0V
Aift


(b) 0 0 o
0 NHI
HN CO, HO OH H3N C2O-
2


(C) 0
S
HN K LCys. 1N
(IN C fl 1:r~02
CO2- N
O" H
0P^-
Alp"


0
H N
+ H20 2< I
CO_. S" N,
H transfer J H
Cysi O"

Asp3


Fig. 1.1 (a) Currently accepted mechanism for the hydrolysis
of L-glutamine to yield ammonia and an acylenzyme 1 by
analogy with purF enzyme, GPA. (b) Synthesis of L-asparagine
by reaction of ammonia with activated aspartyl derivative 2.
(c) Hydrolysis reaction to yield L-glutamate from the
acylenzyme 1. Residue numbering to that of human AS. This
diagram was taken from Richards and Schuster, 1992).


+ AMP















0 1 10 20
AsnB M CjSi F G VJD I K TI5A V E L A K KA L E L SR L
1HAm U G s W AL G S D C LSVQ C LS AMK I A-
30 40
AmB mRRGPDSG IYASD - -[iN A I LAH E nir
HA -- G P 2 AF R r ZN VN G C C FGF HR
s0 60 70
AsB SI F V NAJAj[0P LY N 0 KT HVLAV fNGE
HAm A V VJD P L F MJQ J I RVKKYP Y LWLCYN G E
so 90 1Wo
AsnB I Y M H 0 A L R A E Y G D [~ R j DY 0 F3] IG S Z f V Z
HAM 1 Y N H1 K K M Q H F -E F EVY - KQ 1 :|
110 120
AanB AR~I E=_pirLS S ?-FT ~ "~
HAmD K G[E T I C M L Z 3 7 F v 7 T A

130 140 150
AsnB K DA Y L I G RD H LrG I I UPL Y M G Y DEEi RUG Q0 Y
HAm N K K V F L a S TdFA
160 170 180
AaAB pl Mp]A rL ? V : R T : K ZmP A3S-YDW S
HABI C S JA K G L K H S A T ?L K V G P F p P




Fig. 1.2 Sequence alignment of the N-terminal domains of E.
coli ASB and human AS as deduced from oligonucleotide
sequencing. Identical residues are boxed. The conserved
residues in members of the purF family, (Cys-1, Asp-29 and
His-102) in human AS are underlined. In the mature form of
both human AS and ASB, the N-terminal methionine is absent
and this is therefore numbered as zero. Amino acids are
denoted using the standards one-letter code. This was taken
from Richards and Schuster, 1992.
















S OM 0* c0H .
C oS 0
0)

AIP3 (wS Aq)33


0 0 0*
S S N
__,1 A'Th Cri1 Asp,3

0H-0
H3N H s*+n H20 0 Cys, H i- C02




H lransfr nP
.H*


H

112N

C02i' S'

ASP"



Fig. 1.3 Proposed mechanism for the synthesis of L-
asparagine by E. coli ASB, via an imide intermediate 3.
attack of the primary amide occurs directly upon activated L-
aspartate 2. Cys-1 and asp-33 are then involved in the
hydrolysis of the imide to yield L-glutamate and L-asparagine
in subsequent steps, residues numbering corresponds to that
of the Asn B gene product. This was taken from Richards and
Schuster, 1992.















CHAPTER 2
INVESTIGATION OF THE MECHANISM OF E. COLI ASPARAGINE
SYNTHETASE USING ISOTOPE PARTITIONING


Introduction


Asparagine synthetase B from E. coli catalyzes the

following reactions.

L-Asp + ATP + NH 3 s L-Asn + AMP + PP. (1)
3 '

L-Asp + ATP + L-GIn L-Asn + L-Glu + AMP + PP1 (2)


L-GIn + 1-20 0 L-Glu + NH3 (3)

The E. coli ASA which catalyzes strictly the ammonia-

dependent synthesis of L-asparagine is the most extensively

characterized in terms of kinetic and chemical mechanism. The

order of substrate addition and product release has been

determined, and evidence for the existence of an aspartyl-AMP

intermediate has also been provided (Ceder and Schwartz,

1969a & 1969b).

Studies on the mammalian AS also indicated that the

enzyme produces an aspartyl-AMP intermediate (Luehr and

Schuster, 1985, Horowitz and Meister, 1972). Using steady

state kinetics, the kinetic mechanism of the glutamine-

dependent synthesis of AS was deduced for enzymes from

different mammalian tissues (Chou, 1970, Milman et al., 1980,










Markin et al., 1981 and Hongo and Sato, 1985). However, there

were disagreements in the conclusions presented as discussed

in chapter 1. The differences regarding the order of

substrate binding and product release could be due to

different enzyme sources used. In addition, none of the

studies used a pure enzyme preparation. Although it is

generally accepted that the degree of enzymatic purity will

not alter a kinetic mechanism, the presence of any

contaminants that break down the substrates or the measured

product could have contributed to the variety of results

obtained. This is especially a potential problem with

asparaginase and glutaminase activities. The differences

between the studies could also be due to a limited

experimental approach, relying solely on steady state kinetic

methods. No other technique was used. For example, the

isotope trapping method, isotope exchange and pre-steady-

state kinetic methods were not used.

The kinetic mechanism of E. coli ASB, a member of the

purF family of glutamine-dependent amidotransferase, has not

been studied. Cloning, overexpression and characterization of

ASB was performed by Scofield et al. (1990) and Boehlein et

al. (1994). The overexpression of the enzyme allows us to

obtain a large quantity of highly pure and stable enzyme

which in turn offers a unique opportunity to determine

accurately the reaction mechanism of ASB.

This chapter describes isotope partitioning experiments

that were performed to obtain information on the mechanism of










ASB. The isotope trapping methodology has been used in

studies of glutamine synthetase (Meister et al., 1962),

argininosuccinate synthetase (Rochovansky and Ratner, 1967)

phosphofructokinase (Uyeda, 1970), hexokinase (Rose et al.,

1974, 1979,1981) and CTP synthetase (Lewis and Villafranca,

1989). Meister and his co-workers used the technique to show

if an activated form of L-glutamate (glutamyl-P) was an

intermediate in glutamine synthetase. Other investigators

used the technique to determine the order of addition of

substrates (Rochovansky and Ratner, 1967, Uyeda, 1970 and

Lewis, Rose et al., 1974, 1979, 1981 and Villafranca, 1989).

In this technique, enzyme and one radioactive substrate are

incubated (pulse), followed by the simultaneous addition of a

solution (chase) containing any other substrates necessary

for reaction, as well as a large excess of unlabelled

substrate to dilute the radioactive substrate with the

carrier substrate, therefore diminishing the effect of

additional catalytic turnovers. The reaction is also

terminated as soon as possible after mixing to avoid

utilization of free labelled substrate. The formation of a

labelled product establishes the catalytic competency of the

initial enzyme-substrate complex. We applied the isotope

partitioning technique to determine the sequence of addition

of substrates for both the glutamine- and ammonia-dependent

reactions of AS. The technique was also employed to obtain

the Kd for L-aspartate and ATP.











Materials and Methods



Chemicals and Reagents


L-[14C(U)]Aspartate (224.8 mCi/minmol) was purchased from

NEN Research Products (Boston, MA). [2,8-3H] Adenosine 5'-

Triphosphate (30.0 Ci/mmol) was bought from Amersham Life

Science (England). Scintillation Liquid ScintiVersTm II and

trichloroacetic acid (TCA) were purchased from Fisher

Scientific (Orlando, FL). The ion exchange mono Q columns

were obtained from Bio-Rad. The pyrophosphate reagent,

containing fructose-6-phosphate kinase pyrophosphate

dependent (PPi-PFK), aldolase, triosephosphate isomerase

(TPI), and glycerophosphate dehydrogenase (GDH), for

following PPi production, MgCl2, ATP, L-aspartate, L-

asparagine, L-glutamine, AMP, ammonium acetate, ninhydrin,

ethylenediaminetetraacetic acid (EDTA), Tris(hydroxymethyl)

aminomethane (Tris-HCl), Bis(2-hydroxylethyl)iminotris

(hydroxymethyl)methane (Bis Tris), isopropyl-l-thio-f-D-

galactopyranoside (IPTG), and glycerol were all purchased

from Sigma. DE-81 anion-exchange chromatography paper was

supplied by Whatman (Hillsboro, OR). Dithiothreitol (DTT) was

obtained from Promega Corporation (Madison, Wisconsin).











Expression of the Protein and Purification


An overexpression vector for E. coli ASB, pET-B, has

been constructed by Hinchman and Schuster (1994). The E. coli

B strain, BL21DE3plys S (F-, ompT, rb-, mb-) was transformed

with pET-B plasmid. Protein expression, cell culture and

enzyme purification was carried out as described before

(Boehlein et al., 1994). Transformed cells were plated onto

Luria Broth agar plates supplemented with bacto-tryptone (10

g/Liter), bacto yeast extract (5 g/Liter), NaCl (5 g/Liter),

pH 7.0, ampicillin (100 gg/ml), and chloroamphenicol (30

gg/ml). Plates were incubated at 37C overnight. Single

colonies were used to inoculate fresh minimal media

supplemented with tryptone (10 g/ liter), ampicillin (100 gg/

ml), chloroamphenical (30 gg/ml), and D-glucose (0.75%). The

cultures were grown in an environmental shaker at 370C. When

they reached an absorbance (X=600 nm) of 0.7-1.0, the

cultures were induced by adding IPTG to a final concentration

of 1 mM. Cells were harvested after 2.5-hours by

centrifugation at 15,000 rpm for 5 min in a Beckman Model J2-

21 centrifuge. The supernatant fluid was discarded and the

pellets were stored at -700C until needed. Cells were lysed

by vortexing the pellets in enzyme buffer (50 mM Bis-Tris pH

6.5, 1 mM DTT, 0.5 mM EDTA, and 10% glycerol) using one-tenth

of the original cell culture volume. DNase was added and

cells were left on ice for 0.5 hr. The cell debris was










removed by centrifugation at 15,000 rpm for 20 min in Beckman

centrifuge. The soluble cell extract containing L-asparagine

synthetase B was purified by ion-exchange chromatography, as

described previously (Boehlein et al., 1994). Glycerol was

then added to a final concentration of 10%. The activity of

L-asparagine synthetase B was monitored

spectrophotometrically at 340 run according to the following

coupled reactions, developed by O'Brian (1976). Two moles of

NADH are oxidized to NAD per mole of pyrophosphate produced.

PPi + F-6-P PP'.PFK 0 F-l,6-DP + Pi

F-1,6-DP Ala GAP + DHAP

GAP T- DHAP
2DHAP + 2 O-NADH + 2 H+ L 2-glycerol-3-phosphate + 2 O-NAD+

The standard conditions for assay were 50 mM Tris-HCl (pH

8.0), 20 mM L-glutamine, 10 mM aspartic acid, 10 mM ATP, and
17 mM MgCl2. The soluble cell extract was stored at -700C

until needed.


Protein Concentration Determination


Protein concentration was measured using Bio-Rad Protein

Assay (Bradford, 1976). Mouse immunoglobulin G was used to

obtain a standard curve.


Isotope Partitioning Experiments with Radioactive L-
AsDartate. Glutamine- and Ammonia-Dependent Reactions

A 105-pl solution of 50 mM Tris-HCl (pH 8.0), containing

0.50 mM L-[14C(UJ)I]aspartate (800 cpm/nmole), 2 mM MgCl2, and










1 nmole of ASB was incubated at 37C for three minutes after

which 62 gl chase solution was added such that the final

concentrations of substrates were: 50 mM Tris-HCl, pH 8.0,

5.0 mM ATP, 8.0 mM MgCI2, 10 mM L-glutamine, and 30 mM

unlabelled L-aspartate. The mixture was rapidly mixed, using

vortex, for 3 sec. after which it was quenched by addition of

20 gi of 4 M TCA. A control was done in a similar manner

except that TCA was added prior to incubation to account for

the background from labelled L-aspartate. A blank was also

done in which labelled L-aspartate was added to the chase

solution.

In experiments that included both ATP and labelled L-

aspartate in the pulse solution, the concentration of ATP and

MgCl2 in the pulse were 2.0 mM and 3.0 mM, respectively.

In experiments that included both L-glutamine and

labelled L-aspartate in the pulse solution, the concentration

of L-glutamine in the pulse was 3.0 mM.

In experiments that included both NH3 and labelled L-

aspartate in the pulse solution, the concentration of NH3 in

the pulse was 100 mM. The final concentration of NH3

following the addition of the chase solution was 165 mM.

After quenching, the reactions were neutralized by
addition of 10 gl of 3 M Tris buffer (pH not adjusted) and

centrifuged for five minutes to remove precipitated proteins.

It was determined that this brought pH to 6.5. Aliquots (5
gi) of the reaction mixtures were spotted on DE-81 anion

exchange chromatography paper (28 cm wide and 21 cm long). A










mixture of unlabelled L-aspartate and L-asparagine was also

loaded on both edges of the paper to serve as standards. When

the spots are air dried, the paper were placed vertically in

a chromatography tank containing distilled water as eluant.

The chromatography was stopped after 8 hours, and the paper

air dried. The edges, where standards were spotted, were

excised and sites of L-aspartate and L-asparagine were

determined by spraying with a solution of 0.5% ninhydrin in

absolute ethanol. The bands of radioactive L-aspartate and L-

asparagine, located according to the unlabelled standards,

were excised and put separately into vials with ScintiVersTM

II fluid. The radioactivity and quantity of L-aspartate and

L-asparagine were determined by Beckman 60001C scintillation

counter. The total radioactivity incorporated into L-

asparagine were normalized to the same amount of enzyme used.

The specific activity of the labelled L-aspartate (800

cpm/nmole) was determined after taking into account dilution

factor and the percent of quenching associated with 14C

(25%), using paper chromatography.

All experiments were done in triplicate, and the data

collected were evaluated by taking average of the samples

from which the background (blank, labelled L-aspartate added

to the chase) was subtracted.











Determination of Kd for L-AsDartate


In experiments to determine the Kd for L-aspartate, the

following protocol was used. A 160 li solution of 50 mM

Tris-HCl, (pH 8.0), containing 10.0 mM ATP, 12.0 mM MgCl2, 2

nmole ASB, and labelled L-aspartate (0.15-0.53 mM) was

incubated at 37C for two minutes. A 65 p.1 chase solution was

added such that the final concentrations of substrates were:

50 mM Tris-HCl, pH 8.0, 12.0 mM ATP, 15.0 mM MgCl2, 10 mM L-

glutamine, and 30 mM unlabelled L-aspartate. The

radioactivity and quantity of L-aspartate and L-asparagine

were determined as before.


Isotope Partitioning Experiments with Radioactive ATP.
Glutamine- and Ammonia-DeDendent Reactions


A 120-gl solution of 50 mM Tris-HCl (pH 8.0), containing

0.5 mM [2,8-3H]ATP (1300 cpm/nmole), 2 mM MgCl2, and 1 nmole

of ASB was incubated at 370C for three minutes after which 65

p.1 chase solution was added such that the final

concentrations of substrates were: 50 mM Tris-HCl, pH 8.0, 30

mM ATP ,31 mM MgCl2, 10 mM L-glutamine, and 10 mM unlabelled

L-aspartate was added. The mixture was rapidly mixed by

vortex for 3 sec. after which it was quenched by addition of

20 pI of 4 M TCA. A control was done in which L-aspartate was

omitted from the chase solution to account for other

contaminating ATPase activities.










In experiments that included both L-aspartate and

labelled ATP in the pulse solution, the concentration of L-

aspartate in the pulse was 2.0 mM.

In experiments that included both L-glutamine and

labelled ATP in the pulse solution, the concentration of L-

glutamine in the pulse was 3.0 mM.

In experiments that included both NH3 and labelled ATP

in the pulse solution, the concentration of NH3 in the pulse

was 100 mM. The final concentration of NH3 following the

addition of the chase solution was 165 mM.

Following the quenching, the reactions were centrifuged

for 5 min. Aliquots (5 gi) of the reaction mixtures were

spotted on DE-81 anion-exchange chromatography paper. A

mixture of unlabelled ATP, ADP and AMP was also loaded on

both edges of the paper to serve as standards. After drying,

the paper was developed using 1/50 saturated ammonium acetate

(pH 2.8) as eluant. The chromatography was stopped after 3

hours, and the paper air dried, and unlabelled nucleotides

were visualized on the paper by UV absorbance at 254 nm. The

bands of radioactive ADP, ATP and AMP, located according to

the unlabelled standards, were cut, put into vials with

ScintiVersTm II fluid. The radioactivity and quantity of AMP,

ADP and ATP were determined by Beckman 60001C scintillation

counter. The total radioactivity associated with AMP were

normalized to the same amount of enzyme used. The specific

activity of the labelled ATP (1300 cpm/nmole) was determined

after taking into account the dilution factor and the percent










of quenching associated with tritium (90%), using paper

chromatography.

All studies were done in triplicate, and the data

collected were evaluated by taking average of the samples

from which the background was subtracted.


AsDartate-DeDendent ATP Hydrolysis


The assay mixture (in all experiments, the reaction

volume was 100 gl) was consisted of the following: 50 mM

Tris-HCl (pH 8.0), 2 mM L-aspartate, 2 mM MgCl2, 0.5 mM [2,8-

3H]ATP (600 cpm/ nmole) and 0.5 nmole ASB. Assay mixtures

were incubated at 37C, and reactions were terminated by

addition of 25 p1 of 4 M TCA at the indicated times. A

control was done in a similar manner except the TCA was added

prior to the incubation to account for the background from

labelled ATP. A second control was done in which L-aspartate

was omitted from the reaction mixture to account for other

contaminating ATPase activity. The specific activity of

labelled ATP was determined as described above.

Two other experiments were performed in the same manner

except inorganic pyrophosphatase and/or pyrophosphate reagent

was added to the reaction mixtures (the reaction volume was

100 gi). Assay mixtures were treated as described before.

The ATP Km for this reaction was obtained under the
following conditions. The assay mixtures (100 il) contained

50 mM tris-HCl (pH 8.0), 1 mM L-aspartate, 2 mM MgCl2 and










varying concentration of ATP (0.005-0.4 mM) (30,000

cpm/reaction). The assay mixtures were incubated 37C for 3

min, and reactions were terminated by addition of 15 gi of 4

M TCA at the indicated times. The control was done in which

L-aspartate was removed from the reaction mixture to account

for other contaminating ATPase activity.

Following the quenching, the reactions were treated as

described before, and radioactivity and quantity of AMP, ADP

and ATP were determined by Beckman 60001C scintillation

counter. The total counts (30,000 cpm per reaction) was

calculated by taking into account the dilution factor and the

percent of quenching associated with tritium (90%), using

paper chromatography.


Theory


Isotope trapping was used to obtain information about the

order of addition of substrates for both the glutamine- and

ammonia-dependent reactions of ASB. The following scheme

illustrates the experimental procedure used. Enzyme and

labelled substrate (*A) are incubated (pulse). This is

followed by rapid dilution with a chase solution, containing

a large excess of unlabelled substrate A and of the

complementary reactants, B and C, which are allowed to react

and stopped by a denaturant (acid or base) (Rose, 1980).








p*
BC
EA* 40 EA*BC A


I i p
E+ A* EBC + A*




The labelled substrate has three fates; it can dissociate
from the EA* complex or it can dissociate from EA*BC complex
or it can incorporate into the product. The ability to trap
A* as P* indicates that EA is formed in a catalytically
competent manner and that A can bind first. If, however,
product formation requires that B and/or C binds the enzyme
before A, as in the case of an ordered mechanism, no labelled
product will form. The trapping of the labelled substrate as
product also requires that dissociation of A* from any of the
complexes,(E.A*), or (E.A*BC) be slow enough that a
measurable quantity of the labelled substrate can proceed
toward product formation. Therefore, the failure to trap A*
as P* could be due to the following: the E.A complex may not
be catalytically competent (A must bind after B and C), A may
dissociate from E.A complex, or it may dissociate from E.ABC
complex (Rose, 1980).















IsotoDe Partitioning Experiments with Radioactive L-
Aspartate. Glutamine- and Ammonia-Deuendent Reactions


Table 2.1 summarizes the results for both the glutamine-

and ammonia-dependent AS reactions. A solution containing 1

nmole of AS, 0.50 mM 14C-L-aspartate and 2 mM MgCl2 was

incubated for two minutes. Then a chase solution, containing

a 60-fold excess of unlabelled L-aspartate and saturating

concentrations of ATP, L-glutamine, and MgCl2, was added.

This was followed by quenching and product analysis. When

radioactive L-aspartate was the only substrate in the pulse,

very little radioactive L-aspartate (0.10 0.03 nmole), for

every nmole of AS, was found trapped as L-asparagine. When

the above experiment was modified to include ATP in the

pulse, for every nmole of AS enzyme, about 0.90 0.03 nmole

of L-aspartate was trapped as labelled L-asparagine. When L-

glutamine was included in the pulse with radioactive L-

aspartate, about 0.12 + 0.06 nmole of L-aspartate was trapped

as labelled L-asparagine.

The isotope partitioning experiment with labelled L-

aspartate was also performed for the ammonia-dependent

reaction of AS. In the case where NH3 (100 mM) was used in

the pulse with radioactive L-aspartate, the chase solution

was as described above except NH3 was substituted for L-

glutamine to a final concentration of 165 mM. L-aspartate










(0.15 0.10 nmole) was trapped as radiolabelled L-

asparagine, for every nmole of AS. When the above experiment

was modified to include ATP in the pulse, about 0.70 + 0.10

nmole of L-aspartate was trapped as labelled L-asparagine.

Using the isotope trapping methods, the Kd for L-

aspartate was found to be 0.065 mM ( 0.02). The Kd was

obtained from a double reciprocal plot of 1/[nmole of trapped

L-asparagine] versus 1/L-aspartate (Fig. 2.1). The fact that

reactivity is measured during a single catalytic turnover,

therefore measuring the binding of L-aspartate (labelled),

verifies that this is the Kd.


Isotope Partitioning Experiments with Radioactive ATP.
Glutamine- and Ammonia-Dependent Reactions


Table 2.2 summarizes the results for both the glutamine-

and ammonia-dependent AS reactions. A solution containing 1

nmole of AS, 0.50 mM labelled ATP and 2 mM MgCl2 was

incubated for three minutes. Then a chase solution was added

to it, containing a 60-fold excess of unlabelled ATP and

saturating concentrations of L-aspartate, L-glutamine, and

MgCl2, followed by quenching and product analysis. When

radioactive ATP was the only substrate in the pulse, 0.43

0.10 nmole of ATP was found trapped as AMP, for every nmole

of AS. When the above experiment was modified to include L-

aspartate (2 mM) in the pulse, about 2.2 0.25 nmole of ATP

was trapped as AMP. When L-glutamine (3.0 mM) was included in










the pulse with radioactive ATP about 0.92 0.06 nmole of ATP

was trapped as AMP.

In the case where NH3 (100 mM) was used in the pulse

with radioactive ATP, the chase solution was as described

above, except 65 mM NH3 was substituted for L-glutamine, 0.2

+ 0.07 nmole of ATP was trapped as AMP, for every nmole of

AS. The amount of ATP trapped in this experiment (0.2 nmole)

is half of the amount trapped when labelled ATP was the only

substrate in the pulse, for glutamine-dependent reaction

(0.43 nmole).


Asoartate-Dependent ATP Hydrolysis


In the isotope partitioning experiment where both L-

aspartate and labelled ATP were included in the pulse, for

every nmole of AS enzyme, 2.2 0.25 nmole of labelled ATP

was trapped as labelled AMP but theoretically no more than 1

nmole of ATP should be trapped as AMP. This suggests that ATP

hydrolysis may be occurring in the pulse, prior to the

addition of chase solution that contains all the substrates.

If this is the case, it should be possible to detect labelled

ATP as labelled AMP in the absence of any nitrogen source.

The following experiments were performed using labelled ATP,

as described under Materials and Methods, and formation of

labelled AMP was measured as a function of time in the

absence of any nitrogen source. The rate of AMP formation was

0.3 nmole/min, which was linear with time for 15 min (Fig.










2.1). The rate of AMP formation was increased when

pyrophosphate reagent or inorganic pyrophosphatase was

present (3-4 times). No ATP hydrolysis (AMP formation) was

observed in the control where L-aspartate was omitted from

the reaction mixture (data not shown), suggesting that ATP

hydrolysis is dependent on the presence of L-aspartate. It

was possible that the ATP hydrolysis, in the presence of L-

aspartate, was due to contaminating ammonia present in the

reaction mixture, allowing the synthesis reaction (ammonia-

dependent) to occur. To account for any possible

contaminating NH3 that could contribute to the ATP hydrolysis

(AMP formation), the above experiment was carried out (for 40

minutes), using unlabelled ATP, and formation of L-asparagine

was measured as a function of time using HPLC amino acid

analysis. Amino acid analysis is performed on an applied

Biosystems 130 A separation system (Schuster et al., 1993).

No L-asparagine was detected even after 15 minutes. Very

little L-asparagine (0.03 nmole/min) was detected after 20

minutes of incubation, suggesting that the contaminating NH3

is responsible for the formation of only 1/10 of the AMP. If

ATP hydrolysis was due to the presence of contaminating

ammonia, stoichiometry would be observed between formation of

L-asparagine and AMP.










Table 2.1 Trapping of L-aspartate from Complexes in the
Steady State, Using (14C) L-aspartate, Ammonia- and Glutamine-
Dependent Reactions.

Glutamine-DeDendent Pulse Condition nmle (C) Asn


L-Asp* 0.10 + 0.03


L-Asp* + ATP 0.90 + 0.03


L-Asp* + L-Gln 0.12 + 0.06


Ammonia-DeDendent Pulse Condition nmole of (ik) Asn


L-Asp* + NH3 0.15 0.07


L-Asp* + ATP 0.70 0.10


Isotope trapping experiments were performed under conditions
described in Materials and Methods. All variations were done
in triplicate and evaluated by taking the average from which
the background was subtracted.










Table 2.2 Trapping of ATP from Complexes in the Steady
State, Using (3H) ATP, Ammonia- and Glutamine- Dependent
Reactions.

Glutamine-DeDendent Pulse Condition nmole of (3H) AMP


ATP* 0.43 + 0.10


ATP* + Asp 2.20 + 0.25


ATP* + L-Gln 0.92 + 0.06


Ammonia-Dependent Pulse Condition nmole of (3iH) AMP


ATP* + NH3 0.19 + 0.07


Isotope trapping experiments were performed under conditions
described in Materials and Methods. All variations were done
in triplicate and evaluated by taking the average from which
the background was subtracted.













0.6.

0.5._

0 .4-

00.3.

0.2.

0.1.

0.0.-y- I-iI -- --
0 1 2 3 4 5 6 7
1/[Asp] (mM-1)



Fig. 2.1 Determination of Kd for Aspartate. A 160 p.1
solution of 50 mM Tris-HCl, (pH 8.0), containing 10.0 mM ATP,
12.0 mM MgCl2, 2 nmole ASB, and labelled L-aspartate (0.15-
0.53 mM) was incubated at 370C for two minutes. A 65 p.1 chase
solution was added such that the final concentrations of
substrates were: 50 mM Tris-HCl, pH 8.0, 12.0 rM ATP, 15.0 mM
MgCl2, 10 mM L-glutamine, and 30 mM unlabelled L-aspartate.
Following the quenching, the reactions were treated as
described under Materials and Methods, and radioactivity and
quantity of L-aspartate and L-asparagine were determined as
described before.






46





5-


4.



3.

2.

1.



1 2 31 4 1 6 7 A o
Time (min)



Fig. 2.2 Rate of AMP formation as a function of time.
The assay mixtures (100 pi) contained 50 mM Tris-HCi (pH
8.0), 0.5 mM ATP (600 cpm/ nmole), 2 mM L-aspartate, 2 mM
MgCI2 and 0.5 nmole of ASB. Assay mixtures were incubated at
370C, and reactions were terminated by addition of TCA at the
indicated times. Following the quenching, the reactions were
treated as described under Materials and Methods, and
radioactivity and quantity of AMP, ADP and ATP were
determined as described before.











Discussion


To determine the order of substrate binding, isotope

trapping experiments were done using either L-

[14C(UJ)]Aspartate or [2,8-3H] ATP in the presence or absence

of other substrates. Our results, for the glutamine-dependent

AS reaction, revealed little trapping (10%) of L-Asp* as L-

Asn* when radioactive L-aspartate was used in the pulse in

the presence or absence of L-glutamine. These data suggest

that either L-aspartate can bind the free enzyme in a

catalytically competent manner, or that it dissociates from

binary (E-Asp*) or the ternary complex (E-Asp*-Gln) faster

than it goes on to form the product. When isotope

partitioning experiments were done with ATP in addition to

the L-Asp* in the pulse solution, 90% of E-Asp*-ATP was

trapped as L-Asn*. The ability to trap radioactive L-

aspartate when ATP was included in the pulse solution

suggests that the E-Asp*-ATP complex is formed in a

catalytically competent manner. This strongly suggests that

ATP binds free enzyme first followed by L-aspartate binding.

An alternative suggestion is that in the presence of ATP, the

rate of dissociation of L-Asp* decreased compared to the rate

of overall product formation. What is very clear from these

observations is that the binding and hydrolysis of L-

glutamine is not required prior to the addition of ATP and L-

aspartate, or no labelled L-asparagine would have been










trapped in the absence of L-glutamine. This is strikingly

different from the mechanisms for AS suggested by Chou, 1970,

Milman et al., 1980, Markin et al., 1981 and Hongo and Sato,

1985, who proposed that the binding of L-glutamine had to

occur prior to ATP and L-aspartate binding, in order for the

overall synthesis of L-asparagine to occur.

When isotope partitioning experiments were done with

labelled ATP, in the absence of L-aspartate or L-glutamine,

50% of the E-ATP* complex was trapped as AMP*. This shows

that the E-ATP* complex is formed in a catalytically

competent manner, further suggesting that ATP binds free

enzyme first. Under these conditions, only half of the enzyme

bound ATP* was converted to AMP*. It is possible that ATP*

was bound to the enzyme and partly equilibrated with the

subsequently added unlabelled ATP, or that it dissociated

from binary (E-ATP*) faster than it would go on to form the

product. It is also possible that ATP* was bound to an

additional site (an allosteric site, for which there is no

evidence at this time), that was inhibitory to the ATP

binding for the synthesis, therefore causing less trapping.

Using the isotope trapping techniques, we were unable to

obtain the Kd for ATP. To determine the Kd for ATP, the

concentration of the labelled ATP was varied (0.01-0.5 mM) in

the pulse. Surprisingly, no trapping was detected below 0.5

mM ATP. It is possible that some ATP* equilibrated with the

subsequently added unlabelled ATP, or that it dissociated










from binary (E-ATP*) faster than it would go on to form the

product.

When isotope partitioning experiments were done with L-

aspartate in addition to the ATP* in the pulse solution, E-

ATP*-Asp was trapped as AMP*. These data suggested that the

E-ATP*-Asp complex was formed in a catalytically competent

manner, but the amount of product was puzzling. The fact that

"trapped" AMP* was not stoichiometric with the amount of

enzyme (220%) present suggested that either ATP hydrolysis

was occurring prior to the addition of chase solution or

product formation was occurring in the time scale of the

pulse. We attempted to determine if the net ATP hydrolysis

observed was an activity of the synthetase itself or that of

some other contaminant. Our first results showed that the

production of AMP was found to be L-aspartate dependent and

was not stimulated by aspartate-tRNA. This suggests that the

reaction was not a contaminant and not surely aspartyl tRNA

synthetase. The fact that very little L-asparagine was

detected during this reaction ruled out the possibility of E.

coli ASA, with an apparent Km for ammonia of less than 1 gM

(unpublished data, Boehlein), being involved. The aspartate-

dependent hydrolysis of ATP represented in Reaction 4,


MgATP2- L-aspartate AMP + PP

which is linear with time, is only 1/100 of the overall

reaction rate. As shown in the Figure 2.2, a value of about










3.8 nmole was reached indicating that the L-aspartate-

dependent hydrolysis, in the 15 minutes period, was about 8-

fold turnover of the enzyme. The effect of the presence of

the contaminating NH3, determined by L-asparagine formation,

seems to be responsible for the formation of only 1/10 of

AMP*. This would eliminate the possibility of synthesis

reaction by ASB being responsible for the most of the AMP*

formation. Extrapolation of the line (Fig. 2.2) predicts an

intersection point of approximately 0.5 nmole AMP/0.5 nmole

AS, stoichiometric with the amount of enzyme, further

supporting that the reaction is not a contaminant, and that

the chemistry is very fast compared to the release of the

products. However, whether L-aspartate attacks ATP, forming

aspartyl-AMP, or L-aspartate stimulates ATP hydrolysis is not

clear at this moment. The addition of pyrophosphate reagent

or pyrophosphatase to the reaction mixture produced some

increase in the rate of this reaction (3-4 fold). The

slowness of the partial reaction to the overall synthetase

reaction in not a limit to its acceptance as bona fide

partial reaction of AS, because enzyme theory recognizes the

phenomenon of "substrate synergism" (Bridger et al., 1968).

That is the accelerating effect on a partial reaction of the

presence of the complementary substrate, in this case L-

glutamine, of the enzyme. The hydrolysis seems to be

irreversible as shown in reaction 4, since it has not been

possible to detect 32PPi-ATP exchange for ASB (unpublished










data), presumably because PPi and AMP dissociate from the

enzyme during this catalytic process.

Given the above data it was somewhat surprising that

when L-glutamine was included in the pulse, about 90% of E-

ATP*-Gln was trapped as AMP*. This suggests that an E-ATP*-

Gin complex is formed in a catalytically competent manner.

The E-ATP* complex seems to be tightly bound in the presence

of L-glutamine since the amount of the AMP* trapped is

stoichiometric with the amount of enzyme present. No

stimulation of ATP hydrolysis was observed by L-glutamine

(data not shown). But rather, it seems that in the presence

of L-glutamine, ATP binding to the active site is stabilized.

The lines of evidence presented here indicate that for

the glutamine-dependent reaction of AS the mechanism is

ordered, with ATP binding first to free enzyme. However, we

can not rule out some alternative ordered mechanism with L-

aspartate binding free enzyme, because a small amount of L-

Asp* was trapped as L-Asn* when no ATP was included in the

pulse.

In the case of the ammonia-dependent AS reaction, very

little L-Asp* (15%) was trapped as L-Asn* when NH3 was

included in the pulse solution with radioactive L-aspartate,

suggesting that the presence of NH3 most likely increased the

rate of product formation relative to the dissociation rate

of L-Asp* from free enzyme. The ability to trap radioactive

L-aspartate (70%) when ATP was included in the pulse

solution, shows that an E-Asp*-ATP complex is formed in a










catalytically competent manner, implying that ATP has to bind

first. When isotope partitioning experiments were done with

labelled ATP and NH3 in the pulse, 20% of the E-ATP*-NH3 was

trapped as AMP*. This suggests that the NH3 binding is not

required prior to the binding of ATP* and/or L-Asp* or the

level of AMP* and/or L-Asn* trapped should be close to the

amount of enzyme present. The data also suggest that the

mechanism is ordered such that ATP binds first followed by L-

aspartate binding.

The amount of *ATP trapped as *AMP when NH3 was included

in the pulse with labelled ATP (0.2 nmole), for every nmole

AS, is less than the amount trapped when *ATP alone was in

the pulse for the glutamine-dependent reaction (0.43 nmole).

It is possible that NH3 (100 mM) interfered with the ATP

binding. It is also possible that presence of L-glutamine in

the chase, by stabilizing the ATP binding, prevented the

dissociation of the bound *ATP from the binary (E-ATP*)

complex, so that it would go on to product.

Major conclusions derived from the isotope partitioning

experiments regarding the kinetic mechanism of ASB are as

follows:

1. The enzyme catalyzes aspartate-dependent ATP

hydrolysis. This partial reaction, although very slow (1/100

of the overall rate), provides a direct evidence for the

mechanism of E. coli ASB, in which the hydrolysis requires no

nitrogen source. Argininosuccinate synthetase is another

example where such partial reaction has been observed










(Rochovansky and Ratner, 1967). The enzyme catalyzed

conversion of citrulline to arginine involves an ATP-

dependent condensation between citrulline and L-aspartate.

Both L-aspartate and citrulline could induce cleavage of ATP.

The time course of the aspartate-dependent hydrolysis

differed from that of citrulline-dependent. The aspartate-

dependent hydrolysis was linear over the course of experiment

(30 min), while the citrulline-dependent hydrolysis reached a

maximum in 2 minutes. The aspartate-dependent hydrolysis

reached a value that was about 3-fold turnover of the enzyme.

The cleavage was suggested to be irreversible, since it had

not been possible to detect an aspartate-dependent PPi-ATP

exchange reaction. However, the citrulline-dependent

hydrolysis reached a value that was stoichiometric with the

amount of enzyme present. Basically, the reaction had come to

a halt, probably because the products remained enzyme bound.

2. The mechanism is probably ordered, with ATP binding

first and L-aspartate second, the preferred, but not

mandatory order.

3. The binding and hydrolysis of L-glutamine or the

binding of ammonia is not required prior to ATP and L-

aspartate binding for the synthesis reaction.

These are significantly different from what were

reported in the past for the AS mechanism. The differences

seem to be due to the fact that the previous investigators

relied solely on steady state kinetic methods, and isotope

trapping was not used. In addition, they failed to consider






54



other models to explain their data, as discussed in the next

chapter.















CHAPTER 3
SUBSTRATE BINDING AND PRODUCT RELEASE OF ASPARAGINE
SYNTHETASE B STUDIED BY STEADY STATE KINETICS


Introduction


The proposed kinetic mechanisms for the glutamine-

dependent reaction of AS has been reported for the enzymes

from beef (Schuster et al., 1981) and mouse (Milman et al.,

1980) pancreases, and rat liver (Hongo and Sato, 1985), as

discussed in chapter 1. In all three cases, steady state

kinetic methods were used to elucidate the kinetic mechanism.

There are dramatic differences between the data presented,

but all three reports agree that L-glutamine is the first

substrate to bind. Milman et al. (1980), Markin et al. (1981)

and Hongo and Sato, (1985) all present mechanisms that start

with a glutaminase reaction mainly because AS exhibits a

glutaminase activity.

There are major differences between the data presented

in the past and that for E. coli ASB. Using isotope

partitioning techniques, evidence was presented regarding the

kinetic mechanism of ASB. The results suggested that, (a) the

kinetic mechanism for glutamine- and ammonia-dependent

reactions is preferentially ordered, with ATP binding first

followed by addition of L-aspartate, (b) the enzyme catalyzes

an aspartate-dependent ATP hydrolysis that requires no










nitrogen source, and (c) glutamine binding and hydrolysis is

not required prior to the binding of the other substrates. To

determine whether the difference in the conclusions was

associated with using a different enzyme or with the

techniques employed, the following initial velocity

experiments, including studies with inhibition by products

and substrate analogs, were performed. The experimental

approach used here was that of Frieden (1959), which also

allowed us to distinguish the sequential mechanism from ping-

pong mechanisms.


Materials and Methods



Chemicals and Reaaents


Trichloroacetic acid (TCA) was purchased from Fisher

Scientific (Orlando, FL). The pyrophosphate reagent for

following PPi production, MgCl2, ATP, L-aspartate, L-

asparagine, L-glutamine, L-glutamate, AMP, PPi, ammonium

acetate, L-glutamic acid y-monohydroxamate, ninhydrin,

Tris(hydroxymethyl) aminomethane (Tris-HCl), and Bis(2-

hydroxylethyl)iminotris (hydroxymethyl)methane (Bis Tris),

were all purchased from Sigma. Dithiothreitol (DTT) was

obtained from Promega Corporation (Madison, Wisconsin).











Expression of the Protein and Purification


Protein expression, cell culture and enzyme purification

was carried out as described before (chapter 2).


Protein Concentration Determination


Protein concentration was measured using Bio-Rad Protein

Assay (Bradford, 1976). Mouse immunoglobulin G was used to

obtain a standard curve.

Enzyme Assays

The velocities were measured spectrophotometrically by

assaying for PPi production (O'Brian, 1976), as described in

chapter 2. The assay mixture contained the following

components: 50 mM Tris-HCl (pH 8.0), and varying amounts of

ATP (0.2-0.5 mM), L-aspartate (0.2-0.7 mM), L-glutamine (0.2-

0.7 mM), AMP (3-15 mM), L-glutamate (5-50 mM), L-

asparagine(0.05-0.25 mM), and ammonium acetate (3-50 mM).

MgCl2 concentration was kept constant (3 mM), unless stated

otherwise. The volume of the total reaction mixture was kept

at 160 gi. All reactions were carried out at 37C, using 7.4

gg of enzyme per reaction. In all experiments, L-glutamine

solutions were freshly prepared using recrystallized L-

glutamine (Sheng et al., 1993). Under all experimental

conditions initial velocity was verified.

All studies were done in duplicate, and the data

collected were evaluated by taking average of the samples










from which the background (representing the ATPase activity)

was subtracted. Velocities were reported as nmole of PPi

produced per minute per milligram of protein. The data were

then plotted in the form of double reciprocal plot or a Dixon

plot (Dixon, 1953) depending upon the type of experiments and

weighted using the computer program, Ultrafit, purchased from

Biosoft.

When PPi was used in product-inhibition studies, the

velocities were measured by monitoring the conversion of L-

aspartate to L-asparagine by HPLC amino acid analysis (Sheng

et al., 1993). The assay mixture (in all experiments, the
reaction volume was 200 il) was consisted of the following:

50 mM Tris-HCl (pH 7.0), and varying amounts of ATP, L-

aspartate, L-glutamine, and PPi. Value of pH 7.0 was chosen

to overcome precipitation of MgCl2 by PPi. Concentration of

MgCl2 was kept so that it would be 1 mM above ATP and PPi.

Assay mixtures were incubated for 15 minutes at 370C after

addition of enzyme (1 gg/reaction mixture) and terminated by

addition of 50 pl of 20% (1.22 M) trichloroacetic acid (TCA)

containing 0.2 mM L-histidine as an internal control. The

controls were performed in a similar pattern except that TCA

was added prior to the incubation. An aliquot of the reaction

mixture (10 gl) was then injected into the amino acid

analyzer. Amino acid analysis is performed on an applied

Biosystems 130 A separation system (Sheng et al., 1993).

Under all experimental conditions enzyme stability and










initial velocity were verified, doing time course

experiments.

All studies were done in triplicate, and the data

collected were evaluated by taking average of the samples.

Velocities were reported as nmole of L-asparagine produced

per minute per milligram of protein.


Stoichiometrv of PPj and L-Glutamate


In this experiment the assay mixtures (100 p.) contained

50 mM Tris-HCl (pH 8.0), 1 mM ATP, 1 mM L-aspartate, 5 mM

MgCl2 and varying concentration of L-glutamine (0.1-20 mM).

The assay mixtures and the enzyme were preincubated at 370C

for 3 min. The reactions were initiated by the addition of

the enzyme (3 ig/reaction) and were incubated 370C for 4 min

before being terminated by addition of 20% TCA (15 li).

Modified glutaminase assay (Bernt and Bergmeyer, 1974) was

used to measure L-glutamate concentrations. PPi production

was measured by modifying the continuous spectrophotometric

assay (O'Brian, 1976) to an end point assay. In this case,

385 gi of the coupling buffer (50 mM imidazole, pH not

adjusted, and 20 pi of pyrophosphate reagent, which was

originally reconstituted in 1 ml of ddH20), was added to the

reaction mixtures, following TCA kill, and incubated at room

temperature for 30 min. The absorbance of the resulting

solution was measured at 340 nm, and the amount of PPi

produced in the reaction determined from a standard curve.










The ratio of L-glutamate/PPi versus concentration of L-

glutamine was used in plotting.


Results



Initial Rate Studies


For determination of the order of the addition of the

substrates in the E. coli ASB, initial velocity studies were

performed and substrates varied in a systematic way. In

particular, to determine the order of addition of substrates

in the glutamine-dependent reaction, the concentrations of

two of the three substrates, L-aspartate, L-glutamine and ATP

were varied while maintaining the third substrate fixed and

subsaturating in the absence of any products (Frieden, 1959).

Keeping the concentration of L-glutamine fixed and

subsaturating, the plot of 1/v versus. I/[L-aspartate] at

different ATP concentrations (or vice versa (data not shown))

was found to be intersecting (Fig. 3.1A) which was confirmed

by slope and intercept plots (Fig. 3.1B). Parallel lines were

obtained when 1/v versus 1/[L-Glutamine] (constant and

subsaturating L-aspartate) was plotted at varied ATP

concentrations, and 1/v versus 1/[L-aspartate] (constant and

subsaturating ATP) was plotted at varied L-glutamine

concentrations (Fig. 3.2A and Fig. 3.3A, respectively). These

were confirmed by slope and intercept plots (Fig. 3.2B and

Fig. 3.3B). Interestingly, this suggests that a product is










released between the addition of each pair of substrates.

Kinetic parameters determined from these data were KAsp of

0.05 mM, KATP of 0.05 mM, and KGn of 0.20 mM.

Of all the possible mechanisms available, two mechanisms that

have been shown from the literature, to be associated with AS

are as follows: Scheme A, the bi-uni-uni addition of

substrates, and Scheme B, the uni-uni-bi addition of

substrates.


(Scheme A)
A


(Scheme B)
A


Q R S


Q R S


Using the method of Fromm (1975), the steady-state initial

velocity equations were derived for the two mechanisms shown

above, assuming no product present. Equations 1 and 2 were

derived from Schemes A and B, respectively.










equation 1) 1/v = 1/V (0/[A] + 0/[B] + 0/[AB] + 0/[C] + 1)

equation 2) 1/v = 1/V (0/[A] + 0/[B] + 0/[C] + 0/[BC] + 1)

Equation 1 predicts that substrate C (L-glutamine) affects

only the intercept of the 1/v versus I/[A] or versus 1/[B]

(ATP and L-aspartate or vice versa) which would result in

parallel lines, characteristic of a ping-pong mechanism,

which indicates that the addition of each pair of substrates

are separated by the release of a product. However, equation

1 indicates that substrates A and B (ATP and L-aspartate or

vice versa) affect both the slope and intercept of 1/v versus

1/[B] and versus I/[A] plots, respectively, resulting in

intersecting lines, which suggests that the addition of A and

B is sequential. Similarly we see from equation 2 that

substrate A (L-glutamine) affects only the intercept of the

1/v versus I/[B] or I/[C] plots, predicting parallel lines.

Equation 2 also predicts that substrates B and C (ATP and L-

aspartate or vice versa) affect both the slope and intercept

of the plots of 1/v versus 1/[C] and versus 1/[B],

respectively, resulting in intersecting lines. In other

words, our initial rate studies agree with the theoretical

results obtained for both equations 1 and 2!

The ammonia-dependent AS reaction was examined

kinetically. It was found that the plot of 1/v versus.

I/[ATP] at different L-aspartate concentrations, keeping the

concentration of ammonium acetate fixed (50 mM), shows

intersecting lines (Fig. 3.4A). This was verified by plotting

the slope and intercept (Fig. 3.4B). A similar pattern was










observed when 1/v versus. I/[ATP] at different L-aspartate

concentrations was plotted (data not shown). These data

suggest that for the ammonia-dependent reaction, the addition

of ATP and L-aspartate is sequential. The plots of 1/v

versus. I/[ATP] at varied ammonium acetate concentrations,

keeping L-aspartate constant and subsaturating (1 mM), (Fig.

3.5A) and of 1/v versus. 1/[L-aspartate] at varied ammonium

acetate concentrations, keeping ATP constant and

subsaturating (1 mM), (Fig. 3.6A) show parallel lines. This

was confirmed by slope and intercept plots (Fig. 3.5B and

Fig. 3.6B, respectively). These data suggest that the

addition of each pair of substrates are separated by the

release of a product. The results for the ammonia-dependent

AS reaction are in agreement with the theoretical predictions

of equation 1. Equation 1 predicts that NH3 (C) affects only

the intercept of the 1/v versus I/[ATP] or versus I/[L-

aspartate] plot (A and B or vice versa) which would result in

parallel lines. Equation 1 also predicts that substrates ATP

and L-aspartate (A and B or vice versa) affect both the slope

and intercept of 1/v versus i/[L-aspartate] and versus

I/[ATP] plots, respectively, resulting in intersecting lines.

The data indicate that the addition of ATP and L-aspartate is

sequential, therefore suggesting that ATP and L-aspartate

will bind first, presumably forming aspartyl-AMP. This is

followed by release of PPi (p) and addition of the NH3.

However, the result for the ammonia-dependent reaction would

not be comparable with those predicted for equation 2 (see










discussion). Kinetic parameters determined from these data

were KAp of 0.05 mM, KATP of 0.05 mM, and KNH3 of 10.0 mM.

The initial rate study of beef pancreas AS (Markin et

al., 1981) showed intersecting lines for the plot of 1/v

versus. 1/ATP at varied L-glutamine concentration, suggesting

ATP and L-glutamine bind sequentially. This was completely

different from the observations made by Milman et al., 1980,

Sato, 1985 and by us. The double reciprocal plot of the

velocity dependence on ATP at various constant concentrations

of the L-glutamine was shown to be parallel (Fig. 3.2A),

which indicates that the addition of each pair of substrates

is separated by the release of a product. To resolve this

difference, the following initial rate experiment was

performed using L-glutamine and an alternative substrate, L-
glutamic acid y-monohydroxamate (LGH). Control experiments

showed that LGH was a competitive inhibitor of L-glutamine

(Ki = 0.2 mM ) (Fig. 3.7). For the initial rate experiment,

the concentration of ATP was varied (0.1-0.7 mM), while

keeping L-aspartate constant and subsaturating (1 mM), at

fixed concentration of L-glutamine and LGH (0.2 mM). The

velocities were measured spectrophotometrically as described

before. The plot of 1/v versus. 1/ATP at fixed concentration

of L-glutamine and LGH (Fig. 3.8) shows that the substitution

of LGH for L-glutamine did not not have any effect on the

slope of the line, characteristic of a ping-pong mechanism.

This indicates that the addition of ATP and L-glutamine are

separated by the release of a product (see discussion).










In order to determine the effect of substrate saturation

on the kinetic behavior, the initial rate studies were also

performed under saturating conditions of the substrates. When

the concentration of L-glutamine was kept constant and

saturating (20 mM), the plot of 1/v versus. 1/[L-aspartate]

at different ATP concentrations (Fig. 3.9) showed parallel

lines, suggesting that the addition of each pair of

substrates was separated by the release of a product (see

discussion). In the case where L-aspartate was kept constant

and saturating (> 2 mM), the plot of 1/v versus. I/[L-

glutamine] at varied ATP concentration showed what appears to

be substrate inhibition (Fig. 3.10). The pattern changed from

parallel to intersecting. In other words, the slope of the

double reciprocal plot began to increase (but not

intercepts), which represents competitive substrate

inhibition. The same type of observation became evident for

the plot of 1/v versus 1/[L-glutamine] at varied L-aspartate

concentrations when ATP was kept constant and saturating (>5

mM). Very high ATP concentration made L-aspartate become an

inhibitor of synthetase reaction.


Inhibition bv Substrate Analogs


Substrate analogs were tested as inhibitors of ASB to

obtain more information about the substrate binding order

(data provided with Dr. S. Boehlein). This alternative

approach requires that a competitive inhibitor be available










for each substrate. Experimentally, the concentration of the

substrate to be studied is varied at fixed varied

concentration of the analog. The remaining substrates are

held at fixed and subsaturating concentrations. Experimental

details are described in the table legend. The substrate

analogs used in this set of experiments were, AMP-PNP, 3-

methyl aspartate and L-glutamic acid y-methyl ester. Table 3.1

shows the inhibition patterns for all the substrate analogs

tested for the glutamine-dependent reaction of ASB. Initial

velocity studies in the presence of AMP-PNP demonstrated

competitive inhibition with respect to ATP, suggesting it

competes with the ATP for the same site on the enzyme. AMP-

PNP was found to be noncompetitive with respect to L-

aspartate and uncompetitive with respect to L-glutamine. 3-

methyl aspartate was found to be competitive with L-aspartate

and noncompetitive with respect to ATP and L-glutamine. L-
glutamic acid y-methyl ester was competitive with L-glutamine,

uncompetitive with ATP and noncompetitive with L-aspartate.

The rate equations for the effect of analogs (competitive

inhibitors) of each substrate were derived for the two

mechanisms (A and B). Although there were some disagreements,

initial rate results (Table 3.1) in most cases agreed with

predicted patterns obtained from rate equations for both

mechanisms (see discussion).











Stoichiometrv of Glutamine-deoendent Reaction


E. coli ASB can function as a glutaminase (reaction 3)

when L-aspartate and ATP are not present, showing that L-

glutamine can be bound by the free enzyme. In order to

determine the relevance of the glutaminase reaction to the

glutamine-dependent synthetase mechanism, a stoichiometry

experiment was performed. In this experiment, the

concentration of L-glutamine was varied while keeping ATP and

L-aspartate constant and subsaturating (1 mM). The

concentrations of L-glutamate and PPi produced were measured

simultaneously. The synthesis of L-asparagine has been shown

to be stoichiometric with PPi formation (data not shown)

under all circumstances. Stoichiometry should also be

observed between formation of L-glutamate and PPi if no

glutaminase was occurring during the synthesis of L-

asparagine. However, as shown in Figure 11, plotting the

ratio of L-glutamate/PPi versus L-glutamine concentration

resulted in a hyperbolic curve, approaching a plateau with

increasing concentrations of L-glutamine. This shows that the

L-glutamate and PPi production is non-stoichiometric,

indicating that glutaminase reaction is occurring at the same

time as the synthetase reaction and at a much faster rate

than the synthetase reaction as concentration of L-glutamine

increases, approaching approximately 2:1.











Product Inhibition Studies


Our initial velocity studies proved somewhat

inconclusive information regarding the order of addition of

the substrates. To obtain additional information about the

order of product release, product inhibition of the initial

velocity was studied for both the L-glutamine-dependent and

ammonia-dependent L-asparagine synthetase reactions. The

velocities were measured spectrophotometrically by assaying

for PPi production (O'Brian, 1976). Figures 12 through 17 and

Table 3.2 show the product inhibition patterns and constants

for the ammonia-dependent AS reaction. When the concentration

of one substrate is varied, the remaining substrates are held

at fixed concentrations: ATP, 1 mM, L-aspartate, 1 mM, and

ammonium acetate, 50 mM. L- asparagine was found to be

competitive with respect to ammonia (Ki = 0.08 0.025 mM),

and noncompetitive with respect to ATP and L-aspartate (Ki =

0.190 + 0.002 and 0.26 0.002 mM, respectively). AMP was

noncompetitive with respect to all the three substrates. The

Ki for ATP, L-aspartate and ammonia were 3 0.001, 8 0.001

and 15 0.002 mM. The fact that L-asparagine was competitive

with ammonia, suggesting they bind to the same enzyme form,

allows us to place L-asparagine after ammonia (Q) followed by

AMP (R).

Figures 18 through 28 and Table 3.3 show the product

inhibition patterns and constants for the glutamine-dependent










AS reaction. When the concentration of one substrate was

varied, the remaining substrates were held at fixed and

subsaturating concentrations (1 imM). The MgCl2 concentration

was kept constant (3 mM), unless stated otherwise. L-

asparagine was found to be competitive with respect to L-

glutamine (Ki = 0.015 0.003 mM), suggesting it binds to the

same enzyme form as L-glutamine. L-asparagine was

noncompetitive with respect to ATP (Ki = 0.09 0.001 mM) and

L-aspartate (Ki = 0.27 0.001 mM), suggesting it binds the

free enzyme and enzyme-substrate complex. PPi was competitive

with respect to ATP and L-aspartate (Ki = 0.05 0.002 and

0.39 0.01 mM, respectively). L-glutamate was a poor

inhibitor and was shown to be noncompetitive with respect ATP

(Ki = 47 + 0.001 mM). L-glutamate was found to be competitive

with respect to L-aspartate (Ki = 23 0.001 mM) and

noncompetitive with respect to L-glutamine (Ki = 52 0.001

mM). AMP was also a poor inhibitor. AMP was noncompetitive

with respect to all the three substrates. The Ki for L-

glutamine, L-aspartate and ATP were 14 0.001, 9 0.001,

and 5 + 0.001 mM

For further clarification of the order of product

release, more specifically the first product, the following

initial velocity experiments were performed. L-aspartate and

L-glutamine concentrations were varied against each other in

the presence of L-glutamate (50 mM), keeping the ATP

concentration constant and subsaturating (1 mM). In another

experiment, concentrations of L-aspartate and L-glutamine










were varied in the presence of PPi (0.4 mM), while keeping

the ATP concentration constant and subsaturating (1 mM). If

L-glutamate is the first product released between L-aspartate

and L-glutamine, then a double reciprocal plot of 1/v versus

1/[L-aspartate] at various L-glutamine concentrations will

produce intersecting lines in the presence of L-glutamate. If

however, L-glutamate is not the product released between the

addition of L-glutamine and L-aspartate, then the double

reciprocal plot will result in parallel lines. On the other

hand, if PPi is is the first product released between L-

aspartate and L-glutamine, intersecting lines will be

observed for the plot of 1/v versus I/[L-aspartate], varying

the L-glutamine concentration. Figures 29 and 30 show the

results of 1/v versus i/[L-aspartate] at various the L-

glutamine concentrations in the presence of L-glutamate or

PPi. The parallel lines in the presence of L-glutamate (Fig.

3.29) suggests that L-glutamate can not be the first product

(P) released between L-aspartate and L-glutamine. Parallel

lines were also observed in the presence of PPi (Fig 3.30)

which suggest that PPi cannot be the first product off (see

discussion).

To obtain more information about the order of product

release (Q, R, S), double-inhibition studies were performed

for the glutamine-dependent AS reaction. This set of

experiments examines the relation between the products,

determining whether the two products interact with each other

on the enzyme's surface. For this set of experiments, the










concentration of all the substrates was held constant (1 mM),

and concentrations of two of the products were varied against

each other. If L-asparagine and AMP interact with each other

(are next to each other in the release order), then a Dixon

plot (1953) of 1/v versus [L-asparagine] at various AMP

concentrations will give intersecting lines (Segel, 1975).

If, however, their binding is separated by another product,

then the Dixon plot will show parallel lines. Figures 3.31,

3.32, and 3.33 show the dual-inhibition studies of the 1/v

versus. [L-asparagine] and [L-glutamate] at different

concentrations of L-asparagine, and AMP, accordingly. The

plot of 1/v versus. [L-asparagine] at varied AMP

concentrations (Fig. 3.31) shows intersecting lines. This

shows that the presence of L-asparagine enhances AMP

inhibition or vice versa, therefore, indicating that they can

combine sequentially with the enzyme. The plots of 1/v

versus. [L-glutamate] at different AMP and L-asparagine

concentrations (Fig. 3.32 and Fig. 3.33) revealed parallel

lines. The presence of L-glutamate would not enhance L-

asparagine or AMP inhibition or vice versa, which suggests

that the products are mutually exclusive. Based on these

observations we are now able to place the AMP and L-glutamate

release steps (R and S, respectively) after the release of L-

asparagine (Q).

The rate equation for the proposed mechanism (Scheme A)

was derived assuming the products P, Q, R and S were present,

and the product inhibition patterns predicted were compared










with experimental results. There are some disagreements

between the predicted patterns and the experimental results.

Other considerations are necessary to explain the

experimental data (see discussion). One of the most obvious

possibilities is the existence of an isomerization step

following the release of the last product shown below.




(Scheme C)
A B P C R
KiA k2B k3 K4C K5 K6 K7

K-1i k-2 k-3p K-4 K-5Q K-6R K-7S


F<^
k8






The iso-mechanism predicts a non-competitive inhibition

pattern between the last product and first substrate which

would otherwise be competitive. In other words, after the

release of S, L-glutamate, the enzyme is in a conformation

that is not accessible to A, as is indicated by the symbol F

(Segel, 1975). The F form has to convert back to E by a

reaction indicated by rate constants k8/k-8. The rate

equation for the proposed mechanism with the iso step (Scheme

C) was also derived assuming all products were present and

was compared with the rate equation for the same mechanism

with no iso step (Scheme A). The equation for the Scheme C

was very different and much more complex than that of the










equation for Scheme A (see discussion). The differences that

were useful for the comparison of the two mechanisms were as

such:

equation 3) 1/v = 1/V (O[RS]/[AB])

equation 4) 1/v = 1/V (O[RS] / [AB] + O[RS])

The equation for the mechanism with no iso step (equation 3)

predicts that the RS term affects only the slope of 1/v

versus. I/A (or 1/B) which would result in a competitive

pattern (lines intersecting on 1/v axis). Similarly the

equation for the mechanism with iso step (equation 4)

predicts that RS term affects both the slope and the

intercept of the plot of 1/v versus. I/A (or 1/B) which

results in a noncompetitive pattern (lines intersecting to

the left of the 1/v axis). According to our proposed

mechanism (Scheme A), A and B are ATP and L-aspartate or vice

versa, and R and S are AMP and L-glutamate, respectively.

To verify if there is an isomerization step in our

proposed mechanism leading to unusual product inhibition

patterns, the following initial velocity experiment was

performed: The concentration of L-aspartate was varied, at

fixed varied concentrations of L-glutamate and AMP in a

constant ratio ([L-glutamate] = 5[AMP]), while keeping ATP

and L-glutamine constant and subsaturating (1 mM). The plot

of 1/v versus. 1/L-aspartate (Fig. 3.34) shows intersecting

lines whose intersection is to the left of 1/v axis

(noncompetitive). This result agrees with the theoretical










prediction of equation 4, supporting a proposed mechanism

with an iso step (Scheme C).


Discussion


The initial velocity patterns of the relationship between

pairs of substrate obtained for both the ammonia- and

glutamine- dependent reactions were similar. From these data

we proposed the two mechanisms (A and B). Scheme A, the bi-

uni-uni addition of substrates, and Scheme B, the uni-uni-bi

addition of substrates.


(Scheme A)
A


(Scheme B)
A


The data presented for the ammonia-dependent reaction are

only consistent with mechanism A (bi-uni-uni-bi ping-pong).










According to mechanism A, ATP and L-aspartate bind first,

forming aspartyl-AMP. This is followed by release of PPi (P)

and addition of NH3. The results for the ammonia-dependent

reaction do not support mechanism B (uni-uni-bi-bi ping-

pong). According to equation 2 (derived to fit mechanism B),

substrate A affects only the intercept of the 1/v versus

1/[B] or 1/[C] plots, which predicts parallel lines. This

indicates that the addition of each pair of substrates is

separated by the release of a product. Therefore, if

mechanism B is the correct mechanism, following the addition

of NH3, a product must be released prior to the addition of

the second substrate, which is not possible.

When the ammonia-dependent reaction was studied for the

beef pancreatic AS (Markin et al., 1981), the kinetic

mechanism was proposed to be significantly different from

that of ASB. According to their previous proposed mechanism,

NH3 bound first followed by a random addition of ATP and L-

aspartate. When evaluating the ammonia-dependent AS reaction,

these workers found that the plots of 1/v versus 1/NH3 at

varied L-aspartate concentrations (at saturating ATP) and 1/v

versus 1/NH3 at varied ATP concentrations (with saturating L-

aspartate) resulted in parallel lines. This was proposed to

indicate either a random addition of ATP and L-aspartate or

the release of a product between their additions. No

information was provided regarding the pattern between ATP

and L-aspartate (with saturating NH3), and no experimental










evidence was provided to support the notion that the addition

of ATP and L-aspartate was random.

Although our data from initial velocity studies

supported a bi-uni-uni-bi ping-pong mechanism (mechanism A),

for the ammonia-dependent reaction of ASB, we were still

unable to determine whether the addition of ATP and L-

aspartate is ordered or random. The data from isotope

trapping experiments clearly indicated that the mechanism is

ordered such that ATP binds first followed by L-aspartate

binding (see Chapter 2). These data together support

mechanism A (ordered bi-uni-uni-bi ping-pong) for the

ammonia-dependent AS reaction. The rate equation and kinetic

parameters have been worked out (Fromm, 1975). For this

reaction the rate equation is as follows:
equation 5) 1/v = 1/V (I/[ABC]-[PQRS] (01[C] + 02[AB] +

*3[AC] + 04[BC] + 05[ABC] + 06[P] + 47[PQ] + 08[AP] + 9[CR] +

010[PR] + 011[QR] + 012[ABP] + 13[ABQ] + 14[APQ] + 015[BQR]
+ 016[PQR] + 017[BCR] + 018[CQR] + 019[ABCQ] + 020[ABPQ] +

021[BCQR] + 022[BPQR]))

The data for the glutamine-dependent AS reaction,

however, are consistent with both mechanisms (A and B).

According to mechanism A, a bi-uni-uni-ter ping-pong

mechanism, ATP and L-aspartate sequentially bind followed by

PPi release. This is followed by L-glutamine which is the

last substrate to bind. According to mechanism B, a uni-uni-

bi-ter ping-pong mechanism, L-glutamine binds first followed

by the release of the L-glutamate. ATP is then the second










substrate to add, followed by the addition of L-aspartate or

vice versa.

Kinetic studies done under saturating condition provided

evidence of substrate inhibition (Fig. 3.10). Under these

conditions substrate inhibition was observed with ATP and L-

aspartate. This is an artifact that is common in ping-pong

mechanisms and has been reported for other enzymes including

fatty acid synthetase (Katiyar et al., 1975), 5-

enolpyruvoylshikimate-3-phosphate synthase (EPSPS) (Gruys et

al., 1992), UDP-N-Acetylenolpyruvylglucosamine reductase

(Dhalla et al., 1995) and nucleoside diphosphate kinase

(Garces and Cleland, 1969). Substrate inhibition usually is

caused by the substrates binding to the improper forms of the

enzyme which would result in dead-end complexes. The

inhibition effect is competitive because the substrates

interact with the same enzyme form (A and B with E), which is

characterized by an increase in the slopes and not the

intercepts of reciprocal plot. Interestingly, when the

concentration of L-glutamine was kept constant and high (20.0

mM) (100 x Km), a parallel pattern was seen for the plot of

L-aspartate versus ATP (Fig. 3.9), suggesting that the

addition of this pair of substrates are separated by the

release of a product. If we were to accept the parallel

pattern for the plot of L-aspartate versus ATP, and not the

intersecting pattern obtained under nonsaturating

concentration of L-glutamine (1 mM), we could argue that the

mechanism of glutamine-dependent reaction of ASB is totally










random with respect to all the three substrates. Therefore,

this would rule out both proposed mechanisms A and B.

However, according to Fromm (1975), the nonvaried substrate,

L-glutamine in this case, must be kept above its respective

Km, but nonsaturating. This is because if it is raised to a

saturating concentration (100 x Km), artifactual parallel

lines may be observed in the double-reciprocal plot. In other

words, the magnitude of either O/AB (mechanism A) or O/BC

(mechanism B) would change such that the slope term

disappears. This in fact can explain some of the

discrepancies between our data and the data presented by

Markin et al., (1981). In their study, the plot of L-

aspartate versus ATP showed parallel lines when they kept L-

glutamine constant and saturating (16.67 mM).

The double-reciprocal plot of ATP versus L-glutamine, at

constant concentrations of L-aspartate, was parallel for

E.coli ASB (Fig. 3.3A), rat liver AS (Hongo and Sato, 1985)

and mouse pancreatic AS (Milman et al.,1980). Yet, the beef

pancreatic AS displayed an intersecting pattern (Markin et

al., 1981). This discrepancy was the motivation for using an

alternative substrate. LGH was substituted for L-glutamine in

the initial rate studies. We showed that the plot of 1/v

versus. I/ATP at varied fixed concentration of L-glutamine or

LGH (0.2 mM), resulted in parallel lines. It appeared that

the substitution of the LGH for L-glutamine did not alter the
magnitude of either the O/AB or the O/BC, therefore, no slope

effect become evident. In other words, according to equation










1 (derived to fit mechanism A), the slope term 4/AB (A = ATP

and B = L-aspartate or vice versa) is independent of C.

Therefore, regardless of what C is (L-glutamine or LGH), the

plot of 1/v versus 1/ATP would result in parallel lines. On

the other hand, according to equation 2 (derived to fit
mechanism B), the slope term O/BC (B = ATP and C = L-

aspartate or vice versa) is independent of A. Therefore, no

matter what A is (L-glutamine or LGH), parallel pattern will

be observed for the plot of 1/v versus l/ATP, indicating that

the addition of ATP and L-glutamine are separated by the

release of a product.

Substrate analogs were used to obtain more information

about the reaction mechanism, mainly the substrate binding

order. Rate equations for the effect of analogs (competitive

inhibitors) of the substrates for mechanism A were derived.

The rate equation for the effect of AMP-PNP with respect to A

(ATP) is described as follows:
equation 6) 1/v = 1/Vmax + Ka/Vmax(A) (1 + I/Ki) +

Kb/Vmax(B) + KiaKb/Vmax(A)(B) (1 + I/Ki) + Kc/VmaxC

The equation shows that for the double reciprocal plot of 1/v

versus 1/ATP at different concentrations of AMP-PNP, only the

the slope term is altered; i.e.,

equation 7) Slope = Ka/Vmax(A) (1 + I/Ki) +

KiaKb/Vmax(A) (B) (1 +I/Ki)

therefore, equation 6 predicts that AMP-PNP is a competitive

inhibitor of ATP.










On the other hand when B (L-aspartate) is the varied

substrate, the double reciprocal plot at different fixed

concentrations of AMP-PNP will show an increase in both the

slope and intercept terms,

equation 8) Intercept = 1/Vmax {1 + Ka/(A) (1 + I/Ki)}

equation 9) Slope = l/Vmax {Kb + KiaKb/(A) (1 + I/Kji)}

and when C (L-glutamine) is the varied substrate, only the

intercept term increases.

equation 10) Intercept = 1/Vmax{l + Ka/(A) (1 + Ki) +

KiaKb/(A)(B) (1 + I/Ki)}

Therefore, equation 6 predicts that AMP-PNP is a

noncompetitive inhibitor of B (L-aspartate) and an

uncompetitive inhibitor of C (L-glutamine). These predicted

patterns obtained from equation 6 are therefore comparable

with the experimental results obtained with AMP-PNP with

respect to ATP, L-aspartate and L-glutamine, respectively

(see Table 3.1).

The rate equation for the effect of P-methyl aspartate

of L-aspartate is described by eq. (11).
equation 11) 1/v = 1/Vmax + Ka/Vmax(A) + Kb/Vmax(B) (1 +

I/Ki) + KiaKb/Vmax(A)(B) + Kc/VmaxC

As can be seen from equation 11, a competitive inhibitor with

respect to L-aspartate will show uncompetitive inhibition

with respect to ATP and L-glutamine. However, there is some
disagreement with the experimental result, in that P-methyl

aspartate was found to be noncompetitive with respect to ATP

and L-glutamine.










When the rate equation for the effect of L-glutamic acid
y-methyl ester of L-glutamine is derived, the following

relationship is obtained.

equation 12) 1/v = l/Vmax + Ka/Vmax(A) + Kb/Vmax(B) +

KiaKb/Vmax(A)(B) + Kc/VmaxC (1 + I/Ki)

The equation predicts that a competitive inhibitor with

respect to C (L-glutamine) should be uncompetitive with

respect ATP and L-aspartate. However, L-glutamic acid y-methyl

ester was found to be noncompetitive with respect to L-

aspartate.

The rate equations were also obtained for the effect of

analogs of the substrates for mechanism B. The rate equation
for the effect of L-glutamic acid y-methyl ester on L-

glutamine is described as follows:

equation 13) 1/v = 1/Vmax + Ka/Vmax(A) (1 +I/Ki) +

Kb/Vmax(B) + Kc/Vmax(C) + KibKc/Vmax(B)C)

According to the equation, the inhibitor that is competitive

with respect to A (L-glutamine), will be uncompetitive with

respect to (B) ATP and (C) L-aspartate. However, the

experimental result is somewhat different in that L-glutamic
acid y-methyl ester was found to be noncompetitive with

respect to L-aspartate.

The rate equation was derived for the effect of AMP-PNP

with respect with B (ATP) as shown below:

equation 14) 1/v = 1/Vmax + Ka/Vmax(A) + Kb/Vmax (B) (1 +

I/Ki) + Kc/Vmax(C) + KibKc/Vmax(B)(C) (1 + I/Ki)










Equation 14 predicts that a competitive inhibitor with

respect to ATP (namely AMP-PNP) should be noncompetitive with

respect to L-aspartate and uncompetitive with respect to L-

glutamine. These predictions are comparable with the

experimental results (see Table 3.1).

The rate equation for the effect of 5-methyl aspartate

of L-aspartate is described as follows:

equation 15) 1/v = 1/Vmax + Ka/Vmax(A) + Kb/Vmax (B) +

Kc/Vmax(C) (1 + I/Ki) + KibKc/Vmax(B)(C) 1/Vmax

The equation predicts that a competitive inhibitor with

respect to L-aspartate should be noncompetitive with respect

to ATP and uncompetitive with respect to L-glutamine. This

disagrees with the experimental results (see Table 3.1) in

that it was found to be noncompetitive with respect to ATP

and L-glutamine. For most cases the experimental data agreed

with the theoretical results, however it was not possible to

differentiate between bi-uni-uni ter ping-pong (Scheme A) and

uni-uni-bi ter ping-pong (Scheme B) mechanisms using this

approach.

AS can function as a glutaminase in the presence or

absence of the ATP and L-aspartate. Other glutamine

amidotransferases such as carbamyl phophate synthetase

(Nagano et al., 1970), cytidine triphosphate synthetase

(Levitzki and Koshland, 1971) and glutamine

phosphoribosylpyrophosphate amidotransferase (Mei and Zalkin,

1989) have also been shown to have this activity. In most

cases, however, the relative amount of glutaminase activity










appeared to be far less than the overall reaction rate,

except for the AS from leukemia cells that was shown to have

higher glutaminase activity compared to synthetase (2:1)

activity. Our data showed that the formation of L-glutamate
was higher than that of PPi for ASB. It also showed that

ratio of L-glutamate to PPi increased with increasing

concentration of L-glutamine. This suggests that glutaminase

activity is happening at the same time as synthetase

activity, and by increasing the concentration of L-glutamine,

the glutaminase activity increases significantly over the

synthetase activity.

The rate equation for the proposed mechanism A (ordered

bi-uni-uni-ter ping-pong) was derived assuming the products

P, Q, R and S were present, as shown below.
equation 16) 1/v = 1/V (I/[ABC]-[PQRS] (11[C] + 02[AB] + 03

[BC] + 04[AC] + 05[ABC] + 06[P] + 0|7[PQ] + 08[PS] + 09[PQR] +

010[PQS] + 011[SPR] + 012[SPQR] + 013[SQR] + I14[AP] + 015[CS]
+ 0I16[APQ] + 017[APB] + 0l8[ABQ] + 419[CRS] + 020[BCS] +

021[APQR] + 022[ABPQ] + 023[ABQR] + 024[BQRS] + 025[ABCQ] +

026[CQRS] + 027[ABCR] + 028[BCRS] + 029[ABPQR] + 030[BPQRS] +

031[ABCQR] + 032[BCQRSI))

Assuming Q, R, and S = 0, equation 16 becomes
equation 17) 1/v = 1/V (01/[AB] + 02/[C] + 03/[A] + 04/[B]
05 + 06[P]/[ABC] + 014[P]/[BC] +417[P]/[C])

Equation 17 predicts that P (PPi) is noncompetitive with

respect to ATP and L-aspartate and competitive with respect

to L-glutamine which are contrary to the experimental










results. PPi was found to be competitive with respect to ATP

(Fig. 3.21) and L-aspartate (Fig. 3.22). No definite

inhibition pattern was obtained versus L-glutamine, however.

Therefore, the AS mechanism is different from mechanism (A)

in which competitive inhibition of PPi versus L-glutamine is

the expected pattern.

Assuming P, R, and S = 0,
equation 18) 1/v = 1/V (01/[AB] + 02/[C] + 03/[A] + 04/[B]

+ 05 + 018[IQ]/[C])

Equation 18 predicts that Q (L-asparagine) is competitive

with respect to L-glutamine and uncompetitive with respect to

ATP and L-aspartate. L-asparagine was found to be competitive

with respect to L-glutamine (Fig. 3.18), however. It was

found to be noncompetitive with respect to ATP and L-

aspartate (Fig. 3.19 & Fig. 3.20).

Assuming P, Q, and S = 0,
equation 19) 1/v = 1/V (01/[AB] + 0I2/[C] + 03/[A] + 04/[B]

+ 05 + 027[R])

Equation 19 predicts that R (AMP) is uncompetitive with

respect to all the three substrates. These are different from

experimental results in that noncompetitive inhibition

patterns were observed (Fig. 3.26, Fig. 3.27. & Fig. 3.28)

Assuming P, R, and S = 0,
equation 20) 1/v = 1/V (01/[AB] + 02/[C] + 03/[A] + 04/[B]

+ 05 + 15[S]/[AB] + 020[S]/[A])

According to equation 20, L-glutamate would be competitive

with respect to ATP, noncompetitive with respect to L-










aspartate and uncompetitive with respect to L-glutamine.

However, L-glutamate was found to be uncompetitive,

competitive and noncompetitive with respect to ATP, L-

aspartate and L-glutamine, respectively (Fig. 3.23, Fig. 24,

& Fig. 3.25).

The rate equation for the mechanism B (ordered uni-uni-

bi-ter ping-pong) was also derived assuming the products P,
Q, R and S were present.
equation 21) 1/v = 1/V (I/[ABC]-[PQRS] (0l[PQR] + 02[PQ] +
03[P] + 04[BC] + 05[PC] + 06[APQR] + 07{PQRS] + 08[APQ] +

09[AP] + lo0[ABC] + 11[ACP] + 0l2[AQR] + I13[QRS] + 14[AQ] +

015[A] + 416[AC] + 017[ABCQR] + 018[BCQRS] + 019[CPQRS] +

020[ABQR] + 021[ABCQ] + 422[BQRS] + 023[PRS] + 024[ABCR] +

025[BCRS] + 026[CPRS] + 027[ABQ] + 028[PQS] + 029[PS] +

430[AB] + 03l[BCS] + 032[CPS]))

The product inhibition equation (equation 21), from which the

terms containing two or three products have been omitted, has

the expression:
equation 22) 1/v = 1/V (I/[ABC] (03[P] + 04[BC] + |5[PC] +

09[AP] + 010[ABC] + 0II[ACP] + 014[AQ] + 15[A] + 016[AC] +

021[ABCQ] + 024[ABCR] + 027[ABQ] + 030[AB] 031[BCS]))

assuming A= L-glutamine, B = ATP, C = L-aspartate, P = L-
glutamate, Q = PPi, R = AMP and S = L-asparagine.

The rate equation was also derived for the uni-uni-bi-

ter Theorell-Chance mechanism assuming the products P, Q, R

and S were present. The product inhibition equation from










which the terms containing two or three products have been

omitted, has the expression:
equation 23) 1/v = 1/V (I/[ABC] (03[P] + 04[BC] + 05[PC] +

09[AP] + 010[ABC] + 011[ACP] + 014[AQ] + 015[A] + 016[AC] +

024[ABCR] + 027[ABQ] + 030[AB] + 031[BCS]))

There is basically one difference between the predicted

patterns obtained from equation 22 (ordered uni-uni-bi-ter

ping-pong) and those from equation 23 (uni-uni-bi-ter

Theorell-Chance). According to the ordered mechanism

(equation 22), PPi should be noncompetitive with respect to

ATP, whereas according to the Theorell-Chance mechanism

(equation 23), it should be competitive with respect to ATP.

Although PPi was shown to be competitive with respect to ATP

(Fig. 3.21), other disagreements with Theorell-Chance

mechanism were found to exist.

The product inhibition studies did not allow us

unequivocally to rule out any of the proposed mechanisms.

Yet, they provided information as to where some of the

products can be placed in the scheme. L-asparagine was found

to be competitive with respect to L-glutamine (competing for

the same enzyme form), therefore they have to be next to each

other. PPi was competitive versus ATP and L-aspartate, which

can be placed following the addition of these two substrates.

To determine the product released between the addition of L-

aspartate and L-glutamine, initial velocity experiments were

performed where L-aspartate and L-glutamine were varied in

the presence of L-glutamate or PPi. The parallel lines in the










presence of L-glutamate (Fig. 3.29) suggests that L-glutamate

cannot be the first product (P) released between L-aspartate

and L-glutamine. However, given the fact that L-glutamate is

a poor inhibitor (Ki >> 20 mM), this can also suggest that L-

glutamate cannot bind tightly enough for the reverse reaction

to occur. Parallel lines were also observed in the presence

of PPi (Fig 3.30) which suggest that PPi cannot be the first

product off either. However, this can also suggest that PPi

binds the enzyme such that the reverse reaction can not take

place. In other words, it can not form the complex that is

catalytically competent for the reverse reaction. The double

inhibition studies were performed to determine the

relationship between the products, therefore trying to define

their release order. The results in Figure. 3.31 show that

the presence of L-asparagine enhances AMP inhibition. This

suggests that they combine sequentially with the enzyme,

therefore they must be next to each other (Segel, 1975) in

order. However, the presence of L-glutamate shows no

enhancement to either L-asparagine or AMP inhibition (Fig.

3.32 and 3.33), suggesting that their binding is separated by

another product or some other step.

Our data from initial velocity studies supported a bi-

uni-uni-bi ping-pong mechanism (mechanism A), for ammonia-

dependent reaction of ASB. Initial velocity, product

inhibition (single or double) studies were performed in an

attempt to come up with a mechanism, for glutamine-dependent

AS reaction. Initial velocity experiments were also performed