<%BANNER%>

Intramolecular Tunnel and Regulatory Mechanisms of Asparagine Synthetase (ASNS)

Permanent Link: http://ufdc.ufl.edu/UFE0021729/00001

Material Information

Title: Intramolecular Tunnel and Regulatory Mechanisms of Asparagine Synthetase (ASNS)
Physical Description: 1 online resource (153 p.)
Language: english
Creator: Li, Kai
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: amidotransferase, asns, inhibition, tunneling
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: So far, more and more evidence from several actively studied polyfunctional enzymes indicate the existence of an intramolecular tunnel. Among those enzymes, ammonia/ammonium is the most common intermediate thus far. As one of these enzymes, the tunnel in asparagine synthetase B (ASB) from Escherichia coli is about 20 ? long. It connects two active sites, a glutaminase site and a synthetase site. My work on asparagine synthetase (ASNS), an enzyme involved in leukemia resistance to the treatment using asparaginase, has focused on the following aspects: (1) kinetic characterization of ASNS; (2) structure activity relationships; (3) regulatory mechanisms. In this work, I developed a first quantitative NMR based assay (Li, K., et al, Biochemistry, 2007, 46(16), 4840-4849) and studied the efficiency of ammonia tunneling in ASB catalyzed reaction. The function of bioactive macromolecule can only be understood under the structural context. Therefore, to study the function of the intramolecular tunnel in ASB, first I identified the tunnel residues by computational methods. Most of the tunnel residues are conserved from prokaryote to eukaryote species. The studies of several tunnel mutant showed that one of the mutants may have a blocked tunnel. Among the family of amidotransferases, the two half reactions are strictly coupled. While ASNS can catalyze the hydrolysis of glutamine without other substrates, this raises the interest in the regulatory mechanisms of this enzyme, which prevents it from consuming glutamine and releasing free ammonia. As part of my PhD research, I investigated the mechanisms of product inhibition to this enzyme. Under physiological conditions, the activity of this enzyme may be inhibited significantly by asparagine, which prevents cells from wasting glutamine and producing toxic ammonia in vivo. So far, only one crystal structure of glutamine dependent asparagine synthetase was reported. To further understand the function of ASNS, especially of human enzyme, more crystal structures are needed. Therefore, co-working with Alexandria Berry, I investigated the binding of substrate/transition state analog to hASNS and successfully prepared inhibitor bound enzyme. The prepared enzyme has been sent for crystal analysis.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Kai Li.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Richards, Nigel G.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021729:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021729/00001

Material Information

Title: Intramolecular Tunnel and Regulatory Mechanisms of Asparagine Synthetase (ASNS)
Physical Description: 1 online resource (153 p.)
Language: english
Creator: Li, Kai
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: amidotransferase, asns, inhibition, tunneling
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: So far, more and more evidence from several actively studied polyfunctional enzymes indicate the existence of an intramolecular tunnel. Among those enzymes, ammonia/ammonium is the most common intermediate thus far. As one of these enzymes, the tunnel in asparagine synthetase B (ASB) from Escherichia coli is about 20 ? long. It connects two active sites, a glutaminase site and a synthetase site. My work on asparagine synthetase (ASNS), an enzyme involved in leukemia resistance to the treatment using asparaginase, has focused on the following aspects: (1) kinetic characterization of ASNS; (2) structure activity relationships; (3) regulatory mechanisms. In this work, I developed a first quantitative NMR based assay (Li, K., et al, Biochemistry, 2007, 46(16), 4840-4849) and studied the efficiency of ammonia tunneling in ASB catalyzed reaction. The function of bioactive macromolecule can only be understood under the structural context. Therefore, to study the function of the intramolecular tunnel in ASB, first I identified the tunnel residues by computational methods. Most of the tunnel residues are conserved from prokaryote to eukaryote species. The studies of several tunnel mutant showed that one of the mutants may have a blocked tunnel. Among the family of amidotransferases, the two half reactions are strictly coupled. While ASNS can catalyze the hydrolysis of glutamine without other substrates, this raises the interest in the regulatory mechanisms of this enzyme, which prevents it from consuming glutamine and releasing free ammonia. As part of my PhD research, I investigated the mechanisms of product inhibition to this enzyme. Under physiological conditions, the activity of this enzyme may be inhibited significantly by asparagine, which prevents cells from wasting glutamine and producing toxic ammonia in vivo. So far, only one crystal structure of glutamine dependent asparagine synthetase was reported. To further understand the function of ASNS, especially of human enzyme, more crystal structures are needed. Therefore, co-working with Alexandria Berry, I investigated the binding of substrate/transition state analog to hASNS and successfully prepared inhibitor bound enzyme. The prepared enzyme has been sent for crystal analysis.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Kai Li.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Richards, Nigel G.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021729:00001


This item has the following downloads:


Full Text





INTRAMOLECULAR TUNNEL AND REGULATORY MECHANISMS OF
ASPARAGINE SYNTHETASE (ASNS)




















By

Kai Li













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


2007


































2007 Kai Li

































To my wife, Zhen, and my parents















ACKNOWLEDGMENTS

I acknowledge the financial support for the travel grants from College of Liberal Arts and

Sciences and Graduate Student Council, which made it possible to present my work at several

conferences.

I thank Dr. Nigel G. J. Richards, my supervisor, who guided me into enzymology research

field by working on asparagine synthetase. I am grateful to my doctoral dissertation committee,

Dr. Nicole A. Horenstein, Dr. Daniel L. Purich, Dr. Adrian E. Roitberg, and Dr. Jon D. Stewart,

for their helpful advices on my project. Furthermore, I would like to thank Dr. Gail E. Fanucci,

Dr. Ion Ghiviriga, Dr. Thomas J. Lyons and Dr. David N. Silverman for their helps.

Special thanks go to Dr. Drazenka Svedruzic, Dr. Jemy A. Gutierrez, Dr. Tania C6rdova de

Sintjago and Dr. Patricia Moussatche, for their helpful discussions with my projects; William

Beeson and Alexandria Berry, two undergraduates, with whom I accomplished this work; Xiao

He, Dr. Hui Jiang, Sangbae Lee, Mario E. Moral, and Dr. Yong Ran, for their friendship and

helps.

I am grateful to my mom and stepfather for their unconditional support of any decision I

made, to my parents-in-law for their care by sending me email everyday and to my wife for the

happy life we shared for almost ten years.











TABLE OF CONTENTS



A C K N O W L E D G M E N T S ..............................................................................................................4

L IST O F TA B LE S ...... .. .... ........................................ .............................. 8

LIST O F FIG U RE S ................................................................. 9

ABSTRAC T ................................................. ............... 12

CHAPTER

1 IN TR OD U CTION .......................................................................... .. ... .... 14

Asparagine Synthetase (ASNS) and the Intramolecular Tunnel .........................................14
The Intram molecular Tunnel .............................................................................................. 17
Glutaminase Active Site and the Mouth of the Tunnel........... ...................................17
The Body of the Tunnel ................ ........... ............... .. .............. 18
The B bottom of the Tunnel .............................. .... ..... ....... .................. ............... 19
The Interface of the Two Domains and Water ............... ........... ..................................19

2 A gHMQC-BASED NMR ASSAY FOR INVESTIGATING AMMONIA
CHANNELING IN GLUTAMINE DEPENDENT AMIDOTRANSFERASES ................... 35

Introduction ................... ............................................................. ................35
M material and M methods .............. ................. ........... ................... ......... 37
M materials ...............................................................................................3 7
NM R M easurem ents .................. ..................................................... .. 38
C om petition E xperim ent ........................................................................ ...................38
15N Exchanging Experim ent..................... ............. .................................................... 39
HPLC-Based Determination of Total L-Asparagine.....................................................40
Kinetic Simulations .................................. .. .......... ........... 41
Results and D discussion ..................................... .. .... ...... .. ............41
N M R A ssay ..............................................................................4 2
C om petition E xperim ents ...................................................................... ..... ................42
15N exchange E xperim ents ........................................... .................. ............... 46
C onclu sions.......... .............................. ................................................48

3 CHARACTERIZATION OF A TUNNEL MUTANT.........................................................61

In tro d u ctio n ................................................... ........................... ................6 1
M material and M methods .......................... ........... .. ........... ........... ..... 62
M a te ria ls .................................................................................................................... 6 2
C om petition E xperim ent ........................................................................ ...................63
NMR Measurements and HPLC assay ......................................................64









R e su lts an d D iscu ssio n ....................................................... ........................ ....................6 4
C o n c lu sio n s ............................................................................................................................. 6 5

4 ASPARAGINE INHIBITION TO ASPARAGINE SYNTHETASES AND ITS
EFFECT ON THE QUATERNARY STRUCTURE OF THE ENZYMES...........................74

In tro du ctio n ...............................................................7 4
M materials and M methods ........................ .. ........................ .. .... ........ ........ 75
M a te ria ls .........................................................................................................7 5
E n zy m e A ssay s ......................................................................7 5
A sparagine Inhibition Studies ........................................................................................ 76
Size Exclusion Chromatography to Determine the Quaternary Structure of
A sparagine Synthestases ........................................................................................ 76
The Effect of Asparagine on the Quaternary Structure of ASB and hAS .......................76
SEC Studies of ASB with Different Amount of ASB ..............................................77
R e su lts ...............................................................................................7 7
P ro d u ct In h ib itio n .................................................. ...... ...... ......... ........................ 7 7
Asparagine Affects the Quaternary Structure of ASB, But Not Human Enzyme...........77
The Equilibrium between the Dimer and Monomer Form of ASB...............................79
Construction of A sparagine Inhibition M odel ...................................... ................... 80
D iscu ssio n ........... ................. .......................................................................................... 8 3
C onclu sions.......... ..........................................................86

5 THE UTILIZATION OF DIFFERENT NITROGEN SOURCES BY ASPARAGINE
SYNTHETASE: EVOLUTION FROM Escherichia coli TO HUMAN .............................103

In tro d u ctio n ................... ...................1.............................3
M material and M methods ..................................................................... ..................... 105
M a te ria ls ......................................................................................1 0 5
Com petition Experim ent .............. ..... ........... ..... .................... ............... 105
NMR Measurements, HPLC assay and Kinetic Simulations ............. .................105
R results and D discussion ........... .. ....................................... ..................... 106
C onclusions.....................................................................109

6 PREPARATION OF LIGAND BOUND hASNS INHIBIITION AND BINDING
S T U D IE S .......................................................................................................1 1 5

In tro du ctio n ................... ...................1...................1.........5
M materials and M methods ............... ............................ .............. .. ...... ..... ...... 116
M a te ria ls .................................................................................................................. 1 1 6
E nzym e A ssay s .....................................................................................116
Loss of Glutaminase Activity with Time at Different Temperature ...........................17
D ON Inhibition .................. .... .............................. ..................... 117
Loss of Synthetase Activity with Time at different Temperature ........... .....................118
Inhibition of the DON Inhibited Enzyme by Sulfoximine ................. .............. .......118
R e su lts ............... ........... ...................................................................................................... 1 18
Stability Experim ent .................. .................................... ................. 118



6









hASNS lost glutaminase activity with time at room temperature and 37 C .........119
The loss of NH3 dependent synthetase activity with time at room temperature
and 37 .......................................................................... .... .. ........ ........ ...............119
Inhibition to hA SN S by D O N .................... .................................................. ... 120
DON inhibits 90% of glutaminase activity of hASNS after incubating for 20
m minutes ....................... ..... ......... .. .. .............. ................120
DON had little effect on the rate of ammonia dependent synthesis of
asparagine. ...........................................................120
Inhibition by Adenylated Sulfoximine ...................................................................121
Inhibition to the ammonia dependent synthetase activity of stock hASNS ...........121
Inhibition to the ammonia dependent synthetase activity of DON incubated
h A S N S ........................................................................ 12 2
D iscu ssion .. .... ............................................. .......................................... 123

APPENDIX

A SEQUENCE ALIGNMENT OF TUNNEL RESIDUES IN ASB ......................................142

B THE KINETIC MODEL FOR EXCHANGE EXPERIMENT .........................................143

R E F E R E N C E L IS T ..................................................................................................... .... 14 6

B IO G R A PH IC A L SK E T C H ......................................................................... ........................ 153









LIST OF TABLES


Table page

1-1 Residues in Gin binding pocket and possible roles ................................. ............... 31

1-2 Size of narrow part of the tunnel ............................................................. .....................32

1-3 Residues in the interface between two domains of E. coli ASB.......................................33

2-1 AS-B catalyzed incorporation of 15N into L-asparagine in the steady-state
com p petition assay s ....................................................... ................ 6 0

3-1 The kinetic parameters for glutaminase activity, ammonia dependent synthetase
activity and glutamine dependent synthetase activity of both wild type ASB and
A3 88L mutant ................................... .................. ............... ........... 73

4-1 Retention time of ASB peaks in SEC studies ............ ........... ........................96

4-2 The calculated MW and predicted structure of ASB corresponding to each peak............97

5-1 hASNS catalyzed production of 15N into L-asparagine in the steady-state competition
assay s .. ....................... ..... .............. ..... ......................... ................. 1 14

6-1 Percent glutaminase activity at different incubation time without DON .........................136

6-2 Percent ammonia dependent synthetase activity without inhibitors ...............................137

6-3 The loss of glutaminase activity due to DON inhibition .................................................138

6-4 Synthetase activity of the DON inhibited and control enzyme after incubation
determined using the pyrophosphate assay............... ................. ............... 139

6-5 Kinetical constant for sulfoximine inhibition to ammonia dependent synthetase
activity of hA SN S ..................................... ................. ........... .. ............ 140

6-6 Parameters for inhibition of stock hASNS and control hASBS by sulfoximine ............141









LIST OF FIGURES


Figure p e

1-1 Proposed mechanism of hydrolysis of glutamine by E. coli asparagine synthetase B. .....21

1-2 Synthesis of asparagines through an intermediate P-aspartyl-adenylate by ASB..............22

1-3 Cartoon of glutaminase domain of E. coli ASB. Glutamine is showed by spacefill.
Alal(Cysl) is showed by ball and stick............................................................... 23

1-4 The tunnel w within A SB .......................................................................... .....................24

1-5 Cysl-Arg-Gly-Asn-Asp is a common motif of the glutamine binding site for Ntn
subfamily of glutamine dependent amidotransferases ..................................................25

1-6 T he m south of the tunnel .............................................................................. .............. 26

1-7 The narrowest part of the tunnel formed by M120, 1143, N389, A399 and carbonyl
group of A 3 88 .............................................................................................................27

1-8 T he tunnel in E coli A SB ......................................................................... ...................28

1-9 The body of the tunnel ..................................... ................. .......... ...... ....... 29

1 1 0 T h e G lu 3 5 2 ........................................................................................................3 0

2-1 Two possible pathways for the ammonia transfer .................................. .................50

2-2 gH M Q C 1H -N M R spectrum ..................................................................... ..................51

2 -3 S tan d ard cu rv e s ............................................................................................................ 5 2

2-4 14N/15N incorporation ratios in L-asparagine formed by the AS-B synthetase reaction
under com petition assay conditions ........................................................ ............. 55

2-5 N orm al H -N M R spectrum ......................... ...... .................................... ............... 56

2-6 Glutamine-dependence of the exchange of 15NH3 into L-glutamine under exchange
assay co n d itio n s ......................................................................... 5 7

2-7 Exchange of 15NH3 into L-glutamine during the glutaminase reaction catalyzed by
A S -B ......................................................................................... . 5 8

3-1 One of the conserved tunnel residues, Ala388 ............................. .....................67

3-2 Competition results for A388L ASB mutant, as well as WT ASB............................... 68









3-3 Production of asparagine catalyzed by A388L ASB mutant and WT ASB with varied
g lu tam in e ................... ......................................................... ................ 6 9

3-4 Production of asparagine catalyzed by A388L ASB mutant and WT ASB with varied
am m o n iu m ............................................................................. 7 0

3-5 The nitrogen exchange catalyzed by A388L mutant and WT ASB...............................71

3-6 The polarity of the tunnel residues ......................................................... .....................72

4-1 Kinetic studies of asparagine inhibition to ASNS .................................. ............... 90

4-2 SEC standard curve............. ..... ......................... ........ 93

4-3 SEC analysis of quaternary structure of ASNS with different concentration of
asparagine in m obile phase ........................................................................... 94

4-4 SEC analysis of quaternary structure of ASB with different amount of enzyme ..............98

4-5 Fraction of dimer vs concentration of asparagine............... ...........................................99

4-6 The interface of the tw o subunits ...................................................................... 100

4-7 Sequence alignm ent of the interface residues............................................................... 101

4-8 The glutaminase activity of ASNS around physiological conditions............................102

5-1 Quantification of 15N-asparagine and 14N-asparagine using NMR assay and HPLC
assay ................... ........... ......................... ...........................1 10

5-2 Simulation model for the competition reactions catalyzed by hASNS ............................111

5-3 Simulations of competition reactions catalyzed by hASNS ..........................................112

5-4 Utilization of ammonia and glutamine as nitrogen source by ASB and hASNS .............113

6-1 The m echanism s of A SN S catalyzed reaction ................................................................. 127

6-2 DON reacts with Cysl residue of ASNS and forms covalent adduct............................128

6-3 The inhibitor adenylated sulfoximine (right) mimics the nucleophilic attacking of P-
aspartyl-AM P by am m onia ........................................................................... .. 129

6-4 The glutaminase activity of hASNS decreased exponentially with time.........................130

6-5 The ammonia dependent synthetase activity of hASNS decreased with time ...............131

6-6 The glutaminase activity of DON inhibited hASNS decreased exponentially with
tim e .......................................................... .............................. . .1 3 2









6-7 The inhibition to free hASNS by sulfoximine.........................................133

6-8 Inhibition to control hASNS by sulfoximine...........................................134

6-9 Inhibition to DON inhibited hASNS by sulfoximine .............. ................................. 135









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 INTRAMOLECULAR TUNNEL AND REGULATORY MECHANISMS OF
ASPARAGINE SYNTHETASE (ASNS)

By

Kai Li

December 2007

Chair: Nigel G. J. Richards
Major: Chemistry

So far, more and more evidence from several actively studied polyfunctional enzymes

indicate the existence of an intramolecular tunnel. Among those enzymes, ammonia/ammonium

is the most common intermediate thus far. As one of these enzymes, the tunnel in asparagine

synthetase B (ASB) from Escherichia coli is about 20 A long. It connects two active sites, a

glutaminase site and a synthetase site.

My work on asparagine synthetase (ASNS), an enzyme involved in leukemia resistance to

the treatment using asparaginase, has focused on the following aspects: (1) kinetic

characterization of ASNS; (2) structure activity relationships; (3) regulatory mechanisms.

In this work, I developed a first quantitative NMR based assay (K. Li et al, Biochemistry,

2007, 46[16], 4840-4849) and studied the efficiency of ammonia tunneling in ASB catalyzed

reaction.

The function of bioactive macromolecule can only be understood under the structural

context. Therefore, to study the function of the intramolecular tunnel in ASB, first I identified the

tunnel residues by computational methods. Most of the tunnel residues are conserved from

prokaryote to eukaryote species. The studies of several tunnel mutant showed that one of the

mutants may have a blocked tunnel.









Among the family of amidotransferases, the two half reactions are strictly coupled. While

ASNS can catalyze the hydrolysis of glutamine without other substrates, this raises the interest in

the regulatory mechanisms of this enzyme, which prevents it from consuming glutamine and

releasing free ammonia. As part of my PhD research, I investigated the mechanisms of product

inhibition to this enzyme. Under physiological conditions, the activity of this enzyme may be

inhibited significantly by asparagine, which prevents cells from wasting glutamine and

producing toxic ammonia in vivo.

So far, only one crystal structure of glutamine dependent asparagine synthetase was

reported. To further understand the function of ASNS, especially of human enzyme, more crystal

structures are needed. Therefore, co-working with Alexandria Berry, I investigated the binding

of substrate/transition state analog to hASNS and successfully prepared inhibitor bound enzyme.

The prepared enzyme has been sent for crystal analysis.









CHAPTER 1

INTRODUCTION

Asparagine Synthetase (ASNS) and the Intramolecular Tunnel

With the advances in molecular biology and crystallography, many complex proteins have

been structurally determined. More and more evidences show that intramolecular tunnel exists in

many enzymes with multiple catalytic sites. Since the first intramolecular tunnel within

tryptophan synthase was found in 1988 (1), tunnels within at least 11 enzymes have been

demonstrated (2, 3). Among these tunnels, NH3/NH4+ tunnels in amidotransferases are the most

common. It has been determined in carbamoyl phosphate synthetase (CPS), GMP synthetase,

glutamine phosphor-ribosylpyrophosphate amidotransferase (glutamine PRPP amidotransferase),

asparagines synthetase B (ASB), glutamate synthase, imidazole glycerol phosphate synthase,

glucosamine 4-phosphate synthase, and recently, CTP synthase (4). All these enzymes use

glutamine as nitrogen source donor. First, glutamine is hydrolyzed into glutamate (product) and

NH3/NI4+ (intermediate) at one active site. Then NH3/NI4+ is transported to another active site

through an intramolecular tunnel for next half-reaction. It seems that all proteins that use

glutamine to produce NH3/NH4+ as intermediate will have this kind of intramolecular tunnel.

Among those enzymes with NH3/NH4+ tunnel identified thus far, asparagine synthetase

(include ASB and mammalian AS), Glucosamine 4-phosphate synthase, Glutamine PRPP

amidotransferase, and glutamate synthase belong to Ntn hydrolases / class II glutamine

amidotransferase subfamily, since they all have an N-terminal nucleophile cysteine, which

catalyzes the hydrolysis of glutamine (5, 6).

Asparagine synthetase catalyzes ATP-dependent conversion of aspartate to asparagine. The

discovery of asparagine synthetase resulted from the studies about the metabolism of amino

acids. Asparagine is one of the amino acids that have been lately studied. After the first









discovery of proteins with asparagine synthetase activity in the early 1950s, asparagine

synthetase has been found in many species including prokaryote and eukaryote. The studies for

asparagine synthetase raised even more interests because of the correlation between asparagine

level in human cells and the resistance to the treatment of asparaginase in certain types of cancer

cells such as acute lymphoblastic leukemia (ALL) (7, 8).

In E. coli, two unlinked genes with little similarities, asnA and asnB, encode asparagine

synthetase (9-12). While asparagine synthetase A (ASA) from asnA can only use ammonium as

nitrogen source, asparagine synthetase B (ASB, EC 6.3.5.4) can use both ammonium and

glutamine, with a preference of glutamine. ASB belongs to class II or Ntn amidotransferase

subfamily and catalyzes following reactions.

Gln + H20 + ATP + Asp 4 Glu + AMP + PPi+ Asn (R1)
NH3 + ATP + Asp AMP + PPi + Asn (R2)
Gln + H20 Glu + NH3 (R3)


Because of its 37% identity to human asparagine synthetase and the same reactions that

catalyzes, ASB has served as a model system to understand the biosynthesis and metabolic

control of asparagine. Up to now, the function of ASB from E. coli has been extensively studied

(13). Boehlein et al (1994) reported that both C1A and C1S mutant of ASB lose glutaminase

activity completely, although these two mutants still can bind glutamine. This result confirms the

important catalytic role of N-terminal cysteine residue in hydrolysis of glutamine. By analogy

with cysteine proteases, the thiolate anion of Cysl residue act as nucleophile (Figure 1-1) and

attack the y-carbonyl group of glutamine to form a thioester intermediate and ammonia. The

thioester then reacts with water to produce glutamate. This mechanism is strongly supported by

the report that a covalent intermediate during the glutaminase reaction were isolated and

proposed to be y-glutamyl thioester (14). In 1998, the same research group reported their studies









about the kinetic mechanism of E. coli ASB. Scheme 1-1 shows the kinetic mechanism based on

their report. They also provided the evidence by 180 transfer experiment that supports the

existence of intermediate, P-aspartyl-adenylate (P-aspartyl-AMP), which reacts with NH3/NH4+

produced by hydrolysis of glutamine (or NH3i/NH4 from bulk solution) to synthesize asparagine

(Figure 1-2).

After the crystal structure of C1A mutant of ASB was reported (15), there is no doubt

about the existence of an intramolecular tunnel in ASB. The crystal structure clearly shows two

active sites, named glutaminase site and synthetase site. Glutaminase domain comprises the N-

terminal half of the protein (1-190) and synthetase domain the C-terminal half (208-553). The

two halves are connected by a coil of 17 residues. The distance between two active sites is about

20 A. Like other members of Ntn subfamily, glutamine domain is a appa four layer sandwich

structure. Glutamine binding pocket is between the two antiparallel P-sheets at one edge of the

domain (Figure 1-3).

Glutaminase domain catalyzes the hydrolysis of glutamine (R3). Synthetase domain

catalyzes ATP-dependent synthesis of asparagine (R2), using NH3/NH4+ produced by R3. There

are two possible pathways for NH3/NH4+ produced at glutaminase site to get to synthetase site.

One is intermolecular pathway. That is, NH3/NH4+ first is released by ASB and then equilibrates

with bulk solution before the synthesis reaction. The other pathway is through intramolecular

tunneling. The kinetic results rule out the first one as the major pathway, if it does exist. Suppose

NH3/NH4+ goes through bulk solution, it will take some time for NH3/NH4+ cumulating and

reaching a concentration at which the synthetase activity can be detected. This means there

would be a lag for the glutamine dependent synthetase activity. Furthermore, the kcat value of

glutamine dependent synthetase activity should be same as that of ammonia dependent activity.









This is inconsistent with the kinetic results. There must be an intramolecular pathway (tunneling)

to transport NH3/NH4+.

To date, ASB has been extensively studied by steady state studies (16-24). Although there

is no doubt for the existence of tunnel in ASB, the mechanisms of ammonia tunneling are still

unclear. In order to understand vectorial transport of NH3/NH4+ under structural context, I

explored the structure of the tunnel,

The Intramolecular Tunnel

Determination of residues that involved in NH3/NI4+ tunnel by experiment is the essential

step for understanding the transportational mechanisms of NH3/NH4+, synchronization of the two

active sites, and therefore, the catalytical mechanisms of asparagine synthetase.

41 residues were determined involving in the NH3/NH4+ tunnel within ASB based on the

crystal structure (PDB code: 1CT9) of C1A mutant of E. coli ASB. These residues can be

arbitrarily divided into 3 parts, the mouth, the body, and the bottom. Figure 1-4 shows the cavity

formed by these residues with Gln and AMP binding. By aligning sequences from 21 species,

including bacteria, plant and animal, prokaryote and eukaryote, 21 out of 41 residues mentioned

above are identical. 7 are almost absolutely conserved with only one exception. Most of the rest

are substituted by similar residue such as L by I or by V (Appendix A).

Glutaminase Active Site and the Mouth of the Tunnel

Table 1-1 lists all the residues in glutamine active site of E. coli ASB. The distinctive

feature of Ntn subfamily is that they all contain several invariant residues. Cysl-Arg49-Asn74

Gly75-Asp98 motif is distinctive feature of glutamine binding site of Ntn subfamily (E. coli ASB

numbering) (Figure 1-5). Arg49 and Asp98 account for the specific binding of glutamine.

Carboxyl group of glutamine forms ionic bond with Arg49 as well as hydrogen bonds with

backbone of Ile52 and Val53, while the amino group interacts with two negative-charged









residues Glu76, Asp98 and the backbone of Gly75. The distance between carboxyl O atom of

Asp98 and amino N of Gin is 2.56 A, indicating a very strong interaction. Meanwhile, Gly75

also forms a strong hydrogen bond with the carbonyl/amide group of Gln with an N-O distance

of 2.53 A. Carbonyl/amide group of Gln is also hydrogen-bonding with sidechain of Asn74.

With the amino group of Cysl, these two fingerprint residues are believed to provide an

oxyanion hole that stables transition state during the hydrolysis of glutamine. N/amide of Gln

also interacts with carbonyl O of backbone of Leu50.

The amide group of Gln are surrounded by Ala 1 (Cys 1), Leu 50, Ile 52, Asn 74, Gly 75,

Glu 76 (Figure 1-6), which form the mouth of the tunnel. These residues are highly conserved

(Table 1-1, highlighted residues) and probably play an important role not only in binding of Gln

and catalysis, but also in NH3 releasing.

The Body of the Tunnel

In the interface of the two domains, Met 120, Ile 143, Asn 389, Ala 399, plus the carbonyl

group of the backbone of Ala 388 form the narrow part of the tunnel, with a diameter of about 5-

6 A (Figure 1-4, table 1-2 & Figure 1-7).

His 29, Arg 30, Gly 31, Pro 32, Tyr78, Met392, Ser 393, Gly 396, Val 397, Glu 398 and

Arg400 locate between the narrow part and the mouth of the tunnel (Figure 1-8A, B) and form

the upper part of the tunnel. This part of the tunnel contains an absolutely conserved residue

cluster in asparagine synthetases from different species, His29 -- Arg30 Gly31 -- Pro32 -- Asp33.

Arg30 and Gly31 are also fingerprint residues of Ntn subfamily. Other conserved residues in this

part include Gly396 and Glu398. The residues in this part can not form a ring by themselves like

the residues in the other part. They fill the gap between mouth and tunnel part 2. The size of

tunnel here is bigger than other parts.









The lower part of the tunnel comprises Ile142, Leu232, Met329, Val344, Leu345, Ser346,

Cys385, Ala388, Val401, Leu404 and sidechain of Glu348. Some of these residues are also in

the ATP binding site (Figure 1-8C), such as Leu232, Leu345 and Ser346. Most of the residues in

this part are hydrophobic (Figure 1-9) and conserved.

The Bottom of the Tunnel

This part of tunnel is formed by Asp 238, Gly 347, Glu 348, Gly 349, Ser 350, Asp 351,

Glu 352, Tyr 357, Leu 380 and Asp 384. Most of the residues are in ATP binding site (Figure 1-

8D). An interesting feature of this part is that almost all the residues are polar and 5 out of 10 are

negative-charged residues (Figure 1-9). These negative-charged residues, plus Tyr357 form the

end of the tunnel, which is presumed as the NH3/NH4 binding site. The carboxyl group of Glu

352 is right at the bottom of the tunnel and about 4-7 A far away from AMP. This negative

charge may provide the driving force to transport NH3/NH4 The a-phosphate group of AMP

(ATP) is on the side of this part of tunnel (Figure 1-10).

The Interface of the Two Domains and Water

The fact that binding of ATP stimulates glutaminase activity of E. coli ASB suggests

conformational changes during the catalysis (17). The signal between catalytic domains should

arrive through those residues in the interface, especially those that interact with other residues

from different domain (Table 1-3).

7 water molecules are in the tunnel, water 2, 3, 50, 190, 690, 709 and 861. Water 709 is the

only one in the ATP binding site with AMP as the ligand. Since the kinetic data suggest no ATP

hydrolyzing during catalysis, water molecules are supposed to be excluded when ATP and Asp

bind the enzyme. The rest are all in the interface between the two domains.

ASB is an ideal system to study highly conserved NH3/NI4+ tunnel in asparagine

synthetase. With successful techniques of cloning, expression and purification (16), coupled with









the well-developed assay methods such as Glutaminase assay (17), PPi assay (22), and HPLC

assay (15), the intramolecular tunnel of ASB were studied and the structure activity relationship

were investigated.













H3Gin I IN
Gn H




EH
Cys1


H3N H HN H


H H
0 E0


Tunneling


Thioester


Cys1 Cysl Cys1


Proposed mechanism of hydrolysis of glutamine by E. coli asparagine synthetase B. The
thiolate anion group of Cys 1 acts as nucleophile and form an ES thioester.


Figure 1-1.

























ATP PPi

H3N+ 0


O
0

Asp


NH, AMP


H W NH

0

Asn


Figure 1-2. Synthesis of asparagines through an intermediate j-aspartyl-adenylate by ASB.























Cartoon of glutaminase domain of E. coli ASB. Glutamine is showed by spacefill.
Alal(Cysl) is showed by ball and stick.


Figure 1-3.



























The tunnel within ASB. Tunnel residues are showed in sticks, colored by cpk. The cavity
(tunnel) is brown. Glutamine (left) and AMP (right) are showed in spacefill, colored by cpk.
Water (red) and heavy metal ions (grey) are showed in small-size of spacefill.


Figure 1-4.














ASP98 ASP123
ARG9 ARG73



GLY7 GLN106 GLU Li5300







A B "






GLY228 ONL2511 GLY ONL1

ASN227 CYS1 GLY4 CYS1







Cysl-Arg-Gly-Asn-Asp is a common motif of the glutamine binding site for Nt subfamily
of glutamine dependent amidotransferases A. asparagine synthetase B from Escherichia
coli (PDB code: ICT9, modified). B. Glucosamine 4-Phosphate Synthase from Escherichia
coli (PDB code: 1JXA and 1XFF, modified). C. Glutamate Synthase From Synechocystis
Sp (PDB code: O1FE). D. Glutamine Phosphoribosyl-pyrophosphate (Prpp)
Amidotransferase from Escherichia coli (PDB code: 1ECC).


Figure 1-5.























The mouth of the tunnel. 6 highly conserved residues, Ala 1 (Cys 1), Leu 50, Ile 52, Asn 74,
Gly 75, Glu 76, surround the amide group of glutamine and form the mouth of the tunnel.
A. Mouth of the tunnel with glutamine. Probe radius is 1.4 A. B. Relative position to AMP.
The mouth of tunnel is showed in green.


Figure 1-6.
























Figure 1-7. The narrowest part of the tunnel formed by M120, 1143, N389, A399 and carbonyl group of
A388.

















































Figure 1-8. The tunnel in E. coli ASB. Mouth (green); Upper part (yellow); narrow part (brown);
Lower part cyann); Bottom part (pink). Gin (left) and AMP (right) are showed in spacefill;
Glu352 are showed in ball and stick form and colored by cpk.


I


I. J
























The body of the tunnel is mainly formed by hydrophobic residues (yellow) and the bottom
almost totally by polar (green) and acidic (red) residues. Several basic residues (blue) close
to glutamine site.


Figure 1-9.























Figure 1-10. The Glu352 is right at the bottom of the tunnel. AMP (spacefill) is on one side. The hole
shows the tunnel. Mouth (green); Upper part (yellow); narrow part (brown); Lower part
cyann); Bottom part (pink).











Table 1-1. Residues in Gln binding pocket and possible roles


Re e Functional H-bond/salt Distance Cn
Resname ( Conservation Roles
group bridge (A)

Alal backbone -- -- Forming tunnel
----------------------- ---------- conserved
(Cysl) sidechain -- Nucleophilic catalysis
His47 sidechain -- -- H or F Active site
sidechain EN-H...01 3.09
Arg49 idh -------- ---- conserved specificity
sidechain TIN-H...01 2.97
backbone O...H-NE 3.07 Binding of Gln, catalysis and
Leu50 ----------- ------------------ ------ conserved
sidechain -- --forming tunnel?
Ser51 whole -- -- Not cons. Active site
backbone N-H...02 2.79 or Binding of Gln and forming
Ile52 --------------- ------------------- ------------- l or V
sidechain -- --tunnel?
backbone N-H...02 2.77
Val53 ---- ----------------- Not cons. Gln binding?
sidechain -- --
Gly58 whole -- -- conserved Active site?
Gln60 whole -- -- conserved Active site?
Leu62 sidechain -- -- Cons. subst. Active site
Val73 whole -- -- Not cons. Active site
Asn74 backbone -- -- conserved Specificity and forming
Asn74 ------------- conserved
sidechain ND-H...OE 2.97 oxyanion hole and tunnel
-G 5 O...H-N 3.12 Specificity and forming
Gly75 -- ------------------------------ conserved tne
-- N-H...OE 2.53 oxyanion hole and tunnel
backbone Gin binding and forming
Glu76 -------------------------------------- conserved n
sidechain OE1...H-N 3.09 tunnel?
Ser97 whole -- S, C, V Active site
backbone -- --
Asp98 --------------- --------------- conserved Specificity
sidechain OD1...H-N 2.56
Cys99 whole -- -- Not cons. Active site
H203 -- O-H...OE 2.80 Reactant?
H2030 -- -H...01 2.89
Note: Those residues forming mouth of the tunnel are highlighted by yellow. Fingerprint residues of Ntn
subfamily are showed in red










Table 1-2. Size of narrow part of the tunnel


Residues Atoms Distance (A)
Metl20 Asn389 CE -- ND 5.58
Metl20 Asn389 SD --ND 5.51
Ala399 Ile143 CD1 6.27
Ala399 Ile143 CG1 6.85













Table 1-3. Residues in the interface between two domains ofE. coli ASB

(Those residues forming tunnel are showed in red)


Residues Atoms Distance (A) Conservation

As 5 OD1 2.77 not
--- ----l~~~~~~~~01~~ 2.77
Arg193 NH1 not
Asp54 N D or G
----- ------ 2.81 --- ---- D-D or G-E
Asp226 OD1 Dor E
Ile52 O 161, 5V
2.74
Asp226 N 17D, 4E
Glu76 OE1 Conserved
2.51 ---------Conserved----------
Glu398 OE1 Conserved
Asp33 OD1 2.76 Conserved
~~~~~~p'--- ----1~~D~~ 2.76
Lys342 NZ 18K except mammalian
Lys342 2.76 18K except mammalian
------------------ 2.76 ------------
Glu398 N Conserved
Gln 18 NE 92 14D, 4Q, 1R, 1T, 1N
Asp405 OD1 conserved
His139 NE2 2 7P, 6A, 5H, 3T
2.85 -------------
Leu404 O Conserved
Glu173 OE1 not
2.23 --------------------------
His381 ND1 15H, 5Y (mammalian), 1N
Lysl62 NZ Conserved
2.89 ---------Conserved----------
Glu316 OE1 Conserved
Lys162 N Conserved
Glu316 OE1 Conserved
Metl61 N 10L, 9M, 3A (mammalian)
Glu316 OE1 Conserved
Metl61 N 10L, 9M, 3A (mammalian)
2.89
Glu316 OE2 Conserved
Lys390 NZ 2.18K, 3R (mammalian)
2.66
Glu316 OE2 Conserved
Tyrl46 OH 18Y, 3F (mammalian)
Lys390 NZ 18K, 3R (mammalian)
..... ^ ^ ---------^ ----- 2. /1 -- ^ -^-- ^ ^ ^--- --


Arg30 ---- NH1 2.84 conserved
Asn389 OD1 17N, 4D (mammalian)
Arg30 NH2 3.05 Conserved
-------- ArgO ----------'--- NH------ 3.05 ----------Cn v----------
Asn389 OD1 17N, 4D (mammalian)
His192 -- -- not
L sM-- -- 13K, 3R, 3H (mamm.), 2T








Scheme 1-1


Synthetase site
(C-terminal)






Glutaminase site
(N-terminal)


ATP
1


Asp


AMP
t


NH3

S Tunneling

Gin NH3 Glu
1 t t


Synthetase site
(C-terminal)






Glutaminase site
(N-terminal)









CHAPTER 2
A gHMQC-BASED NMR ASSAY FOR INVESTIGATING AMMONIA CHANNELING IN
GLUTAMINE DEPENDENT AMIDOTRANSFERASES

Introduction *

Glutamine-dependent amidotransferases catalyze ammonia transfer from the amide of

glutamine to a variety of acceptors (3, 5, 6, 25). These enzymes first hydrolyze glutamine into

glutamate and ammonia. The ammonia intermediate is then transferred to a second active site

and reacts with other substrates to synthesize different nitrogen containing product. X-ray crystal

structures of many amidotransferases has showed that this family of enzymes has multiple active

sites which are separated from each other and connected with an intramolecular tunnel (2, 15,

26-33). The reactions catalyzed at both active sites are integrated together by intermediate

transfer through this tunnel. It seems that all the glutamine dependent amidotransferases have an

ammonia tunnel, with the glutaminase active site at one end and the synthetase active site at the

other end. Although it has been proposed that ammonia produced by hydrolysis of glutamine is

channeled efficiently into a second active site to complete the overall enzymatic reaction (6, 34-

38), kinetic studies for relatively few glutamine-dependent amidotransferases have provide

evidences to support this hypothesis (39-42). Perhaps the most convincing results for efficient

ammonia channeling has come from 15N NMR studies of carbamoyl phosphate synthetase (CPS)

(39, 43-45), which demonstrated that no exchange happens between 14NH3 produced in the

glutamine domain of CPS and 15NH3 present in bulk solution (39). In addition, pre-steady state

kinetic measurements have shown the rates of ammonia utilization and glutaminase activity in

CPS to be coupled, implying that only a single ammonia molecule is present in the

intramolecular tunnel of the enzyme during catalytic turnover (43). Integrated structural,



SThis work was finished with the helpful discussion of an undergraduate William Beeson.









computational and experimental measurements have also provided support for active site

coupling and intramolecular ammonia transfer in imidazole glycerol phosphate synthase (IGPS)

(40, 46-49). Both CPS and IGPS are Class I amidotransferases (6, 50), however, and their

molecular behavior may not therefore be representative of the evolutionarily unrelated Class II (6,

51) or Class III amidotransferases (52).

Glutamine-dependent asparagine synthetase (ASNS) (EC 6.3.5.4) belongs to the Class II

amidotransferases, which also includes Glucosamine 4-phosphate synthase, Glutamine

phosphoribosylpyrophosphate (PRPP) amidotransferase, and glutamate synthase (5, 6). The two

domains of asparagine synthetase, glutaminase domain and synthetase domain, catalyzes the

hydrolysis of glutamine (R3) and the ATP-dependent synthesis of L-asparagine from L-aspartic

acid respectively (R2) (13), with R1 as the overall enzymatic reaction.

Gln + H20 + ATP + Asp 4 Glu + AMP + PPi+ Asn (R1)
NH3 + ATP + Asp AMP + PPi + Asn (R2)
Gln+H20 Glu + NH3 (R3)

Both the mammalian and bacterial forms of the enzyme can accept free ammonia as an alternate

substrate in vitro (24, 53). With ATP and aspartate, glutamine is hydrolyzed into glutamate and

ammonia at glutaminase site through a thioester enzyme-substrate complex (14) and then the

ammonia intermediate is transported through a 20 A long, intramolecular tunnel to the synthetase

site, which is accepted by P-Aspartyl-AMP, an intermediate that formed at synthetase site using

ATP and Asp, to produce asparagine. In sharp contrast to the behavior of other Class II

amidotransferases for which the independent catalytic sites are tightly coupled (54-56), without

the ammonia acceptor, asparagine synthetase still can hydrolyze glutamine (22, 57). This lack of

active site coupling in ASNS is surprising in light of observations on other amidotransferases,

and raises questions about the structural integrity of the solvent inaccessible, intramolecular









tunnel that is seen in the AS-B crystal structure as the enzyme proceeds through its catalytic

cycle.

We now report the use of a very convenient isotope-edited H NMR-based assay (58) that

we have developed to probe ammonia transfer between the two active sites in glutamine

dependent asparagine synthetase from Escherichia coli (AS-B). This gradient heteronuclear

multiple quantum coherence (gHMQC) method, which should be generally applicable for studies

of other glutamine-dependent amidotransferases, has provided quantitative information on the

extent to which 15N is incorporated into the side chain of asparagine formed in the enzyme-

catalyzed reaction. In addition, although our studies of AS-B have employed end-point

measurements, the gHMQC strategy may be amenable for use in continuous assays of nitrogen

transfer catalyzed by other members of this enzyme family. At least in the case of AS-B, the

results of these gHMQC NMR studies show that (i) high glutamine concentrations do not

suppress ammonia-dependent asparagine formation by this enzyme, and (ii) ammonia in bulk

solution can undergo reaction with a thioester intermediate formed during the glutaminase half-

reaction (14, 59). These observations are consistent with a model in which exogenous ammonia

can access the tunnel in AS-B during glutamine-dependent asparagine synthesis, in contrast to

expectations based on studies of Class I amidotransferases (39, 46, 47).

Material and Methods

Materials

Unless otherwise stated, all chemicals and reagents were purchased from Sigma (St. Louis,

MO), and were of the highest available purity. 1,3-15N2-Uracil, 1N-L-asparagine, and d6_DMSO

were purchased from Sigma-Aldrich (St. Louis, MO). The isotopic incorporation in these

samples was greater than 99%. All experiments employed freshly prepared solutions of

recrystallized L-glutamine (17). Recombinant, wild type AS-B was expressed and purified









following literature procedures (16). Protein concentrations were determined using a modified

Bradford assay (Pierce, Rockford, IL) (60), for which standard curves were constructed with

bovine serum albumin, and corrected as previously reported (55). NMR spectra were recorded on

a Varian INOVA 500 instrument equipped with a 5 mm triple resonance indirect detection probe

(z-axis gradients) operating at 500 and 50 MHz for 1H and 15N, respectively. Chemical shifts are

reported in ppm relative to sodium 2,2-dimethyl-2-silapentane-5-sulfonate (DSS). The co-axial

inner cell (catalog: NE-5-CIC-V) containing the external NMR standards was purchased from

New Era Enterprises (Vineland, NJ).

NMR Measurements

Concentrations of 15N-L-asparagine were determined using a solution of 1,3-15N2-uracil

dissolved in d6_DMSO as a standard, which was placed in a coaxial inner cell inserted into the 5

mm tube containing the assay sample. The amount of 15N-L-asparagine was measured using a

phase-sensitive, 1-D gHMQC pulse sequence (58), as implemented in the Varian VNMR

software package (version 6.1C). NMR spectra were acquired on a Varian INOVA500

spectrometer, at a fixed temperature of 25 C, in 256 transients with a digital resolution of 0.25

Hz/point (25966 points in the FID over a spectral window of 6492 Hz) using a relaxation delay

of 1 s and an acquisition time of 2 s. The total time for acquiring each spectrum was therefore 13

min. The encoding gradient level was 37 Gauss/cm of 2.5 ms duration, and the corresponding

values for the decoding gradient were 18.7 Gauss/cm and 1 ms, with a gradient recovery time

was 0.5 ms. 900 pulse times used in the experiment for 1H and 15N were 9.6 |tsec and 25.8 ts,

respectively, and the 1H-15N coupling constant was set to a value of 87 Hz. The FID was

weighted with a line broadening of 10 Hz and a Gaussian of 0.278 Hz.

Competition Experiment

Reaction mixtures consisted of 100 mM HEPPS buffer (pH 8.0), 60 mM MgC12, 10 mM









ATP, 20 mM aspartate and different concentrations of 15NH4C1 (pH 8.0) or glutamine in total

volume of 2 mL. In experiments where the concentration of 15NH4C1 (pH 8.0) was varied as 2.0

mM, 5.0 mM, 25 mM, 50 mM and 100 mM, L-glutamine was fixed at 20 mM. Alternatively,

when L-glutamine was varied as 0 mM, 2.5 mM, 10 mM, 20 mM and 40 mM, 15NH4C1 was

added at an initial concentration of 100 mM. Reactions were initiated by the addition of AS-B

(31 [tg) and the resulting samples incubated for 10 min at 37 C before being quenched by the

addition of trichloroacetic acid (TCA) (60 tL). After centrifugation for 5 min at 3000 rpm to

remove precipitated protein, the supernatant was adjusted to pH 5 by the addition of 10 M aq.

NaOH and an aliquot of this solution (650 tL) transferred to a 5 mm NMR tube for analysis. At

higher pH values, amide NH exchange precluded the derivation of any quantitative relationship

between peak area and 15N-Asn concentration. The standard samples contained 100 mM HEPPS

buffer (pH 8.0), 60 mM MgC12, 10 mM ATP, 100 mM 15NH4C1 (pH 8.0), 20 mM glutamine and

different concentrations of 15N-L-Asn from 0.25 mM to 5 mM, without adding AS-B.

15N Exchanging Experiment

Reaction mixtures consisted of 100 mM HEPPS buffer (pH 8.0), 60 mM MgC12, 10 mM

ATP and different concentrations of 15NH4C1 (pH 8.0) or glutamine in total volume of 2 mL. In

experiments where the concentration of 15NH4C1 (pH 8.0) was varied as 2.0 mM, 5.0 mM, 25

mM, 50 mM and 100 mM, L-glutamine was fixed at 20 mM. Alternatively, when L-glutamine

was varied as 0 mM, 2.0 mM, 5.0 mM, 10 mM, 20 mM, 40 mM and 80 mM, 15NH4C1 was added

at an initial concentration of 100 mM. Reactions were initiated by the addition of AS-B (62 [tg)

and the resulting samples incubated for 10 min at 37 C before being quenched by the addition of

trichloroacetic acid (TCA) (60 tL). After centrifugation for 5 min at 3000 rpm to remove

precipitated protein, the supernatant was adjusted to pH 5 by the addition of 10 M aq. NaOH and

an aliquot of this solution (650 tL) transferred to a 5 mm NMR tube for analysis. At higher pH









values, amide NH exchange precluded the derivation of any quantitative relationship between

peak area and 1N-Asn concentration. The standard samples contained 100 mM HEPPS buffer

(pH 8.0), 60 mM MgC12, 10 mM ATP, 100 mM 5NH4C1 (pH 8.0), 20 mM glutamine and

different concentrations of 15N-L-Asn from 0.10 mM to 5 mM, without adding AS-B.

HPLC-Based Determination of Total L-Asparagine

In order to obtain an estimate of total L-asparagine in the final reaction mixtures, we

employed an HPLC-based end-point assay (15). Hence, an aliquot of each mixture (40 pL) was

diluted (200 tL final volume) with 400 mM aq. Na2CO3, pH 9, containing 10% DMSO and 30%

dinitrofluorobenzene (DNFB) (as a saturated solution in EtOH). The resulting solutions were

heated at 50 OC for 45 min to permit reaction of DNFB with the amino acids to yield their

dinitrophenyl (DNP) derivatives. Caution: Extreme care should be taken when handling

solutions of 2,4-dinitrofluorobenzene in organic solvents because this reagent is a potent allergen

and will penetrate many types of laboratory gloves (46). Aliquots of each assay mixture (20 tL)

were analyzed by reverse-phase HPLC (RP-HPLC) using a C8i column and a flow-rate of 0.7

mL/min. The DNP-derivatized amino acids were eluted using a step gradient of 40 mM formic

acid buffer, pH 3.6, and CH3CN. In this procedure, the initial concentration of the organic phase

(CH3CN) was 14%, which was maintained over a period of 26 min before the amount of CH3CN

was increased to 80% over a period of 30 s, and elution continued for a further 8 min. Eluted

amino acid DNFB derivatives were monitored at 365 nm and identified by comparison to

authentic standards. Under these conditions, DNP-asparagine exhibited a retention time of

approximately 25 min, and could be quantified on the basis of its peak area. Calibration curves

were constructed using solutions of pure L-asparagine derivatized in the same manner as the

samples.









Kinetic Simulations

Simulations were performed using the GEPASI software package (61, 62).

Results and Discussion

Two mechanisms can be envisaged for nitrogen transfer from the glutaminase to the

synthetase active sites in AS-B (Figure 2-1). In the first, ammonia is directly transferred between

the two active sites through an intramolecular tunnel (2, 15, 26-33). This proposal, which is

hypothesized to occur in all other glutamine-dependent amidotransferases (6, 34-38), is

supported by X-ray crystallographic observations on the bacterial enzyme (15). A second model

can be envisaged, however, in which ammonia is released into bulk solution prior to re-entering

the synthetase site and reacting with the 0-aspartyl-AMP intermediate. We therefore sought a

simple and rapid kinetic assay to distinguish between these possibilities, which might also be

applicable in experiments aimed at developing structure-function relationships for residues

defining the tunnel observed in AS-B. Early work on CTP synthetase (41, 63, 64) and GMP

synthetase (65) aimed at investigating this problem employed the pH-dependence of synthetase

activity when glutamine and ammonia (or alternate substrates such as hydroxylamine) were both

present in solution. We elected to employ an alternate, and somewhat more straightforward,

strategy, however, in which the glutamine-dependent asparagine synthetase reaction was

performed in the presence of exogenous 15NH4C1. In particular, we hoped to employ 15N NMR

spectroscopy to determine the extent of 15N incorporation into asparagine as a function of

glutamine concentration, as previously reported in studies on CPS (19). The low sensitivity of

the 15N nucleus proved to be a significant limitation to our efforts, mandating a substantial

investment of spectrometer time in order to obtain spectra with sufficiently high signal-to-noise

for any quantitative measurements, thereby severely limiting the number of conditions under

which competition studies could be carried out. We therefore investigated the use of a phase-









sensitive ID gHMQC experiment to measure the extent to which 1N was incorporated into

asparagine when AS-B was incubated with aspartate and ATP in the presence of both glutamine

(containing nitrogen isotopes at natural abundance) and 15NH4C1.

NMR Assay

In gHMQC spectra, resonances are only observed for hydrogen nuclei that are (i) attached

to 1N nuclei, and (ii) are not in fast exchange. The chemical shifts for 15N-amide protons are:

15N-Asn (6NHcis = 7.58 (dd) and 6NHtrans = 6.86 (dd), J = 90 Hz); 15N-Gln (6NH1 = 7.53 (dd) and

6NH2 = 6.81 (dd), J = 90 Hz). Hence, this NMR strategy permits a 30-fold increase in sensitivity

over direct acquisition of 15N spectra. In addition, and very importantly for the goals of these

experiments, the time needed to acquire 15N-edited H NMR spectra was approximately three

orders of magnitude faster than for the equivalent measurements using 5N NMR spectroscopy.

We were therefore able to obtain very "clean" 1H NMR spectra for assay mixtures containing

AS-B because the signals for protons on free 15NH3 and 15NH4 were not observed (Figure 2-2).

Accurate integration of signals relative to those from an internal standard (15N2-uracil dissolved

in d4-DMSO) was also possible, and standard curves relating the observed peak area of amide

proton signals to the concentration of 15N-asparagine present in solution under a variety of

conditions were easily obtained (Figure 2-3). All of these results were carefully validated by

independent measurements of asparagine using an HPLC-based assay (23).

Competition Experiments

Having established that the amount of 15N-asparagine formed in the reaction could be

quantitated using gHMQC spectroscopy, we studied whether the incorporation of 15N from

exogenous 15NH4C1 into asparagine could be suppressed by L-glutamine in the synthetase

reaction catalyzed by AS-B. If ASB was incubated with both 14N-glutamine and 15NH4C1 at the

presence of other substrates and cofactor, both 14N-asparagine and 15N-asparagine would be









produced. 14N-asparagine was produced by 14NH3 intermediate that goes through an

intramolecular tunnel. And 15N-asparagine was formed by 15NH3 from free solution. 15NH3 may

compete with 14NH3 from 14N-glutamine as nitrogen donor. Therefore, quantitative analysis of

both 15N-asparagine and 14N-asparagine in the reaction mixture gives information about the

capability of ASB to utilize different nitrogen source (scheme 2-1). If ammonia can not access

the active site with the presence of glutamine, all the asparagine would be 14N product at

saturating concentration of glutamine. In this case, free ammonia is suppressed by glutamine.

The ratio of 14N asparagine to 15N asparagine would be infinite. This ratio will be decreasing

with the extent of suppression becomes less. When there is no suppression at all, the 14NH3 will

compete with 14NH3 freely. If ammonia tunneling is fast and efficient enough, then the ratio of

14N asparagine to 1N asparagine would be determined by the ratio of production rates using

ammonia and glutamine. That is,



[14N Asn] v14 (k,,, /Kmtunnehng)[Gln][E] [Gln] [Gln]
["5N-Asn] v15 (kt /Km bndng )[15NH3][E] [15NH3] [15NH3],

Therefore,

[14N Asn] [G n] [14N Asn] [G In]
oc or =Ax
[15N Asn] [15NH3 i, [15N Asn] [ NH3 ,



Alternatively, if the tunnel is not efficient, the ratio of 14N asparagine to 15N asparagine would

decreases. Most of the asparagine would be labeled form and the ratio of 14N/15N in asparagine

would be close to zero if the 14NH3 intermediate leaks and equilibrates with 15NH3 from bulk

solution, since [15NH3] was much bigger than [14NH3]. Hence, we studied the competition in the

presence of both 14NH3 and 14N-glutamine as nitrogen donors. Under experimental conditions,









the final concentration of asparagine formed was 1-2 mM. The dependence of 1N incorporation

into the product amide on the initial concentration of L-glutamine was determined using the

NMR assay we developed and HPLC assay (Table 1), after correcting the amount of 15N in L-

asparagine to allow for the natural abundance of 15N in L-glutamine. A similar set of experiments

in which L-glutamine was fixed at 20 mM and 15NH4C1 varied over a range of initial

concentrations was also carried out (Table 2-1).

Somewhat unexpectedly, in light of the behavior reported for CPS in a similar competition

experiment (19), we did not observe complete suppression of 15N incorporation at saturating

concentrations of L-glutamine (KM(app): 0.69 mM (53)). Instead, the 14N/15N incorporation ratio

exhibited saturation behavior, reaching a limiting value of 1.2 + 0.2 as the concentration of L-

glutamine was increased (Figure 2-4A). The ammonia-dependence of the 14N/15N incorporation

ratio when the initial concentration of L-glutamine was fixed at 20 mM was also examined and

again showed substantial 15N incorporation even at non-saturating concentrations of 15NH3

(Figure 2-4B). Any calculation of the expected 14N/15N ratio in the side chain of the product by

the simple comparison of the V/K values for the two nitrogen sources is complicated, however,

by the fact that L-asparagine inhibits the glutaminase activity of ASNS with an apparent Ki of

50-60 tM (34, 41) while having no significant impact on ammonia-dependent synthetase activity

at 1 mM concentration (41; also see supporting information). Thus, the rate of ammonia-

dependent L-asparagine synthesis is unaffected by the presence of L-asparagine over the period

of our competition experiment, while that for glutamine-dependent synthetase activity decreases

over time (Figure 2-4C). The effects of this differential inhibition are reflected in the total

amount of L-asparagine formed after a given time under the assay conditions, which depends on

whether one or both nitrogen sources are present. As a result, the total amount of L-asparagine









formed in 10 minutes is greater when both exogenous ammonia and glutamine are present in

solution than when glutamine is employed as the sole nitrogen source (Table 2-1). This also

explains the observation that total L-asparagine is increased when ammonia is added to a solution

containing a saturating amount (20 mM) of L-glutamine because ammonia-dependent activity is

not inhibited by the reaction product binding to the glutaminase active site. Given this

complication in understanding the apparent inability of L-glutamine to prevent 15N incorporation

into asparagine, we sought to model the theoretical 14N/15N incorporation ratio as a function of

either L-glutamine or ammonia concentration by kinetic simulations of the competition

experiment. In keeping with hypotheses developed from previous studies on CPS (19) and IGPS

(20), we assumed that exogenous 15NH3 and L-glutamine could not bind simultaneously to the

E.ATP.Asp ternary complex required for synthetase activity (40, 41). The assignment of rate

constants was accomplished (40) on the basis of (i) direct NMR and HPLC measurements of L-

asparagine production rate under the assay conditions (k2, k4, k7 and kio), and (ii) literature data

on the steady state kinetics of AS-B (k-l, k-3, k-6 and k-9) (40, 53). All second order "on" rate

constants were defined as 108 M-^s^, although this does not account for conformational changes

that might take place on substrate binding. Perhaps more importantly, we assumed that (i) the

presence of L-asparagine in the glutaminase active site of the enzyme does not affect rate

constants associated with ammonia-dependent synthetase activity (k4, k7 and kio), and (ii) the

"off' rate of asparagine from the inhibitory site is not affected by bound ammonia (k11 and k-12).

The only remaining unknown rate constants (k-L and k-12) in this simple model could then be

estimated from the observed KI for L-asparagine in the glutamine-dependent synthetase reaction.

Although this kinetic model exhibits a time-dependent decrease in the rate of nitrogen

incorporation from L-glutamine to give 14N-L-asparagine in the competition assay (Figure 2-4C),









it does not qualitatively reproduce the saturation behavior of the 14N/1N incorporation ratio as

the concentration of L-glutamine is increased in the presence of a fixed amount of 15NH4C1

(Figure 2-4A). The lack of quantitative agreement between the experimental and simulated

14N/15N incorporation ratios likely arises from an over-estimation of the affinity of the AS-B

glutaminase domain for L-asparagine, and we therefore conclude that both nitrogen sources can

be present on the enzyme simultaneously.

15N exchange Experiments

Evidence to support (i) the conclusion from the kinetic simulations, and (ii) the hypothesis

that 15NH3 might be able to access the N-terminal glutaminase site (and presumably the

intramolecular tunnel) when L-glutamine was also present came from the observation of peaks in

the isotope-edited gHMQC spectrum with chemical shift values that were different from those of

the resonances arising from the amide protons of 15N-asparagine (Figure 2-2B). The acquisition

of 1H spectra for the reaction mixture in the absence of isotopic editing, but with suppression of

the water signal, showed that these unexpected peaks were satellites of the broad signal from the

amide protons in 14N-L-glutamine (Figure 2-5). As a result, we concluded that these signals arose

from amide protons bonded to 15N in the side chain of L-glutamine. Control experiments

established that 14N/15N exchange did not occur in the absence of AS-B, and so this reaction was

examined in more detail by incubating the enzyme with 15NH4C1 in the presence of glutamine

and ATP. Under these conditions, the rate of 1N-substituted glutamine formation (i) showed

saturation behavior with respect to glutamine at a fixed concentration (100 mM) of 15NH4C1

(Figure 2-6), and (ii) was linearly dependent on the concentration of exogenous 15NH4C1 (data

not shown).

A kinetic model was constructed to describe the 14N/15N exchange (Figure 2-7A), which

gave the following equation for partitioning of the thioester intermediate at the steady-state to









yield 15N-L-glutamine (see supporting information):

kk2 [E], [Gln] ['NH']
v = (Eqn. 1)
(k, [GIn] + k215NH ] + k3 [H20])

Here v is the rate of 15N-glutamine production via the exchange reaction in the presence of ATP

but the absence of aspartate, and kl, k2 and k3 are microscopic rate constants with estimated

values of 5.1 mM-1s-, 7.8 x 10-3 mM-1s1 and 1.06 x 10-4 mM-ls-, respectively, assuming that

hydrolysis of the thioester intermediate is the rate-limiting step in the glutaminase reaction.

Using these values in our model for the exchange reaction gave excellent agreement between the

observed concentrations of 15N-glutamine and those expected from kinetic simulation at a variety

of NH4C1 concentrations (Figure 2-7B). Having established a firm understanding of this reaction

in the absence of aspartate, we next examined whether the overall rate of 14N/15N exchange was

altered when AS-B was undergoing catalytic turnover to yield asparagine. A similar analysis to

that described above showed that our kinetic model gave excellent agreement with experimental

values (data not shown) when the microscopic rate constants kl, k2 and k3 were assigned values

of 9.6 mM1 s 1.8 x 10-2 mM-1 s and 8.8 x 10-5 mM-ls-, respectively. The similarity of these

rate constants to those determined for 14N/15N exchange in the absence of aspartate suggests that

the molecular mechanism of the reaction is identical under both sets of conditions.

The simplest explanation for 14N/15N exchange is that 15N-substituted glutamine is formed

by reaction of exogenous 15NH3 with a thioester intermediate formed in the N-terminal active site

during the enzyme-catalyzed hydrolysis of glutamine (Scheme 2-2) (14, 59). This is an intriguing

finding because a similar 14N/15N exchange reaction involving the side chain amide of L-

glutamine was not observed in similar competition experiments employing CPS (19), although it

is possible that the amounts of 15N-substituted glutamine formed might have been too small for

detection by 15N NMR spectroscopy. In addition, this phenomenon has not been described









previously, to the best of our knowledge, for any other glutamine-dependent amidotransferase.

The molecular pathway by which 15NH3 gains access to the N-terminal, glutaminase site remains

to be determined. Hence, it is possible that exogenous ammonia can access the N-terminal

domain directly as a consequence of the enzyme adopting a conformation in which the tunnel

linking the two active sites is solvent accessible. Alternatively, ammonia might enter the

intramolecular tunnel linking the two active sites after entering the C-terminal domain. In both

mechanisms, ammonia could displace water molecules in the tunnel prior to glutamine binding

and the adoption of the "closed" conformation observed in the crystal structure of the enzyme.

Conclusions

In summary, we have reported a sensitive, quantitative and reproducible isotope-edited 1H

NMR assay for monitoring the incorporation of 15N into the side chain amide of L-asparagine.

Spectra can be obtained rapidly thereby permitting the examination of a wide variety of

conditions for these steady-state competition experiments. The application of this assay to the

reaction catalyzed by AS-B has shown, somewhat unexpectedly, that high concentrations of

glutamine can not suppress the incorporation of 15N from exogenous 15NH3. This result, in

combination with our observation that exogenous ammonia can trap the thioester intermediate in

the glutaminase reaction, is consistent with a model in which the putative ammonia channel

linking the N- and C-terminal active sites can be accessed by 15NH3 molecules at some point

during catalytic turnover. This behavior seems to contrast with previous findings concerning

ammonia channeling in Class I amidotransferases, such as CPS (19) and IGPS (20), which

appear to have been optimized during evolution for (i) high nitrogen transfer efficiency (19), and

(ii) tight kinetic coupling of glutaminase and synthetase activities (20, 23). In part, this may be a

consequence of the ability of asparagine to inhibit non-productive glutaminase activity, which

might have relaxed any evolutionary pressure to construct intra-domain interactions needed for









efficient kinetic coupling of catalysis in the two active sites of asparagine synthetase.

Any interpretation of the inability of saturating glutamine to suppress ammonia-dependent

synthetase activity in terms of molecular structure is complicated by the observed solvent

inaccessibility of the intramolecular tunnel that is observed in the high-resolution crystal

structure of AS-B (12). We have shown previously, however, that coupling of the glutaminase

and synthetase activities of the enzyme appears to break down as the glutamine concentration is

increased (40), perhaps because of a conformational change that permits the release of ammonia

(14NH3) from one or both of the glutaminase sites. It is therefore possible that molecules of

exogenous ammonia (15NH3) might be able to gain access to the synthetase site when the enzyme

adopts this conformation, with the result that high concentrations of L-glutamine fail to suppress

1sN incorporation to the extent seen for other amidotransferases.











Free solution pathway
(Accumulation)

49
4 I I
I I
*l
I I
*I
/
L


- -20A


Two possible path ai\\ for the ammonia transfer to reach synthetase site. One is
intramolecular pathway, in which ammonia intermediate go through an intramolecular
tunnel after being released from glutaminase site, without exposes to solution. The second
pathway is free solution pathway. The ammonia is released to the free solution first, and
then binds to synthetase site. In the latter pathway, the concentration of ammonia in the
solution is very low at beginning, increases with time, and finally reaches maximum. This
would produce a lag in the plot of synthetase activity versus time, which can be ruled out
by kinetic studies after separate active sites are confirmed.


Figure 2-1.










(A)













12 11 10 9 8 7 6 5 4 3 2 1 ppm
Chemical shift 6
(B)
N N N N





Q Q Q Q





7.9 7.8 7.7 7.6 7.5 7.4 7.3 7.2 7.1 7.0 6.9 6.8 6.7 6.6 ppm

Chemical shift
Figure 2-2. gHMQC 'H-NMR spectrum. (A) "Clean" 'H-NMR spectrum obtained by gHMQC. Only
5sN-amide protons were showed in the spectrum. For the samples used in this work, these
peaks (from left to right) stands for "1N2-uracil (6 > 11.00 ppm), "N-amides (glutamine or
asparagine, 6: 6.50 7.80 ppm) and solvent (H20, 6 = 4.70 ppm) respectively. The peak
areas of "N-glutamine or 15N-asparagine are relative to the total peak area of "Nz-uracil in
d4-DMSO, which was set to 100. (B) The "N-amide peaks. "N-Asn (6NHcs = 7.58 ppm (dd)
and NHtrans = 6.86 ppm (dd), J = 90 Hz); "N-Gln (6NH = 7.53 ppm (dd) and 6NH2 = 6.81
ppm (dd), J = 90 Hz)





























1 2 3 4 5


115N-Asnl mM


250 -


200 -


150 -
-


S100 -
-

50 -


0-


2 3
['IN-Asnl mM


Standard curves. (A) Standard curve for exchange experiment. The concentrations of 15N-
Asn added were 0.05 5.0 mM. R2 = 0.9986. Detection limit < 50 gM. (B) Standard curve
for competition experiment. The concentrations of 15N-Asn was varied from 0.25 5.0 mM.
R2= 0.9991. Detection limit < 50 gM.


Figure 2-3.











H3N+ O

Scheme 2-1 0 SNH2
mNH3 O0
sNH3 N-Asn


ATP PPi H2
H3N 0 H3N+ 0 IN ,
o-Q- y ---O- -O U N
0 0 I"
Asp
p-Aspartyl-AMP OH OH


NINH, H3N 0

SNH2
NH3 Tunneling
H3N+ O
H A O "N-Asn


Glu
H3N
-0 0

O NH2
Gin


































U
U
U
U


20 40 60


L-Glutamine (mM)


20 40 60 80 100


15NH4C1 (mM)


o



20 40 60 8O o0 120
"NH4C1 (mM)












1 -

0.9

0 0.8


C

S 0.5

0.4

U 0.3

0.2

0.1

0
0 100 200 300 400 500 600

Time (s)

Figure 2-4. 14N/1N incorporation ratios in L-asparagine formed by the AS-B synthetase reaction under
competition assay conditions. (A) Comparison of the experimentally observed (n) and
simulated (m) '4N/'N incorporation ratio in 100 mM HEPPS buffer, pH 8, containing 10
mM MgATP, 20 mM aspartate and 100 mM 5NH4C1. (B) Comparison of the
experimentally observed (n) and simulated (m) '4N/5N incorporation ratio in 100 mM
HEPPS buffer, pH 8, containing 10 mM MgATP, 20 mM aspartate and 20 mM L-
glutamine. The inset shows an expanded view of the plot for the experimental 14N/5N
ratios. (C) Kinetic simulation of the time-dependent formation of total L-asparagine (blue),
14N-L-asparagine (red) and "5N-L-asparagine (green) showing the impact of differential
inhibition of the two synthetase reactions by L-asparagine.


























7.8 7.6 7.4 7.2 7.0 6.8 6.6 ppm

Normal 'H-NMR spectrum that shows amide protons from both "N-glutamine (product,
sharp peak) and 14N-glutamine (reactant, sharp peak). The spectrum was acquired after
water suppression at 25 C. Acquisition time was 5s. Number of transients was 128.


Figure 2-5.













" 0.8



0.6


0
0.4



0.2
Z-
tt


0 20 40 60 80


L-Glutamine (mM)


Glutamine-dependence of the exchange of 1NH3 into L-glutamine under exchange assay
conditions. Exchange reactions were performed in 100 mM HEPPS buffer, pH 8,
containing 60 mM MgC12, 10 mM ATP and 100 mM 1NH4C1 (2 mL total volume), and
each data point represents an average of duplicate experiments.


Figure 2-6.













5100 M-1s-1
E + 14GIn
7.8 x 10-3 [14NH s-1


E + Glu
S 1.06 x 10-4 [H20] s-1

Sx 10-3 [15NH] s-1

E + 15GIn


I I
40 60
L-Glutamine (mM)


Figure 2-7. Exchange of 1NH3 into L-glutamine during the glutaminase reaction catalyzed by AS-B.
(A) Kinetic model employed in the simulations of this exchange process. (B) Comparison
of the experimentally observed (n) and simulated (m) rate of 1NH3 exchange into L-
glutamine when AS-B is incubated in 100 mM HEPPS buffer, pH 8, with L-glutamine and
100 mM "1NH4C1 in the absence of ATP and aspartate.











H3N' NH2

O 0
S- H
Enzyme
H. H
H.N O

Enzyme


H-n A
ENO0

Enzyme


H -O+NH3

0 0-
H N
-NO

Enzyme


NH3 (solution)


NH3 (tunnel)


Scheme 2-2.


0 90
O


H3NI ,NH2
O0 O- 0


Enzyme

Hypothetical mechanism for formation of a thioester intermediate during the ASNS-
catalyzed hydrolysis of L-glutamine. Note that the N-terminal amino group is thought to
function as the general acid/base in the reaction (17, 55), although direct evidence for this
proposal remains to be obtained in the case of ASNS. The thioester intermediate (TE),
which subsequently reacts with water to give glutamate, is drawn within a "shadowed" box.










Table 2-1 AS-B catalyzed incorporation of 1N into L-asparagine in the steady-state competition assays.
Gin 15 15b 15 bc Total Asn 14N-Asn/"N-
(mM) 5NH4C1 (mM) N-Gln (mM) N-Asn (mM)M)d Asn
(mM) (mM)d Asn

0 100 NDe 0.83 0.03 0.83 0.03 0
2.5 100 NDe 0.71 + 0.02 1.22 + 0.06 0.65 0.08
10 100 0.18 0.01 0.66 0.02 1.30 + 0.06 1.0 + 0.1
20 100 0.26 + 0.01 0.67 + 0.01 1.4 + 0.1 1.1 + 0.2
40 100 0.34 0.01 0.63 0.03 1.4 0.1 1.2 0.2
20 0 NDe NDe 0.76 0.04 NAe
20 25 0.12 0.01 0.29 0.01 0.94 0.03 2.2 0.1
20 50 0.16 + 0.02 0.44 + 0.02 1.09 + 0.05 1.5 + 0.1
20 75 0.19 + 0.01 0.62 + 0.07 1.19 + 0.07 1.1 + 0.1
20 100 0.26 0.01 0.67 0.01 1.4 0.1 1.1 0.2


a All reactions contained 10 mM MgATP, 20 mM aspartate, 60 mM MgC12 and 490 nM AS-B in 100 mM HEPPS
buffer, pH 8 (2 mL total volume).
b As determined by gHMQC NMR spectroscopy. Errors are estimated on the basis of measurements employing
known concentrations of authentic 5N-Asn.
This value is corrected for the amount of 15N-Asn that would be formed from 15N-Gln at natural abundance, but
the contribution of 14N-Asn formed from any 14NH3 released from the enzyme, as a result of the glutaminase
activity of AS-B, is assumed to be negligible.
d Mean value and standard deviation computed from two separate determinations on duplicate samples using
reverse-phase HPLC.
e ND not detected; NA not applicable.









CHAPTER 3
CHARACTERIZATION OF A TUNNEL MUTANT

Introduction

The crystal structure of C1A mutant of Escherichia coli asparagine synthetase B showed

that the two active sites are separated by a 20 A distance. The biochemical studies on ASB, with

structural evidence, support the presence of an intramolecular ammonia tunnel in this enzyme.

To investigate the channeling mechanism of ammonia in asparagine synthetase under the

structural context, the tunnel residues need to be identified and their key roles in ammonia

tunneling need to be elucidated. The tunnel residues were determined by computing the

molecular surface of the protein, with a probing radius at 1.4 A. Do these residues form the real

tunnel that ammonia intermediate goes through? Futhermore, since most of them are conserved

in different species including bacterial, fungus, plants, insect and mammalian, which indicates

they are important to either the structure or the function of the tunnel. Then what kind of roles do

they play in ammonia tunneling? To provide some insight to these questions, I performed

biochemical studies on ASB. The strategy here is to characterize tunnel mutants in which

ammonia transfer is affected by either blockage or leakage of the tunnel. The hypothesis are: if

the tunnel is blocked or perforated by mutations without affecting the active sites, glutaminase

activity and ammonia dependent synthetase activity of the mutants will not be significantly

affected, compared to that of wild type, while its glutamine dependent synthetase activity will

decrease.

The putative tunnel is arbitrarily divided into three parts: the mouth, the body and the

bottom. The residues in the mouth and the bottom part are within the active sites. Therefore, only

those residues in the body part were considered. Among these residues, Ala388 is one of the

conserved residues. Its backbone forms the narrow part of the tunnel, with other residues. And its









side chain points into the tunnel (Figure 3-1). Since the side chain of alanine is a methyl group,

mutation of this residue to another one with bigger side chain such as Leu may block the tunnel.

Based on this hypothesis, the A388L mutant was prepared and characterized.

Material and Methods

Materials

The experiments were performed as described in Chapter 2 Materials and Methods part.

The primers of A388L mutant were ordered from idtdna.com. QuikChange Site-Directed

Mutagenesis Kit for PCR were from Stratagene. PfuTurboTM DNA polymerase, 10x reaction

buffer, dNTP mix and XL1-blue supercompetent cells were from the mutagenesis kit. Dpn I

restriction enzyme was from New England Biolab. Wizard Plus Midipreps DNA Purification

System for plasmid purification was from Promega.

Unless otherwise stated, all chemicals and reagents were purchased from Sigma (St. Louis,

MO), and were of the highest available purity. 1,3-15N2-Uracil, 1N-L-asparagine, and d6-DMSO

were purchased from Sigma-Aldrich (St. Louis, MO). The isotopic incorporation in these

samples was greater than 99%. All experiments employed freshly prepared solutions of

recrystallized L-glutamine (17). NMR spectra were recorded on a Varian INOVA 500 instrument

equipped with a 5 mm triple resonance indirect detection probe (z-axis gradients) operating at

500 and 50 MHz for 1H and 15N, respectively. Chemical shifts are reported in ppm relative to

sodium 2,2-dimethyl-2-silapentane-5-sulfonate (DSS). The co-axial inner cell (catalog: NE-5-

CIC-V) containing the external NMR standards was purchased from New Era Enterprises

(Vineland, NJ).

Primers were designed and analyzed using Gene Runner ver3.05 (Hastings Software). The

sequences of A388L primers are 5'-GAC TGC GCG CGT CTG AAC AAA GCG ATG TC-3'

and 5'-GA CAT CGC TTT GTT CAG ACG CGC GCA GTC-3 for sense and antisense primer









respectively. The plasmid for mutant expression was prepared by PCR using a pET-21c(+)

plasmid with Escherichia coli asnB gene inserted into its Nde I restriction site as the DNA

template. The template plasmid contains total 7103 bps (5441 bps for empty plasmid plus 1662

bps for asnB gene) with a marker for ampicillin resistance. QuikChange Site-Directed

Mutagenesis Kit was used for PCR and cloning. The plasmid with asnB gene mutant was

purified using Wizard Plus Midipreps DNA Purification System. Recombinant, A388L ASB

mutant was expressed and purified following literature procedures (16). Protein concentrations

were determined using a modified Bradford assay (Pierce, Rockford, IL) (60), for which standard

curves were constructed with bovine serum albumin, and corrected as previously reported (55).

Competition Experiment

Reaction mixtures consisted of 100 mM HEPPS buffer (pH 8.0), 60 mM MgC12, 10 mM

ATP, 20 mM aspartate and different concentrations of 15NH4C1 (pH 8.0) or glutamine in total

volume of 2 mL. In experiments where the concentration of 15NH4C1 (pH 8.0) was varied as 10

mM, 25 mM, 50 mM, 75 mM and 100 mM, L-glutamine was fixed at 20 mM. Alternatively,

when L-glutamine was varied as 0 mM, 2.5 mM, 5mM, 10 mM, 20 mM and 40 mM, 15NH4C1

was added at an initial concentration of 100 mM. Reactions were initiated by the addition of

A388L ASB mutant (30 tg) and the resulting samples incubated for 10 min at 37 C before

being quenched by the addition of trichloroacetic acid (TCA) (60 tL). After centrifugation for 5

min at 3000 rpm to remove precipitated protein, the supernatant was adjusted to pH 5 by the

addition of 10 M aq. NaOH and an aliquot of this solution (650 tL) transferred to a 5 mm NMR

tube for analysis. At higher pH values, amide NH exchange precluded the derivation of any

quantitative relationship between peak area and 15N-Asn concentration. The standard samples

contained 100 mM HEPPS buffer (pH 8.0), 60 mM MgC12, 10 mM ATP, 100 mM 15NH4C1 (pH









8.0), 20 mM glutamine and different concentrations of 15N-L-Asn from 0.05 mM to 2.5 mM,

without adding ASNS.

NMR Measurements and HPLC assay

See Chapter 2 Materials and Methods part.

Results and Discussion

After the A388L ASB mutant was prepared, the kinetic parameters were first determined

using enzyme assays (Beeson, unpublished data). Since the pyrophosphate assay, which is used

for determination of synthetase activity of the enzyme, gave a low kcat value due to a possible

pyrophosphatase contamination, the kcat values for both glutamine dependent and ammonia

dependent synthetase activity were reexamined to fit NMR results. As expected, the specificity

of A388L mutant to glutamine based on its glutaminase activity changed by less than 2 fold

compared to the results for wild type ASB. This change resulted mainly from the decrease of Km

values. Its specificity for ammonia based on the ammonia dependent synthetase activity had little

change. Both kcat and Km decreased a little. For its glutamine dependent synthetase activity, the

specificity for glutamine dropped by 8 fold, with kcat decreased by 5 fold (Table 3-1).

When both 15NH4C1 and glutamine present in reaction mixture, the 14N-asparagine/15N-

asparagine ratio for A388L mutant was less than that of WT ASB, with either varied glutamine

from 2.5 mM to 40 mM (Figure 3-2A) or varied ammonia from 50 mM to 100 mM (Figure 3-

2B). In general, these results are consistent with the result above that glutamine dependent

synthetase activity decreased significantly, with no significant change of its ammonia dependent

synthetase activity. However, the 14N-asparagine/5N-asparagine production ratio catalyzed by

the A388L mutant is not proportional to that catalyzed by the WT enzyme. With 15NH4+ at 100

mM, the 14N-asparagine/15N-asparagine ratio decreased by about 10 folds when the concentration

of glutamine was below 10 mM, while with higher glutamine, it decreased by about 4 folds.









When varying ammonium, this ratio didn't change proportionally either. Especially at low

concentration of ammonium, its value was close to that of WT ASB. These results may be

caused by the overall effect of 1) the asparagine inhibition of glutaminase activity; 2) the

suppression of the ammonia utilization by glutamine, and 3) glutamine inhibition.

The production of 14N-asparagine by A388L mutant decreased significantly compared to

WT ASB, while the production of 1N-asparagine had no big change, when either varying

glutamine or varying ammonium (Figure 3-3, 3-4). Furthermore, the production of 15N-glutamine

increased slightly for the mutant (Figure 3-5), which is consistent with the increased specificity

of glutamine for its glutaminase activity. The 14N-asparagine/15N-asparagine ratio, and the

production of 14N-asparagine, 15N-asparagine and 15N-glutamine, as well as the kinetic results

determined by enzyme assays, support that the mutation of Ala388 to Leu may block the tunnel.

Ala388 is in the narrowest part of the tunnel. It is not only at the interface of the two domains,

but also one of the non-polar residues close to the glutaminase sites that consists of mainly polar

residues (Figure 3-6). Its side chain, a methyl group, is not likely to involve in interactions with

other residues, as aromatic residues or charged residues usually do. However, mutation of this

alanine to leucine resulted in the possible blockage of the tunnel, which may suggest that this

part of the tunnel may act as the gate of the tunnel.

Conclusions

Based on the tunnel residues determined by computational methods, a tunnel mutant was

prepared, in which a conserved alanine residue was mutated to leucine. The glutaminase activity

and ammonia dependent synthetase activity of this A388L mutant are similar to that of WT ASB,

while its glutamine dependent synthetase activity dropped by about 8 fold. The kinetic studies of

this mutant experimentally support that the identified tunnel residues by computational methods

are those forming the intramolecular ammonia tunnel. Furthermore, the NMR studies on this









mutant, together with the kinetic results, suggest that the mutation may result in the blockage of

the tunnel.





























One of the conserved tunnel residues, Ala388. The backbone of A388 forms the narrow
part of the tunnel with several other tunnel residues. Its side chain points into the tunnel
close to the glutaminase site.


Figure 3-1.











1.4

1.2 -

1 -

0.8

0.6 -

0.4

0.2 -

0 4


U U


[Gin] mM


DASB
SA388L


40 60


100


120


[NH4] mM


Competition results for A388L ASB mutant, as well as WT ASB. Filled square: A388L;
blank square: WT ASB. In general, the 14N-asparagine/"N-asparagine ratio decreased due
to the mutation of alanine to leucine. A) varying glutamine; B) varying ammonium.


o ASB
m A388L


4-

3.5

3-

2.5

2-

1.5

1 -

0.5

0


Figure 3-2.









0.85

0.8


0.75 -

0.7

0.65

0.6 -

0.55 -

0.5


0.9
0.8
0.7
0.6-
0.5
0.4
0.3
0.2
0.1
0


SASB
SA388L


I ~


[Gin] mM


OASB
SA388L


[Gin] mM


Figure 3-3. Production of asparagine catalyzed by A388L ASB mutant and WT ASB with varied
glutamine. Filled square: A388L; blank square: WT ASB. In general, the 14N-asparagine
decreased significantly due to the mutation of alanine to leucine. A) production of 1N-
asparagine; B) production of 14N-asparagine.


i





























60

[NH4*] mM


[NH4] mM


Figure 3-4. Production of asparagine catalyzed by A388L ASB mutant and WT ASB with varied
ammonium. Filled square: A388L; blank square: WT ASB. In general, the 14N-asparagine
decreased significantly due to the mutation of alanine to leucine. A) production of 1N-
asparagine; B) production of 14N-asparagine.


o ASB
* A388L


0.7 -

0.6-

0.5-

0.4-

0.3-

0.2 -


0.1

0


100


120


0.9

0.8 -
-
0.7 -

0.6

0.5

0.4


SASB
* A388L


0.3


0.2 -

0.1 -

0


100


120


Ir





.











0.35

0.3 -

0.25 -

E 0.2 m
SOASB
0.15 A388L
10
0.1

0.05

0 i -- i
0 10 20 30 40 50

[Gin] mM



0.3
o ASB
0.25 A388L


0.2 M


= 0.15


0.1 -


0.05 -



0 20 40 60 80 100 120

[NH4+] mM


Figure 3-5. The nitrogen exchange catalyzed by A388L mutant and WT ASB. Filled square: A388L;
blank square: WT ASB. In general, the production of 1N-glutamine increased due to the
mutation of alanine to leucine. A) varying glutamine; B) varying ammonium.

































The polarity of the tunnel residues. Yellow: non-polar residues; green: polar residues; blue:
basic residues; red: acidic residues. Both AMP (bottom) and glutamine (top) showed in
balls.


Figure 3-6.











Table 3-1. The kinetic parameters for glutaminase activity, ammonia dependent synthetase activity and
glutamine dependent synthetase activity of both wild type ASB and A388L mutant


WT ASB A388L mutant
S1 kcat/Km (M kcat/Km (M
kct (s-) Km (mM) ls ) kt (s-1) Km (mM) ls )

Glutaminase activity 3.38 1.67 2000 3.12 0.87 3600

Ammonia dependent 3.19 17 190 2.34 15 160
synthetase activity
Glutaine dependent 2.94 0.69 4300 0.57 1.06 540
synthetase activity









CHAPTER 4
ASPARAGINE INHIBITION TO ASPARAGINE SYNTHETASES AND ITS EFFECT ON
THE QUATERNARY STRUCTURE OF THE ENZYMES

Introduction *

Glutamine-dependent (EC 6.3.5.4) asparagine synthetase belongs to amidotransferase

superfamily(5, 6), and was found in all kinds of species including archaeon, bacteria, fungi,

plant, and animals (13). Based on the structure and function of its glutaminase domain,

asparagine synthetase was classified as class II amidotransferase, which consists of three other

amidotransferases, glutamine PRPP amidotransferase, glucosamine 4-phosphate synthase, and

glutamate synthase. This enzyme has two domains, glutaminase domain and synthetase domain,

which are connected by a single loop (15). It catalyzes ATP-dependent synthesis of L-asparagine

from L-aspartic acid using either glutamine or ammonium as nitrogen source. At glutaminase

site, the amide nitrogen of glutamine is hydrolyzed into ammonia intermediate through a

thioester enzyme-substrate complex, with glutamate as one of the products (14). Ammonia is

then transported through an intramolecular tunnel to synthetase site and reacts with P-Aspartyl-

AMP, an intermediate that formed by ATP and Asp at synthetase site, to produce asparagine.

Alternatively, this enzyme can also use free ammonia as nitrogen donor. Unlike other class II

amidotransferases (54, 66-68), the hydrolysis of glutamine is not strictly coupled with the

synthesis of final product in vitro (17). Without ATP and Asp, asparagine synthetase can

consume glutamine and release toxic ammonia. Furthermore, the glutaminase activity of

bacterial enzyme is even stimulated by the presence of ATP without producing asparagine (17).

These observations raise the interest in the regulation of glutaminase activity of asparagine



SLi designed the inhibition experiments. Inhibition to ASB part was performed by Li and an undergraduate, William
Beeson. Inhibition to hASNS was performed by William Beeson himself. Li and Beeson did the kinetic analysis. Li
did the simulation. Li designed and performed the SEC experiments and analyed the results.









synthetase in vivo to prevent the waste of glutamine and the releasing of free ammonia.

Kinetic studies of asparagine synthetase have been extensively performed in the last 20

years, especially for Escherichia coli enzyme (ASB) (2, 3, 7, 8, 11, 22, 24, 35, 57, 69-72). The

glutaminase activity of asparagine synthetase can be affected by chloride ions (72), ATP or

asparagine (22, 24). With a KI value of about 100 riM, asparagine may play an important role in

preventing unnecessary hydrolysis of glutamine under physiological conditions. Therefore, we

investigated the possible mechanisms for asparagine inhibition to the glutaminase activity of

Escherichia coli ASB and hASNS from both structural and kinetic aspects. Our results showed

that asparagine inhibited the hydrolysis of glutamine by ASB and hASNS through different

mechanisms and had different effect on the quaternary structure of these two enzymes.

Materials and Methods

Materials

Unless otherwise stated, all chemicals and reagents were purchased from Sigma (St. Louis,

MO). L-Glutamine was recrystallized prior to use as previously described (53). Wild-type

Escherichia coli ASB and recombinant, C-terminally tagged human AS were expressed in and

purified from a baculo virus/insect cell expression system as previously described (16, 57).

Enzyme Assays

Kinetic constants were determined by incubating purified, recombinant asparagine

synthetase in reaction mixtures in which one substrate was varied and all others were at

saturating concentrations. Asparagine synthetase was added to 200 [tL reaction mixtures

containing 100 mM HEPPS, pH 8.0, 100 mM NaC1, 8mM MgC12, and varying amounts of L-

glutamine (0-50 mM). The reaction mixture was incubated at 37 C for 10 minutes and then

quenched with 30 [tL of 20% w/v trichloroacetic acid. Each quenched reaction mixture was

added to a premixed 770 uL solution containing 300 mM glycine, 250 mM hydrazine, pH 9.0,









1.5 mM NAD+, and 1 mM ADP. Glutamate production was assayed by adding 2 [L of L-

glutamic acid dehydrogenase to the mixture and monitoring absorption spectrophotometrically at

340 nm for 60 minutes. Measurements at specific substrate concentrations were performed in

duplicate or triplicate, and the initial rate data was analyzed by curve-fitting using computer-

based methods to determine the Vmax and KM constants.

Asparagine Inhibition Studies

Product inhibition studies were carried out on the wild-type E. coli ASB (3.1 [g) and

recombinant, C-terminally tagged hAS (3.0 [tg) using the enzyme assay described above, but

with a fixed amount of asparagine (0-0.5 mM for ASB and 0-1.0 mM for hASNS) added to each

run. The apparent KM and Vmax values were analyzed by curve-fitting using standard computer-

based methods.

Size Exclusion Chromatography to Determine the Quaternary Structure of Asparagine
Synthestases.

Gel filtration chromatography was performed on Rainin Dynamax HPLC system using a

Phenomenex BIOSEP SEC-S2000 column (300 x 7.8 mm with 75 x 7.8 mm guard column,

particle size: 5 am; pore size: 145 A) with detection wavelength at 280nm and flow rate at 1.0

mL/min. The molecular weight is calculated based on a standard curve of a series of marker

proteins with different molecular weights purchased from Sigma. Unless stated otherwise, stock

enzymes for Escherichia coli asparagine synthetase B (ASB) or human asparagine synthetase

were directly injected for analysis with a volume of 5 aL.

The Effect of Asparagine on the Quaternary Structure of ASB and hAS

The enzyme was analyzed as described above using a mobile phase that contained 50mM

Tris-H2SO4 buffer at pH 7.0 with different concentrations of asparagine (0-5.0 mM). All sample

points have at least two repeats.









SEC Studies of ASB with Different Amount of ASB

The enzyme was analyzed by SEC using a mobile phase that contained 50mM Tris-H2SO4

buffer at pH 7.0 with different injection volume of 3.1 mg/mL stock solution (2.5-40 PL).

Results

Product Inhibition

The glutaminase activity of ASB was studied with varying amounts of asparagine present

(0.025 0.5 mM). The apparent KM for glutaminase reaction increased with increasing

asparagine concentration. Figure 4-1A shows a replot of apparent KM respect to glutamine vs

concentration of asparagine. The increase of KM with increasing concentration of inhibitor results

from the binding of asparagine to the free enzyme. However, the regression line is not linear but

more like polynomial. The presence of asparagine in the reaction mixture affected not only the

apparent KM value, but also the kcat value of the glutaminase reaction, which decreased as the

concentration of asparagine was increased (Figure 4-1B). This result provides the evidence that

asparagine binds to enzyme-substrate complex. Therefore, asparagine is a mixed inhibitor to

ASB with respect to glutamine. The glutaminase activity of human enzyme was also studied with

varying amounts of asparagine present (0.05 1.0 mM). The inhibition results showed clearly

that asparagine is a competitive inhibitor to glutamine. The maximum glutaminase activity

stayed constant with a value of 3.6 s-1 in term of monomer. And the apparent KM versus

concentration of asparagine is a straight line with determined inhibition constant as 0.11 0.2

mM (Figure 4-1C).

Asparagine Affects the Quaternary Structure of ASB, But Not Human Enzyme

The nonlinear relationship of apparent KM for ASB vs concentration of asparagine

indicates more than one binding site for asparagine. In the crystal structure of ASB C1A mutant

(PDB code: 1CT9), the enzyme exists as dimer form with glutamine bound to it (15). Each









subunit has two separated active sites, glutaminase site and synthetase sites. The distance

between the two sites is about 20 A. For better understanding of the inhibition mechanisms to

ASB by asparagine from the structural point of view, we studied the quaternary structure of the

enzyme by SEC with or without the presence of asparagine in the mobile phase.

It was reported that the activity of asparagine synthetases was affected by chloride ion and

pH (72). There was no significant difference for the SEC results of asparagine synthetases using

either Tris-HCl or Tris-H2S04 at pH 7.0, or within the pH range of 7.0-7.9 (unpublished data).

Here we choose Tris-H2S04 at pH 7.0 as the mobile phase. The molecular weight (MW) was

calculated using a standard curve (Figure 4-2). Using 50 mM Tris-H2S04 buffer at pH 7.0, two

peaks were observed in the size exclusion chromatogram of ASB, with retention times of 9.15

and 9.51 minutes respectively. The calculated MWs for these two peaks are 120kD and 84kD

respectively (Figure 4-3A). After adding asparagine in the mobile phase, the retention time of the

higher MW ASB peak changed, which indicates asparagine affects the quaternary structure of

ASB. With the presence of asparagine in the mobile phase, two peaks with RT at 8.79 and 9.49

minutes were observed. The calculated MWs for the two peaks are 171kD and 85kD

respectively, with higher MW for the peak with shorter retention time (Figure 4-3A). Under

experimental conditions, the retention times of these two peaks were independent of the

concentration of asparagine added, with the lowest concentration at 0.1 mM (Table 4-1). The

true MW of one subunit of ASB is about 62.5kD based on its gene sequence. Therefore, the 9.50

minutes peak (calculated MW: -85 kD) with or without asparagine is very likely the monomer

form of ASB, with the 8.79 minutes peak (calculated MW: 171 kD) the dimer form (Table 4-2).

And the intermediate peak at 9.15 minutes (calculated MW: 120 kD) without the presence of

asparagine may result from quick exchanges between the forms. That is, equilibrium exists









between the monomer and dimer form. This conclusion was supported by the SEC results using

different amount of enzyme without asparagine in mobile phase (Figure 4-4).

Besides the retention time, asparagine also affects the individual peak area and height.

While the total area of the monomer and dimer peaks had almost no change with different

concentrations of asparagine in the mobile phase, the dimer peak area and height increased with

increasing asparagine, which supports that asparagine thermodynamically stabilizes the dimer

form. The plot of percentage of dimer peak area in total ASB peak area vs. concentration of

asparagine behaves like a Michaelis-Menten curve (Figure 4-5). At low concentrations of

asparagine, the peak area of dimer ASB and the peak height ratio increase rapidly with the

increasing concentration of asparagine. At high concentrations of asparagine, the curve gradually

levels off and indicates the saturating status. The estimated binding constant of asparagine is

about 0.10 mM. In our results, a monomer peak was always observed. In other words, ASB did

not completely converted to dimer at high concentration of asparagine (5 mM). One possible

reason is that partial protein has no chance to collide with each other and form dimer during

elution. Regardless of this, we can conclude that with asparagine present in the mobile phase,

both dimer and monomer peak showed in the chromatogram and the dimer is stabilized by

asparagine binding at the experimental conditions.

Under the same conditions, the quaternary structure of hASNS was studied without

asparagine or with 5 mM asparagine in the mobile phase. Most of enzyme exists as dimer form

with retention time at 8.75 minutes. With 5 mM asparagine, no detectable changes were

observed for the retention time and the peak area (Figure 4-3B).

The Equilibrium between the Dimer and Monomer Form of ASB.

When the mobile phase contains no asparagine, the SEC chromatogram showed one









monomer peak and one intermediate peak, while no dimer peak. Based on these results, we came

up the idea that the ASB monomer and dimer form equilibrium at experimental conditions.

However, this peak may results from the overlapping of the two forms of enzyme peaks. To

confirm our hypothesis, we performed SEC studies using different amount of the enzyme

without the presence of asparagine in the mobile phase.

The ratio of both peak area and peak height for the two peaks (intermediate/monomer)

increased with more enzyme injected. Furthermore, the retention time of intermediate peak is

closer to that of dimer peak. All these results support that more dimer formed at higher

concentration of enzyme and the two forms of enzyme are in a dynamic equilibrium (Figure 4-4).

Construction of Asparagine Inhibition Model

The inhibition model for ASB (scheme 4-1A) was constructed based on the following

considerations: 1. The dimer form is the active form. The active form of ASB was seldom

addressed in previous studies. To date, the only 3D structural information of ASB was reported

by Larsen et al(15). In their studies, the C1A mutant of ASB is a homogenous dimer with one

glutamine bound to each glutaminase site. Since the mutated residue is at the end of N-terminal

domain and plays catalytic role rather than structural role, we can consider this structure is close

to that of the wild type enzyme. Therefore, it is reasonable to assume that the dimer is the active

form of the enzyme. Furthermore, this hypothesis was supported by our result that glutamine also

stabilizes the dimer form of ASB like asparagine (unpublished data). Because the dimer form of

ASB was stabilized at low concentration of asparagine and its substrate glutamine has the similar

effect on its quaternary structure, the amount of dimer form can be considered constant at most

of the experimental conditions with both asparagine and glutamine present. Therefore, although

our results showed equilibrium exists between the two forms of enzyme, we didn't include the









equilibrium part in the model for the purpose of simplicity. 2. One subunit of asparagine

synthetase has two possible asparagine binding sites, glutaminase active site and synthetase

active site because the space-filled model of the crystal structure of C1A mutant showed no third

cavities that may bind the third asparagine molecule. Hence, the dimer form of asparagine has

four possible asparagine binding sites. 3. The first asparagine binding site to the free enzyme is a

glutaminase site. The first asparagine binding site is not likely to be at the synthetase domain

because ammonia dependent synthetase activity is much less affected by 1 mM of asparagine

than that to the inhibition glutamine dependent synthetase activity. This is also supported by the

fact that asparagine is an analog of glutamine. 4. The result that kcat value decreased with

increasing concentrations of asparagine (Figure 4-1B) indicates that asparagine binds to enzyme-

substrate (EEGln) complex. No matter how many asparagine binding sites there are in one

molecules of free enzyme, the apparent kcat can be deduced as:


1 + k3 [1]
cat' xk[ cat
1+
ykcat +k4


Where y is a factor resulting from the effect of asparagine binding on the turn over number for

the reaction from AsnEEGln complex to glutamate; k3 andn k4 are the forward and backward rate

constants for the binding of asparagine to enzyme-substrate complex (EEGln) respectively. By


fitting y, and kc, to the experimental results using KaleidaGraph, we got y value close
ykat +k4

to zero (0 + 0.3 mM). This means that the AsnEEGln complex is hardly turn over to glutamate.

Therefore, we can simplify the model as showed in scheme 4-1B. The corresponding apparent

kcat is:









1 1
kat' k x kcat x kcat
1+k3[I] 1+[]
k4 K,

Where Kli is the dissociation constant of AsnEEGln complex into asparagine and EEGln, which

was fitted as 0.33 0.04 mM; kcat was fitted as 7.4 0.2 s-1 for dimer form, consistent with the

previously reported value. 5. Based on the results that apparent KM value versus the

concentration of asparagine is positively related but not a straight line, several possible models

were proposed regarding the number of asparagine molecule binding to the free enzyme (Scheme

4-2). If there is only one asparagine binding site in the free enzyme (Scheme 4-2A), the apparent

KM versus the concentration of asparagine is a hyperpola when Kis is different than Kii.


1+ [I]
K,'= x K
1I+ []
K1+

If more than one asparagine binding sites in the free enzyme, the apparent KM versus the

concentration of asparagine is more like a polynomial line with second order for two binding

sites model (Scheme 4-2B), third order for three sites (Scheme 4-2C), and so on.

1+ [V] [1]2
K,, KK
KM'= 1 -Is Is2 x KM two binding sites model
1+V



1+ + +
K KK KKK
Km'= ~ I s2 IKI 2 K s3 x K. three binding sites model
1+ []
K,

Among these models, the best fit to the experimental results is the three sites model. The two

sites model fits the data with the concentration of asparagine below 0.2 mM, not for higher









concentration, which implys the third order term. We could not directly get all three dissociation

constants by fitting except Kis1 (0.09 0.01 mM) and KisiKIs2KIs3 (0.0039 0.0002 mM3).

However, we can estimate Kis2 and KIs3 using the fitted value of Kisi. The final value for Kis2 and

KIs3 are 0.7 0.2 mM and 0.06 0.02 mM respectively.

The model for asparagine inhibition to hASNS is much simpler since it is a competitive

inhibitor to the enzyme. Only one molecule of asparagine can bind to the free enzyme. And it

doesn't bind EGln complex. The inhibition constant was determined as 0.11 0.02 mM.

Discussion

As mentioned above, the crystal structure of C1A mutant of ASB suggests the active form

of this enzyme is dimer. This is supported by our results from chromatography studies, in which

the binding of glutamine drove the enzyme to the dimer form, similar to asparagine. Based on

this, we constructed the kinetic model for the asparagine inhibition to ASB. Our kinetic results

and the simulated model tells us the apparent Km values of ASB for glutamine was affected

significantly by the presence of asparagine and this may results from one molecule of free

enzyme can binds at least three asparagine molecules. Consider the fact that no other possible

asparagine binding site was found in the crystal structure of C1A mutant except the two active

sites of each subunit, the simulated model further support that the active form of ASB is a dimer.

Communication between domains and/or subunits and allosteric regulation is a common

phenomenon for biological macromolecules. Previous studies showed that binding of ATP at

synthetase domain of ASB stimulates its glutaminase activity, indicating a communication

between its two catalytic domains(]7). In this study, asparagine stabilized the dimer form of

ASB and drove the monomer/dimer equilibrium to the dimer side tells us that the interaction

between subunits became stronger with asparagine binding to the enzyme, implying

conformational changes and possible communication at the interface. This is also supported by









the three binding sites model for asparagine inhibition since all these asparagine binding sites are

impossible in the same subunit of ASB. The inhibition mechanism of hASNS by asparagine is

different from that of ASB. Aspargine is a simple competitive inhibitor for hASNS under the

same experimental conditions. The SEC results showed hASNS is present as dimer at the

experimental condition. Its quaternary structure was not affected by 5 mM asparagine, which is

about 50 times of inhibition constant. Therefore, the two subunits work independently and do not

communicate with each other.

In order to find possible structural basis for the communication during asparagine

inhibition, we checked those interacting residues in the interface of the two subunits. The major

interaction between the two subunits of ASB is that the guanidium group of Arg334 forms H-

bonds with a loop that closes to glutaminase active site (Figure 4-6). This loop includes a cluster

of absolutely conserved amino acid residues (H29R30G31P32D33), which involves in formation

of ammonia tunnel. The backbone of G31 and P32 and the side chain of S35 interact with the 3

nitrogen atoms in the guanidium group of R334 (Figure 4-6A). Close to Arg334, two other

residues from synthetase domain and one from glutaminase domain may also play a role in the

interaction of the two subunits. Arg334 is followed by another positive residue, Lys335. Its

amine group does not contact with the HRGPD loop, but exposed to bulk solution and interact

with several water molecules. However, the indole ring of Trp34 from glutaminase domain of the

other subunits inserts into the two long side chains of Arg334 and Lys335. With another indole

ring from Trp395, a sandwich structure is formed with the side chain of Arg334 in the middle,

probably helping Arg334 position correctly and forming hydrogen bonds with HRGPD loop,

especially Trp395 with the indole ring parallel to the side chain of Arg334. All these residues are

in highly conserved sequence region for all kinds of species, from bacterial to fungus, from plant









to mammals (Figure 4-7). It is also important to point out that the HRGPD loop has

XGPXXXGX sequence. The two glycine residues make this loop more flexible to adopt different

conformations. Based on sequence alignment, most of the species uses the same set of residues

as Escherichia coli, from bacterial to plant, while mammals have different residues at the

corresponding positions. In mammalian species, Trp34 is replaced by an Ala. This small side

chain causes less steric hindrance compared to the indole ring of Trp34 and makes the dimer

form more favorable for hASNS. Interestingly, a lysine residue in mammalian enzyme is at the

corresponding position of Arg334 in ASB. The one carbon shorter side chain may suggest a

closer contact between the two subunits in human enzyme. Consequently, the tryptophan 395 is

changed to a smaller aromatic residue, histidine. In the HRGPD loop, no second glycine residue

is present for mammalian enzyme, which has less conformational flexibility. All these

differences may provide a structural basis for the different quaternary structures of ASB and

hASNS and allosteric effect upon asparagine binding of ASB.

Using the inhibition models we constructed above, the apparent activity was simulated

with the concentration of asparagine and glutamine varying from 0.1 mM to 2 mM for both ASB

and hASNS (Figure 4-8). This ranged of concentration was selected based on the physiological

concentration of glutamine and asparagine in blood, which is about 0.6 mM and 0.1 mM

respectively. Here for the convenience of comparison with the glutaminase activity without any

inhibition, the enzyme activity with no asparagine was also simulated, although it is impossible

for cell to have no asparagine at all. The hydrolysis of glutamine is depressed significantly for

both enzymes, with more inhibition effect on ASB. At concentration of 0.6 mM glutamine and

0.1 mM asparagine, about 55% of enzyme activity is lost because of product inhibition for ASB

and 40% for hASNS. This means even at normal conditions, the glutaminase activity of









asparagine synthetase is suppressed significantly. If the asparagine concentration is above 0.3

mM, less than 10% activity will be left for ASB and less than 30% for hASNS. The presence of

asparagine prevents the enzyme from hydrolyzing glutamine at its maximum rate and producing

less toxic ammonia. As a regulator, asparagine regulates not only the function of asparagine

synthetase, but also the expression of the corresponding gene (73). In some prokaryote cells,

such as Escherichia coli, asparagine can be synthesized by ammonia dependent asparagine

synthetase using ammonia instead of glutamine. Experimental results showed that ammonia

utilizing enzyme is also regulated by asparagine through similar strategy (74, 75). Like

glutamine, asparagine is one of the principal and non-essential amino acids involved in the

storage and transport of nitrogen. The balance between glutamine and asparagine is directly

affected by this enzyme. Recently, the regulatory patterns of asparagine synthetases in

Helianthus annuus (sunflower) were reported by Herrera-Rodriguez et al (76-79). Three genes

encode asparagine synthetase, HAS1, HAS1.1 and HAS2. The expression of these genes is

regulated by light, carbon and nitrogen availability with different sensitivities. The asparagine

level is higher than glutamine during germination and cotyledon senescence. During these stages,

most of asparagine synthetases was encoded from HAS2. For leaf senescence, all three genes

work together and the asparagine level is lower than glutamine. These results imply the

importance of asparagine synthetase in keeping the relative asparagine and glutamine level.

Conclusions

Product inhibition of asparagine synthetase is a common strategy in all species to regulate

the conversion from glutamine to asparagine. However, the inhibition mechanism for

Escherichia coli ASB is different from that of hASNS. The multi-site inhibition to ASB by

asparagine and the changes of its quaternary structure after asparagine binding suggest

conformational changes at the interface of the two subunits and communications between them.









Under same conditions, hASNS showed no such properties. At room temperature, ASB and

hASNS showed different quaternary structure. The former is a mixture of dimer and monomer

with equilibrium between these two forms, while the latter presents as dimer. This structural

difference may result from the different residues at the interface of the subunits. The flexible

conformation of ASB provides the structural basis for allosteric communication. Glutamine and

asparagine are the two major sources to storage and transport nitrogen in vivo. The regulation of

the reaction that synthesizes asparagine from glutamine by asparagine inhibition may have

important physiological meanings.









Scheme 4-1


k, kcat,
A EE +GIn k EEGIn kat EE + Glu
+ k2 +
n Asn Asn

Sk41k3 kcat
EEAsn AsnEEGIn EEAsn + Glu
n

k, kat
B EE +GIn ~ EEGIn -- EE + Glu
+ k2 +
n Asn Asn

,1 K.il[
EEAsn AsnEEGIn
n









Scheme 4-2


k, kca
A EE + GIn = EEGIn- a EE + Glu
+ k2 +
Asn Asn

Kis f K]iR
EEAsn AsnEEGIn

k, kca
B EE + GIn EEGIn EE + Glu
+ k2 +
Asn Asn

Klsl K li
EEAsn AsnEEGIn
+
Asn

KIs21
AsnEEAsn

k, kat
C EE +GIn EEGIn EE + Glu
+ k2 +
Asn Asn

Ksi 1 Klil
EEAsn AsnEEGIn
+
Asn

Kls21 [
AsnEEAsn
+
Asn


KAsnsns
AsnEAsnEAsn


















25-


S20-
200 0 2 04 06 08 /
L L-asparagine (mM)

15


C 10-


5-


0
0 0.1 0.2 0.3

L-asparagine (mM)

Figure 4-1. Kinetic studies of asparagine inhibition to ASNS
A. apparent KM (mM) vs [Asn] mM


0.4 0.5
















3.5- a


3-


2.5


2


1.5


1 --- i--------
0 0.1 0.2 0.3

L-asparagine (mM)

Figure 4-1 B. apparent kcat (s-1) vs [Asn] mM













80

70

60

50

40

30 +

20


0 I I I
10


0 0.5 1 1.5 2
1/[Gln] (mM-1)

[Asn]: 0, 0.05, 0.1, 0.25, 0.5, 1 mM

Figure 4-1 C. 1/v vs [Gln] mM-1















U


5.5-








4.5-



4-I I
0 0.1 0.2 0.3
KD
Figure 4-2. SEC standard curve: log(MW) vs KD


KD V,
Sis elution volume; V is column volume; V is void volume
Ve is elution volume; Vc is column volume; Vo is void volume







































I I I I I I
7 8 9 10 11 12
Retention time (min)

SEC analysis of quaternary structure of ASNS with different concentration of asparagine
in mobile phase
A. SEC analysis of quaternary structure of ASB


Figure 4-3.


































7 8 9 10 11 12
Retention time (min)
Figure 4-3 B. SEC analysis of quaternary structure of hASNS with (black) or without 5 mM asparagine
(red) in mobile phase










Table 4-1. Retention time of ASB peaks in SEC studies


The peaks of ASB and their retention time (RT) with/without asparagine in mobile phase


[Asn] R.T. (min)
mM
Peak 1 Average (min) Peak2 Average (min) Peak3 Average (min)

0 9.410 9.086
0 9.485 9.098
0 9.458 9.131
0 9.551 9.51 0.06 9.180 9.15 0.05
0 9.596 9.225
0 9.551 9.200
0 9.498 9.146
0.10 9.561 8.861
0.10 9.556 8.873
0.25 9.546 8.831
0.25 9.516 8.801
0.50 9.491 8.771
0.50 9.481 9.49 0.05 8.768 8.79 0.05
1.0 9.520 8.800
1.0 9.493 8.788
5.0 9.416 8.726
5.0 9.410 8.748
5.0 9.425 8.713









Table 4-2. The calculated MW and predicted structure of ASB corresponding to each peak


Mobile phase With Asn Without Asn With/without Asn
R.T. (min) 8.79 9.15 9.50
Calculated MW (kD) 171 120 85
Quick exchange
Possible structure Dimer between dimer and Monomer
monomer form
True MW (kD) 125 -62.5
Calc. MW/true MW 1.37 1.36







































1 I i L I 1
7 8 9 10 11 12

Retention time (min)
Figure 4-4. SEC analysis of quaternary structure of ASB with different amount of enzyme injected
(2.5-40 gL)







98












0.8-

0.75-


0.7-
"E E

0.65-0

t 0.6-

0.55-


0.5-

0.45 I
0 1 2 3 4 5

L-asparagine (mM)

Figure 4-5. Fraction of dimer vs concentration of asparagine





















































































































































































Figure 4-6. The interface of the two subunits
















100


I

I
::::,
'
~::::"`
""
"
::: :.

'




'''

'""''' '"
















B_Eco_kl2
B_Shigellal
B_Salmonella
B_Erwinia
B_Yersinia
B_Vibrio
B_Shewanellal
F_yeas t
F_Candida
F_Schizo
P_wheat
P H vulgare
P_Aspargus
P_Chris_bells
P_S_parvifloral
P_Lcorniculatus2
P_fava_bean
P M sativa
P A sinicus2
P_soybean
P_soybean2
P_ Pvulgaris
P_S_parviflora2
PP vulgarisl
orangutan
human
cow
dog
brown
mouse
hamster
chicken
zebra


....HRGPDWSG........PMYLMSRKI........AWGVEARVPFLD...
....HRGPDWSG........PMYLMSRKI........AWGVEARVPFLD...
....HRGPDWSG........PMYLMSRKI........AWGVEARVPFLD...
....HRGPDWSG........PMYLMSRKI........AWGVEARVPFLD...
....HRGPDWSG........PMYLMSRKI........AWGVEARVPFLD...
....HRGPDWSG........PMFLMGRKI........AWGVEGRVPFLD...
....HRGPDWSG........PMYLMARKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLLSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMYLLSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMYLLSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMARKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........VWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........AWGLEARVPFLD...
....HRGPDWSG........PMFLMSRKI........AWGLEARVPFLD...
....HRGPDAFR........GMYLISKYI........AHGLELRVPFLD...
....HRGPDAFR........GMYLISKYI........AHGLELRVPFLD...
....HRGPDAFR........GMYLISKYI........AHGLELRVPFLD...
....HRGPDAFR........GMYLVSKYI........AHGLELRVPFLD...
....HRGPDAFR........GMYLISKYI........AHGLELRVPFLD...
....HRGPDAFR........GMYLISKYI........AHGLELRVPFLD...
....HRGPDAFR........GMYLISKYI........AHGLELRVPFLD...
....HRGPDAFR........GMYLVSKYI........AHGLELRVPFLD...
....HRGPDAFR........GMYLVSKYI........AHGLELRVPFLD...


Figure 4-7. Sequence alignment of the interface residues






















1.75
1.5
1.25
0.9 1
0.75 >
0.7 0.5
0.25
0.5 0

0.3 05040.3 0.2 0"' 0C
00.5 0.4
0.1 0.9 0.7 0.6







2.75
.2.5

2.25
2
1.75
1.5
1.25
1
0.75
0.5
0.25
0.7 0

0.4 C dhd
0.1 00 o
"- D o


Figure 4-8. The glutaminase activity of ASNS around physiological conditions






102










CHAPTER 5
THE UTILIZATION OF DIFFERENT NITROGEN SOURCES BY ASPARAGINE
SYNTHETASE: EVOLUTION FROM Escherichia coli TO HUMAN

Introduction

Glutamine-dependent asparagine synthetase (EC 6.3.5.4) catalyzes ATP-dependent

synthesis of asparagine from aspartate, utilizing either glutamine or ammonia as nitrogen source.

For more than 15 years, the mechanisms of enzymatic reaction and ammonia tunneling by

asparagine synthetase was extensively studied using asparagine synthetase B from Escherichia

coli. Furthermore, due to the difficulties in expression of human asparagine synthetase (hASNS),

ASB also served as a model system for designing inhibitor of asparagine synthetase, since the

amount of human asparagine synthetase is directly related to the resistance of leukemia cell to

the asparaginase chemotherapy. So far, only one crystal structure was resolved for the C1A

mutant of ASB. It is a homodimer with two active sites in each subunit. The two sites,

glutaminase site and synthetase site, are separated with an intramolecular tunnel that is about 20

A long. The way substrates bind to the active sites is like from back door and front door.

Glutaminase site is a appa four layer sandwich structure, which is common in the class II

glutamine dependent amidotransferase. The synthetase site consists mostly of a-helices. Human

enzyme has 37% identity to ASB based on their primary structure. And most of the functional

residues, including those in the active sites and those tunnel residues for ammonia channeling,

are conserved in both enzymes. Both hASNS and ASB are supposed to have same catalytic

mechanism and overall structure. On the other hand, since they are from different sources, one

from bacterial, the other from human, the two enzymes are evolutionarily different from each

other. With more studies on hASNS after it was successfully prepared using a baculovirus

expression system in Richards' lab (57), more results showed the differences between them. In









previous chapter, studies about asparagine inhibition showed different regulatory mechanisms to

the glutaminase activity of hASNS and ASB. Asparagine inhibits the glutaminase activity of

hASNS by a simple competitive mechanism to glutamine, while it regulates the glutaminase

activity of ASB through a cooperative binding mechanism. The first binding constant is close to

that of hASNS, implying asparagine binds to the same active site first in both enzymes.

Consequently, asparagine has higher inhibition effect on ASB due to consecutively binding of

three asparagine molecules to the enzyme. Since ASB is a homodimer and each subunit has only

two possible asparagine binding sites, the cooperative binding of three asparagine molecules

suggests the communications between the two subunits, as well as between the two domains. In

the latter case, the binding of asparagine to the glutaminase active site affects synthetase site.

Furthermore, ATP itself, or ATP with aspartate, stimulates the glutaminase activity of ASB,

which support the idea that the two catalytic domains communicate with each other and substrate

binding to synthetase site results in the changes in the glutaminase site. Therefore, the

communication between the two domains in ASB is two-way effect. The studies on the

quaternary structure of ASB showed that the asparagine binding stimulates dimerization,

implying the conformational changes for the interface residues between the two subunits. This

may provide the structural basis for the inter-subunit communication. As for hASNS, its

glutaminase activity is further decreased with both ATP and aspartate present (Beeson,

unpublished results). The communication between the two domains is from synthetase site to

glutaminase site. The results about asparagine binding showed no communication from

glutaminase site to synthetase site. This is also supported by the fact that asparagine binding did

not change the quaternary structure of human enzyme. In this Chapter, I investigated the

utilization of two different nitrogen sources, glutamine or ammonium, by hASNS using the new









developed NMR assay. Compared with the results of ASB, I will discuss the evolutionary

differences between hASNS and Escherichia coli enzyme (ASB).

Material and Methods

The experiments were performed as described in Chapter 2 Materials and Methods part.

Materials

See Chapter 2 Materials and Methods part. Recombinant, wild type hASNS was expressed

and purified following literature procedures (57).

Competition Experiment

Reaction mixtures consisted of 100 mM HEPPS buffer (pH 8.0), 60 mM MgC12, 10 mM

ATP, 20 mM aspartate and different concentrations of 15NH4C1 (pH 8.0) or glutamine in total

volume of 2 mL. In experiments where the concentration of 15NH4C1 (pH 8.0) was varied as 10

mM, 25 mM, 50 mM, 75 mM and 100 mM, L-glutamine was fixed at 20 mM. Alternatively,

when L-glutamine was varied as 0 mM, 2.5 mM, 5mM, 10 mM, 20 mM and 40 mM, 15NH4C1

was added at an initial concentration of 100 mM. Reactions were initiated by the addition of

hASNS (18 [tg) and the resulting samples incubated for 10 min at 37 C before being quenched

by the addition oftrichloroacetic acid (TCA) (60 pL). After centrifugation for 5 min at 3000 rpm

to remove precipitated protein, the supernatant was adjusted to pH 5 by the addition of 10 M aq.

NaOH and an aliquot of this solution (650 atL) transferred to a 5 mm NMR tube for analysis. At

higher pH values, amide NH exchange precluded the derivation of any quantitative relationship

between peak area and 15N-Asn concentration. The standard samples contained 100 mM HEPPS

buffer (pH 8.0), 60 mM MgC12, 10 mM ATP, 100 mM 15NH4C1 (pH 8.0), 20 mM glutamine and

different concentrations of 15N-L-Asn from 0.05 mM to 2.5 mM, without adding hASNS.

NMR Measurements, HPLC assay and Kinetic Simulations

See Chapter 2 Materials and Methods part.











Results and Discussion

When both ammonium and glutamine are present in the reaction mixture, the two nitrogen

sources compete with each other for the P-aspartyl-AMP intermediate to produce asparagine. If

15NH4+ and 14N-glutamine are used, the ability of utilization of different nitrogen source by

hASNS and the efficiency of ammonia transfer can be investigated by quantification of 15N-

asparagine, produced from 15NH4+, and 14N-asparagine, from glutamine, using the new

developed NMR assay, combining with HPLC assay reported before (Fig. 5-1).

As shown in Table 5-1, the incorporation of 15N into asparagine is increased with more

15NH4+ present in the solution, with constant concentration of glutamine (20 mM). The 14N-

asparagine/ N-asparagine ratio dropped down from 4.6 to 0.4. When the concentration of

15NH4C1 was fixed as 100 mM, at low concentration of glutamine, the 14N-asparagine produced

increased with glutamine. Unfortunately, due to the low amount of 14N-Asn, the error is big.

However, we can see the positive relationship between the concentration of glutamine and the

14N-Asn/15N-Asn ratio. While at high concentration of glutamine, the ratio dropped down, which

suggests that the production of 14N-Asn slowed down, based on the results that ammonia

dependent synthetase activity keep constant. This may result from the inhibition to the

glutaminase activity of this enzyme. Both glutamine and asparagine are potential inhibitors. The

glutamine inhibition (substrate inhibition) was oberserved when I determined kinetic parameters

for its glutaminase activity by varing the concentration of glutamine (0-80 mM), which suggest

that a second molecule of glutamine binds to the enzyme and inhibits its glutaminase activity.

Previous work also showed that asparagine is a good inhibor to its glutaminase activity with Ki

value of 120 piM. However, under 1 mM concentration, the inhibition appears competitive to

glutamine and didn't affect kcat value. Since the total production of asparagine is less than 0.5









mM, and relatively small compared to the concentration of the substrate, this inhibition is very

likely caused by glutamine inhibition.

As described in Chapter 2, if ammonia can not access the active site with the presence of

glutamine, all the asparagine would be 14N product at saturating concentration of glutamine. In

this case, free ammonia is suppressed by glutamine. The ratio of 14N asparagine to 15N

asparagine would be infinite. This ratio will be decreasing with the extent of suppression

becomes less. When there is no suppression at all, the 14NH3 will compete with 14NH3 freely. If

ammonia tunneling is fast and efficient enough, then the ratio of 14N asparagine to 15N

asparagine would be determined by the ratio of production rates using ammonia and glutamine.

That is,

[14N- Asn] ,14 (k,, /1Km-tunnehng)[Gln][E]G n [Gln][E]G0n [Gln][E]G0n
[5 N-Asn] V15 (kct / Km- bndng)[15NH3][E]NH3 [15NH 3][E]NH3 [15NH3],[E]NH3

Therefore,

[14N Asn] [Gln] [E]0ln [15N Asn] [15NH3 [E]NH
oc or oc
[5N Asn] [15NH3 [E]NH3 [14N- Asn] [Gln][E]G1n
Where,

[E]NH is the enzyme form that can bind exogenous ammonia and produce asparagine.
[E]G0n is the enzyme form that can bind glutamine and produce asparagine.

Alternatively, if the tunnel is not efficient, the ratio of 14N asparagine to 5N asparagine would

decreases. Most of the asparagine would be labeled form and the ratio of 14N/15N in asparagine

would be close to zero if the 14NH3 intermediate leaks and equilibrates with 15NH3 from bulk

solution, since [15NH3] was much bigger than [14NH3].

The kinetic model was constructed based on the assumption that glutamine competes with

exogenous ammonium, which includes 1) asparagine inhibition, 2) glutamine inhibition, and 3)









competition between 14N source and 15N source (Figure 5-2). The inhibition constants for

aspargine and glutamine were set to 120 [iM and 120 mM respectively. The simulation results

were consistent with the experimental results (Figur 5-3AB), which support the hypothesis that

glutamine is a competitive substrate to exgenous ammonium and the ammonia tunneling is fast

and efficient enough in hASNS so that ammonia intermediate produced from the hydrolysis of

glutamine is competing with 15NH3/15NH4+. The binding of glutamine did not suppress the

utilization of ammonium by hASNS.

Compared with the results of ASB, hASNS showed less preference of using glutamine as

nitrogen source. Under the same experimental conditions, the 14N-Asn/15N-Asn ratio in the ASB

catalyzed reaction is bigger than that in hASNS catalyzed reaction (Figure 5-4A), or inversely,

the 15N-Asn/14N-Asn ratio is less (Figure 5-4B). Relatively less 15N was incorporated into

asparagine by ASB, suggesting more preference to glutamine by ASB relative to human enzyme.

The presence of glutamine partially suppresses the utilization of exogenous ammonia by ASB. In

another word, ASB prefers using glutamine than ammonium, while human enzyme is more

capable of using ammonium as nitrogen source to synthesize asparagine than ASB. This

difference may evolve from the need of utilizing different nitrogen sources. In Escherichia coli,

both ammonia dependent and glutamine dependent asparagine synthetase were discovered. The

two enzymes have little similarities in structure and not evolutionary related (11). The ammonia

dependent asparagine synthetase (ASA), encoded from asnA, can only use ammonium as

nitrogen source. And glutamine dependent asparagine synthetase (ASB) can use both ammonium

and glutamine, with a preference of glutamine. In human cell, only glutamine dependent

asparagine synthetase was found, which is homologous to the Escherichia coli enzyme, ASB.

The human enzyme can also use both ammonium and glutamine, with higher ability for using









ammonium, relative to ASB. Since no ammonia dependent asparagine synthetase was found in

human genome, this kinetic difference may result from the need of ammonia utilizing enzyme in

human cells. During the process of evolution, human enzyme gradually developed more ability

to utilize exogenous ammonium, and kept its glutamine dependent asparagine synthetase activity,

while ASB functions mainly as glutamine dependent enzyme since Escherichia coli cell can

express ammonia dependent asparagine synthetase.

Conclusions

The studies on human asparagine synthetase showed that when both ammonia and

glutamine are present in the reaction mixture, ammonia competes freely with glutamine as the

nitrogen source. The simulated results based on this study are consistent with the experimental

results. In previous chapter, Escherichia coli asparagine synthetase showed partial suppression of

ammonia by glutamine. Compared to hASNS, the bacterial enzyme has higher preference to use

glutamine, and relatively less preference of ammonia. This may result from evolution due to the

fact that in Escherichia coli, there is ammonia dependent asparagine synthetase. It is not

necessary for glutamine dependent enzyme to utilize ammonia. However, no ammonia

dependent AS was found in human cells. The human enzyme may function as both ammonia and

glutamine dependent enzyme during the evolution.












pH 8 buffer + NIHCI + 4'-
(+nbstfate) +enzyme
10 os atl 3" C




Ajust pH to .0 by addi:


liqu ccettrLont
aliqtot


pH9 buffer + DMSO
+DNFB in EtOH
45 mis at 50 C'


quencd. neItralize, spin.
lnalyzeusini2 RP-HPLC


* I. '-
1 ,


I 1N-.Pi+14N-.ii 1


Figure 5-1. Quantification of 1N-asparagine and 14N-asparagine using NMR assay and HPLC assay.
The enzyme can use both "NNH4C1 and 14N-glutamine as nitrogen source. With both in the
reaction mixture, both "N-asparagine and 14N-asparagine were produced. The amount of
"N-asparagine produced can be directly determined using NMR assay. The total amount of
asparagine can be quantified by HPLC assay. The difference is the amount of 14N-
asparagine produced.











15 E1SNAsn kj1 15N k-Asn 1 kl
N-Asn + EEN-Asn EN-AsnNH, 1NH, + E IN-Asn EGInGIn + 5NHI- EGInGInNSNH3
k-9 k14,
kl21k-12 kk8 klk-13 Ik15

15N-Asn 15N-Asn Gin EGInGIn + 5N-Asn
+ + +
k4 k5-Asn E kl
15N-Asn + E E INH,3 NH3 + + Gin EGIn
"-3 -1
+ + k2j
Asn Asn
k E + Asn+ Glu
k"1 k.11 k5 k.5
k k Asn: 14N-Asn
N-Asn + EAsn EAsndNH3 15NH, + EAsn E: EATPAsp complex
Figure 5-2. Simulation model for the competition reactions catalyzed by hASNS.
Figure 5-2. Simulation model for the competition reactions catalyzed by hASNS.

















0.45

0.4

0.35

0.3

0.25

0.2

0.15

0.1

0.05

0 1


5 10 15 20 25 30 35 40 45

[Gin] mM


3


2.5


2


1.5


1


0.5


0


0 20 40 60
[15NH4 mM


80 100


Figure 5-3. Simulations of competition reactions catalyzed by hASNS. A) varing glutamine, brown
square: experimental results; blue, simulation. B) varing ammonium, blue: simulation;
purple square: experimental


i














1.4

1.2

1

0.8

0.6

0.4

0.2


DASB

xhASNS


'V


[Gin] mM


0 20 40 60
[15NH4] mM


80 100 120


Figure 5-4. Utilization of ammonia and glutamine as nitrogen source by ASB and hASNS. In general,
ASB shows higher preference to use glutamine as nitrogen source. A) varing glutamine,
square: ASB; cross, hASNS. B) varing ammonium, blank diamond: ASB; square: hASNS











Table 5-1. hASNS catalyzed production of 1N into L-asparagine in the steady-state competition assays.

Gin 15 b 1c Total Asn 14N-Asn/"N-
(mM) 1NH4C1 (mM) 1N-Gln (mM)b N-Asn (mM) M)d Asn


0 100 NDe 0.21 + 0.02 0.215+ 0.009 0
2.5 100 0.025 0.004 0.19 0.05 0.21+ 0.02 0.1 0.2
5.0 100 0.05 + 0.02 0.21 + 0.03 0.244 0.007 0.2 0.2
10 100 0.060 + 0.002 0.21 + 0.02 0.28 0.02 0.3 + 0.1
20 100 0.054 0.006 0.198 0.006 0.28+ 0.02 0.4 0.1
40 100 0.07 + 0.01 0.228 + 0.003 0.269 0.002 0.2 0.02
20 0 NDe NDe 0.105 0.004 NAe
20 10 NDe 0.031 + 0.005 0.175 + 0.002 4.6 0.7
20 25 NDe 0.074 + 0.010 0.222 0.009 2.0 + 0.3
20 50 0.02 0.01 0.14 0.02 0.26 0.02 0.9 0.2
20 75 0.047 + 0.002 0.1721 0.0007 0.25 0.02 0.4 + 0.1
20 100 0.054 + 0.006 0.198 + 0.006 0.28 0.02 0.4 + 0.1


a All reactions contained 10 mM MgATP, 20 mM aspartate, 60 mM MgC12 and 280 nM hASNS in 100 mM HEPPS
buffer, pH 8 (2 mL total volume).
b As determined by gHMQC NMR spectroscopy. Errors are estimated on the basis of measurements employing
known concentrations of authentic 5N-Asn.
This value is corrected for the amount of 15N-Asn that would be formed from 15N-Gln at natural abundance, but
the contribution of 14N-Asn formed from any 14NH3 released from the enzyme, as a result of the glutaminase
activity of AS-B, is assumed to be negligible.
d Mean value and standard deviation computed from two separate determinations on duplicate samples using
reverse-phase HPLC.
e ND not detected; NA not applicable.









CHAPTER 6
PREPARATION OF LIGAND BOUND hASNS INHIBIITION AND BINDING STUDIES

Introduction 1

Glutamine dependent asparagine synthetase (ASNS, EC 6.3.5.4) catalyzes ATP dependent

conversion of aspartate to asparagine using glutamine as the nitrogen source. One of the interests

in asparagine synthetase lies in the correlation of this enzyme to the resistance of the acute

lymphoblastic leukemia cells after treated with L-asparaginase. For a long time, the Escherichia

coli asparagine synthetase B (ASB) was studied as a model for the human enzyme because of its

readiness to be expressed in a large amount. ASNS belongs to class II amidotransferase, with

two active sites that are separated by an intramolecular tunnel. Each of the two active sites

catalyzes a half reaction. At glutaminase site, ammonia is released from glutamine through a

thioester intermediate formed by the nucleophilic attack of the thiolate group of Cysl residue to

the amide group of glutamine. At the other end of the tunnel, the y-carboxylic acid group of

aspartate is activated by ATP and form the P-aspartyl-AMP intermediate, followed by ammonia

attacking to its y-carbonyl group to produce asparagine (Figure 6-1). Although both human

ASNS and ASB show same properties from the point view of chemistry, the kinetic studies of

human enzyme revealed several differences from that of E. coli enzyme, which were discussed in

the previous chapter. So far, the crystal structure of C1A mutant of Escherichia coli asparagine

synthetase B (ASB) is the only one that was reported for asparagine synthetase (ASNS). Since

the function of biomolecules can only be fully understood under the structural context, more

crystal structure of ASNS, especially of hASNS, need to be studied. Recently, human asparagine

synthetase (hASNS) was successfully expressed in a baculovirus system using sf9 insect cells.



1 This work was done with Alexandria Berry, an undergraduate. Li designed the experiments and analyzed the
results. Li and Berry performed the work.









This makes possible to have enough protein for X-ray crystallographic studies of human enzyme.

In this work, efforts were made to prepare inhibitor-bound hASNS. 4-diazo-5-oxo-L-

norleucine (DON) is a glutamine analog. It irreversibly inhibits the glutaminase activity of class

II amidotransferases by reacting with the thiol group of the catalytic Cysl residue in the

glutaminase active site and forming a covalent C-S bond (Figure 6-2). The transition state analog

Adenylated sulfoximine, mimicing that P-aspartyl-AMP is attacked by ammonia, was a good

inhibitor for human asparagine synthetase with KI at nM level (71) (Figure 6-3). My results

showed that both inhibitors can bind to the enzyme, with DON in the glutaminase site and

sulfoximine in the other.

Materials and Methods

Materials

Unless otherwise stated, all chemicals and reagents were purchased from Sigma (St. Louis,

MO). L-Glutamine was recrystallized prior to use as previously described (53). C-terminally

tagged human ASNS were expressed in and purified from a baculovirus/insect cell expression

system as previously described (16, 57). Adenylated sulfoximine (as a 1:1 mixture of

diasterioisomers) were obtained from Japan.

Enzyme Assays

The glutaminase activity of hASNS was determined by measuring the production of

glutamate using glutamate dehydrogenase (L-GLDH) coupled assay. ASNS reaction mixtures

(200 [tL total volume) contained 100 mM NaC1, 10 mM MgC12, and 25 mM L-glutamine in 100

mM EPPS. Reactions were initiated upon the addition of hASNS and incubated at 370C for 10

minutes, and terminated by addition of 30 [tL 20% (v/v) trichloroacetic acid (TCA). The reaction

mixtures was added to 770 [tL coupling assay mixture, which consisted of pH 9.0 buffer (300

mM glycine, 250 mM hydrazine), 1.5 mM NAD+, and 1 mM ADP. The absorbance was recorded









before and after the addition of 2.2 units of glutamate dehydrogenase (0 and 30 minutes).

Glutamate was quantified using standard curve. The synthetase activity with either glutamine or

ammonia as nitrogen source was assayed by determining the formation of PPi using PPi reagent

(Sigma Technical Bulletin BI-100), based on the fact that PPi and asparagine were produced at

1:1 ratio in the hASNS catalyzed reaction.

Loss of Glutaminase Activity with Time at Different Temperature

The stability of hASNS was first investigated to determine the best incubation time. Time-

based glutaminase activity of hASNS was tested at three different incubation temperatures, in ice,

room temperature and 37 C. After 10 iL of deionized water being added to one tube

(approximately 100 tL) of the stock enzyme solution (1 mg/mL, 50 mM EPPS, pH 8, 5 mM

DTT, and 20% glycol), the solution was aliquoted into 3 tubes, which were incubated on ice(0-4

oC, at room temperature (approximately 220C), and at 370C respectively. At a series of time

points (0-61 minutes), 2 pL incubated enzyme from each aliquot were removed and immediately

added into 198 pL of glutaminase reaction mixture to test its glutaminase activity.

DON Inhibition

The inhibition mixture was prepared by mixing stock enzyme solution (1.0 mg/mL) and

DON aqueous solution (50 mM), which contained 0.9 mg/mL hASNS and 5.88 mM DON, while

the control was made by mixing stock enzyme solution and deionized water. EPPS solution was

filtered using Fisherbrand 25 mm Syringe Filter, 0.2 pm, nylon. Solutions were incubated at

room temperature and 15 pL aliquots were removed simultaneously from both inhibition mixture

and control at varying time points (4-30 minutes), and then the enzyme was separated from free

DON by using 30,000 nominal molecular weight limit (Daltons) Microcon Centrifugal Filter

Devices spin columns (Millipore Corporation, Bedford, MA). After each aliquot was spun down

for 1 minute at 10,000 rpm, the enzyme was rinsed twice using 100 aL of 100 mM filtered EPPS









(4 minutes at 13,000 rpm). The purified enzyme was recovered in a new centrifuge tube by 30

tL of 100 mM filtered EPPS, which was added to the bottom side of the filter (1 minute at 7,000

rpm). All enzymes recovered were immediately assayed for their glutaminase activity (10 gL

recovered enzyme). The DON inhibited enzyme was also checked for synthetase activity versus

control enzyme synthetase activity using the pyrophosphate assay and concentration using the

Bradford assay.

Loss of Synthetase Activity with Time at different Temperature

The time dependent loss of synthetase activity was determined at both room temperature

and 37 C. About 12% deionized water was added to the stock enzyme. The enzyme solution

then was aliquoted into 8 different tubes and stored in a -800C freezer. These tubes were removed

from the freezer one by one and sit at room temperature (approximately 220C) for different times

(0-70 minutes), or at 370C over a 60 minute period, followed by testing their synthetase activity

(incubated at 370C in 100 mM EPPS, pH 8, containing 10 mM MgC12, 100 mM NH4C1, 20 mM

aspartate, and 2.5 mM ATP over a period of 10 minutes with a 1 mL final volume).

Inhibition of the DON Inhibited Enzyme by Sulfoximine

The DON inhibited enzyme was prepared as described above, with incubation time being

20 minutes. The synthetase activity of the DON inhibited enzyme was assayed using the

pyrophosphate assay with 0-10 VM sulfoximine. As for the control enzyme, the synthetase

activity was measured with 0 gM or 10 tM sulfoximine inhibitor.

Results

Stability Experiment

Previous studies by size exclusion chromatography showed that hASNS dissociates into

monomer from dimer with time. The structural changes may result in the loss of its activity.

Furthermore, the unstable tholate group of Cysl is readily oxidized in the solution, which causes









loss of the glutaminase activity of the enzyme. Here we investigated the stability of the enzyme

before we characterized the inhibition to hASNS by DON and sulfoximine.

hASNS lost glutaminase activity with time at room temperature and 37 C

Aliquots of hASNS were incubated in ice, at room temperature (about 22 C) and at 37 C

respectively. The glutaminase activity of hASNS was immediately tested after being incubated

for a certain time by previously reported assay method (17). Our results showed that hASNS lost

its glutaminase activity faster at higher temperature. When incubated in ice, its activity decreased

by 15% after 60 minutes (data not shown). While after 20 minutes' incubation at 37 oC, the

enzyme lost 50% of its glutaminase activity (Figure 6-4, Table 6-1). After 60 minutes at 37 oC,

more than 85% activity was lost. At room temperature, hASNS still kept 57% of its activity after

60 minutes. The percent activity left decreased exponentially with time, which indicates a first

order reaction for the lost of glutaminase activity. The calculated half-life for glutaminase

activity of hASNS are 74 minutes and 20 minutes for room temperature and 37 oC respectively.

The loss of NH3 dependent synthetase activity with time at room temperature and 37 C

The ammonia dependent synthetase activity was also checked after hASNS was incubated

at room temperature or at 37 oC (Figure 6-5, Table 6-2). After being incubated for certain time,

hASNS was added to assay mixture to initiate the reaction. The results showed that hASNS lost

its ammonia dependent activity in different patterns for short time incubation and long time

incubation. At room temperature, the enzyme lost little synthetase activity until being incubated

for more than 30 minutes, and then its activity dropped exponentially. Similar results were

observed for the enzyme incubated at 37 oC. When the incubation time was shorter than 15

minutes, the synthetase activity went down more like linearly than exponentially. After being

incubated for more than 15 minutes, the activity decrease exponentially.









Inhibition to hASNS by DON

It has been reported that DON reacts with thiol group of Cysl residue of several other class

II amidotransferases and forms covalent bonded complex (E-ON). While few studies about DON

inhibition to asparagine synthetase were reported, we did this experiment to support the

hypothesis that DON is a glutaminase inhibitor for all the class II amidotransferases and has little

effect on synthetase activity of hASNS.

DON inhibits 90% of glutaminase activity of hASNS after incubating for 20 minutes.

k[DON] = 0.0026 s-1
E EDON
[DON] = 5.88 mM


Before measuring its glutaminase activity after hASNS was incubated with 5.88 mM for a

series of time, we separated E-ON from DON by using spin column (see methods part) because

our results suggested that DON is a substrate of coupling enzyme glutamate dehydrogenase (L-

GLDH) and affect the glutaminase assay. The protein concentration was tested using modified

Coomassie assay and the percent recovery of hASNS for separation step was 71 + 5 %. For each

time point, we used incubated enzyme without DON as control to eliminate the effect caused by

separation step (Figure 6-6, Table 6-3). After 20 minutes' incubation, the glutaminase activity of

DON inhibited hASNS was only about 10% of control enzyme. The percent activity decreased

exponentially with time. This indicates a first-order reaction for the inactivation of hASNS by

5.88 mM DON and the rate constant was calculated as 0.0026 s-1.

DON had little effect on the rate of ammonia dependent synthesis of asparagine.

In this work, we did not try to characterize the ammonia dependent synthetase activity of

DON incubated enzyme. We simply compared the production rate of pyrophosphate by DON

incubated enzyme with that by control enzyme (incubated without DON) at saturating level of all









the substrates (except ammonia at 100 mM). Several reports showed that Asn:PPi ratio close to

1:1 for free enzyme or inhibited enzyme. Therefore, we here assume that DON incubated hASNS

will not affect Asn:PPi ratio during the enzyme catalyzed synthesis. Note: In order to fully

characterize the inhibition mechanism, this assumption need to be confirmed by using HPLC-

based end point assay (23) and PPi assay (22). The concentration of the recovered enzyme

solutions was determined to be 0.78 mg/mL using the Bradford assay. The volume of recovered

enzyme solution varied from about 12 to 18 pL (approximately 0.0117 mg of enzyme), versus

the initial volume of 15 pL of 1.0 mg/mL of stock enzyme (0.015 mg). With all the substrate at

saturating level, the production rate of pyrophosphate had no significant changes between E-ON

complex and control enzyme, which implies little changes for ammonia dependent synthetase

activity (Table 6-4).

Inhibition by Adenylated Sulfoximine

The transition state analog adenylated sulfoximine is a good inhibitor with KI at nanomolar

level. It was reported to suppress proliferation of an L-asparginase-resistant leukemia cell line

and is a potential solution for drug resistance in lymphoblastic and myeloblastic leukemia cells

(71). Here we examined its inhibition to DON incubated hASNS and checked the difference

between its inhibition effect on DON incubated hASNS and on free enzyme.

Inhibition to the ammonia dependent synthetase activity of stock hASNS

The inhibition to hASNS by sulfoximine has been studied regarding its ammonia

dependent synthetase activity as well as glutaminase activity. While sulfoximine was reported as

a good inhibitor to synthetase activity of hASNS, little effect on glutaminase activity of hASNS

by sulfoximine was discovered. As control experiment for the studies of its inhibition to DON

incubated hASNS, we tested ammonia dependent synthetase activity of stock hASNS with 3

different concentrations of sulfoximine, 0, 5, and 10 uM respectively. The initial rate was 0.0404









+ 0.0007 pM/s. The production of PPi catalyzed by hASNS without sulfoximine linearly

increased with time. With the presence of inhibitor, the production gradually slowed down with

time (Figure 6-7).

Data were processed based on the slow onset inhibition model as reported in literature (71,

80). Three parameters, vss, vo and k were determined by fitting the data using KaleidaGraph v3.5

from Synergy Software, where vss and vo are initial and steady-state velocities respectively, and k

is the apparent first-order rate constant for isomerization of El to El*. k6 was calculated using the

following equations with these three parameter determined at different concentrations of

inhibitor:


k6 =kx (1)
V0

The average value of k6 then was used to estimate Ki and k5 by fitting to the equation below.

[I]
K,
k = k6 +5 k x (2)
+[ATP] [I]
1+ +
Ka K,

The calculated kinetic results were listed in Table 6-5 as well as those reported in the literature.

Inhibition to the ammonia dependent synthetase activity of DON incubated hASNS

The ammonia dependent synthetase activity of DON incubated hASNS was tested with

different concentrations of sulfoximine (0-10uM), immediately after purified E-ON was prepared.

The incubated hASNS without DON was used as control. The production of PPi without

sulfoximine linearly increased with time for both DON incubated enzyme and control enzyme,

with kcat value of 0.09 0.01 uM/s and 0.047 0.005 uM/s respectively (Figure 6-8 and 6-9).

With sulfoximine, the production of PPi versus time was consistent with the slow onset

inhibition model, same as what we got before for stock enzyme (Figure 6-8). Based on the results









for control enzyme, the rate constants for conversion of El to EI* were calculated and listed in

Table. Here we assume the control enzyme has same Ki value as stock enzyme. Little difference

was found between these rate constants for control enzyme and for stock enzyme (Table 6-6).

This showed that the binding property of hASNS to sulfoximine had no significant changes after

the control enzyme was prepared. Unfortunately, for DON incubated enzyme, we could only

accurately determine the initial velocity vo by fitting the data using KaleidaGraph, not vss and k.

All the fitting gave a big error for vss and k, which has no practical meaning, except k was fitted

as 4.9x10-4 S-1 at 10 tM of sulfoximine (Figure 6-9).

Discussion

The catalytic mechanism can only be completely understood under structural context.

hASNS has two active sites, glutaminase site and synthetase site. In order to prepare inhibitor

bound hASNS and make crystal structure, we did kinetic studies about hASNS using two

inhibitors, DON and sulfoximine. DON is a glutamine analog. DON incubated enzyme showed

significant decrease for its glutaminase activity after being separated from DON and diluted for

more than 30 fold in the enzyme catalyzed reaction. The first order rate constant was calculated

as 0.0026 s-1 in the presence of 5.88 mM DON at room temperature. After 20 minutes'

incubation with DON, 90% of hASNS was inhibited based on its glutaminase activity while its

ammonia dependent synthetase activity decreased about 10%. This result supports the idea that

DON reacts with the thiolate group of Cysl residue of hASNS and form covalently bound

enzyme, although no direct evidence shows the inhibition is irreversible for hASNS. The

covalently bound enzyme complex has been observed for other class II amidotransferases, such

as glutamine phosphoribosylpyrophosphate (PRPP) amidotransferase, glutamate synthase and

glucosamine-4-phosphate synthetase (GlmS) (28, 54, 81, 82). All this evidence indicates that

DON incubated hASNS is E-ON complex. The fact that DON incubated enzyme lost









glutaminase activity without affecting its synthetase activity significantly proves that DON, as a

glutamine analog, binds glutaminase active site, not synthetase site. This conclusion was further

supported by the result that the ammonia dependent synthetase activity of DON incubated

enzyme was inhibited by sulfoximine, an analog of transition state that formed in the synthetase

site, which had little effect on the glutaminase activity of hASNS by previous studies. The

apparent first order rate constant was calculated as 0.00049 s-1 for the inhibition of DON

incubated enzyme with the presence of 10 gM sulfoximine. The calculated half-life was 24

minutes and the time for 90% inhibition was 78 minutes.

hASNS is not stable. After one hour at room temperature, the enzyme lost 43% of its

glutaminase activity and 11% of synthetase activity. When the enzyme was kept at 37 C, these

numbers are 87% and 67% respectively. The lost of its glutaminase activity may result from the

oxidation of the thiolate group of Cysl. However, it is unexpected that its ammonia dependent

synthetase activity also decreased with time. Maybe this is caused by structural changes of the

enzyme at higher temperature because its molecular weight changed from dimer form to

monomer form by size exclusion chromatography.

The sulfoximine inhibition to DON incubated hASNS is different from the control. Only k

value for the highest concentration of sulfoximine (10 tM) was estimated as 0.000489 s-1. It is

decreased by 5 times compared to that of control enzyme, which means the formation of EI*

became slower for DON incubated enzyme. Since stock enzyme and control enzyme showed

similar binding affinity to sulfoximine. We can exclude the possibility that this was caused by

incubation for 20 minutes at room temperature as well as separation step. Therefore, we can

conclude that the formation of the covalent complex from the reaction of DON with enzyme at

glutaminase site resulted in some conformational changes of the synthetase site. Since k can be









expressed as a function of k5, k6 and Ki at certain concentration of ATP, and k6 << k or k5 for

control enzyme, the decreasing of the k value is likely caused by the decreasing of k5, or

increasing of Ki or both. This implies that DON inhibited enzyme has some structural changes

either that affect the binding of sulfoximine or that make the "on" rate of El to EI* slower. This

can be supported by the following analysis.

If incubation with DON does not affect its binding to sulfoximine, that is, Ki is a constant

with value of 0.181 aM, so

I+ [ATP] [I] 1+ [ATP] [I]
Ka K K, K1
k, =(k- k6)x ] [I] [I]
K1 K1

+ 5mM 10uVM
1+--+
1 /in m-4 -i 0.2mM 0.181pMf _
k5 <4.9x10-4S1 0. 2mM 0.=18l 7.2x10 4s1
10 \pM
0.18 l/M

Therefore, k5 is less than 7.2 x 10-4 s-1. This value is one fourth of that for control.

[I]
K,
k6 k-k5 x S[ATP] [I]
Ka Ki

So k6 must be less than 4.9 x 10-4 s-1. Compared to original value of 1.9 x 10-4 s-1, the change of

k6 cannot significantly affect k value to be changed from 2.3 x 10-3 s-1 to 4.9 x 10-4 S-1. So k5 is

the one of the factors that affect k value significantly.

Since k6 is not a major factor that affect k value, we can neglect the k6 term in the equation

2. Then









[I]
K
k kk x K
1+[ATP] [I]
Ka K,

So k is negatively related to KI value. The increasing of KI results in the decreasing of k.

For DON incubated enzyme, it is more difficult to determine vss and k by fitting the data

measured within 15 minutes. The concentration of the pyrophosphate produced can be expressed

using following equation for slow onset inhibition model.




V0 -V
P =vt+(Ie kt)x 0-
k


When time is long enough, P = vt + v This means the production of pyrophosphate is
k


linearly related to the time with vss as slope and v v as intercept on y-axis. Therefore, to
k

determine vs and k, especially vss, we need to measure the production of pyrophosphate for

enough time. Unfortunately, hASNS is not stable enough for a long time assay.

















H
SY H


Cysi


ATP PPi




Asp
Asp


Figure 6-1.


NH,








NH3


S H

HCys
LNy



Cysi


Tunneling







AMP


NH3 AMP


0

YH
0
Asn


The mechanisms of ASNS catalyzed reaction. At glutaminase site, ammonia is released
from glutamine through a thioester intermediate formed by the nucleophilic attack of the
thiolate group of Cysl residue to the amide group of glutamine. Then the thioester is
hydrolyzed into glutamate and reproduce free enzyme. At the other end of the tunnel, the y-
carboxylic acid group of aspartate is activated by ATP and form the P-aspartyl-AMP
intermediate, followed by ammonia attacking to its y-carbonyl group to produce asparagine.













H 0

21 o H o


S + HS+ +

O CH E-Cysl
I N


6-diazo-5-oxo-L-norleucine 6-(Enzyme-cysteinyl)-5-oxo-L-norl
(DON)


Figure 6-2. DON reacts with Cysl residue of ASNS and forms covalent adduct.


-N


ne





eucine










NHN


O H-N OIasp MP O
I0 I [



0-aspartyl-AMP OH OH


Figure 6-3.


NH2

0 H.C N


+ II 1 /-
NH3 O O

adenylated sulfoximine OH OH


The inhibitor adenylated sulfoximine (right) mimics the nucleophilic attacking of 3-
aspartyl-AMP by ammonia.













120-


100


80-


1000


\ '
.3 )

nC3


2000


3000


4000


time (s)

Figure 6-4. The glutaminase activity of hASNS decreased exponentially with time.
Solid line (data: filled square): 22 C; Dashed line (data: blank square): 37 C..



















"mA



NN


1000


2000


3000


4000


5000


time (s)


Figure 6-5. The ammonia dependent synthetase activity of hASNS decreased with time. Solid line (data:
filled square): 22 C; Dashed line (data: blank square): 37 C.






























0 200 400


800 1000 1200


time (s)


Figure 6-6. The glutaminase activity of DON inhibited hASNS decreased exponentially with time. The
curve was fitted using the first four data.
















--- hAS control
- hAS + 5 uM I
--8--hAS+ 10uM I


0 200 400 600 800 1000


Time (s)


The inhibition to free hASNS by sulfoximine








































133


Figure 6-7.


1200


















-- E control
-- E + 10 uM I

60-


==

S40-






20-




0 100 200 300 400 500 600 700 800

Time (s)

Figure 6-8. Inhibition to control hASNS by sulfoximine

















- EDON
-- EDON +10uM I
--- EDON + 8 uM I
--X--EDON+6uM I
--+--EDON+4uM I
- A-EDON+2 uM I


0 200 400 600 800 1000 1200

time (s)

Figure 6-9. Inhibition to DON inhibited hASNS by sulfoximine












Table 6-1. Percent glutaminase activity at different incubation time without DON


Room Temperature (-22 OC) 37 OC

Time (s) %activity Time (s) %activity
0 100 0 100
630 83.8 300 69.4
1230 76.2 600 69.6
1830 81.2 900 66.3
2430 61.0 1200 51.5
3030 67.1 1500 39.6
3630 57.3 3600 13.4










Table 6-2. Percent ammonia dependent synthetase activity without inhibitors


Room Temperature (-22 OC) 37 OC

Time (s) %activity Time (s) %activity
0 100 0 100
120 99.12864
300 95.05294
600 99.59384 598 89.15113
893 84.74562
1200 99.69538 1201 77.24632
1500 68.41563
1800 99.72076 1800 59.92223
2400 97.56304 2400 49.46501
3000 41.84203
3600 88.95752 3600 33.09285
4200 88.06905










Table 6-3. The loss of glutaminase activity due to DON inhibition


Room Temperature (-22 OC)

Time (s) %activity
0 100
254 45
365 29
630 25
1070 11
1324 13
1883 12










Table 6-4. Synthetase activity of the DON inhibited and control enzyme after
incubation determined using the pyrophosphate assay

Rate of Reaction uM/s
Enzyme Solution
Trial 1 Trial 2
hASNS + 5.88 mM DON 3.99E-02 4.37E-02
hASNS control 4.80E-02 4.60E-02










Table 6-5. Kinetical constant for sulfoximine inhibition to ammonia dependent synthetase activity of
hASNS
data source KI (nM) k5 (s-1) k6 (s 1) k(M-1 l K (nM)
repeat 181 3.51 x 103 2.6 x 104 13.4
literature 285 2.98 x 10-3 2.6 x 10-5 436 2.46










Table 6-6. Parameters for inhibition of stock hASNS and control hASBS by sulfoximine


hASNS KI (nM) k5 (s-') k6 (S-1) Kb/[Itt K* (nM)
stock 181 3.51 x 10-3 2.6 x 10-4 -13.4
control 181* 3.12 x 10-3 1.9 x 10-4 11.0











APPENDIX A
SEQUENCE ALIGNMENT OF TUNNEL RESIDUES IN ASB





Tunnel Residues

1 1 2 3 3 3 3 3 3 3 3 3 3 3 4
2 3 5 7 7 2 4 3 4 4 4 5 5 8 8 8 9 9 9 0
9 1 0 4 6 0 3 8 4 6 8 0 2 0 5 9 3 7 9 1


plant 456 e-133
plant 525 e-153
mustard plant 456 e-125
pea 586 e-160
bacteria 564 e-180
plant 586 e-168
fruit fly 558 e-120
Ecoli
bacteria 554
Salmonella 554
Yersinia 554
Paramecium 588 e-132
fungus 571 e-133
fungus2 574 e-136
Neurospora 581 e-148
sunflower 589 e-163
mosquito 533 e-112
C elegans 567 e-105
hamster 561 e-81
mouse 561 e-81
Human 562 e-80


3 3 5 7 7
1 0 2 2 5 8
33577
102258
CHRGPLINGEY
CHRGPLINGEY
CHRGPLVNGEY
CHRGPLINGEY
CHRGPLINGEY
CHRGPLINGEY
CHRGPLINGEY
CHRGPLINGEY
CHRGPLINGEY
CHRGPLINGEY
CHRGPLINGEY
CHRGPLINGEY
CHRGPLINGEY
CHRGPLINGEY
CHRGPLINGEY
CHRGPLVNGEY
CHRGPLINGEY
CHRGPLINGEY
CHRGPLVNGEY
CHRGPLVNGEY
CHRGPLVNGEY
****** ****


1 2 3 3 3 3 3 3 3 3
4 3 2 4 4 4 5 5 8 8
2 2 9 5 7 9 1 7 4 8
2295791748.
TLDMV SGEGSDEYLDCAN
TLDMV SGEGSDEYLDCAN
TLDMV SGEGADEYLDCAN
TLDMV SGEGSDEYLDCAN
LDMV SGEGSDEYLDCAN
CLDMV SGEGSDEYLDCAN
ILDMI SGEGADEYLDCAN
ILDMV SGEGSDEYLDCAN
ILDMV SGEGSDEYLDCAN
ILDMV SGEGSDEYLDCAN
ILDMV SGEGSDEYLDCAN
IFDMC TGEGSDEYLDLAN
TLDMV SGEGSDEYLDCAN
TLDMV SGEGSDEYLDCAN
TLDMV SGEGSDEYLDCAN
TLDMV SGEGSDEYLDCAN
LD VSGEGADEYLDCAN
KLDMV SGEGADEYLDCAD
RLDMI SGEGSDEYLDVAD
RLDMI SGEGSDEYLDVAD
RLDMI SGEGSDEYLDVAD
*** **** ***** *


Red asterisk shows absolutely conserved residue; black asterisk shows those residues that
conserved in 20 out 21 species.


Red letter:


conserved residue in at least 20 out of 21 species. Blue background letter


show different residue(s) other than conserved residues
Yellow column: conserved residue in at least 15 out of 21 species.
Purple column: conserved residue in at least 10 out of 21 species.


Note: 1. An absolutely conserved Gly is right before Met 120 and Ile 142.
2. Ile 52 (16 I, 5 V), Met 120 (12 M, 5 I, 4 V), Ile 142 (11 I, 10 V)
Asn 389 (17 N, 4 D), Val 397 (12 L, 8 V, 1 I), Cys 385 (17 C, 3 V, 1 L)
Val 344 (16 V, 4 I, 1 C), Ser 350 (17 S, 4 A)


SPECIES


4 4
0 0
0 4
RVL
RVL
RVL
RVL
RVL
RVL
RVL
RVL
RVL
RVL
RVL
RPM
RVL
RVL
RVL
RVL
RVL
RVL
RVL
RVL
RVL









APPENDIX B
THE KINETIC MODEL FOR EXCHANGE EXPERIMENT



The model for exchanging experiment was analyzed as following:

E + Gin Eester + NH4 (i)
k2

15 + 15
Eester + NH4 E + N-GIn (ii)
ki

k3
Eester + H20 k -E + Glu (iii)

Since [Gln] >> [1N-Gln] and [15NH4+] >> [NH4 ], the reverse reaction of(i) and (ii) are

negligible. So the model is simplified as:

k, +
E + Gin Eester + NH4 (I)


15 + k215
Eester + NH4 E + 1N-GIn (II)


k3
Eester + H20 k -E + Glu (III)

Based on the steady state theory,

d[Eester]
= k, [E][Gln] k[Eester] [NH] k3[Eester][H20]= 0
dt

k, [E][Gn] = k2 [Eester] [5NH4 ] + k3 [Eester][H20]

k, [E] [G n] = [Eester](k2 [5NH4 ] + k3 [H20])

[Eester](k2 [NH ]+ k3[H20])
[E] =
k, [G In]









Since [E], = [E] + [Eester]


[Eester](k [15NH4 ] + k3 [H20])
Then [E], 4 + [Eester]
k, [G n]

[Eester](k, [G n] + k2 [15NH4 ] + k3[H20])
k,[Gln]

Therefore,

Eester] k, [E], [G n]
[Eester] =
(k, [G In] + k2 [15NH4 ] + k3 [H20])

So the production rate of [15N-GIn] is:

kk 2[E], [GIn][4NHJ]
v = k2[Eester][NH] = [NH
(k, [G n] + k2 [15NH + k3[H20])

When [15NH4+] is kept constant and [Gin] is varied,

(k2 [E], ['1NH4 ])[G In]
k k
(-[ 15NH ] + 3 [HO])+ [G n]


Vmax-Gln[Gln]
KMGln + [G n]

Where

VmaxGln= k2 [E]T [15NH4]

KM = 2[ 15NH]+3 [H2O]


When [Gin] is kept constant and [15NH4] is varied,

S(k, [E], [G n])[1NH ]
k k 0])+[NH
( [GIn]+ 3[H20])+[5NH,]
k2 k2









Vmax 15NH4 15NH4]
K -15NH+ +[]SNH4]

Where

Vmax ; = k2 [E], [G In]

k k
KM1-' H = ["1NH4]+ k[H20]
SNH k2 k









LIST OF REFERENCES


1. Hyde, C. C.; Ahmed, S. A.; Padlan, E. A.; Miles, E. W.; Davies, D. R., Three-Dimensional
Structure of the Tryptophan Synthase R2.2, Multienzyme Complex from Salmonella
typhimurium. JBiol Chem 1988, 263, 17857-71.

2. Huang, X.; Holden, H. M.; Raushel, F. M., Channeling of substrates and intermediates in
enzyme-catalyzed reactions. Annu Rev Biochem 2001, 70, 149-80.

3. Raushel, F. M.; Thoden, J. B.; Holden, H. M., Enzymes with molecular tunnels. Ace Chem Res
2003, 36, (7), 539-48.

4. Lunn, F. A.; Beame, S. L., Alternative substrates for wild-type and L109A E. coli CTP
synthases -- Kinetic evidence for a constricted ammonia tunnel. Eur JBiochem 2004, 271, 4204-
12.

5. Zalkin, H., The amidotransferases. Adv Enzymol Relat Areas Mol Biol 1993, 66, 203-309.

6. Zalkin, H.; Smith, J. L., Enzymes utilizing glutamine as an amide donor. Adv Enzymol Relat
Areas Mol Biol 1998, 72, 87-144.

7. Haskell, C. M.; Canellos, G. P., 1-asparaginase resistance in human leukemia--asparagine
synthetase. Biochem Pharmacol 1969, 18, (10), 2578-80.

8. Prager, M. D.; Peters, P. C.; Janes, J. O.; Derr, I., Asparagine synthetase activity in malignant
and non-malignant human kidney and prostate specimens. Nature 1969, 221, (185), 1064-5.

9. Nakamura, M.; Yamada, M.; Hirota, Y.; Sugimoto, K.; Oka, A.; Takanami, M., Nucleotide
sequence of the asnA gene coding for asparagine synthetase of E. coli K-12. Nucleic Acids Res
1981, 9, (18), 4669-76.

10. Felton, J.; Michaelis, S.; Wright, A., Mutations in two unlinked genes are required to produce
asparagine auxotrophy in Escherichia coli. JBacteriol 1980, 142, (1), 221-8.

11. Scofield, M. A.; Lewis, W. S.; Schuster, S. M., Nucleotide sequence of Escherichia coli asnB
and deduced amino acid sequence of asparagine synthetase B. J Biol Chem 1990, 265, (22),
12895-902.

12. Humbert, R.; Simoni, R. D., Genetic and biomedical studies demonstrating a second gene
coding for asparagine synthetase in Escherichia coli. JBacteriol 1980, 142, (1), 212-20.

13. Richards, N. G.; Schuster, S. M., Mechanistic issues in asparagine synthetase catalysis. Adv
Enzymol Relat Areas Mol Biol 1998, 72, 145-98.

14. Schnizer, H. G.; Boehlein, S. K.; Stewart, J. D.; Richards, N. G.; Schuster, S. M., Formation
and isolation of a covalent intermediate during the glutaminase reaction of a class II
amidotransferase. Biochemistry 1999, 38, (12), 3677-82.









15. Larsen, T. M.; Boehlein, S. K.; Schuster, S. M.; Richards, N. G.; Thoden, J. B.; Holden, H.
M.; Rayment, I., Three-dimensional structure of Escherichia coli asparagine synthetase B: a short
journey from substrate to product. Biochemistry 1999, 38, (49), 16146-57.

16. Boehlein, S. K.; Richards, N. G.; Schuster, S. M., Glutamine-dependent nitrogen transfer in
Escherichia coli asparagine synthetase B. Searching for the catalytic triad. JBiol Chem 1994a,
269, (10), 7450-7.

17. Boehlein, S. K.; Richards, N. G.; Walworth, E. S.; Schuster, S. M., Arginine 30 and
asparagine 74 have functional roles in the glutamine dependent activities of Escherichia coli
asparagine synthetase B. JBiol Chem 1994b, 269, (43), 26789-95.

18. Boehlein, S. K.; Schuster, S. M.; Richards, N. G., Glutamic acid gamma-monohydroxamate
and hydroxylamine are alternate substrates for Escherichia coli asparagine synthetase B.
Biochemistry 1996, 35, (9), 3031-7.

19. Stoker, P. W.; O'Leary, M. H.; Boehlein, S. K.; Schuster, S. M.; Richards, N. G., Probing the
mechanism of nitrogen transfer in Escherichia coli asparagine synthetase by using heavy atom
isotope effects. Biochemistry 1996, 35, (9), 3024-30.

20. Boehlein, S. K.; Walworth, E. S.; Richards, N. G.; Schuster, S. M., Mutagenesis and
chemical rescue indicate residues involved in beta-aspartyl-AMP formation by Escherichia coli
asparagine synthetase B. JBiol Chem 1997, 272, (19), 12384-92.

21. Boehlein, S. K.; Walworth, E. S.; Schuster, S. M., Identification of cysteine-523 in the
aspartate binding site of Escherichia coli asparagine synthetase B. Biochemistry 1997, 36, (33),
10168-77.

22. Boehlein, S. K.; Stewart, J. D.; Walworth, E. S.; Thirumoorthy, R.; Richards, N. G.; Schuster,
S. M., Kinetic mechanism of Escherichia coli asparagine synthetase B. Biochemistry 1998, 37,
(38), 13230-8.

23. Tesson, A. R.; Soper, T. S.; Ciustea, M.; Richards, N. G., Revisiting the steady state kinetic
mechanism of glutamine-dependent asparagine synthetase from Escherichia coli. Arch Biochem
Biophys 2003, 413, (1), 23-31.

24. Fresquet, V.; Thoden, J. B.; Holden, H. M.; Raushel, F. M., Kinetic mechanism of asparagine
synthetase from Vibrio cholerae. Bioorg Chem 2004, 32, (2), 63-75.

25. Buchanan, J. M., The amidotransferases. Adv Enzymol Relat Areas Mol Biol 1973, 39, 91-
183.

26. Nakamura, A.; Yao, M.; Chimnaronk, S.; Sakai, N.; Tanaka, I., Ammonia channel couples
glutaminase with transamidase reactions in GatCAB. Science 2006, 312, (5782), 1954-8.

27. Oshikane, H.; Sheppard, K.; Fukai, S.; Nakamura, Y.; Ishitani, R.; Numata, T.; Sherrer, R. L.;
Feng, L.; Schmitt, E.; Panvert, M.; Blanquet, S.; Mechulam, Y.; Soll, D.; Nureki, O., Structural









basis of RNA-dependent recruitment of glutamine to the genetic code. Science 2006, 312, (5782),
1950-4.

28. Mouilleron, S.; Badet-Denisot, M. A.; Golinelli-Pimpaneau, B., Glutamine binding opens the
ammonia channel and activates glucosamine-6P synthase. JBiol Chem 2006, 281, (7), 4404-12.

29. Anand, R.; Hoskins, A. A.; Stubbe, J.; Ealick, S. E., Domain organization of Salmonella
typhimurium formylglycinamide ribonucleotide amidotransferase revealed by X-ray
crystallography. Biochemistry 2004, 43, (32), 10328-42.

30. Endrizzi, J. A.; Kim, H.; Anderson, P. M.; Baldwin, E. P., Crystal structure of Escherichia
coli cytidine triphosphate synthetase, a nucleotide-regulated glutamine amidotransferase/ATP-
dependent amidoligase fusion protein and homologue of anticancer and antiparasitic drug targets.
Biochemistry 2004, 43, (21), 6447-63.

31. Myers, R. S.; Jensen, J. R.; Deras, I. L.; Smith, J. L.; Davisson, V. J., Substrate-induced
changes in the ammonia channel for imidazole glycerol phosphate synthase. Biochemistry 2003,
42, (23), 7013-22.

32. van den Heuvel, R. H.; Ferrari, D.; Bossi, R. T.; Ravasio, S.; Curti, B.; Vanoni, M. A.;
Florencio, F. J.; Mattevi, A., Structural studies on the synchronization of catalytic centers in
glutamate synthase. JBiol Chem 2002, 277, (27), 24579-83.

33. Douangamath, A.; Walker, M.; Beismann-Driemeyer, S.; Vega-Fernandez, M. C.; Sterner, R.;
Wilmanns, M., Structural evidence for ammonia tunneling across the (beta alpha)(8) barrel of the
imidazole glycerol phosphate synthase bienzyme complex. Structure 2002, 10, (2), 185-93.

34. Raushel, F. M.; Thoden, J. B.; Holden, H. M., The amidotransferase family of enzymes:
molecular machines for the production and delivery of ammonia. Biochemistry 1999, 38, (25),
7891-9.

35. Richards, N. G.; Kilberg, M. S., Asparagine synthetase chemotherapy. Annu Rev Biochem
2006, 75, 629-54.

36. Binda, C.; Bossi, R. T.; Wakatsuki, S.; Arzt, S.; Coda, A.; Curti, B.; Vanoni, M. A.; Mattevi,
A., Cross-talk and ammonia channeling between active centers in the unexpected domain
arrangement of glutamate synthase. Structure 2000, 8, (12), 1299-308.

37. Massiere, F.; Badet-Denisot, M. A., The mechanism of glutamine-dependent
amidotransferases. CellMolLife Sci 1998, 54, (3), 205-22.

38. Weeks, A.; Lund, L.; Raushel, F. M., Tunneling of intermediates in enzyme-catalyzed
reactions. Curr Opin Chem Biol 2006, 10, (5), 465-72.

39. Mullins, L.; Raushel, F., Channeling of ammonia through the intermolecular tunnel contained
within carbamoyl phosphate synthetase. JOURNAL OF THE AMERICAN CHEMICAL SOCIETY
1999, 121, (15), 3803-3804.









40. Myers, R. S.; Amaro, R. E.; Luthey-Schulten, Z. A.; Davisson, V. J., Reaction coupling
through interdomain contacts in imidazole glycerol phosphate synthase. Biochemistry 2005, 44,
(36), 11974-85.

41. Willemoes, M., Competition between ammonia derived from internal glutamine hydrolysis
and hydroxylamine present in the solution for incorporation into UTP as catalysed by
Lactococcus lactis CTP synthase. Arch Biochem Biophys 2004, 424, (1), 105-11.

42. Bera, A. K.; Chen, S.; Smith, J. L.; Zalkin, H., Temperature-dependent function of the
glutamine phosphoribosylpyrophosphate amidotransferase ammonia channel and coupling with
glycinamide ribonucleotide synthetase in a hyperthermophile. JBacteriol 2000, 182, (13), 3734-
9.

43. Miles, B. W.; Raushel, F. M., Synchronization of the three reaction centers within carbamoyl
phosphate synthetase. Biochemistry 2000, 39, (17), 5051-6.

44. Kim, J.; Raushel, F. M., Perforation of the tunnel wall in carbamoyl phosphate synthetase
derails the passage of ammonia between sequential active sites. Biochemistry 2004, 43, (18),
5334-40.

45. Thoden, J. B.; Huang, X.; Raushel, F. M.; Holden, H. M., Carbamoyl-phosphate synthetase.
Creation of an escape route for ammonia. JBiol Chem 2002, 277, (42), 39722-7.

46. Chaudhuri, B. N.; Lange, S. C.; Myers, R. S.; Davisson, V. J.; Smith, J. L., Toward
understanding the mechanism of the complex cyclization reaction catalyzed by imidazole
glycerolphosphate synthase: crystal structures of a ternary complex and the free enzyme.
Biochemistry 2003, 42, (23), 7003-12.

47. Amaro, R. E.; Myers, R. S.; Davisson, V. J.; Luthey-Schulten, Z. A., Structural elements in
IGP synthase exclude water to optimize ammonia transfer. Biophys J 2005, 89, (1), 475-87.

48. Luthey-Schulten, Z. A.; Amaro, R., Molecular dynamics simulations of substrate channeling
through an a/1 barrel protein. Chem. Phys. 2004, 307, 147-155.

49. Amaro, R.; Tajkhorshid, E.; Luthey-Schulten, Z., Developing an energy landscape for the
novel function of a (beta/alpha)8 barrel: ammonia conduction through HisF. Proc Natl Acad Sci
USA 2003, 100, (13), 7599-604.

50. Smith, J. L., Structures of glutamine amidotransferases from the purine biosynthetic pathway.
Biochem Soc Trans 1995, 23, (4), 894-8.

51. Brannigan, J. A.; Dodson, G.; Duggleby, H. J.; Moody, P. C.; Smith, J. L.; Tomchick, D. R.;
Murzin, A. G., A protein catalytic framework with an N-terminal nucleophile is capable of self-
activation. Nature 1995, 378, (6555), 416-9.

52. Bieganowski, P.; Pace, H. C.; Brenner, C., Eukaryotic NAD+ synthetase Qnsl contains an
essential, obligate intramolecular thiol glutamine amidotransferase domain related to nitrilase. J
Biol Chem 2003, 278, (35), 33049-55.









53. Sheng, S.; Moraga-Amador, D. A.; van Heeke, G.; Allison, R. D.; Richards, N. G.; Schuster,
S. M., Glutamine inhibits the ammonia-dependent activities of two Cys-1 mutants of human
asparagine synthetase through the formation of an abortive complex. J Biol Chem 1993, 268,
(22), 16771-80.

54. Kim, J. H.; Krahn, J. M.; Tomchick, D. R.; Smith, J. L.; Zalkin, H., Structure and function of
the glutamine phosphoribosylpyrophosphate amidotransferase glutamine site and communication
with the phosphoribosylpyrophosphate site. JBiol Chem 1996, 271, (26), 15549-57.

55. Badet, B.; Vermoote, P.; Haumont, P. Y.; Lederer, F.; LeGoffic, F., Glucosamine synthetase
from Escherichia coli: purification, properties, and glutamine-utilizing site location.
Biochemistry 1987, 26, (7), 1940-8.

56. van den Heuvel, R. H.; Curti, B.; Vanoni, M. A.; Mattevi, A., Glutamate synthase: a
fascinating pathway from L-glutamine to L-glutamate. CellMol Life Sci 2004, 61, (6), 669-81.

57. Ciustea, M.; Gutierrez, J. A.; Abbatiello, S. E.; Eyler, J. R.; Richards, N. G., Efficient
expression, purification, and characterization of C-terminally tagged, recombinant human
asparagine synthetase. Arch Biochem Biophys 2005, 440, (1), 18-27.

58. Hurd, R. E.; John, B. K., Gradient-enhanced proton-detected heteronuclear multiple-quantum
coherence spectroscopy. J. Magn. Reson. 1991, 91, 648-653.

59. Schnizer, H. G.; Boehlein, S. K.; Stewart, J. D.; Richards, N. G.; Schuster, S. M., gamma-
Glutamyl thioester intermediate in glutaminase reaction catalyzed by Escherichia coli asparagine
synthetase B. Methods Enzymol 2002, 354, 260-71.

60. Bradford, M. M., A rapid and sensitive method for the quantitation of microgram quantities
of protein utilizing the principle of protein-dye binding. Anal Biochem 1976, 72, 248-54.

61. Mendes, P., Biochemistry by numbers: simulation of biochemical pathways with Gepasi 3.
Trends Biochem Sci 1997, 22, (9), 361-3.

62. Mendes, P., GEPASI: a software package for modelling the dynamics, steady states and
control of biochemical and other systems. Comput Appl Biosci 1993, 9, (5), 563-71.

63. Beame, S. L.; Hekmat, O.; Macdonnell, J. E., Inhibition of Escherichia coli CTP synthase by
glutamate gamma-semialdehyde and the role of the allosteric effector GTP in glutamine
hydrolysis. Biochem J2001, 356, (Pt 1), 223-32.

64. Levitzki, A.; Koshland, D. E., Jr., Cytidine triphosphate synthetase. Covalent intermediates
and mechanisms of action. Biochemistry 1971, 10, (18), 3365-71.

65. Zalkin, H.; Truitt, C. D., Characterization of the glutamine site of Escherichia coli guanosine
5'-monophosphate synthetase. JBiol Chem 1977, 252, (15), 5431-6.









66. Dossena, L.; Curti, B.; Vanoni, M. A., Activation and coupling of the glutaminase and
synthase reaction of glutamate synthase is mediated by E1013 of the ferredoxin-dependent
enzyme, belonging to loop 4 of the synthase domain. Biochemistry 2007, 46, (15), 4473-85.

67. Isupov, M. N.; Obmolova, G.; Butterworth, S.; Badet-Denisot, M. A.; Badet, B.; Polikarpov,
I.; Littlechild, J. A.; Teplyakov, A., Substrate binding is required for assembly of the active
conformation of the catalytic site in Ntn amidotransferases: evidence from the 1.8 A crystal
structure of the glutaminase domain of glucosamine 6-phosphate synthase. Structure 1996, 4, (7),
801-10.

68. Vanoni, M. A.; Curti, B., Structure--function studies on the iron-sulfur flavoenzyme
glutamate synthase: an unexpectedly complex self-regulated enzyme. Arch Biochem Biophys
2005, 433, (1), 193-211.

69. Su, N.; Pan, Y. X.; Zhou, M.; Harvey, R. C.; Hunger, S. P.; Kilberg, M. S., Correlation
between asparaginase sensitivity and asparagine synthetase protein content, but not mRNA, in
acute lymphoblastic leukemia cell lines. Pediatr Blood Cancer 2007.

70. Cui, H.; Darmanin, S.; Natsuisaka, M.; Kondo, T.; Asaka, M.; Shindoh, M.; Higashino, F.;
Hamuro, J.; Okada, F.; Kobayashi, M.; Nakagawa, K.; Koide, H.; Kobayashi, M., Enhanced
expression of asparagine synthetase under glucose-deprived conditions protects pancreatic
cancer cells from apoptosis induced by glucose deprivation and cisplatin. Cancer Res 2007, 67,
(7), 3345-55.

71. Gutierrez, J. A.; Pan, Y. X.; Koroniak, L.; Hiratake, J.; Kilberg, M. S.; Richards, N. G., An
inhibitor of human asparagine synthetase suppresses proliferation of an L-asparaginase-resistant
leukemia cell line. Chem Biol 2006, 13, (12), 1339-47.

72. Horowitz, B.; Meister, A., Glutamine-dependent asparagine synthetase from leukemia cells.
Chloride dependence, mechanism of action, and inhibition. J Biol Chem 1972, 247, (20), 6708-
19.

73. Gong, S. S.; Guerrini, L.; Basilico, C., Regulation of asparagine synthetase gene expression
by amino acid starvation. Mol Cell Biol 1991, 11, (12), 6059-66.

74. Cedar, H.; Schwartz, J. H., The asparagine synthetase of Escherhic coli. I. Biosynthetic role
of the enzyme, purification, and characterization of the reaction products. JBiol Chem 1969, 244,
(15), 4112-21.

75. Cedar, H.; Schwartz, J. H., The asparagine synthetase of Escherichia coli. II. Studies on
mechanism. JBiol Chem 1969, 244, (15), 4122-7.

76. Herrera-Rodriguez, M. B.; Perez-Vicente, R.; Maldonado, J. M., Expression of asparagine
synthetase genes in sunflower (Helianthus annuus) under various environmental stresses. Plant
Physiol Biochem 2007, 45, (1), 33-8.









77. Herrera-Rodriguez, M. B.; Maldonado, J. M.; Perez-Vicente, R., Role of asparagine and
asparagine synthetase genes in sunflower (Helianthus annuus) germination and natural
senescence. JPlant Physiol 2006, 163, (10), 1061-70.

78. Herrera-Rodriguez, M. B.; Maldonado, J. M.; Perez-Vicente, R., Light and metabolic
regulation of HAS 1, HAS 1.1 and HAS2, three asparagine synthetase genes in Helianthus annuus.
Plant Physiol Biochem 2004, 42, (6), 511-8.

79. Herrera-Rodriguez, M. B.; Carrasco-Ballesteros, S.; Maldonado, J. M.; Pineda, M.; Aguilar,
M.; Perez-Vicente, R., Three genes showing distinct regulatory patterns encode the asparagine
synthetase of sunflower (Helianthus annuus). New Phytologist 2002, 155, 33-45.

80. Koroniak, L.; Ciustea, M.; Gutierrez, J. A.; Richards, N. G., Synthesis and characterization of
an N-acylsulfonamide inhibitor of human asparagine synthetase. Org Lett 2003, 5, (12), 2033-6.

81. Krahn, J. M.; Kim, J. H.; Burns, M. R.; Parry, R. J.; Zalkin, H.; Smith, J. L., Coupled
formation of an amidotransferase interdomain ammonia channel and a phosphoribosyltransferase
active site. Biochemistry 1997, 36, (37), 11061-8.

82. van den Heuvel, R. H.; Svergun, D. I.; Petoukhov, M. V.; Coda, A.; Curti, B.; Ravasio, S.;
Vanoni, M. A.; Mattevi, A., The active conformation of glutamate synthase and its binding to
ferredoxin. JMolBiol 2003, 330, (1), 113-28











BIOGRAPHICAL SKETCH

Kai Li was born in Zouping, China on August 2, 1974. After he got his Bachelor of

Science degree in biochemistry from Nankai University in Tianjin, China, he went to Beijing,

China, and became a graduate student in Plant medicine in China Academy of Agriculture

Sciences (CAAS), where he got his Master of Science degree. From 1998 to 2001, he worked in

China Institute of Veterinary Drug Control (IVDC) and did residue analysis of veterinary drugs

in food such as meat, egg and milk. In September 2002, he was enrolled in graduate school of

State University of New York at Albany. Due to family reason, he transferred to University of

Florida after being in Albany for about 9 months. From 2003 to 2007, he studied asparagine

synthetase under the supervision of Dr. Nigel G. J. Richards.





PAGE 1

INTRAMOLECULAR TUNNEL AND REGULATORY MECHANISMS OF ASPARAGINE SYNTHETASE (ASNS) By Kai Li A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2007 Kai Li

PAGE 3

To my wife, Zhen, and my parents

PAGE 4

4 ACKNOWLEDGMENTS I acknowledge the financial support for the travel grants from College of Liberal Arts and Sciences and Graduate Student Council, which made it possible to present my work at several conferences. I thank Dr. Nigel G. J. Richards, my superv isor, who guided me into enzymology research field by working on asparagine synthetase. I am grateful to my doctoral dissertation committee, Dr. Nicole A. Horenstein, Dr. Dani el L. Purich, Dr. Adrian E. Ro itberg, and Dr. Jon D. Stewart, for their helpful advices on my project. Furtherm ore, I would like to thank Dr. Gail E. Fanucci, Dr. Ion Ghiviriga, Dr. Thomas J. Lyons and Dr. David N. Silverman for their helps. Special thanks go to Dr. Drazenka Svedruzic, Dr. Jemy A. Gutierrez, Dr. Tania Crdova de Sintjago and Dr. Patricia Moussatche, for their helpful discussions with my projects; William Beeson and Alexandria Berry, two undergraduates, with whom I accomplished this work; Xiao He, Dr. Hui Jiang, Sangbae Lee, Mario E. Mora l, and Dr. Yong Ran, for their friendship and helps. I am grateful to my mom and stepfather fo r their unconditional suppo rt of any decision I made, to my parents-in-law for their care by send ing me email everyday and to my wife for the happy life we shared for almost ten years.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4 LIST OF TABLES ...........................................................................................................................8 LIST OF FIGURES .........................................................................................................................9 ABSTRACT ...................................................................................................................... .............12 CHAPTER 1 INTRODUCTION .................................................................................................................. 14 Asparagine Synthetase (ASNS) and the Intramolecular Tunnel ............................................ 14 The Intramolecular Tunnel ..................................................................................................... 17 Glutaminase Active Site and the Mouth of the Tunnel ...................................................17 The Body of the Tunnel ...................................................................................................18 The Bottom of the Tunnel ............................................................................................... 19 The Interface of the Two Domains and Water ....................................................................... 19 2 A gHMQC-BASED NMR ASSAY FOR INVESTIGATING AMMONIA CHANNELING IN GLUTAMINE DEPE NDENT AMIDOTRANSFERASES ................... 35 Introduction .................................................................................................................. ...........35 Material and Methods .............................................................................................................37 Materials ..................................................................................................................... .....37 NMR Measurements ........................................................................................................ 38 Competition Experiment ................................................................................................. 38 15N Exchanging Experiment ............................................................................................ 39 HPLC-Based Determination of Total L-Asparagine ....................................................... 40 Kinetic Simulations .........................................................................................................41 Results and Discussion ........................................................................................................ ...41 NMR Assay .....................................................................................................................42 Competition Experiments ................................................................................................ 42 15N exchange Experiments .............................................................................................. 46 Conclusions .............................................................................................................................48 3 CHARACTERIZATION OF A TUNNEL MUTANT ........................................................... 61 Introduction .................................................................................................................. ...........61 Material and Methods .............................................................................................................62 Materials ..................................................................................................................... .....62 Competition Experiment ................................................................................................. 63 NMR Measurements and HPLC assay ............................................................................64

PAGE 6

6 Results and Discussion ........................................................................................................ ...64 Conclusions .............................................................................................................................65 4 ASPARAGINE INHIBITI ON TO ASPARAGINE SYNT HETASES AND ITS EFFECT ON THE QUATERNARY ST RUCTURE OF THE ENZYMES ........................... 74 Introduction ............................................................................................................................74 Materials and Methods ...........................................................................................................75 Materials ..................................................................................................................... .....75 Enzyme Assays ................................................................................................................ 75 Asparagine Inhibition Studies ......................................................................................... 76 Size Exclusion Chromatography to Dete rmine the Quaternary Structure of Asparagine Synthestases. ............................................................................................. 76 The Effect of Asparagine on the Quat ernary Structure of ASB and hAS ....................... 76 SEC Studies of ASB with Different Amount of ASB ..................................................... 77 Results .....................................................................................................................................77 Product Inhibition ............................................................................................................ 77 Asparagine Affects the Quaternary Stru cture of ASB, But Not Huma n Enzyme ........... 77 The Equilibrium between the Dimer and Monomer Form of ASB. ................................ 79 Construction of Aspara gine Inhibition Model ................................................................. 80 Discussion .................................................................................................................... ...........83 Conclusions .............................................................................................................................86 5 THE UTILIZATION OF DIFFERENT NI TROGEN S OURCES BY ASPARAGINE SYNTHETASE: EVOLUTION FROM Escherichia coli TO HUMAN .............................103 Introduction .................................................................................................................. .........103 Material and Methods ...........................................................................................................105 Materials ..................................................................................................................... ...105 Competition Experiment ............................................................................................... 105 NMR Measurements, HPLC assay and Kinetic Simulations ........................................ 105 Results and Discussion ........................................................................................................ .106 Conclusions ...........................................................................................................................109 6 PREPARATION OF LIGAND BOUND hA SNS INHIBIITION AND BINDING STUDIES ....................................................................................................................... .......115 Introduction ..........................................................................................................................115 Materials and Methods .........................................................................................................116 Materials ..................................................................................................................... ...116 Enzyme Assays .............................................................................................................. 116 Loss of Glutaminase Activity with Time at Different Temperature ............................. 117 DON Inhibition ..............................................................................................................117 Loss of Synthetase Activity with Time at different Temperature ................................. 118 Inhibition of the DON Inhibited Enzyme by Sulfoximine ............................................ 118 Results ...................................................................................................................................118 Stability Experiment ...................................................................................................... 118

PAGE 7

7 hASNS lost glutaminase activity with time at room temperature and 37 oC .........119 The loss of NH3 dependent synthetase activity with time at room temperature and 37 oC ............................................................................................................. 119 Inhibition to hASNS by DON ....................................................................................... 120 DON inhibits 90% of glutaminase activ ity of hASNS after incubating for 20 mi nutes. ............................................................................................................... 120 DON had little effect on the rate of ammonia dependent synthesis of asparagine. .......................................................................................................... 120 Inhibition by Adenylated Sulfoximine ..........................................................................121 Inhibition to the ammonia dependent synthetase activity of stock hASNS ........... 121 Inhibition to the ammonia dependent synthetase activity of DON incubated hASNS ................................................................................................................ 122 Discussion .................................................................................................................... .........123 APPENDIX A SEQUENCE ALIGNMENT OF TUNNEL RESIDUES IN ASB ....................................... 142 B THE KINETIC MODEL FOR EXCHANGE EXPERIMENT ............................................ 143 REFERENCE LIST .....................................................................................................................146 BIOGRAPHICAL SKETCH .......................................................................................................153

PAGE 8

8 LIST OF TABLES Table page 1-1 Residues in Gln binding pocket and possible roles ...........................................................31 1-2 Size of narrow part of the tunnel........................................................................................ 32 1-3 Residues in the interface between two domains of E. coli ASB ........................................ 33 2-1 AS-B catalyzed incorporation of 15N into L-asparagine in the steady-state competition assays ............................................................................................................ .60 3-1 The kinetic parameters for glutaminas e activity, ammonia dependent synthetase activity and glutamine depende nt synthetase activity of both wild type AS B and A388L mutant .................................................................................................................. ..73 4-1 Retention time of ASB peaks in SEC studies .................................................................... 96 4-2 The calculated MW and predicted struct ure of ASB corresponding to each peak ............97 5-1 hASNS catalyzed production of 15N into L-asparagine in the steady-state competition assays ...............................................................................................................................114 6-1 Percent glutaminase activity at di fferent incubation time without DON ......................... 136 6-2 Percent ammonia dependent synthe tase activity without inhibitors ................................ 137 6-3 The loss of glutaminase ac tivity due to DON inhibition ................................................. 138 6-4 Synthetase activity of the DON inhibited and control enzyme after incubation determined using the pyrophosphate assay ...................................................................... 139 6-5 Kinetical constant for sulfoximine i nhibition to ammonia dependent synthetase activity of hASNS ............................................................................................................140 6-6 Parameters for inhibition of stock hASNS and control hASBS by sulfoximine ............. 141

PAGE 9

9 LIST OF FIGURES Figure page 1-1 Proposed mechanism of hydrolysis of glutamine by E. coli asparagine synthetase B. ..... 21 1-2 Synthesis of asparagines through an intermediate -aspartyl-adenylate by ASB. ............. 22 1-3 Cartoon of glutaminase domain of E. coli ASB. Glutamine is s howed by spacefill. Ala1(Cys1) is showed by ball and stick. ............................................................................ 23 1-4 The tunnel within ASB. .................................................................................................... .24 1-5 Cys1-Arg-Gly-Asn-Asp is a common mo tif of the glutamine binding site f or Ntn subfamily of glutamine dependent amidotransferases ....................................................... 25 1-6 The mouth of the tunnel. .................................................................................................. ..26 1-7 The narrowest part of the tunnel formed by M120, I143, N389, A399 and carbonyl group of A388. ................................................................................................................ ...27 1-8 The tunnel in E. coli ASB .................................................................................................. 28 1-9 The body of the tunnel .......................................................................................................29 1-10 The Glu352 ............................................................................................................... .........30 2-1 Two possible pathways for the ammonia transfer ............................................................. 50 2-2 gHMQC 1H-NMR spectrum. .............................................................................................51 2-3 Standard curves ........................................................................................................... .......52 2-4 14N/15N incorporation ratios in L-asparagine formed by the AS-B synthetase reaction under competition assay conditions ...................................................................................55 2-5 Normal 1H-NMR spectrum ................................................................................................56 2-6 Glutamine-dependence of the exchange of 15NH3 into L-glutamine under exchange assay conditions .................................................................................................................57 2-7 Exchange of 15NH3 into L-glutamine during the gl utaminase reaction catalyzed by AS-B .......................................................................................................................... ........58 3-1 One of the conserved tunnel residues, Ala388................................................................... 67 3-2 Competition results for A388L AS B mu tant, as well as WT ASB .................................... 68

PAGE 10

10 3-3 Production of asparagine catalyzed by A388L ASB mutant and W T ASB with varied glutamine...................................................................................................................... ......69 3-4 Production of asparagine catalyzed by A388L ASB mutant and W T ASB with varied ammonium ...................................................................................................................... ...70 3-5 The nitrogen exchange catalyzed by A388L mut ant and WT ASB ................................... 71 3-6 The polarity of the tunnel residues .....................................................................................72 4-1 Kinetic studies of aspara gine inhibition to ASNS ............................................................. 90 4-2 SEC standard curve ............................................................................................................93 4-3 SEC analysis of quaternary structure of ASNS with different concentration of asparagine in mobile phase ................................................................................................94 4-4 SEC analysis of quaternary structure of ASB with different amount of enzyme .............. 98 4-5 Fraction of dimer vs con centration of asparagine ............................................................... 99 4-6 The interface of the two subunits ..................................................................................... 100 4-7 Sequence alignment of the interface residues .................................................................. 101 4-8 The glutaminase activity of ASNS around physiological conditions .............................. 102 5-1 Quantification of 15N-asparagine and 14N-asparagine using NMR assay and HPLC assay ......................................................................................................................... ........110 5-2 Simulation model for the competition reactions catalyzed by hASNS ............................ 111 5-3 Simulations of competition r eactions catalyzed by hASNS ............................................ 112 5-4 Utilization of ammonia and glutamine as nitrogen source by ASB and hASNS ............. 113 6-1 The mechanisms of AS NS catalyzed reaction .................................................................127 6-2 DON reacts with Cys1 residue of ASNS and forms covalent adduct ..............................128 6-3 The inhibitor adenylated sulfoximi ne (right) mimics the nucleophilic attacking of aspartyl-AMP by ammonia ..............................................................................................129 6-4 The glutaminase activity of hASN S decreased exponen tially with time ......................... 130 6-5 The ammonia dependent s ynthetase activity of hASNS decreased with time .................131 6-6 The glutaminase activity of DON inhi bited hASNS decreased exponentially with time .......................................................................................................................... ........132

PAGE 11

11 6-7 The inhibition to free hASNS by sulfoximine .................................................................133 6-8 Inhibition to contro l hASNS by sulfoximine ................................................................... 134 6-9 Inhibition to DON inhib ited hASNS by sulfoximi ne ...................................................... 135

PAGE 12

12 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE INTRAMOLECULAR TUNNEL AND REGULATORY MECHANISMS OF ASPARAGINE SYNTHETASE (ASNS) By Kai Li December 2007 Chair: Nigel G. J. Richards Major: Chemistry So far, more and more evidence from se veral actively studied polyfunctional enzymes indicate the existence of an intramolecular tunnel. Among those enzymes, ammonia/ammonium is the most common intermediate thus far. As one of these enzymes, the tunnel in asparagine synthetase B (ASB) from Escherichia coli is about 20 long. It co nnects two active sites, a glutaminase site and a synthetase site. My work on asparagine synthe tase (ASNS), an enzyme involve d in leukemia resistance to the treatment using asparaginase, has focu sed on the following aspects: (1) kinetic characterization of ASNS; (2) st ructure activity relationships; (3) regulatory mechanisms. In this work, I developed a first quantit ative NMR based assay (K. Li et al, Biochemistry, 2007, 46[16] 4840-4849) and studied the efficiency of ammonia tunneling in ASB catalyzed reaction. The function of bioactive macromolecule can only be understood under the structural context. Therefore, to study the function of the intramolecular tunnel in ASB, first I identified the tunnel residues by computational methods. Most of the tunnel residues are conserved from prokaryote to eukaryote species. The studies of se veral tunnel mutant showed that one of the mutants may have a blocked tunnel.

PAGE 13

13 Among the family of amidotransferases, the tw o half reactions are st rictly coupled. While ASNS can catalyze the hydrolysis of glutamine without other substrates, this raises the interest in the regulatory mechanisms of this enzyme, which prevents it from consuming glutamine and releasing free ammonia. As part of my PhD research, I investigated the mechanisms of product inhibition to this enzyme. Under physiological conditions, the activ ity of this enzyme may be inhibited significantly by asparagine, which pr events cells from wasting glutamine and producing toxic ammonia in vivo So far, only one crystal structure of glutamine dependent asparagine synthetase was reported. To further understand the function of ASNS, especially of human enzyme, more crystal structures are needed. Therefore, co-working w ith Alexandria Berry, I investigated the binding of substrate/transition state analog to hASNS a nd successfully prepared inhibitor bound enzyme. The prepared enzyme has been sent for crystal analysis.

PAGE 14

14 CHAPTER 1 INTRODUCTION Asparagine Synthetase (ASNS) and the Intramolecular Tunnel With the advances in molecular biology and crystallography, many co mplex proteins have been structurally determined. More and more eviden ces show that intramolecular tunnel exists in many enzymes with multiple catalytic sites. Since the first intramolecular tunnel within tryptophan synthase was found in 1988 (1), tunnels within at l east 11 enzymes have been demonstrated (2, 3). Among these tunnels, NH3/NH4 + tunnels in amidotransferases are the most common. It has been determined in carbamoyl phosphate synthetase (CPS), GMP synthetase, glutamine phosphor-ribosylpyrophosphate amidotransfe rase (glutamine PRPP amidotransferase), asparagines synthetase B (ASB), glutamate sy nthase, imidazole glycerol phosphate synthase, glucosamine 4-phosphate synthase and recently, CTP synthase ( 4). All these enzymes use glutamine as nitrogen source donor. First, glutam ine is hydrolyzed into glutamate (product) and NH3/NH4 + (intermediate) at one active site. Then NH3/NH4 + is transported to another active site through an intramolecular tunnel fo r next half-reaction. It seems that all proteins that use glutamine to produce NH3/NH4 + as intermediate will have this kind of intramolecular tunnel. Among those enzymes with NH3/NH4 + tunnel identified thus far, asparagine synthetase (include ASB and mammalian AS), Glucosam ine 4-phosphate synthase, Glutamine PRPP amidotransferase, and glutamate synthase be long to Ntn hydrolases / class II glutamine amidotransferase subfamily, since they all have an N-terminal nucleophile cysteine, which catalyzes the hydrolysis of glutamine ( 5, 6). Asparagine synthetase catalyzes ATP-dependent conversion of aspartate to asparagine. The discovery of asparagine synthetase resulted from the studies about the metabolism of amino acids. Asparagine is one of the amino acids th at have been lately studied. After the first

PAGE 15

15 discovery of proteins with as paragine synthetase activity in the early 1950s, asparagine synthetase has been found in many species incl uding prokaryote and eukaryote. The studies for asparagine synthetase raised even more interest s because of the correla tion between asparagine level in human cells and the resistance to the trea tment of asparaginase in certain types of cancer cells such as acute lymphoblastic leukemia (ALL) (7, 8). In E. coli two unlinked genes with little similarities, asnA and asnB encode asparagine synthetase ( 9-12). While asparagine synthetase A (ASA) from asnA can only use ammonium as nitrogen source, asparagine synthetase B (ASB, EC 6.3.5.4) can use both ammonium and glutamine, with a preference of glutamine. AS B belongs to class II or Ntn amidotransferase subfamily and catalyzes following reactions. Gln + H2O + ATP + Asp Glu + AMP + PPi + Asn (R1) NH3 + ATP + Asp AMP + PPi + Asn (R2) Gln + H2O Glu + NH3 (R3) Because of its 37% identity to human asparagi ne synthetase and the same reactions that catalyzes, ASB has served as a model system to understand the bios ynthesis and metabolic control of asparagine. Up to now, the function of ASB from E. coli has been extensively studied ( 13). Boehlein et al (1994) repor ted that both C1A and C1S muta nt of ASB lose glutaminase activity completely, although these tw o mutants still can bind glutamine. This result confirms the important catalytic role of N-terminal cysteine residue in hydrolysis of glutamine. By analogy with cysteine proteases, the thiolate anion of Cys1 residue act as nucl eophile (Figure 1-1) and attack the -carbonyl group of glutamine to form a th ioester intermediate and ammonia. The thioester then reacts with water to produce gl utamate. This mechanism is strongly supported by the report that a covalent intermediate durin g the glutaminase reaction were isolated and proposed to be -glutamyl thioester ( 14). In 1998, the same research group reported their studies

PAGE 16

16 about the kinetic mechanism of E. coli ASB. Scheme 1-1 shows the kinetic mechanism based on their report. They also provided the evidence by 18O transfer experiment that supports the existence of intermediate, -aspartyl-adenylate ( -aspartyl-AMP), which reacts with NH3/NH4 + produced by hydrolysis of glutamine (or NH3/NH4 + from bulk solution) to synthesize asparagine (Figure 1-2). After the crystal structure of C1A mutant of ASB was reported ( 15), there is no doubt about the existence of an intram olecular tunnel in ASB. The crys tal structure clearly shows two active sites, named glutaminase site and synthetase site. Glut aminase domain comprises the Nterminal half of the protein (1-190) and synt hetase domain the C-terminal half (208-553). The two halves are connected by a coil of 17 residues. The distance between two active sites is about 20 Like other members of Ntn subfamily, glutamine domain is a four layer sandwich structure. Glutamine binding pocket is between the two antiparallel -sheets at one edge of the domain (Figure 1-3). Glutaminase domain catalyzes the hydrolysis of glutamine (R3). Synthetase domain catalyzes ATP-dependent synthesi s of asparagine (R2), using NH3/NH4 + produced by R3. There are two possible pathways for NH3/NH4 + produced at glutaminase site to get to synthetase site. One is intermolecular pathway. That is, NH3/NH4 + first is released by ASB and then equilibrates with bulk solution before the synthesis reaction. The other pathway is through intramolecular tunneling. The kinetic results rule out the first one as the major pathway, if it does exist. Suppose NH3/NH4 + goes through bulk solution, it will take some time for NH3/NH4 + cumulating and reaching a concentration at which the synthetase activity can be detected. This means there would be a lag for the glutamine dependent synthetase activity. Furthermore, the kcat value of glutamine dependent synthetase ac tivity should be same as that of ammonia dependent activity.

PAGE 17

17 This is inconsistent with the kinetic results. Th ere must be an intramol ecular pathway (tunneling) to transport NH3/NH4 +. To date, ASB has been extensivel y studied by steady state studies (16-24 ). Although there is no doubt for the existence of tunnel in ASB, the mechanisms of ammonia tunneling are still unclear. In order to understa nd vectorial transport of NH3/NH4 + under structural context, I explored the struct ure of the tunnel, The Intramolecular Tunnel Determination of residues that involved in NH3/NH4 + tunnel by experiment is the essential step for understanding the tran sportational mechanisms of NH3/NH4 +, synchronization of the two active sites, and therefore, the catalytical mechanisms of asparagine synthetase. 41 residues were determined involving in the NH3/NH4 + tunnel within ASB based on the crystal structure (PDB code : 1CT9) of C1A mutant of E. coli ASB. These residues can be arbitrarily divided into 3 parts, the mouth, the body, and the botto m. Figure 1-4 shows the cavity formed by these residues with Gln and AMP bi nding. By aligning sequen ces from 21 species, including bacteria, plant and an imal, prokaryote and eukaryote, 21 out of 41 residues mentioned above are identical. 7 are almost absolutely conserved with only one exception. Most of the rest are substituted by similar residue su ch as L by I or by V (Appendix A). Glutaminase Active Site and the Mouth of the Tunnel Table 1-1 lists all the residues in glutamine active site of E. coli ASB. The distinctive feature of Ntn subfamily is that they a ll contain several invariant residues. Cys1-Arg49-Asn74Gly75-Asp98 motif is distinctive feature of glutamine binding site of Ntn subfamily ( E. coli ASB numbering) (Figure 1-5). Arg49 and Asp98 account for the specific bi nding of glutamine. Carboxyl group of glutamine forms ionic bon d with Arg49 as well as hydrogen bonds with backbone of Ile52 and Val53, while the ami no group interacts with two negative-charged

PAGE 18

18 residues Glu76, Asp98 and the backbone of Gly75. The distance between carboxyl O atom of Asp98 and amino N of Gln is 2.56 indicating a very strong interaction. Meanwhile, Gly75 also forms a strong hydrogen bond with the carbonyl/a mide group of Gln with an N-O distance of 2.53 Carbonyl/amide group of Gln is also hydrogen-bondi ng with sidechain of Asn74. With the amino group of Cys1, these two finge rprint residues are believed to provide an oxyanion hole that stables transi tion state during the hydrolysis of glutamine. N/amide of Gln also interacts with carbonyl O of backbone of Leu50. The amide group of Gln are surrounded by Al a 1 (Cys 1), Leu 50, Ile 52, Asn 74, Gly 75, Glu 76 (Figure 1-6), which form the mouth of th e tunnel. These residues are highly conserved (Table 1-1, highlighted residues) and probably play an important role not only in binding of Gln and catalysis, but also in NH3 releasing. The Body of the Tunnel In the interface of the two domains, Met 120, Ile 143, Asn 389, Ala 399, plus the carbonyl group of the backbone of Ala 388 form the narrow part of the tunnel, with a diameter of about 56 (Figure 1-4, table 1-2 & Figure 1-7). His 29, Arg 30, Gly 31, Pro 32, Tyr78, Me t392, Ser 393, Gly 396, Val 397, Glu 398 and Arg400 locate between the narrow part and the m outh of the tunnel (Figure 1-8A, B) and form the upper part of the tunnel. This part of the tunnel contains an absolutely conserved residue cluster in asparagine synthetases from different species, His29 -Arg30 -Gly31 -Pro32 -Asp33. Arg30 and Gly31 are also fingerprint residues of Ntn subfamily. Other conserved residues in this part include Gly396 and Glu398. The residues in this part can not form a ring by themselves like the residues in the other part. They fill the gap between mouth and tunnel part 2. The size of tunnel here is bigger than other parts.

PAGE 19

19 The lower part of the tunnel comprise s Ile142, Leu232, Met329, Val344, Leu345, Ser346, Cys385, Ala388, Val401, Leu404 and sidechain of Gl u348. Some of these residues are also in the ATP binding site (Figure 1-8C ), such as Leu232, Leu345 and Se r346. Most of the residues in this part are hydrophobic (Fi gure 1-9) and conserved. The Bottom of the Tunnel This part of tunnel is formed by Asp 238, Gly 347, Glu 348, Gly 349, Ser 350, Asp 351, Glu 352, Tyr 357, Leu 380 and Asp 384. Most of the residues are in ATP binding site (Figure 18D). An interesting feature of this part is that almost all the re sidues are polar and 5 out of 10 are negative-charged residues (Figure 1-9). These nega tive-charged residues, plus Tyr357 form the end of the tunnel, which is presumed as the NH3/NH4 + binding site. The carboxyl group of Glu 352 is right at the bottom of the tunnel and about 4-7 far away from AMP. This negative charge may provide the driv ing force to transport NH3/NH4 +. The -phosphate group of AMP (ATP) is on the side of this part of tunnel (Figure 1-10). The Interface of the Two Domains and Water The fact that binding of ATP s timulates glutaminase activity of E. coli ASB suggests conformational changes during the catalysis ( 17 ). The signal between catalytic domains should arrive through those residues in the interface, esp ecially those that interact with other residues from different domain (Table 1-3). 7 water molecules are in the tunnel, wate r 2, 3, 50, 190, 690, 709 and 861. Water 709 is the only one in the ATP binding site with AMP as the ligand. Since the kine tic data suggest no ATP hydrolyzing during catalysis, wate r molecules are supposed to be excluded when ATP and Asp bind the enzyme. The rest are all in the interface between the two domains. ASB is an ideal system to study highly conserved NH3/NH4 + tunnel in asparagine synthetase. With successful techniques of cloning, expressi on and purification (16), coupled with

PAGE 20

20 the well-developed assay methods such as Glutaminase assay ( 17), PPi assay ( 22), and HPLC assay ( 15), the intramolecular tunnel of ASB were studied and the structur e activity relationship were investigated.

PAGE 21

21 Figure 1-1. Proposed mechanism of hydrolysis of glutamine by E. coli asparagine synthetase B The thiolate anion group of Cys1 acts as nucleophile and form an ES thioester. Thioester

PAGE 22

22 Figure 1-2. Synthesis of asparagines through an intermediate -aspartyl-adenylate by ASB

PAGE 23

23 Figure 1-3. Cartoon of glutaminase domain of E. coli ASB. Glutamine is showed by spacefill. Ala1(Cys1) is showed by ball and stick.

PAGE 24

24 Figure 1-4. The tunnel within ASB Tunnel residues are showed in sticks, colored by cpk. The cavity (tunnel) is brown. Glutamine (left) and AMP (right) are showed in spacefill, colored by cpk. Water (red) and heavy metal ions (grey) are showed in small-size of spacefill.

PAGE 25

25 Figure 1-5. Cys1-Arg-Gly-Asn-Asp is a common motif of the glutamine binding site for Ntn subfamily of glutamine dependent amidotransfe rases A. asparagine synthetase B from Escherichia coli (PDB code: 1CT9, modified). B. Glucosamine 4-Phosphate Synthase from Escherichia coli (PDB code: 1JXA and 1XFF, modified). C. Glutamate Synthase From Synechocystis Sp (PDB code: 1OFE). D. Glutamine Phosphoribosyl-pyrophosphate (Prpp) Amidotransferase from Escherichia coli (PDB code: 1ECC). A B C D

PAGE 26

26 Figure 1-6. The mouth of the tunnel 6 highly conserved residues, Ala 1 (Cys 1), Leu 50, Ile 52, Asn 74, Gly 75, Glu 76, surround the amide group of glutamine and form the mouth of the tunnel. A. Mouth of the tunnel with glutamine. Probe radius is 1.4 B. Relative position to AMP. The mouth of tunnel is showed in green. A B

PAGE 27

27 Figure 1-7. The narrowest part of the tunnel formed by M120, I143, N389, A399 and carbonyl group of A388.

PAGE 28

28 Figure 1-8. The tunnel in E. coli ASB. Mouth (green); Upper part (yellow); narrow part (brown); Lower part (cyan); Bottom part (pink). Gln (l eft) and AMP (right) are showed in spacefill; Glu352 are showed in ball and s tick form and colored by cpk. A B C D

PAGE 29

29 Figure 1-9. The body of the tunnel is mainly fo rmed by hydrophobic residu es (yellow) and the bottom almost totally by polar (green) and acidic (red) residues. Several basic residues (blue) close to glutamine site.

PAGE 30

30 Figure 1-10. The Glu352 is right at the bottom of the tunnel. AMP (spacefill) is on one side. The hole shows the tunnel. Mouth (green); Upper part (yellow); narrow part (brown); Lower part (cyan); Bottom part (pink).

PAGE 31

31Table 1-1. Residues in Gln binding pocket and possible roles Resname Functional group H-bond/salt bridge Distance () Conservation Roles Ala1 (Cys1) backbone --conserved Forming tunnel sidechain --Nucleophilic catalysis His47 sidechain --H or F Active site Arg49 sidechain N-HO1 3.09 conserved specificity sidechain N-HO1 2.97 Leu50 backbone OH-NE 3.07 conserved Binding of Gln, catalysis and forming tunnel? sidechain --Ser51 whole --Not cons. Active site Ile52 backbone N-HO2 2.79 I or V Binding of Gln and forming tunnel? sidechain --Val53 backbone N-HO2 2.77 Not cons. Gln binding? sidechain --Gly58 whole --conserved Active site? Gln60 whole --conserved Active site? Leu62 sidechain --Cons. subst. Active site Val73 whole --Not cons. Active site Asn74 backbone --conserved Specificity and forming oxyanion hole and tunnel sidechain ND-HOE 2.97 Gly75 -OH-N 3.12 conserved Specificity and forming oxyanion hole and tunnel -N-HOE 2.53 Glu76 backbone conserved Gln binding and forming tunnel? sidechain OE1H-N 3.09 Ser97 whole --S, C, V Active site Asp98 backbone --conserved Specificity sidechain OD1H-N 2.56 Cys99 whole --Not cons. Active site H2O3 -O-HOE 2.80 Reactant? H2O30 -O-HO1 2.89 Note : Those residues forming mouth of the tunnel are highlighted by yellow. Fingerprint residues of Ntn subfamily are showed in red

PAGE 32

32Table 1-2. Size of narrow part of the tunnel Residues Atoms Distance () Met120 Asn389 CE -ND 5.58 Met120 Asn389 SD -ND 5.51 Ala399 Ile143 O CD1 6.27 Ala399 Ile143 O CG1 6.85

PAGE 33

33 Table 1-3. Residues in the interface between two domains of E. coli ASB (Those residues forming tunnel are showed in red) Residues Atoms Distance () Conservation Asp115 OD1 2.77 not Arg193 NH1 not Asp54 N 2.81 D or G D-D or G-E Asp226 OD1 D or E Ile52 O 2.74 16I, 5V Asp226 N 17D, 4E Glu76 OE1 2.51 Conserved Glu398 OE1 Conserved Asp33 OD1 2.76 Conserved Lys342 NZ 18K except mammalian Lys342 O 2.76 18K except mammalian Glu398 N Conserved Gln118 NE 2.92 14D, 4Q, 1R, 1T, 1N Asp405 OD1 conserved His139 NE2 2.85 7P, 6A, 5H, 3T Leu404 O Conserved Glu173 OE1 2.23 not His381 ND1 15H, 5Y (mammalian), 1N Lys162 NZ 2.89 Conserved Glu316 OE1 Conserved Lys162 N 2.76 Conserved Glu316 OE1 Conserved Met161 N 3.07 10L, 9M, 3A (mammalian) Glu316 OE1 Conserved Met161 N 2.89 10L, 9M, 3A (mammalian) Glu316 OE2 Conserved Lys390 NZ 2.66 18K, 3R (mammalian) Glu316 OE2 Conserved Tyr146 OH 2.71 18Y, 3F (mammalian) Lys390 NZ 18K, 3R (mammalian) Arg30 NH1 2.84 conserved Asn389 OD1 17N, 4D (mammalian) Arg30 NH2 3.05 Conserved Asn389 OD1 17N, 4D (mammalian) His192 --not Lys406 --13K, 3R, 3H (mamm.), 2T

PAGE 34

34 Synthetase site ( C-terminal ) Synthetase site ( C-terminal ) ATP Asp PPi Asn AMP Glutaminase site (N-terminal) Gln NH3 Glu Glutaminase site (N-terminal) NH3 Tunneling Scheme 1-1

PAGE 35

35 CHAPTER 2 A gHMQC-BASED NMR ASSAY FOR INVESTIGATING AMMONIA CHANNELING IN GLUTAMINE DEPENDENT AMIDOTRANSFERASES Introduction Glutamine-dependent amidotransferases cata lyze ammonia transfer from the amide of glutamine to a variety of acceptors ( 3, 5, 6, 25). These enzymes first hydrolyze glutamine into glutamate and ammonia. The ammonia intermediate is then transferred to a second active site and reacts with other substrates to synthesize different nitrogen containing product. X-ray crystal structures of many amidotransferases has showed th at this family of enzymes has multiple active sites which are separated from each other and connected with an intramolecular tunnel ( 2, 15, 26-33). The reactions catalyzed at both active sites are integrated together by intermediate transfer through this tunnel. It seems that all the glutamine depe ndent amidotransferases have an ammonia tunnel, with the glutaminase active site at one end and the synthetase active site at the other end. Although it has been proposed that ammonia produced by hydrolysis of glutamine is channeled efficiently into a second active site to complete the overall enzymatic reaction ( 6, 3438), kinetic studies for relative ly few glutamine-dependent amidotransferases have provide evidences to support this hypothesis ( 39-42). Perhaps the most convincing results for efficient ammonia channeling has come from 15N NMR studies of carbamoyl phosphate synthetase (CPS) ( 39, 43-45), which demonstrated that no exchange happens between 14NH3 produced in the glutamine domain of CPS and 15NH3 present in bulk solution ( 39). In addition, pre-steady state kinetic measurements have shown the rates of ammonia utilization and glutaminase activity in CPS to be coupled, implying that only a si ngle ammonia molecule is present in the intramolecular tunnel of the enzyme during catalytic turnover ( 43). Integrated structural, This work was finished with the helpful discussion of an undergraduate William Beeson.

PAGE 36

36 computational and experimental measurements have also provided s upport for active site coupling and intramolecular ammonia transfer in imidazole glycerol phosphate synthase (IGPS) ( 40, 46-49). Both CPS and IGPS are Class I amidotransferases ( 6, 50), however, and their molecular behavior may not therefore be representa tive of the evolutionar ily unrelated Class II ( 6, 51) or Class III amidotransferases (52 ). Glutamine-dependent asparagine synthetase (ASNS) (EC 6.3.5.4) belongs to the Class II amidotransferases, which also includes Gl ucosamine 4-phosphate synthase, Glutamine phosphoribosylpyrophosphate (PRPP) amidotra nsferase, and glutamate synthase (5, 6 ). The two domains of asparagine synthetase, glutaminas e domain and synthetase domain, catalyzes the hydrolysis of glutamine (R3) and the ATP-dependent synthesis of L-aspara gine from L-aspartic acid respectively (R2) (13 ), with R1 as the overa ll enzymatic reaction. Gln + H2O + ATP + Asp Glu + AMP + PPi + Asn (R1) NH3 + ATP + Asp AMP + PPi + Asn (R2) Gln + H2O Glu + NH3 (R3) Both the mammalian and bacterial forms of the enzyme can accept free ammonia as an alternate substrate in vitro ( 24, 53 ). With ATP and aspartate, glutamine is hydrolyzed into glutamate and ammonia at glutaminase site through a thioester enzyme-substrate complex ( 14) and then the ammonia intermediate is transported through a 20 long, intramolecular t unnel to the synthetase site, which is accepted by -Aspartyl-AMP, an intermediate that formed at synthetase site using ATP and Asp, to produce asparagine. In sharp c ontrast to the behavior of other Class II amidotransferases for which the independent catalytic sites are tightly coupled ( 54-56 ), without the ammonia acceptor, asparagine synthetase still can hydrolyze glutamine ( 22, 57). This lack of active site coupling in ASNS is surprising in li ght of observations on other amidotransferases, and raises questions about the st ructural integrity of the solv ent inaccessible, intramolecular

PAGE 37

37 tunnel that is seen in the AS-B crystal structure as the enzyme proceeds through its catalytic cycle. We now report the use of a very convenient isotope-edited 1H NMR-based assay ( 58 ) that we have developed to probe ammonia transfer between the two active sites in glutamine dependent asparagine synthetase from Escherichia coli (AS-B). This gradient heteronuclear multiple quantum coherence (gHMQC) method, which should be generally applicable for studies of other glutamine-dependent amidotransferases, has provided quantitative information on the extent to which 15N is incorporated into the side chain of asparagine formed in the enzymecatalyzed reaction. In addition, although our studies of AS-B have employed end-point measurements, the gHMQC strategy may be amenab le for use in continuous assays of nitrogen transfer catalyzed by other members of this enzy me family. At least in the case of AS-B, the results of these gHMQC NMR studies show that (i) high glutamine concentrations do not suppress ammonia-dependent asparagine formation by this enzyme, and (ii) ammonia in bulk solution can undergo reaction with a thioester intermediate formed during the glutaminase halfreaction ( 14, 59). These observations are consistent with a model in which exogenous ammonia can access the tunnel in AS-B during glutamine-depe ndent asparagine synthesis, in contrast to expectations based on studies of Class I amidotransferases ( 39, 46, 47). Material and Methods Materials Unless otherwise stated, all chemicals and reagen ts were purchased from Sigma (St. Louis, MO), and were of the hi ghest available purity. 1,3-15N2-Uracil, 15N-L-asparagine, and d6-DMSO were purchased from Sigma-Aldrich (St. Loui s, MO). The isotopic incorporation in these samples was greater than 99%. All experiment s employed freshly prepared solutions of recrystallized L-glutamine ( 17). Recombinant, wild type AS -B was expressed and purified

PAGE 38

38 following literature procedures ( 16 ). Protein concentrations were determined using a modified Bradford assay (Pierce, Rockford, IL) ( 60), for which standard curv es were constructed with bovine serum albumin, and correc ted as previously reported ( 55). NMR spectra were recorded on a Varian INOVA 500 instrument equipped with a 5 mm triple resonance indirect detection probe (z-axis gradients) operati ng at 500 and 50 MHz for 1H and 15N, respectively. Chemical shifts are reported in ppm relative to so dium 2,2-dimethyl-2-silapentane-5 -sulfonate (DSS). The co-axial inner cell (catalog: NE-5-CIC-V) containing the external NMR standards was purchased from New Era Enterprises (Vineland, NJ). NMR Measurements Concentrations of 15N-L-asparagine were determined using a solution of 1,3-15N2-uracil dissolved in d6-DMSO as a standard, which was placed in a coaxial inner cell inserted into the 5 mm tube containing the assay sample. The amount of 15N-L-asparagine was measured using a phase-sensitive, 1-D gHMQC pulse sequence ( 58), as implemented in the Varian VNMR software package (version 6.1C). NMR spectra were acquired on a Varian INOVA500 spectrometer, at a fixed temperature of 25 oC, in 256 transients with a digital resolution of 0.25 Hz/point (25966 points in the FI D over a spectral window of 6492 Hz) using a relaxation delay of 1 s and an acquisition time of 2 s. The total time for acquiring each spectrum was therefore 13 min. The encoding gradient level was 37 Gauss/cm of 2.5 ms duration, and the corresponding values for the decoding gradient were 18.7 Gauss/ cm and 1 ms, with a gradient recovery time was 0.5 ms. 90 pulse times us ed in the experiment for 1H and 15N were 9.6 sec and 25.8 s, respectively, and the 1H-15N coupling constant was set to a value of 87 Hz. The FID was weighted with a line broadening of 10 Hz and a Gaussian of 0.278 Hz. Competition Experiment Reaction mixtures consisted of 100 mM HEPPS buffer (pH 8.0), 60 mM MgCl2, 10 mM

PAGE 39

39 ATP, 20 mM aspartate and di fferent concentrations of 15NH4Cl (pH 8.0) or glutamine in total volume of 2 mL. In experiment s where the concentration of 15NH4Cl (pH 8.0) was varied as 2.0 mM, 5.0 mM, 25 mM, 50 mM and 100 mM, L-glutam ine was fixed at 20 mM. Alternatively, when L-glutamine was varied as 0 mM, 2.5 mM, 10 mM, 20 mM and 40 mM, 15NH4Cl was added at an initial concentration of 100 mM. Reactions were initiated by the addition of AS-B (31 g) and the resulting samples incubated for 10 min at 37 oC before being quenched by the addition of trichloro acetic acid (TCA) (60 L). After centrifugation fo r 5 min at 3000 rpm to remove precipitated protein, th e supernatant was adjusted to pH 5 by the addition of 10 M aq. NaOH and an aliquot of this solution (650 L) transferred to a 5 mm NMR tube for analysis. At higher pH values, amide NH exchange precluded the derivation of any qua ntitative relationship between peak area and 15N-Asn concentration. The standard samples contained 100 mM HEPPS buffer (pH 8.0), 60 mM MgCl2, 10 mM ATP, 100 mM 15NH4Cl (pH 8.0), 20 mM glutamine and different concentrations of 15N-L-Asn from 0.25 mM to 5 mM, without adding AS-B. 15N Exchanging Experiment Reaction mixtures consisted of 100 mM HEPPS buffer (pH 8.0), 60 mM MgCl2, 10 mM ATP and different concentrations of 15NH4Cl (pH 8.0) or glutamine in total volume of 2 mL. In experiments where the concentration of 15NH4Cl (pH 8.0) was varied as 2.0 mM, 5.0 mM, 25 mM, 50 mM and 100 mM, L-glutamine was fixed at 20 mM. Alternatively, when L-glutamine was varied as 0 mM, 2.0 mM, 5.0 mM, 10 mM, 20 mM, 40 mM and 80 mM, 15NH4Cl was added at an initial concentration of 100 mM. Reactions were initia ted by the addition of AS-B (62 g) and the resulting samples incubated for 10 min at 37 oC before being quenched by the addition of trichloroacetic acid (TCA) (60 L). After centrifugation for 5 min at 3000 rpm to remove precipitated protein, the supernat ant was adjusted to pH 5 by the addition of 10 M aq. NaOH and an aliquot of this solution (650 L) transferred to a 5 mm NMR t ube for analysis. At higher pH

PAGE 40

40 values, amide NH exchange precluded the deriva tion of any quantitative relationship between peak area and 15N-Asn concentration. The standard sa mples contained 100 mM HEPPS buffer (pH 8.0), 60 mM MgCl2, 10 mM ATP, 100 mM 15NH4Cl (pH 8.0), 20 mM glutamine and different concentrations of 15N-L-Asn from 0.10 mM to 5 mM, without adding AS-B. HPLC-Based Determination of Total L-Asparagine In order to obtain an estimate of total L-asparagine in the final reaction mixtures, we employed an HPLC-based end-point assay ( 15). Hence, an aliquot of each mixture (40 L) was diluted (200 L final volume) with 400 mM aq. Na2CO3, pH 9, containing 10% DMSO and 30% dinitrofluorobenzene (DNFB) (as a saturated solution in EtOH). The resulting solutions were heated at 50 C for 45 min to permit reaction of DNFB with the amino acids to yield their dinitrophenyl (DNP) derivatives. Caution : Extreme care should be taken when handling solutions of 2,4-dinitrofluorobenzene in organic solv ents because this reagen t is a potent allergen and will penetrate many types of laboratory gloves ( 46). Aliquots of each assay mixture (20 L) were analyzed by reverse-phase HPLC (RP-HPLC) using a C18 column and a flow-rate of 0.7 mL/min. The DNP-derivatized amino acids were elut ed using a step gradient of 40 mM formic acid buffer, pH 3.6, and CH3CN. In this procedure, the initial concentration of the organic phase (CH3CN) was 14%, which was maintained over a period of 26 min before the amount of CH3CN was increased to 80% over a period of 30 s, a nd elution continued for a further 8 min. Eluted amino acid DNFB derivatives were monitored at 365 nm and identified by comparison to authentic standards. Under these conditions, DNP-asparagine exhibited a retention time of approximately 25 min, and could be quantified on th e basis of its peak area. Calibration curves were constructed using solutions of pure L-asparagine derivatized in the same manner as the samples.

PAGE 41

41 Kinetic Simulations Simulations were performed using the GEPASI software package ( 61, 62 ). Results and Discussion Two mechanisms can be envisaged for nitrog en transfer from the glutaminase to the synthetase active sites in AS-B (Figure 2-1). In th e first, ammonia is directly transferred between the two active sites through an intramolecular tunnel ( 2, 15, 26-33). This proposal, which is hypothesized to occur in all other glut amine-dependent amidotransferases ( 6, 34-38), is supported by X-ray crystallographic observations on the bacterial enzyme ( 15). A second model can be envisaged, however, in whic h ammonia is released into bul k solution prior to re-entering the synthetase site and reacting with the -aspartyl-AMP intermediate. We therefore sought a simple and rapid kinetic assay to distinguish be tween these possibilities, which might also be applicable in experiments aimed at developi ng structure-function rela tionships for residues defining the tunnel observed in AS-B. Early work on CTP synthetase ( 41, 63, 64 ) and GMP synthetase ( 65) aimed at investigating this problem em ployed the pH-dependence of synthetase activity when glutamine and ammonia (or alternat e substrates such as hydroxylamine) were both present in solution. We elected to employ an a lternate, and somewhat more straightforward, strategy, however, in which th e glutamine-dependent asparagi ne synthetase reaction was performed in the presence of exogenous 15NH4Cl. In particular, we hoped to employ 15N NMR spectroscopy to determ ine the extent of 15N incorporation into asparagine as a function of glutamine concentration, as previo usly reported in studies on CPS ( 19). The low sensitivity of the 15N nucleus proved to be a significant limitati on to our efforts, mandating a substantial investment of spectrometer time in order to obta in spectra with sufficiently high signal-to-noise for any quantitative measurements, thereby se verely limiting the num ber of conditions under which competition studies could be carried out. We therefore investigated the use of a phase-

PAGE 42

42 sensitive 1D gHMQC experiment to measure the extent to which 15N was incorporated into asparagine when AS-B was incubated with aspa rtate and ATP in the presence of both glutamine (containing nitrogen isotopes at natural abundance) and 15NH4Cl. NMR Assay In gHMQC spectra, resonances are only observe d for hydrogen nuclei that are (i) attached to 15N nuclei, and (ii) are not in fast exchange. The chemical shifts for 15N-amide protons are: 15N-Asn ( NHcis = 7.58 (dd) and NHtrans = 6.86 (dd), J = 90 Hz); 15N-Gln ( NH1 = 7.53 (dd) and NH2 = 6.81 (dd), J = 90 Hz). Hence, this NMR stra tegy permits a 30-fold increase in sensitivity over direct acquisition of 15N spectra. In addition, and very importantly for the goals of these experiments, the time needed to acquire 15N-edited 1H NMR spectra was approximately three orders of magnitude faster than fo r the equivalent measurements using 15N NMR spectroscopy. We were therefore able to obtain very clean 1H NMR spectra for assa y mixtures containing AS-B because the signals for protons on free 15NH3 and 15NH4 + were not observed (Figure 2-2). Accurate integration of signals relative to those from an internal standard (15N2-uracil dissolved in d4-DMSO) was also possible, and standard curves relating the observed peak area of amide proton signals to the concentration of 15N-asparagine present in solution under a variety of conditions were easily obtained (Figure 2-3). All of these result s were carefully validated by independent measurements of aspara gine using an HPLC-based assay ( 23 ). Competition Experiments Having established that the amount of 15N-asparagine formed in the reaction could be quantitated using gHMQC spectroscopy, we studied whether the incorporation of 15N from exogenous 15NH4Cl into asparagine could be suppressed by L-glutamine in the synthetase reaction catalyzed by AS-B. If ASB was incubated with both 14N-glutamine and 15NH4Cl at the presence of other substr ates and cofactor, both 14N-asparagine and 15N-asparagine would be

PAGE 43

43 produced. 14N-asparagine was produced by 14NH3 intermediate that goes through an intramolecular tunnel. And 15N-asparagine was formed by 15NH3 from free solution. 15NH3 may compete with 14NH3 from 14N-glutamine as nitrogen donor. Theref ore, quantitative analysis of both 15N-asparagine and 14N-asparagine in the reaction mixt ure gives information about the capability of ASB to utilize different nitroge n source (scheme 2-1). If ammonia can not access the active site with the presence of gl utamine, all the asparagine would be 14N product at saturating concentration of glutamine. In this case, free ammonia is suppressed by glutamine. The ratio of 14N asparagine to 15N asparagine would be infinite. This ratio will be decreasing with the extent of suppressi on becomes less. When there is no suppression at all, the 14NH3 will compete with 14NH3 freely. If ammonia tunneling is fast and efficient enough, th en the ratio of 14N asparagine to 15N asparagine would be determined by the ratio of production rates using ammonia and glutamine. That is, i bindingmcat tunneling mcatNH NHENH Kk E Kk v v AsnN AsnN ][ ]Gln[ ][ ]Gln[ ]][)[/( ]][Gln)[ /( ][ ][3 15 3 15 3 15 15 14 15 14 Therefore, iNH G AsnN AsnN ][ ln][ ][ ][3 15 15 14 or iNH G A AsnN AsnN ][ ln][ ][ ][3 15 15 14 Alternatively, if the tunnel is not efficient, the ratio of 14N asparagine to 15N asparagine would decreases. Most of the asparagine w ould be labeled form and the ratio of 14N/15N in asparagine would be close to zero if the 14NH3 intermediate leaks and equilibrates with 15NH3 from bulk solution, since [15NH3] was much bigger than [14NH3]. Hence, we studied the competition in the presence of both 14NH3 and 14N-glutamine as nitrogen donors. Under experimental conditions,

PAGE 44

44 the final concentration of asparagine formed was 1-2 mM. The dependence of 15N incorporation into the product amide on the initial concentration of L-glutamine was determined using the NMR assay we developed and HPLC assay (Table 1), after correcting the amount of 15N in Lasparagine to allow for the natural abundance of 15N in L-glutamine. A similar set of experiments in which L-glutamine was fixed at 20 mM and 15NH4Cl varied over a range of initial concentrations was also car ried out (Table 2-1). Somewhat unexpectedly, in light of the behavior reported for CPS in a similar competition experiment (19), we did not observe complete suppression of 15N incorporation at saturating concentrations of L-glutamine (KM(app): 0.69 mM ( 53)). Instead, the 14N/15N incorporation ratio exhibited saturation behavior, reaching a limiting value of 1.2 0.2 as the concentration of Lglutamine was increased (Figure 2-4A). The ammonia-dependence of the 14N/15N incorporation ratio when the initial concentration of L-glutamine was fixed at 20 mM was also examined and again showed substantial 15N incorporation even at non-sa turating concentrations of 15NH3 (Figure 2-4B). Any calculation of the expected 14N/15N ratio in the side chain of the product by the simple comparison of the V/K values for the two nitrogen sources is complicated, however, by the fact that L-asparagine inhibits the glutaminase act ivity of ASNS with an apparent KI of 50-60 M ( 34 41) while having no significant impact on am monia-dependent synthetase activity at 1 mM concentration (41; also see supporting informati on). Thus, the rate of ammoniadependent L-asparagine synthesis is una ffected by the presence of L-asparagine over the period of our competition experiment, while that for gl utamine-dependent synthetase activity decreases over time (Figure 2-4C). The effects of this di fferential inhibition are reflected in the total amount of L-asparagine formed after a given time under the assay conditions, which depends on whether one or both nitrogen sources are pr esent. As a result, the total amount of L-asparagine

PAGE 45

45 formed in 10 minutes is greater when both ex ogenous ammonia and glutamine are present in solution than when glutamine is employed as th e sole nitrogen source (Table 2-1). This also explains the observation that total L-asparagine is increased when ammonia is added to a solution containing a saturati ng amount (20 mM) of L-glutamine because ammonia-dependent activity is not inhibited by the reaction product binding to the glutaminase active site. Given this complication in understanding the apparent inability of L-glutamine to prevent 15N incorporation into asparagine, we sought to model the theoretical 14N/15N incorporation ratio as a function of either L-glutamine or ammonia concentration by kinetic simulations of the competition experiment. In keeping with hypotheses developed from previous studies on CPS ( 19) and IGPS ( 20), we assumed that exogenous 15NH3 and L-glutamine could not bind simultaneously to the E.ATP.Asp ternary complex requi red for synthetase activity ( 40, 41). The assignment of rate constants was accomplished (40) on the basis of (i) direct NMR and HPLC measurements of Lasparagine production rate under the assay conditions (k2, k4, k7 and k10), and (ii) literature data on the steady state kinetics of AS-B (k-1, k-3, k-6 and k-9) ( 40 53). All second or der on rate constants were defined as 108 M-1s-1, although this does not account for conformational changes that might take place on substrate binding. Perhap s more importantly, we assumed that (i) the presence of L-asparagine in the glutaminase active site of the enzyme does not affect rate constants associated with ammoni a-dependent synthetase activity (k4, k7 and k10), and (ii) the off rate of asparagine from the inhib itory site is not affected by bound ammonia (k-11 and k-12). The only remaining unknown rate constants (k-11 and k-12) in this simple model could then be estimated from the observed KI for L-asparagine in the glutamine-de pendent synthetase reaction. Although this kinetic model exhibits a time-de pendent decrease in the rate of nitrogen incorporation from L-glutamine to give 14N-L-asparagine in the competition assay (Figure 2-4C),

PAGE 46

46 it does not qualitatively reproduce the saturation behavior of the 14N/15N incorporation ratio as the concentration of L-glutamine is increased in th e presence of a fixed amount of 15NH4Cl (Figure 2-4A). The lack of quantitative agreement between th e experimental and simulated 14N/15N incorporation ratios likely arises from an over-estimation of the affinity of the AS-B glutaminase domain for L-asparagine, and we therefore concl ude that both nitrogen sources can be present on the enzyme simultaneously. 15N exchange Experiments Evidence to support (i) the conclusion from th e kinetic simulations, and (ii) the hypothesis that 15NH3 might be able to access the N-terminal glutaminase site (and presumably the intramolecular tunnel) when L-glutamine was also present came from the observation of peaks in the isotope-edited gHMQC spectrum with chemical shift values that were different from those of the resonances arising from the amide protons of 15N-asparagine (Figure 2-2B). The acquisition of 1H spectra for the reaction mixture in the absence of isotopic editing, but with suppression of the water signal, showed that these unexpected p eaks were satellites of the broad signal from the amide protons in 14N-L-glutamine (Figure 2-5). As a result, we concluded that these signals arose from amide protons bonded to 15N in the side chain of L-glutamine. Control experiments established that 14N/15N exchange did not occur in the absenc e of AS-B, and so this reaction was examined in more detail by incubating the enzyme with 15NH4Cl in the presence of glutamine and ATP. Under these conditions, the rate of 15N-substituted glutamine formation (i) showed saturation behavior with respect to glutam ine at a fixed concentration (100 mM) of 15NH4Cl (Figure 2-6), and (ii) was linearly depe ndent on the concentration of exogenous 15NH4Cl (data not shown). A kinetic model was constructed to describe the 14N/15N exchange (Figure 2-7A), which gave the following equation for partitioning of the thioester intermediate at the steady-state to

PAGE 47

47 yield 15N-L-glutamine (see supporting information): ])[][ln][( ]ln][[][234 15 2 1 4 15 21OHkNHkGk NHGEkk vT (Eqn. 1) Here v is the rate of 15N-glutamine production via the exchange reaction in the presence of ATP but the absence of aspartate, and k1, k2 and k3 are microscopic rate constants with estimated values of 5.1 mM-1s-1, 7.8 x 10-3 mM-1s-1 and 1.06 x 10-4 mM-1s-1, respectively, assuming that hydrolysis of the thioester intermediate is the rate-limiting step in the glutaminase reaction. Using these values in our model for the exchange reaction gave excellent agreement between the observed concentrations of 15N-glutamine and those expected from kinetic simulation at a variety of NH4Cl concentrations (Figure 2-7B). Having esta blished a firm understanding of this reaction in the absence of aspartate, we next examined whether the overall rate of 14N/15N exchange was altered when AS-B was undergoing catalytic turnove r to yield asparagine. A similar analysis to that described above showed that our kinetic model gave excellent agreement with experimental values (data not shown) when the microscopic rate constants k1, k2 and k3 were assigned values of 9.6 mM-1s-1, 1.8 x 10-2 mM-1s-1 and 8.8 x 10-5 mM-1s-1, respectively. The similarity of these rate constants to those determined for 14N/15N exchange in the absence of aspartate suggests that the molecular mechanism of the reaction is identical under both sets of conditions. The simplest explanation for 14N/15N exchange is that 15N-substituted glutamine is formed by reaction of exogenous 15NH3 with a thioester intermediate form ed in the N-terminal active site during the enzyme-catalyzed hydrolys is of glutamine (Scheme 2-2) ( 14, 59 ). This is an intriguing finding because a similar 14N/15N exchange reaction involving the side chain amide of Lglutamine was not observed in similar competition experiments employing CPS ( 19), although it is possible that the amounts of 15N-substituted glutamine formed might have been too small for detection by 15N NMR spectroscopy. In addition, this phenomenon has not been described

PAGE 48

48 previously, to the best of our knowledge, for a ny other glutamine-dependent amidotransferase. The molecular pathway by which 15NH3 gains access to the N-terminal, glutaminase site remains to be determined. Hence, it is possible that exogenous ammonia can access the N-terminal domain directly as a consequence of the enzy me adopting a conformation in which the tunnel linking the two active sites is solvent accessible. Alternatively, a mmonia might enter the intramolecular tunnel linking the tw o active sites after entering th e C-terminal domain. In both mechanisms, ammonia could displace water molecu les in the tunnel prior to glutamine binding and the adoption of the closed conformation obs erved in the crystal structure of the enzyme. Conclusions In summary, we have reported a sensitive, quantitative and reproducible isotope-edited 1H NMR assay for monitoring the incorporation of 15N into the side chain amide of L-asparagine. Spectra can be obtained rapidly thereby perm itting the examination of a wide variety of conditions for these steady-state competition experi ments. The application of this assay to the reaction catalyzed by AS-B has shown, somewhat unexpectedly, that high concentrations of glutamine can not suppre ss the incorporation of 15N from exogenous 15NH3. This result, in combination with our observation that exogenous ammonia can trap the thioester in termediate in the glutaminase reaction, is consistent with a model in which the putative ammonia channel linking the Nand C-terminal active sites can be accessed by 15NH3 molecules at some point during catalytic turnover. This behavior seems to contrast w ith previous findings concerning ammonia channeling in Class I ami dotransferases, such as CPS ( 19 ) and IGPS ( 20), which appear to have been optimized during evolution for (i) high nitrogen transfer efficiency ( 19), and (ii) tight kinetic coupling of glutam inase and synthetase activities ( 20, 23 ). In part, this may be a consequence of the ability of asparagine to inhibit non-productive glutaminase activity, which might have relaxed any evolutionary pressure to construct intra-domain interactions needed for

PAGE 49

49 efficient kinetic coupling of catalysis in th e two active sites of as paragine synthetase. Any interpretation of the inability of satura ting glutamine to suppress ammonia-dependent synthetase activity in terms of molecular stru cture is complicated by the observed solvent inaccessibility of the intramolecular tunnel that is observed in the high-resolution crystal structure of AS-B ( 12). We have shown previo usly, however, that coupling of the glutaminase and synthetase activities of the enzyme appears to break down as the glutamine concentration is increased ( 40 ), perhaps because of a conformational cha nge that permits the release of ammonia (14NH3) from one or both of the glutaminase sites. It is therefore possible that molecules of exogenous ammonia (15NH3) might be able to gain access to th e synthetase site when the enzyme adopts this conformation, with the result that hi gh concentrations of L-glutamine fail to suppress 15N incorporation to the extent s een for other amidotransferases.

PAGE 50

50 Figure 2-1. Two possible pathways for the ammonia transfer to reach synthetase site. One is intramolecular pathway, in which ammonia intermediate go throug h an intramolecular tunnel after being released from glutaminase site, without exposes to solution. The second pathway is free solution pathway. The ammonia is released to the free solution first, and then binds to synthetase site. In the latter pathway, the concentration of ammonia in the solution is very low at beginning, increases w ith time, and finally reaches maximum. This would produce a lag in the plot of synthetase activity versus time, which can be ruled out by kinetic studies after separate active sites are confirmed.

PAGE 51

51 Figure 2-2. gHMQC 1H-NMR spectrum. (A) Clean 1H-NMR spectrum obtained by gHMQC. Only 15N-amide protons were showed in the spectrum. For the samples used in this work, these peaks (from left to right) stands for 15N2-uracil ( > 11.00 ppm), 15N-amides (glutamine or asparagine, : 6.50 7.80 ppm) and solvent (H2O, = 4.70 ppm) respectively. The peak areas of 15N-glutamine or 15N-asparagine are relative to the total peak area of 15N2-uracil in d4-DMSO, which was set to 100. (B) The 15N-amide peaks. 15N-Asn ( NHcis = 7.58 ppm (dd) and NHtrans = 6.86 ppm (dd), J = 90 Hz); 15N-Gln ( NH1 = 7.53 ppm (dd) and NH2 = 6.81 ppm (dd), J = 90 Hz) Chemical shift (A) C h e mi ca l s hif t N N N N Q Q Q Q (B)

PAGE 52

52 Figure 2-3. Standard curves. (A) Standard curve for exchange experiment. The concentrations of 15NAsn added were 0.05 5.0 mM. R2 = 0.9986. Detection limit < 50 M. (B) Standard curve for competition experiment. The concentrations of 15N-Asn was varied from 0.25 5.0 mM. R2 = 0.9991. Detection limit < 50 M.

PAGE 53

53 Scheme 2-1

PAGE 54

54 A 0 0.2 0.4 0.6 0.8 1 1.2 1.4 02 04 06 0 B 0 2 4 6 8 1 0 1 2 020406080100120 L-Glutamine (mM) 14N/15N ratio in L-Asn side chain 14N/15N ratio in L-Asn side chain 15NH4Cl (mM) 0 1 2 3 020406080100120 15NH4Cl (mM) 14N/15N

PAGE 55

55 C Figure 2-4. 14N/15N incorporation ratios in L-asparagine form ed by the AS-B synthetase reaction under competition assay conditions. (A) Comparison of the experimentally observed ( ) and simulated ( ) 14N/15N incorporation ratio in 100 mM HEPPS buffer, pH 8, containing 10 mM MgATP, 20 mM aspartate and 100 mM 15NH4Cl. (B) Comparison of the experimentally observed ( ) and simulated ( ) 14N/15N incorporation ratio in 100 mM HEPPS buffer, pH 8, containing 10 mM MgATP, 20 mM aspartate and 20 mM Lglutamine. The inset shows an expanded view of the plot for the experimental 14N/15N ratios. (C) Kinetic simulation of the time-depe ndent formation of total L-asparagine (blue), 14N-L-asparagine (red) and 15N-L-asparagine (green) showing the impact of differential inhibition of the two synthetase reactions by L-asparagine. Time (s) Product concentration (mM)

PAGE 56

56 Figure 2-5. Normal 1H-NMR spectrum that shows amide protons from both 15N-glutamine (product, sharp peak) and 14N-glutamine (reactant, sharp peak). The spectrum was acquired after water suppression at 25 oC. Acquisition time was 5s. Number of transients was 128.

PAGE 57

57 Figure 2-6. Glutamine-dependence of the exchange of 15NH3 into L-glutamine under exchange assay conditions Exchange reactions were performed in 100 mM HEPPS buffer, pH 8, containing 60 mM MgCl2, 10 mM ATP and 100 mM 15NH4Cl (2 mL total volume), and each data point represents an average of duplicate experiments. L-Glutamine (mM) 15N-L-glutamine formation rate ( M/s)

PAGE 58

58 7.8x10-3[14NH3]s-1E+14Gln TE E+Glu E+15Gln 7.8x10-3[15NH3]s-11.06x10-4[H2O]s-15100M-1s-1 Figure 2-7. Exchange of 15NH3 into L-glutamine during the glutam inase reaction catalyzed by AS-B. (A) Kinetic model employed in the simulations of this exchange process. (B) Comparison of the experimentally observed ( ) and simulated ( ) rate of 15NH3 exchange into Lglutamine when AS-B is incubated in 100 mM HEPPS buffer, pH 8, with L-glutamine and 100 mM 15NH4Cl in the absence of ATP and aspartate. 15N/14N exchange rate (mM/s) L-Glutamine (mM) B A

PAGE 59

59 NH2 O O H3N+ OH S-N+H H N H H Enzyme O S O H3N+ OH N+H H N Enzyme O H H NH2 -O S O H3N+ OH NH2 H N Enzyme O +NH3 -O S O H3N+ OH NH2 H N Enzyme O O NH3(tunnel) NH3(solution) Scheme 2-2. Hypothetical mechanism for formation of a thioester intermediate during the ASNScatalyzed hydrolysis of L-glutamine. Note that the N-terminal amino group is thought to function as the general acid/base in the reaction ( 17, 55), although direct evidence for this proposal remains to be obtained in the case of ASNS. The thioester intermediate (TE), which subsequently reacts with water to give glutamate, is drawn within a shadowed box.

PAGE 60

60Table 2-1 AS-B catalyzed incorporation of 15N into L-asparagine in the steady-state competition assays.a a All reactions contained 10 mM MgATP, 20 mM aspartate, 60 mM MgCl2 and 490 nM AS-B in 100 mM HEPPS buffer, pH 8 (2 mL total volume). b As determined by gHMQC NMR spectroscopy. Errors ar e estimated on the basis of measurements employing known concentrations of authentic 15N-Asn. c This value is corrected for the amount of 15N-Asn that would be formed from 15N-Gln at natural abundance, but the contribution of 14N-Asn formed from any 14NH3 released from the enzyme, as a result of the glutaminase activity of AS-B, is assumed to be negligible. d Mean value and standard deviation computed from two separate determinations on duplicate samples using reverse-phase HPLC. e ND not detected; NA not applicable. Gln (mM) 15NH4Cl (mM) 15N-Gln (mM)b 15N-Asn (mM)b,c Total Asn (mM)d 14N-Asn/15NAsn 0 100 NDe 0.83 0.03 0.83 0.03 0 2.5 100 NDe 0.71 0.02 1.22 0.06 0.65 0.08 10 100 0.18 0.01 0.66 0.02 1.30 0.06 1.0 0.1 20 100 0.26 0.01 0.67 0.01 1.4 0.1 1.1 0.2 40 100 0.34 0.01 0.63 0.03 1.4 0.1 1.2 0.2 20 0 NDe NDe 0.76 0.04 NAe 20 25 0.12 0.01 0.29 0.01 0.94 0.03 2.2 0.1 20 50 0.16 0.02 0.44 0.02 1.09 0.05 1.5 0.1 20 75 0.19 0.01 0.62 0.07 1.19 0.07 1.1 0.1 20 100 0.26 0.01 0.67 0.01 1.4 0.1 1.1 0.2

PAGE 61

61 CHAPTER 3 CHARACTERIZATION OF A TUNNEL MUTANT Introduction The crystal structure of C1A mutant of Escherichia coli asparagine synthetase B showed that the two active sites are separated by a 20 distance. The biochemical studies on ASB, with structural evidence, s upport the presence of an intramolecular ammonia tunnel in this enzyme. To investigate the channeling mechanism of ammonia in asparagine synthetase under the structural context, the tunnel re sidues need to be identified and their key ro les in ammonia tunneling need to be elucidated. The tunnel residues were determined by computing the molecular surface of the protein, with a probing radius at 1.4 Do these residues form the real tunnel that ammonia intermediate goes through? Futhermore, sinc e most of them are conserved in different species including b acterial, fungus, plants, insect and mammalian, which indicates they are important to eith er the structure or the function of th e tunnel. Then what kind of roles do they play in ammonia tunneling? To provide some insight to these questions, I performed biochemical studies on ASB. The strategy here is to characterize t unnel mutants in which ammonia transfer is affected by either blockage or leakage of the tunnel. The hypothesis are: if the tunnel is blocked or perforated by mutations without affecting the ac tive sites, glutaminase activity and ammonia dependent synthetase activ ity of the mutants will not be significantly affected, compared to that of wild type, whil e its glutamine dependent synthetase activity will decrease. The putative tunnel is arbitrarily divided in to three parts: the mouth, the body and the bottom. The residues in the mouth and the bottom pa rt are within the active sites. Therefore, only those residues in the body part were considered. Among these residues, Ala388 is one of the conserved residues. Its backbone fo rms the narrow part of the tunnel, with other re sidues. And its

PAGE 62

62 side chain points into the tunne l (Figure 3-1). Since the side ch ain of alanine is a methyl group, mutation of this residue to anothe r one with bigger side chain su ch as Leu may block the tunnel. Based on this hypothesis, the A388L muta nt was prepared and characterized. Material and Methods Materials The experiments were performed as describe d in Chapter 2 Materials and Methods part. The primers of A388L mutant were ordered from idtdna.com. QuikChange Site-Directed Mutagenesis Kit for PCR were from Stratagene. Pfu TurboTM DNA polymerase, 10 reaction buffer, dNTP mix and XL1-blue supercompete nt cells were from the mutagenesis kit. Dpn I restriction enzyme was from New England Bi olab. Wizard Plus Midipreps DNA Purification System for plasmid purification was from Promega. Unless otherwise stated, all chemicals and reagen ts were purchased from Sigma (St. Louis, MO), and were of the hi ghest available purity. 1,3-15N2-Uracil, 15N-L-asparagine, and d6-DMSO were purchased from Sigma-Aldrich (St. Loui s, MO). The isotopic incorporation in these samples was greater than 99%. All experiment s employed freshly prepared solutions of recrystallized L-glutamine (17). NMR spectra were recorded on a Varian INOVA 500 instrument equipped with a 5 mm triple resonance indirect detection probe (z-axis gradients) operating at 500 and 50 MHz for 1H and 15N, respectively. Chemical shifts are reported in ppm relative to sodium 2,2-dimethyl-2-silapentan e-5-sulfonate (DSS). The co-axi al inner cell (catalog: NE-5CIC-V) containing the external NMR standard s was purchased from New Era Enterprises (Vineland, NJ). Primers were designed and analyzed using Ge ne Runner ver3.05 (Hastings Software). The sequences of A388L primers are 5-GAC TGC GCG CGT CTG AAC AAA GCG ATG TC-3 and 5-GA CAT CGC TTT GTT C AG ACG CGC GCA GTC-3 for sense and antisense primer

PAGE 63

63 respectively. The plasmid for mutant expre ssion was prepared by PCR using a pET-21c(+) plasmid with Escherichia coli asnB gene inserted into its Nde I restriction site as the DNA template. The template plasmid contains tota l 7103 bps (5441 bps for empty plasmid plus 1662 bps for asnB gene) with a marker for ampicillin resistance. QuikChange Site-Directed Mutagenesis Kit was used for PCR and cloning. The plasmid with asnB gene mutant was purified using Wizard Plus Midipreps DNA Purification Syst em. Recombinant, A388L ASB mutant was expressed and purified following literature procedures ( 16). Protein concentrations were determined using a modified Brad ford assay (Pierce, Rockford, IL) ( 60), for which standard curves were constructed with bovine serum albumin, and corrected as previously reported ( 55 ). Competition Experiment Reaction mixtures consisted of 100 mM HEPPS buffer (pH 8.0), 60 mM MgCl2, 10 mM ATP, 20 mM aspartate and di fferent concentrations of 15NH4Cl (pH 8.0) or glutamine in total volume of 2 mL. In experiment s where the concentration of 15NH4Cl (pH 8.0) was varied as 10 mM, 25 mM, 50 mM, 75 mM and 100 mM, L-glutam ine was fixed at 20 mM. Alternatively, when L-glutamine was varied as 0 mM, 2.5 mM, 5mM, 10 mM, 20 mM and 40 mM, 15NH4Cl was added at an initial concentration of 100 mM Reactions were initiated by the addition of A388L ASB mutant (30 g) and the resulting samples incubated for 10 min at 37 oC before being quenched by the addition of trichloroacetic acid (TCA) (60 L). After centrifugation for 5 min at 3000 rpm to remove preci pitated protein, the supernatan t was adjusted to pH 5 by the addition of 10 M aq. NaOH and an aliquot of this solution (650 L) transferred to a 5 mm NMR tube for analysis. At higher pH values, amid e NH exchange precluded the derivation of any quantitative relationship between peak area and 15N-Asn concentration. The standard samples contained 100 mM HEPPS buf fer (pH 8.0), 60 mM MgCl2, 10 mM ATP, 100 mM 15NH4Cl (pH

PAGE 64

64 8.0), 20 mM glutamine and diffe rent concentrations of 15N-L-Asn from 0.05 mM to 2.5 mM, without adding ASNS. NMR Measurements and HPLC assay See Chapter 2 Materials and Methods part. Results and Discussion After the A388L ASB mutant was prepared, the kinetic parameters were first determined using enzyme assays (Beeson, unpublished data). Since the pyrophosphate a ssay, which is used for determination of synthetase activity of the enzyme, gave a low kcat value due to a possible pyrophosphatase contamination, the kcat values for both glutamin e dependent and ammonia dependent synthetase activity were reexamined to fit NMR results. As expected, the specificity of A388L mutant to glutamine based on its glut aminase activity changed by less than 2 fold compared to the results for wild type ASB. This change resulted mainly from the decrease of Km values. Its specificity for ammonia based on the a mmonia dependent syntheta se activity had little change. Both kcat and Km decreased a little. For its glutamin e dependent synthe tase activity, the specificity for glutamine dropped by 8 fold, with kcat decreased by 5 fold (Table 3-1). When both 15NH4Cl and glutamine present in reaction mixture, the 14N-asparagine/15Nasparagine ratio for A388L mutant was less than that of WT ASB, with either varied glutamine from 2.5 mM to 40 mM (Figure 3-2A) or vari ed ammonia from 50 mM to 100 mM (Figure 32B). In general, these results are consistent with the result above that glutamine dependent synthetase activity decreased significantly, with no significant change of its ammonia dependent synthetase activity. However, the 14N-asparagine/15N-asparagine producti on ratio catalyzed by the A388L mutant is not proportional to that catalyzed by th e WT enzyme. With 15NH4 + at 100 mM, the 14N-asparagine/15N-asparagine ratio decreased by about 10 folds when the concentration of glutamine was below 10 mM, while with higher glutamine, it decreased by about 4 folds.

PAGE 65

65 When varying ammonium, this ratio didnt cha nge proportionally either. Especially at low concentration of ammonium, its value was close to that of WT ASB. These results may be caused by the overall effect of 1) the asparagi ne inhibition of glutaminase activity; 2) the suppression of the ammonia utilization by gl utamine, and 3) glutamine inhibition. The production of 14N-asparagine by A388L mutant decr eased significantly compared to WT ASB, while the production of 15N-asparagine had no big ch ange, when either varying glutamine or varying ammonium (Figure 33, 3-4). Furthermore, the production of 15N-glutamine increased slightly for the mutant (Figure 3-5), wh ich is consistent with the increased specificity of glutamine for its glutaminase activity. The 14N-asparagine/15N-asparagine ratio, and the production of 14N-asparagine, 15N-asparagine and 15N-glutamine, as well as the kinetic results determined by enzyme assays, support that the mutation of Ala388 to Leu may block the tunnel. Ala388 is in the narrowest part of the tunnel. It is not only at the interface of the two domains, but also one of the non-polar residues close to the glutaminase sites that consists of mainly polar residues (Figure 3-6). Its side chai n, a methyl group, is not likely to involve in interactions with other residues, as aromatic re sidues or charged residues usually do. However, mutation of this alanine to leucine resulted in the possible blockage of the tunnel, which may suggest that this part of the tunnel may act as the gate of the tunnel. Conclusions Based on the tunnel residues de termined by computational me thods, a tunnel mutant was prepared, in which a conserved alanine residue was mutated to leucine. The glutaminase activity and ammonia dependent synthetase activity of this A388L mutant are similar to that of WT ASB, while its glutamine dependent synthetase activit y dropped by about 8 fold. The kinetic studies of this mutant experimentally suppo rt that the identified tunnel residues by computational methods are those forming the intramolecular ammonia t unnel. Furthermore, the NMR studies on this

PAGE 66

66 mutant, together with the kinetic results, suggest that the mutation may result in the blockage of the tunnel.

PAGE 67

67 Figure 3-1. One of the conserved tunnel residues, Ala388. The backbone of A388 forms the narrow part of the tunnel with several other tunnel r esidues. Its side chain points into the tunnel close to the glutaminase site.

PAGE 68

68 0 0.2 0.4 0.6 0.8 1 1.2 1.4 01020304050 [Gln] mM14N/15N ASB A388L 0 0.5 1 1.5 2 2.5 3 3.5 4 020406080100120 [NH4 +] mM15N/14N ASB A388L Figure 3-2. Competition results for A388L ASB mu tant, as well as WT ASB. Filled square: A388L; blank square: WT ASB. In general, the 14N-asparagine/15N-asparagine ratio decreased due to the mutation of alanine to leucine. A) varying glutamine; B) varying ammonium.

PAGE 69

69 0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 01020304050 [Gln] mM15N-Asn (mM) ASB A388L 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 01020304050 [Gln] mM14N-Asn (mM) ASB A388L Figure 3-3. Production of asparagine catalyzed by A388L ASB mutant and WT ASB with varied glutamine. Filled square: A388L; blank square: WT ASB. In general, the 14N-asparagine decreased significantly due to the mutation of alanine to leucine. A) production of 15Nasparagine; B) production of 14N-asparagine.

PAGE 70

70 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 020406080100120 [NH4 +] mM15N-Asn (mM) ASB A388L 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 020406080100120 [NH4 +] mM14N-Asn (mM) ASB A388L Figure 3-4. Production of asparagine catalyzed by A388L ASB mutant and WT ASB with varied ammonium. Filled square: A388L; blank square: WT ASB. In general, the 14N-asparagine decreased significantly due to the mutation of alanine to leucine. A) production of 15Nasparagine; B) production of 14N-asparagine.

PAGE 71

71 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 01020304050 [Gln] mM15N-Gln (mM) ASB A388L 0 0.05 0.1 0.15 0.2 0.25 0.3 020406080100120 [NH4 +] mM15N-Gln (mM) ASB A388L Figure 3-5. The nitrogen exchange catalyzed by A388L mutant and WT ASB. Filled square: A388L; blank square: WT ASB. In general, the production of 15N-glutamine increased due to the mutation of alanine to leucine. A) varying glutamine; B) varying ammonium.

PAGE 72

72 Figure 3-6. The polarity of the tunnel residues. Yellow: non-polar residues; green: polar residues; blue: basic residues; red: acidic residues. Both AMP (bottom) and glutamine (top) showed in balls.

PAGE 73

73 Table 3-1. The kinetic parameters for glutaminase activi ty, ammonia dependent synthetase activity and glutamine dependent synthetase activity of both wild type ASB and A388L mutant WT ASB A388L mutant kcat (s-1) Km (mM) kcat/Km (M1s-1) kcat (s-1) Km (mM) kcat/Km (M1s-1) Glutaminase activity 3.38 1.67 2000 3.12 0.87 3600 Ammonia dependent synthetase activity 3.19 17 190 2.34 15 160 Glutamine dependent synthetase activity 2.94 0.69 4300 0.57 1.06 540

PAGE 74

74 CHAPTER 4 ASPARAGINE INHIBITION TO ASPARAGINE SYNTHETASES AND ITS EFFECT ON THE QUATERNARY STRUCT URE OF THE ENZYMES Introduction Glutamine-dependent (EC 6.3.5.4) asparagine s ynthetase belongs to amidotransferase superfamily(5, 6), and was found in all kinds of speci es including archaeon, bacteria, fungi, plant, and animals ( 13). Based on the structure and func tion of its glutaminase domain, asparagine synthetase was classified as class II amidotransferase, which consists of three other amidotransferases, glutamine PRPP amidotrans ferase, glucosamine 4-phosphate synthase, and glutamate synthase. This enzyme has two domains, glutaminase domain and synthetase domain, which are connected by a single loop ( 15). It catalyzes ATP-dependent synthesis of L-asparagine from L-aspartic acid using either glutamine or ammonium as nitrogen source. At glutaminase site, the amide nitrogen of glutamine is hydr olyzed into ammonia intermediate through a thioester enzyme-substrate complex, with glutamate as one of the products ( 14 ). Ammonia is then transported through an intramolecular tunnel to synthetase site and reacts with -AspartylAMP, an intermediate that formed by ATP and Asp at synthetase site to produce asparagine. Alternatively, this enzyme can also use free a mmonia as nitrogen donor. Unlike other class II amidotransferases ( 54, 66-68), the hydrolysis of glutamine is not strictly coupled with the synthesis of final product in vitro ( 17). Without ATP and Asp, as paragine synthetase can consume glutamine and release toxic ammonia. Furthermore, the glutaminase activity of bacterial enzyme is even stimulated by the presence of ATP without producing asparagine (17). These observations raise the interest in the re gulation of glutaminase activity of asparagine Li designed the inhibition experiments. Inhibition to ASB part was performed by Li and an undergraduate, William Beeson. Inhibition to hASNS was perfor med by William Beeson himself. Li and Beeson did the kinetic analysis. Li did the simulation. Li designed and performed the SEC experiments and analyed the results.

PAGE 75

75 synthetase in vivo to prevent the waste of glutamine and the releasing of free ammonia. Kinetic studies of asparagine synthetase have been extensiv ely performed in the last 20 years, especially for Escherichia coli enzyme (ASB) ( 2, 3, 7, 8, 11, 22, 24, 35, 57, 69-72). The glutaminase activity of asparagine syntheta se can be affected by chloride ions ( 72), ATP or asparagine ( 22, 24). With a KI value of about 100 M, asparagine may play an important role in preventing unnecessary hydrolysis of glutamine under physiological conditions. Therefore, we investigated the possible mechanisms for aspara gine inhibition to the glutaminase activity of Escherichia coli ASB and hASNS from both structural and kinetic aspects. Our results showed that asparagine inhibited the hydrolysis of glutamine by ASB and hASNS through different mechanisms and had different effect on the qua ternary structure of these two enzymes. Materials and Methods Materials Unless otherwise stated, all chemicals and reagen ts were purchased from Sigma (St. Louis, MO). L-Glutamine was recrystallized prior to use as previously described ( 53). Wild-type Escherichia coli ASB and recombinant, C-terminally tagged human AS were expressed in and purified from a baculo virus/insect cell e xpression system as previously described ( 16, 57 ). Enzyme Assays Kinetic constants were determined by in cubating purified, recombinant asparagine synthetase in reaction mixtures in which one substrate was varied and all others were at saturating concentrations. Asparagine synthetase was added to 200 L reaction mixtures containing 100 mM HEPPS, pH 8.0, 100 mM NaCl, 8mM MgCl2, and varying amounts of Lglutamine (0-50 mM). The reactio n mixture was incubated at 37 oC for 10 minutes and then quenched with 30 L of 20% w/v trichloroacetic acid. Each quenched reaction mixture was added to a premixed 770 uL solution containi ng 300 mM glycine, 250 mM hydrazine, pH 9.0,

PAGE 76

76 1.5 mM NAD+, and 1 mM ADP. Glutamate pr oduction was assayed by adding 2 L of Lglutamic acid dehydrogenase to the mixture and monitoring abso rption spectrophotometrically at 340 nm for 60 minutes. Measuremen ts at specific substrate concentrations were performed in duplicate or triplicate, and the initial rate data was analyzed by curve-fitting using computerbased methods to determine the Vmax and KM constants. Asparagine Inhi bition Studies Product inhibition studies were carried out on the wild-type E. coli ASB (3.1 g) and recombinant, C-terminally tagged hAS (3.0 g) using the enzyme a ssay described above, but with a fixed amount of asparagine (0-0.5 mM fo r ASB and 0-1.0 mM for hASNS) added to each run. The apparent KM and Vmax values were analyzed by curve-fitting using standard computerbased methods. Size Exclusion Chromatography to Determin e the Quaternary Structure of Asparagine Synthestases. Gel filtration chromatography was performed on Rainin Dynamax HPLC system using a Phenomenex BIOSEP SEC-S2000 column (300 x 7.8 mm with 75 x 7.8 mm guard column, particle size: 5 m; pore size: 145 ) with detection wave length at 280nm and flow rate at 1.0 mL/min. The molecular weight is calculated base d on a standard curve of a series of marker proteins with different molecula r weights purchased from Sigma. Unless stated otherwise, stock enzymes for Escherichia coli asparagine synthetase B (ASB) or human asparagine synthetase were directly injected for analysis with a volume of 5 L. The Effect of Asparagine on the Quaternary Structure of ASB and hAS The enzyme was analyzed as described above using a mobile phase that contained 50mM Tris-H2SO4 buffer at pH 7.0 with different concentrati ons of asparagine (0-5.0 mM). All sample points have at least two repeats.

PAGE 77

77 SEC Studies of ASB with Different Amount of ASB The enzyme was analyzed by SEC using a mo bile phase that c ontained 50mM Tris-H2SO4 buffer at pH 7.0 with different injecti on volume of 3.1 mg/mL stock solution (2.5-40 L). Results Product Inhibition The glutaminase activity of AS B was studied with varying am ounts of asparagine present (0.025 0.5 mM). The apparent KM for glutaminase reaction increased with increasing asparagine concentration. Figure 41A shows a replot of apparent KM respect to glutamine vs concentration of asparagine. The increase of KM with increasing concentra tion of inhibitor results from the binding of asparagine to the free enzyme However, the regression line is not linear but more like polynomial. The presence of asparagine in the reaction mixture affected not only the apparent KM value, but also the kcat value of the glutaminase r eaction, which decreased as the concentration of asparagine was increased (Figure 4-1B). This result provides the evidence that asparagine binds to enzyme-substrate complex. Th erefore, asparagine is a mixed inhibitor to ASB with respect to glutamine. The glutaminase activity of human enzyme was also studied with varying amounts of asparagine present (0.05 1.0 mM). The inhibition results showed clearly that asparagine is a competitive inhibitor to glutamine. The maximum glutaminase activity stayed constant with a value of 3.6 s-1 in term of monomer. And the apparent KM versus concentration of asparagine is a straight line with determined inhibition constant as 0.11 0.2 mM (Figure 4-1C). Asparagine Affects the Quaternary Stru cture of ASB, But Not Human Enzyme The nonlinear relationsh ip of apparent KM for ASB vs concentration of asparagine indicates more than one binding site for asparagi ne. In the crystal structure of ASB C1A mutant (PDB code: 1CT9), the enzyme exists as dimer form with glutamine bound to it ( 15). Each

PAGE 78

78 subunit has two separated active s ites, glutaminase site and synthetase sites. The distance between the two sites is about 20 For better understanding of the in hibition mechanisms to ASB by asparagine from the structural point of vi ew, we studied the quaternary structure of the enzyme by SEC with or without the presence of asparagine in the mobile phase. It was reported that the activity of asparagine synthetases was affect ed by chloride ion and pH ( 72). There was no significant difference for the SEC results of asparagi ne synthetases using either Tris-HCl or Tris-H2SO4 at pH 7.0, or within the pH range of 7.0-7.9 (unpublished data). Here we choose Tris-H2SO4 at pH 7.0 as the mobile phase. The molecular weight (MW) was calculated using a standard curve (Figure 4-2). Using 50 mM Tris-H2SO4 buffer at pH 7.0, two peaks were observed in the size exclusion chroma togram of ASB, with retention times of 9.15 and 9.51 minutes respectively. The calculated MWs for these two peaks are 120kD and 84kD respectively (Figure 4-3A). After adding asparagine in the mobile phase, the retention time of the higher MW ASB peak changed, which indicates as paragine affects the qua ternary structure of ASB. With the presence of aspara gine in the mobile phase, two peaks with RT at 8.79 and 9.49 minutes were observed. The calculated MWs for the two peaks are 171kD and 85kD respectively, with higher MW for the peak with shorter retention time (Figure 4-3A). Under experimental conditions, the retention times of these two peaks were independent of the concentration of asparagine adde d, with the lowest concentrati on at 0.1 mM (Table 4-1). The true MW of one subunit of ASB is about 62.5kD based on its gene sequence. Therefore, the 9.50 minutes peak (calculated MW: ~85 kD) with or without asparagine is very likely the monomer form of ASB, with the 8.79 minutes peak (calculated MW: 171 kD) the dimer form (Table 4-2). And the intermediate peak at 9.15 minutes (calculated MW: 120 kD) without the presence of asparagine may result from quick exchanges betw een the forms. That is, equilibrium exists

PAGE 79

79 between the monomer and dimer form. This conc lusion was supported by the SEC results using different amount of enzyme without aspara gine in mobile phase (Figure 4-4). Besides the retention time, asparagine also affects the individual peak area and height. While the total area of the monomer and dime r peaks had almost no ch ange with different concentrations of asparagine in the mobile phase the dimer peak area and height increased with increasing asparagine, which supports that asparagine thermodynamically stabilizes the dimer form. The plot of percentage of dimer peak area in total ASB peak area vs. concentration of asparagine behaves like a Michaelis-Menten curv e (Figure 4-5). At low concentrations of asparagine, the peak area of dimer ASB and the peak height ratio increase rapidly with the increasing concentration of aspara gine. At high concentrations of asparagine, the curve gradually levels off and indicates the saturating status. Th e estimated binding constant of asparagine is about 0.10 mM. In our results, a monomer peak was always obser ved. In other words, ASB did not completely converted to dimer at high con centration of asparagine (5 mM). One possible reason is that partial protein has no chance to collide with each other and form dimer during elution. Regardless of this, we can conclude that with asparagine present in the mobile phase, both dimer and monomer peak showed in the chromatogram and the dimer is stabilized by asparagine binding at the experimental conditions. Under the same conditions, the quaternary structure of hASNS was studied without asparagine or with 5 mM asparagine in the mob ile phase. Most of enzyme exists as dimer form with retention time at 8.75 minutes. With 5 mM asparagine, no detectable changes were observed for the retention time and the peak area (Figure 4-3B). The Equilibrium between the Dimer and Monomer Form of ASB. When the mobile phase contains no aspara gine, the SEC chromatogram showed one

PAGE 80

80 monomer peak and one intermediate peak, while no dimer peak. Based on these results, we came up the idea that the ASB monomer and dimer fo rm equilibrium at experimental conditions. However, this peak may results from the overl apping of the two forms of enzyme peaks. To confirm our hypothesis, we performed SEC studi es using different amount of the enzyme without the presence of asparagine in the mobile phase. The ratio of both peak area and peak height for the two peaks (intermediate/monomer) increased with more enzyme injected. Furthermore, the retention time of intermediate peak is closer to that of dimer peak. All these resu lts support that more dimer formed at higher concentration of enzyme and the two forms of enzyme are in a dynamic equilibrium (Figure 4-4). Construction of Asparagine Inhibition Model The inhibition model for ASB (scheme 4-1A) was constructed based on the following considerations: 1. The dimer form is the activ e form. The active form of ASB was seldom addressed in previous studies. To date, the only 3D structural in formation of ASB was reported by Larsen et al(15 ). In their studies, the C1A mutant of ASB is a homogenous dimer with one glutamine bound to each glutaminase site. Since the mutated residue is at the end of N-terminal domain and plays catalytic role rather than structur al role, we can consider this structure is close to that of the wild type enzyme. Therefore, it is reasonable to assume that the dimer is the active form of the enzyme. Furthermore, this hypothesi s was supported by our result that glutamine also stabilizes the dimer form of ASB like asparagine (unpublished data). Because the dimer form of ASB was stabilized at low concen tration of asparagine and its su bstrate glutamine has the similar effect on its quaternary st ructure, the amount of dimer form can be considered constant at most of the experimental conditions with both asparagine and glutamine present. Therefore, although our results showed equilibrium exists between the two forms of enzyme, we didnt include the

PAGE 81

81 equilibrium part in the model for the purpose of simplicity. 2. One subunit of asparagine synthetase has two possible aspa ragine binding sites, glutaminas e active site and synthetase active site because the space-filled model of the crystal structure of C1A mutant showed no third cavities that may bind the third asparagine molecule. Hence, the dimer form of asparagine has four possible asparagine binding si tes. 3. The first asparagine binding site to the free enzyme is a glutaminase site. The first aspara gine binding site is not likely to be at the synthetase domain because ammonia dependent synthetase activity is much less affected by 1 mM of asparagine than that to the inhibition glutamine dependent synthetase activity. This is also supported by the fact that asparagine is an analog of glutamine. 4. The result that kcat value decreased with increasing concentrations of asparagine (Figure 4-1B) indicates th at asparagine binds to enzymesubstrate (EEGln) complex. No matter how many asparagine binding sites there are in one molecules of free enzyme, the apparent kcat can be deduced as: cat cat cat catk kk Ik kk Ik k 4 3 4 3][ 1 ][ 1 Where is a factor resulting from the effect of as paragine binding on the turn over number for the reaction from AsnEEGln complex to glutamate; k3 andn k4 are the forward and backward rate constants for the bindiing of asparagine to en zyme-substrate complex (EEGln) respectively. By fitting 4 3kk kcatand catk to the experimental results using KaleidaGraph, we got value close to zero (0 0.3 mM). This means that the AsnEEG ln complex is hardly turn over to glutamate. Therefore, we can simplify the model as showed in scheme 4-1B. The corresponding apparent kcat is:

PAGE 82

82 cat Ii cat catk K I k k Ik k ][ 1 1 ][ 1 1 '4 3 Where KIi is the dissociation constant of AsnEEGln complex into asparagine and EEGln, which was fitted as 0.33 0.04 mM; kcat was fitted as 7.4 0.2 s-1 for dimer form, consistent with the previously reported value. 5. Based on the results that apparent KM value versus the concentration of asparagine is positively related but not a straight line, several possible models were proposed regarding the number of asparagine molecule binding to the free enzyme (Scheme 4-2). If there is only one asparagine binding site in the free enzyme (Scheme 4-2A), the apparent KM versus the concentration of asparagine is a hyperpola when KIs is different than KIi. M Ii Is MK K I K I K ][ 1 ][ 1 If more than one asparagine binding s ites in the free enzyme, the apparent KM versus the concentration of asparagine is more like a polynomial line with second order for two binding sites model (Scheme 4-2B), third order fo r three sites (Scheme 4-2C), and so on. M Ii IsIs Is MK K I KK I K I K ][ 1 ][][ 1 '21 2 1 M Ii IsIsIs IsIs Is MK K I KKK I KK I K I K ][ 1 ][][][ 1 '321 3 21 2 1 Among these models, the best fit to the experiment al results is the thr ee sites model. The two sites model fits the data with the concentrat ion of asparagine below 0.2 mM, not for higher two binding sites model three binding sites model

PAGE 83

83 concentration, which implys the th ird order term. We could not directly get all three dissociation constants by fitting except KIs1 (0.09 0.01 mM) and KIs1KIs2KIs3 (0.0039 0.0002 mM3). However, we can estimate KIs2 and KIs3 using the fitted value of KIs1. The final value for KIs2 and KIs3 are 0.7 0.2 mM and 0.06 0.02 mM respectively. The model for asparagine inhibition to hASNS is much simpler since it is a competitive inhibitor to the enzyme. Only one molecule of asparagine can bind to the free enzyme. And it doesnt bind EGln complex. The inhibition constant was determined as 0.11 0.02 mM. Discussion As mentioned above, the crysta l structure of C1A mutant of ASB suggests the active form of this enzyme is dimer. This is supported by our results from chromatography studies, in which the binding of glutamine drove th e enzyme to the dimer form, similar to asparagine. Based on this, we constructed the kinetic model for the as paragine inhibition to ASB. Our kinetic results and the simulated model tells us the apparent Km values of ASB for glutamine was affected significantly by the presence of as paragine and this may results from one molecule of free enzyme can binds at least three asparagine molecules. Consider the fact that no other possible asparagine binding site was found in the crystal structure of C1A mutant except the two active sites of each subunit, the simulated model further support that the active form of ASB is a dimer. Communication between domains and/or subunits and allosteric re gulation is a common phenomenon for biological macromolecules. Previ ous studies showed that binding of ATP at synthetase domain of ASB st imulates its glutaminase activity, indicating a communication between its two catalytic domains( 17). In this study, asparagine st abilized the dimer form of ASB and drove the monomer/dimer equilibrium to the dimer side tells us that the interaction between subunits became str onger with asparagine binding to the enzyme, implying conformational changes and possible communication at the interface. This is also supported by

PAGE 84

84 the three binding sites model for asparagine inhib ition since all these asparagine binding sites are impossible in the same subunit of ASB. The i nhibition mechanism of hASNS by asparagine is different from that of ASB. Aspargine is a simple competitive inhibitor for hASNS under the same experimental conditions. The SEC results showed hASNS is present as dimer at the experimental condition. Its quatern ary structure was not affected by 5 mM asparagine, which is about 50 times of inhibition consta nt. Therefore, the two subunits work independently and do not communicate with each other. In order to find possible structural basis for the communicati on during asparagine inhibition, we checked those interacting residues in the interface of the two subunits. The major interaction between the two subunits of ASB is that the guanidium group of Arg334 forms Hbonds with a loop that closes to glutaminase active site (Figure 4-6). This loop includes a cluster of absolutely conserved amino acid residues (H29R30G31P32D33), which involves in formation of ammonia tunnel. The backbone of G31 and P32 a nd the side chain of S35 interact with the 3 nitrogen atoms in the guanidium group of R3 34 (Figure 4-6A). Close to Arg334, two other residues from synthetase domain and one from gl utaminase domain may also play a role in the interaction of the two subunits Arg334 is followed by anothe r positive residue, Lys335. Its amine group does not contact with the HRGPD l oop, but exposed to bulk solution and interact with several water molecules. However, the indo le ring of Trp34 from glutaminase domain of the other subunits inserts into the two long side chains of Arg334 a nd Lys335. With another indole ring from Trp395, a sandwich structure is formed with the side chain of Arg334 in the middle, probably helping Arg334 position correctly and forming hydrogen bonds with HRGPD loop, especially Trp395 with the indole ring parallel to the side chai n of Arg334. All these residues are in highly conserved sequence region for all kinds of species, from bacteria l to fungus, from plant

PAGE 85

85 to mammals (Figure 4-7). It is also impor tant to point out that the HRGPD loop has XGPXXXGX sequence. The two glycin e residues make this loop more flexible to adopt different conformations. Based on sequence alignment, most of the species uses the same set of residues as Escherichia coli from bacterial to plant, while mamm als have different residues at the corresponding positions. In mammalian species, Tr p34 is replaced by an Ala. This small side chain causes less steric hindrance compared to the indole ring of Trp34 and makes the dimer form more favorable for hASNS. Interestingly, a lysine residue in mammalian enzyme is at the corresponding position of Arg334 in ASB. The one carbon shorter side chain may suggest a closer contact between the two subunits in human enzyme. Consequently, the tryptophan 395 is changed to a smaller aromatic residue, histidin e. In the HRGPD loop, no second glycine residue is present for mammalian enzyme, which has less conformational flexibility. All these differences may provide a structural basis for the different quaternary structures of ASB and hASNS and allosteric effect upon asparagine binding of ASB. Using the inhibition models we constructed above, the apparent activity was simulated with the concentration of aspara gine and glutamine varying from 0.1 mM to 2 mM for both ASB and hASNS (Figure 4-8). This ranged of concen tration was selected ba sed on the physiological concentration of glutamine and asparagine in blood, which is about 0.6 mM and 0.1 mM respectively. Here for the conve nience of comparison with the gl utaminase activity without any inhibition, the enzyme activity w ith no asparagine was also simulated, although it is impossible for cell to have no asparagine at all. The hydrolysis of glutamine is depr essed significantly for both enzymes, with more inhibition effect on AS B. At concentration of 0.6 mM glutamine and 0.1 mM asparagine, about 55% of enzyme activity is lost because of product inhibition for ASB and 40% for hASNS. This means even at nor mal conditions, the glutaminase activity of

PAGE 86

86 asparagine synthetase is suppre ssed significantly. If the asparagi ne concentration is above 0.3 mM, less than 10% activity will be left for ASB and less than 30% for hASNS. The presence of asparagine prevents the enzyme from hydrolyzi ng glutamine at its maxi mum rate and producing less toxic ammonia. As a regula tor, asparagine regulates not only the function of asparagine synthetase, but also the expres sion of the corresponding gene ( 73). In some prokaryote cells, such as Escherichia coli asparagine can be synthesized by ammonia dependent asparagine synthetase using ammonia instead of glutamine. Experimental results showed that ammonia utilizing enzyme is also regulated by asparagine through similar strategy ( 74, 75 ). Like glutamine, asparagine is one of the principa l and non-essential amino acids involved in the storage and transport of nitroge n. The balance between glutamin e and asparagine is directly affected by this enzyme. Recently, the regulato ry patterns of asparagine synthetases in Helianthus annuus (sunflower) were reported by Herrera-Rodrguez et al ( 76-79). Three genes encode asparagine synthetase, HAS1, HAS1.1 and HAS2. The expression of these genes is regulated by light, carbon and nitr ogen availability with different sensitivities. The asparagine level is higher than glutamine during germination and cotyledon senescence. During these stages, most of asparagine synthetases was encoded fr om HAS2. For leaf senescence, all three genes work together and the asparagine level is lower than glutamine. These results imply the importance of asparagine synthetase in keeping the relative asparagine and glutamine level. Conclusions Product inhibition of asparagine synthetase is a common strategy in all species to regulate the conversion from glutamine to asparagi ne. However, the inhibition mechanism for Escherichia coli ASB is different from that of hASNS. The multi-site inhibition to ASB by asparagine and the changes of its quaternary structure afte r asparagine binding suggest conformational changes at the interface of the two subunits and communications between them.

PAGE 87

87 Under same conditions, hASNS showed no such properties. At room temperature, ASB and hASNS showed different quaternary structure. The former is a mixture of dimer and monomer with equilibrium between these two forms, while the latter presents as dimer. This structural difference may result from the different residues at the interface of the subunits. The flexible conformation of ASB provides the structural basis for allosteric communication. Glutamine and asparagine are the two major sources to storage and transport nitrogen in vivo The regulation of the reaction that synthesizes asparagine from glutamine by asparagine inhibition may have important physiological meanings.

PAGE 88

88 Scheme 4-1 EE + Gln + n Asn EEAsnn+ Asn AsnEEGln EE + Glu EEAsn + Gl u EEGln k1k2kcatkcatKI' k4k3 EE + Gln + n Asn EEAsnn+ Asn AsnEEGln EE + Glu EEGln k1k2kcatKI' KIi A B

PAGE 89

89 Scheme 4-2 EE + Gln + Asn EEAsn + Asn AsnEEGln EE + Glu EEGln k1k2kcatKIsKIi EE + Gln + Asn EEAsn AsnEEAsn + Asn + Asn AsnEEGln EE + Glu EEGln k1k2kcatKIs1KIs2KIi EE + Gln + Asn EEAsn AsnEEAsn + Asn + Asn AsnEEGln EE + Glu EEGln k1k2kcatAsnEAsnEAsn + Asn KIs2KIs3KIiKIs1 A B C

PAGE 90

90 0 5 10 15 20 25 30 00.10.20.30.40.5 Figure 4-1. Kinetic studies of aspara gine inhibition to ASNS A. apparent KM (mM) vs [Asn] mM apparent KM (mM) L-asparagine (mM) apparent KM (mM) L-asparagine (mM) 0 5 10 15 20 00.20.40.60.81

PAGE 91

91 1 1.5 2 2.5 3 3.5 4 00.10.20.30.40.5 Figure 4-1 B. apparent kcat (s-1) vs [Asn] mM apparent kcat (s-1) L-asparagine (mM)

PAGE 92

92 0 10 20 30 40 50 60 70 80 00 511 52 [Asn]: 0, 0.05, 0.1, 0.25, 0.5, 1 mM Figure 4-1 C. 1/v vs [Gln] mM-1 1/v 1/[Gln] (mM-1)

PAGE 93

93 4 4.5 5 5.5 6 0 0.1 0.2 0.3 Figure 4-2. SEC standard curve: log(MW) vs KD oc oe DVV VV K Ve is elution volume; Vc is column volume; Vo is void volume log(MW) KD

PAGE 94

94 Figure 4-3. SEC analysis of quaternary structure of ASNS with different c oncentration of asparagine in mobile phase A. SEC analysis of quaternary structure of ASB

PAGE 95

95 Figure 4-3 B. SEC analysis of quaternary structure of hASNS with (black) or without 5 mM asparagine (red) in mobile phase

PAGE 96

96Table 4-1. Retention time of ASB peaks in SEC studies The peaks of ASB and their retention time (RT) with/without asparagine in mobile phase [Asn] mM R.T. (min) Peak 1 Average (min) Peak2 Average (min) Peak3 Average (min) 0 9.410 9.51 0.06 9.086 9.15 0.05 0 9.485 9.098 0 9.458 9.131 0 9.551 9.180 0 9.596 9.225 0 9.551 9.200 0 9.498 9.146 0.10 9.561 9.49 0.05 8.861 8.79 0.05 0.10 9.556 8.873 0.25 9.546 8.831 0.25 9.516 8.801 0.50 9.491 8.771 0.50 9.481 8.768 1.0 9.520 8.800 1.0 9.493 8.788 5.0 9.416 8.726 5.0 9.410 8.748 5.0 9.425 8.713

PAGE 97

97Table 4-2. The calculated MW and predicted structure of ASB corresponding to each peak Mobile phase With Asn Without Asn With/without Asn R.T. (min) 8.79 9.15 9.50 Calculated MW (kD) 171 120 85 Possible structure Dimer Quick exchange between dimer and monomer form Monomer True MW (kD) 125 62.5 Calc. MW/true MW 1.37 1.36

PAGE 98

98 Figure 4-4. SEC analysis of quaternary structure of ASB with different amount of enzyme injected (2.5-40 L)

PAGE 99

99 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8 012345 Figure 4-5. Fraction of dimer vs concentration of asparagine fraction of dimer L-asparagine (mM)

PAGE 100

100 Figure 4-6. The interface of the two subunits A B

PAGE 101

101 Figure 4-7. Sequence alignment of the interface residues

PAGE 102

102 0.1 0.3 0.5 0.7 0.9 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75v e l o c i t y (s1)[G l n ] mM[ As n ] m M 0.1 0.4 0.7 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75v e l o c i t y ( s1)[ Gl n ] m M[ A s n ] mM Figure 4-8. The glutaminase activity of ASNS around physiological conditions

PAGE 103

103 CHAPTER 5 THE UTILIZATION OF DIFFERENT NI TROGEN SOURCES BY ASPARAGINE SYNTHETASE: EVOLUTION FROM Escherichia coli TO HUMAN Introduction Glutamine-dependent asparagine synthe tase (EC 6.3.5.4) catalyzes ATP-dependent synthesis of asparagine from aspartate, utilizing either glutamine or ammonia as nitrogen source. For more than 15 years, the mechanisms of enzymatic reaction and ammonia tunneling by asparagine synthetase was extensively studied using asparagine synthetase B from Escherichia coli Furthermore, due to the difficulties in expres sion of human asparagine synthetase (hASNS), ASB also served as a model system for designing inhibitor of asparagine synthetase, since the amount of human asparagine synthetase is directly related to th e resistance of leukemia cell to the asparaginase chemotherapy. So far, only one crystal structure was resolved for the C1A mutant of ASB. It is a homodimer with two active sites in each s ubunit. The two sites, glutaminase site and synthetase site, are separate d with an intramolecular tunnel that is about 20 long. The way substrates bind to the active sites is like from back door and front door. Glutaminase site is a four layer sandwich structure, which is common in the class II glutamine dependent amidotransferase. The synthetase site consists mostly of -helices. Human enzyme has 37% identity to ASB based on their primary structure. And most of the functional residues, including those in the active sites a nd those tunnel residues for ammonia channeling, are conserved in both enzymes. Both hASNS a nd ASB are supposed to have same catalytic mechanism and overall structure. On the other ha nd, since they are from different sources, one from bacterial, the other from human, the two enzymes are evolutionari ly different from each other. With more studies on hASNS after it was successfully prepared using a baculovirus expression system in Richards lab ( 57), more results showed the differences between them. In

PAGE 104

104 previous chapter, studies about asparagine inhi bition showed different regulatory mechanisms to the glutaminase activity of hASNS and ASB. Asparagine inhibits the gl utaminase activity of hASNS by a simple competitive mechanism to glut amine, while it regulates the glutaminase activity of ASB through a cooperative binding mech anism. The first binding constant is close to that of hASNS, implying asparagine binds to the same active site first in both enzymes. Consequently, asparagine has higher inhibition effect on ASB due to consecutively binding of three asparagine molecules to the enzyme. Sin ce ASB is a homodimer and each subunit has only two possible asparagine binding sites, the cooperative binding of three asparagine molecules suggests the communications between the two subun its, as well as between the two domains. In the latter case, the bind ing of asparagine to the glutaminase active site affects synthetase site. Furthermore, ATP itself, or ATP with aspartate, stimulates the glutaminase activity of ASB, which support the id ea that the two catal ytic domains communicate with each other and substrate binding to synthetase site resu lts in the changes in the glut aminase site. Therefore, the communication between the two domains in AS B is two-way effect. The studies on the quaternary structure of ASB showed that the asparagine binding s timulates dimerization, implying the conformational changes for the inte rface residues between the two subunits. This may provide the structural basis for the in ter-subunit communication. As for hASNS, its glutaminase activity is furt her decreased with both ATP and aspartate present (Beeson, unpublished results). The communication between the two domains is from synthetase site to glutaminase site. The results about aspara gine binding showed no communication from glutaminase site to synthetase site. This is also supported by the fact that asparagine binding did not change the quaternary structure of human en zyme. In this Chapter, I investigated the utilization of two different nitrogen sources, gl utamine or ammonium, by hASNS using the new

PAGE 105

105 developed NMR assay. Compared with the results of ASB, I will disc uss the evolutionary differences between hASNS and Escherichia coli enzyme (ASB). Material and Methods The experiments were performed as describe d in Chapter 2 Materials and Methods part. Materials See Chapter 2 Materials and Methods part. R ecombinant, wild type hASNS was expressed and purified following literature procedures ( 57 ). Competition Experiment Reaction mixtures consisted of 100 mM HEPPS buffer (pH 8.0), 60 mM MgCl2, 10 mM ATP, 20 mM aspartate and di fferent concentrations of 15NH4Cl (pH 8.0) or glutamine in total volume of 2 mL. In experiment s where the concentration of 15NH4Cl (pH 8.0) was varied as 10 mM, 25 mM, 50 mM, 75 mM and 100 mM, L-glutam ine was fixed at 20 mM. Alternatively, when L-glutamine was varied as 0 mM, 2.5 mM, 5mM, 10 mM, 20 mM and 40 mM, 15NH4Cl was added at an initial concentration of 100 mM Reactions were initiated by the addition of hASNS (18 g) and the resulting samples incubated for 10 min at 37 oC before being quenched by the addition of trichloroacetic acid (TCA) (60 L). After centrifugation for 5 min at 3000 rpm to remove precipitated protein, the supernatant was adjusted to pH 5 by the addition of 10 M aq. NaOH and an aliquot of this solution (650 L) transferred to a 5 mm NMR tube for analysis. At higher pH values, amide NH exchange precluded the derivation of any qua ntitative relationship between peak area and 15N-Asn concentration. The standard samples contained 100 mM HEPPS buffer (pH 8.0), 60 mM MgCl2, 10 mM ATP, 100 mM 15NH4Cl (pH 8.0), 20 mM glutamine and different concentrations of 15N-L-Asn from 0.05 mM to 2.5 mM, without adding hASNS. NMR Measurements, HPLC assay and Kinetic Simulations See Chapter 2 Materials and Methods part.

PAGE 106

106 Results and Discussion When both ammonium and glutamine are presen t in the reaction mixture, the two nitrogen sources compete with each other for the -aspartyl-AMP intermediate to produce asparagine. If 15NH4 + and 14N-glutamine are used, the ab ility of utilization of different nitrogen source by hASNS and the efficiency of ammonia transfer can be inve stigated by quantification of 15Nasparagine, produced from 15NH4 +, and 14N-asparagine, from glutamine, using the new developed NMR assay, combining with HP LC assay reported before (Fig. 5-1). As shown in Table 5-1, the incorporation of 15N into asparagine is increased with more 15NH4 + present in the solution, with constant concentration of glutamine (20 mM). The 14Nasparagine/15N-asparagine ratio dropped down from 4.6 to 0.4. When the concentration of 15NH4Cl was fixed as 100 mM, at low concentration of glutamine, the 14N-asparagine produced increased with glutamine. Unfortunately, due to the low amount of 14N-Asn, the error is big. However, we can see the positiv e relationship between the concen tration of glutamine and the 14N-Asn/15N-Asn ratio. While at high concentration of glutamine, the ratio droped down, which suggests that the production of 14N-Asn slowed down, based on the results that ammonia dependent synthetase activity keep constant. This may result from the inhibition to the glutaminase acitivity of this enzyme. Both glutam ine and asparagine are potential inhibitors. The glutamine inhibition (substrate inhibition) was oberserved when I determined kinetic parameters for its glutaminase activity by varing the concentration of glutamine (0-80 mM), which suggest that a second molecule of glutamine binds to the enzyme and inhibits its glutaminase activity. Previous work also showed that asparagine is a good inhibor to its glut aminase activity with KI value of 120 M. However, under 1 mM concentr ation, the inhibition appears competitive to glutamine and didnt affect kcat value. Since the total production of asparagine is less than 0.5

PAGE 107

107 mM, and relatively small compared to the concentra tion of the substrate, th is inhibition is very likely caused by glutamine inhibition. As described in Chapter 2, if ammonia can not access the active site with the presence of glutamine, all the as paragine would be 14N product at saturating concentration of glutamine. In this case, free ammonia is suppre ssed by glutamine. The ratio of 14N asparagine to 15N asparagine would be infinite. This ratio will be decreasing w ith the extent of suppression becomes less. When there is no suppression at all, the 14NH3 will compete with 14NH3 freely. If ammonia tunneling is fast and effi cient enough, then the ratio of 14N asparagine to 15N asparagine would be determined by the ratio of production rates using ammonia and glutamine. That is, 3 3 3][][ ]][Gln[ ]][[ ]][Gln[ ]][)[/( ]][Gln)[ /( ][ ][3 15 ln 3 15 ln 3 15 ln 15 14 15 14NHi G NH G NH bindingmcat G tunneling mcatENH E ENH E ENH Kk E Kk v v AsnN AsnN Therefore, 3][][ ]ln][[ ][ ][3 15 ln 15 14NHi GENH EG AsnN AsnN or ln 3 15 14 15]ln][[ ][][ ][ ][3G NHiEG ENH AsnN AsnN Where, 3][NHE is the enzyme form that can bind e xogenous ammonia and produce asparagine. ln][GE is the enzyme form that can bind glutamine and produce asparagine. Alternatively, if the tunnel is not efficient, the ratio of 14N asparagine to 15N asparagine would decreases. Most of the asparagine w ould be labeled form and the ratio of 14N/15N in asparagine would be close to zero if the 14NH3 intermediate leaks and equilibrates with 15NH3 from bulk solution, since [15NH3] was much bigger than [14NH3]. The kinetic model was constructed based on th e assumption that glutamine competes with exogenous ammonium, which includes 1) asparagine inhibition, 2) glutam ine inhibition, and 3)

PAGE 108

108 competition between 14N source and 15N source (Figure 5-2). The inhibition constants for aspargine and glutamine were set to 120 M and 120 mM respectively. The simulation results were consistent with the experimental results (Figur 5-3AB), which support the hypothesis that glutamine is a competitive substrate to exgenou s ammonium and the ammonia tunneling is fast and efficient enough in hASNS so that ammonia intermediate pr oduced from the hydrolysis of glutamine is competing with 15NH3/15NH4 +. The binding of glutamin e did not suppress the utilization of ammonium by hASNS. Compared with the results of ASB, hASNS sh owed less preference of using glutamine as nitrogen source. Under the same experimental conditions, the 14N-Asn/15N-Asn ratio in the ASB catalyzed reaction is bigger than that in hASNS catalyzed reaction (Figure 5-4A), or inversely, the 15N-Asn/14N-Asn ratio is less (Figure 5-4B). Relatively less 15N was incorporated into asparagine by ASB, suggesting more preference to glutamine by ASB relative to human enzyme. The presence of glutamine part ially suppresses the ut ilization of exogenous ammonia by ASB. In another word, ASB prefers using glutamine th an ammonium, while human enzyme is more capable of using ammonium as nitrogen source to synthesize asparagine than ASB. This difference may evolve from the need of utilizing different nitrogen sources. In Escherichia coli both ammonia dependent and glutamine dependent asparagine synthetase were discovered. The two enzymes have little similarities in structure and not evolutionary related ( 11). The ammonia dependent asparagine synthetase (ASA), encoded from asnA can only use ammonium as nitrogen source. And glutamine dependent aspara gine synthetase (ASB) can use both ammonium and glutamine, with a preferen ce of glutamine. In human cell, only glutamine dependent asparagine synthetase was f ound, which is homologous to the Escherichia coli enzyme, ASB. The human enzyme can also use both ammonium and glutamine, with higher ability for using

PAGE 109

109 ammonium, relative to ASB. Si nce no ammonia dependent aspara gine synthetase was found in human genome, this kinetic difference may result from the need of ammoni a utilizing enzyme in human cells. During the process of evolution, human enzyme gradually developed more ability to utilize exogenous ammonium, and kept its glut amine dependent asparagine synthetase activity, while ASB functions mainly as gl utamine dependent enzyme since Escherichia coli cell can express ammonia dependent asparagine synthetase. Conclusions The studies on human asparagine syntheta se showed that when both ammonia and glutamine are present in the reaction mixture, a mmonia competes freely with glutamine as the nitrogen source. The simulated results based on th is study are consistent with the experimental results. In previous chapter, Escherichia coli asparagine synthetase showed partial suppression of ammonia by glutamine. Compared to hASNS, th e bacterial enzyme has higher preference to use glutamine, and relatively less pref erence of ammonia. This may result from evolution due to the fact that in Escherichia coli, there is ammonia dependent aspa ragine synthetase. It is not necessary for glutamine dependent enzyme to utilize ammonia. Ho wever, no ammonia dependent AS was found in human cells. The hu man enzyme may functi on as both ammonia and glutamine dependent enzyme during the evolution.

PAGE 110

110 Figure 5-1. Quantification of 15N-asparagine and 14N-asparagine using NMR assay and HPLC assay. The enzyme can use both 15NH4Cl and 14N-glutamine as nitrogen source. With both in the reaction mixture, both 15N-asparagine and 14N-asparagine were produced. The amount of 15N-asparagine produced can be directly determ ined using NMR assay. The total amount of asparagine can be quantified by HPLC ass ay. The difference is the amount of 14Nasparagine produced.

PAGE 111

111 E +Gln EGln EAsnGlu++ +NH3 15 E NH3 15 E+N15-Asn+Asn EAsn+NH3 15 NH3 15EAsn +Asn++NH3 15 NH3 15 +-Asn -Asn N15N15N15N15-Asn -Asn E E N15-Asn E EAsn N15-Asn N15-Asn+ +Asn:N14-Asn E: EATPAsp complex k1k-1k-12k-5k-6k-8k-3k-11k-9k2k3k4k5k6k7k8k9k10k11k12k13+Gln EGlnGln+NH3 15 EGlnGln NH3 15 EGlnGln+N15-Asn k14k15k-13k-14 Figure 5-2. Simulation model for the competition reactions catalyzed by hASNS.

PAGE 112

112 A) B) Figure 5-3. Simulations of competition reactions catalyzed by hASNS. A) varing glutamine, brown square: experimental results; blue, simulation. B) varing ammonium, blue: simulation; purple square: experimental

PAGE 113

113 A) B) Figure 5-4. Utilization of ammonia and glutamine as nitrogen source by ASB a nd hASNS. In general, ASB showes higher preference to use glutamine as nitrogen source. A) varing glutamine, square: ASB; cross, hASNS. B) varing ammonium, blank diamond: ASB; square: hASNS

PAGE 114

114 Table 5-1. hASNS catalyzed production of 15N into L-asparagine in the steady-state competition assays.a a All reactions contained 10 mM MgATP, 20 mM aspartate, 60 mM MgCl2 and 280 nM hASNS in 100 mM HEPPS buffer, pH 8 (2 mL total volume). b As determined by gHMQC NMR spectroscopy. Errors ar e estimated on the basis of measurements employing known concentrations of authentic 15N-Asn. c This value is corrected for the amount of 15N-Asn that would be formed from 15N-Gln at natural abundance, but the contribution of 14N-Asn formed from any 14NH3 released from the enzyme, as a result of the glutaminase activity of AS-B, is assumed to be negligible. d Mean value and standard deviation computed from two separate determinations on duplicate samples using reverse-phase HPLC. e ND not detected; NA not applicable. Gln (mM) 15NH4Cl (mM) 15N-Gln (mM)b 15N-Asn (mM)b,c Total Asn (mM)d 14N-Asn/15NAsn 0 100 NDe 0.21 0.02 0.215 0.009 0 2.5 100 0.025 0.004 0.19 0.05 0.21 0.02 0.1 0.2 5.0 100 0.05 0.02 0.21 0.03 0.244 0.007 0.2 0.2 10 100 0.060 0.002 0.21 0.02 0.28 0.02 0.3 0.1 20 100 0.054 0.006 0.198 0.006 0.28 0.02 0.4 0.1 40 100 0.07 0.01 0.228 0.003 0.269 0.002 0.2 0.02 20 0 NDe NDe 0.105 0.004 NAe 20 10 NDe 0.031 0.005 0.175 0.002 4.6 0.7 20 25 NDe 0.074 0.010 0.222 0.009 2.0 0.3 20 50 0.02 0.01 0.14 0.02 0.26 0.02 0.9 0.2 20 75 0.047 0.002 0.1721 0.0007 0.25 0.02 0.4 0.1 20 100 0.054 0.006 0.198 0.006 0.28 0.02 0.4 0.1

PAGE 115

115 CHAPTER 6 PREPARATION OF LIGAND BOUND hASNS INHIBIITION AN D BINDING STUDIES Introduction 1 Glutamine dependent asparagine synthetase (ASNS, EC 6.3. 5.4) catalyzes ATP dependent conversion of aspartate to asparagi ne using glutamine as the nitrog en source. One of the interests in asparagine synthetase lies in the correlation of this enzyme to the resistance of the acute lymphoblastic leukemia cells after treated with L-asparaginase. For a long time, the Escherichia coli asparagine synthetase B (ASB ) was studied as a model for the human enzyme because of its readiness to be expressed in a large amount. ASNS belongs to class II amidotransferase, with two active sites that are separated by an intramolecular tunnel. Each of the two active sites catalyzes a half reaction. At glutaminase site, ammonia is released from glutamine through a thioester intermediate formed by the nucleophilic at tack of the thiolate gr oup of Cys1 residue to the amide group of glutamine. At the other end of the tunnel, the -carboxylic acid group of aspartate is activated by ATP and form the -aspartyl-AMP intermediate, followed by ammonia attacking to its -carbonyl group to produce asparagine (Figure 6-1). Although both human ASNS and ASB show same properties from the point view of chemistry, the kinetic studies of human enzyme revealed several differences from that of E. coli enzyme, which were discussed in the previous chapter. So far, the crystal structure of C1A mutant of Escherichia coli asparagine synthetase B (ASB) is the only one that was re ported for asparagine s ynthetase (ASNS). Since the function of biomolecules can only be fully understood under the stru ctural context, more crystal structure of ASNS, especi ally of hASNS, need to be st udied. Recently, human asparagine synthetase (hASNS) was successfully expressed in a baculovirus system using sf9 insect cells. 1 This work was done with Alexandria Berry, an undergraduate. Li designed the experiments and analyzed the results. Li and Berry performed the work.

PAGE 116

116 This makes possible to have e nough protein for X-ray crystallogra phic studies of human enzyme. In this work, efforts were made to pr epare inhibitor-bound hA SNS. 4-diazo-5-oxo-Lnorleucine (DON) is a glutamine an alog. It irreversibly inhibits the glutaminase activity of class II amidotransferases by reacting with the thiol group of the ca talytic Cys1 residue in the glutaminase active site and forming a covalent C-S bond (Figure 6-2). The transition state analog Adenylated sulfoximine, mimicing that -aspartyl-AMP is attacked by ammonia, was a good inhibitor for human asparagine synthetase with KI at nM level ( 71) (Figure 6-3). My results showed that both inhibitors can bind to the enzyme, with DON in the glutaminase site and sulfoximine in the other. Materials and Methods Materials Unless otherwise stated, all chemicals and reagen ts were purchased from Sigma (St. Louis, MO). L-Glutamine was recrystallized pr ior to use as prev iously described ( 53). C-terminally tagged human ASNS were expressed in and purifie d from a baculovirus/in sect cell expression system as previously described ( 16, 57). Adenylated sulfoximine (as a 1:1 mixture of diasterioisomers) were obtained from Japan. Enzyme Assays The glutaminase activity of hASNS was de termined by measuring the production of glutamate using glutamate dehydrogenase (L-GL DH) coupled assay. ASNS reaction mixtures (200 L total volume) contained 100 mM NaCl, 10 mM MgCl2, and 25 mM L-glutamine in 100 mM EPPS. Reactions were in itiated upon the addition of hASN S and incubated at 37C for 10 minutes, and terminated by addition of 30 L 20% (v/v) trichloroaceti c acid (TCA). The reaction mixtures was added to 770 L coupling assay mixture, which c onsisted of pH 9.0 buffer (300 mM glycine, 250 mM hy drazine), 1.5 mM NAD+, and 1 mM ADP. The absorbance was recorded

PAGE 117

117 before and after the addition of 2.2 units of glutamate dehydrogenase (0 and 30 minutes). Glutamate was quantified using standard curve. Th e synthetase activity with either glutamine or ammonia as nitrogen source was assayed by determining the formation of PPi using PPi reagent (Sigma Technical Bulletin BI-100), based on the fact that PPi a nd asparagine were produced at 1:1 ratio in the hASNS catalyzed reaction. Loss of Glutaminase Activity with Time at Different Temperature The stability of hASNS was first investigated to determine the best incubation time. Timebased glutaminase activity of hASNS was tested at three different incubation temperatures, in ice, room temperature and 37 oC. After 10 L of deionized water being added to one tube (approximately 100 L) of the stock enzyme solution (1 mg/mL, 50 mM EPPS, pH 8, 5 mM DTT, and 20% glycol), the soluti on was aliquoted into 3 tubes, which were incubated on ice(0-4 oC, at room temperature (approximately 22C), a nd at 37C respectively. At a series of time points (0-61 minutes), 2 L incubated enzyme from each ali quot were removed and immediately added into 198 L of glutaminase reaction mixture to test its glutaminase activity. DON Inhibition The inhibition mixture was pr epared by mixing stock enzyme solution (1.0 mg/mL) and DON aqueous solution (50 mM), which contained 0.9 mg/mL hA SNS and 5.88 mM DON, while the control was made by mixing stock enzyme solution and deionized water. EPPS solution was filtered using Fisherbrand 25 mm Syringe Filter, 0.2 m, nylon. Solutions were incubated at room temperature and 15 L aliquots were removed simultaneously from both inhibition mixture and control at varying time point s (4-30 minutes), and then the enzyme was separated from free DON by using 30,000 nominal molecular weight li mit (Daltons) Microcon Centrifugal Filter Devices spin columns (Millipore Corporation, Be dford, MA). After each aliquot was spun down for 1 minute at 10,000 rpm, the en zyme was rinsed twice using 100 L of 100 mM filtered EPPS

PAGE 118

118 (4 minutes at 13,000 rpm). The purified enzyme was recovered in a new centrifuge tube by 30 L of 100 mM filtered EPPS, which was added to the bottom side of the filter (1 minute at 7,000 rpm). All enzymes recovered were immediatel y assayed for their glutaminase activity (10 L recovered enzyme). The DON inhibited enzyme was also checked for synthetase activity versus control enzyme synthetase activity using the pyrophosphate assay and c oncentration using the Bradford assay. Loss of Synthetase Activity with Time at different Temperature The time dependent loss of synthetase activity was determined at both room temperature and 37 oC. About 12% deionized water was added to the stock enzyme. The enzyme solution then was aliquoted into 8 diffe rent tubes and stored in a -80C freezer. These tubes were removed from the freezer one by one and sit at room temper ature (approximately 22C) for different times (0-70 minutes), or at 37C over a 60 minute peri od, followed by testing their synthetase activity (incubated at 37C in 100 mM EPPS pH 8, containing 10 mM MgCl2, 100 mM NH4Cl, 20 mM aspartate, and 2.5 mM ATP ove r a period of 10 minutes with a 1 mL final volume). Inhibition of the DON Inhibi ted Enzyme by Sulfoximine The DON inhibited enzyme was prepared as described above, with incubation time being 20 minutes. The synthetase activity of the D ON inhibited enzyme was assayed using the pyrophosphate assay with 0-10 M sulfoximine. As for the control enzyme, the synthetase activity was measured with 0 M or 10 M sulfoximine inhibitor. Results Stability Experiment Previous studies by size exclusion chromat ography showed that hASNS dissociates into monomer from dimer with time. The structural changes may result in the loss of its activity. Furthermore, the unstable tholate group of Cys1 is readily oxidized in the solution, which causes

PAGE 119

119 loss of the glutaminase activity of the enzyme. Here we investigated the stability of the enzyme before we characterized the inhibiti on to hASNS by DON and sulfoximine. hASNS lost glutaminase activity with time at room temperature and 37 oC Aliquots of hASNS were incubated in ice, at room temperature (about 22 oC) and at 37 oC respectively. The glutaminase activity of hASNS was immediately tested after being incubated for a certain time by previous ly reported assay method ( 17 ). Our results showed that hASNS lost its glutaminase activity faster at higher temperature. When incubate d in ice, its activity decreased by 15% after 60 minutes (data not shown). While after 20 minutes incubation at 37 oC, the enzyme lost 50% of its glutaminase activity (F igure 6-4, Table 6-1). After 60 minutes at 37 oC, more than 85% activity was lost. At room temperat ure, hASNS still kept 57% of its activity after 60 minutes. The percent activity left decreased e xponentially with time, which indicates a first order reaction for the lost of glutaminase activ ity. The calculated half-life for glutaminase activity of hASNS are 74 mi nutes and 20 minutes for room temperature and 37 oC respectively. The loss of NH3 dependent synthetase activity with time at room temperature and 37 oC The ammonia dependent synthetase activity wa s also checked after hASNS was incubated at room temperature or at 37 oC (Figure 6-5, Table 6-2). After being incubated for certain time, hASNS was added to assay mixture to initiate the reaction. The re sults showed that hASNS lost its ammonia dependent activity in different pa tterns for short time incubation and long time incubation. At room temperature, the enzyme lost little synthetase activity until being incubated for more than 30 minutes, and then its activity dropped exponentially. Similar results were observed for the enzyme incubated at 37 oC. When the incubation time was shorter than 15 minutes, the synthetase activity went down more like linearly than exponentially. After being incubated for more than 15 minutes, the activity decrea se exponentially.

PAGE 120

120Inhibition to hASNS by DON It has been reported that DON reacts with thio l group of Cys1 residue of several other class II amidotransferases and forms covalent bonded complex (E-ON). While few studies about DON inhibition to asparagine synthe tase were reported, we did this experiment to support the hypothesis that DON is a glutaminase inhibitor for all the class II amidotra nsferases and has little effect on synthetase ac tivity of hASNS. DON inhibits 90% of glutaminase activity of hASNS after incubating for 20 minutes. E EDONk[DON] = 0.0026 s-1[DON] = 5.88 mM Before measuring its glutaminase activity af ter hASNS was incubate d with 5.88 mM for a series of time, we separated E-ON from DON by using spin column (see methods part) because our results suggested that DON is a substrate of coupling enzyme glutamate dehydrogenase (LGLDH) and affect the glutaminase assay. The pr otein concentration was tested using modified Coomassie assay and the percent recovery of hASNS for separati on step was 71 5 %. For each time point, we used incubated enzy me without DON as control to eliminate the effect caused by separation step (Figure 6-6, Table 6-3). After 20 minutes incubation, the glutaminase activity of DON inhibited hASNS was only a bout 10% of control enzyme. Th e percent activity decreased exponentially with time. This indicates a firstorder reaction for the in activation of hASNS by 5.88 mM DON and the rate consta nt was calculated as 0.0026 s-1. DON had little effect on the rate of a mmonia dependent synthe sis of asparagine. In this work, we did not try to characterize the ammonia dependent synthetase activity of DON incubated enzyme. We simply compared the production rate of pyrophosphate by DON incubated enzyme with that by control enzyme (i ncubated without DON) at saturating level of all

PAGE 121

121 the substrates (except ammonia at 100 mM). Several reports showed that Asn:PPi ratio close to 1:1 for free enzyme or inhibited enzyme. Therefor e, we here assume that DON incubated hASNS will not affect Asn:PPi ratio during the enzyme catalyzed synthesis. Note: In order to fully characterize the inhibition mechanism, this as sumption need to be confirmed by using HPLCbased end point assay ( 23) and PPi assay (22 ). The concentration of the recovered enzyme solutions was determined to be 0.78 mg/mL usi ng the Bradford assay. The volume of recovered enzyme solution varied from about 12 to 18 L (approximately 0.0117 mg of enzyme), versus the initial volume of 15 L of 1.0 mg/mL of stock enzyme (0. 015 mg). With all the substrate at saturating level, the producti on rate of pyrophosphate had no significant changes between E-ON complex and control enzyme, which implies little changes for ammonia dependent synthetase activity (Table 6-4). Inhibition by Adenylated Sulfoximine The transition state analog adenylated sulfoximine is a good inhibitor with KI at nanomolar level. It was reported to suppre ss proliferation of an L-aspargin ase-resistant leukemia cell line and is a potential solution for drug resistance in lymphoblastic and myeloblastic leukemia cells ( 71). Here we examined its inhibition to DON incubated hASNS and ch ecked the difference between its inhibition effect on DON in cubated hASNS and on free enzyme. Inhibition to the ammonia dependent synthetase activity of stock hASNS The inhibition to hASNS by sulfoximine has been studied regarding its ammonia dependent synthetase activity as well as glutaminase acitivity. While sulfoximine was reported as a good inhibitor to synthetase act ivity of hASNS, little effect on glutaminas e activity of hASNS by sulfoximine was discovered. As control experiment for the st udies of its inhibition to DON incubated hASNS, we tested ammonia dependent synthetase activity of stock hASNS with 3 different concentrations of sulf oximine, 0, 5, and 10 uM respectiv ely. The initial rate was 0.0404

PAGE 122

122 0.0007 M/s. The production of PPi catalyzed by hASNS without sulfoximine linearly increased with time. With the presence of inhibi tor, the production gradually slowed down with time (Figure 6-7). Data were processed based on the slow onset i nhibition model as report ed in literature ( 71, 80). Three parameters, ss, 0 and k were determined by fitti ng the data using KaleidaGraph v3.5 from Synergy Software, where ss and 0 are initial and steady-state velocities respectively, and k is the apparent first-order rate cons tant for isomerization of EI to EI*. k6 was calculated using the following equations with these three parameter determined at different concentrations of inhibitor: 0 6v v kkss (1) The average value of k6 then was used to estimate KI and k5 by fitting to the equation below. I a IK I K ATP K I kkk][][ 1 ][56 (2) The calculated kinetic results were listed in Table 65 as well as those report ed in the literature. Inhibition to the ammonia dependent synt hetase activity of DON incubated hASNS The ammonia dependent synthe tase activity of DON incubated hASNS was tested with different concentrations of sulf oximine (0-10uM), immediately af ter purified E-ON was prepared. The incubated hASNS without DON was used as control. The production of PPi without sulfoximine linearly increased with time for bot h DON incubated enzyme and control enzyme, with kcat value of 0.09 0.01 uM/s and 0.047 0.005 uM/s respectively (F igure 6-8 and 6-9). With sulfoximine, the production of PPi versus time was consistent with the slow onset inhibition model, same as what we got before fo r stock enzyme (Figure 6-8). Based on the results

PAGE 123

123 for control enzyme, the rate constants for convers ion of EI to EI* were calculated and listed in Table. Here we assume the control enzyme has same KI value as stock enzyme. Little difference was found between these rate constants for control enzyme and for stock enzyme (Table 6-6). This showed that the binding property of hASNS to sulfoximine had no significant changes after the control enzyme was prepared. Unfortunate ly, for DON incubated enzyme, we could only accurately determine the initial velocity 0 by fitting the data using KaleidaGraph, not ss and k. All the fitting gave a big error for ss and k, which has no practical meaning, except k was fitted as 4.9-4 s-1 at 10 M of sulfoximine (Figure 6-9). Discussion The catalytic mechanism can only be completely understood under structural context. hASNS has two active sites, glutaminase site and synthetase site. In order to prepare inhibitor bound hASNS and make crystal structure, we did kinetic studies about hASNS using two inhibitors, DON and sulfoximine. DON is a glut amine analog. DON incubated enzyme showed significant decrease for its glut aminase activity after being separated from DON and diluted for more than 30 fold in the enzyme catalyzed reaction. The first order rate constant was calculated as 0.0026 s-1 in the presence of 5.88 mM DON at room temperature. After 20 minutes incubation with DON, 90% of hASNS was inhibite d based on its glutaminase activity while its ammonia dependent synthetase activity decreased about 10%. This result supports the idea that DON reacts with the thiolate group of Cys1 residue of hASNS and form covalently bound enzyme, although no direct evidence shows the in hibition is irreversible for hASNS. The covalently bound enzyme complex has been observe d for other class II amidotransferases, such as glutamine phosphoribosylpyrophosphate (PRPP) amidotransferase, glutamate synthase and glucosamine-4-phosphate synthetase (GlmS) (28, 54, 81, 82). All this evidence indicates that DON incubated hASNS is E-ON complex. The fact that DON incubated enzyme lost

PAGE 124

124 glutaminase activity without affecting its synthe tase activity significantl y proves that DON, as a glutamine analog, binds glutaminase active site, not synthetase site. This conclusion was further supported by the result that the ammonia dependent syntheta se activity of DON incubated enzyme was inhibited by sulfoximine, an analog of transition state that formed in the synthetase site, which had little effect on the glutaminas e activity of hASNS by previous studies. The apparent first order rate cons tant was calculated as 0.00049 s-1 for the inhibition of DON incubated enzyme with the presence of 10 M sulfoximine. The calculated half-life was 24 minutes and the time for 90% inhibition was 78 minutes. hASNS is not stable. After one hour at room temperature, the enzyme lost 43% of its glutaminase activity and 11% of synthetase activity. When the enzyme was kept at 37 oC, these numbers are 87% and 67% respectively. The lost of its glutaminase activity may result from the oxidation of the thiolate group of Cys1. However, it is unexpected that its ammonia dependent synthetase activity also decreased with time. Maybe this is caused by structural changes of the enzyme at higher temperature because its mol ecular weight changed from dimer form to monomer form by size exclusion chromatography. The sulfoximine inhibition to DON incubated hASN S is different from the control. Only k value for the highest concen tration of sulfoximine (10 M) was estimated as 0.000489 s-1. It is decreased by 5 times compared to that of c ontrol enzyme, which means the formation of EI* became slower for DON incubated enzyme. Since stock enzyme and control enzyme showed similar binding affinity to sulfoximine. We can exclude the possibility that this was caused by incubation for 20 minutes at room temperature as well as separation step. Therefore, we can conclude that the formation of the covalent complex from the reaction of DON with enzyme at glutaminase site resulted in some conformational changes of the synthetase site. Since k can be

PAGE 125

125 expressed as a function of k5, k6 and KI at certain concentration of ATP, and k6 << k or k5 for control enzyme, the decreasing of the k valu e is likely caused by the decreasing of k5, or increasing of KI or both. This implies that DON inhibite d enzyme has some structural changes either that affect the binding of sulfoximine or that make the on rate of EI to EI* slower. This can be supported by the following analysis. If incubation with DON does not affect its binding to sulfoximine, that is, KI is a constant with value of 0.181 M, so I I a I I aK I K I K ATP k K I K I K ATP kkk][ ][][ 1 ][ ][][ 1 )(6 5 14 14 5102.7 181.0 10 181.0 10 2.0 5 1 109.4 s M M M M mM mM s k Therefore, k5 is less than 7.2 10-4 s-1. This value is one fourth of that for control. k K I K ATP K I kkkI a I ][][ 1 ][5 6 So k6 must be less than 4.9 10-4 s-1. Compared to original value of 1.9 10-4 s-1, the change of k6 cannot significantly affect k valu e to be changed from 2.3 10-3 s-1 to 4.9 10-4 s-1. So k5 is the one of the factors that affect k value significantly. Since k6 is not a major factor that aff ect k value, we can neglect the k6 term in the equation 2. Then

PAGE 126

126 I a IK I K ATP K I kk][][ 1 ][5 So k is negatively related to KI value. The increasing of KI results in the decreasing of k. For DON incubated enzyme, it is more difficult to determine ss and k by fitting the data measured within 15 minutes. The concentrati on of the pyrophosphate produced can be expressed using following equation for sl ow onset inhibition model. k vv etvPss kt ss 0)1 ( When time is long enough, k vv tvPss ss 0. This means the production of pyrophosphate is linearly related to the time with ss as slope and k vvss 0 as intercept on y-axis. Therefore, to determine ss and k, especially ss, we need to measure the production of pyrophosphate for enough time. Unfortunately, hASNS is not stable enough for a long time assay.

PAGE 127

127 O-O-NH3+O O H Asp ATPPPiN N N N NH2O O OHOH P O O-O-O NH3+O O H -Aspartyl-AMP O-NH2NH3+O O H AsnNH3 AMP O O N N H3 +O-H H H OEN+S-H H H Cys1O O NH3 +O-H OEN S H H NH3H2O O O O NH3 +O-H H OEN+S-H H H Gln Glu Tunneling NH3Cys1 Cys1 NH3 AMP Figure 6-1. The mechanisms of ASNS catalyzed reaction. At glutaminase site, ammonia is released from glutamine through a thioester intermediate formed by the nucleophilic attack of the thiolate group of Cys1 residue to the amide group of glutamine. Then the thioester is hydrolyzed into glutamate and reproduce free enzyme. At the other end of the tunnel, the carboxylic acid group of aspartate is activated by ATP and form the -aspartyl-AMP intermediate, followed by ammonia attacking to its -carbonyl group to produce asparagine.

PAGE 128

128 O OCH NH2+O N+N H 6-diazo-5-oxo-L-norleucine (DON) SH E-Cys1+O O NH2+O H S +NN 6-(Enzyme-cysteinyl)-5-oxo-L-norleucine Figure 6-2. DON reacts with Cys1 residue of ASNS and forms covalent adduct.

PAGE 129

129N N N N NH2O O OHOH P O O N O NH3+O O S CH3 N N N N NH2O O OHOH P O O O O NH3+O C O NH3 -aspartyl-AMP adenylated sulfoximine Figure 6-3. The inhibitor adenylated sulfoximine (right) mimics the nucleophilic attacking of aspartyl-AMP by ammonia.

PAGE 130

130 0 20 40 60 80 100 120 0 1000200030004000 time (s)%activity Figure 6-4. The glutaminase activity of hASNS decreased exponentially with time. Solid line (data: filled square): 22 oC; Dashed line (data: blank square): 37 oC..

PAGE 131

131 20 40 60 80 100 120 010002000300040005000 time (s)% activity Figure 6-5. The ammonia dependent synthetase activity of hASNS decreased with time. Solid line (data: filled square): 22 oC; Dashed line (data: blank square): 37 oC.

PAGE 132

132 0 20 40 60 80 100 120 020040060080010001200 time (s)%activity Figure 6-6. The glutaminase activity of DON inhib ited hASNS decreased exponentially with time. The curve was fitted using the first four data.

PAGE 133

133 0 5 10 15 20 25 30 35 40 020040060080010001200 hAS control hAS + 5 uM I hAS + 10 uM I Time (s) Figure 6-7. The inhibition to free hASNS by sulfoximine

PAGE 134

134 0 20 40 60 80 0100200300400500600700800 E control E + 10 uM I Time (s) Figure 6-8. Inhibition to control hASNS by sulfoximine

PAGE 135

135 0 10 20 30 40 50 60 70 020040060080010001200 EDON EDON + 10 uM I EDON + 8 uM I EDON + 6 uM I EDON + 4 uM I EDON + 2 uM I time (s) Figure 6-9. Inhibition to DON inhibited hASNS by sulfoximine

PAGE 136

136 Table 6-1. Percent glutaminase activity at different incubation time without DON Room Temperature (~22 oC) 37 oC Time (s) %activity Time (s) %activity 0 100 0 100 630 83.8 300 69.4 1230 76.2 600 69.6 1830 81.2 900 66.3 2430 61.0 1200 51.5 3030 67.1 1500 39.6 3630 57.3 3600 13.4

PAGE 137

137Table 6-2. Percent ammonia dependent synthe tase activity without inhibitors Room Temperature (~22 oC) 37 oC Time (s) %activity Time (s) %activity 0 100 0 100 120 99.12864 300 95.05294 600 99.59384 598 89.15113 893 84.74562 1200 99.69538 1201 77.24632 1500 68.41563 1800 99.72076 1800 59.92223 2400 97.56304 2400 49.46501 3000 41.84203 3600 88.95752 3600 33.09285 4200 88.06905

PAGE 138

138Table 6-3. The loss of glutaminase activity due to DON inhibition Room Temperature (~22 oC) Time (s) %activity 0 100 254 45 365 29 630 25 1070 11 1324 13 1883 12

PAGE 139

139Table 6-4. Synthetase activity of the DON inhibited and control enzyme after incubation determined using the pyrophosphate assay Enzyme Solution Rate of Reaction uM/s Trial 1 Trial 2 hASNS + 5.88 mM DON 3.99E-02 4.37E-02 hASNS control 4.80E-02 4.60E-02

PAGE 140

140Table 6-5. Kinetical constant for sulfoximine inhibition to ammonia dependent synthetase activity of hASNS data source KI (nM) k5 (s-1) k6 (s-1) kobs/[I]tot (M-1 s-1) K* I (nM) repeat 181 3.51 10-32.6 10-413.4 literature 285 2.98 10-32.6 10-5436 2.46

PAGE 141

141Table 6-6. Parameters for inhibition of stock hA SNS and control hASBS by sulfoximine hASNS KI (nM) k5 (s-1) k6 (s-1) Kobs/[I]tot (M-1 s-1) K* I (nM) stock 181 3.51 10-32.6 10-413.4 control 181* 3.12 10-31.9 10-411.0

PAGE 142

142 APPENDIX A SEQUENCE ALIGNMENT OF TUNNEL RESIDUES IN ASB SPECIES Tunnel Residues 1 1 2 3 3 3 3 3 3 3 3 3 3 3 4 2 3 5 7 7 2 4 3 4 4 4 5 5 8 8 8 9 9 9 0 9 1 0 4 6 0 3 8 4 6 8 0 2 0 5 9 3 7 9 1 1 2 3 3 3 3 3 3 3 3 3 3 3 4 4 3 3 5 7 7 4 3 2 4 4 4 5 5 8 8 9 9 9 0 0 1 0 2 2 5 8 2 2 9 5 7 9 1 7 4 8 2 6 8 0 4 plant_456_e-133 CHRGPL I NGEYIVTLDMVISGEGS DEYLDC A NTFG L E A RVL plant2_525_e-153 CHRGPL I NGEYIVTLDMVISGEGS DEYLDC A NTSG L E A RVL mustard_plant_456_e-125 CHRGPL V NGEYIVTLDMVLSGEGA DEYLDC A NTSG L E A RVL pea_586_e-160 CHRGPL I NGEYIVTLDMVISGEGS DEYLDC A NTYG L E A RVL bacteria2_564_e-180 CHRGPL I NGEYIVVLDMVLSGEGS DEYLDC A NMMG V E P RVL plant3_586_e-168 CHRGPL I NGEYMICLDMVISGEGS DEYLDC A NTSG V E A RVL fruit_fly_558_e-120 CHRGPL I NGEYMIILDMILSGEGA DEYLDC A NAMG V E L RVL Ecoli_ CHRGPL I NGEYMIILDMVLSGEGS DEYLDC A NMSG V E A RVL bacteria_554 CHRGPL I NGEYMIILDMVLSGEGS DEYLDC A NMSG V E A RVL Salmonella_554 CHRGPL I NGEYMIILDMVLSGEGS DEYLDC A NMSG V E A RVL Yersinia_554 CHRGPL I NGEYMIILDMVLSGEGS DEYLDC A NMSG V E A RVL Paramecium_588_e-132 CHRGPL I NGEYMIIFDMCLTGEGS DEYLDL A NCLG I E T R PM fungus_571_e-133 CHRGPL I NGEYMVTLDMVLSGEGS DEYLDC A NTMG L E A RVL fungus2_574_e-136 CHRGPL I NGEYMITLDMVLSGEGS DEYLDC A NTMG L E A RVL Neurospora_581_e-148 CHRGPL I NGEYMITLDMVLSGEGS DEYLDC A NTSG L E A RVL sunflower_589_e-163 CHRGPL V NGEYMITLDMVISGEGS DEYLDC A NTSG L E A RVL mosquito_533_e-112 CHRGPL I NGEYMIVLDVVLSGEGA DEYLDC A NCAG L E L RVL C_elegans_567_e-105 CHRGPL I NGEYVVKLDMVLSGEGA DEYLDC A DSMSVE V RVL hamster_561_e-81 CHRGPL V NGEYVVRLDMIFSGEGS DEYLDV A DTAG L E L RVL mouse_561_e-81 CHRGPL V NGEYVVRLDMIFSGEGS DEYLDV A DTAG L E L RVL Human_562_e-80 CHRGPL V NGEYVVRLDMIFSGEGS DEYLDV A DTAG L E L RVL ****** **** ** **** ***** * ** Red asterisk shows absolutely conserved residu e; black asterisk show s those residues that conserved in 20 out 21 species. Red letter: conserved residue in at leas t 20 out of 21 species. Blue background letter show different residue(s) ot her than conserved residues Yellow column: conserved residue in at least 15 out of 21 species. Purple column: conserved residue in at least 10 out of 21 species. Note: 1. An absolutely conserved Gly is right before Met 120 and Ile 142. 2. Ile 52 (16 I, 5 V), Met 120 (12 M, 5 I, 4 V), Ile 142 (11 I, 10 V) Asn 389 (17 N, 4 D), Val 397 (12 L, 8 V, 1 I), Cys 385 (17 C, 3 V, 1 L) Val 344 (16 V, 4 I, 1 C), Ser 350 (17 S, 4 A)

PAGE 143

143 APPENDIX B THE KINETIC MODEL FOR EXCHANGE EXPERIMENT The model for exchanging experiment was analyzed as following: EGln EesterNH4 +NH4 + 15E Eester Glu OH2+ + k1k2N15Gln + + k1k2E Eester+ + k3(i) (ii) (iii) Since [Gln] >> [15N-Gln] and [15NH4 +] >> [NH4 +], the reverse reaction of (i) and (ii) are negligible. So the model is simplified as: EGln EesterNH4 +NH4 + 15E Eester Glu OH2+ + k1N15Gln + + k2E Eester+ + k3(I) (II) (III) Based on the steady state theory, 0]][[]][[ln]][[ ][2 34 15 2 1 OHEesterkNH EesterkGEk dt Eesterd ]][[]][[ln]][[2 34 15 2 1OHEesterkNH EesterkGEk ])[][]([ln]][[234 15 2 1OHkNHkEester GEk ln][ ])[][]([ ][1 234 15 2Gk OHkNHkEester E

PAGE 144

144 Since ][][][ EesterEET Then ][ ln][ ])[][]([ ][1 234 15 2Eester Gk OHkNHkEester ET ln][ ])[][ln][]([1 234 15 2 1Gk OHkNHkGkEester Therefore, ])[][ln][( ln][][ ][234 15 2 1 1OHkNHkGk GEk EesterT So the production rate of [15N-Gln] is: ])[][ln][( ]ln][[][ ]][[234 15 2 1 4 15 21 4 15 2OHkNHkGk NHGEkk NH EesterkvT When [15NH4 +] is kept constant and [Gln] is varied, ln][])[][( ln]])[[][(2 1 3 4 15 1 2 4 15 2GOH k k NH k k GNHEk vT ln][ ln][ln lnmaxGK GVGM G Where ][][4 15 2lnmax NHEkVT G ][][2 1 3 4 15 1 2 lnOH k k NH k k KGM When [Gln] is kept constant and [15NH4 +] is varied, ][])[ln][( ]ln])[[][(4 15 2 2 3 2 1 4 15 1 NHOH k k G k k NHGEk vT

PAGE 145

145 ][ ][4 15 4 15 max4 15 4 15 NH K NH VNHM NH Where ln][][2 max4 15GEk VT NH ][][2 2 3 4 15 2 14 15OH k k NH k k KNHM

PAGE 146

146 LIST OF REFERENCES 1. Hyde, C. C.; Ahmed, S. A.; Padlan, E. A.; Miles, E. W.; Davies, D. R., Three-Dimensional Structure of the Tryptophan Synthase R2 Multienzyme Complex from Salmonella typhimurium. J Biol Chem 1988, 263, 17857-71. 2. Huang, X.; Holden, H. M.; Raushel, F. M., Channeling of substrates and intermediates in enzyme-catalyzed reactions. Annu Rev Biochem 2001, 70, 149-80. 3. Raushel, F. M.; Thoden, J. B.; Holden, H. M., Enzymes with molecular tunnels. Acc Chem Res 2003, 36, (7), 539-48. 4. Lunn, F. A.; Bearne, S. L., Alternative su bstrates for wild-type and L109A E. coli CTP synthases -Kinetic evidence fo r a constricted ammonia tunnel. Eur J Biochem 2004, 271, 420412. 5. Zalkin, H., The amidotransferases. Adv Enzymol Relat Areas Mol Biol 1993, 66, 203-309. 6. Zalkin, H.; Smith, J. L., Enzymes ut ilizing glutamine as an amide donor. Adv Enzymol Relat Areas Mol Biol 1998, 72, 87-144. 7. Haskell, C. M.; Canellos, G. P., l-asparaginase resistance in human leukemia--asparagine synthetase. Biochem Pharmacol 1969, 18, (10), 2578-80. 8. Prager, M. D.; Peters, P. C.; Janes, J. O.; De rr, I., Asparagine synthetase activity in malignant and non-malignant human kidney and prostate specimens. Nature 1969, 221 (185), 1064-5. 9. Nakamura, M.; Yamada, M.; Hirota, Y.; Sugi moto, K.; Oka, A.; Takanami, M., Nucleotide sequence of the asnA gene coding for as paragine synthetase of E. coli K-12. Nucleic Acids Res 1981, 9, (18), 4669-76. 10. Felton, J.; Michaelis, S.; Wright, A., Mutatio ns in two unlinked gene s are required to produce asparagine auxotrophy in Escherichia coli. J Bacteriol 1980, 142, (1), 221-8. 11. Scofield, M. A.; Lewis, W. S.; Schuster, S. M., Nucleotide sequence of Escherichia coli asnB and deduced amino acid sequence of asparagine synthetase B. J Biol Chem 1990, 265, (22), 12895-902. 12. Humbert, R.; Simoni, R. D., Genetic and bi omedical studies demonstrating a second gene coding for asparagine synthe tase in Escherichia coli. J Bacteriol 1980, 142 (1), 212-20. 13. Richards, N. G.; Schuster, S. M., Mechanistic issues in asparagine synthetase catalysis. Adv Enzymol Relat Areas Mol Biol 1998, 72, 145-98. 14. Schnizer, H. G.; Boehlein, S. K.; Stewart, J. D.; Richards, N. G.; Schuster, S. M., Formation and isolation of a covalent intermediate du ring the glutaminase r eaction of a class II amidotransferase. Biochemistry 1999, 38, (12), 3677-82.

PAGE 147

147 15. Larsen, T. M.; Boehlein, S. K.; Schuster, S. M.; Richards, N. G.; Thoden, J. B.; Holden, H. M.; Rayment, I., Three-dimensional structure of Es cherichia coli asparagine synthetase B: a short journey from substrate to product. Biochemistry 1999, 38, (49), 16146-57. 16. Boehlein, S. K.; Richards, N. G.; Schuster, S. M., Glutamine-dependent nitrogen transfer in Escherichia coli asparagine synthetase B. Searching for the catalytic triad. J Biol Chem 1994a, 269, (10), 7450-7. 17. Boehlein, S. K.; Richards, N. G.; Walworth, E. S.; Schuster, S. M., Arginine 30 and asparagine 74 have functional roles in the glut amine dependent activities of Escherichia coli asparagine synthetase B. J Biol Chem 1994b, 269 (43), 26789-95. 18. Boehlein, S. K.; Schuster, S. M.; Richar ds, N. G., Glutamic acid gamma-monohydroxamate and hydroxylamine are alternate s ubstrates for Escherichia coli asparagine synthetase B. Biochemistry 1996, 35, (9), 3031-7. 19. Stoker, P. W.; O'Leary, M. H.; Boehlein, S. K.; Schuster, S. M.; Richards, N. G., Probing the mechanism of nitrogen transfer in Escherichia coli asparagine synthetase by using heavy atom isotope effects. Biochemistry 1996, 35 (9), 3024-30. 20. Boehlein, S. K.; Walworth, E. S.; Richar ds, N. G.; Schuster, S. M., Mutagenesis and chemical rescue indicate residues involved in beta-aspartyl-AMP formation by Escherichia coli asparagine synthetase B. J Biol Chem 1997, 272, (19), 12384-92. 21. Boehlein, S. K.; Walworth, E. S.; Schuster S. M., Identification of cysteine-523 in the aspartate binding site of Escherichi a coli asparagine synthetase B. Biochemistry 1997, 36, (33), 10168-77. 22. Boehlein, S. K.; Stewart, J. D.; Walworth, E. S.; Thirumoorthy, R.; Richards, N. G.; Schuster, S. M., Kinetic mechanism of Escheric hia coli asparagine synthetase B. Biochemistry 1998, 37, (38), 13230-8. 23. Tesson, A. R.; Soper, T. S.; Ciustea, M.; Ri chards, N. G., Revisiting the steady state kinetic mechanism of glutamine-dependent asparagi ne synthetase from Escherichia coli. Arch Biochem Biophys 2003, 413, (1), 23-31. 24. Fresquet, V.; Thoden, J. B.; Holden, H. M.; Ra ushel, F. M., Kinetic mechanism of asparagine synthetase from Vibrio cholerae. Bioorg Chem 2004, 32, (2), 63-75. 25. Buchanan, J. M., The amidotransferases. Adv Enzymol Relat Areas Mol Biol 1973, 39, 91183. 26. Nakamura, A.; Yao, M.; Chimnaronk, S.; Sakai, N.; Tanaka, I., Ammonia channel couples glutaminase with transamida se reactions in GatCAB. Science 2006, 312, (5782), 1954-8. 27. Oshikane, H.; Sheppard, K.; Fukai, S.; Nakamura, Y.; Ishitani, R.; Numata, T.; Sherrer, R. L.; Feng, L.; Schmitt, E.; Panvert, M.; Blanquet, S.; Mechulam, Y.; Soll, D.; Nureki, O., Structural

PAGE 148

148 basis of RNA-dependent recruitment of glutamine to the genetic code. Science 2006, 312, (5782), 1950-4. 28. Mouilleron, S.; Badet-Denisot, M. A.; Golin elli-Pimpaneau, B., Glutamine binding opens the ammonia channel and activates glucosamine-6P synthase. J Biol Chem 2006, 281, (7), 4404-12. 29. Anand, R.; Hoskins, A. A.; Stubbe, J.; Ealic k, S. E., Domain organization of Salmonella typhimurium formylglycinamide ribonucleotid e amidotransferase revealed by X-ray crystallography. Biochemistry 2004, 43, (32), 10328-42. 30. Endrizzi, J. A.; Kim, H.; Anderson, P. M.; Baldwin, E. P., Crystal structure of Escherichia coli cytidine triphosphate synthetase, a nucleot ide-regulated glutamine amidotransferase/ATPdependent amidoligase fusion protein and homologue of anticancer and antiparasitic drug targets. Biochemistry 2004, 43, (21), 6447-63. 31. Myers, R. S.; Jensen, J. R.; Deras, I. L.; Smith, J. L.; Davisson, V. J., Substrate-induced changes in the ammonia channel for im idazole glycerol phosphate synthase. Biochemistry 2003, 42, (23), 7013-22. 32. van den Heuvel, R. H.; Ferrari, D.; Bossi, R. T.; Ravasio, S.; Curti, B.; Vanoni, M. A.; Florencio, F. J.; Mattevi, A., Structural studies on the synchronization of catalytic centers in glutamate synthase. J Biol Chem 2002, 277, (27), 24579-83. 33. Douangamath, A.; Walker, M.; Beismann-Driemeye r, S.; Vega-Fernandez, M. C.; Sterner, R.; Wilmanns, M., Structural evidence for ammonia t unneling across the (beta al pha)(8) barrel of the imidazole glycerol phosphate synthase bienzyme complex. Structure 2002, 10, (2), 185-93. 34. Raushel, F. M.; Thoden, J. B.; Holden, H. M., The amidotransferase family of enzymes: molecular machines for the produc tion and delivery of ammonia. Biochemistry 1999, 38, (25), 7891-9. 35. Richards, N. G.; Kilberg, M. S., Asparagine synthetase chemotherapy. Annu Rev Biochem 2006, 75, 629-54. 36. Binda, C.; Bossi, R. T.; Wakatsuki, S.; Arzt, S.; Coda, A.; Curti, B.; Vanoni, M. A.; Mattevi, A., Cross-talk and ammonia channeling between active centers in the unexpected domain arrangement of glutamate synthase. Structure 2000, 8, (12), 1299-308. 37. Massiere, F.; Badet-Denisot, M. A., The mechanism of glutamine-dependent amidotransferases. Cell Mol Life Sci 1998, 54, (3), 205-22. 38. Weeks, A.; Lund, L.; Raushel, F. M., T unneling of intermediates in enzyme-catalyzed reactions. Curr Opin Chem Biol 2006, 10, (5), 465-72. 39. Mullins, L.; Raushel, F., Channeling of a mmonia through the intermol ecular tunnel contained within carbamoyl phosphate synthetase. JOURNAL OF THE AMERICA N CHEMICAL SOCIETY 1999, 121, (15), 3803-3804.

PAGE 149

149 40. Myers, R. S.; Amaro, R. E.; Luthey-Schul ten, Z. A.; Davisson, V. J., Reaction coupling through interdomain contacts in imid azole glycerol phosphate synthase. Biochemistry 2005, 44, (36), 11974-85. 41. Willemoes, M., Competition between ammonia derived from internal glutamine hydrolysis and hydroxylamine present in th e solution for incorporation into UTP as catalysed by Lactococcus lactis CTP synthase. Arch Biochem Biophys 2004, 424, (1), 105-11. 42. Bera, A. K.; Chen, S.; Smith, J. L.; Zalkin, H., Temperature-dependent function of the glutamine phosphoribosylpyrophosphate amidotransfe rase ammonia channel and coupling with glycinamide ribonucleotide synthetase in a hyperthermophile. J Bacteriol 2000, 182, (13), 37349. 43. Miles, B. W.; Raushel, F. M., Synchronizatio n of the three reaction ce nters within carbamoyl phosphate synthetase. Biochemistry 2000, 39, (17), 5051-6. 44. Kim, J.; Raushel, F. M., Pe rforation of the tunnel wall in carbamoyl phosphate synthetase derails the passage of ammonia between sequential active sites. Biochemistry 2004, 43, (18), 5334-40. 45. Thoden, J. B.; Huang, X.; Raushel, F. M. ; Holden, H. M., Carbamoyl-phosphate synthetase. Creation of an escape route for ammonia. J Biol Chem 2002, 277, (42), 39722-7. 46. Chaudhuri, B. N.; Lange, S. C.; Myers, R. S.; Davisson, V. J.; Smith, J. L., Toward understanding the mechanism of the complex cy clization reaction catal yzed by imidazole glycerolphosphate synthase: crystal structures of a ternary complex and the free enzyme. Biochemistry 2003, 42, (23), 7003-12. 47. Amaro, R. E.; Myers, R. S.; Davisson, V. J.; Luthey-Schulten, Z. A., Structural elements in IGP synthase exclude water to optimize ammonia transfer. Biophys J 2005, 89, (1), 475-87. 48. Luthey-Schulten, Z. A.; Amaro, R., Molecu lar dynamics simulations of substrate channeling through an / barrel protein. Chem. Phys. 2004, 307, 147-155. 49. Amaro, R.; Tajkhorshid, E.; Luthey-Schulten, Z., Developing an energy landscape for the novel function of a (beta/ alpha)8 barrel: ammonia conduction through HisF. Proc Natl Acad Sci U S A 2003, 100, (13), 7599-604. 50. Smith, J. L., Structures of glutamine amidot ransferases from the purine biosynthetic pathway. Biochem Soc Trans 1995, 23, (4), 894-8. 51. Brannigan, J. A.; Dodson, G. ; Duggleby, H. J.; Moody, P. C. ; Smith, J. L.; Tomchick, D. R.; Murzin, A. G., A protein catalytic framework with an N-terminal nucleophile is capable of selfactivation. Nature 1995, 378, (6555), 416-9. 52. Bieganowski, P.; Pace, H. C.; Brenner, C., Eukaryotic NAD+ synthetase Qns1 contains an essential, obligate intramolecular thiol glutamine amidotransferase domain related to nitrilase. J Biol Chem 2003, 278, (35), 33049-55.

PAGE 150

150 53. Sheng, S.; Moraga-Amador, D. A.; van Heeke, G.; Allison, R. D.; Richards, N. G.; Schuster, S. M., Glutamine inhibits the ammonia-depend ent activities of two Cys-1 mutants of human asparagine synthetase through the formation of an abortive complex. J Biol Chem 1993, 268, (22), 16771-80. 54. Kim, J. H.; Krahn, J. M.; Tomchick, D. R.; Smith, J. L.; Zalkin, H., Structure and function of the glutamine phosphoribosylpyrophosphate amidotra nsferase glutamine site and communication with the phosphoribos ylpyrophosphate site. J Biol Chem 1996, 271, (26), 15549-57. 55. Badet, B.; Vermoote, P.; Haumont, P. Y.; Le derer, F.; LeGoffic, F., Glucosamine synthetase from Escherichia coli: purification, proper ties, and glutamine-utilizing site location. Biochemistry 1987, 26, (7), 1940-8. 56. van den Heuvel, R. H.; Curti, B.; Vanoni M. A.; Mattevi, A., Glutamate synthase: a fascinating pathway from Lglutamine to L-glutamate. Cell Mol Life Sci 2004, 61, (6), 669-81. 57. Ciustea, M.; Gutierrez, J. A.; Abbatiello, S. E.; Eyler, J. R.; Richards, N. G., Efficient expression, purification, and characterization of C-terminal ly tagged, recombinant human asparagine synthetase. Arch Biochem Biophys 2005, 440, (1), 18-27. 58. Hurd, R. E.; John, B. K., Gradient-enhanced proton-detected heteronuclear multiple-quantum coherence spectroscopy. J. Magn. Reson. 1991, 91, 648-653. 59. Schnizer, H. G.; Boehlein, S. K.; Stewart, J. D.; Richards, N. G.; Schuster, S. M., gammaGlutamyl thioester intermediate in glutaminase reaction catalyzed by Escherichia coli asparagine synthetase B. Methods Enzymol 2002, 354, 260-71. 60. Bradford, M. M., A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the princi ple of protein-dye binding. Anal Biochem 1976, 72, 248-54. 61. Mendes, P., Biochemistry by numbers: simula tion of biochemical pathways with Gepasi 3. Trends Biochem Sci 1997, 22, (9), 361-3. 62. Mendes, P., GEPASI: a software package for modelling the dynamics, steady states and control of biochemical and other systems. Comput Appl Biosci 1993, 9, (5), 563-71. 63. Bearne, S. L.; Hekmat, O.; Macdonnell, J. E., Inhibition of Escherichia coli CTP synthase by glutamate gamma-semialdehyde and the role of the allosteric effector GTP in glutamine hydrolysis. Biochem J 2001, 356, (Pt 1), 223-32. 64. Levitzki, A.; Koshland, D. E., Jr., Cytidine triphosphate synthetase. Covalent intermediates and mechanisms of action. Biochemistry 1971, 10 (18), 3365-71. 65. Zalkin, H.; Truitt, C. D., Characterization of the glutamine site of Escherichia coli guanosine 5'-monophosphate synthetase. J Biol Chem 1977, 252, (15), 5431-6.

PAGE 151

151 66. Dossena, L.; Curti, B.; Vanoni, M. A., Activation and coupling of the glutaminase and synthase reaction of glutamate synthase is mediated by E1013 of the ferredoxin-dependent enzyme, belonging to loop 4 of the synthase domain. Biochemistry 2007, 46, (15), 4473-85. 67. Isupov, M. N.; Obmolova, G.; Butterworth, S.; Badet-Denisot, M. A.; Badet, B.; Polikarpov, I.; Littlechild, J. A.; Teplyakov, A., Substrat e binding is required for assembly of the active conformation of the catalytic si te in Ntn amidotransferases: evidence from the 1.8 A crystal structure of the glutaminase domain of glucosamine 6-phosphate synthase. Structure 1996, 4, (7), 801-10. 68. Vanoni, M. A.; Curti, B., Structure--func tion studies on the iron-sulfur flavoenzyme glutamate synthase: an unexpectedly complex self-regulated enzyme. Arch Biochem Biophys 2005, 433, (1), 193-211. 69. Su, N.; Pan, Y. X.; Zhou, M.; Harvey, R. C.; Hunger, S. P.; Kilb erg, M. S., Correlation between asparaginase sensitivity and asparagine synthetase protein content, but not mRNA, in acute lymphoblastic leukemia cell lines. Pediatr Blood Cancer 2007. 70. Cui, H.; Darmanin, S.; Natsuisaka, M.; Kondo, T.; Asaka, M.; Shindoh, M.; Higashino, F.; Hamuro, J.; Okada, F.; Kobayashi, M.; Nakagawa, K.; Koide, H.; Kobayashi, M., Enhanced expression of asparagine synthetase under gluc ose-deprived conditions protects pancreatic cancer cells from apoptosis induced by glucose deprivation and cisplatin. Cancer Res 2007, 67, (7), 3345-55. 71. Gutierrez, J. A.; Pan, Y. X.; Koroniak, L.; Hi ratake, J.; Kilberg, M. S.; Richards, N. G., An inhibitor of human asparagine sy nthetase suppresses proliferation of an L-asparaginase-resistant leukemia cell line. Chem Biol 2006, 13 (12), 1339-47. 72. Horowitz, B.; Meister, A., Glutamine-dependent asparagine synthetase from leukemia cells. Chloride dependence, mechanis m of action, and inhibition. J Biol Chem 1972, 247 (20), 670819. 73. Gong, S. S.; Guerrini, L.; Basilico, C., Regula tion of asparagine synthetase gene expression by amino acid starvation. Mol Cell Biol 1991, 11, (12), 6059-66. 74. Cedar, H.; Schwartz, J. H., Th e asparagine synthetase of Escherhic coli. I. Biosynthetic role of the enzyme, purificati on, and characterization of the reaction products. J Biol Chem 1969, 244 (15), 4112-21. 75. Cedar, H.; Schwartz, J. H., The asparagine synthetase of Escheric hia coli. II. Studies on mechanism. J Biol Chem 1969, 244, (15), 4122-7. 76. Herrera-Rodriguez, M. B.; Perez-Vicente, R.; Maldonado, J. M., Expr ession of asparagine synthetase genes in sunflower (Helianthus annuus) under various environmental stresses. Plant Physiol Biochem 2007, 45, (1), 33-8.

PAGE 152

152 77. Herrera-Rodriguez, M. B.; Maldonado, J. M. ; Perez-Vicente, R., Role of asparagine and asparagine synthetase genes in sunflower (Helianthus annuus) germination and natural senescence. J Plant Physiol 2006, 163 (10), 1061-70. 78. Herrera-Rodriguez, M. B.; Maldonado, J. M.; Perez-Vicente, R., Light and metabolic regulation of HAS1, HAS1.1 and HAS2, three asparagine synthetase genes in Helianthus annuus. Plant Physiol Biochem 2004, 42, (6), 511-8. 79. Herrera-Rodrguez, M. B.; Ca rrasco-Ballesteros, S.; Maldonado, J. M.; Pineda, M.; Aguilar, M.; Prez-Vicente, R., Three genes showing distin ct regulatory patterns encode the asparagine synthetase of sunflower (Helianthus annuus ). New Phytologist 2002, 155 33-45. 80. Koroniak, L.; Ciustea, M.; Gutierrez, J. A.; Ri chards, N. G., Synthesis and characterization of an N-acylsulfonamide inhibitor of human asparagine synthetase. Org Lett 2003, 5, (12), 2033-6. 81. Krahn, J. M.; Kim, J. H.; Burns, M. R.; Parry, R. J.; Zalkin, H.; Smith, J. L., Coupled formation of an amidotransferase interdomai n ammonia channel and a phosphoribosyltransferase active site. Biochemistry 1997, 36, (37), 11061-8. 82. van den Heuvel, R. H.; Svergun, D. I.; Pet oukhov, M. V.; Coda, A.; Curti, B.; Ravasio, S.; Vanoni, M. A.; Mattevi, A., The active conformati on of glutamate synthase and its binding to ferredoxin. J Mol Biol 2003, 330, (1), 113-28

PAGE 153

153 BIOGRAPHICAL SKETCH Kai Li was born in Zouping, China on August 2, 1974. After he go t his Bachelor of Science degree in biochemistry from Nankai Univer sity in Tianjin, China, he went to Beijing, China, and became a graduate student in Plan t medicine in China Academy of Agriculture Sciences (CAAS), where he got his Master of Sc ience degree. From 1998 to 2001, he worked in China Institute of Veterinary Drug Control (IVDC) and did residue analysis of veterinary drugs in food such as meat, egg and milk. In September 2002, he was enrolled in graduate school of State University of New York at Albany. Due to family reason, he transferred to University of Florida after being in Albany for about 9 mont hs. From 2003 to 2007, he studied asparagine synthetase under the supervision of Dr. Nigel G. J. Richards.