Báo cáo khoa học: Alternative substrates for wild-type and L109A E. coli CTP synthases Kinetic evidence for a constricted ammonia tunnel doc

Alternative substrates for wild-type and L109A
E. coli
CTP synthases
Kinetic evidence for a constricted ammonia tunnel
Faylene A. Lunn and Stephen L. Bearne
Department of Biochemistry and Molecular Biology, Dalhousie University, Halifax, Nova Scotia, Canada
Cytidine 5¢-triphosphate (CTP) synthase catalyses t he ATP-
dependent formation of CTP from uridine 5¢-triphosphate
using either NH
3
or
L
-glutamine as the nitrogen s ource. The
hydrolysis of glutamine is c atalysed in the C-terminal glu-
tamine amide t ransfer domain a nd the n ascent NH
3
that is
generated is transferred via an NH
3
tunnel [Endrizzi, J.A.,
Kim, H., Anderson, P.M. & Baldwin, E.P. (2004)
Biochemistry 43, 6447–6463] to the active site of the N-ter-
minal synthase domain where the a mination reaction occurs.
ReplacementofLeu109byalanineinEscherichia coli CTP
synthase causes an uncoupling of g lutamine hydrolysis and
glutamine-dependent CTP formation [Iyengar, A . & Bearne,
S.L. (2003) Biochem. J. 369 , 497–507]. To test our hypot hesis
that L109A CTP synthase has a constricted or a leaky NH
3
tunnel, we e xamined the ability of wild-type and L109A
CTP synthases to utilize NH

3
,NH
2
OH, and NH
2
NH
2
as
exogenous substrates, and as nascent substrates generated
via the hydrolysis of glutamine, c-glutamyl hydroxamate,
and c-glutamyl hydrazide, respectively. We show that the
uncoupling of the hydrolysis of c-glutamyl hydroxamate and
nascent NH
2
OH production from N
4
-hydroxy-CTP for-
mation is more pronounced with the L109A enzyme, relative
to the wild-type C TP synthase. These results suggest that the
NH
3
tunnel o f L109A, in the presence of bound allosteric
effector guanosine 5¢-triphosphate, is not leaky but contains
a constriction that discriminates b etween NH
3
and NH
2
OH
on the basis of size.
Keywords: amidotransferase; ammonia tunnel; CTP syn-

thase; glutaminase; alternative substrates.
Cytidine 5¢-triphosphate (CTP) synthase [CTPS;
EC 6.3.4.2; UTP:ammonia ligase (ADP-forming)] catalyses
the ATP-dependent formation of CTP from UTP using
either
L
-glutamine (Gln) or NH
3
as the nitrogen source [1,2].
This Gln amidotransferase is a single polypeptide chain
consisting of two domains. The C-terminal Gln amide
transfer (GAT) domain utilizes a C ys-His-Glu triad to
catalyse the rate-limiting hydrolysis of Gln (glutaminase
activity) [3–5], and the nascent NH
3
derived from this
glutaminase a ctivity is transferred to the N-terminal
synthase domain wh ere the a mination of a phosphorylated
UTP intermediate is catalysed [6,7]. The reactions catalysed
by CTPS are summarized in Scheme 1.
CTPS catalyses the final step in the de novo synthesis o f
cytosine nu cleotides. As CTP h as a central role in the
biosynthesis of nucleic acids [8] and membrane phospho-
lipids [9], C TPS is a recognized target for the development
of antineoplastic agents [8,10], antiviral agents [ 10–12], and
antiprotozoal agents [13–15]. The Escherichia coli enzyme
is one of the most thoroughly characterized CTP synthases
with respect to its physical and kinetic p roperties, and i s
regulated in a c omplex fashion [1]. GTP i s required as a
positive allosteric effector to increase the efficiency of the

glutaminase activity and Gln-dependent CTP synthesis
[3,16] but inhibits CTP synthesis at concentrations
>0.15 m
M
[17]. I n a ddition, the enzyme is inhibited by the
product CTP [18] and displays positive cooperativity for
ATP and UTP [ 18–20]. ATP and UTP act synergistically t o
promote tetramerization of the enzyme to its active form [20].
Recently, the X -ray crystal structure of E. coli CTP S was
solved at a resolution of 2 .3 A
˚
[21]. The enzyme crystallised
as a tetramer, presumably because of the high protein
concentrations used as bound nucleotides were not present
in the structure (i.e. a po-E. coli CTPS) [21]. The authors
identified a solvent-filled ÔvestibuleÕ ( 230 A
˚
3
) that connects
the GAT active site and the GAT/synthase interface. This
vestibule is connected to a t ubular passage that leads into the
synthase site. The presence of this vestibule a nd NH
3
tunnel
in CTPS is consistent with the identification of NH
3
tunnels
in the X-ray structures of other amidotransferases inclu-
ding carbamoyl phosphate synth ase (CPS) [22–24], Gln
phosphoribosylpyrophosphate [25,26], GMP synthase [27],

glucosamine-6-phosphate synthase [28–30], asparagine
synthase B [31], and anthranilate synthase [32,33].
Previously, we reported that amino acid residues between
Arg105 and Gly110 of E. coli CTPS are important for
efficient coupling of Gln hydrolysis in the GAT domain to
CTP formation in the s ynthase domain. Replacement of the
highly conserved L eu109 residue by alanine produced an
enzyme that exhibited w ild-type levels of NH
3
-dependent
CTP formation, affinity for Gln, glutaminase activity,
Correspondence to S. L. Bearne, Department of Biochemistry and
Molecular Biology, Dalhousie University, Halifax, Nova Scotia,
B3H 1X5, Canada. Fax: +1 902 494 1355, Te l.: +1 902 494 1974,
E-mail:
Abbreviations: CPS, carbamoyl phosphate synthase; CTPS, CTP
synthase; GAT, Gln amide transfer; Gln,
L
-glutamine; Gln-OH,
L
-c-glutamyl hydroxamate; Gln-NH
2
,
L
-c-glutamyl hydrazide;
OPA, o-phthaldialdehyde.
Enzyme: CTP synthase (EC 6 .3.4.2)
(Received 1 8 August 2 004, revised 3 September 2004,
accepted 6 September 2004)
Eur. J. Biochem. 271, 4204–4212 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04360.x

affinity for GTP, and activation by GTP. Most interest-
ingly, however, the L109A mutant exhibited impaired
Gln-dependent CTP formation. These observations were
consistent with the h ypothesis that L eu109 plays a role i n
either the structure or formation of a n NH
3
tunnel and
ensures efficient c oupling o f th e Gln h ydrolysis a nd amina-
tion reactions. In the present report, we show that hydroxyl-
amine,
L
-c-glutamy l hydroxamate (Gln-OH), hydrazine,
and
L
-c-glutamyl hydrazide (Gln-NH
2
) are alternative
substrates for E. coli CTPS. Comparison of the kinetic
parameters of Gln and NH
3
with those of the co rresponding
bulkier substrates Gln-OH and NH
2
OH suggests that the
impaired Gln-dependent CTP formation exhibited by the
L109A mutant is caused by a constriction of the NH
3
tunnel. This is the fi rst functional evidence implicating a
constriction in the N H
3

tunnel of E. coli CTPS.
Experimental procedures
General materials and methods
All chemicals were purchased from Sigma-Aldrich Canad a
Ltd. (Oakville, ON, Canada), except where mentioned
otherwise. For HPLC experiments, a Waters 510 pump and
680 controller were used for solvent delivery. Injections were
made using a Rheodyne 7725i sample injector fitted with a
20 lL injection loop.
Enzyme expression and purification
Wild-type and L109A recombinant E. coli CTPS were
expressed in and purified from E. coli strain BL21(DE3)
cells transformed with the plasmid p ET15b-CTPS1 or the
mutated plasmid as described previously [3,34]. These
constructs encode the E. coli pyrG gene product with an
N-terminal His
6
-tag. The BL21(DE3) cells were grown in
Luria–Bertani medium at 37 °C, induced using isopropyl
thio-b-
D
-galactoside according to the Novagen expression
protocol [35], and lysed using sonication on ice ( 5 · 10 s
bursts w ith 30 s intervals at output setting 5 using a Branson
Sonifier 250). The crude lysate was clarified by centrifugation
(39 000 g,20min,4°C) and the soluble histidine-tagged
CTPS was purified using metal ion affinity chromatography
as described in the Novagen protocols [35]. The resulting
enzyme solution was dialysed into HEPES buffer (70 m
M

,
pH 8.0) containing EGTA (0.5 m
M
). All enzyme purifica-
tion procedures were conducted at 4 °C.
Thrombin-catalysed cleavage of the histidine tag f rom
soluble enzyme (new N-terminus, GSHMLEM
1
…)was
conducted in HEPES buffer (70 m
M
, pH 8.0) containing
EGTA (0.5 m
M
) using a thrombin ratio of 0.5 unitsÆmg
)1
of
target protein. After 8 h at 25 °C, cleavage was complete
and the biotinylated thrombin was removed from the
reaction mixture using streptavidin agarose r esin (Novagen,
EMD Biosciences, Inc., Madison, WI, USA) at a ratio of
32 lL settled re sin per unit of thrombin following the
Novagen protocol [35]. Cleaved CTPS, free of biotinylated
thrombin, was then dialysed against HEPES buffer (70 m
M
,
pH 8.0) containing EGTA (0.5 m
M
)andMgCl
2

(10 m
M
)
(assay buffer). The results o f the purification and cleavage
procedures were routinely monitored using SDS/PAGE.
Typically, enzyme preparations were P 98% pure. The
amino acid residues in the recombinant wild-type and
mutant enzymes are numbered according to the s equence of
the w ild-t ype E. co li enzyme starting w ith M
1
as position 1.
Cyclization of Gln-OH
The c onversion of Gln-OH to 2-pyrrolidone-5-carboxylic
acid [ 36] a t 37 °C w as followed using a Bruker AVANCE
500 M Hz NMR spectrometer. A solution of Gln-OH
(20 m
M
) in deuterated potassium phosphate buffer
(100 m
M
, pD 8.0) was prepared and t he ionic strength
was adjusted to 0.30
M
using KCl. At various times (5, 7, 16,
26, 36, and 4 6 m in) the
1
H N MR spectrum was recorded.
The relative concentrations of Gln-OH and 2-pyrrolidone-
5-carboxylic acid were determined by integration of the
signals at 3.80 p.p.m. (triplet) and 4.22 p.p.m. (multiplet)

corresponding to the proton on the carbon adjacent to the
carboxylate carbon on Gln-OH and 2-pyrrolidone-5-carb-
oxylic acid, r espectively. (Chemical s hifts are relative to the
D
2
O lock signal.)
L-glutamine
R = H, OH, NH
2
; R' = ribose-5'-triphosphate
UTP R = H CTP
R = OH N
4
-hydroxy CTP
R = NH
2
N
4
-amino CTP
L-glutamate
H
2
O
tunnel with
constriction
or leak
exogenous H
2
N–R
[3]

[2]
[1]
N
N
R'
O
HN
HN
N
R'
O
O
O
NH
3
O
OO
–
+
H
O
O
N
NH
3
O
–
+
R
–

HN R
2
glutaminase
reaction
synthase
reaction
R
N
N
R'
O
OPO
3
2
AT P A DP
–
P
i
Scheme 1
nascent H
2
N–R from leak
equilibrates with solvent
[4]
phosphorylated UTP
intermediate
Scheme 1. Reactions catalysed by E. coli CTP synthase.
Ó FEBS 2004 Alternative substrates for E. coli CTP synthase (Eur. J. Biochem. 271) 4205
For t hose experiments utilizing Gln-OH a s t he substrate,
we found that it was essential to maintain the Gln-OH stock

solution at 4 °C and add this solution directly to t he assay
cocktail to initiate the reaction. At 37 °C, the observed
first order rate constant for cyclization of Gln-OH to
form 2-pyrrolidone-5-carboxylate and NH
2
OH was
7.7 ( ± 0.4) · 10
)5
s
)1
(i.e. t
1/2
 2.5h)atpD8.0(data
not shown). H ence, significant production of NH
2
OH occurs
in Gln-OH solutions at 37 °C and the resulting NH
2
OH c an
complicate kinetic experiments if the Gln-OH solutions are
not kept on ice prior to addition to the assay solution.
Enzyme assays and protein determinations
CTPS activity was d etermined at 3 7 °C u sing a c ontinuous
spectrophotometric a ssay by following the rate of increase
in absorbance at 291 n m r esulting from either the c onver-
sion of UTP to CTP (De ¼ 1338
M
)1
Æcm
)1

) [18], to
N
4
-hydroxy-CTP (De ¼ 40 23
M
)1
Æcm
)1
)[37],orto
N
4
-amino-CTP (De ¼ 1364
M
)1
Æcm
)1
;estimatedfromthe
difference of the spectra of uridine and N
4
-amino cytidine).
Substrates (NH
4
Cl, NH
2
OH, NH
2
NH
2
, Gln, G ln-OH, and
Gln-NH

2
) were dissolved in assay buffer and the pH was
adjusted to 8.0 using 6
M
KOH. The standard assay m ixture
consisted of HEPES buffer (70 m
M
, pH 8.0) containing
EGTA (0.5 m
M
), Mg Cl
2
(10 m
M
), CTPS, and saturating
concentrations of UTP (1 m
M
)andATP(1m
M
)inatotal
volume of 1 mL. Enzyme and nucleotides were p reincu-
bated together f or 2.5 m in at 37 °C followed b y addition of
substrate to initiate the reaction. Total NH
4
Cl concentra-
tions in the assays were 5, 10, 20, 30, 50, 60, 80, and 100 m
M
;
total NH
2

OHÆHCl concentrations in the assays were 5, 10,
15, 20, 30, 40, 50, 75 , and 100 m
M
;totalNH
2
NH
2
Æ2HCl
concentrations in the assays were 5, 10, 20, 30, 40, 50, 60, 70,
80, 90, and 100 m
M
; and CTPS concentrations were
20 lgÆmL
)1
for wild-type and 20–24 lgÆmL
)1
for L109A.
For assays of Gln- or Gln analogue-dependent CTP
formation, concentrations of Gln were 0.1, 0.2, 0 .3, 0.5,
1.0, 2.0, 3.0, and 6.0 m
M
; concentrations o f Gln-OH were
0.1, 0.2, 0.3, 0.5, 1.0, 2.0, 5.0, 10.0, and 15.0 m
M
; concen-
trations of Gln-NH
2
were 0.5, 1.0, 5.0, 10.0, 15.0, 20.0, 25.0,
40.0, 60.0, 80.0, and 100.0 m
M

; and CTPS concentrations
ranged betw een 28 and 120 lgÆmL
)1
for wild-type and
40–160 lgÆmL
)1
for L109A. The concentration of GTP was
maintained at 0.25 m
M
forallassayswhenGlnorGln
analogues were used as the substrate. For assays conducted
using Gln-OH, a freshly prepared stock solution w as stored
on ice and added cold to each a ssay. This protocol was
necessary to minimize cyclization o f Gln-OH with concom-
itant production of NH
2
OH (see above).
The i onic strength was maintained at 0.30
M
in all assays
by the addition of KCl. All kinetic parameters were
determined in triplicate and average values are reported.
The reported errors are standard deviations. Initial rate
kinetic data was fit to Eqn (1) by nonlinear regression
analysis using the program
PRISM
(GraphPad Software,
Inc., San Diego, CA).
v
i

¼
V
max
½S
K
m
þ½S
ð1Þ
In Eqn (1), v
i
is the initial velocity, V
max
(¼ k
cat
[E]
T
)isthe
maximal velocity a t saturating substrate concentrations,
[S] is the substrate concentration, and K
m
is the M ichaelis
constant for t he substrate. Values of K
m
for NH
3
,NH
2
OH,
and NH
2

NH
2
were calculated using the concentration of
these species present at pH 8 .0 (i.e. pK
a
(NH
4
+
) ¼ 9.24;
pK
a
(
+
NH
3
OH) ¼ 5.97; p K
a
(NH
2
NH
3
+
) ¼ 8.10 [38]). Val-
ues of k
cat
(per subunit) were calculated for CTPS variants
with the His
6
-tag removed using the molecular masses ( Da)
of 61 0 29 (wild-type) and 60 987 (L109A). Protein concen-

trations were determined using the Bio-Rad Protein Assay
(Bio-Rad Laboratories, Hercules, CA, USA) with bovine
serum albumin standards.
Glutaminase assay
The abilities of wild-type and L109A CTP synthases to
catalyse Gln hydrolysis were determined by following the
production of glutamate using reversed-phase HPLC
separation of the o-phthaldialdehyde (OPA) derivatives
of glutamate, Gln, Gln-OH, and Gln-NH
2
with fluores-
cence detection [39]. Assays were conducted at 37 °Cin
HEPES buffer (70 m
M
, pH 8.0) containing EGTA
(0.5 m
M
), MgCl
2
(10 m
M
), ATP (1 m
M
), UTP (1 m
M
),
GTP (0.25 m
M
), and either Gln (0.25, 0.50, 1.0, 5.0, and
10.0 m

M
), Gln-OH (1.0, 3.0, 5 .0, 7.0, and 10 m
M
), or Gln-
NH
2
(3.0, 7.0, 10.0, 15.0, and 20.0 m
M
). CTPS concen-
trations ranged between 5 and 56 lgÆmL
)1
for wild-type
and 5–54 lgÆmL
)1
for L109A in a total volume of
2.5 m L.
All components were preincubated for 2.5 min at 37 °C
prior to initiation of the reaction by addition of substrate
(Gln, Gln-OH, or Gln-NH
2
). To minimize cyclization of
Gln-OH, stock solutions (1 mL) were prepared at appro-
priate concentrations and fl ash-frozen in liquid n itrogen.
These Gln-OH solutions were thawed for 2.5 min at
37 °C and then used to initiate the reaction. At various
time po ints (0, 1, 3, 5, 7, and 10 min), a liquots (20 lL) of
the assay solution were transferred to 1.5 mL polypropy-
lene tubes and reacted immediately with an equal volume
of OPA reagent (40 m
M

) [39]. Derivatization with OPA
was shown t o effectively t erminate the reaction. (Boiling
of the reaction led to rapid cyclization of the Gln-OH
[36].) After 1 min a t r oom temperature the reaction was
neutralized by addition of sodium acetate buffer (160 lL,
0.1
M
, pH 6.2) and an aliquot (20 lL) was analysed u sing
reversed-phase HPLC.
Separation of the isoindole derivatives of Gln, g lutamate,
Gln-OH, and Gln-NH
2
were conducted using a Synergi
Fusion-RP column (4 lm; 80 A
˚
;50· 4.6 mm; Pheno-
menex, Torrance, C A) eluted under isocratic conditions
using 0 .1
M
sodium acet ate (adjusted to pH 6.2 with g lacial
acetic acid)/methanol/tetrahydrofuran (800 : 190 : 10; v/v/
v) at a flow rate of 1.5 mLÆmin
)1
. The solvent was degassed
prior to use. The fluorescence of the isoindole derivatives
formed from reactio n of Gln, glutamate, Gln-OH, and Gln-
NH
2
with OPA reagent was monitored using a Waters
474 scanning fluorescence detector (k

ex
¼ 343 nm, k
em
¼
440 n m). These derivatives eluted with retention times equal
to 5.6, 2.1 , 4.4, and 3.8 min, respectively. Peak areas were
determined by integration of the resulting c hromatograms
using
PEAKSIMPLE
software from Mandel Scientific (Guelph,
ON, C anada). Concentrations of glutamate were deter-
mined u sing a standard curve prepared by derivatization of
4206 F. A. Lunn and S. L. Bearne (Eur. J. Biochem. 271) Ó FEBS 2004
standard glutamate solutions (0.025, 0.050, 0.075, 0.100,
0.150, 0.200, and 0.250 m
M
).
Calculations
Geometry optimizations and electrostatic potential sur-
faces were calculated for N H
3
,NH
2
OH, and NH
2
NH
2
by performing self-consistent-field calculations at the
6–31 G** level using
SPARTAN

¢04
WINDOWS
(Wavefunction,
Inc., Irvine, CA). This software was a lso used to c alculate
the molecular surface areas and volumes of these molecules.
Results and discussion
Previously, we reported that replacement of the highly
conserved L eu109 residue in E. coli CTPS by alanine yields
an enzyme that has kinetic prop erties similar to t hose of
wild-type CTPS with respect to NH
3
-dependent CTP
formation, affinity for Gln, glutaminase activity, affinity
for G TP, and activation by GTP [34]. However, unlike wild-
type CTPS, the L109A mutant exhibited impaired Gln-
dependent CTP f ormation. These observations suggested
that Leu109 plays a role in either the structure or formation
of an NH
3
tunnel and ensures efficient coupling of the Gln
hydrolysis and amination reactions. In the present study, we
use bulky analogues of exogenous N H
3
(i.e. NH
2
OH and
NH
2
NH
2

) and nascent NH
3
(i.e. NH
2
OH and NH
2
NH
2
derived from the hydrolysis of Gln-OH and Gln-NH
2
,
respectively) to test our hypothesis that the presence of a n
alanine at position 109 introduces a constriction in the NH
3
tunnel of E. coli CTPS. T his approach has been u sed to
demonstrate that t he G359S m utant o f CPS has a partially
blocked NH
3
tunnel that prevents diffusion of NH
2
OH
while still allowing some NH
3
to diffuse through [40]. T he
hypothesis that r eplacement of the bulky Leu109 by the
smaller a lanine could cause a tunnel blockage has precedent.
For example, the F334A m utant of glutamine phosphorib-
osylpyrophosphate amidotransferase exhibited kinetic prop-
erties expecte d fo r a blocked or disrupted NH
3

tunnel [41].
Many amidotransferases can utilize N H
2
OH and
NH
2
NH
2
in place of NH
3
[42–44]. Both NH
2
NH
2
and
NH
2
OH (and its derivatives NH
2
OCH
3
and CH
3
NHOH)
are substrates for CTPS from Ehrlich ascites tumour cells
[45], a nd NH
2
OH has been sh own to be a substrate for
E. coli [46] and Lactococcus lactis [47] CTP synthases. W ith
the exception of Lieberman’s work in 1956 [46], little is

known about the ability o f E. coli CTPS to utilize a lter-
native NH
3
sources. In a ddition, some amidotransferases
such as CPS [40] a nd asparagine synthase B [48] have been
showntohydrolyseGln-OHandGln-NH
2
to give rise to
NH
2
OH and NH
2
NH
2
, respectively. Although E. coli
CTPS has be en shown to utilize Gln-OH as a substrate
[16], the present study describes the first detailed kinetic
characterization of the ability of E. coli CTPS to utilize
alternative substrates. We show that replacement of Leu109
by alanine in E. coli CTPS causes the enzyme to discrim-
inate between nascent NH
3
and the bulkier analogue
NH
2
OH based on size but does not lead to discrimination
between exogenous NH
3
and bulkier analogues (i.e.
NH

2
OH and NH
2
NH
2
). Our findings are consistent with
the L109A mutation causing constriction of an NH
3
tunnel.
Exogenous NH
3
and its analogues
Exogenous NH
3
,NH
2
OH, and NH
2
NH
2
all served as
substrates for wild-type and L109A CTP synthases
(Table 1). However, both NH
2
OH and NH
2
NH
2
exhibited
k

cat
/K
m
values with both enzymes that were approximately
30-fold less t han t he k
cat
/K
m
value for NH
3
. This r eduction
in k
cat
/K
m
wascausedbyanincreasedK
m
value for NH
2
OH
and NH
2
NH
2
.RelativetoNH
3
,theK
m
value for NH
2

OH
was increased approximately 40-fold while the k
cat
value
was s lightly greater than that for NH
3
. T his observation is in
accord with the slightly greater nucle ophilicity o f NH
2
OH
relative to NH
3
[49,50]. Hence, it appears that once NH
2
OH
is bound it reacts readily with the phosphorylated UTP
intermediate. The individual K
m
and k
cat
values for
NH
2
NH
2
could not be determined for either wild-type
CTPS or L109A CTPS because saturation was not
observed, indicating that the K
m
for this s ubstrate w as also

markedly increased relative to that observed for NH
3
.
Three possible routes that exogenous NH
3
or its
analogues m ight traverse to reach t he site of reaction with
the phosphorylated UTP intermediate are shown in
Table 1. Kinetic Param eters for w ild-type and L 109A CTP synthases. – , Not determined.
Substrate
Wild-type CTPS L109A CTPS
K
m
(m
M
) k
cat
(s
)1
) k
cat
/K
m
(m
M
)1
Æs
)1
) K
m

(m
M
) k
cat
(s
)1
) k
cat
/K
m
(m
M
)1
Æs
)1
)
Kinetic parameters for CTP formation
NH
3
2.15 ± 0.14 9.50 ± 0.53 4.43 ± 0.12 2.17 ± 0.09 10.1 ± 0.3 4.63 ± 0.04
NH
2
OH 82.8 ± 6.8 14.0 ± 1.8 0.169 ± 0.016 75.3 ± 9.8 14.1 ± 1.9 0.187 ± 0.003
NH
2
NH
2
–
a
–

a
0.147 ± 0.019 –
a
–
a
0.128 ± 0.015
Gln 0.354 ± 0.057 6.10 ± 0.80 17.8 ± 2.3 0.497 ± 0.132 1.86 ± 0.34 3.85 ± 0.82
Gln-OH 0.165 ± 0.017 0.453 ± 0.001 2.77 ± 0.28 0.250 ± 0.091 0.063 ± 0.014 0.260 ± 0.061
Gln-NH
2
39.4 ± 0.5 1.41 ± 0.04 0.036 ± 0.001 – – –
Kinetic parameters for the glutaminase activity
Gln 0.327 ± 0.002 5.62 ± 0.12 17.2 ± 0.1 0.550 ± 0.012 5.06 ± 0.24 9.22 ± 0.62
Gln-OH 0.324 ± 0.101 0.930 ± 0.040 3.06 ± 0.90 0.260 ± 0.061 0.310 ± 0.033 1.26 ± 0.40
Gln-NH
2
–
a
–
a
0.034 ± 0.004 –
b
–
b
–
b
a
Saturation could not be achieved and k
cat
/K

m
values were determined from measurements conducted with [S] << K
m
.
b
Activity too low to
measure reliably.
Ó FEBS 2004 Alternative substrates for E. coli CTP synthase (Eur. J. Biochem. 271) 4207
Scheme 1. Route 1 represents a bimolecular reaction with
the reactive intermediate. This route is unlikely as saturation
kinetics are observed when NH
3
is a substrate, su ggesting
the formatio n of an initial enzyme-NH
3
complex. Routes 2
and 3 involve the binding of NH
3
at a site on CTPS followed
by either direct reaction with the phosphorylated UTP
intermediate (route 2) or passage through an internal tunnel
to its site of reaction with the phosphorylated UTP
intermediate (route 3). Although structural studies of many
different a midotransferases h ave suggested the presence of
NH
3
tunnels to shuttle the nascent NH
3
from the site of Gln
hydrolysis to the synthase domain [51], it is not always clear

what route is f ollowed b y exogenous NH
3
.Foranygiven
exogenous substrate (i.e. NH
3
,NH
2
OH, or NH
2
NH
2
), the
kinetic p arameters ( K
m
, k
cat
, a nd/or k
cat
/K
m
) a re sim ilar f or
both wild-type a nd L109A E. coli CTP synthases. Thus,
replacement of Leu109 by alanine does not cause any
discrimination between exogenous substrates of a g iven size
with respect to binding affinity, t urnover, and efficiency. In
addition, once the bulkier, exogenous NH
2
OH enters the
enzyme, i t is transferred to the synthase active site and reacts
with the phosphorylated UTP i ntermediate as e fficiently as

NH
3
as indicated by the similar k
cat
values for either the
wild-type o r L109A CTP synthases. B ased on their recently
solved crystal structure of wild-type E. coli CTPS, Endrizzi
et al . [21] suggested that exogenous NH
3
could access the
active site via a ÔholeÕ on the p rotein’s surface that resides
midway between the Gln and UTP binding sites (Fig. 1).
Our observations suggest that, after binding to L109A
CTPS, p assage of exogenous NH
3
or its analogues through
the NH
3
tunnel (i.e. route 3) are not inhibited by a
constriction if it is present. Alternatively, a constriction may
be present a t a location within the NH
3
tunnel that i s closer
to the site of Gln hydrolysis so that exogenous substrates
entering through the hole bypass the constriction.
It is important to note t hat both the wild-type and L109A
enzymes do discriminate between NH
3
and the bulkier
substrates in terms of binding (i.e. elevated K

m
values for
NH
2
OH and N H
2
NH
2
relative to NH
3
for both w ild-type
and L109A CTP s ynthases). The entrance for exogenous
NH
3
is approximately 3 A
˚
in diameter thereby p ermitting
access of NH
3
(surface area ¼ 43.65 A
˚
2
; volume ¼
26.52 A
˚
3
) [21]. On the other hand, entrance of bulkier
substrates such as NH
2
OH (surface area ¼ 54.89 A

˚
2
;
volume ¼ 35.62 A
˚
3
)andNH
2
NH
2
(surface area ¼
60.23 A
˚
2
; volume ¼ 40.62 A
˚
3
) may be more difficult and
require proper orientation of these molecules a long their
longitudinal a xis in order to pass through the hole and avoid
unfavourable steric interactions. This r equirement for
correct orientation could, in part, be responsible for the
elevated K
m
values observed for the bulkier substrates. The
electrostatic potential surfaces of NH
3
,NH
2
OH, and

NH
2
NH
2
(data not shown), and their ability to act as
hydrogen bond donors and acceptors are similar, and hence
they are expected to behave similarly within the proteins,
provided no adverse steric interactions are encountered.
Nascent NH
3
and its analogues
The abilities of wild-type and L109A CTP synthases to
catalyse the hydrolysis o f Gln, Gln-OH, and Gln-NH
2
(i.e.
glutaminase activity) and to subsequently catalyse the
formation o f C TP, N
4
-hydroxy-CTP, and N
4
-amino-CTP,
respectively, were examined (Table 1). R elative to Gln, the
k
cat
/K
m
values for Gln-OH and Gln-NH
2
hydrolysis
catalysed by wild-type CTPS were reduced approximately

Fig. 1. Location of Leu109 relative to the opening for exogenous NH
3
(PDB code 1S1M [21]). (A) Amino acid residues comprising the walls of the
entryway for exogenous NH
3
include residues 50–55, Val60, Glu68, Lys297, Tyr298, Ala304, Phe353, Gly354, Arg356, Glu403, and Arg468 (shown
in green, space-filling representation). The sulphur of t he catalytic nucleophile Cys379 is yellow. The loop comprised o f r esidues 104–110 f rom the
adjacent subunit is shown in red with Leu109 shown in space-filling representation. (B) Viewed from the side, relative t o (A), Leu109 is poised above
the o pening for e xogenous NH
3
. G TP is shown modelled i nto the cleft [21], however, this m odel probably does not ac curately reflect th e change in
conformation associated with GTP binding. Movement o f t he 104–110 loop may occur upon GTP binding so that Leu109 is repositioned to pack
against bound GTP and perhaps h elp further seal the e ntryway for exogenous NH
3
.
4208 F. A. Lunn and S. L. Bearne (Eur. J. Biochem. 271) Ó FEBS 2004
six-fold and 500-fold, respectively. The same trend is also
observed for wild-type CTPS-catalysed formation of
N
4
-hydroxy-CTP and N
4
-amino-CTP. Comparison of the
K
m
and k
cat
values for Gln-OH hydrolysis with those
observed for Gln hydrolysis reveals that this reduction in
efficiency a rises f rom a six-fold reduction in k

cat
while there
is no change in the K
m
value. The marked reduction in t he
efficiency (k
cat
/K
m
) of wild-type CTPS-catalysed formation
of N
4
-hydroxy-CTP resulted mainly from a 111-fold
increase in the K
m
value. A similar trend is also observed
with L109A C TPS. Unfortunately, w e were unable to detect
any significant amount of glutaminase activity using L109A
CTPS with Gln-NH
2
as a substrate. Consequen tly, we were
not able to employ nasc ent NH
2
NH
2
in our analysis f or
tunnel constriction.
The values of k
cat
/K

m
and k
cat
for wild-type CTPS-
catalysed Gln hydrolysis and CTP formation are experi-
mentally equal. This indicates t hat there is total coupling of
the reaction s forming the nascent N H
3
and i ts reac tion to
form CTP a t both l ow (i.e. k
cat
/K
m
conditions) and high (i.e.
k
cat
conditions) concentrations of Gln. However, when Gln-
OH is the substrate, N
4
-hydroxy-CTP formation is only
fully coupled to Gln-OH hydrolysis when the concentration
of Gln-OH is subsaturating (Table 1). To illustrate how this
coupling is altered wh en either the nature o f the substrate o r
enzyme is altered, w e employ two Ôcoupling ratiosÕ as defined
in Eqns 2 a nd 3, and reported in Figs 2 and 3. Such ratios
have been used to characterize the c hannelling efficiency of
amidotransferases [41].
Subsaturating coupling ratio ¼
ðk
cat

=K
m
Þ
CTP for mation
ðk
cat
=K
m
Þ
glutaminase activity
ð2Þ
Saturating coupling ratio ¼
ðk
cat
Þ
CTP f ormat ion
ðk
cat
Þ
glutami nasea c tivit y
ð3Þ
For wild-type CTPS, these ratios are both unity for Gln and
Gln-OH at subsaturating concentrations of the substrate
(Fig. 2 ) indicating that the n ascent NH
3
is consumed in the
amination reaction as rapidly as it is produced at all
concentrations of glutamine ( i.e. r eactions are fully coupled
as mentioned above). Unlike wild-type CTPS-catalysed
hydrolysis of Gln, Gln-OH hydrolysis is only fully coupled

to N
4
-hydroxy-CTP formation at low substrate concentra-
tions (Fig. 2) with uncoupling (coupling ratio ¼ 0.487)
being observed at saturating concentrations of Gln-OH
(Fig. 3 ). The k
cat
value for the wild-type CTPS-catalysed
formation of N
4
-hydroxy-CTP from nascent NH
2
OH (Gln-
OH as the su bstrate) is reduced 13-fold relative t o that for
nascent N H
3
(Gln as the substrate). The bulkier nascent
NH
2
OH must either encounter some unfavourable steric
interactions or a ÔbottleneckÕ as it traverses t he NH
3
tunnel,
or the kinetic expression for k
cat
for the hydrolysis of Gln-
OH contains terms that include rate constants for the
hydrolysis reaction, production of NH
2
OH and Glu, and

release o f G lu. (The e xact kinetic mechanism [i.e. order of
addition of substrates] is not known because the coopera-
tivity displayed by CTPS makes initial velocity studies
difficult to interpret [52] and hence the expression for k
cat
cannot presently be derived.) However, because the k
cat
value for wild-type hydrolysis of Gln-OH is reduced only
six-fold relative to the k
cat
value for Gln hydrolysis, it
appears t hat t he additional reduction in k
cat
(to 13-fold as
mentioned above) that is observed for Gln-OH-dependent
N
4
-hydroxy-CTP formation results from some other limit-
ing effect such as a Ôb ottleneckÕ.
Examination of the coupling ratios in Figs 2 and 3
reveals that at all substrate concentrations, L109A CTPS
exhibits u ncoupling (i.e. coupling ratios < 1). At saturating
substrate concentrations (Fig. 3 ), replacement of Leu109 by
alanine causes the coupling r atios to be reduced by factor s
of 2.95 and 2.40 for the Gln- and Gln-OH-dependent
reactions, respectively. Interestingly, the coupling ratios for
the G ln- a nd Gln-OH-dependent reactions are also both
reduced approximately two-fold for both the wild-type
(1.09 fi 0.487) and L109A (0.368 fi 0.203) enzymes.
Hence, L109A i s no more s ensitive to the increase d size of

NH
2
OH than wild-type CTPS when substrate concentra-
tions are saturating. Therefore, the rate of transfer of the
bulkier, nascent NH
2
OH under k
cat
conditions appears to
be limit ed by a ÔbottleneckÕ that is not affected by
replacement o f Leu109 by alanine. For this reason, only
the k
cat
/K
m
data (Fig. 2 ) are used to determine if the mutant
enzyme is sensitive to the larger siz e of the nasce nt NH
2
OH.
Previously, we reported that L109A exhibited uncoup-
ling of Gln hydrolysis from CTP formation [34]. We
wild-type
L109A
substrate
Gln
Glu-OH
0.905 ± 0.281
1.03 ± 0.13
0.206 ± 0.081
0.418 ± 0.093

2.03 ± 0.9
2
1.14 ± 0.38
2.46 ± 0.63
P = 0.0028
P = 0.0144
P = 0.5229 P = 0.0408
4.39 ± 2.20
Fig. 2. Coupling r atios for wild-type and L109A CTP synthases at
subsaturating substrate concentrations. Subsaturatin g coup ling ratios
(Eqn 2) are shown in boldface. The f actors by which the ratios change
upon altering either the substrate (vertical arrows) o r enzyme (hori-
zontal arrows) are shown i n i talics. The s tatistical signifi cance o f t he
changes in the coupling ratios is indicated by the corresponding
P value based on an unpaired, 2-tailed t-test (P < 0.05 is s tatistically
significant).
wild-type
L109A
substrate
Gln
Glu-OH
0.487 ± 0.021
1.09 ± 0.14
0.203 ± 0.050
0.368 ± 0.069
1.81 ± 0.5
6
2.23 ± 0.31
2.95 ± 0.68
2.40 ± 0.60

P = 0.0008
P = 0.0018 P = 0.0285
P = 0.0013
Fig. 3. Coupling r atios for wild-type and L109A CTP synthases at
saturating substrate concentrations. Saturating coupling ratios (Eqn 3)
are shown in b oldface. The factors by w hich the r atios change upon
altering either the s ubstrate ( vert ical arrows) o r en zyme ( horizo ntal
arrows) are shown in italics. The statistical significance of the changes
in the coupling rat ios is indic ated by th e corre spondin g P value based
on an unpaired, 2-tailed t-test (P < 0.05 i s statistically significant).
Ó FEBS 2004 Alternative substrates for E. coli CTP synthase (Eur. J. Biochem. 271) 4209
hypothesized t hat t his uncoupling could a rise from (a) a
leaky NH
3
tunnel, (b) a constricted NH
3
tunnel, or (c) the
failure of a transient tunnel to form. Our comprehensive
kinetic characterization of the ability of wild-type and
L109A CTP synthases to utilize bulkier analogues of both
NH
3
and Gln now permits us to r efine our hyp othesis. As
shown in Fig. 2 , L109A CTPS exhibits more pronounced
uncoupling with Gln-OH than with Gln. Hence, the
uncoupling obs erved w ith L 109A CTPS appears to depend
on the size of the nascent NH
3
analogue. This o bservation is
most consistent with the presence of a constricted NH

3
tunnel. If a leaky tunnel were present, we would e xpect the
bulkier nascent NH
2
OH to either leak out to bulk solvent,
like the nascent NH
3
(route 4 in Scheme 1), and therefore
exhibit the same degree of uncoupling, or be retained within
the tunnel for steric reasons and subsequently form N
4
-
hydroxy-CTP. In this latter case, less uncoupling would be
expected for the L109A enzyme, resulting in a higher
coupling ratio for nascent NH
2
OH relative to nascent N H
3
.
Structural aspects of uncoupling
In the crystal structure of apo-E. coli CTPS, L eu109 is
located o n a loop (residues 105–114) from an adjacent
subunit that extends over a deep cleft that separates the
GAT and synthase sites (Fig. 1) [21]. Interestingly, Leu109
is poised over t his cleft and above the opening that Endrizzi
et al . [21] identified as a putative entry point for e xogenous
NH
3
to access a solven t-filled vestibule that c onnects the
GAT active s ite and the GAT/synthase interface. Modelling

studies conducted by E ndrizzi et al. [ 21] s uggest tha t GTP
binds in the cleft that overlies the entry point for exogenous
NH
3
. T his finding is in accord with our recent report t hat
GTP binding inhibits CTP formation from exogenous NH
3
[17]. Studies also suggest that GTP binding induces a
conformational c hange in E. coli [3,16,17,52,53] and L. lac-
tis [54] CTP synthases. In the absence o f bound ligands, the
structure of apo-E. coli CTPS does not provide much
insight into what conformational changes might o ccur upon
GTP binding.
As replacement of Leu109 by Ala does not affect k
cat
values for t he reaction of bound exogenous substrates, the
size discrimination that is observed between nascent NH
3
and NH
2
OH must arise from differences between the
conformations that result when GTP i s bound to wild-type
CTPS relative to L109A CTPS. We propose that upon
binding GTP (perhaps concomitant with Gln binding) i n
the cleft between the GAT and synthase domains, the two
domains are drawn together. Consequently, the loop
comprised of residues 105–114 would move inward so that
Leu109 either packs against the bound GTP and/or helps to
occlude the entryway for exogenous NH
3

during catalysis of
Gln-dependent reactions; and the internal NH
3
tunnel/
vestibule may become ÔkinkedÕ.ThisÔkinkÕ could be
responsible for the ÔbottleneckÕ which leads to uncoupling
with wild-type CTPS when NH
2
OH is the s ubstrate at
saturating concentrations (Fig. 3 and see above). Such
significant conformational changes would be expected
because GTP binding causes conformational changes in
the GAT domain to promote stabilization of the tetrah edral
intermediates and transition states formed during Gln
hydrolysis [3].
This sc enario is consistent with the l ack of equilibration
of the nascent NH
3
derived from Gln hydrolysis with the
bulk solvent [4], the failure of L109F to catalyse glutamine
hydrolysis at wild-type rates [34], and the observation that
GTP binding inhibits NH
3
-dependent CTP formation [17].
It is probable that the phenyl group in L109F is too large
to pack properly against GTP thereby disrupting the
appropriate change in conformation required for full
coupling and glutaminase activity [34]. Although it is not
clear how the L 109A mutation leads to uncoupling, one
possibility is that a conformational ÔkinkÕ arises via the

mechanism mention ed above so that a functional tunnel
that efficiently couples the glutaminase and amination
reactions is not properly formed. Formation of a
competent NH
3
tunnel upon ligand binding has been
suggested by structural studies on GMP synthase [27,55]
and Gln phosphoribosylpyrophosphate amidotransferase
[25], and the same may be true for CTPS. While the
presence of a phenylalanine at position 109 may i mpede
the appropriate conformational changes required for
catalysis, substitution by alanine might permit Ôtoo muchÕ
of a conformational change b ecause of differences b etween
the packing of the leucine vs. alanine side chains with
GTP leading to a more significant ÔkinkÕ. Although the
kink/constriction could occur at any point along the r oute
traversed b y t he nascent NH
3
, one possible location is the
narrow ÔgateÕ between Pro54 and Val60, identified by
Endrizzi et al. [21], that resides at the base of the proposed
entryway fo r e xogenous NH
3
. F urther narrowing of this
Ôga teÕ upon GTP b inding could lead to a constriction that
discriminates between nascent NH
3
and the bulkier
NH
2

OH within L109A but does not affect the use of
exogenous NH
3
and its analogues (at least under k
cat
/K
m
conditions). Both explanations are f ully consistent with the
kinetic properties exhibited by L109A CTPS with alter-
native, bulkier substrates.
In conclusion, we have shown that L109A CTPS exhibit s
greater uncoupling with the bulkier, nascent NH
2
OH,
derived from Gln-OH hydrolysis, than with NH
3
derived
from Gln hydrolysis. This uncoupling is not caused by a
leaky NH
3
tunnel but arises because of a constriction within
the t unnel a s demonstrated by the ability of L 109A CTPS to
discriminate between nascent substrates based on size,
relative to the wild-type enzyme.
Acknowledgements
This work was supported, in part, by an operating grant from the
Canadian Institutes of He alth Research (S .L.B.), a Natural Sciences
and E ngineering Research Council (NSERC) of C anada Collaborative
Health Research Project grant (S.L.B.), and a graduate student
fellowship from the Nova Scotia Health Research Foundation

(F.A.L.). We express our thanks to Professor Enoch Baldwin
(University of California, Davis, CA, USA) for kindly providing us
with the PDB file f or apo-E. coli CTPS and the coordinates for GTP
modelled into t he GTP-binding site.
References
1. Koshland, D.E. Jr & Levitzki, A. (1974) CTP Synthetase and
Related Enzymes. In T he E nzym es (Boyer, P.D., ed.), pp. 539–559.
Academic Press, New York.
4210 F. A. Lunn and S. L. Bearne (Eur. J. Biochem. 271) Ó FEBS 2004
2. Long, C . & K oshland, D.E. Jr (1978) Cytidine triphosphate syn-
thetase. M ethods Enzymol. 51, 79–83.
3. Bearne, S.L., Hekmat, O. & MacDonnell, J.E. (2001) Inh ibition of
Escherichia c oli CTP synthase by glutamate gamma-semialdehyde
and the role of the allosteric effector GTP in glut amine hydrolysis.
Biochem. J . 356, 2 23–232.
4. Levitzki,A.&Koshland,D.E.Jr(1971)Cytidinetriphosphate
synthetase. Covalent intermediates and mechanisms of action.
Biochemistry 10 , 3365–3371.
5. Willemoe
¨
s, M. (2004) Competition b etween ammonia d erived
from internal glutamine hydrolysis and hydroxylamine present
in the solution for incorporationintoUTPascatalysedby
Lactococcus lactis CTP synthase. Arch. Biochem. Biophys. 424,
105–111.
6. Weng, M., Makaroff, C.A. & Z alkin, H. (1986) Nucleotide se-
quence of Escherichia coli pyrG encoding C TP synthetase. J. Biol.
Chem. 261, 5568–5574.
7. Weng, M .L. & Zalkin, H. (1987) Structural role for a conserved
region in the CTP syn the tase glutamine amide transfer domain.

J. Bacteriol. 169, 3 023–3028.
8. Hatse, S., De C lercq, E. & Balzarini, J . (1999) Role of anti-
metabolites of purine and pyrimidine nucleotide metabolism i n
tumor cell differentiation. Biochem. Pharmacol. 58, 539–555.
9. Kennedy, E.P. (1986) The biosynthesis of p hosp holipids. In Lipids
and Membranes: Past, Present and Future (Op den Kamp, J.A.F.,
Roelofsen, B. & Wirtz, K.W.A., eds), pp. 171–206. Elsevier
Scientific Publishers, Amsterdam.
10. Kensler, T.W. & Cooney, D.A. (1989) Inhibitors of the de novo
pyrimidine pathway. In Design of Enzy me Inhibitors as Drugs
(Sandler, M. & Smith, H.J., eds), pp. 379–401. Oxford University
Press, New Y ork.
11. Gao, W.Y., Johns, D.G. & Mitsuya, H. (2000) Poten tiation of the
anti-HIV activity of zalcitab ine and lami vudine by a CTP sy nthase
inhibitor, 3-deazauridine. Nucleosides Nucleotides Nucleic Acids
19, 371–377.
12. De Clercq, E. (1993) Antiviral agents: characteristic activity
spectrum depending on the molecular target with which they
interact. Adv. Virus. Res. 42 , 1–55.
13. Hendriks, E.F., O’Sullivan, W .J. & Stewart, T.S. (1998) Molecular
cloning and characterization of the Plasmodium falciparum cy ti-
dine triphosphate synthetase gene. Biochim. Biophys. A cta 1399,
213–218.
14. Hofer,A.,Steverding,D.,Chabes,A.,Brun,R.&Thelander,L.
(2001) Trypanosoma brucei CTP synthetase: a target for the
treatment of A frican sleeping sickness. Proc. Natl Acad. Sci. USA
98, 6412–6416.
15. Lim, R.L., O’Sullivan, W.J. & Stewart, T.S. (1 996) Isolation,
characterization and expression of the gene en cod ing cytidine
triphosphate synthetase from Giard ia intestinalis. Mol. Biochem.

Parasitol. 78 , 249–257.
16. Levitzki, A. & Koshland, D.E. Jr (1972) Role of an allosteric
effector. Guanosine triphosphate activation in cytosine triphos-
phate synthetase. Biochem istry 11, 241–246.
17. MacDonnell, J.E., Lunn, F.A. & Bearne, S.L. (2004) Inhibition of
E. coli CTP synthase by the ÔpositiveÕ allosteric effec tor GTP.
Biochim. Bioph ys. Acta 1699, 213–220.
18. Long, C.W. & Pardee, A.B. (1967) Cytidine triphosphate syn-
thetase of Escherichia coli B. I. Purification and ki netics. J. Biol.
Chem. 242, 4715–4721.
19. Levitzki, A. & Koshland, D .E. Jr (1969) Negative c ooperativity in
regulatory enzymes. Proc. Natl Acad. Sci. USA 62, 1 121–1128.
20. Levitzki, A. & Koshland, D.E. Jr (1972) Ligand-induced dimer-
to-tetramer transformation in cytosine triphosphate synthetase.
Biochemistry 11 , 247–253.
21. Endrizzi, J.A., Kim, H., Anderson, P.M. & B aldwin, E.P. (2004)
Crystal structure of Esche richia coli cytidine triphosphate synthe-
tase, a nucleotide-regulated glutamine amidotransferase/ATP-
dependent amidoligase fusion protein and homologue of anti-
cancer and antiparasitic drug targets. Bioch em ist ry 43, 6447–6463.
22. Holden, H .M., Thoden, J.B. & Ra ushel, F.M. (1998) Carbamoyl
phosphate synthetase: a tunnel runs th rough it. Curr. Opin. Struct.
Biol. 8, 679–685.
23. Mullins, L.S. & R aushel, F.M. (1999) Channeling of ammonia
through the intermolecular tunnel contained within carbamoyl
phosphate synthetase. J. Am. Chem. Soc. 121, 3803–3804.
24. Thoden, J.B ., H olde n, H.M., Wesenberg, G., Raushel, F .M. &
Rayment, I. (1997) Structure of carbamoyl phosphate synthetase:
a journey of 96 A
˚

from substrate to product. Biochemistry 36,
6305–6316.
25. Krahn, J.M., K im, J .H., Burns, M.R.,Parry,R.J.,Zalkin,H.&
Smith, J.L. (1997) Coupled formation of an amidotransferase
interdomain ammonia c hannel and a phosphoribosyltransferase
active site. Biochemistry 36, 11061–11068.
26. Muchmore, C.R., Krahn, J.M., Kim, J.H., Zalkin, H. & Smith,
J.L. (1998) Crystal s tructure of glutamine phosphoribosylpyr-
ophosphate amidotransferase from Escherichia coli. Protein Sci. 7,
39–51.
27. Tesmer, J.J., Klem, T .J., Deras, M.L., Davisson, V.J. & Smith,
J.L. (1996) The crystal structure of GMP synthetase reveals a
novel catalytic triad and is a s tructural paradigm for two enzyme
families. Nat. Struct. Biol. 3, 74–86.
28. Teplyakov, A., Obmolova, G., Badet,B.&Badet-Denisot,M.A.
(2001) Ch annelin g of ammonia i n glucosamine-6-phosphate syn -
thase. J. Mol. Biol. 313, 1 093–1102.
29. Teplyakov, A., Obmolova, G., B adet-Denisot, M.A. & Badet, B.
(1999) The mechanism of sugar p ho sphate isomerization by glu-
cosamine 6-phosphate synthase. Protein S ci. 8, 596–602.
30. Teplyakov, A., Obmolova, G., Badet-Denisot, M.A., Badet, B. &
Polikarpov, I. (1998) Involvement of the C terminus in
intramolecular nitrogen channeling in glucosamine 6-phosphate
synthase: evidence from a 1.6 A
˚
crystal structure of the i somerase
domain. Structure 6, 1047–1055.
31. Larsen, T.M., Boehlein, S.K., Schuster, S.M., Richards, N.G.,
Thoden, J.B., Holden, H.M. & Rayment, I. (1999) Three-
dimensional s tructure of Escherichia coli asparagine synthetase B:

a short journey from substrate to product. Biochemistry 38,
16146–16157.
32. Morollo, A.A. & Eck, M.J. (2001) Structure of the cooperative
allosteric anthranilate synthase from Salmonella typhimurium.
Nat. Struct. B iol. 8, 243–247.
33. Spraggon, G., Kim, C., Nguyen-Huu, X., Yee, M.C., Yanofsky,
C. & Mills, S.E. (2001) The s tructures of anthranilate synthase of
Serratia marcescens crystallized in the presence of (i) its substrates,
chorismate and g lutamine, and a product, glutamate, and (ii) its
end-product inh ibitor, 1-trypto phan. Proc.NatlAcad.Sci.USA
98, 6021–6026.
34. Iyengar, A. & Bearne, S.L. (2003) Aspartate 107 and leucine 10 9
facilitate efficient coupling of glutamine hydrolysis to CTP
synthesis b y E. coli CTP synthase. Biochem. J. 369, 497–507.
35. Novagen. (1997) pET System Manual, 7th edn. TB055, 18–64.
Novagen, Madison, WI, U SA.
36. Levintow, L., Me ister, A., Hogeboom, G.H. & Kuff, E.L. (1955)
Studies on the relationship between the enzymatic synthesis of
glutamine and the glutamyl t ransfer r eaction. J. Am. C hem. Soc.
77, 5304–5308.
37. Willemoe
¨
s, M. & Larsen, S. (2003) Su bstrat e inhibition of Lac-
tococcus lactis cytidine 5¢-triphosphate synthase by ammonium
chloride is e nhanced by salt-dependent tetramer asso ciation. Arch.
Biochem. Biop hys. 413, 17– 22.
38. Jencks, W.P. & Regenstein , J. (1968) Ionization constants of acids
and bases. In Handbook of Biochemistry (Sober, H.A., ed.), p p.
J150–J189, The Chemical Rubber C o, Cleveland, Oh io.
Ó FEBS 2004 Alternative substrates for E. coli CTP synthase (Eur. J. Biochem. 271) 4211

39. Iyengar, A. & Bearne, S.L. (2002) An assay for CTP synthetase
glutaminase activity using high performance liquid chromato-
graphy. Anal. Biochem. 308, 396–400.
40. Huang, X. & Raushel, F .M. (2000) Restricted passage of reaction
intermediates t hrough th e ammonia tunnel o f carb am oyl p hos-
phate synthetase. J. Biol. C hem. 275, 26233–26240.
41. Bera, A.K., Smith, J.L. & Zalkin, H. (2000) Dual role for
the glutamine phosphoribosylpyrophosphate amidotransferase
ammonia channel. I nterdo main signaling a nd intermediate chan-
neling. J. Biol. Chem. 27 5, 7975–7979.
42. Buchanan, J.M. ( 1973) The amidotransferases. Adv. Enzymol.
Relat. Areas M ol. Biol. 39 , 91–183.
43. Massiere,F.&Badet-Denisot,M.A.(1998)Themechanismof
glutamine-depend ent amidotransferases. Cell. Mol. Life Sci. 54,
205–222.
44. Zalkin, H. (1993) The amidotransferases. Adv. Enzymol. Relat.
Areas Mol. B iol. 66, 203 –309.
45. Kizaki, H., Ohsaka, F. & Sakurada, T. (1987) Synthesis of
N
4
-substituted CTP by mammalian CTP synthetase. Biochem.
Biophys. R es. Commun. 14 5 , 569–574.
46. Lieberman, I. (1956) En zymatic amination of uridine triphosphate
to cytidine triph osph ate. J. Biol. Chem. 222, 765–775.
47. Willemoe
¨
s, M. (2003) Thr-431 and Arg-433 are part of a conserved
sequence mo tif of the glutamine amidotransferase domain of CTP
synthases and are involved in GTP activation of the La c toco ccus
lactis enzyme. J. Biol. Chem. 27 8, 9407–9411.

48. Boehlein, S.K., Schuster, S.M. & Richards, N .G. ( 1996) Glutamic
acid gamma-monohydroxamate a nd hydroxylamine are alternate
substrates for Escherichia coli asparagine synthetase B. Bioche-
mistry 35, 3031–3037.
49. Pearson, R.G., Sobel, H. & Songstad, J. (1968) N ucleophilic
reactivity constants toward methyl iodide a nd trans-[Pt(py)
2
Cl
2
].
J. Am. Chem. Soc. 90, 319–326.
50. Bunting, J.W., Mason, J.M. & H eo, C.K.M. (1994) Nucleophili-
city towards a satu rated carbon atom: rate constan ts for the
aminolysis of methyl 4-nitrobenzenesulfonate in aqu eous s olu-
tion. A comparison of the n and N
+
parameters for amine
nucleophilicity. J. Chem. Soc. P erkin Trans. 2, 2291–2300.
51. Huang, X., Holden, H.M. & Raushel, F.M. (2001) Channeling of
substrates and intermediates in enzyme-catalyzed reactions. Annu.
Rev. Biochem. 70 , 149–180.
52. Lewis, D.A. & Villafranca, J.J. (1989) Investigation of the
mechanism of CTP synthetase using rapid quench and isotope
partitioning methods. Biochemistry 28, 8454–8459.
53. Levitzki, A., Stallcup, W.B . & Koshland, D .E. Jr (1971) Half-
of-the-sites reactivity and the conformational states of cytidine
triphosphate synthetase. Biochemistry 10, 3371–3378.
54. Willemoe
¨
s, M. & Sigurskjold, B.W. (2002) Steady-state kinetics o f

the glutaminase reaction of CTP synthase from Lactococcus lactis.
Eur. J. Biochem. 269, 4 772–4779.
55. Zalkin, H. & Smith, J.L. (1998) Enzymes utilizing glutamine as an
amide donor. Adv. Enzymol. Relat. Areas Mol. Biol. 72 , 87–144.
4212 F. A. Lunn and S. L. Bearne (Eur. J. Biochem. 271) Ó FEBS 2004