The Role of the glutamine-amidotransfer domain in the chemical mechanism of human asparagine synthetase

MISSING IMAGE

Material Information

Title:
The Role of the glutamine-amidotransfer domain in the chemical mechanism of human asparagine synthetase
Physical Description:
xiv, 231 leaves : ill. ; 29 cm.
Language:
English
Creator:
Sheng, Shijie, 1963-
Publication Date:

Subjects

Subjects / Keywords:
Aspartate-Ammonia Ligase -- isolation & purification   ( mesh )
Aspartate-Ammonia Ligase -- physiology   ( mesh )
Cysteine -- physiology   ( mesh )
Glutamine   ( mesh )
Asparagine -- isolation & purification   ( mesh )
Mutagenesis, Site-Directed   ( mesh )
Saccharomyces cerevisiae   ( mesh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1993.
Bibliography:
Includes bibliographical references (leaves 215-230).
Statement of Responsibility:
by Shijie Sheng.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 49645163
ocm49645163
System ID:
AA00011157:00001

Full Text

















THE ROLE OF THE GLUTAMINE-AMIDOTRANSFER DOMAIN IN
THE CHEMICAL MECHANISM OF HUMAN ASPARAGINE SYNTHETASE













BY

SHIJIE SHENG


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


1993















ACKNOWLEDGEMENTS

I want to thank my supervisor, Dr. Sheldon Schuster, for his patience,

support and confidence in me. I am especially grateful for his allowing me

so much freedom to pursue my research interests and for his being there

whenever I needed guidance and help. My sincere thanks and great respect

are also extended to the members of my supervisory committee for their

guidance through my graduate work: Dr. Daniel Purich, Dr. Ben Dunn,

Dr. Brian Cain and Dr. David Richardson. Thanks go also to Dr. Nigel

Richards and Dr. Don Allison for providing me, through many discussions,

with insights into the chemical and kinetic mechanisms of asparagine

synthetase. I am especially appreciative of their critical reviews of two

manuscripts included in my thesis.

My gratitude also goes to my friends and colleagues in Dr. Schuster' s

laboratory who helped and supported my research in numerous ways: Dr.

Gino van Heeke, Dr. Susan Hinchman, Mr. Richard Schnizer, Ms. Ellen

Wolworth, Dr. Lori Rice, Ms. Holly Grey and Ms. Pouran Tari. I also

want to thank Mr. Jamie Kraft for his skillful help on the development of

the colorimetric assay for asparagine. A special word of appreciation goes

to Dr. David Moraga for his friendship and his patient and fruitful

collaborations.















DEDICATION

To my parents, Yixiin Sheng and Yiishu Hou, who gave me the fortune

to be loved and the ability to love; who gave me the free spirit to dream and

the courage to make the dreams come true; who gave me my wonderful

sisters and little brother. Since I was born, I have never stood alone.

To my husband Daming Jiang, for his love, support and understanding

all the endurance is worthwhile.

To three special people, Uncle Guodong Zai, Professor Wei-Chin Liu

and Dr. Sheldon M. Schuster, who selflessly enlightened me with their

wisdom.














PREFACE

Portions of this dissertation have been published or have been

submitted for publication. Except for the Introduction and Conclusion

(Chapter 1 and 9), each chapter represents a manuscript and is therefore

written to be complete within itself.

Chapter 2: Shijie Sheng, David A. Moraga, Gino van Heeke, and

Sheldon M. Schuster (1992) Protein Expression and Purification 3: 337-

346

Chapter 3: David A. Moraga, Shijie Sheng, Gino van Heeke, and

Sheldon M. Schuster (1992) submitted for publication.

Chapter 4: Shijie Sheng, David A. Moraga, R. Donald Allison, and

Sheldon M. Schuster (1992) submitted for publication.

Chapter 5: Shijie Sheng, David A. Moraga, Nigel G. J. Richards, and

Sheldon M. Schuster (1992) submitted for publication.

Chapter 6: Shijie Sheng, Jamie J. Kraft, and Sheldon M. Schuster

(1992) submitted for publication.

Chapter 7: Shijie Sheng, and Sheldon M. Schuster (1992)

Biotechniques 13:704-708

Chapter 8: Shijie Sheng, and Sheldon M. Schuster (1992) J. Biol.

Chem. (in press)
















TABLE OF CONTENTS



ACKNOWLEDGEMENTS ................................................... ii

PR EFAC E .............................................. ................... iv

LIST OF TABLES ..................................................... Vii

LIST OF FIGURES........................................................ Xiii

ABBREVIATIONS ............................................................ Xi

ABSTRACT. ........................ ........... ......................... Xiii

CHAPTERS

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

An Historical Overview..................................... 1
Contemporary Closely Related Studies .................. 6

2 HIGH LEVEL EXPRESSION OF HUMAN ASPARAGINE
SYNTHETASE AND PRODUCTION OF MONOCLONAL
ANTIBODIES FOR ENZYME PURIFICATION.............. 13

Introduction...................................... ........ 13
Materials and Methods................................... 14
Results.................................................... 21
Discussion................................................ 34

3 CHARACTERIZATION OF CYS1 MUTANTS OF HUMAN
ASPARAGINE SYNTHETASE................................. 41

Introduction.................................................. 41
Materials and Methods................................... 43
R results .................................................... 49
Discussion................................................ 64

4 GLUTAMINE INHIBITS THE AMMONIA-DEPENDENT
ACTIVITIES OF TWO N-TERMINAL SUBSTITUTION
MUTANTS OF HUMAN ASPARAGINE SYNTHETASE
BY FORMING AN ABORTIVE COMPLEX .................. 72

Introduction ...... .......... ......................... ..... 72
Materials and Methods.................................... 76










4 Results..................................................... 78
Discussion........... ...................................... 96

5 EVIDENCE FOR AN IMIDE INTERMEDIATE IN THE
GLUTAMINE-DEPENDENT ASPARAGINE
SYNTHETASE REACTION................................ 108

Introduction............................................. 108
Materials and Methods.................................. 112
Results .................................................. 114
Discussion................................................ 124

6 A SIMPLE QUANTITATIVE COLORIMETRIC
ASSAY FOR L-ASPARAGINE.............................. 127

Introduction............................................. 127
Materials and Methods................................... 129
Results................................................... 135
Discussion............................................... 148

7 SIMPLE MODIFICATIONS OF A PROTEIN
IMMUNOBLOTTING PROTOCOL TO REDUCE
NONSPECIFIC BACKGROUND..............................159

Introduction ....................................... ....... 159
Materials and Methods...................................159
Protein Immunoblotting Protocols......................161
Results and Discussion ........................ ....... 162

8 PURIFICATION AND CHARACTERIZATION OF
Saccharomyces cerevisiae DNA-DAMAGE
RESPONSIVE PROTEIN 48 (DDRP 48 )................. 169

Introduction............................................. 169
Materials and Methods ................................... 172
Results ....................... .......................... 179
Discussion .............................................. 196

9 DISCUSSIONS AND FUTURE DIRECTIONS .................. 201
Discussions ................................................201
Future Directions........................................ 205

APPENDIX DERIVATION OF INITIAL VELOCITY EQUATIONS ..........209

BIBLIOGRAPHY .................... ............... ....................... 215

BIOGRAPHICAL SKETCH.................................................. 231















LIST OF TABLES


Tables Pae

2-1 rHAS purification from S. cerevisiae
BJ2168/pVTXAS 1 and AB 116/pBS24.1GAS........... 33

3-1 Activity, N-terminal amino acid identity and
pH optima of human asparagine synthetase
and Cys mutants purified from a S. cerevisiae
expression system..................... ................... 55

3-2 Apparent Km's (Kmapp) and Kcat for human
asparagine synthetase and Cys1 mutants................ 58

6-1 Enzyme activities of asparagine synthetases and
asparaginase. a comparison between the ninhydrin
colorimetric assay and HPLC amino acid analysis..... 151

8-1 Yeast Saccharomyces cerevisiae strains................. 174

8-2 Amino acid composition of "YP 75" compared with
that predicted for DDRP 48............................. 183

8-3 Hydrolysis of Nucleotides by DDRP 48 .................. 188















LIST OF FIGURES


Figures Page

1-1 Sequence alignment of the N-terminal domains of
several asparagine synthetases............................ 8

1-2 The mechanism for AS proposed according to
a currently accepted hypothesis.......................... 10

1-3 A mechanism for AS recently proposed by
Richards and Schuster ............ ...................... 11

2-1 Construction of pBS24. GAS............................. 23

2-2 Immunoblots of S. cerevisiae total cell extracts from
cultures grown in media containing various
carbon sources................................... .... ..... 25

2-3 Cross-reactivity of monoclonal antibodies against
rHAS with YP 75, E. coli Asn B and E. coli Asn A.... 28

2-4 Silver Staining of SDS-PAGE of purified rHAS........... 32

3-1 Yeast plasmids for the expression of Cys1
mutants of HAS........................................ 51

3-2 Immunoblots of total cell extracts from S. cerevisiae
containing plasmids with one of several HAS mutants..54

3-3 Effects of 6-diazo-5-oxo-L-norleucine (DON)
concentration on the activity of mutants C1A
and C1A, and wild-type (WT)...........................61

3-4 Effects of 6-diazo-5-oxo-L-norleucine (DON) on
the specific activity of mutant C1A and wild-type
(WT) versus incubation time............................ 62

4-1A Double-reciprocal plots of initial velocities of C1A
with ammonium acetate at various fixed
concentrations of glutamine .................................. 80

4-1B Double-reciprocal plots of initial velocities of CIS
with ammonium acetate at various fixed
concentrations of glutamine............................ 82











4-1C Replots of the slopes of the double-reciprocal plots
for C1A and for CIS versus glutamine concentration... 84

4-2A Double-reciprocal plots of initial velocities of C1A
with aspartate at various fixed concentrations of
glutam ine.................................................. 88

4-2B Double-reciprocal plots of initial velocities of CIS
with aspartate at various fixed concentrations of
glutamine.................................................. 90

4-3A Double-reciprocal plots of initial velocities of CIA
with ATP at various fixed concentrations of
glutamine.................................................... 93

4-3B Double-reciprocal plots of initial velocities of CIS
with ATP at various fixed concentrations of
glutam ine.............. ........ ............... ... .......... 95

4-4 Plots of enzyme initial velocities versus
ammonium acetate concentration........................ 98

4-5 Plots of enzyme initial velocities versus
ATP concentration..................................... 100

4-6 Proposed kinetic mechanism for the glutamine
effects on CIA and CIS and the initial velocity
equations derived from this mechanism ............... 104

5-1 Proposed mechanism for the glutamine-dependent
nitrogen transfer step in asparagine synthetase........ 110

5-2 Correlation of radioactivity and Rf values for
experimental conditions leading to the production
and hydrolysis of asparagine synthetase ............... 118

5-3 Formation and degradation of reaction intermediate
by CIS ......................................... .........121

5-4 Correlation of intermediate formation with glutamine
inhibition if the ammonia-dependent synthetase
activity of mutant CIS............................... 123

6-1 The UV-visible absorption spectra of the mixtures of
ninhydrin with asparagine, glutamine, glutamate and
aspartate.............. ................................... 137

6-2 The effects of the ninhydrin concentration on the
asparagine-ninhydrin reaction.......................... 140











6-3 The effects of temperature on the
asparagine-ninhydrin reaction......................... 142

6-4 The effects of pH on the asparagine-ninhydrin
reaction............................................ ...... 145

6-5 The time course of the production of A340 in the
asparagine-ninhydrin reaction ....................... 147

6-6 The standard curves of A340 versus asparagine
concentration................................... ........... 150

6-7 The proposed structure for the proline-ninhydrin
complexes and the asparagine-ninhydrin complex..... 154

7-1 SDS-PAGE of yeast proteins ..................................... 163

7-2 Western immunoblotting detection of rHAS
from total cell extract of yeast S. cerevisiae
BJ2168/pVTXAS 1...................................... 165

7-3 Western immunoblotting detection of the a-subunit
of yeast mitochondrial FiATPase....................... 167

8-la SDS-PAGE of DDRP 48 when the acrylamide
concentration was 10% .................................. 180

8-lb SDS-PAGE of DDRP 48 when the acrylamide
concentration was 15%................... ............... 181

8-2 The sequence of first 30 N-terminal amino acids
of "YP 75" compared with the N-terminal sequence
predicted for DDRP 48................................. 182

8-3 Enzymatic deglycosylation analysis of DDRP 48......... 185

8-4 Characterization of ATP and GTP hydrolysis
by DDRP 48..................... ............. ........... 190

8-5 Inhibition of DDRP 48-ATP hydrolysis activity
by nucleotides........................................ ... 191

8-6 Western blotting of DDRP 48 from untreated
S. cerevisiae strains with polyclonal antisera........ 193

8-7 Western blotting of DDRP 48 from untreated
S. cerevisiae strains and DDRP 48 from heat shock
or EMS treated S. cerevisiae strains................... 195

9-1 The structure of a potential inhibitor of AS............. 207















ABBREVIATIONS
ALL: acute lymphoblastic leukemia

E. coli : Escherichia coli

S. cerevisiae : Saccharomyces cerevisiae

AS: asparagine synthetase (s)

HAS: human asparagine synthetase

rHAS (or wtHAS): recombinant wild type human asparagine synthetase

C1A: mutant of rHAS with Cysl replaced by alanine

C1S: mutant of rHAS with Cys1 replaced by serine

C1Y: mutant of rHAS with Cys1 replaced by tyrosine

C1K: mutant of rHAS with Cys1 replaced by lysine

DC1: mutant of rHAS with Cys1 deletion

CI(A): mutant of rHAS with an alanine insertion between Cys1 and Gly2

C1G:G2C: mutant of rHAS with the position of the Gly2 and Cys1 being
reversed

DON: 6-diazo-5-oxo-L-norleucine

CONV: L-2-amino-4-oxo-5-chloro-pentanoic acid

DONV: 5-diazo-4-oxo-L-norvaline

GAT: glutamine amidotransfer

CHO: Chinese hamster ovary

FMDP: N3-(4-methoxyfumaroyl)-2, 3-diaminopropanoic acid

ADHI and II: S. cerevisiae alcohol dehydrogenase I and II

GAPp: promoter of the S. cerevisiae Glyceraldehyde
phosphatedehydrogenase gene











SDS-PAGE: sodium-dodecyl-sulfate polyacrylmide gel electrophoresis

PMSF: phenylmethyisulfonylfluoride

Mab: monoclonal antibody

EDTA: ethylenediamine-tetraacetic acid

EGTA: ethyleneglycol-bis-(b-amino-ethyl ether) N, N'-tetraacetic acid

DTT: Dithiothreitol

TBS: Tris buffered saline containing 10 mM Tris/ 0.9% NaCl(w/v), pH
7.2.

kD(a): kilodaltons

DIN: DNA-damage inducible genes

DDR: DNA-damage responsive genes

NQO: 4-nitroquinoline- I-oxide

MNNG: N-methyl-N' -nitro-N-nitrosoguanidine

EMS: ethylmethane sulfonate

PVDF: polyvinylidene difluoride

PNGase F: N-glycosidase F*

POGase: O-glycosidase*

IEF: Isoelectric focusing

ELISA: Enzyme-linked immunosorbent assay

OPA: o-Phthaldialdehyde

E. coli Asn A: E. coli asparagine synthetase A

E. coli Asn B: E. coli asparagine synthetase B















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

THE ROLE OF THE GLUTAMINE-AMIDOTRANSFER DOMAIN IN
THE CHEMICAL MECHANISM OF HUMAN ASPARAGINE SYNTHETASE

BY

SHIJIE SHENG

May, 1993
Chairman: Dr. Sheldon M. Schuster
Major Department: Biochemistry and Molecular Biology

Recombinant wild-type human asparagine synthetase (rHAS) is

overexpressed in yeast S. cerevisiae AB116/pBS24.1GAS and is purified

using the monoclonal antibody affinity chromatography procedure. As a

result, approximately 12.7 mg of pure rHAS is obtained from one liter of

yeast culture in a minimal medium. The purified rHAS has both the

glutamine-dependent and the ammonia-dependent asparagine synthetase

activities, as well as glutaminase activity. In order to study the role of the

N-terminal cysteine in the glutamine amidotransfer of asparagine synthesis,

a series of Cys1 mutants of HAS are created using side-directed

mutagenesis and are overexpressed in another yeast expression system.

Among these mutants only C1A, CIS and C1(A) can be purified to

homogeneity by the monoclonal antibody affinity chromatography

procedure. These purified stable mutants lack the glutamine-dependent

asparagine synthetase activities, but still have the ammonia-dependent

asparagine synthetase activity comparable to that of the wild-type enzyme.











It is found that the ammonia-dependent activities of C1A and CIS, as well

as the wild-type enzyme, are all inhibited by 6-diazo-5-oxo-L-norieucine

(DON), a glutamine analog and an alkylating agent. Detailed kinetic

analyses demonstrate that glutamine inhibits the ammonia-dependent

activities of C1A and CIS by forming an abortive complex. In experiments

on CIS, a possible reaction intermediate in the glutamine-dependent

asparagine synthesis of asparagine has been detected using DEAE anion-

exchange paper chromatography. The properties of this intermediate are

consistent with those of an asymmetric imide which has been recently

proposed as an intermediate in the glutamine-dependent activity of

asparagine synthetase.

In addition to the studies on the role of the glutamine amidotransfer

domain of HAS, a quantitative colorimetric assay for L-asparagine is

developed, an immunoblotting protocol is modified to reduce the

nonspecific backgrounds, and a yeast DNA-damage responsive protein,

DDRP 48, is purified and characterized.















CHAPTER 1
INTRODUCTION

An Historical Overview

The general interest in L-asparagine metabolism was greatly stimulated

by the discovery of the antitumor effects of L-asparaginase (Kidd, 1953a

and 1953b; Broome, 1961, 1963a and 1963b). Since then L-asparaginase

has been used to treat a wide variety of human neoplasms and is found to

be uniquely effective against acute lymphoblastic leukemia (ALL) (Dolowy

etal.,1966; Hill, 1967; Oettgen et al.,1967; Cooney and Handschumacher,

1970; Burchenal, 1970; and Hersh, 1971). The effect of L-asparaginase in

remission induction is often dramatic. When Escherichia coli (E. coli ) L-

asparaginase is used as a single-chemotherapeutic agent for ALL, complete

remission in 40-68% of the patients is achieved (Uren and

Handschumacher, 1977; Capizzi et al., 1971). Chemotherapy using a

combination of L-asparaginase, vincristine and prednisone results in a

remission rate of nearly 95% in previously untreated ALL (Uren and

Handschumacher, 1977; Ortega et al. 1977). It is found that the patients

receiving L-asparaginase at the end of the remission induction period had

greater disease-free survival than those receiving only vincristine and

steroid (Jones et al. 1977). Unfortunately, while L-asparaginase is a potent

and effective agent, three major drawbacks limit its general utility. First,

this therapy results in various kinds of adverse effects, ranging from

immunosuppression to liver and central nervous system dysfunction

(Haskell et al., 1969), anemia, life-threatening serum ammonia









2

concentrations, and hypoglycemia (Terebello et al., 1986). Although these

side effects are usually reversible and dose dependent (Crowther, 1971),

they often preclude use of the drug. Second, many patients who achieve

remission suffer a relapse with tumors resistant to further L-asparaginase

therapy (Terebello et al., 1986; Kiriyama et al., 1989; Lobel et al., 1979).

Finally, although 5-10 % of all solid tumors are sensitive to L-asparaginase

(Capizzi et al., 1971), the use of L-asparaginase remains confined to ALL,

because the enzyme enhances the growth of resistant tumors and increases

their metastatic activity (Tallal et al., 1970; Capizzi et al., 1971). Since the

effectiveness of L-asparaginase therapy is related to decreasing the

extracellular supplies of asparagine (Broome, 1961; Calabresi and Parks,

1980; Haskell et al., 1969; Uren and Handschumacher, 1977), an

alternative proposed approach has been to inhibit the biosynthesis of

asparagine. A prerequisite for this type of approach is to understand the

kinetic and chemical mechanism of the enzyme responsible, namely

asparagine synthetase (AS).

The first partial purification of AS was reported in 1962 (Ravel and

Norton, 1962). Since that time the biochemical and mechanistic properties

of AS have been investigated in a variety of species, including mouse

leukemia cells (Horowitz et al., 1968; Horowitz and Meister, 1972), rat

liver (Patterson and Orr, 1968 and 1969; Hongo et al., 1979), Chinese

hamster ovary cells (Gantt and Arfin, 1981), mouse pancreas (Milman and

Cooney, 1979), Lupine seedings (Rognes, 1970), guinea pig (Holcenberg,

1969), Streptococcus bovis (Burchall et al., 1964), E. coli (Cedar and

Schwartz, 1969a and 1969b), Saccharomyces cerevisiae (S. cerevisiae)








3

(Ramos and Wiame, 1979 and 1980), Neurospora (MacPhee, Nelson and

Schuster, 1983), corn root (Stulen and Oaks, 1977), soybean leaves

(Stewart, 1979), beef pancreas (Markin, Luehr and Schuster, 1981; Beran

and Schuster, 1981; Luehr and Schuster, 1985; Mehlhaff, Luehr and

Schuster, 1985; Mehlhaff and Schuster, 1991; Pfeiffer et al., 1986), rat

testicle (Hongo et al., 1992) and human (Van Heeke and Schuster, 1989a

and 1989b). Except for some prokaryotic AS such as E. coli AS A that only

catalyzed the ammonia-dependent asparagine synthesis reaction (Cedar and

Schwartz, 1969a and 1969b), most AS can catalyze the following reactions

(Patterson and Orr, 1968 and 1969):

(1) L-Asp + L-Gln + ATP -> L-Asn + L-Glu + AMP + PPi
(2) L-Asp + NH3 + ATP -> L-Asn + AMP + PPi

(3) L-Gln -> L-Glu + NH3

The molecular weight, subunit composition and distribution of AS

vary widely among different species and organs (Milman and Cooney,

1974; Hongo et al., 1992; Beran and Schuster, 1981; Hongo and Sato,

1983; Markin and Schuster, 1979). In mammalian cells, the abundance and

activity of AS also depend on the diet (Patterson and Orr, 1969; Hongo et

al., 1979) and the stage of animal development (Huang and Knox, 1975).

Some tumor cells exhibited little or no AS activity. In contrast, elevated

AS activity is commonly observed in L-asparaginase resistant cells derived

from L-asparaginase sensitive tumor cell (Patterson and Orr, 1968 and

1969; Gantt and Arfin, 1981; Kiriyama et al., 1989; Sugiyama, Arfin and

Harris, 1983). Some studies indicate that the expression of AS may be

regulated by gene amplification and methylation (Andrulis et al., 1983;











Andrulis, Evans-Blackler and Siminovitch, 1985; Andrulis, Argoza and

Cairney, 1990). Although no causal relationship between the AS gene

expression and the L-aspraginase sensitivity can be established, some

recent studies suggest that the expression and activity of AS is implicated

in the control of cell proliferation and differentiation (Colletta and Cirafici,

1992; Gong and Basilico, 1990; Saito et al., 1991; Hongo et al., 1990;

Hongo et al., 1989).

Despite the different kinetic mechanisms obtained for AS isolated from

different species (Cedar and Schwartz, 1969b; Hongo and Sato, 1985;

Markin, Luehr and Schuster, 1981; Milman, Cooney and Huang, 1980;

Mehlhaff, Luehr and Schuster, 1985), most investigators agree that

asparagine synthesis reaction involves a P-aspartyl-AMP intermediate. This

intermediate was first proposed by Cedar and Schwartz, based on the

observation that an 180 isotope was transferred from uniformly labeled L-

aspartate to the phosphate group of AMP, but not to the inorganic

pyrophosphate (Cedar and Schwartz, 1969b). Solid evidence for the
existence of the P-aspartyl-AMP intermediate was also obtained from the

studies of mouse leukemia AS (Horowitz and Meister, 1972) and beef

pancreatic AS (Luehr and Schuster, 1985).

Many early studies aiming at searching for potent AS inhibitors

involved screening a large number of compounds that bear some structural

resemblance to the substrates or the products of AS (Cooney et al., 1976,

1980; Mokotoff et al., 1975a, 1975b). Unfortunately only a few
compounds weakly inhibited AS in vitro. Among these inhibitors, 3-

aspartyl methylamide was moderately effective in increasing the survival of











mice carrying L-asparaginase resistant tumors (Uren, Chang and

Handschumacher, 1977). Several 3-aspartyl-AMP analogs prove to be

highly potent inhibitors but lack specificity (Zhukof, Biriukof and

Khomutov, 1988; Pike and Beefers, 1982). The observation that AS
requires a reducing agent such as 3-mercaptoethanol or dithiothreitol for

stability and activity (Hongo and Sato, 1983; Milman and Cooney, 1979)

has led to a hypothesis that AS has an essential sulfhydryl group that could

be modified (Pinkus and Meister, 1972; Cooney et al., 1976; Cooney et al.,

1980). Sulfhydryl modifiers are found to have various kinds of effects on

AS activity. Some sulfhydryl modifiers, such as maleimide, 6-diazo-5-oxo-

norleucine (DON) and L-2-amino-4-oxo-5-chloro-pentanoic acid (CONV),

inhibit the glutamine-dependent reaction of AS specifically (Pike and

Beefers, 1982). Other sulfhydryl modifiers, such as 5-diazo-4-oxo-L-

norvaline (DONV), are considered to be asparagine analogs and can inhibit

both the ammonia-dependent and the glutamine-dependent AS activity (Pike

and Beefers, 1982). Sulfhydryl modifiers, such as methionine

sulfonximine, inhibit the ammonia-dependent AS activity only (Stewart,

1979). It must be pointed out that even though some AS inhibitors have

antineoplastic activity, attempts to correlate antitumor potency and enzyme-

inhibition potency have been unsuccessful (Cooney et al., 1976). Many AS

inhibitors are homologous to asparagine or glutamine. Since both glutamine

and asparagine are of great importance for many metabolic pathways, these

inhibitors would affect both normal and transformed cells to a great extent

(Uren and Handschumacher, 1977).











Contemporary Closely Related Studies

The failure of previous attempts to develop potent AS inhibitors is

mainly due to a lack of knowledge about the detailed kinetic and chemical

mechanisms of the enzyme. One of the difficulties in studying the AS from

mammalian sources is isolation of sufficient amounts of pure enzyme. The

successful isolation of human asparagine synthetase (HAS) cDNA led to the

overexpression of recombinant HAS (rHAS) in L-asparagine-requiring

Jensen rat sarcoma cells (Andrulis, Chen and Ray, 1987), E. coli (Van

Heeke and Schuster, 1989a) and S. cerevisiae (Van Heeke and Schuster,

1990). A large quantity of rHAS can be obtained from the E. coii

expression system. Unfortunately, the overexpressed protein was insoluble

at the normal growth temperature of E. coli. More importantly, the N-

terminal methionine was still present on rHAS while this is not the case for

HAS and beef pancreatic AS. Soluble and properly processed rHAS can be

obtained from the yeast S. cerevisiae expression system. Previously,

monoclonal antibodies against beef pancreatic asparagine synthetase were

produced (Pfeiffer et al.. 1986) and used to make the monoclonal antibody

affinity chromatography columns for rHAS purification. Using this

procedure, approximately 1 mg of pure rHAS is obtained from 1 liter of

yeast minimal medium culture. SDS-PAGE revealed that the purified rHAS

had an approximate molecular mass of 64 kDa, which is in good agreement

with the calculated result: 64,437 Da (for 562 amino acid residues). The

rHAS obtained from S. cerevisiae has both the glutamine-dependent and

the ammonia-dependent asparagine synthetase activities (Van Heeke and

Schuster, 1990).











The ability of most AS, including HAS, to use either ammonia or

glutamine as the nitrogen source is also found in other well characterized

amidotransferases, such as glutamine phosphoribosylpyrophosphate

amidotransferase (Tso et al., 1982), CTP synthetase (Levitzki and

Koshland, 1971) and GMP synthetase (Zalkin and Truit, 1977). When the

protein sequences of AS from human (Andrulis, Chen and Ray, 1987),

Chinese hamster ovary (CHO) (Andrulis et al., 1989) Pisum sativum

(Tsai and Coruzzi, 1990), rat testicle (Hongo et al., 1992), and the enzyme

coded by E. coli asparagine synthetase B gene (Scolfield, Lewis and

Schuster, 1990) are compared, a highly conserved N-terminal sequence of

MetCysGlyILe is revealed (see Figure 1-1). According to Zalkin' s

definition, AS belongs to the purF-type of glutamine amidotransferase (Tso

et al., 1982). The purF-type glutamine amidotransferases have two

functionally distinct domains: one is capable of catalyzing the ammonia-

dependent reaction and is referred to as the synthesis domain; another

confers the capacity to utilize the glutamine amide and is referred to as the

glutamine amide transfer (GAT) domain. In AS, the GAT domain is located

at the N-terminus of the mature protein (Zalkin, 1992). It has been

confirmed that during asparagine synthesis, a (3-aspartyl-AMP intermediate

is formed, presumably in the synthesis domain (Walsh, 1979; Luehr and

Schuster, 1985; Pfeiffer et al., 1987). Very little, however, is known about

how the amide is transferred to this intermediate, either via free ammonia

or directly from glutamine.

Based upon site-directed mutagenesis studies of highly conserved

residues in the enzyme glutamine phosphoribosylpyrophosphate










8








Q Q I-
U2L



I I L Iu


,--- __ c,



Cu
S0










w a. >
I I -- -













So-
0000 1
C-







Se m
0 0
I22

S r" ,--
U* *
tn I .JG











amidotransferase (GPA), it has been shown that the conserved N-terminal

cysteine residue is essential for the glutamine-dependent nitrogen transfer

activity. A cysteine-histidine-aspartate catalytic triad working model for

purF-type amidotransferases has been proposed by Mei and Zalkin (1989)

(see Figure 1-2). In this model, the conserved N-terminal cysteine is used

to form a covalent cysteinyl-glutamine tetrahedral intermediate, which is
hydrolyzed in the following step to produce ammonia and a y-

glutamylthioester. The ammonia is then transferred to the synthesis
domain, while the y-glutamylthioester is hydrolyzed so that glutamate can

be released and the active thiol group of Cys can be regenerated. This

hypothesis is supported by the following facts. First, site-directed

mutagenesis has been used to demonstrate that the N-terminal cysteine is

essential for the glutamine-dependent activity of HAS (Van Heeke and

Schuster, 1989b). Second, glutamine analogs such as the alkylating agent

DON and N3-(4-methoxyfumaroyl)-2,3-diaminopropanoic acid (FMDP) can

inhibit the glutamine-dependent activities of purF-type amidotransferases

via either reversible inhibition or irreversible inactivation (Pike and

Beefers, 1982; Hartman, 1963; Badet et al., 1988). Third, the proposed y-

glutamyl-enzyme intermediates have been isolated (Levitzki and Koshland,

1971; Chaparian and Evans, 1991). Evidence supporting the

aforementioned mechanism, however, does not rule out other possibilities.

An alternative mechanism has been proposed by Richards and Schuster

(1992). According to the latter hypothesis, direct nucleophilic attack of the
primary amide of glutamine upon P-aspartyl-AMP is followed by release of

asparagine in a hydrolysis reaction catalyzed by Cyst (see Figure 1-3).













0

H r- NH, H

N 0 H S < l His1
CO2- N

Cys', O
-
I~ASDM


c

. NH, I "NHI


Asp


H
N Hiso2
N
H

2,


0

H S
H3N'-J : NS
1 'Cys, / j Hiso2
CO0z N


k" H
Asp2q


(b) 0 0

HO OH

HO OH


- NH-
.H*


0

H~j~ NH2
H3N._ C02-


Cys, fN His.
N



Asp"


+ H 2

H' transfer


0

H0 r0'



Cys1-
0*


H
N Hism

N
H



Asp


Figure 1-2: The mechanism for AS proposed according to a currently
accepted hypothesis (Mei and Zalkin, 1989)


* AMP
















coa-


Hi tranITl
-AMP
SH. H31


s"

Cyll'


AMP


0

S
H3NO'

C02


-Asn

+ HO20
O-I


H' "'H


0*



Hs3N Co

H-0


H'transr
-H*


0

H 0-
H3N' :



Cys1


0"


Asp"


Figure 1-3: A mechanism for AS recently proposed by Richards and
Schuster (1992)


H,0

0^
AaV
Aspr











This mechanism implies the formation of an imide intermediate in the

glutamine-dependent nitrogen transfer step.

This dissertation focuses on the role of GAT domain in the chemical

mechanism of human AS. The research includes the overexpression,

purification, basic characterization and detailed kinetic analysis of both

wide type HAS (wtHAS) and several site-specific mutants of rHAS.

Evidence for the imide intermediate in the glutamine-dependent asparagine

synthetase reaction proposed by Richards and Schuster (1992) was

obtained. Several AS assays, including a novel ninhydrin colorimetric

assay, were developed to meet the various needs in this study. For routine

protein analysis, a protein immunoblotting protocol was modified to reduce

the nonspecific background. In addition, the attempt to purify human AS

led to the unexpected purification and characterization of a yeast DNA-

damage responsive protein.















CHAPTER 2
HIGH LEVEL EXPRESSION OF HUMAN ASPARAGINE SYNTHETASE AND
PRODUCTION OF MONOCLONAL ANTIBODIES FOR ENZYME PURIFICATION

Introduction

L-Asparagine synthetase (EC 6.3.1.1) catalyzes the synthesis of L-

asparagine by the following reaction:

L-Asp + ATP + NH3 (or L-Gln) ----> L-Asn + AMP +PPi (or + L-Glu)

Asparagine synthetases that can utilize L-glutamine as the ammonia

source for asparagine synthesis are also found to be able to catalyze L-

aspartate- and ATP-independent glutamine hydrolysis producing glutamate

and ammonia (Horowitz and Meister, 1973). No detailed mechanisms for

any of these reactions have been reported. The investigation of the kinetic

and chemical mechanisms of human asparagine synthetase became of

special clinical interest since the discovery of the antitumor effects of L-

asparaginase (Broome, 1963b: Dolowy et al., 1966). Since L-asparaginase

sensitive tumors have low or no asparagine synthetase activities (Broome

and Schwartz, 1967; Horowitz et al., 1968), it is thought that the specific

inhibition of asparagine synthetase could be an alternative approach to the

treatment of these types of tumors. At present, however, no specific

inhibitors of asparagine synthetase exist, except for monoclonal antibodies.

Recombinant human asparagine synthetase (rHAS) has been cloned and

expressed in Saccharomyces cerevisiae (Van Heeke and Schuster, 1990),

partly in an attempt to obtain sufficient amounts of pure enzyme for

detailed kinetic studies. In this yeast expression system, rHAS is











constitutively produced in the presence of glucose. N-terminal protein

sequencing indicated that the enzyme is properly processed in the yeast

host. Immunoaffinity purification of the enzyme was initially accomplished

by using monoclonal antibodies made against beef pancreatic asparagine

synthetase. The optimal recovery of pure enzyme was 1 mg per liter of

yeast culture, which accounted for less than 0.5% of the total yeast extract

protein. The low abundance of rHAS in the crude yeast extract raised a

problem of high nonspecific background in the affinity column purification.

Lengthy washing of the column prior to enzyme elution was required to

remove proteins binding nonspecifically, thus also lowering the yield and

the specific activities of the enzyme.

In this paper, a new rHAS expression system is presented. The

production of monoclonal antibodies to rHAS and a one-step enzyme

purification using an immunoaffinity procedure are also described. This

resulted in the production of more, extremely pure rHAS than ever before

possible.

Materials and Methods
Strains and Plasmids

E. coli K-12 strain XL1-Blue {recAl, endAl, gyrA96, thi-1, hsdR17,

supE44, relAl, lac, [ F' proAB, laclqZDM15, TnlO(tetr) ] } was obtained

from STRATAGENE (La Jolla, CA). Cells were cultured in Luria broth

(Sambrook, Fritsch and Maniatis, 1982) containing tetracycline (15

gg/mL). Ampicillin (50 gg/mL) was also added to the medium for selection
when cells had been transformed with plasmids containing an appropriate

marker. Plasmid pVTXAS 1 is a recombinant human asparagine synthetase











(rHAS) expression vector that had been constructed from the vector

pVTO11U (Vernet, Dignard and Thomas, 1987) for previous studies in this

laboratory. Vector pBS 100.1 (Barr et al., 1987) is a pBR322 derived

plasmid for the fusion of genes directly to the glucose-regulatable alcohol

dehydrogenase II /glyceraldehyde-3-phosphate dehydrogenase promoter

(ADH II/GAPp) (Barr et al., 1987; Cousens et al., 1987). pBS24. 1 is an

E. colilS. cerevisiae shuttle vector (Sabin et al., 1989) containing the

pBR322 sequence for amplification in E. coli. sequences for autonomous

replication in yeast, and the selectable markers for the growth of

transformed yeast in uracil and/or leucine deficient media. Saccharomyces

cerevisiae strain BJ2168 (a, prc 1-407, prbl-1122, pep4-3, leu2, trpl,

ura3-52) was obtained from the Yeast Genetic Stock Center (Berkeley,

CA) and was used for expression of rHAS from the pVTXAS1 plasmid. S.

cerevisiae strain AB116 (a, leu2, irpl, ura3-52, prbl-1122, pep4-3, prcl-

407, ciro) was kindly provided by Dr. Phil Barr (Chiron Corp.) and was

used for rHAS expression from the pBS24.1GAS plasmid.

Chemicals and Reagents

DNA restriction enzymes were obtained from either PROMEGA

(Madison, WI ) or New England Biolabs (Beverly, MA). T4 DNA ligase

(2000 U/ml) was purchased from New England Biolabs. T4 DNA

polymerase and calf intestinal phosphatase (CIP) were purchased from

PROMEGA. Yeast and bacteria culture media were purchased from DIFCO

(Detroit, MI). Cyanogen bromide activated Sepharose-4B resin and PD-10

columns were purchased from Pharmacia (Piscataway, NJ). Gel

electrophoresis reagents and protein concentration assay dye were









16

purchased from Bio-Rad (Melville, NY). HPLC reagents were purchased

from Fisher (Orlando, FL). Glass beads and all other chemicals were

obtained from SIGMA (St. Louis, MO). The purified asparagine

synthetases A and B from E. coli were available in our laboratory.

Construction of the Overexpression Plasmid pBS24.1GAS

The outline for the construction of pBS24.1GAS is shown in Figure 2-

1. The Xbal/Xhol fragment from pVTXAS1 and the Ncol/Sall digestion

product from pBS 100.1 were ligated using T4 DNA ligase. Ligation

reactions were performed in low melting (LM) point agarose (FMC

Bioproducts, Rockland, ME) as follows: The desired fragments from 0.8%

LM agarose gels were excised and placed in separate microcentrifuge tubes.

The gel slices were melted at 65 *C and cooled to 37 *C prior to the

ligation. The reactions were started by mixing molten agarose containing

the bands of interest, 9 4l of 10X T4 ligase buffer, 6 units of T4 DNA

ligase and enough water to a final volume of 90 gL. This mixture was

quickly vortexed and kept at room temperature for 12-15 hours. The

resulting rHAS plasmid DNA was extracted with phenol/chloroform at 65'C

and precipitated with ethanol/sodium acetate. The DNA pellets were

resuspended in water and the entire volume was used to transform E. coli

XL1-Blue cells. Plasmids with the rHAS coding sequence in the desired

orientation (pBS 100.1HAS) were identified by restriction enzyme analysis.

The Sall/partial BamHI fragments from pBS100. 1HAS were ligated to the

SalllBamHI sites from pBS24.1 as described above in order to obtain

pBS24.1GAS.











Transformation

S. cerevisiae cells were transformed using a standard lithium acetate

procedure (Treco, 1987). E. coli transformation was performed according

to the method of Hanahan (1983).

Yeast Cultures and Protein Expression

S. cerevisiae BJ2168 containing plasmid pVTXAS1 (designated as

BJ2168/pVTXAS1) was cultured in 10 mL or 1000 mL uracil deficient

minimum media with 2% glucose as previously described (Van Heeke and

Schuster, 1990). Single colonies of S. cerevisiae AB116 transformed with

pBS24. GAS (designated as AB116/pBS24. IGAS) were picked from a

uracil/leucine omission plate to innoculate 10 mL uracil/leucine deficient

liquid media. The carbon source was one of many nutrients tested as

substrates for yeast growth and expression of rHAS (see Figure 2-2). The

cultures were incubated at 30 'C with vigorous shaking for 40 hours. For

large-scale cultures of AB 116/pBS24. 1GAS, 1000 mL uracil and leucine

deficient media containing 8% of D-galactose were innoculated with 10 mL

fresh AB 16/pBS24.1GAS cultures in the same media at an OD600 of

approximately 3.0. The OD600 values of the large-scale AB 116/pBS24.1

GAS cultures at late-log phase were in the range of 5.8-6.2.

Small-scale cell harvesting and protein extractions were performed as

previously described (Van Heeke and Schuster, 1990). The cell extract

proteins were separated by SDS-PAGE and electroblotted to a

polyvinylidene fluoride (Bio-Rad) membrane using a Bio-Rad western blot

apparatus. The rHAS was detected by immunoblotting coupled with the

alkaline phosphatase color reaction (Towbin, Staehelin and Gordon, 1976).











Large-scale cell harvesting and protein extractions were performed as

follows: The cells were centrifuged at 3000 xg and 4 'C for 5 minutes and

washed once with cell lysis buffer (50 mM Tris-HCI (pH 8.0), 0.5 mM

EDTA, 1 mM EGTA, 1 mM DTT, 1 mM PMSF, 1 mM 3-aminobenzamidine).

The cell pellets were resuspended in 35 mL cell lysis buffer, and the entire

volume was transferred into the 70 mL chamber of a Bead-Beater (Biospec

Products, Bartlesville, OK). Cold, acid-washed glass beads (0.45-0.50

mm in diameter) were added to fill the rest of the space, and the chamber

was installed in a melting ice bath. After a short burst of 2 seconds, a 15

second blending was applied to the cell suspension followed by 5 minutes

of cooling. The burst-blending-cooling process was repeated 5 more times.

The glass beads and the homogenate were separated by decanting the

homogenate into ultracentrifuge tubes. Thirty ml fresh cell lysis buffer

was added to the glass beads to rinse out as thoroughly as possible the

remaining cell homogenate. The two portions of the homogenate were

combined and centrifuged at 30,000 xg and 4 'C for 90 minutes. The clear

cell extract supernatant fluid was carefully collected and immediately

loaded onto the antibody affinity columns.

Production of Monoclonal Antibodies to rHAS

Production of monoclonal antibodies against rHAS was performed in

the Hybridoma Core Facility of the Interdisciplinary Center for

Biotechnology Research at the University of Florida.

rHAS was first purified from S. cerevisiae BJ2168/pVTXASl as

described by Van Heeke and Schuster (Van Heeke and Schuster, 1990).

Female BALB/c mice were immunized with the purified rHAS using a









19

procedure derived from the method of Kohler and Milstein (Kohler and

Milstein, 1975). Hybridomas were produced by fusion of the spleen cells

of the immunized mice and SP 2/0 myeloma cells in the ratio of 7:1

(Kohler, Howe and Milstein, 1976). The positive hybridomas growing in

the hypoxanthine/aminopterin/thymidine selection media were screened by

antibody capturing ELISA and ImmunoDot Blotting tests with purified

rHAS. The selected hybridomas were cloned by the limiting dilution

technique and screened again by antibody capturing ELISA tests with

purified rHAS. The ascites from the BALB/c mice that had been infected

with the antibody producing monoclonal hybridomas were obtained and

delipidized as follows: equal volumes of the ascites and 1,1,2-

trichlorotrifluoroethane were mixed by magnetic stirring for 20 minutes at 4

'C. Following a centrifugation at 3000 xg and 4 *C for 30 minutes, the top

layer supernatant fractions were centrifuged again at 20,000 xg and 4 'C

for 30 minutes to remove the debris. The lipid-free ascites were stored

frozen at -20 'C until needed.

Immunoaffinitv Columns

Monoclonal antibodies to rHAS were partially purified from the fat-

free ascites that had been diluted with 2 volumes of PBS. The protein

pellet from a 45% saturation ammonium sulfate precipitation was

redissolved in PBS to the starting volume and reprecipitated with 50%

saturated ammonium sulfate. The precipitated antibodies were redissolved

in PBS using one-tenth the original ascites volume and dialyzed thoroughly

against 4 changes of PBS during 24 hours at 4 'C. The covalent coupling

of the monoclonal antibodies to cyanogen bromide activated Sepharose-4B









20

was performed according to a standard procedure from the manufacturer

(Pharmacia) except that the gel was swollen in 15 mM Tris-HCl buffer (pH

7.5) instead of 1 mM HCI. The remaining active sites on the gel were

blocked by ethanolamine (pH 8.0) at a final concentration of 0. 1 M.

Immunoaffinity Purification of rHAS

The monoclonal antibody affinity column was equilibrated with 10

column volumes of PBS (pH 7.2). The yeast cell extract was allowed to

flow through the column at a rate of 0.25 mL/min. The column was

washed with 15-20 column volumes of PBS at a flow rate of 1 mL/min until

the protein concentration of the run-through PBS was less than 0.2 mg/mL.

rHAS was eluted with 0.1 M Na2CO3 (pH 10.6) at a flowrate of 0.5

mL/min. The protein was precipitated immediately in 60% saturated

ammonium sulfate and redissolved in a small volume of enzyme buffer (50

mM Tris-HCl [pH 8.01, 0.5 mM EDTA, 20% [v/vl glycerol). Purified

rHAS was desalted on a PD-10 column and stored at -70 'C in enzyme

buffer at a final concentration of approximately 1 mg/mL.

Enzyme Activity Assay Methods

Asparagine produced in the rHAS catalyzed reactions was measured by

HPLC amino acid analysis on an Applied Biosystem 420A derivatizer and

130A separation system. The enzyme reaction conditions were as

described by Luehr and Schuster (Luehr and Schuster, 1985) except that

the total volume of the reaction mixtures was reduced to 160 pL. The

reactions were terminated by the addition of trichloroacetic acid to a final

concentration of 5% (w/v). Each reaction mixture was filtered through a

0.2 p.m nylon membrane before loading on the HPLC. The conditions for











amino acid analysis on the HPLC were those recommended in the standard

procedure provided by the manufacturer.

The glutaminase activity assay was performed as described by Luehr

and Schuster (Luehr and Schuster, 1985).

Protein concentrations were determined by the Bio-Rad dye method

and using mouse immunoglobulin G to construct a standard curve. SDS-

PAGE and silver staining of protein were performed according to a

standard procedure. Protein isoelectric focusing electrophoresis on Phast

gels was carried out by the Protein Chemistry Core Facility of ICBR at the

University of Florida. Western blots of rHAS with monoclonal antibodies

were obtained using a procedure derived from the method of Towbin et al.

(1976).

Results

Construction of pBS24.1GAS

The construction of plasmid pBS24.1GAS is outlined in Figure 2-1.

The rHAS coding region, obtained by restriction enzyme digestion of

pVTXAS 1, was fused directly to the alcohol dehydrogenase II/

glyceraldehyde-3-phosphate dehydrogenase promoter (ADHII/GAPp)

sequence from plasmid pBS 100.1. The resulting plasmid containing the

rHAS gene in the clockwise orientation (pBS100. 1AS) was identified by

restriction enzyme analysis. The DNA fragment containing the

ADHII/GAPp-rHAS sequence was obtained by restriction enzyme digestion

of pBS 100. 1AS with Sail (complete) and with BamHI (partial). This

fragment was ligated to the corresponding sites in the pBS24.1 plasmid.

Since the insertion of ADHII/GAPp-rHAS sequence does not alter the

































Figure 2-1: Construction of pBS24. IGAS (see Materials and
Methods for a detailed description). Only the relevant restriction
enzyme sites are shown on the plasmid maps. The arrow indicates
the orientation of the gene. Abbreviations: rHAS, recombinant
human asparagine synthetase; AP, yeast ADHI promoter; A3', 3'
portion of the ADHI gene; GAPp, Alcohol dehydrogenase II
/Glyceraldehyde-3-phosphate dehydrogenase promoter; Gt, GAPp
terminator; af, a-factor terminator; T4 pol, T4 DNA polymerase; CIP,
Calf intestinal phosphatase; Sp, Sph I; Xb, Xba I; Ba, BamH I; Xh,
XhoI; Nc, Nco I;Sa, Sal I.


















S Xb Ba Xh Sp
I is I I I
ap oHAS
amp Eco ori ura 2u onr


pBS100.1

SNcol/Sall

T4 Pol

CIP


T4 DNA Ligase


Ba

GAmp
amp


Eco on amp
2u
2u leu2-d

pBS24.1

SSall/BamHI


HAS
Eco on

pBS100.1AS


Sa

Gi
83


Sall/partial BamHI
J


T4 DNA Ligase


Sa
L.


Ba Ba
I I


al HAS GAPp Eco ori amp
2u 2u
ura3 2u leu2-d

pBS24.1GAS


pVTXAS1


Xbal/Xhol


T4 Pol
I


Sa Ba


2u
ura3


)


I


I


W











essential features of pBS24.1, the resulting plasmid, pBS24.1GAS, can

propagate in S. cerevisiae ciro strain AB116 at a high copy number. The

presence of the ADHII promoter regulatory sequence upstream from the

GAP promoter allows for glucose repressible expression of rHAS; i.e.,

when glucose is depleted from the culture media, ADHII/GAPp is

derepressed and the transcription of the rHAS gene ensues.

Expression of rHAS from AB 16/pBS24.1GAS

In order to optimize the expression of rHAS in S. cerevisiae

AB116/pBS24. 1GAS, various monosaccharides were tested as the

carbon source in the culture media. Equal amounts of the cell extract

proteins obtained from 40 hour cultures of AB116/pBS24.1GAS were

electrophoresed and immunoblotted with anti-rHAS monoclonal

antibody 1C8. The results are shown in Figure 2-2. The highest

expression of rHAS was observed when either 8% D-galactose (Lane

8), 8% succinate (Lane 5), or 8% D-mannitol (Lane 3) was the carbon

source. Slightly less production of rHAS was observed when the

medium was supplemented with 1% glucose (Lane 7). An intermediate

amount expression was obtained when 8% glycerol or 8% galactose

plus 1% glucose were added as the carbon source (Lanes 2 and 6).

Although the yeast cells grew well in both media, no rHAS was

detected in cell extracts when 8% fructose (Lane 4) or 8% glucose was

used as the carbon source. It appears that some fructose is converted

to glucose in the gluconeogenesis pathway of yeast cells. These

observations were consistent with the mechanism by which

ADHII/GAPp is regulated. Galactose (8%) was the carbon source of














1 2 3 4 5 6 7 8


Figure 2-2: Immunoblots of S. cerevisiae total cell extracts from
cultures grown in media containing various carbon sources. Three
micrograms of immunoaffinity purified rHAS (lane 1), 84.8 pg of
AB116/pBS24.1GAS total cell extract from 40 hours cultures (lanes 2-
8), 8% glycerol (lane 2), 8% mannitol (lane 3), 8% fructose (lane 4),
8% succinate (lane 5), 1% glucose plus 8% galactose (lane 6), 1%
glucose (lane 7), 8% galactose (lane 8). Arrows on the left indicate the
positions of the molecular mass markers; from the top: phosphorylase
B (106 kDa), serum albumin (80 kDa), ovalbumin (49.5 kDa),
carbonic anhydrase (32.5 kDa). See Materials and Methods for further
details.











choice for large scale rHAS production because it sustained the cell growth

to the highest late-log phase cell density (OD600: 5.8-6.2) while allowing

derepressed overexpression of rHAS. Although succinate and mannitol

also allowed high expression of rHAS, they were not good substrates for

cell growth. Culture OD600 readings of only 0.5-0.6 were reached after

60-hour incubation using these compounds as nutrients.

The expression of rHAS in S. cerevisiae AB 16/pBS24.1GAS was

compared with that in S. cerevisiae BJ2168/pVTXAS1 under optimal

culture conditions by western blotting analysis. Even one-tenth the amount

of protein from the AB116/pBS24.1GAS cell extract resulted in a more

intense band than that observed for extracts of BJ2168/pVTXAS1 (data not

shown).

Production of Monoclonal Antibodies to rHAS

rHAS used for the immunization of the BALB/c mice and for the

screening of the antibody producing hybridomas was purified from S.

cerevisiae BJ2168/pVTXASI using an immunoaffinity column procedure

with monoclonal antibody against beef pancreatic asparagine synthetase

(Van Heeke and Schuster, 1990). The protein was approximately 95%

homogeneous as shown by silver-stained SDS-PAGE. The glutamine-

dependent and the ammonia-dependent asparagine synthetase specific

activities of rHAS were in the range of 50-150 nmoles/min.mg.

The spleen cells from the immunized mice were fused with SP2/0

myeloma cells in order to produce hybridomas. The supernatant fluids of

the resulting mass culture were screened by an ELISA and an ImmunoDot

Blotting test with purified rHAS. The results were identical by either test.











Eleven culture supernatant fluids demonstrated moderate to high immune

responses to rHAS. The hybridomas from each one of these eleven mass

cultures were expanded and single-cell cloned and the resulting monoclonal

hybridomas with the highest ELISA titer to rHAS were selected to produce

large amount of monoclonal antibody (Mab) in the form of ascites. The

monoclonal antibodies from the delipidized ascites without further

purification were diluted 1000-fold and used in an antibody capturing

ELISA test to examine the cross-reactivities of each Mab with an S.

cerevisiae 75 kDa protein often co-purified with rHAS (YP75, see below)

and E. coli asparagine synthetases A (AsnA) and B (AsnB). As a positive

control, proteins were reacted with a mouse antiserum against rHAS at a

1000-fold dilution. Wells containing protein and ascites fluid from

nonimmunized HL4 mice at a 1000-fold dilution were used as negative

controls. The ELISA results are illustrated in Figure 2-3. Mab 3G6 had the

highest titer against rHAS and also the highest specificity for this protein.

No significant cross-reactivity of any Mab with E. coli AsnA or AsnB was

observed. Since all the Mabs as well as the positive and negative controls

gave rise to detectable cross-reactions with YP75, it is likely that such

cross-reactivity was caused by nonspecific protein interactions due to some

unique features of YP75. It is not clear, however, why Mab 4A2 had even

higher affinity for YP75 than for rHAS. The ELISA analysis was repeated

with several other dilutions of Mab 4A2. Similar results were obtained

(data not shown). One explanation for this particular phenomenon could be

that Mab 4A2 was contaminated with antibodies to YP75 that was originally

co-purified with rHAS.






























1'


1 0

0.9

0.8

0.7

06

0.5

04

03

0.2

0.1 .i

0.0 ,
Polt- iL4 2C8 IC8 411l 314 4A2 11i 319 5G11 3G6 2G4 5A8


Figure 2-3: Cross-reactivity of monoclonal antibodies against rHAS
(i) with YP75 ( = ), E. coli AsnB ( SB ), E. coli AsnA (ES ). 0.1
gg of protein was coated on the 96-well ELISA plates. Monoclonal
antibodies in the form of ascites were diluted 1000-fold with 1% BSA in
PBS buffer. A polyclonal antiserum raised against pure rHAS was diluted
in the same manner and used as a positive control. The ascites of
nonimmunized HL4 mice was also diluted 1000-fold and used as a negative
control.











Immunoaffinity Columns

Since cross-reactivity analysis showed that Mab 3G6 had the highest

specificity for rHAS, it was therefore chosen for the preparation of

immunoaffinity resins. After a partial purification by ammonium sulfate

precipitations, Mab 3G6 from the mouse ascites had approximately 80%

gamma globulin. A total of 120 mg partially purified Mab 3G6 protein was

covalently coupled to 12 ml swollen cyanogen bromide activated

Sepharose-4B gel by gentle rotation at 4 *C for 20 hours. A protein

concentration assay indicated that 97% of the protein had been coupled to

the gel. Ethanolamine was added to terminate the coupling reaction and

block the remaining active sites. The resulting gel was packed in four

different plastic columns for immunoaffinity purification of rHAS. The

binding capacity of each column was 2.0-2.5 mg rHAS.

When insufficient amounts of anti-HAS Mab were coupled to the

immunoaffinty resin, a 75 kDa S. cerevisiae protein (YP75) was co-

purified with rHAS. YP75 could be purified from the cyanogen bromide

activated column that had been treated with ethanolamine alone (Chapter 8).

It was therefore apparent that the co-purifiaction of YP75 with rHAS was

the result of strong interactions between the yeast protein and

ethanolamine, used to block the remaining sites of the immunoaffinity

resin.

Yeast Protein Extraction and rHAS Immunoaffinity Purification

S. cerevisiae BJ2168/pVTXAS1 and AB116/pBS24.1GAS were

harvested from 1 liter cultures with OD600 readings of 4.0 and 5.8

respectively. The homogenization with glass beads on the Bead Beater for









30

90 seconds resulted in 80-85% cell disruption. After ultracentrifugation,

the cell extract from S. cerevisiae BJ2168/pVTXAS 1 homogenate contained

a total of 324 mg protein, while 220 mg of cell extract protein was obtained

from the homogenate of S. cerevisiae AB 16/pBS24.1GAS.

A total of 81 mg protein extract from S. cerevisiae BJ2168/pVTXAS1,

equivalent to 250 mL liquid culture, was loaded on a 3 mL 3G6 antibody

affinity column. Sixty-five milliliters of PBS buffer (pH 7.2) was used to

remove the nonspecific binding proteins from the column. At this point,

0.1 M Na2CO3 (pH 10.6) was applied to the column and 3 mL fractions

were collected. Fractions 2 and 3, containing most of the rHAS, were

combined. Solid ammonium sulfate was directly added to 60% saturation

in order to precipitate rHAS. Desalting was accomplished by passing the

protein through a PD-10 size exclusion column. Eighty-one mg of

BJ2168/pVTXAS cell extract yielded 0.5 mg of pure rHAS.

rHAS in S. cerevisiae AB116/pBS24. 1GAS extract was purified in a

similar manner except that only 22 mg cell extract proteins, equivalent to

100 mL liquid culture, were loaded on the same 3G6 affinity column.

Forty-five milliliters of PBS buffer (pH 7.2 ) was required to wash the

column. The protein eluted from the column with Na2CO3 was

predominantly collected in fractions 2 and 3. A total of 1.27 mg rHAS

was obtained after the ammonium sulfate precipitation and PD-10 column

desalting.

Protein Gel Electrophoresis

The rHAS purified from S. cerevisiae AB116/pBS24.1GAS was

analyzed by SDS-PAGE. The silver stained gel (shown in Figure 2-4)









31

demonstrated that rHAS purified with 3G6 immunoaffinity column was

homogeneous. The subunit molecular mass of rHAS is 64 kDa which is

identical to that of the rHAS purified from S. cerevisiae BJ2168/pVTXAS1

(Van Heeke and Schuster, 1989b), suggesting that no major differences in

post translational modifications exist between the two different yeast

expression systems. The purified rHAS was also subjected to

electrophoresis on a protein isoelectric focusing gel (IEF). The silver

stained IEF gel exhibited only one band with an isoelectric point of 7.5

(data not shown).

Enzymatic Activities


The asparagine synthetase activities of both the yeast cell extracts and

the purified rHAS were measured by HPLC amino acid analysis. The

results are listed in Table 2-1. rHAS purified from S. cerevisiae

BJ2168/pVTXAS1 had an ammonia-dependent specific activity of 93.5

nmoles Asn/min.mg, while its glutamine-dependent asparagine synthetase

specific activity was 65.7 nmoles Asn/min.mg. These results are similar to

an earlier report when the monoclonal antibodies to beef pancreatic

asparagine synthetase were used for immunoaffinity purification of rHAS

from S. cerevisiae BJ2168/pVTXAS1 (Van Heeke and Schuster, 1990).

The rHAS purified from S. cerevisiae AB116/ pBS24.1GAS had

significantly higher specific asparagine synthetase activities. The

ammonia-dependent and the glutamine-dependent specific activities were

1,870 nmoles Asn/min.mg and 937 nmoles Asn/min.mg respectively. The

glutaminase specific activity of rHAS purified from S. cerevisiae

AB116/pBS24. 1GAS was found to be 1,250 nmoles/min mg, as determined
























0 r


Figure 2-4: Silver staining of SDS-PAGE. Lane 1, Bio-Rad molecular
mass standards, from top to bottom: phosphorylase B (97.4 kDa), serum
albumin (66.2 kDa), ovalbumin (45.0 kDa), carbonic anhydrase (31.0
kDa). Lane 2: 75 ng of rHAS purified from S. cerevisiae
AB116/pBS24.1GAS.


*





























0 r
a <


C .
IIIEI


I


Iii


IC







ii

4 I) I

o









c u 8


*' e .i u uOC


rA
%25-
'0 E



1 m^
E

1c


P.


'5 S
;-,
0.
.C

0 ,,


EE
'5
C 6




SE
,EC











by the OPA derivatization method.

Discussion
Significant efforts have been made in the past to investigate the kinetic

mechanisms of asparagine synthetases from various species. However, the

detailed kinetic and structural studies were greatly limited by the

availability of sufficient amounts of pure enzymes. No detailed

mechanisms have been proposed. With the advances of molecular cloning

technology and the application of sophisticated protein purification methods

such as immunoaffinty column chromatography, large quantities of pure

enzymes have now become a reality.

High yield of recombinant human asparagine synthetase (rHAS) was

first obtained from an E. coli overexpression system (Van Heeke and

Schuster, 1989b). Unfortunately, a large fraction of the enzyme expressed

was insoluble when the E. coli cells were cultured at temperatures higher

than 300C. Furthermore, the enzyme isolated from E. coli still had the N-

terminal methionine residue which was not found in the native form of the

protein. Later, rHAS was cloned in the eukaryotic host S. cerevisiae by

Van Heeke and Schuster (1990). The resulting expression system did not

have the problem of protein solubility. The N-terminal methionine was

removed from the recombinant protein in a similar manner to the native

protein. However, the low production of rHAS (< 0.5% of the total cell

extract protein) was a major drawback and presented a problem for the

enzyme purification. When the monoclonal antibodies to beef pancreatic

asparagine synthetase were used for the immunoaffinity column

purification, only 1 mg rHAS was obtained from each liter of yeast culture.









35

In addition, due to the significantly high abundance of the yeast proteins in

the cell extract, lengthy washing of the column was required and

consequently the loss of the enzymatic activities of rHAS was almost

inevitable.

The low production of rHAS in S. cerevisiae BJ2168/pVTXAS1 is due

to the fact that the transcription of rHAS gene is regulated by the moderate

alcohol dehydrogenase I promoter, a constitutive promoter in the presence

of glucose. When S. cerevisiae BJ2168/pVTXAS1 was cultured in uracil

deficient media containing 2% glucose, the production of rHAS was found

to be constant for the first 24-26 hours and then started to decline. This

phenomenon could be related to the depletion of glucose in the media after

the culture reached the late-log phase. When the glucose concentration in

the culture media was varied in the range of 2%-20% (w/v), no significant

changes in the abundance of rHAS in the yeast cell extract were observed.

In fact, higher glucose concentration resulted in less total cell extractable

proteins. It is believed that glucose at concentrations higher than 2% may

force the yeast cells to grow anaerobically. As a consequence, overall

cellular protein synthesis is slowed down. Meanwhile, anaerobic growth

of the yeast would accelerate the accumulation of lactate which represses

the ADHI promoter, thus repressing the rHAS gene transcription. In

addition, we observed that yeast cells became increasingly difficult to break

when glucose was present in the media at a concentration higher than 2%.

The construction of plasmid pBS24. 1GAS presented in this paper has

two features that are essential for the quantitative improvements of rHAS

expression. First, the rHAS gene was cloned into plasmid pBS24.1 which











can replicate independently of the S. cerevisiae endogenous 241 plasmid.

Thus, the resulting construct can propagate at a high copy number in an S.

cerevisiae ciro strain such as AB116. In order to test the effects on

expression by pBS24.1 alone, rHAS gene along with the ADHI promoter

was subcloned into the vector and was allowed to proliferate in S.

cerevisiae AB116 host cells. The rHAS expression was found to be higher

than that in S. cerevisiae BJ2168/pVTXAS 1. Second, in order to obtain an

even better expression of rHAS than by using pBS24.1 plasmid alone, the

transcription of the rHAS gene was driven by the glyceraldehyde-3-

phosphate dehydrogenase promoter that is linked to the 3' -end of the

regulatory region of the alcohol dehydrogenase II promoter (designated as

ADHII/GAPp). This ADHII /GAPp becomes a strong promoter when

glucose is deficient in the culture media due to the derepression of ADHIIp.

The culture condition for S. cerevisiae AB 116/pBS24.1GAS was

optimized for rHAS expression by utilizing 8% galactose as the carbon

source, and a rHAS production as high as nearly 6% of the total cell extract

protein was obtained. The cell growth was relatively slow, but eventually

a culture OD600 reading of 5.8-6.2, higher than that of S. cerevisiae

BJ2168/pVTXAS1 culture, could be reached. The total protein extracted

from S. cerevisiae AB116/pBS24.1GAS was approximately two-thirds of

that from the same volume of S. cerevisiae BJ2168/pVTXAS1 culture in 2%

glucose medium. Such low protein content is considered to be the

reflection of the physiological condition of yeast cells in galactose medium.

In fact, the combination of high rHAS expression and low abundance of

total cellular protein in S. cerevisiae AB116/ pBS24.1GAS is a desirable











feature that allows for rapid enzyme purification which is critical for

getting active rHAS. While the pure enzyme usually remained active for at

least two months if stored in small aliquots at -700C, cell extracts of S.

cerevisiae ABI116/pBS24. GAS and BJ2168/pVTXAS1 quickly lost

asparagine synthetase activities during the enzyme purification procedures.

The longer the purification lasted, the less specific enzyme activities were

recovered. The rHAS that was inactivated without being degradated could

be co-purified with active rHAS from monoclonal antibody 3G6 affinity

column (data not shown). Compared to the purification of rHAS from S.

cerevisiae AB116/pBS24. 1GAS, longer column washing was required in

order to get pure enzyme from S. cerevisiae BJ2168/pVTXAS1. As

indicated in Table 2-1, the enzyme purified from the latter expression

system had significantly less specific asparagine synthetase activities.

Interestingly, in the enzyme purification from S. cerevisiae

AB116/pBS24. 1GAS, the yield of both ammonia-dependent and glutamine-

dependent asparagine synthetase activities were greater than 100%. It is

possible that the inactivation of rHAS in total cell extract was stimulated by

the presence of yeast cellular proteins. Under the enzyme assay

conditions, in which the reaction mixtures were incubated at 370C for 30

minutes, such enzyme inactivation could be even faster, therefore causing

more severe negative error in the enzyme specific activity determinations.

In addition, endogenous asparaginase activities in the cell extract would

result in an underestimation of the total asparagine synthetase activity. The

possibility that the lower total enzyme activities of cell extracts might be











due to the interactions of rHAS with yeast inhibitory factors was not

explored.

The previous immunoaffinity purification of rHAS from S. cerevisiae

BJ2168/pVTXAS1 using the monoclonal antibodies against beef pancreatic

asparagine synthetase encountered a special problem. A highly hydrophilic

protein with an apparent molecular mass of 75 kDa (designated as YP75)

was often co-purified with rHAS. Later, it was found that YP75 also has a

Mg2+- dependent ATP hydrolysis activity which could confound detailed

kinetic and mechanistic studies of asparagine synthetase. Although the

results of other enzyme activity assays failed to prove that YP75 is an

asparagine synthetase, it became obvious that a better purification system

was required to avoid the YP75 contamination. Our previous results

suggested that the binding of YP75 to the immunoaffinity column was

related to the concentration of ethanolamine which had been introduced for

blocking the remaining active sites on the gel. More YP75 was co-purified

with rHAS when more active sites on the gel were blocked by

ethanolamine. When cyanogen bromide activated Sepharose-4B was simply

coupled with ethanolamine, the resulting column could be used for affinity

purification of YP75. Pure rHAS was obtained when such ethanolamine

affinity column was used as a pre-column in the enzyme immunoaffinity

purification procedure. Unfortunately, the additional pre-column step made

the purification tedious and lengthy. Low yield of enzyme activities was a

severe drawback. An alternative approach to solve the YP75 contamination

problem was to limit the ethanolamine concentration on the gel by reducing

the remaining active sites that were available for blocking reagent. This









39

goal was achieved by partially hydrolyzing the active sites on the gel with

dilute Tris-HCl buffer at pH 7.5 prior to the covalent antibody coupling

reactions.

Monoclonal antibodies (Mabs) to rHAS were produced. The cross-

reactivities of each Mab with YP75, as well as two E. coli asparagine

synthetases were carefully examined. Mab 3G6 was finally chosen for

making the immunoaffinity columns because of its high affinity and high

specificity for rHAS. In fact, Mab 3G6 also had higher specificity for

rHAS than all the Mabs previously produced against beef pancreatic

asparagine synthetase (data not shown). When Mab 3G6 affinity column

was used to purify rHAS from S. cereviciae BJ2168/pVTXAS 1, up to 2 mg

of pure rHAS was obtained from each liter yeast culture. This yield is

twice as much as previously reported when anti-beef pancreatic asparagine

synthetase monoclonal antibody were used for making affinity columns.

Since rHAS was directly purified from the yeast cell extract, it was

necessary to find out whether yeast endogenous asparagine synthetase(s)

was co-purified with rHAS. S. cerevisiae has two asparagine synthetases,

each having a molecular weight of approximately 150,000 (Ramos and

Wiame, 1980). When both ELISA and western blotting tests were

performed, no cross-reactivities between Mab 3G6 and any of yeast

endogenous proteins were observed. When the cell extracts of S.

cerevisiae AB116 or S. cerevisiae BJ2168 were passed through 3G6

immunoaffinity column, no protein was purified. Meanwhile, the

immunoaffinity purified rHAS was examined by silver-stained SDS-PAGE

and IEF, only one band was observed on both gels. Although very little is









40

known about the physical and biological characteristics of those two yeast

asparagine synthetases, our results indicated that neither of them was co-

purified with rHAS.

The development of a better expression system for rHAS was

stimulated by a need for simplification of yeast culture conditions and for

the quantitative improvement of enzyme production. The latter point is

especially important when undertaking structural studies. Twenty

milligrams of pure proteins are routinely used in crystallization procedures

for X-ray crystallography. The overexpression of rHAS by the

AB 16/pBS24.1GAS system, the fast one-step enzyme immunoaffinity

column purification, as well as the monoclonal antibodies to rHAS will

greatly facilitate future kinetic and structural studies of the enzyme.















CHAPTER 3
CHARACTERIZATION OF CYS1 MUTANTS OF
HUMAN ASPARAGINE SYNTHETASE

Introduction

Asparagine synthetase (AS) (EC 6.3.1.1) catalyzes the synthesis of

asparagine from aspartate using either glutamine or ammonia as the

nitrogen donor.

Gin (NH3) + Asp + ATP --------> Asn + Glu + AMP + PPi

The ability to utilize NH3 in the place of glutamine as the nitrogen donor is

a general characteristic of a family of amidotransferases such as glutamine

phosphoribosylpyrophosphate amidotransferase (Tso et al., 1982), CTP

synthetase (Levitzki and Koshland, 1971; Levitzki, Stallcup and Koshland,

1971), GMP synthetase (Zalkin and Truit, 1977), and glucosamine-6-

phosphate synthase (Walker et al., 1984). The amino acid sequence for AS

(Andrulis, Chen and Ray, 1987) cloned from a human cell line reveals a

purF- type glutamine amide transfer domain (Tso et al., 1982) at the amino

terminus. Ten out of the 40 amino acid residues identified as being part of

this domain are highly conserved (Scolfield, Lewis and Schuster, 1990).

One of these residues, a Cys, found at position-1 in AS has drawn

considerable attention after it was determined to be essential for glutamine

amide transfer presumably through the formation of a covalent glutamyl-

enzyme catalytic intermediate.

Treatment of bovine pancreatic asparagine synthetase with the

alkylating agent and glutamine analog, 6-diazo-5-oxo-L-norleucine (DON)











was shown to abolish the the glutaminase activity as well as the glutamine-

dependent synthetase activity with little effect on the ammonia-dependent

activity (Mehlhaff, and Schuster, 1991). Likewise, replacement by

mutagenesis of the Cyst residue with Ala (Van Heeke and Schuster, 1989b)

also eliminated the glutamine-dependent activities of the human enzyme

while leaving the ammonia-dependent activity unaffected.

While these observations indicated the essential role of Cyst in the

amide transfer reaction of AS, questions remain concerning other possible

direct or indirect, functional or structural roles this residue may have on

the enzyme. In the absence of X-ray crystallographic data on AS, one

approach to addressing these questions is by the construction and analysis

of stable Cysl mutants. Previous to the availability of these types of

mutants, functional asparagine synthetase lacking glutamine-dependent

activity could only be obtained by treatment with chemically modifying

agents (e.g., DON) or by the addition of monoclonal antibodies (Pfeiffer et

al., 1987). However, it is difficult to be certain that unwanted

perturbations by either one of these treatments are not occurring. By

contrast, changes with absolute specificity can be introduced on the protein

sequence through site-directed mutagenesis. Although the effects of these

changes can not be interpreted unambiguously, mutagenesis has proven

useful for probing the importance of specific residues on enzyme function,

folding and stability (Van Heeke and Schuster, 1989b; Zhou et al., 1992;

Makaroff et al., 1983).

Active AS lacks the amino-terminal methionine. Similarly, it was

shown that fully active glutamine phosphoribosylpyrophosphate











amidotransferase (GPA) from chicken (Zhou et al., 1992) and from B.

subtilus (Shortle, 1989) is the result of a posttranslational cleavage of an

11-residue propeptide. The Cys1 residue in the mature enzyme was found

to be not only necessary for the glutamine-dependent function of the

enzyme, but also for propeptide processing. Although a comparable

posttranslational modification does not occur in AS, a role for Cyst in

methionine processing or in the structural stability of the enzyme can not be

ruled out.

We have previously reported on both an E. coli (Van Heeke and

Schuster, 1989a) and an S. cerevisiae (Van Heeke and Schuster, 1990)

expression system for recombinant human asparagine synthetase (rHAS).

Only in the yeast system were we able to obtain soluble enzyme in which

the N-terminal methionine had been correctly processed and that was

functionally identical to the native protein. The S. cerevisiae expression

system had been designed for efficient site-directed modifications of one of

the four N-terminal amino acids.

In this paper, we report on the use of such a system for the

construction and characterization of a series of Cyst mutants of HAS for

the purpose of further exploring the role of this residue in the enzyme

function. An attempt is made to describe the results in terms of the

available data on the role of CysI in the chemical mechanism of the

enzyme.

Materials and Methods
Chemicals and Reagents

Yeast culture media was purchased from DIFCO (Detroit, MI).









44

Cyanogen bromide activated Sepharose-4B resin and PD-10 desalting

columns were acquired from Pharmacia (Piscataway, NJ). 14C-[8]-ATP

was purchased from ICN (Costa Mesa, CA). Scintillation fluid

ScintiVersTM II* was obtained from Fisher (Orlando, FL). DE-81 anion-

exchange chromatography paper was purchased from Whatman (Hillsboro,

OR). Gel electrophoresis reagents and protein concentration assay dye

were obtained from BIORAD (Melville, NY). Glass beads Pyrophosphate

Assay Reagents and other chemicals were obtained from SIGMA (St.

Louis, MO). DNA restriction enzymes were acquired from either

PROMEGA (Madison, WI) or New England Biolabs (Beverly, MA). T4

DNA ligase (2000 U/.l) was purchased from New England Biolabs. T4

DNA polymerase and calf intestinal phosphatase were obtained from

PROMEGA.

Glutamine Recrystallization

Solid glutamine was dissolved in a 60-700C 1.5 mM EDTA solution to

the point of saturation. Pure glutamine was precipitated by adding two

volumes of 100% ethanol. Complete precipitation was obtained while the

solution was still warm. Filtrate was discarded and glutamine crystals

were vacuum filtered a 0.45 p.m nylon membrane. Crystals were allowed to

dry by spreading in a clean container and stored at room temperature until

ready for use.

Strains and Plasmids:

E. coli XL1-BlueTet strain (recAl, endAl, gyrA96, thi-1, hsdR17,

supE44, relAl, lac, IF' proAB, laclqZAMI5, TnlO(tetr)] ) was obtained

from STRATAGENE (La Jolla, CA). Cells were cultured in Luria broth









45

(Sambrook, Fritsch and Maniatis, 1982). Ampicillin was added to the

medium for selection when cells had been transformed with plasmids

containing the appropriate marker. Plasmid pVTXAS 1 is a HAS expression

vector that had been constructed from the vector pVT101U (Vernet,

Dignard and Thomas, 1987) for previous studies in our laboratory (Van

Heeke and Schuster, 1989a, 1990). Saccharomyces cerevisiae strain

BJ2168 (aprc 1-407 prbl-1122, pep4-3, leu2, trpl ura3-52 ) was obtained

from the Yeast Genetic Stock Center (Berkeley, CA) and was used for

expression of HAS and the Cys HAS mutants.

Construction of Cysl HAS Mutants

The plasmid pVTXAS was used to obtain all the CysI mutations of

HAS. Figure 3-1 shows a simplified map of pVTXAS and the

oligonucleotides that were used for mutagenesis of HAS. A more detailed

description of this vector can be found in reference 14. All mutant clones

were constructed as previously described for the Cys to Ala mutant with

some modifications (Van Heeke and Schuster, 1989b). Briefly, the vector

fragment of an Xhol/ApaI digestion of pVTXAS 1 was excised from gels

and then ligated to double-stranded, phosphorylated oligonucleotides

containing the desired base-pair changes. An aliquot of the ligation mixture

was used to transform E. coli (Hanahan, 1983) and the DNA of positive

clones was verified by sequencing double-stranded plasmid DNA using the

dideoxy chain termination method. S. cerevisiae strain BJ2168 was

transformed with each one of the clones using the lithium acetate procedure

(Treco, 1987).









46

Large-scale Yeast Culture for HAS Purification

Culturing of BJ2168 S. cerevisiae harboring the various plasmid

clones and HAS immunoaffinity purification were performed as previously

described by Sheng and Moraga, et al.( 1992a).

Enzyme Assays

Assay conditions for asparagine synthetase activity were similar to

those described by Luehr and Schuster (1985), except that the total assay

mixture volume was 60 .tl. Termination of the reactions was accomplished

by the addition of 20 .il of 20% (w/v) trichloroacetic acid solution

containing 0.32 mM histidine as an internal standard for HPLC

analysis.The substrate mixture included 85 mM Tris HC1, 50 mM NaC1,

8.33 mM MgC12, 5 mM ATP, 10 mM aspartate and either 20 mM glutamine

or 75 mM ammonium acetate. For pH optimum determinations, the pH of

Tris HC1 and glutamine or ammonium acetate was adjusted to the desired

pH before used for the preparation of substrate mixtures. Routine assays

were performed on 0.5 to 1.0 4.g enzyme by the measurement of asparagine

production by the HPLC assay (Unnithan, Moraga and Schuster, 1984)

using an Applied Biosystem 420 derivatizer and 130 A separation system.

The glutaminase activity was also measured by HPLC monitoring of

glutamate production. Substrate mixture consisted of 33.3 mM Tris HCI at

pH 8.0 and 21.7 mM glutamine.

Two other assays were used for the product stoichiometry studies of

asparagine synthetase:

PPi production assay. An enzymatic coupling system that utilizes PPi

as a substrate and subsequently converts NAD+ to NADH was used to











determine the PPi production during asparagine synthesis by HAS. This

assay was derived from a method previously described (O'Brian, 1976).

Briefly, a vial of PPi assay reagents (SIGMA) was reconstituted with 1 mL

of deionized, distilled H20. The final reaction mixture of 160 ptl contained

asparagine substrates, 1 Ipg enzyme and 20 .1l reconstituted PPi assay

reagent. Upon the addition of the enzyme and quick mixing, the reaction

mixtures were immediately transferred into quartz microcuvets and their

absorbances at 340 nm monitored at 370C for 20 minutes using a Beckman

DU-64 Spectrophotometer. Blanks were samples in which aspartate was

omitted from the substrate mixture. For each vial of PPi assay reagents, a

standard curve was obtained using the rates of decrease of absorbance at

340 nm for pure PPi in the presence of substrate mixture.

AMP production assay. A radiometric method using 14C-[8]-ATP

(specific activity, 357 pCi/mmol) in the substrate mixture was utilized to

measure 14C-[81-AMP production (Allison, Todhunter and Purich, 1977).

Following incubation at 370C for 30 minutes, the 60 tl reactions were

terminated by boiling for 2 minutes. After centrifugation and cooling, 6 41

were loaded on DE-81 anion-exchange chromatography paper (25 cm long x

29 cm wide) in two separate applications. A mixture of nonradioactive

ATP, ADP and AMP was loaded next to both edges of the paper as

standards. Samples were allowed to dry and then placed vertically in jars

containing ammonium format (1/45 saturated, pH 2.75) as the developing

solvent. Ascending chromatography proceeded for 3 hours and the paper

was allowed to dry. Non-radiolabeled nucleotides were visualized on the

paper by UV absorbance at 254 nm and used as a reference to locate the











corresponding spots on the enzyme assay samples. Bands for AMP and

ADP+ATP were cut, put separately into 7 mL vials containing ScintiVersTM

II* fluid and analyzed in a Beckman LS 60001C scintillation counter.

KmLP Determination
The apparent affinity constants (Kmapp) for the substrates were

determined by incubating the enzymes at initial velocity conditions and

saturating concentrations for all except one of the substrates. In the case

of Kmapp determination for aspartate, the concentration of this substrate

was varied from 0.25 mM to 1.2 mM. In the case of ATP, the

concentration was varied from 0.2 mM to 0.8 mM. For ammonia, the

concentration was varied from 0.6 mM to 8.5 mM. The Kmapp for every

substrate was calculated from double-reciprocal plots of I/varying substrate

concentration versus I/velocity.

Protein Concentration and Immunoblots

Protein concentrations were measured by the method of Bradford

(1976) using pure mouse immuglobulin G to construct the standard curve.

Protein molecular mass estimations by SDS-polyacrylamide gel

electrophoresis (SDS-PAGE) and protein western blotting using the 3G6

anti-HAS monoclonal antibody were performed as described earlier (Sheng

and Schuster, 1992).

Protein N-terminal sequencing

Purified enzymes were subjected to SDS-PAGE (Laemmli, 1970) and

electroblotted onto polyvinylidene difluoride (PVDF) membranes with 10

mM MES buffer (pH 6.0, containing 20% methanol) at 340 mA and 4 oC

for 90 minutes. Blotted proteins were stained with Coomasie Brilliant Blue











R-250 (Nakagawa and Fukuta, 1989) and the band corresponding to the

HAS molecular mass (ca. 64 kDa) was excised for sequencing. N-terminal

protein sequencing was performed by the Protein Chemistry Core facility of

the Interdisciplinary Center for Biotechnology Research (ICBR) at the

University of Florida, using an Applied Biosystem 473A Protein Sequencer

and 610A software for data analysis.

Results
Construction of Cvsl Mutants of HAS

Site-specific mutagenesis of HAS at the amino terminus was

performed using the pVTXAS 1 plasmid following a similar procedure

previously used for the generation of a Cys1 to Ala HAS mutant (Van

Heeke and Schuster, 1989b). Briefly, the DNA sequence between Xhol

and Apal in the vector was replaced with either single-stranded or double-

stranded synthetic oligonucleotides carrying the desired mutations, and

having termini complementary to the restriction enzyme sites in the

plasmid. Double-stranded oligonucleotides were used for the construction

of the C1(A), AC and C1G:G2C mutants because the ligation efficiency

with single-stranded oligonucleotides was found to be very low, thus

limiting the recovery of clones even when highly efficient transformation

procedures were used (Hanahan, 1983). Figure 3-1 shows the

oligonucleotides that were utilized for this purpose. The first three mutants

represent cases where the codon for Cyst was replaced by the codons of

either Ser (CIS), Tyr (C1Y) or Lys (C1K) (replacement mutants). The Ala

(C1A) replacement mutant (number four) was available for this study from

previous work (Van Heeke and Schuster, 1989b). The fifth mutant lacked

































Figure 3-1: Yeast Plasmids for the Expression of Cys1 Mutants
of HAS. The thick arrow indicates the HAS coding region.
Sequences denote part of the coding region (in wild-type,WT) or
oligomers used for the construction of mutants. The XhoI and Apal
sites are 31 base pairs appart and bracket the mutagenesis target area.
The small arrow indicates the translation start codon. Abbreviations
used are: ADHp, yeast alcohol dehydrogenase I promoter; ADH3',
3' sequence of yeast alcohol dehydrogenase. The following primer
was used for sequence confirmation: 5'-AATCTTCATAGCACTCA-
3'. Refer to Materials and Methods for further details.













XholI
5' CMGACATIATATA^
CGIATATAGT

5' TCGAGCATATATACA
CGIT~ATGT
5' TCGAGCAATAIATACA
C TATATGTC


AITG


ATG
TAC


QCA
CT

TGT
ACA


ATIT
TAA


AIG a TCT ATT
T.2 7ACA TAA


Acal
TTG GGC C-3'
AAC


GGA ATT TIG GGC C-3
CCT TAA AAC


TIG lGC C-3'
AAC


5' TCA GCATATATACA AIG AT GGA ATT TIG GC C-3'

5' ITCGAGCATATATACA ATG C GGA ATT T~I GC C-3

5' TCIAGCATATATACA ATG 3 GGA ATT TG GGC C-3


5' TCCAGCATATATACA ATG E


GGA ATT TIG GGC C-3'


5' TGACATAMTATACA ATIG c GA ATT TIG GG3C C-3'


AC


C (R)


C1 G:G2C


CIY

C1K

CIA

CIS

WT











the codon for Cys1 (AC1, deletion mutant). In the last two mutants, either

an Ala codon was inserted in between Cys1 and Gly2 residues (C1[A],

insertion mutant) or the position of the Gly2 and Cys1 codons were

reversed (C1G:G2C, reversion mutant). Clones were initially screened by

restriction enzyme mapping analysis. The mutations were confirmed by

DNA sequencingusing a 17 base-pair primer (see legend of Figure 3-1).

This primer was complementary to a region of HAS 60 base pairs

downstream from the start codon. DNA sequencing of two random clones

revealed that the ClG:G2C mutant had an additional mutation. At the third

position of the amino terminus, the isoleucine codon had been replaced

unintentionally by a codon for glycine.

Expression of the Mutants in S. cerevisiae

Protein expression from the pVTXAS1 plasmid or its derivatives is

driven by the alcohol dehydrogenase I promoter (ADHp) in a constitutive

manner. S. cerevisiae BJ2168 transformed with plasmids carrying one of

the mutations was plated on YMM omission plates (no uracil) and incubated

at 300C. After 3 days, single colonies were picked and grown in 10 mL

liquid omission medium for 30 hours. Aliquots of 0.5 mL of cells were

centrifuged and redissolved in asparagine synthetase isolation buffer

(Sheng et al., 1992a). The cells were then mixed with 1 gram of chilled

glass beads, and lysed by vortexing 6 to 7 times for 15 seconds with 5

minute intervals at OOC. Cell extracts were cleared by centrifugation, and

30 pig of protein was immediately analyzed by SDS-PAGE. Proteins were

transferred onto a membrane were then immunoblotted with an anti-HAS

monoclonal antibody. Immunoblots are shown in Figure 3-2. As can be









53

seen, a HAS band could be observed in cell extracts from each mutant, thus

indicating that all the mutations had yielded expressable proteins.

However, when the large-scale immunoaffinity purification procedure was

carried out, pure protein could only be obtained from the C1A, CIS and

C1(A) mutants in comparable amounts to that obtained from the non mutant

enzyme. A very small amount of pure protein was obtained from the ACI

mutant and no protein could be purified from the rest of the mutants. As

shown in Table 1, all the mutant proteins that could be purified had

ammonia-dependent asparagine synthetase activity and no glutaminase or

glutamine-dependent asparagine synthetase activities. This activity pattern

had been previously reported for the C1A mutant by Van Heeke and

Schuster (1989b). However, based on a mechanistic scheme proposed by

Zalkin for other glutamine amidotransferases (Mei and Zalkin, 1989), we

expected at least some residual glutamine-dependent activity by the C1S

mutant. A nucleophilic attack on glutamine by a hydroxylate rather than a

thiolate ion would lead to the formation of an acylenzyme intermediate and

the release of ammonia. In HAS, the released ammonia participates in a

subsequent nitrogen transfer reaction to an aspartyl-AMP intermediate for

the synthesis of asparagine (Luehr and Schuster, 1985). Contrary to our

expectations, no glutamine-dependent activity was detected in the CIS

mutant even when assayed at a higher pH. It is important to point out that

in order to eliminate spurious residual glutamine-dependent or glutaminase

activities of CIA and CIS mutants, it was necessary to recrystallize

commercially available glutamine and to use freshly-made solutions.

Substantial amounts of ammonia were found in concentrated glutamine
















































Figure 3-2: Immunoblots of Total Cell Extracts from S. cerevisiae
Containing Plasmids with one of Several HAS Mutants. Arrows on the
right indicate the positions of the molecular mass markers; from the top:
phosphorylase B (106 kDa), serum albumin (80 kDa), ovalbumin (49.5
kDa), carbonic anhydrase (32.5 kDa). Thirty gg of total cell extracts were
loaded on each lane of a 10% gel (SDS-PAGE). Lane 1, C1Y; Lane 2, C1K;
Lane 3, C1G:G2C; Lane 4, C1A; Lane 5, ACI; Lane 6, CIS; Lane 7, C1A;
Lane 8, WT; Lane 9, purified HAS (l1ig). See Materials and Methods for
further details.















Table 3-1
Activity, N-terminal Amino Acid Identity and pH Optima of Human Asparal
Synthetase and Cyst Mutants Purified from an S. cerevisiae Expression Sys



Enzymatic activity (nmol Asn/min.mg)a

N-terminal

Enzyme NH3-dependent Gin-dependent Glutaminase Amino Acid pH-op



HAS (wt) 155 111 152 Cys 8

C1A 166 None None Ala 9

CIS 178 None None Ser 8

C1Y NDb NDb NDb NDb N

C1K NDb NDb NDb NDb N

AC 4.8 None None Gly N

C1(A) 156 None None Ala 9

C1G:G2C NDb NDb NDb NDb N


aValues are the average of two to seven determinations from different enzyme
preparations.
bNot determined. These mutants could not be purified in an immunoaffinity
column.











stock solutions kept at -20 oC before use. Ammonia was also formed by

the non-enzymatic hydrolysis of glutamine at 370C during the enzyme

assay. This latter confounding factor was eliminated by performing the

glutamine-dependent assays utilizing large amounts of enzyme (30 gg,

normally 0.5-1.0 gg protein per sample was used) and incubating for a

short time (1-3 minutes).

Native, as well as recombinant HAS expressed in yeast lacks the N-

terminal methionine. Posttranslational processing of methionine is

governed primarily by the identity of the second amino acid residue

(Tsunasawa, Stewart and Sherman, 1985; Ben-Bassat et al., 1987).

Therefore, N-terminal amino acid analysis was performed for the first five

amino acid residues of purifiable mutants in order to determine whether

methionine had been removed. The results are also shown in Table 3-1.

Methionine was processed in the CIA, CIS and AC1 mutant proteins, but

in the CI(A) mutant both the methionine and Cys1 had been removed.

Consequently, expression of C1A and CI(A) resulted in the same protein

by virtue of their posttranslational processing, even though nucleotide

sequencing had shown that Cl(A) still had the codon for Cys'. The

identity of the remaining four amino acids that were sequenced for the

purifiable mutant proteins was in agreement with that of wild-type HAS.

No N-terminal sequence information was obtained from the mutated

proteins which could not be purified.

Stoichiometry. pH Optimum and Kinetic Parameters of CIA and C1S
mutants

Equation 1 predicts that the concentration of products of the reaction

catalyzed by HAS are in a 1:1:1:1 (Asn:Glu:AMP:PPi ) ratio. This











proportionality of the reaction products remained the same when ammonia

was used in the place of glutamine. As shown in Table 3-1, mutations of

the Cysl terminal residue completely abolished the glutamine-dependent

asparagine synthesis by HAS. Therefore, it was appropriate to determine

whether this drastic effect could take place on the amide transfer domain of

the enzyme without causing any significant changes to the synthesis

domain, now using ammonia as the nitrogen source. Clearly, measurement

of the production of asparagine alone by the mutants would not be

sufficient to discard this possibility. One possible effect of the Cys'

mutations could be a decouplingg" of the catalytic use of substrates in the

ammonia-dependent synthesis reaction. Measurement of Asn, AMP and PPi

in three separate assays, revealed that the product stoichiometry by the C1A

and CIS mutant enzymes remained 1:1:1 for Asn, AMP and PPi

respectively (data not shown).

Interestingly, the pH optimum of the ammonia-dependent reaction for

the CIS and C1A mutants was increased by 0.5 to 1.0 pH unit,

respectively, as compared with the non-mutant enzyme (Table 3-1). This

finding suggested the possibility that the amino acid substitutions caused

conformational changes on the mutant proteins, thus perhaps also affecting

affinity for the substrates. Therefore, the apparent affinity constants

(Kmapp) of the mutants for every substrate was determined and compared to
those of the wild-type enzyme. This was done at pH 8.0 (the optimum

activity pH for the wild-type enzyme) and at pH 8.5 or 9.0 (the pH optima

for CIS and CIA respectively). As can be seen in Table 3-2, both mutants

had comparable apparent Km' s for all three substrates when determined at




















E


CL
E
`ji
(9 Q

E m


















2,^
TE









1-5












t|
Q.















go





12 CL


E
I
CL
CL-


0)C -


o 0
0




0,0

C C
00 c
CL ZC

o c



g 0
CI;
3*o









010
ou








E<





Oa o

wQt C0
w '0)
0. 0 0.








S)~ OTw




(B r OT
aJ fc 1 I
(0. U *


0 CD 0
a- CoJ coJ
cm 04











pH 8.0 or at their optimum pH. The apparent Km's of the mutants were

also compared to those of the wild-type enzyme. Similar Kmapp for

aspartate and ammonia at pH 8.0 were obtained for the mutants and wild-

type HAS. But the Kmapp for ATP was approximately 3-fold greater in

wild-type than in the mutant enzymes. An effect of this magnitude has been

regarded only as a small effect in previous single-site mutagenesis studies

(Gibbs and Zoller, 1991; Wilkinson et al., 1983). Further comparison of

the apparent Km's of the mutants and the wild-type enzyme, when

determined at the pH optima of the mutants, once again, revealed no major

differences with the exception of the Kmapp for aspartate. This parameter

showed a decrease of approximately 2.5-fold in the mutants. As expected,

both wild-type HAS and the mutants had a significantly smaller Kmapp for

ammonia when measured at higher pH' s. This could be attributed to an

increase in the [NH3]/[NH4+] ratio proportional to an increase in pH.

Lastly, the ammonia-dependent catalytic potential of wild-type HAS

diminished to less than half its original value by the increase of one pH

unit, while that of the mutants was largely unaffected. In fact, the activity

of non-mutant asparagine synthetase declined precipitously beyond pH

9.0, while that of the CIA and CIS mutants could be easily measured even

at pH 9.6 (data not shown).

Inhibition of the Ammonia-dependent Activity of Mutants by a Glutamine
Analog

We have described above the various approaches that were taken in

order to eliminate possible spurious glutamine-dependent asparagine

synthetase detection in the C1A and CIS mutants of HAS. An additional

approach taken was the measurement of glutaminase and glutamine-









60

dependent asparagine synthetase activities in the presence of 6-diazo-5-

oxo-norleucine (DON). This compound is both a glutamine analog and an

alkylating agent, and has been shown to inhibit only the glutamine-

dependent activity of asparagine synthetase (Mehlhaff and Schuster, 1991)

as well as other glutamine amide transfer enzymes (Tso et al., 1982). The

effect of DON is presumed to be the result of alkylation of the essential

Cys1 residue of the target enzymes (Mei and Zalkin, 1989). It was thought

that if any of the mutants had a measurable amount of glutamine-dependent

activity, this would be neutralized by DON and reflected in a decrease in

the activity even in the presence of non-enzymatic hydrolysis of glutamine.

In addition, the Cys1 mutants of HAS provided us with the unique

opportunity to ask whether DON would have an effect on the enzyme which

was not due exclusively to its binding to Cys1. As expected, no glutamine-

dependent activity was detected in any of the mutant enzymes in the

presence or absence of DON. This, once again, indicated that neither Ala

nor Ser could even partially replace the function of Cys1 in HAS.

Unexpectedly, inhibition was observed in control samples in which the

ammonia-dependent activity of the mutants was measured in the presence of

DON. This result lead us to carefully examine any possible effect of DON

on the same activity of the wild-type enzyme. It was determined that,

similar to its effect on the mutants, DON also inhibited the ammonia-

dependent reaction of the non-mutant enzyme. Figure 3-3 shows the

concentration dependence of the DON effect on the ammonia-dependent

activity of the enzymes. At a concentration of 10 mM, DON inhibited the

ammonia-dependent activity of CIS, CIA and wild-type by 50%, 30% and









61







200




150




100
I
-A



50




0
0 1 3 6 10
DON concentration
(mM)





Figure 3-3: Effect of 6-diazo-5-oxo-norleucine (DON) Concentration
on the Activity of Mutants C1A and CIS, and Wild-Type (WT). The
concentration of ammonia in the samples was 15 mM. All other substrates
were kept at saturating concentrations. Symbols are: C1A, ( ); CIS,
(m ); WT, ( N ).










62




Panel A

15n 4


120




I 60


30


0 3 6 9 12
Time (min)


Panel B
100


80



jjo -



20



0 3 6 9 12
Time (mia)


Figure 3-4: Effect of 6-diazo-5-oxo-norleucine (DON) on the Specific
Activity of Mutant CIA and Wild-type (WT) versus Incubation Time.
Panel A, CIA; panel B, wild-type. The concentration of DON in the
samples was 7 mM. The concentration of ammonia in the samples was 15
mM. All other substrates were kept at saturating concentrations. Symbols
are: C1A without DON, (A); C1A with DON, ( A); WT without DON,
( ); WT with DON, ( ).











41% respectively. When the enzymes were incubated on ice with 7 mM

DON for 30 minutes, applied to a desalting (size-exclusion) column and

then assayed, an inhibition of 17-22% was still present (data not shown).

Therefore, it appears that at least part of the inhibitory effect of DON on

the ammonia-dependent activity of the wild-type enzyme and the mutants

may be due to covalent inactivation. These results appeared

inconsequential since other authors had already reported moderate changes

due to DON treatment of the ammonia-dependent synthesis of enzymes that

otherwise use glutamine as their nitrogen source. In every case, those

changes were largely ignored. Non-covalent binding is a fast

phenomenon and is unlikely to be dependent on the enzyme assay time

scale. Conversely, the specific activity of the enzyme should decrease with

incubation time if inhibition is the result only of covalent inactivation.

Figure 3-4 shows the specific activity of wild-type HAS and the mutants in

the presence or absence of DON versus time of incubation. Clearly, DON

inhibition could be observed as soon as enzyme activity could be measured.

In fact, the specific activity rose during the first few minutes of the assay.

However, this could be explained by the fact that the enzyme was

maintained at 40C previous to starting the assay at 370C. Although

enzymes and substrates were incubated separately at 370C for 5 minutes

prior to mixing, this may not have been sufficient time to reach temperature

equilibrium. In any event, the results were consistent with either a non-

covalent inhibition pattern or with a mixed effect of covalent plus non-

covalent inactivation.











Discussion

In order to more fuller characterize the role of the N-terminal cysteine

in HAS, a series of mutations was created in which the codon for Cys1 was

deleted (AC) or replaced by one of several codons (C1A, CIS, C1K, C1Y).

Two additional mutants were constructed by either inserting an alanine

codon in between Cys' and Gly2 [CI(A)] or by inverting the positions of

Cysi and Gly2 (C1G:G2C). All mutants were overproduced in an S.

cerevisiae expression system, but only C1A, CIS and C1(A) could be

purified from an immunoaffinity procedure with yields comparable to that

of the wild-type enzyme. It is possible that certain amino acid mutations

caused significant conformational changes or even a different folding of the

protein so that this would be reflected in less efficient binding to the

immunoaffinity columns. No experiments were conducted to confirm or

discard this possibility.

An alternative explanation for our inability to purify some of the

mutants could be that the N-terminal cysteine is not only essential for

glutamine-dependent and glutaminase activities, but may also play a role in

the maintenance of enzyme proteolytic stability. This was not an exclusive

role of Cysi since stable mutants containing either Ala (CIA) or Ser (CIS)

at the N-terminal end of the enzyme could be constructed and expressed.

Immunoblots on crude enzyme extracts revealed that all mutants were

expressed and yielded an asparagine synthetase band of comparable

intensity to that of equivalent amounts of protein from the wild-type.

Decreased metabolic stability of some of the mutants would result in

sufficiently altered proteins which could no longer be purified by a









65

procedure involving the use of an antibody. The enzyme purification

procedure took approximately 12 hours, with a 4 to 5 hour interval

between cell disruption and cell extract application to an immunoaffinity

column. Although the enzyme isolation buffer contained several protease

inhibitors (Sheng et al., 1992a), the time of purification was apparently

long enough to allow for proteolysis of susceptible proteins by endogenous

yeast proteases. Increased susceptibility of the AC1 mutant to proteolysis

was suggested by the fact that although only small amounts of the protein

could be purified, its enzymatic activity remained as stable as that of the

wild-type thereafter (data not shown). Work by other authors have shown

that the recognition of an amino-terminal residue on a protein may mediate

its metabolic stability and the set of individual amino acids can be ordered

with respect to the half-lives that they confer on the protein when present at

its amino-terminus. Varshavsky and coworkers (Bachmair, Finly and

Varshavsky, 1986) discovered that B-galactosidase expressed in yeast

initially as a ubiquitin-1-gal fusion, had strikingly different half-lives in

vivo depending on the nature of the amino acid at the amino-terminus of B-

gal. These researchers later extended their findings to many other

intracellular proteins from both prokaryotes and eukaryotes to establish

what was deemed the "N-end Rule" (Gondka et al., 1989). Although the

classification of N-terminal amino acids differed depending on whether the

protein was expressed in rabbit reticulocytes or yeast, the observations are

useful in so far as many exogenous proteins are expressed in yeast. Cys,

Ser, Ala and Gly were classified as part of the stabilizing class in S.

cerevisiae. Tyr and Lys, on the other hand, were found to render proteins









66

highly susceptible to proteolysis regardless of their expression system.

The mutagenesis experiments on the N-terminal Cys of HAS largely agree

with the predictions from the N-end Rule. However, in contrast to it, two

other mutants which had glycine at their N-terminus (AC1 and C1G:G2C)

also appeared to be metabolically unstable. Unfortunately, the half lives of

the enzymes were difficult to determine because of competing asparaginase

activity in BJ2168 yeast cell extracts and the characteristically low specific

activity of asparagine synthetase.

All mutants that could be purified lacked the N-terminal methionine

as in native and recombinant non-mutant HAS (Van Heeke and Schuster,

1989b). This indicated that Cys1 had no exclusive role, if any, in HAS

posttranslational processing of methionine by S. cerevisiae cells.

Surprisingly, the mutant in which an Ala codon was inserted in between

Cys1 and Gly2 [C1(A)], as verified by DNA sequencing, lacked both the

methionine and Cysl thus yielding an enzyme identical to the CIA mutant.

Further studies are required to establish whether other amino acid residues

downstream from Cys1 are important in N-terminal posttranslational

processing of HAS when expressed in yeast.

Of the three mutants that yielded purifiable proteins, it was

hypothesized that the CIS mutant would have residual glutamine-dependent

asparagine synthetase and glutaminase activities. These expectations were

based on insights into possible mechanistic pathways by which nitrogen is

transferred, provided by the pioneering work of Zalkin on glutamine

phosphoribosylpyrophosphate amido transferase (GPA)(Tso et al., 1982;

Zalkin and Truit, 1977; Zhou and Bryles, 1992; Makaroff et al., 1983).











These authors have clearly shown that the essential cysteine residue

required for the proposed formation of the covalent glutamyl (acylenzyme)

intermediate is the N-terminal Cys1 residue in GPA. This suggested that a

similar mechanism might operate in asparagine synthetase. Moreover, for

GPA, a "catalytic triad" composed of Cyst, His101 and Asp29 has been

identified (Mei and Zalkin, 1989) implying that the N-terminal residue

might be involved in an amide hydrolysis step. The corresponding triad

residues in HAS are Cys1, His103 and Asp30. By analogy to proteases,

Zalkin has proposed that Hisl01 might promote the formation of a thiolate

ion, which through nucleophilic attack on glutamine, generates an

acylenzyme. For asparagine synthetase, the ammonia molecule that would

be released in this reaction participates in a subsequent nitrogen transfer to

an aspartyl-AMP intermediate for the synthesis of asparagine. Since the

pKa for the hydroxyl hydrogen of serine is higher than for the sulfhydril

hydrogen of cysteine, the assay performed at a higher pH might have

improved the chance to detect glutamine-dependent asparagine synthetase

activity of the CIS mutant. But neither this nor any glutaminase activity

was not detected in proteins purified from either C1A or CIS.

Significant effort was expended in assuring that all possible spurious

glutamine-dependent synthesis of asparagine would be eliminated. For

instance, an apparent activity was observed if old solutions of glutamine

maintained at -200C were used for the preparation of assay mixtures.

Ammonia also formed from non-enzymatic hydrolysis of glutamine under

assay conditions at 37 oC. Newly prepared solutions and short assay times

on large amounts of enzyme eliminated this problem. The possibility still









68

remains that the mutants had a residual glutamine-dependent activity which

was below the detection limit of the assay. The HPLC assay could detect as

little as 50 picomoles of asparagine produced.

Aside from the absence of the glutamine amide transfer reaction in

mutants C1A and CIS, the synthesis reaction using ammonia exhibited a

large degree of similarity to the same reaction catalyzed by the non-mutant

enzyme. Not only did the stoichiometry of the reaction products remain the

same, but also the catalytic potential of the mutants and the apparent

affinity constants for the substrates were only moderately affected, that is,

no differences greater than 3-fold were observed. Other studies have

shown that mutations on single residues important for enzymatic activity

can result in Kcat and Km differences greater than 3 orders of magnitude

(Gibbs and Zoller, 1991). Not surprisingly, the Kmapp for ammonia of

both the mutants and the wild-type asparagine synthetase decreased

significantly when measured at a higher pH. Although other phenomena

such as pH-dependent conformational changes cannot be excluded, this

effect was most likely due to an increase in the [NH3/NH4+] ratio

proportional to the pH, thus making the effective nucleophilic species

concentration higher. This ratio is 0.056, 0.178 and 0.56 at pH 8.0, 8.5

and 9.0 respectively.

The extent of the similarity between CIA and CIS mutants, and

between mutants and wild-type HAS may be extended to the level of

folding during synthesis and expression. This view is favored by the fact

that all three proteins were purified on the same immunoaffinity column

using the 3G6 anti-HAS monoclonal antibody (Sheng et al., 1992a). Wild-









69

type HAS and C1A could also be purified from immunoaffinity resins made

by using two additional anti-beef pancreatic asparagine synthetase

monoclonal antibodies (Van Heeke and Schuster, 1989b). It was

interesting, however, that both mutants had a higher pH optima than the

non-mutant enzyme. This phenomenon was suggestive of further

differences in the ammonia-dependent reaction between Cyst mutants and

the wild-type enzyme that might become apparent under a closer

examination. For instance, one could ask whether the pattern of substrate

inhibition or product inhibition observed on asparagine synthetase remains

unchanged in the mutants.

The use of 6-diazo-5-oxo-norleucine (DON) in this study originally

had the purpose of further confirming whether mutants had any detectable

glutamine-dependent activity. The assumptions were that since DON is

known to affect only the glutamine-dependent reaction of amide

transferases of the same class of asparagine synthetase, any measurable

activity using glutamine as the amino donor would show a decrease when

measured in the presence of the inhibitor. As expected, no glutamine-

dependent activity was observed in the absence or presence of DON.

However, a sizable inhibition of the ammonia-dependent reaction of both

mutants was observed. This unexpected result appeared to have been at

least partially the result of a covalent modification of the enzymes since

inhibition by DON could not be eliminated by applying a mixture of enzyme

plus inhibitor to a column containing a size-exclusion resin. HAS contains

nine cysteines in addition to the Cyst (in wild-type), and all are found in

the N-terminal half of the enzyme. It is conceivable, provided that any of









70

the cysteines other than Cys1 are available, that DON alkylation could

result in non-specific inactivation of the mutant enzymes. A greater degree

of inhibition was obtained when DON was added to the incubation mixture

during enzyme assay. At a fixed DON concentration, the specific activity

of the inhibited enzymes did not decrease with incubation time (data shown

for C1A only). This seemed to indicate that DON inhibition of the

ammonia-dependent reaction of mutants was due to a non-covalent binding

phenomenon rather than to aikylation. However, since enzyme activity

could only be detected after approximately 90 seconds, one cannot rule out

the possibility of a quick enzyme alkylation reaction or a mixed effect. On

the other hand, given the fact that DON is also an analog of glutamine, a

most intriguing possibility arises in which Cys1 mutants incapable of amide

transfer from glutamine are still capable of some glutamine binding. It was

also clearly shown that the wild-type HAS was inhibited by DON in a

pattern similar to that observed in the C1A and C1S mutants. DON

inhibition of the wild-type enzyme deserves further exploration, since

previous work with beef pancreatic asparagine synthetase actually indicated

a moderate increase of the ammonia-dependent activity as a result of DON

treatment (Mehlhaff and Schuster, 1991).

In summary, we have described the construction and characterization

of Cys1 mutants of HAS which indicates that this residue is not only

essential for the glutamine-dependent activity of the enzyme, but may also

be important in its metabolic stability. In addition, the availability of stably

expressed mutants has opened the door for a more extensive exploration of

the relation between the amide transfer and the synthesis domain of the









71

enzyme since they retain many of the features of the wild-type enzyme in

their ammonia-dependent function.















CHAPTER 4
GLUTAMINE INHIBITS THE AMMONIA-DEPENDENT ACTIVITIES OF TWO N-
TERMINAL SUBSTITUTION MUTANTS OF HUMAN ASPARAGINE SYNTHETASE
BY FORMING AN ABORTIVE COMPLEX

Introduction

Eukaryotic asparagine synthetase (AS) catalyzes the following

reactions (Patterson and Orr, 1968):

NH3 + L-Asp + ATP-Mg >L-Asn + AMP + PPi [1]

L-Gln + L-Asp + ATP-Mg > L-Asn + AMP + PPi + L-Glu [2]

L-Gln + H20 > L-Glu + NH3 [3]

It has been reported that the AS gene is among the early delayed genes

that, together with some early immediate genes such as the c-jun oncogene,

are specifically induced by hormonal factor before cells advance from GI

phase to S phase in the cell cycle (Colletta and Cirafici, 1992; Gong and

Basilico, 1990; Hongo, Taketa and Sato, 1989). It appears that the

expression and the activity of AS is closely related to cell differentiation

and, in some cases, oncogenesis (Gong and Basilico, 1990; Hongo,

Sakagami and Sato, 1990; Huang and Knox, 1975). The role of AS

production and its enzymatic activity in transformed cells is not clear.

However, it is believed that the effectiveness of L-asparaginase

chemotherapy in treating certain types of cancers is due to the depletion of

L-asparagine from the circulation (Uren and Handschumacher 1977). A

thorough understanding of the AS mechanism will guide the search for

potent and highly specific AS inhibitors that might be used as alternative

antitumor agents. Meanwhile, as a result of protein sequence alignments











and enzyme activity comparisons, several enzymes including human

asparagine synthetase (HAS) have been identified as purF-type glutamine

amidotransferases which are responsible for the biosynthesis of amino

acids, purine and pyrimidine nucleotides, amino sugars etc. (Mei and

Zalkin, 1989). Therefore, the results of detailed kinetic studies on AS

will also aid the study of the reaction mechanisms of other purF-type

glutamine amidotransferases. A few preliminary kinetic mechanisms for AS

from several sources have been reported (Cedar and Schwartz, 1969a;

Markin, Luehr and Schuster, 1981; Milman, Cooney and Huang, 1980;

Hongo and Sato, 1985; Mehlhaff, Luehr and Schuster, 1985), but the

detailed kinetic and chemical mechanism of AS are not described.

The purF-type amidotransferases, including HAS, have two

functionally distinct domains: one is capable of catalyzing the ammonia-

dependent reaction, and is referred to as the synthesis domain; another

confers the capacity to utilize the glutamine amide, and is referred to as the

glutamine amide transfer (GAT) domain. In AS, the two domains are

fused together with the GAT domain positioned at the N-terminus of the

mature protein (Zalkin, 1992). It has been confirmed that during

asparagine synthesis, an (3-aspartyl-AMP intermediate is formed,

presumably in the synthesis domain (Walsh, 1979; Luehr and Schuster,

1985; Pfeiffer et al., 1987). Very little, however, is known about how the

amide group is transferred to this intermediate either via free ammonia or

directly from glutamine (in reaction 1 and 2 respectively).

According to a previously proposed working model for purF-type

amidotransferases (referred to as Mechanism I), the conserved N-terminal











cysteine is used to form a covalent cysteinyl-glutamine tetrahedral

intermediate which is hydrolyzed in the following step to produce ammonia

and a y-glutamylthioester. The ammonia is then transferred to the synthesis

domain, while the y-glutamylthioester is hydrolyzed so that glutamate is

released and the active thio group of Cys' is regenerated (Mei and Zalkin,

1989). This hypothesis is consistent with several lines of evidence.

First, site-directed mutagenesis has been used to demonstrate that the N-

terminal cysteine is essential for the glutamine-dependent activities (Moraga

et al, 1992; Van Heeke and Schuster, 1989b). Second, glutamine analogs

such as the alkylating agents 6-diazo-5-oxo-L-norleucine (DON) and N3-(4-

methoxyfumaroyl)-2,3-diaminopropanoic acid (FMDP) can inhibits the

glutamine-dependent activities of purF-type amidotransferases via either

reversible inhibition or irreversible inactivation (Mehlhaff, Luehr and

Schuster, 1985; Hartman, 1963; Badet, Vermoote and LeGoffie, 1988).
Third, the proposed y-glutamyl-enzyme intermediates have been isolated

(Levitzki and Koshland, 1971; Chaparian and Evans, 1991). Nevertheless,

this model is still insufficient to explain many other observations, such as

the fact that the amide nitrogen of the substrate glutamine does not

exchange with the nitrogen of exogenous ammonia (Zalkin and Truit,

1977). In addition, for some enzymes, other substrates of the synthesis

reaction are also required for the glutamyl-enzyme intermediate formation

(Hartman, 1963), suggesting that glutamine hydrolysis may be a multistep-

reaction.

The glutaminase activity (reaction 3) probably reflects an uncoupling

of glutamine hydrolysis from the product synthesis. Furthermore, since the









75

proposed cysteinylglutamine tetrahedral adduct has not been detected or

isolated, there is no evidence excluding the possibility that glutamine

hydrolysis is the consequence of complete amide transfer by a different

mechanism. Recently, Richards and Schuster (1992) proposed an

alternative amide transfer pathway for asparagine synthetase (referred to as

Mechanism II). According to this hypothesis, the glutamine amide attacks
the P-aspartyl-AMP intermediate as a nucleophile producing AMP and a y-

glutamine-P-aspartate covalent intermediate. The N-terminal cysteine of

the enzyme, assisted by a basic group, hydrolyzes this new intermediate
producing L-asparagine and a y-glutamyl-enzyme. The y-glutamyl-enzyme

is then hydrolyzed in a similar manner as previously proposed in

Mechanism I. This recent proposal emphasizes the tight coupling between

GAT domain and synthesis domain and differs from the previous one

mainly in two aspects. First, glutamine, instead of being hydrolyzed

prior to the amidotransfer, donates its amide group by actually

participating in the synthesis reaction as a reactant. Second, the substrate

in N-terminal cysteine mediated hydrolysis reaction is not glutamine, but

an intermediate containing glutamine moiety. As a result, no free ammonia

is ever produced by this mechanism.

Unless the reaction intermediate(s) can be detected or isolated, it is

difficult to distinguish Mechanism I and Mechanism II by steady-state

kinetic studies on wild-type purF-type enzymes. Recently, two HAS

mutants have been obtained using site-directed mutagenesis. The N-

terminal cysteine is replaced by alanine and serine in mutant CIA and CIS

respectively (Moraga et al., 1992). Although the N-terminal mutations









76

result in the abolishment of the glutamine-dependent activities, apparent

Km values of the substrates and other basic kinetic characteristics of the

ammonia-dependent reactions of both mutants are very similar to the

corresponding data obtained for the wild-type enzyme. This indicates that

the substitutions of the N-terminal cysteine in these two cases have little

impact on the enzyme conformational integrity. This opens the opportunity

to use glutamine strictly as an inhibitor for these mutants without having to

be confronted by the difficulties of either glutamine utilization or highly

altered enzymatic property. Glutamine inhibitory effects on the ammonia-

dependent reactions of both mutants are reported in this paper. The

significance of these observations with respect to the two proposed purF-

type glutamine amid transfer mechanisms are discussed.

Materials and Methods
Recombinant wild-type human asparagine synthetase (wt HAS) was

overexpressed in yeast Sacchromyces cerevisiae (S. cerevisiae )

AB116/pBS24.1GAS (Sheng et al., 1992a). Mutant HAS CIA and CIS

were overexpressed in yeast S. cerevisiae BJ2168/C1A and BJ2168/C1S

respectively (Moraga et al., 1992; Van Heeke and Schuster, 1989b).

Homogeneous enzymes were obtained using monoclonal antibody affinity

chromatography procedure as previously described (Moraga, 1992; Sheng

et al., 1992a). 14C-[8]-ATP (specific activity: 0.1 mCi/mL) was purchased

from ICN (Costa Mesa, CA). Scintillation fluid ScintiVersemlI* was

purchased from Fisher (Orlando, FL). DE-81 anion-exchange paper was

purchased from Whatman (Hillsboro, OR). Protein concentration assay dye

was purchased from Bio-Rad (Melville, NY). Other chemicals were all the









77

highest grade products from SIGMA (St. Louis, MO). First grade L-

glutamine and L-aspartic acid were purchased from SIGMA and

recrystallized as previously described (Moraga et al., 1992).

Protein concentrations were determined by the method of Bradford

(1976) using pure mouse immunoglobulin G to construct the standard

curve.

The ammonia-dependent asparagine synthetase activities of HAS were

determined using a radiometric assay derived from the method of Allison et

al. (1977). The glutamine stock solution was always made fresh. ATP

stock solution was made by mixing 14C-[8]-ATP and nonradioactive ATP to

a final concentration of 84 mM with a final specific radioactivity of 3x10-2

mCi/mL. A total of 100 p.L of HAS assay mixture contained 85 mM Tris

HCI and 50 mM NaCl. The pH optima for wt HAS, C1A and CIS were

8.0, 9.0 and 8.5 respectively (Moraga and Sheng, et al., 1992). The pH of

the Tris HCI buffer and the substrate stock solutions were adjusted

accordingly. The concentrations of the substrates in each group of

experiments were as indicated in the results. The reactions were initiated by

adding 0.5 tg of purified enzyme. After the reaction mixtures were

incubated at 37 'C for 30 minutes, the reactions were terminated by boiling

for 2 minutes. The blank of each reaction was performed in a similar

manner except that the enzyme was not added. The reaction mixtures thus

obtained were subjected to DE-81 anion-exchange paper chromatography

and the production of AMP in each reaction was analyzed as previously

described (Allison, Todhunter and Purich, 1977). Initial velocities of the

enzyme reactions were obtained from three parallel experiments. The









78

standard deviations were calculated using the following equation: (S is the

S (Xi- Ct)

b n-1

standard deviation

; Xi is the initial velocity; is the average initial velocity; n is the number

of repeats). The average values were taken for graphical analysis using the

least square method. The derivation of the initial velocity equations is

presented in the Appendix of this dissertation.

Results
The ammonia-dependent activities of CIA and CIS at the optimal assay

conditions (pH 8.5, 9.0 respectively and 85 mM Tris HC1, 50 mM NaC1,

10 mM L-aspartate, 5 mM ATP, 8.33 mM MgCl2 and 50 mM NH40Ac)

were both inhibited by low concentrations of glutamine (0.4 to 2.0 mM).

As the concentration of ammonium acetate was lowered to below 25 mM,

glutamine inhibition became more significant (data not shown). In order to

study the glutamine inhibition with respect to ammonia, the ammonium

acetate concentration was varied in the range of 1 to 25 mM at six fixed

concentrations of glutamine in the range of 0 to 5 mM, while the

concentrations of aspartate, ATP and MgCl2 were maintained at 10 mM, 5

mM and 8.33 mM, respectively. Initial velocities were obtained from three

parallel experiments and the average values were used for graphical data

analyses. As shown in Figure 4-1A, the double-reciprocal plots for C1A

were linear when the ammonium acetate concentrations were lower than 20

mM and reached to the lowest points at 20 mM ammonium acetate. As the

concentration of ammonium acetate was raised higher than 20 mM, the















o




(MOn
= ..- ....
8 2 E o


C4)




--r 4 Z
o s -
.5 0 "
E cE .-
so t






< t
25













..-u,

-E U >
o *- --c- >
E09 Em -0










oE_
~ .0 -,


S 0 -





c is -*












o > '-2
22 -
SEs2 o
-' 0-E^f




^S ^Ul
u e 'C ^























00
0











c0
0







o





0


J o
o6
0


'4, 0


(Nl/,u!.) A/I



















8 3
ise
2 E
som



s-0
on



P5,
*? S"
CUV







U o




o -o
.CU
C .

. --
.s^


(J 2
o -o





.g 0 C
. ^ -

0 .)
ss*-





-.. C..)
U CU
8>C
O- ,


o. E


i(.2
O .

5p-* OC
* -o


2S|


asos
SE.2


S0o
II
SE.

.o
2'$


- u
I-
'pm
C ,

U0
S-.
f*'



-u



CU
" 2

co

CU
L-3




5 ."



E- .-
o -


o ZC
CUo

m >>




o -.
-. C ..)
CU CU 4)
C'3 :
M's





C.)





0 ,
(d








CO"
0 > a


aS <.

4


4 0
5' ..
"Co


i^ >- **-






.c
Ego
cc
Q) C









-1

^ll
Be)Q



















E

o

\\- O









CC
6 0






















*M










CO




-o
a- a











0..



-) ,
-c
L.--
a'*









og









aco
oe
u-

ca a


so




r*








U.









84






o





0






0 o




0


o





O 0











A ^ ^ oc s o c o s


sold ~PXoI Jd!-6qnop jo OdoTs









85

double-reciprocal plots started to exhibit upward curvature, suggesting

substrate inhibition. The slopes of the linear portion double-reciprocal

plots for C1A were obtained using the least square method. When these

slopes were replotted versus the glutamine concentrations, a linear

relationship was obtained (Figure 4-1C). Unlike the plots for C1A, linear

double-reciprocal plots were obtained for CIS over the entire ammonium

acetate concentration range tested (Figure 4-1B). The slopes of these

double-reciprocal plots were obtained using the least square method and

replotted versus the glutamine concentrations. As shown in Figure 4-1C,

the replots for CIS was also linear. The graphical analyses presented in

Figure 4-lA, Figure 4-1B and Figure 4-1C suggested that glutamine was an

inhibitor with respect to ammonia. The calculated Ki values of glutamine

in C1A and CIS are 2.9 mM and 1.4 mM, respectively. It appears that the

glutamine inhibitory effects with respect to ammonia are slightly more

significant in CIS than in C1A. It is worth noting that both the

extrapolated double-reciprocal plots for CIS and the extrapolated linear

portion double-reciprocal plots for C1A intersect to the right of 1/v axis.

The effects of glutamine on C1A and CIS with respect to aspartate

were investigated in a similar manner. Aspartate concentration was varied

in the range of 0.5 to 10 mM at six fixed concentrations of glutamine in

the range of 0 to 5 mM, while the concentrations for ammonium acetate,

ATP and MgCI2 were maintained at 15 mM, 5 mM and 8.33 mM

respectively. Initial velocities were obtained from three parallel

experiments and the average values were used for graphical data analyses.

Compared to the glutamine inhibition with respect to ammonia, the effects











of glutamine with respect to aspartate on C1A and CIS were more

complicated and the double-reciprocal plots for the two mutants exhibited

significant differences (Figure 4-2A and Figure 4-2B). As shown in Figure

4-2A, the double-reciprocal plot for C1A was linear when the glutamine

concentration was zero, and became more and more hyperbolic as the

glutamine concentration was increased. The effects of glutamine on C1A

seemed also to depend on the concentration of aspartate. When the

aspartate concentration was varied in the range of 0.5 to 2 mM, an

introduction of low concentration of glutamine concentration (1 mM)

slightly stimulated the enzyme activity. As glutamine concentration was

continuously raised up from 2 mM to 5 mM, the inhibitory effect was

dominant and became more and more apparent. When the aspartate

concentration was varied in the range of 2 mM to 5 mM, increased

inhibition was observed as the glutamine concentration was raised from 1

mM to 3 mM. The inhibition was then gradually reduced as the glutamine

concentration was further increased from 3 mM to 5 mM, however, the

enzyme activity did not return to its initial level. In this case, the

intersecting points fell to the right of 1/v axis. The double-reciprocal plots

for CIS were similar to the double-reciprocal plots for C1A in the high

concentration range of aspartate, but were more hyperbolic with no

common intersecting points. When the glutamine concentration was varied

in the range of 0 to 2 mM, glutamine inhibited CIS in a glutamine

concentration-dependent fashion. As the glutamine concentration was

further increased to 5 mM, the inhibition was gradually weakened, but the

enzyme activity did not regain its initial level. These results are consistent